*** UNIX MANUAL PAGE BROWSER ***

PF.CONF(5) FreeBSD File Formats Manual PF.CONF(5) NAME pf.conf - packet filter configuration file DESCRIPTION The pf(4) packet filter modifies, drops, or passes packets according to rules or definitions specified in pf.conf. This is an overview of the sections in this manual page: PACKET FILTERING including network address translation (NAT). OPTIONS globally tune the behaviour of the packet filtering engine. QUEUEING provides rule-based bandwidth and traffic control. TABLES provide a method for dealing with large numbers of addresses. ANCHORS are containers for rules and tables. STATEFUL FILTERING tracks packets by state. TRAFFIC NORMALISATION includes scrub, fragment handling, and blocking spoofed traffic. OPERATING SYSTEM FINGERPRINTING is a method for detecting a host's operating system. EXAMPLES provides some example rulesets. GRAMMAR provides a complete BNF grammar reference. The current line can be extended over multiple lines using a backslash (`\'). Comments can be put anywhere in the file using a hash mark (`#'), and extend to the end of the current line. Care should be taken when commenting out multi-line text: the comment is effective until the end of the entire block. Argument names not beginning with a letter, digit, or underscore must be quoted. Additional configuration files can be included with the include keyword, for example: include "/etc/pf/sub.filter.conf" Macros can be defined that will later be expanded in context. Macro names must start with a letter, digit, or underscore, and may contain any of those characters. Macro names may not be reserved words (for example pass, in, out). Macros are not expanded inside quotes. For example: ext_if = "kue0" all_ifs = "{" $ext_if lo0 "}" pass out on $ext_if from any to any pass in on $ext_if proto tcp from any to any port 25 PACKET FILTERING pf(4) has the ability to block, pass, and match packets based on attributes of their layer 3 and layer 4 headers. Filter rules determine which of these actions are taken; filter parameters specify the packets to which a rule applies. Each time a packet processed by the packet filter comes in on or goes out through an interface, the filter rules are evaluated in sequential order, from first to last. For block and pass, the last matching rule decides what action is taken; if no rule matches the packet, the default action is to pass the packet without creating a state. For match, rules are evaluated every time they match; the pass/block state of a packet remains unchanged. Most parameters are optional. If a parameter is specified, the rule only applies to packets with matching attributes. The matching for some parameters can be inverted with the ! operator. Certain parameters can be expressed as lists, in which case pfctl(8) generates all needed rule combinations. By default pf(4) filters packets statefully: the first time a packet matches a pass rule, a state entry is created. The packet filter examines each packet to see if it matches an existing state. If it does, the packet is passed without evaluation of any rules. After the connection is closed or times out, the state entry is automatically removed. The following actions can be used in the filter: block The packet is blocked. There are a number of ways in which a block rule can behave when blocking a packet. The default behaviour is to drop packets silently, however this can be overridden or made explicit either globally, by setting the block-policy option, or on a per-rule basis with one of the following options: drop The packet is silently dropped. return This causes a TCP RST to be returned for TCP packets and an ICMP UNREACHABLE for other types of packets. return-icmp return-icmp6 This causes ICMP messages to be returned for packets which match the rule. By default this is an ICMP UNREACHABLE message, however this can be overridden by specifying a message as a code or number. return-rst This applies only to TCP packets, and issues a TCP RST which closes the connection. An optional parameter, ttl, may be given with a TTL value. Options returning ICMP packets currently have no effect if pf(4) operates on a bridge(4), as the code to support this feature has not yet been implemented. The simplest mechanism to block everything by default and only pass packets that match explicit rules is specify a first filter rule of: block all match The packet is matched. This mechanism is used to provide fine grained filtering without altering the block/pass state of a packet. match rules differ from block and pass rules in that parameters are set every time a packet matches the rule, not only on the last matching rule. For the following parameters, this means that the parameter effectively becomes "sticky" until explicitly overridden: nat-to, binat-to, rdr-to, queue, rtable, and scrub. log is different still, in that the action happens every time a rule matches i.e. a single packet can get logged more than once. pass The packet is passed; state is created unless the no state option is specified. The following parameters can be used in the filter: in or out A packet always comes in on, or goes out through, one interface. in and out apply to incoming and outgoing packets; if neither are specified, the rule will match packets in both directions. log (all | matches | to interface | user) In addition to any action specified, log the packet. Only the packet that establishes the state is logged, unless the no state option is specified. The logged packets are sent to a pflog(4) interface, by default pflog0; pflog0 is monitored by the pflogd(8) logging daemon which logs to the file /var/log/pflog in pcap binary format. The keywords all, matches, to, and user are optional and can be combined using commas, but must be enclosed in parentheses if given. Use all to force logging of all packets for a connection. This is not necessary when no state is explicitly specified. If matches is specified, it logs the packet on all subsequent matching rules. It is often combined with to interface to avoid adding noise to the default log file. The keyword user logs the UID and PID of the socket on the local host used to send or receive a packet, in addition to the normal information. To specify a logging interface other than pflog0, use the syntax to interface. quick If a packet matches a rule which has the quick option set, this rule is considered the last matching rule, and evaluation of subsequent rules is skipped. on interface | any This rule applies only to packets coming in on, or going out through, this particular interface or interface group. For more information on interface groups, see the group keyword in ifconfig(8). any will match any existing interface except loopback ones. on rdomain number This rule applies only to packets coming in on, or going out through, this particular routing domain. inet | inet6 This rule applies only to packets of this address family. proto protocol This rule applies only to packets of this protocol. Common protocols are ICMP, ICMP6, TCP, and UDP. For a list of all the protocol name to number mappings used by pfctl(8), see the file /etc/protocols. from source port source os source to dest port dest This rule applies only to packets with the specified source and destination addresses and ports. Addresses can be specified in CIDR notation (matching netblocks), as symbolic host names, interface names or interface group names, or as any of the following keywords: any Any address. no-route Any address which is not currently routable. route label Any address matching the given route(8) label. self Expands to all addresses assigned to all interfaces. <table> Any address matching the given table. urpf-failed Any source address that fails a unicast reverse path forwarding (URPF) check, i.e. packets coming in on an interface other than that which holds the route back to the packet's source address. Ranges of addresses are specified using the `-' operator. For instance: "10.1.1.10 - 10.1.1.12" means all addresses from 10.1.1.10 to 10.1.1.12, hence addresses 10.1.1.10, 10.1.1.11, and 10.1.1.12. Interface names, interface group names, and self can have modifiers appended: :0 Do not include interface aliases. :broadcast Translates to the interface's broadcast address(es). :network Translates to the network(s) attached to the interface. :peer Translates to the point-to-point interface's peer address(es). Host names may also have the :0 modifier appended to restrict the name resolution to the first of each v4 and v6 address found. Host name resolution and interface to address translation are done at ruleset load-time. When the address of an interface (or host name) changes (under DHCP or PPP, for instance), the ruleset must be reloaded for the change to be reflected in the kernel. Surrounding the interface name (and optional modifiers) in parentheses changes this behaviour. When the interface name is surrounded by parentheses, the rule is automatically updated whenever the interface changes its address. The ruleset does not need to be reloaded. This is especially useful with NAT. Ports can be specified either by number or by name. For example, port 80 can be specified as www. For a list of all port name to number mappings used by pfctl(8), see the file /etc/services. Ports and ranges of ports are specified using these operators: = (equal) != (unequal) < (less than) <= (less than or equal) > (greater than) >= (greater than or equal) : (range including boundaries) >< (range excluding boundaries) <> (except range) `><', `<>' and `:' are binary operators (they take two arguments). For instance: port 2000:2004 means `all ports >= 2000 and <= 2004', hence ports 2000, 2001, 2002, 2003, and 2004. port 2000 >< 2004 means `all ports > 2000 and < 2004', hence ports 2001, 2002, and 2003. port 2000 <> 2004 means `all ports < 2000 or > 2004', hence ports 1-1999 and 2005-65535. The operating system of the source host can be specified in the case of TCP rules with the os modifier. See the OPERATING SYSTEM FINGERPRINTING section for more information. The host, port, and os specifications are optional, as in the following examples: pass in all pass in from any to any pass in proto tcp from any port < 1024 to any pass in proto tcp from any to any port 25 pass in proto tcp from 10.0.0.0/8 port >= 1024 \ to ! 