.\" $OpenBSD: pf.4,v 1.62 2008/09/10 14:57:37 jmc Exp $ .\" .\" Copyright (C) 2001, Kjell Wooding. All rights reserved. .\" .\" Redistribution and use in source and binary forms, with or without .\" modification, are permitted provided that the following conditions .\" are met: .\" 1. Redistributions of source code must retain the above copyright .\" notice, this list of conditions and the following disclaimer. .\" 2. Redistributions in binary form must reproduce the above copyright .\" notice, this list of conditions and the following disclaimer in the .\" documentation and/or other materials provided with the distribution. .\" 3. Neither the name of the project nor the names of its contributors .\" may be used to endorse or promote products derived from this software .\" without specific prior written permission. .\" .\" THIS SOFTWARE IS PROVIDED BY THE PROJECT AND CONTRIBUTORS ``AS IS'' AND .\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE .\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE .\" ARE DISCLAIMED. IN NO EVENT SHALL THE PROJECT OR CONTRIBUTORS BE LIABLE .\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL .\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS .\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) .\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT .\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY .\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF .\" SUCH DAMAGE. .\" .Dd September 6, 2024 .Dt PF 4 .Os .Sh NAME .Nm pf .Nd packet filter .Sh SYNOPSIS .Cd "device pf" .Cd "options PF_DEFAULT_TO_DROP" .Sh DESCRIPTION Packet filtering takes place in the kernel. A pseudo-device, .Pa /dev/pf , allows userland processes to control the behavior of the packet filter through an .Xr ioctl 2 interface. There are commands to enable and disable the filter, load rulesets, add and remove individual rules or state table entries, and retrieve statistics. The most commonly used functions are covered by .Xr pfctl 8 . .Pp Manipulations like loading a ruleset that involve more than a single .Xr ioctl 2 call require a so-called .Em ticket , which prevents the occurrence of multiple concurrent manipulations. .Pp Fields of .Xr ioctl 2 parameter structures that refer to packet data (like addresses and ports) are generally expected in network byte-order. .Pp Rules and address tables are contained in so-called .Em anchors . When servicing an .Xr ioctl 2 request, if the anchor field of the argument structure is empty, the kernel will use the default anchor (i.e., the main ruleset) in operations. Anchors are specified by name and may be nested, with components separated by .Sq / characters, similar to how file system hierarchies are laid out. The final component of the anchor path is the anchor under which operations will be performed. .Sh SYSCTL VARIABLES AND LOADER TUNABLES The following .Xr loader 8 tunables are available. .Bl -tag -width indent .It Va net.pf.states_hashsize Size of hash table that stores states. Should be power of 2. Default value is 131072. .It Va net.pf.source_nodes_hashsize Size of hash table that stores source nodes. Should be power of 2. Default value is 32768. .It Va net.pf.rule_tag_hashsize Size of the hash table that stores tags. .It Va net.pf.udpendpoint_hashsize Size of hash table that store UDP endpoint mappings. Should be power of 2. Default value is 32768. .It Va net.pf.default_to_drop This value overrides .Cd "options PF_DEFAULT_TO_DROP" from kernel configuration file. .It Va net.pf.filter_local This tells .Nm to also filter on the loopback output hook. This is typically used to allow redirect