10.1.2.3 port != ssh pass in proto tcp from any os "OpenBSD" pass in proto tcp from route "DTAG" The following additional parameters can be used in the filter: all This is equivalent to `from any to any'. allow-opts By default, packets with IPv4 options or IPv6 hop-by-hop or destination options header are blocked. When allow-opts is specified for a pass rule, packets that pass the filter based on that rule (last matching) do so even if they contain options. For packets that match state, the rule that initially created the state is used. The implicit pass rule, that is used when a packet does not match any rules, does not allow IP options or option headers. Note that IPv6 packets with type 0 routing headers are always dropped. divert-packet port port Used to send matching packets to divert(4) sockets bound to port port. If the default option of fragment reassembly is enabled, scrubbing with reassemble tcp is also enabled for divert-packet rules. divert-reply Used to receive replies for sockets that are bound to addresses which are not local to the machine. See setsockopt(2) for information on how to bind these sockets. divert-to host port port Used to redirect packets to a local socket bound to host and port. The packets will not be modified, preserving the original destination address for the application to access. SOCK_STREAM connections can access the original destination address using getsockname(2). SOCK_DGRAM sockets can be configured with the ip(4) IP_RECVDSTADDR and IP_RECVDSTPORT socket options when receiving IPv4 packets, or the ip6(4) IPV6_RECVPKTINFO and IPV6_RECVDSTPORT socket options when receiving IPv6 packets. flags a/b | any This rule only applies to TCP packets that have the flags a set out of set b. Flags not specified in b are ignored. For stateful connections, the default is flags S/SA. To indicate that flags should not be checked at all, specify flags any. The flags are: (F)IN, (S)YN, (R)ST, (P)USH, (A)CK, (U)RG, (E)CE, and C(W)R. flags S/S Flag SYN is set. The other flags are ignored. flags S/SA This is the default setting for stateful connections. Out of SYN and ACK, exactly SYN may be set. SYN, SYN+PSH, and SYN+RST match, but SYN+ACK, ACK, and ACK+RST do not. This is more restrictive than the previous example. flags /SFRA If the first set is not specified, it defaults to none. All of SYN, FIN, RST, and ACK must be unset. Because flags S/SA is applied by default (unless no state is specified), only the initial SYN packet of a TCP handshake will create a state for a TCP connection. It is possible to be less restrictive, and allow state creation from intermediate (non-SYN) packets, by specifying flags any. This will cause pf(4) to synchronize to existing connections, for instance if one flushes the state table. However, states created from such intermediate packets may be missing connection details such as the TCP window scaling factor. States which modify the packet flow, such as those affected by af-to, modulate state, nat-to, rdr-to, or synproxy state options, or scrubbed with reassemble tcp, will also not be recoverable from intermediate packets. Such connections will stall and time out. group group Similar to user, this rule only applies to packets of sockets owned by the specified group. icmp-type type [code code] icmp6-type type [code code] This rule only applies to ICMP or ICMP6 packets with the specified type and code. Text names for ICMP types and codes are listed in icmp(4) and icmp6(4). The protocol and the ICMP type indicator (icmp-type or icmp6-type) must match. ICMP responses are not permitted unless they either match an existing request, or unless no state or keep state (sloppy) is specified. label string Adds a label to the rule, which can be used to identify the rule. For instance, `pfctl -s labels' shows per-rule statistics for rules that have labels. The following macros can be used in labels: $dstaddr The destination IP address. $dstport The destination port specification. $if The interface. $nr The rule number. $proto The protocol name. $srcaddr The source IP address. $srcport The source port specification. For example: ips = "{ 1.2.3.4, 1.2.3.5 }" pass in proto tcp from any to $ips \ port > 1023 label "$dstaddr:$dstport" Expands to: pass in inet proto tcp from any to 1.2.3.4 \ port > 1023 label "1.2.3.4:>1023" pass in inet proto tcp from any to 1.2.3.5 \ port > 1023 label "1.2.3.5:>1023" The macro expansion for the label directive occurs only at configuration file parse time, not during runtime. max-pkt-rate number/seconds Measure the rate of packets matching the rule and states created by it. When the specified rate is exceeded, the rule stops matching. Only packets in the direction in which the state was created are considered, so that typically requests are counted and replies are not. For example, to pass up to 100 ICMP packets per 10 seconds: block in proto icmp pass in proto icmp max-pkt-rate 100/10 When the rate is exceeded, all ICMP is blocked until the rate falls below 100 per 10 seconds again. once Create a one shot rule. The first matching packet marks the rule as expired. Expired rules are skipped and hidden, unless pfctl(8) is used in debug or verbose mode. probability number% A probability attribute can be attached to a rule, with a value set between 0 and 100%, in which case the rule is honoured using the given probability value. For example, the following rule will drop 20% of incoming ICMP packets: block in proto icmp probability 20% prio number Only match packets which have the given queueing priority assigned. [!]received-on interface Only match packets which were received on the specified interface (or interface group). any will match any existing interface except loopback ones. rtable number Used to select an alternate routing table for the routing lookup. Only effective before the route lookup happened, i.e. when filtering inbound. set delay milliseconds Packets matching this rule will be delayed at the outbound interface by the given number of milliseconds. set prio priority | (priority, priority) Packets matching this rule will be assigned a specific queueing priority. Priorities are assigned as integers 0 through 7, with a default priority of 3. If the packet is transmitted on a vlan(4) interface, the queueing priority will also be written as the priority code point in the 802.1Q VLAN header. If two priorities are given, TCP ACKs with no data payload and packets which have a TOS of lowdelay will be assigned to the second one. Packets with a higher priority number are processed first, and packets with the same priority are processed in the order in which they are received. For example: pass in proto tcp to port 25 set prio 2 pass in proto tcp to port 22 set prio (2, 5) The interface priority queues accessed by the set prio keyword are always enabled and do not require any additional configuration, unlike the queues described below and in the QUEUEING section. set queue queue | (queue, queue) Packets matching this rule will be assigned to the specified queue. If two queues are given, packets which have a TOS of lowdelay and TCP ACKs with no data payload will be assigned to the second one. See QUEUEING for setup details. For example: pass in proto tcp to port 25 set queue mail pass in proto tcp to port 22 set queue(ssh_bulk, ssh_prio) set tos string | number Enforces a TOS for matching packets. string may be one of critical, inetcontrol, lowdelay, netcontrol, throughput, reliability, or one of the DiffServ Code Points: ef, af11 ... af43, cs0 ... cs7; number may be either a hex or decimal number. tag string Packets matching this rule will be tagged with the specified string. The tag acts as an internal marker that can be used to identify these packets later on. This can be used, for example, to provide trust between interfaces and to determine if packets have been processed by translation rules. Tags are "sticky", meaning that the packet will be tagged even if the rule is not the last matching rule. Further matching rules can replace the tag with a new one but will not remove a previously applied tag. A packet is only ever assigned one tag at a time. Tags take the same macros as labels (see above). [!]tagged string Used with filter or translation rules to specify that packets must already be tagged with the given string in order to match the rule. tos string | number This rule applies to packets with the specified TOS bits set. string may be one of critical, inetcontrol, lowdelay, netcontrol, throughput, reliability, or one of the DiffServ Code Points: ef, af11 ... af43, cs0 ... cs7; number may be either a hex or decimal number. For example, the following rules are identical: pass all tos lowdelay pass all tos 0x10 pass all tos 16 user user This rule only applies to packets of sockets owned by the specified user. For outgoing connections initiated from the firewall, this is the user that opened the connection. For incoming connections to the firewall itself, this is the user that listens on the destination port. When listening sockets are bound to the wildcard address, pf(4) cannot determine if a connection is destined for the firewall itself. To avoid false matches on just the destination port, combine a user rule with source or destination address self. All packets, both outgoing and incoming, of one connection are associated with the same user and group. Only TCP and UDP packets can be associated with users. The user and group arguments refer to the effective (as opposed to the real) IDs, in case the socket is created by a setuid/setgid process. User and group IDs are stored when a socket is created; when a process creates a listening socket as root (for instance, by binding to a privileged port) and subsequently changes to another user ID (to drop privileges), the credentials will remain root. User and group IDs can be specified as either numbers or names. The syntax is similar to the one for ports. The following example allows only selected users to open outgoing connections: block out proto tcp all pass out proto tcp from self user { < 1000, dhartmei } The example below permits users with uid between 1000 and 1500 to open connections: block out proto tcp all pass out proto tcp from self user { 999 >< 1501 } The `:' operator, which works for port number matching, does not work for user and group match. Translation Translation options modify either the source or destination address and port of the packets associated with a stateful connection. pf(4) modifies the specified address and/or port in the packet and recalculates IP, TCP, and UDP checksums as necessary. If specified on a match rule, subsequent rules will see packets as they look after any addresses and ports have been translated. These rules will therefore have to filter based on the translated address and port number. The state entry created permits pf(4) to keep track of the original address for traffic associated with that state and correctly direct return traffic for that connection. Different types of translation are possible with pf: af-to Translation between different address families (NAT64) is handled using af-to rules. Because address family translation overrides the routing table, it's only possible to use af-to on inbound rules, and a source address for the resulting translation must always be specified. The optional second argument is the host or subnet the original addresses are translated into for the destination. The lowest bits of the original destination address form the host part of the new destination address according to the specified subnet. It is possible to embed a complete IPv4 address into an IPv6 address using a network prefix of /96 or smaller. When a destination address is not specified, it is assumed that the host part is 32-bit long. For IPv6 to IPv4 translation this would mean using only the lower 32 bits of the original IPv6 destination address. For IPv4 to IPv6 translation the destination subnet defaults to the subnet of the new IPv6 source address with a prefix length of /96. See RFC 6052 Section 2.2 for details on how the prefix determines the destination address encoding. For example, the following rules are identical: pass in inet af-to inet6 from 2001:db8::1 to 2001:db8::/96 pass in inet af-to inet6 from 2001:db8::1 In the above example the matching IPv4 packets will be modified to have a source address of 2001:db8::1 and a destination address will get prefixed with 2001:db8::/96, e.g. 198.51.100.100 will be translated to 2001:db8::c633:6464. In the reverse case the following rules are identical: pass in inet6 from any to 64:ff9b::/96 af-to inet \ from 198.51.100.1 to 0.0.0.0/0 pass in inet6 from any to 64:ff9b::/96 af-to inet \ from 198.51.100.1 The destination IPv4 address is assumed to be embedded inside the original IPv6 destination address, e.g. 64:ff9b::c633:6464 will be translated to 198.51.100.100. The current implementation will only extract IPv4 addresses from the IPv6 addresses with a prefix length of /96 and greater. binat-to A binat-to rule specifies a bidirectional mapping between an external IP netblock and an internal IP netblock. It expands to an outbound nat-to rule and an inbound rdr-to rule. nat-to A nat-to option specifies that IP addresses are to be changed as the packet traverses the given interface. This technique allows one or more IP addresses on the translating host to support network traffic for a larger range of machines on an "inside" network. Although in theory any IP address can be used on the inside, it is strongly recommended that one of the address ranges defined by RFC 1918 be used. Those netblocks are: 10.0.0.0 - 10.255.255.255 (all of net 10, i.e. 10/8) 172.16.0.0 - 172.31.255.255 (i.e. 172.16/12) 192.168.0.0 - 192.168.255.255 (i.e. 192.168/16) nat-to is usually applied outbound. If applied inbound, nat-to to a local IP address is not supported. rdr-to The packet is redirected to another destination and possibly a different port. rdr-to can optionally specify port ranges instead of single ports. For instance: match in ... port 2000:2999 rdr-to ... port 4000 redirects ports 2000 to 2999 (inclusive) to port 4000. match in ... port 2000:2999 rdr-to ... port 4000:* redirects port 2000 to 4000, port 2001 to 4001, ..., port 2999 to 4999. rdr-to is usually applied inbound. If applied outbound, rdr-to to a local IP address is not supported. In addition to modifying the address, some translation rules may modify source or destination ports for TCP or UDP connections; implicitly in the case of nat-to options and explicitly in the case of rdr-to ones. Port numbers are never translated with a binat-to rule. Translation options apply only to packets that pass through the specified interface, and if no interface is specified, translation is applied to packets on all interfaces. For instance, redirecting port 80 on an external interface to an internal web server will only work for connections originating from the outside. Connections to the address of the external interface from local hosts will not be redirected, since such packets do not actually pass through the external interface. Redirections cannot reflect packets back through the interface they arrive on, they can only be redirected to hosts connected to different interfaces or to the firewall itself. However packets may be redirected to hosts connected to the interface the packet arrived on by using redirection with NAT. For example: pass in on $int_if proto tcp from $int_net to $ext_if port 80 \ rdr-to $server pass out on $int_if proto tcp to $server port 80 \ received-on $int_if nat-to $int_if Note that redirecting external incoming connections to the loopback address will effectively allow an external host to connect to daemons bound solely to the loopback address, circumventing the traditional blocking of such connections on a real interface. For example: pass in on egress proto tcp from any to any port smtp \ rdr-to 127.0.0.1 port spamd Unless this effect is desired, any of the local non-loopback addresses should be used instead as the redirection target, which allows external connections only to daemons bound to this address or not bound to any address. For af-to, nat-to and rdr-to options for which there is a single redirection address which has a subnet mask smaller than 32 for IPv4 or 128 for IPv6 (more than one IP address), a variety of different methods for assigning this address can be used: bitmask The bitmask option applies the network portion of the redirection address to the address to be modified (source with nat-to, destination with rdr-to). least-states [sticky-address] The least-states option selects the address with the least active states from a given address pool and considers given weights associated with address(es). Weights can be specified between 1 and 65535. Addresses with higher weights are selected more often. sticky-address can be specified to ensure that multiple connections from the same source are mapped to the same redirection address. Associations are destroyed as soon as there are no longer states which refer to them; in order to make the mappings last beyond the lifetime of the states, increase the global options with set timeout src.track. random [sticky-address] The random option selects an address at random within the defined block of addresses. sticky-address is as described above. round-robin [sticky-address] The round-robin option loops through the redirection address(es) and considers given weights associated with address(es). Weights can be specified between 1 and 65535. Addresses with higher weights are selected more often. sticky-address is as described above. source-hash [key] [sticky-address] The source-hash option uses a hash of the source address to determine the redirection address, ensuring that the redirection address is always the same for a given source. An optional key can be specified after this keyword either in hex or as a string; by default pfctl(8) randomly generates a key for source-hash every time the ruleset is reloaded. sticky-address is as described above. static-port With nat-to rules, the static-port option prevents pf(4) from modifying the source port on TCP and UDP packets. When more than one redirection address or a table is specified, bitmask is not permitted as a pool type. Routing If a packet matches a rule with one of the following route options set, the packet filter will route the packet according to the type of route option. When such a rule creates state, the route option is also applied to all packets matching the same connection. dup-to The dup-to option creates a duplicate of the packet and routes it like route-to. The original packet gets routed as it normally would. reply-to The reply-to option is similar to route-to, but routes packets that pass in the opposite direction (replies) to the specified address. Opposite direction is only defined in the context of a state entry, and reply-to is useful only in rules that create state. It can be used on systems with multiple paths to the internet to ensure that replies to an incoming network connection to a particular address are sent using the path associated with that address (symmetric routing enforcement). route-to The route-to option routes the packet to the specified destination address instead of the destination address in the packet header. When a route-to rule creates state, only packets that pass in the same direction as the filter rule specifies will be routed in this way. Packets passing in the opposite direction (replies) are not affected and are routed normally. For the dup-to, reply-to, and route-to route options for which there is a single redirection address which has a subnet mask smaller than 32 for IPv4 or 128 for IPv6 (more than one IP address), the methods least-states, random, round-robin, and source-hash, as described above, can be used. OPTIONS pf(4) may be tuned for various situations using the set command. set block-policy drop | return The block-policy option sets the default behaviour for the packet block action: drop Packet is silently dropped. return A TCP RST is returned for blocked TCP packets, an ICMP UNREACHABLE is returned for blocked UDP packets, and all other packets are silently dropped. The default value is drop. set debug level Set the debug level, which limits the severity of log messages printed by pf(4). This should be a keyword from the following ordered list (highest to lowest): emerg, alert, crit, err, warning, notice, info, and debug. These keywords correspond to the similar (LOG_) values specified to the syslog(3) library routine. The default value is err. set fingerprints filename Load fingerprints of known operating systems from the given filename. By default fingerprints of known operating systems are automatically loaded from pf.os(5), but can be overridden via this option. Setting this option may leave a small period of time where the fingerprints referenced by the currently active ruleset are inconsistent until the new ruleset finishes loading. The default location for fingerprints is /etc/pf.os. set hostid number The 32-bit hostid number identifies this firewall's state table entries to other firewalls in a pfsync(4) failover cluster. By default the hostid is set to a pseudo-random value, however it may be desirable to manually configure it, for example to more easily identify the source of state table entries. The hostid may be specified in either decimal or hexadecimal. set limit limit-item number Sets hard limits on the memory pools used by the packet filter. See pool(9) for an explanation of memory pools. Limits can be set on the following: states Set the maximum number of entries in the memory pool used by state table entries (those generated by pass rules which do not specify no state). The default is 100000. src-nodes Set the maximum number of entries in the memory pool used for tracking source IP addresses (generated by the sticky-address and src.track options). The default is 10000. frags Set the maximum number of entries in the memory pool used for fragment reassembly. The maximum may not exceed, and should be well below, the maximum number of mbuf clusters (sysctl kern.maxclusters) in the system. The default is NMBCLUSTERS/32. NMBCLUSTERS defines the total number of packets which can exist in-system at any one time. Refer to <machine/param.h> for the platform-specific value. tables Set the number of tables that can exist. The default is 1000. table-entries Set the number of addresses that can be stored in tables. The default is 200000, or 100000 on machines with less than 100MB of physical memory. pktdelay_pkts Set the maximum number of packets that can be held in the delay queue. The default is 10000. anchors Set the number of anchors that can exist. The default is 512. Multiple limits can be combined on a single line: set limit { states 20000, frags 2000, src-nodes 2000 } set loginterface interface | none Enable collection of packet and byte count statistics for the given interface or interface group. These statistics can be viewed using: # pfctl -s info In this example pf(4) collects statistics on the interface named dc0: set loginterface dc0 One can disable the loginterface using: set loginterface none The default value is none. set optimization environment Optimize state timeouts for one of the following network environments: aggressive Aggressively expire connections. This can greatly reduce the memory usage of the firewall at the cost of dropping idle connections early. conservative Extremely conservative settings. Avoid dropping legitimate connections at the expense of greater memory utilization (possibly much greater on a busy network) and slightly increased processor utilization. high-latency A high-latency environment (such as a satellite connection). normal A normal network environment. Suitable for almost all networks. satellite Alias for high-latency. The default value is normal. set reassemble yes | no [no-df] The reassemble option is used to enable or disable the reassembly of fragmented packets, and can be set to yes (the default) or no. If no-df is also specified, fragments with the "dont-fragment" bit set are reassembled too, instead of being dropped; the reassembled packet will have the "dont-fragment" bit cleared. The default value is yes. set ruleset-optimization level basic Enable basic ruleset optimization. This is the default behaviour. Basic ruleset optimization does four things to improve the performance of ruleset evaluations: 1. remove duplicate rules 2. remove rules that are a subset of another rule 3. combine multiple rules into a table when advantageous 4. reorder the rules to improve evaluation performance none Disable the ruleset optimizer. profile Uses the currently loaded ruleset as a feedback profile to tailor the ordering of quick rules to actual network traffic. It is important to note that the ruleset optimizer will modify the ruleset to improve performance. A side effect of the ruleset modification is that per-rule accounting statistics will have different meanings than before. If per-rule accounting is important for billing purposes or whatnot, either the ruleset optimizer should not be used or a label field should be added to all of the accounting rules to act as optimization barriers. Optimization can also be set as a command-line argument to pfctl(8), overriding the settings in pf.conf. set skip on ifspec List interfaces for which packets should not be filtered. Packets passing in or out on such interfaces are passed as if pf was disabled, i.e. pf does not process them in any way. This can be useful on loopback and other virtual interfaces, when packet filtering is not desired and can have unexpected effects. PF filters traffic on all interfaces by default. set state-defaults state-option, ... The state-defaults option sets the state options for states created from rules without an explicit keep state. For example: set state-defaults pflow, no-sync set state-policy if-bound | floating The state-policy option sets the default behaviour for states: if-bound States are bound to an interface. floating States can match packets on any interfaces (the default). set syncookies never | always | adaptive When syncookies are active, pf will answer each and every incoming TCP SYN with a syncookie SYNACK, without allocating any resources. Upon reception of the client's ACK in response to the syncookie SYNACK, pf will evaluate the ruleset and create state if the ruleset permits it, complete the three way handshake with the target host, and continue the connection with synproxy in place. This allows pf to be resilient against large synflood attacks, which could otherwise exhaust the state table. Due to the blind answers to each and every SYN, syncookies share the caveats of synproxy: seemingly accepting connections that will be dropped later on. never pf will never send syncookie SYNACKs (the default). always pf will always send syncookie SYNACKs. adaptive pf will enable syncookie mode when a given percentage of the state table is used up by half-open TCP connections, such as those that saw the initial SYN but didn't finish the three way handshake. The thresholds for entering and leaving syncookie mode can be specified using: set syncookies adaptive (start 25%, end 12%) set timeout variable value frag Seconds before an unassembled fragment is expired (60 by default). interval Interval between purging expired states and fragments (10 seconds by default). src.track Length of time to retain a source tracking entry after the last state expires (0 by default, which means there is no global limit. The value is defined by the rule which creates the state.). When a packet matches a stateful connection, the seconds to live for the connection will be updated to that of the protocol and modifier which corresponds to the connection state. Each packet which matches this state will reset the TTL. Tuning these values may improve the performance of the firewall at the risk of dropping valid idle connections. Alternatively, these values may be adjusted collectively in a manner suitable for a specific environment using set optimization (see above). tcp.closed (90 seconds by default) The state after one endpoint sends an RST. tcp.closing (900 seconds by default) The state after the first FIN has been sent. tcp.established (24 hours by default) The fully established state. tcp.finwait (45 seconds by default) The state after both FINs have been exchanged and the connection is closed. Some hosts (notably web servers on Solaris) send TCP packets even after closing the connection. Increasing tcp.finwait (and possibly tcp.closing) can prevent blocking of such packets. tcp.first (120 seconds by default) The state after the first packet. tcp.opening (30 seconds by default) The state after the second packet but before both endpoints have acknowledged the connection. tcp.tsdiff (30 seconds by default) Maximum allowed time difference between RFC 1323 compliant packet timestamps. ICMP and UDP are handled in a fashion similar to TCP, but with a much more limited set of states: icmp.error (10 seconds by default) The state after an ICMP error came back in response to an ICMP packet. icmp.first (20 seconds by default) The state after the first packet. udp.first (60 seconds by default) The state after the first packet. udp.multiple (60 seconds by default) The state if both hosts have sent packets. udp.single (30 seconds by default) The state if the source host sends more than one packet but the destination host has never sent one back. Other protocols are handled similarly to UDP: other.first (60 seconds by default) other.multiple (60 seconds by default) other.single (30 seconds by default) Timeout values can be reduced adaptively as the number of state table entries grows. adaptive.start (60000 states by default) When the number of state entries exceeds this value, adaptive scaling begins. All timeout values are scaled linearly with factor (adaptive.end - number of states) / (adaptive.end - adaptive.start). adaptive.end (120000 states by default) When reaching this number of state entries, all timeout values become zero, effectively purging all state entries immediately. This value is used to define the scale factor; it should not actually be reached (set a lower state limit, see below). Adaptive timeouts are enabled by default, with an adaptive.start value equal to 60% of the state limit, and an adaptive.end value equal to 120% of the state limit. They can be disabled by setting both adaptive.start and adaptive.end to 0. The adaptive timeout values can be defined both globally and for each rule. When used on a per-rule basis, the values relate to the number of states created by the rule, otherwise to the total number of states. For example: set timeout tcp.first 120 set timeout tcp.established 86400 set timeout { adaptive.start 60000, adaptive.end 120000 } set limit states 100000 With 90000 state table entries, the timeout values are scaled to 50% (tcp.first 60, tcp.established 43200). "pfctl -F Reset" restores default values for the following options: debug, all limit options, loginterface, reassemble, skip, syncookies, all timeouts. QUEUEING Packets can be assigned to queues for the purpose of bandwidth control. At least one declaration is required to configure queues, and later any packet filtering rule can reference the defined queues by name. When filtering, the last referenced queue name is where any passed packets will be queued, while for blocked packets it specifies where any resulting ICMP or TCP RST packets should be queued. If the referenced queue does not exist on the outgoing interface, the default queue for that interface is used. Queues attached to an interface build a tree, thus each queue can have further child queues. Only leaf queues, i.e. queues without children, can be used to assign packets to. The root queue must specifically reference an interface, all other queues pick up the interfaces they should be created on from their parent queues. In the following example, a queue named std is created on the interface em0, with 3 child queues ssh, mail and http: queue std on em0 bandwidth 100M queue ssh parent std bandwidth 10M queue mail parent std bandwidth 10M queue http parent std bandwidth 80M default The specified bandwidth is the target bandwidth, every queue can receive more bandwidth as long as the parent still has some available. The maximum bandwidth that should be assigned to a given queue can be limited using the max keyword. If a limitation isn't imposed on the root queue, borrowing can result in saturating the bandwidth of the outgoing interface. Similarly, a minimum (reserved) bandwidth can be specified: queue ssh parent std bandwidth 10M min 5M max 25M For each of these 3 bandwidth specifications an additional burst bandwidth and time can be specified: queue ssh parent std bandwidth 10M burst 90M for 100ms All bandwidth values are specified as bits per second or using the suffixes K, M, and G to represent kilobits, megabits, and gigabits per second, respectively. The value must not exceed the interface bandwidth. If multiple connections are assigned the same queue, they're not guaranteed to share the queue bandwidth fairly. An alternative flow queue manager can be used to achieve fair sharing by indicating how many simultaneous states are expected with a flows option, unless a minimum bandwidth has been specified as well. When packets are classified by the stateful inspection engine, a flow identifier is assigned to all packets belonging to the state, thus limiting the number of individual flows that can be recognized by the resolution of a flow identifier. The current implementation is able to classify traffic into 32767 distinct flows. However, efficient fair sharing is observed even with a much smaller number of flows. For example on a 10Mbit/s DSL or a cable modem uplink, the following simple configuration can be used: queue outq on em0 bandwidth 9M max 9M flows 1024 qlimit 1024 \ default It's important to specify the upper bound within 90-95% of the expected bandwidth and raise the default queue limit. If a flows option appears without a bandwidth specification, the flow queue manager is selected as the queueing discipline for the corresponding interface acting as a default queue for all outgoing packets. In such a scenario, a queueing hierarchy is not supported. In addition to the bandwidth and flow specifications, queues support the following options: default Packets not matched by another queue are assigned to this queue. Exactly one default queue per interface is required. on interface Specifies the interface the queue operates on. If not given, it operates on all matching interfaces. parent name Defines which parent queue the queue should be attached to. Mandatory for all queues except root queues. The parent queue must exist. quantum size Specifies the quantum of service for the flow queue manager. The lower the quantum size the more advantage is given to streams of smaller packets at the expense of bulk transfers. The default value is set to the configured Maximum Transmission Unit (MTU) of the specified interface. qlimit limit The maximum number of packets held in the queue. The default is 50. Packets can be assigned to queues based on filter rules by using the queue keyword. Normally only one queue is specified; when a second one is specified it will instead be used for packets which have a TOS of lowdelay and for TCP ACKs with no data payload. To continue the previous example, the examples below would specify the four referenced queues, plus a few child queues. Interactive ssh(1) sessions get a queue with a minimum bandwidth; scp(1) and sftp(1) bulk transfers go to a separate queue. The queues are then referenced by filtering rules. queue rootq on em0 bandwidth 100M max 100M queue http parent rootq bandwidth 60M burst 90M for 100ms queue developers parent http bandwidth 45M queue employees parent http bandwidth 15M queue mail parent rootq bandwidth 10M queue ssh parent rootq bandwidth 20M queue ssh_interactive parent ssh bandwidth 10M min 5M queue ssh_bulk parent ssh bandwidth 10M queue std parent rootq bandwidth 20M default block return out on em0 inet all set queue std pass out on em0 inet proto tcp from $developerhosts to any port 80 \ set queue developers pass out on em0 inet proto tcp from $employeehosts to any port 80 \ set queue employees pass out on em0 inet proto tcp from any to any port 22 \ set queue(ssh_bulk, ssh_interactive) pass out on em0 inet proto tcp from any to any port 25 \ set queue mail TABLES Tables are named structures which can hold a collection of addresses and networks. Lookups against tables in pf(4) are relatively fast, making a single rule with tables much more efficient, in terms of processor usage and memory consumption, than a large number of rules which differ only in IP address (either created explicitly or automatically by rule expansion). Tables can be used as the source or destination of filter or translation rules. They can also be used for the redirect address of nat-to and rdr-to and in the routing options of filter rules, but not for bitmask pools. Tables can be defined with any of the following pfctl(8) mechanisms. As with macros, reserved words may not be used as table names. manually Persistent tables can be manually created with the add or replace option of pfctl(8), before or after the ruleset has been loaded. pf.conf Table definitions can be placed directly in this file and loaded at the same time as other rules are loaded, atomically. Table definitions inside pf.conf use the table statement, and are especially useful to define non-persistent tables. The contents of a pre-existing table defined without a list of addresses to initialize it is not altered when pf.conf is loaded. A table initialized with the empty list, { }, will be cleared on load. Tables may be defined with the following attributes: const The const flag prevents the user from altering the contents of the table once it has been created. Without that flag, pfctl(8) can be used to add or remove addresses from the table at any time, even when running with securelevel(7) = 2. counters The counters flag enables per-address packet and byte counters, which can be displayed with pfctl(8). persist The persist flag forces the kernel to keep the table even when no rules refer to it. If the flag is not set, the kernel will automatically remove the table when the last rule referring to it is flushed. This example creates a table called "private", to hold RFC 1918 private network blocks, and a table called "badhosts", which is initially empty. A filter rule is set up to block all traffic coming from addresses listed in either table: table <private> const { 10/8, 172.16/12, 192.168/16 } table <badhosts> persist block on fxp0 from { <private>, <badhosts> } to any The private table cannot have its contents changed and the badhosts table will exist even when no active filter rules reference it. Addresses may later be added to the badhosts table, so that traffic from these hosts can be blocked by using the following: # pfctl -t badhosts -Tadd 204.92.77.111 A table can also be initialized with an address list specified in one or more external files, using the following syntax: table <spam> persist file "/etc/spammers" file "/etc/openrelays" block on fxp0 from <spam> to any The files /etc/spammers and /etc/openrelays list IP addresses, one per line. Any lines beginning with a `#' are treated as comments and ignored. In addition to being specified by IP address, hosts may also be specified by their hostname. When the resolver is called to add a hostname to a table, all resulting IPv4 and IPv6 addresses are placed into the table. IP addresses can also be entered in a table by specifying a valid interface name, a valid interface group, or the self keyword, in which case all addresses assigned to the interface(s) will be added to the table. ANCHORS Besides the main ruleset, pf.conf can specify anchor attachment points. An anchor is a container that can hold rules, address tables, and other anchors. When evaluation of the main ruleset reaches an anchor rule, pf(4) will proceed to evaluate all rules specified in that anchor. The following example blocks all packets on the external interface by default, then evaluates all rules in the anchor named "spam", and finally passes all outgoing connections and incoming connections to port 25: ext_if = "kue0" block on $ext_if all anchor spam pass out on $ext_if all pass in on $ext_if proto tcp from any to $ext_if port smtp Anchors can be manipulated through pfctl(8) without reloading the main ruleset or other anchors. This loads a single rule into the anchor, which blocks all packets from a specific address: # echo "block in quick from 1.2.3.4 to any" | pfctl -a spam -f - The anchor can also be populated by adding a load anchor rule after the anchor rule. When pfctl(8) loads pf.conf, it will also load all the rules from the file /etc/pf-spam.conf into the anchor. anchor spam load anchor spam from "/etc/pf-spam.conf" An anchor rule can also contain a filter ruleset in a brace-delimited block. In that case, no separate loading of rules into the anchor is required. Brace delimited blocks may contain rules or other brace- delimited blocks. When an anchor is populated this way, the anchor name becomes optional. Since the parser specification for anchor names is a string, double quote characters (`"') should be placed around the anchor name. anchor "external" on egress { block anchor out { pass proto tcp from any to port { 25, 80, 443 } } pass in proto tcp to any port 22 } Anchor rules can also specify packet filtering parameters using the same syntax as filter rules. When parameters are used, the anchor rule is only evaluated for matching packets. This allows conditional evaluation of anchors, like: block on $ext_if all anchor spam proto tcp from any to any port smtp pass out on $ext_if all pass in on $ext_if proto tcp from any to $ext_if port smtp The rules inside anchor "spam" are only evaluated for TCP packets with destination port 25. Hence, the following will only block connections from 1.2.3.4 to port 25: # echo "block in quick from 1.2.3.4 to any" | pfctl -a spam -f - Matching filter and translation rules marked with the quick option are final and abort the evaluation of the rules in other anchors and the main ruleset. If the anchor itself is marked with the quick option, ruleset evaluation will terminate when the anchor is exited if the packet is matched by any rule within the anchor. An anchor references other anchor attachment points using the following syntax: anchor name Evaluates the filter rules in the specified anchor. An anchor has a name which specifies the path where pfctl(8) can be used to access the anchor to perform operations on it, such as attaching child anchors to it or loading rules into it. Anchors may be nested, with components separated by `/' characters, similar to how file system hierarchies are laid out. The main ruleset is actually the default anchor, so filter and translation rules, for example, may also be contained in any anchor. Anchor rules are evaluated relative to the anchor in which they are contained. For example, all anchor rules specified in the main ruleset will reference anchor attachment points underneath the main ruleset, and anchor rules specified in a file loaded from a load anchor rule will be attached under that anchor point. Anchors may end with the asterisk (`*') character, which signifies that all anchors attached at that point should be evaluated in the alphabetical ordering of their anchor name. For example, the following will evaluate each rule in each anchor attached to the "spam" anchor: anchor "spam/*" Note that it will only evaluate anchors that are directly attached to the "spam" anchor, and will not descend to evaluate anchors recursively. Since anchors are evaluated relative to the anchor in which they are contained, there is a mechanism for accessing the parent and ancestor anchors of a given anchor. Similar to file system path name resolution, if the sequence `..' appears as an anchor path component, the parent anchor of the current anchor in the path evaluation at that point will become the new current anchor. As an example, consider the following: # printf 'anchor "spam/allowed"\n' | pfctl -f - # printf 'anchor "../banned"\npass\n' | pfctl -a spam/allowed -f - Evaluation of the main ruleset will lead into the spam/allowed anchor, which will evaluate the rules in the spam/banned anchor, if any, before finally evaluating the pass rule. STATEFUL FILTERING pf(4) filters packets statefully, which has several advantages. For TCP connections, comparing a packet to a state involves checking its sequence numbers, as well as TCP timestamps if a rule using the reassemble tcp parameter applies to the connection. If these values are outside the narrow windows of expected values, the packet is dropped. This prevents spoofing attacks, such as when an attacker sends packets with a fake source address/port but does not know the connection's sequence numbers. Similarly, pf(4) knows how to match ICMP replies to states. For example, to allow echo requests (such as those created by ping(8)) out statefully and match incoming echo replies correctly to states: pass out inet proto icmp all icmp-type echoreq Also, looking up states is usually faster than evaluating rules. If there are 50 rules, all of them are evaluated sequentially in O(n). Even with 50000 states, only 16 comparisons are needed to match a state, since states are stored in a binary search tree that allows searches in O(log2 n). Furthermore, correct handling of ICMP error messages is critical to many protocols, particularly TCP. pf(4) matches ICMP error messages to the correct connection, checks them against connection parameters, and passes them if appropriate. For example if an ICMP source quench message referring to a stateful TCP connection arrives, it will be matched to the state and get passed. Finally, state tracking is required for nat-to and rdr-to options, in order to track address and port translations and reverse the translation on returning packets. pf(4) will also create state for other protocols which are effectively stateless by nature. UDP packets are matched to states using only host addresses and ports, and other protocols are matched to states using only the host addresses. If stateless filtering of individual packets is desired, the no state keyword can be used to specify that state will not be created if this is the last matching rule. Note that packets which match neither block nor pass rules, and thus are passed by default, are effectively passed as if no state had been specified. A number of parameters can also be set to affect how pf(4) handles state tracking, as detailed below. State Modulation Much of the security derived from TCP is attributable to how well the initial sequence numbers (ISNs) are chosen. Some popular stack implementations choose very poor ISNs and thus are normally susceptible to ISN prediction exploits. By applying a modulate state rule to a TCP connection, pf(4) will create a high quality random sequence number for each connection endpoint. The modulate state directive implicitly keeps state on the rule and is only applicable to TCP connections. For instance: block all pass out proto tcp from any to any modulate state pass in proto tcp from any to any port 25 flags S/SFRA \ modulate state Note that modulated connections will not recover when the state table is lost (firewall reboot, flushing the state table, etc.). pf(4) will not be able to infer a connection again after the state table flushes the connection's modulator. When the state is lost, the connection may be left dangling until the respective endpoints time out the connection. It is possible on a fast local network for the endpoints to start an ACK storm while trying to resynchronize after the loss of the modulator. The default flags settings (or a more strict equivalent) should be used on modulate state rules to prevent ACK storms. Note that alternative methods are available to prevent loss of the state table and allow for firewall failover. See carp(4) and pfsync(4) for further information. SYN Proxy By default, pf(4) passes packets that are part of a TCP handshake between the endpoints. The synproxy state option can be used to cause pf(4) itself to complete the handshake with the active endpoint, perform a handshake with the passive endpoint, and then forward packets between the endpoints. No packets are sent to the passive endpoint before the active endpoint has completed the handshake, hence so-called SYN floods with spoofed source addresses will not reach the passive endpoint, as the sender can't complete the handshake. The proxy is transparent to both endpoints; they each see a single connection from/to the other endpoint. pf(4) chooses random initial sequence numbers for both handshakes. Once the handshakes are completed, the sequence number modulators (see previous section) are used to translate further packets of the connection. synproxy state includes modulate state. Rules with synproxy state will not work if pf(4) operates on a bridge(4). Also they act on incoming SYN packets only. Example: pass in proto tcp from any to any port www synproxy state Stateful Tracking Options A number of options related to stateful tracking can be applied on a per- rule basis. One of keep state, modulate state, or synproxy state must be specified explicitly to apply these options to a rule. floating States can match packets on any interfaces (the opposite of if-bound). This is the default. if-bound States are bound to an interface (the opposite of floating). max number Limits the number of concurrent states the rule may create. When this limit is reached, further packets that would create state are dropped until existing states time out. no-sync Prevent state changes for states created by this rule from appearing on the pfsync(4) interface. pflow States created by this rule are exported on the pflow(4) interface. sloppy For TCP, uses a sloppy connection tracker that does not check sequence numbers at all, which makes insertion and ICMP teardown attacks way easier. This is intended to be used in situations where one does not see all packets of a connection, e.g. in asymmetric routing situations. It cannot be used with modulate state or synproxy state. For ICMP, this option allows states to be created from replies, not just requests. timeout seconds Changes the timeout values used for states created by this rule. For a list of all valid timeout names, see OPTIONS above. Multiple options can be specified, separated by commas: pass in proto tcp from any to any \ port www keep state \ (max 100, source-track rule, max-src-nodes 75, \ max-src-states 3, tcp.established 60, tcp.closing 5) When the source-track keyword is specified, the number of states per source IP is tracked. source-track global The number of states created by all rules that use this option is limited. Each rule can specify different max-src-nodes and max-src-states options, however state entries created by any participating rule count towards each individual rule's limits. source-track rule The maximum number of states created by this rule is limited by the rule's max-src-nodes and max-src-states options. Only state entries created by this particular rule count toward the rule's limits. The following limits can be set: max-src-nodes number Limits the maximum number of source addresses which can simultaneously have state table entries. max-src-states number Limits the maximum number of simultaneous state entries that a single source address can create with this rule. For stateful TCP connections, limits on established connections (connections which have completed the TCP 3-way handshake) can also be enforced per source IP. max-src-conn number Limits the maximum number of simultaneous TCP connections which have completed the 3-way handshake that a single host can make. max-src-conn-rate number/seconds Limit the rate of new connections over a time interval. The connection rate is an approximation calculated as a moving average. When one of these limits is reached, further packets that would create state are dropped until existing states time out. Because the 3-way handshake ensures that the source address is not being spoofed, more aggressive action can be taken based on these limits. With the overload <table> state option, source IP addresses which hit either of the limits on established connections will be added to the named table. This table can be used in the ruleset to block further activity from the offending host, redirect it to a tarpit process, or restrict its bandwidth. The optional flush keyword kills all states created by the matching rule which originate from the host which exceeds these limits. The global modifier to the flush command kills all states originating from the offending host, regardless of which rule created the state. For example, the following rules will protect the webserver against hosts making more than 100 connections in 10 seconds. Any host which connects faster than this rate will have its address added to the <bad_hosts> table and have all states originating from it flushed. Any new packets arriving from this host will be dropped unconditionally by the block rule. block quick from <bad_hosts> pass in on $ext_if proto tcp to $webserver port www keep state \ (max-src-conn-rate 100/10, overload <bad_hosts> flush global) TRAFFIC NORMALISATION Traffic normalisation is a broad umbrella term for aspects of the packet filter which deal with verifying packets, packet fragments, spoof traffic, and other irregularities. Scrub Scrub involves sanitising packet content in such a way that there are no ambiguities in packet interpretation on the receiving side. It is invoked with the scrub option, added to regular rules. Parameters are specified enclosed in parentheses. At least one of the following parameters must be specified: max-mss number Reduces the maximum segment size (MSS) on TCP SYN packets to be no greater than number. This is sometimes required in scenarios where the two endpoints of a TCP connection are not able to carry similar sized packets and the resulting mismatch can lead to packet fragmentation or loss. Note that setting the MSS this way can have undesirable effects, such as interfering with the OS detection features of pf(4). min-ttl number Enforces a minimum TTL for matching IP packets. no-df Clears the "dont-fragment" bit from a matching IPv4 packet. Some operating systems have NFS implementations which are known to generate fragmented packets with the "dont-fragment" bit set. pf(4) will drop such fragmented "dont-fragment" packets unless no-df is specified. Unfortunately some operating systems also generate their "dont-fragment" packets with a zero IP identification field. Clearing the "dont-fragment" bit on packets with a zero IP ID may cause deleterious results if an upstream router later fragments the packet. Using random-id is recommended in combination with no-df to ensure unique IP identifiers. random-id Replaces the IPv4 identification field with random values to compensate for predictable values generated by many hosts. This option only applies to packets that are not fragmented after the optional fragment reassembly. reassemble tcp Statefully normalises TCP connections. reassemble tcp performs the following normalisations: TTL Neither side of the connection is allowed to reduce their IP TTL. An attacker may send a packet such that it reaches the firewall, affects the firewall state, and expires before reaching the destination host. reassemble tcp will raise the TTL of all packets back up to the highest value seen on the connection. Timestamp Modulation Modern TCP stacks will send a timestamp on every TCP packet and echo the other endpoint's timestamp back to them. Many operating systems will merely start the timestamp at zero when first booted, and increment it several times a second. The uptime of the host can be deduced by reading the timestamp and multiplying by a constant. Also observing several different timestamps can be used to count hosts behind a NAT device. And spoofing TCP packets into a connection requires knowing or guessing valid timestamps. Timestamps merely need to be monotonically increasing and not derived off a guessable base time. reassemble tcp will cause scrub to modulate the TCP timestamps with a random number. Extended PAWS Checks There is a problem with TCP on long fat pipes, in that a packet might get delayed for longer than it takes the connection to wrap its 32-bit sequence space. In such an occurrence, the old packet would be indistinguishable from a new packet and would be accepted as such. The solution to this is called PAWS: Protection Against Wrapped Sequence numbers. It protects against it by making sure the timestamp on each packet does not go backwards. reassemble tcp also makes sure the timestamp on the packet does not go forward more than the RFC allows. By doing this, pf(4) artificially extends the security of TCP sequence numbers by 10 to 18 bits when the host uses appropriately randomized timestamps, since a blind attacker would have to guess the timestamp as well. For example: match in all scrub (no-df random-id max-mss 1440) Fragment Handling The size of IP datagrams (packets) can be significantly larger than the maximum transmission unit (MTU) of the network. In cases when it is necessary or more efficient to send such large packets, the large packet will be fragmented into many smaller packets that will each fit onto the wire. Unfortunately for a firewalling device, only the first logical fragment will contain the necessary header information for the subprotocol that allows pf(4) to filter on things such as TCP ports or to perform NAT. One alternative is to filter individual fragments with filter rules. If packet reassembly is turned off, it is passed to the filter. Filter rules with matching IP header parameters decide whether the fragment is passed or blocked, in the same way as complete packets are filtered. Without reassembly, fragments can only be filtered based on IP header fields (source/destination address, protocol), since subprotocol header fields are not available (TCP/UDP port numbers, ICMP code/type). The fragment option can be used to restrict filter rules to apply only to fragments, but not complete packets. Filter rules without the fragment option still apply to fragments, if they only specify IP header fields. For instance: pass in proto tcp from any to any port 80 The rule above never applies to a fragment, even if the fragment is part of a TCP packet with destination port 80, because without reassembly this information is not available for each fragment. This also means that fragments cannot create new or match existing state table entries, which makes stateful filtering and address translation (NAT, redirection) for fragments impossible. In most cases, the benefits of reassembly outweigh the additional memory cost, so reassembly is on by default. The memory allocated for fragment caching can be limited using pfctl(8). Once this limit is reached, fragments that would have to be cached are dropped until other entries time out. The timeout value can also be adjusted. When forwarding reassembled IPv6 packets, pf refragments them with the original maximum fragment size. This allows the sender to determine the optimal fragment size by path MTU discovery. Blocking Spoofed Traffic Spoofing is the faking of IP addresses, typically for malicious purposes. The antispoof directive expands to a set of filter rules which will block all traffic with a source IP from the network(s) directly connected to the specified interface(s) from entering the system through any other interface. For example: antispoof for lo0 Expands to: block drop in on ! lo0 inet from 127.0.0.1/8 to any block drop in on ! lo0 inet6 from ::1 to any For non-loopback interfaces, there are additional rules to block incoming packets with a source IP address identical to the interface's IP(s). For example, assuming the interface wi0 had an IP address of 10.0.0.1 and a netmask of 255.255.255.0: antispoof for wi0 inet Expands to: block drop in on ! wi0 inet from 10.0.0.0/24 to any block drop in inet from 10.0.0.1 to any Caveat: Rules created by the antispoof directive interfere with packets sent over loopback interfaces to local addresses. One should pass these explicitly. OPERATING SYSTEM FINGERPRINTING Passive OS fingerprinting is a mechanism to inspect nuances of a TCP connection's initial SYN packet and guess at the host's operating system. Unfortunately these nuances are easily spoofed by an attacker so the fingerprint is not useful in making security decisions. But the fingerprint is typically accurate enough to make policy decisions upon. The fingerprints may be specified by operating system class, by version, or by subtype/patchlevel. The class of an operating system is typically the vendor or genre and would be OpenBSD for the pf(4) firewall itself. The version of the oldest available OpenBSD release on the main FTP site would be 2.6 and the fingerprint would be written as: "OpenBSD 2.6" The subtype of an operating system is typically used to describe the patchlevel if that patch led to changes in the TCP stack behavior. In the case of OpenBSD, the only subtype is for a fingerprint that was normalised by the no-df scrub option and would be specified as: "OpenBSD 3.3 no-df" Fingerprints for most popular operating systems are provided by pf.os(5). Once pf(4) is running, a complete list of known operating system fingerprints may be listed by running: # pfctl -so Filter rules can enforce policy at any level of operating system specification assuming a fingerprint is present. Policy could limit traffic to approved operating systems or even ban traffic from hosts that aren't at the latest service pack. The unknown class can also be used as the fingerprint which will match packets for which no operating system fingerprint is known. Examples: pass out proto tcp from any os OpenBSD block out proto tcp from any os Doors block out proto tcp from any os "Doors PT" block out proto tcp from any os "Doors PT SP3" block out from any os "unknown" pass on lo0 proto tcp from any os "OpenBSD 3.3 lo0" Operating system fingerprinting is limited only to the TCP SYN packet. This means that it will not work on other protocols and will not match a currently established connection. Caveat: operating system fingerprints are occasionally wrong. There are three problems: an attacker can trivially craft packets to appear as any operating system; an operating system patch could change the stack behavior and no fingerprints will match it until the database is updated; and multiple operating systems may have the same fingerprint. EXAMPLES In this example, the external interface is kue0. We use a macro for the interface name, so it can be changed easily. All incoming traffic is "normalised", and everything is blocked and logged by default. ext_if = "kue0" match in all scrub (no-df max-mss 1440) block return log on $ext_if all Here we specifically block packets we don't want: anything coming from source we have no back routes for; packets whose ingress interface does not match the one in the route back to their source address; anything that does not have our address (157.161.48.183) as source; broadcasts (cable modem noise); and anything from reserved address space or invalid addresses. block in from no-route to any block in from urpf-failed to any block out log quick on $ext_if from ! 157.161.48.183 to any block in quick on $ext_if from any to 255.255.255.255 block in log quick on $ext_if from { 10.0.0.0/8, 172.16.0.0/12, \ 192.168.0.0/16, 255.255.255.255/32 } to any For ICMP, pass out/in ping queries. State matching is done on host addresses and ICMP ID (not type/code), so replies (like 0/0 for 8/0) will match queries. ICMP error messages (which always refer to a TCP/UDP packet) are handled by the TCP/UDP states. pass on $ext_if inet proto icmp all icmp-type 8 code 0 For UDP, pass out all UDP connections. DNS connections are passed in. pass out on $ext_if proto udp all pass in on $ext_if proto udp from any to any port domain For TCP, pass out all TCP connections and modulate state. SSH, SMTP, DNS, and IDENT connections are passed in. We do not allow Windows 9x SMTP connections since they are typically a viral worm. pass out on $ext_if proto tcp all modulate state pass in on $ext_if proto tcp from any to any \ port { ssh, smtp, domain, auth } block in on $ext_if proto tcp from any \ os { "Windows 95", "Windows 98" } to any port smtp Here we pass in/out all IPv6 traffic: note that we have to enable this in two different ways, on both our physical interface and our tunnel. pass quick on gif0 inet6 pass quick on $ext_if proto ipv6 This example illustrates packet tagging. There are three interfaces: $int_if, $ext_if, and $wifi_if (wireless). NAT is being done on $ext_if for all outgoing packets. Packets in on $int_if are tagged and passed out on $ext_if. All other outgoing packets (i.e. packets from the wireless network) are only permitted to access port 80. pass in on $int_if from any to any tag INTNET pass in on $wifi_if from any to any block out on $ext_if from any to any pass out quick on $ext_if tagged INTNET pass out on $ext_if proto tcp from any to any port 80 In this example, we tag incoming packets as they are redirected to spamd(8). The tag is used to pass those packets through the packet filter. match in on $ext_if inet proto tcp from <spammers> to port smtp \ tag SPAMD rdr-to 127.0.0.1 port spamd block in on $ext_if pass in on $ext_if inet proto tcp tagged SPAMD This example maps incoming requests on port 80 to port 8080, on which a daemon is running (because, for example, it is not run as root, and therefore lacks permission to bind to port 80). match in on $ext_if proto tcp from any to any port 80 \ rdr-to 127.0.0.1 port 8080 If a pass rule is used with the quick modifier, packets matching the translation rule are passed without inspecting subsequent filter rules. pass in quick on $ext_if proto tcp from any to any port 80 \ rdr-to 127.0.0.1 port 8080 In the example below, vlan12 is configured as 192.168.168.1; the machine translates all packets coming from 192.168.168.0/24 to 204.92.77.111 when they are going out any interface except vlan12. This has the net effect of making traffic from the 192.168.168.0/24 network appear as though it is the Internet routable address 204.92.77.111 to nodes behind any interface on the router except for the nodes on vlan12. Thus, 192.168.168.1 can talk to the 192.168.168.0/24 nodes. match out on ! vlan12 from 192.168.168.0/24 to any nat-to 204.92.77.111 In the example below, the machine sits between a fake internal 144.19.74.* network, and a routable external IP of 204.92.77.100. The last rule excludes protocol AH from being translated. pass out on $ext_if from 144.19.74.0/24 nat-to 204.92.77.100 pass out on $ext_if proto ah from 144.19.74.0/24 In the example below, packets bound for one specific server, as well as those generated by the sysadmins are not proxied; all other connections are. pass in on $int_if proto { tcp, udp } from any to any port 80 \ rdr-to 127.0.0.1 port 80 pass in on $int_if proto { tcp, udp } from any to $server port 80 pass in on $int_if proto { tcp, udp } from $sysadmins to any port 80 This example maps outgoing packets' source port to an assigned proxy port instead of an arbitrary port. In this case, proxy outgoing isakmp with port 500 on the gateway. match out on $ext_if inet proto udp from any port isakmp to any \ nat-to ($ext_if) port 500 One more example uses rdr-to to redirect a TCP and UDP port to an internal machine. match in on $ext_if inet proto tcp from any to ($ext_if) port 8080 \ rdr-to 10.1.2.151 port 22 match in on $ext_if inet proto udp from any to ($ext_if) port 8080 \ rdr-to 10.1.2.151 port 53 In this example, a NAT gateway is set up to translate internal addresses using a pool of public addresses (192.0.2.16/28). A given source address is always translated to the same pool address by using the source-hash keyword. The gateway also translates incoming web server connections to a group of web servers on the internal network. match out on $ext_if inet from any to any nat-to 192.0.2.16/28 \ source-hash match in on $ext_if proto tcp from any to any port 80 \ rdr-to { 10.1.2.155 weight 2, 10.1.2.160 weight 1, \ 10.1.2.161 weight 8 } round-robin The bidirectional address translation example uses a single binat-to rule that expands to a nat-to and an rdr-to rule. pass on $ext_if from 10.1.2.120 to any binat-to 192.0.2.17 The previous example is identical to the following set of rules: pass out on $ext_if inet from 10.1.2.120 to any \ nat-to 192.0.2.17 static-port pass in on $ext_if inet from any to 192.0.2.17 rdr-to 10.1.2.120 In the example below, a router handling both address families translates an internal IPv4 subnet to IPv6 using the well-known 64:ff9b::/96 prefix: pass in on $v4_if inet af-to inet6 from ($v6_if) to 64:ff9b::/96 Paired with the example above, the example below can be used on another router handling both address families to translate back to IPv4: pass in on $v6_if inet6 to 64:ff9b::/96 af-to inet from ($v4_if) GRAMMAR Syntax for pf.conf in BNF: line = ( option | pf-rule | antispoof-rule | queue-rule | anchor-rule | anchor-close | load-anchor | table-rule | include ) option = "set" ( [ "timeout" ( timeout | "{" timeout-list "}" ) ] | [ "ruleset-optimization" [ "none" | "basic" | "profile" ] ] | [ "optimization" [ "default" | "normal" | "high-latency" | "satellite" | "aggressive" | "conservative" ] ] [ "limit" ( limit-item | "{" limit-list "}" ) ] | [ "loginterface" ( interface-name | "none" ) ] | [ "block-policy" ( "drop" | "return" ) ] | [ "state-policy" ( "if-bound" | "floating" ) ] [ "state-defaults" state-opts ] [ "fingerprints" filename ] | [ "skip on" ifspec ] | [ "debug" ( "emerg" | "alert" | "crit" | "err" | "warning" | "notice" | "info" | "debug" ) ] | [ "reassemble" ( "yes" | "no" ) [ "no-df" ] ] ) pf-rule = action [ ( "in" | "out" ) ] [ "log" [ "(" logopts ")"] ] [ "quick" ] [ "on" ( ifspec | "rdomain" number ) ] [ af ] [ protospec ] [ hosts ] [ filteropts ] logopts = logopt [ [ "," ] logopts ] logopt = "all" | "matches" | "user" | "to" interface-name filteropts = filteropt [ [ "," ] filteropts ] filteropt = user | group | flags | icmp-type | icmp6-type | "tos" tos | ( "no" | "keep" | "modulate" | "synproxy" ) "state" [ "(" state-opts ")" ] | "scrub" "(" scrubopts ")" | "fragment" | "allow-opts" | "once" | "divert-packet" "port" port | "divert-reply" | "divert-to" host "port" port | "label" string | "tag" string | [ "!" ] "tagged" string | "max-pkt-rate" number "/" seconds | "set delay" number | "set prio" ( number | "(" number [ [ "," ] number ] ")" ) | "set queue" ( string | "(" string [ [ "," ] string ] ")" ) | "rtable" number | "probability" number"%" | "prio" number | "af-to" af "from" ( redirhost | "{" redirhost-list "}" ) [ "to" ( redirhost | "{" redirhost-list "}" ) ] | "binat-to" ( redirhost | "{" redirhost-list "}" ) [ portspec ] [ pooltype ] | "rdr-to" ( redirhost | "{" redirhost-list "}" ) [ portspec ] [ pooltype ] | "nat-to" ( redirhost | "{" redirhost-list "}" ) [ portspec ] [ pooltype ] [ "static-port" ] | [ route ] | [ "set tos" tos ] | [ [ "!" ] "received-on" ( interface-name | interface-group ) ] scrubopts = scrubopt [ [ "," ] scrubopts ] scrubopt = "no-df" | "min-ttl" number | "max-mss" number | "reassemble tcp" | "random-id" antispoof-rule = "antispoof" [ "log" ] [ "quick" ] "for" ifspec [ af ] [ "label" string ] table-rule = "table" "<" string ">" [ tableopts ] tableopts = tableopt [ tableopts ] tableopt = "persist" | "const" | "counters" | "file" string | "{" [ tableaddrs ] "}" tableaddrs = tableaddr-spec [ [ "," ] tableaddrs ] tableaddr-spec = [ "!" ] tableaddr [ "/" mask-bits ] tableaddr = hostname | ifspec | "self" | ipv4-dotted-quad | ipv6-coloned-hex queue-rule = "queue" string [ "on" interface-name ] queueopts-list anchor-rule = "anchor" [ string ] [ ( "in" | "out" ) ] [ "on" ifspec ] [ af ] [ protospec ] [ hosts ] [ filteropt-list ] [ "{" ] anchor-close = "}" load-anchor = "load anchor" string "from" filename queueopts-list = queueopts-list queueopts | queueopts queueopts = ([ "bandwidth" bandwidth ] | [ "min" bandwidth ] | [ "max" bandwidth ] | [ "parent" string ] | [ "default" ]) | ([ "flows" number ] | [ "quantum" number ]) | [ "qlimit" number ] bandwidth = bandwidth-spec [ "burst" bandwidth-spec "for" number "ms" ] bandwidth-spec = number ( "" | "K" | "M" | "G" ) action = "pass" | "match" | "block" [ return ] return = "drop" | "return" | "return-rst" [ "(" "ttl" number ")" ] | "return-icmp" [ "(" icmpcode [ [ "," ] icmp6code ] ")" ] | "return-icmp6" [ "(" icmp6code ")" ] icmpcode = ( icmp-code-name | icmp-code-number ) icmp6code = ( icmp6-code-name | icmp6-code-number ) ifspec = ( [ "!" ] ( interface-name | interface-group ) ) | "{" interface-list "}" interface-list = [ "!" ] ( interface-name | interface-group ) [ [ "," ] interface-list ] route = ( "route-to" | "reply-to" | "dup-to" ) ( redirhost | "{" redirhost-list "}" ) af = "inet" | "inet6" protospec = "proto" ( proto-name | proto-number | "{" proto-list "}" ) proto-list = ( proto-name | proto-number ) [ [ "," ] proto-list ] hosts = "all" | "from" ( "any" | "no-route" | "urpf-failed" | "self" | host | "{" host-list "}" | "route" string ) [ port ] [ os ] "to" ( "any" | "no-route" | "self" | host | "{" host-list "}" | "route" string ) [ port ] ipspec = "any" | host | "{" host-list "}" host = [ "!" ] ( address [ "weight" number ] | address [ "/" mask-bits ] [ "weight" number ] | "<" string ">" ) redirhost = address [ "/" mask-bits ] address = ( interface-name | interface-group | "(" ( interface-name | interface-group ) ")" | hostname | ipv4-dotted-quad | ipv6-coloned-hex ) host-list = host [ [ "," ] host-list ] redirhost-list = redirhost [ [ "," ] redirhost-list ] port = "port" ( unary-op | binary-op | "{" op-list "}" ) portspec = "port" ( number | name ) [ ":" ( "*" | number | name ) ] os = "os" ( os-name | "{" os-list "}" ) user = "user" ( unary-op | binary-op | "{" op-list "}" ) group = "group" ( unary-op | binary-op | "{" op-list "}" ) unary-op = [ "=" | "!=" | "<" | "<=" | ">" | ">=" ] ( name | number ) binary-op = number ( "<>" | "><" | ":" ) number op-list = ( unary-op | binary-op ) [ [ "," ] op-list ] os-name = operating-system-name os-list = os-name [ [ "," ] os-list ] flags = "flags" ( [ flag-set ] "/" flag-set | "any" ) flag-set = [ "F" ] [ "S" ] [ "R" ] [ "P" ] [ "A" ] [ "U" ] [ "E" ] [ "W" ] icmp-type = "icmp-type" ( icmp-type-code | "{" icmp-list "}" ) icmp6-type = "icmp6-type" ( icmp-type-code | "{" icmp-list "}" ) icmp-type-code = ( icmp-type-name | icmp-type-number ) [ "code" ( icmp-code-name | icmp-code-number ) ] icmp-list = icmp-type-code [ [ "," ] icmp-list ] tos = ( "lowdelay" | "throughput" | "reliability" | [ "0x" ] number ) state-opts = state-opt [ [ "," ] state-opts ] state-opt = ( "max" number | "no-sync" | timeout | "sloppy" | "pflow" | "source-track" [ ( "rule" | "global" ) ] | "max-src-nodes" number | "max-src-states" number | "max-src-conn" number | "max-src-conn-rate" number "/" number | "overload" "<" string ">" [ "flush" [ "global" ] ] | "if-bound" | "floating" ) timeout-list = timeout [ [ "," ] timeout-list ] timeout = ( "tcp.first" | "tcp.opening" | "tcp.established" | "tcp.closing" | "tcp.finwait" | "tcp.closed" | "tcp.tsdiff" | "udp.first" | "udp.single" | "udp.multiple" | "icmp.first" | "icmp.error" | "other.first" | "other.single" | "other.multiple" | "frag" | "interval" | "src.track" | "adaptive.start" | "adaptive.end" ) number limit-list = limit-item [ [ "," ] limit-list ] limit-item = ( "states" | "frags" | "src-nodes" | "tables" | "table-entries" ) number pooltype = ( "bitmask" | "least-states" | "random" | "round-robin" | "source-hash" [ ( hex-key | string-key ) ] ) [ "sticky-address" ] include = "include" filename FILES /etc/hosts Host name database. /etc/pf.conf Default location of the ruleset file. /etc/examples/pf.conf Example ruleset file. /etc/pf.os Default location of OS fingerprints. /etc/protocols Protocol name database. /etc/services Service name database. SEE ALSO pf(4), pflow(4), pfsync(4), pf.os(5), pfctl(8), pflogd(8) HISTORY The pf.conf file format first appeared in OpenBSD 3.0. FreeBSD 14.1-RELEASE-p8 April 15, 2024 FreeBSD 14.1-RELEASE-p8