rules to adjust the source address. .It Va net.pf.request_maxcount The maximum number of items in a single ioctl call. .It Va net.pf.rdr_srcport_rewrite_tries The maximum number of times to try and find a free source port when handling redirects. Such rules are typically applied to external traffic, so an exhaustive search may be too expensive. .El .Pp Read only .Xr sysctl 8 variables with matching names are provided to obtain current values at runtime. .Sh KERNEL OPTIONS The following options in the kernel configuration file are related to .Nm operation: .Pp .Bl -tag -width ".Dv PF_DEFAULT_TO_DROP" -compact .It Dv PF_DEFAULT_TO_DROP Change default policy to drop by default .El .Sh IOCTL INTERFACE .Nm supports the following .Xr ioctl 2 commands, available through .Aq Pa net/pfvar.h : .Bl -tag -width xxxxxx .It Dv DIOCSTART Start the packet filter. .It Dv DIOCSTOP Stop the packet filter. .It Dv DIOCSTARTALTQ Start the ALTQ bandwidth control system (see .Xr altq 9 ) . .It Dv DIOCSTOPALTQ Stop the ALTQ bandwidth control system. .It Dv DIOCBEGINADDRS Fa "struct pfioc_pooladdr *pp" .Bd -literal struct pfioc_pooladdr { u_int32_t action; u_int32_t ticket; u_int32_t nr; u_int32_t r_num; u_int8_t r_action; u_int8_t r_last; u_int8_t af; char anchor[MAXPATHLEN]; struct pf_pooladdr addr; }; .Ed .Pp Clear the buffer address pool and get a .Va ticket for subsequent .Dv DIOCADDADDR , .Dv DIOCADDRULE , and .Dv DIOCCHANGERULE calls. .It Dv DIOCADDADDR Fa "struct pfioc_pooladdr *pp" .Pp Add the pool address .Va addr to the buffer address pool to be used in the following .Dv DIOCADDRULE or .Dv DIOCCHANGERULE call. All other members of the structure are ignored. .It Dv DIOCADDRULE Fa "struct pfioc_rule *pr" .Bd -literal struct pfioc_rule { u_int32_t action; u_int32_t ticket; u_int32_t pool_ticket; u_int32_t nr; char anchor[MAXPATHLEN]; char anchor_call[MAXPATHLEN]; struct pf_rule rule; }; .Ed .Pp Add .Va rule at the end of the inactive ruleset. This call requires a .Va ticket obtained through a preceding .Dv DIOCXBEGIN call and a .Va pool_ticket obtained through a .Dv DIOCBEGINADDRS call. .Dv DIOCADDADDR must also be called if any pool addresses are required. The optional .Va anchor name indicates the anchor in which to append the rule. .Va nr and .Va action are ignored. .It Dv DIOCADDALTQ Fa "struct pfioc_altq *pa" Add an ALTQ discipline or queue. .Bd -literal struct pfioc_altq { u_int32_t action; u_int32_t ticket; u_int32_t nr; struct pf_altq altq; }; .Ed .It Dv DIOCGETRULES Fa "struct pfioc_rule *pr" Get a .Va ticket for subsequent .Dv DIOCGETRULE calls and the number .Va nr of rules in the active ruleset. .It Dv DIOCGETRULE Fa "struct pfioc_rule *pr" Get a .Va rule by its number .Va nr using the .Va ticket obtained through a preceding .Dv DIOCGETRULES call. If .Va action is set to .Dv PF_GET_CLR_CNTR , the per-rule statistics on the requested rule are cleared. .It Dv DIOCGETADDRS Fa "struct pfioc_pooladdr *pp" Get a .Va ticket for subsequent .Dv DIOCGETADDR calls and the number .Va nr of pool addresses in the rule specified with .Va r_action , .Va r_num , and .Va anchor . .It Dv DIOCGETADDR Fa "struct pfioc_pooladdr *pp" Get the pool address .Va addr by its number .Va nr from the rule specified with .Va r_action , .Va r_num , and .Va anchor using the .Va ticket obtained through a preceding .Dv DIOCGETADDRS call. .It Dv DIOCGETALTQS Fa "struct pfioc_altq *pa" Get a .Va ticket for subsequent .Dv DIOCGETALTQ calls and the number .Va nr of queues in the active list. .It Dv DIOCGETALTQ Fa "struct pfioc_altq *pa" Get the queueing discipline .Va altq by its number .Va nr using the .Va ticket obtained through a preceding .Dv DIOCGETALTQS call. .It Dv DIOCGETQSTATS Fa "struct pfioc_qstats *pq" Get the statistics on a queue. .Bd -literal struct pfioc_qstats { u_int32_t ticket; u_int32_t nr; void *buf; int nbytes; u_int8_t scheduler; }; .Ed .Pp This call fills in a pointer to the buffer of statistics .Va buf , of length .Va nbytes , for the queue specified by .Va nr . .It Dv DIOCGETRULESETS Fa "struct pfioc_ruleset *pr" .Bd -literal struct pfioc_ruleset { u_int32_t nr; char path[MAXPATHLEN]; char name[PF_ANCHOR_NAME_SIZE]; }; .Ed .Pp Get the number .Va nr of rulesets (i.e., anchors) directly attached to the anchor named by .Va path for use in subsequent .Dv DIOCGETRULESET calls. Nested anchors, since they are not directly attached to the given anchor, will not be included. This ioctl returns .Er ENOENT if the parent anchor given at .Va path does not exist. .It Dv DIOCGETRULESET Fa "struct pfioc_ruleset *pr" Get a ruleset (i.e., an anchor) .Va name by its number .Va nr from the given anchor .Va path , the maximum number of which can be obtained from a preceding .Dv DIOCGETRULESETS call. This ioctl returns .Er ENOENT if the parent anchor given by .Va path does not exist or .Er EBUSY if the index passed in by .Va nr is greater than the number of anchors. .It Dv DIOCADDSTATE Fa "struct pfioc_state *ps" Add a state entry. .Bd -literal struct pfioc_state { struct pfsync_state state; }; .Ed .It Dv DIOCGETSTATENV Fa "struct pfioc_nv *nv" Extract the entry identified by the .Va id and .Va creatorid fields of the .Va state nvlist from the state table. .It Dv DIOCKILLSTATESNV Fa "struct pfioc_nv nv" Remove matching entries from the state table. This ioctl returns the number of killed states in .Va "killed" . .Bd -literal nvlist pf_state_cmp { number id; number creatorid; number direction; }; nvlist pf_kill { nvlist pf_state_cmp cmp; number af; number proto; nvlist pf_rule_addr src; nvlist pf_rule_addr dst; string ifname[IFNAMSIZ]; string label[PF_RULE_LABEL_SIZE]; }; .Ed .It Dv DIOCCLRSTATESNV Fa "struct pfioc_nv nv" Clear all states. It works like .Dv DIOCKILLSTATESNV , but ignores the .Va af , .Va proto , .Va src , and .Va dst fields of the .Vt pf_kill nvlist. .It Dv DIOCSETSTATUSIF Fa "struct pfioc_if *pi" Specify the interface for which statistics are accumulated. .Bd -literal struct pfioc_if { char ifname[IFNAMSIZ]; }; .Ed .It Dv DIOCGETSTATUS Fa "struct pf_status *s" Get the internal packet filter statistics. .Bd -literal struct pf_status { u_int64_t counters[PFRES_MAX]; u_int64_t lcounters[LCNT_MAX]; u_int64_t fcounters[FCNT_MAX]; u_int64_t scounters[SCNT_MAX]; u_int64_t pcounters[2][2][3]; u_int64_t bcounters[2][2]; u_int32_t running; u_int32_t states; u_int32_t src_nodes; u_int32_t since; u_int32_t debug; u_int32_t hostid; char ifname[IFNAMSIZ]; u_int8_t pf_chksum[MD5_DIGEST_LENGTH]; }; .Ed .It Dv DIOCCLRSTATUS Clear the internal packet filter statistics. .It Dv DIOCNATLOOK Fa "struct pfioc_natlook *pnl" Look up a state table entry by source and destination addresses and ports. .Bd -literal struct pfioc_natlook { struct pf_addr saddr; struct pf_addr daddr; struct pf_addr rsaddr; struct pf_addr rdaddr; u_int16_t sport; u_int16_t dport; u_int16_t rsport; u_int16_t rdport; sa_family_t af; u_int8_t proto; u_int8_t direction; }; .Ed .It Dv DIOCSETDEBUG Fa "u_int32_t *level" Set the debug level. .Bd -literal enum { PF_DEBUG_NONE, PF_DEBUG_URGENT, PF_DEBUG_MISC, PF_DEBUG_NOISY }; .Ed .It Dv DIOCGETSTATESV2 Fa "struct pfioc_states_v2 *ps" Get state table entries. .Bd -literal struct pfioc_states_v2 { int ps_len; uint64_t ps_req_version; union { void *ps_buf; struct pf_state_export *ps_states; }; }; struct pf_state_export { uint64_t version; uint64_t id; char ifname[IFNAMSIZ]; char orig_ifname[IFNAMSIZ]; struct pf_state_key_export key[2]; struct pf_state_peer_export src; struct pf_state_peer_export dst; struct pf_addr rt_addr; uint32_t rule; uint32_t anchor; uint32_t nat_rule; uint32_t creation; uint32_t expire; uint32_t spare0; uint64_t packets[2]; uint64_t bytes[2]; uint32_t creatorid; uint32_t spare1; sa_family_t af; uint8_t proto; uint8_t direction; uint8_t log; uint8_t state_flags_compat; uint8_t timeout; uint8_t sync_flags; uint8_t updates; uint16_t state_flags; uint16_t qid; uint16_t pqid; uint16_t dnpipe; uint16_t dnrpipe; int32_t rtableid; uint8_t min_ttl; uint8_t set_tos; uint16_t max_mss; uint8_t set_prio[2]; uint8_t rt; char rt_ifname[IFNAMSIZ]; uint8_t spare[72]; }; .Ed .It Dv DIOCCHANGERULE Fa "struct pfioc_rule *pcr" Add or remove the .Va rule in the ruleset specified by .Va rule.action . .Pp The type of operation to be performed is indicated by .Va action , which can be any of the following: .Bd -literal enum { PF_CHANGE_NONE, PF_CHANGE_ADD_HEAD, PF_CHANGE_ADD_TAIL, PF_CHANGE_ADD_BEFORE, PF_CHANGE_ADD_AFTER, PF_CHANGE_REMOVE, PF_CHANGE_GET_TICKET }; .Ed .Pp .Va ticket must be set to the value obtained with .Dv PF_CHANGE_GET_TICKET for all actions except .Dv PF_CHANGE_GET_TICKET . .Va pool_ticket must be set to the value obtained with the .Dv DIOCBEGINADDRS call for all actions except .Dv PF_CHANGE_REMOVE and .Dv PF_CHANGE_GET_TICKET . .Va anchor indicates to which anchor the operation applies. .Va nr indicates the rule number against which .Dv PF_CHANGE_ADD_BEFORE , .Dv PF_CHANGE_ADD_AFTER , or .Dv PF_CHANGE_REMOVE actions are applied. .\" It Dv DIOCCHANGEALTQ Fa "struct pfioc_altq *pcr" .It Dv DIOCCHANGEADDR Fa "struct pfioc_pooladdr *pca" Add or remove the pool address .Va addr from the rule specified by .Va r_action , .Va r_num , and .Va anchor . .It Dv DIOCSETTIMEOUT Fa "struct pfioc_tm *pt" .Bd -literal struct pfioc_tm { int timeout; int seconds; }; .Ed .Pp Set the state timeout of .Va timeout to .Va seconds . The old value will be placed into .Va seconds . For possible values of .Va timeout , consult the .Dv PFTM_* values in .Aq Pa net/pfvar.h . .It Dv DIOCGETTIMEOUT Fa "struct pfioc_tm *pt" Get the state timeout of .Va timeout . The value will be placed into the .Va seconds field. .It Dv DIOCCLRRULECTRS Clear per-rule statistics. .It Dv DIOCSETLIMIT Fa "struct pfioc_limit *pl" Set the hard limits on the memory pools used by the packet filter. .Bd -literal struct pfioc_limit { int index; unsigned limit; }; enum { PF_LIMIT_STATES, PF_LIMIT_SRC_NODES, PF_LIMIT_FRAGS, PF_LIMIT_TABLE_ENTRIES, PF_LIMIT_MAX }; .Ed .It Dv DIOCGETLIMIT Fa "struct pfioc_limit *pl" Get the hard .Va limit for the memory pool indicated by .Va index . .It Dv DIOCRCLRTABLES Fa "struct pfioc_table *io" Clear all tables. All the ioctls that manipulate radix tables use the same structure described below. For .Dv DIOCRCLRTABLES , .Va pfrio_ndel contains on exit the number of tables deleted. .Bd -literal struct pfioc_table { struct pfr_table pfrio_table; void *pfrio_buffer; int pfrio_esize; int pfrio_size; int pfrio_size2; int pfrio_nadd; int pfrio_ndel; int pfrio_nchange; int pfrio_flags; u_int32_t pfrio_ticket; }; #define pfrio_exists pfrio_nadd #define pfrio_nzero pfrio_nadd #define pfrio_nmatch pfrio_nadd #define pfrio_naddr pfrio_size2 #define pfrio_setflag pfrio_size2 #define pfrio_clrflag pfrio_nadd .Ed .It Dv DIOCRADDTABLES Fa "struct pfioc_table *io" Create one or more tables. On entry, .Va pfrio_buffer must point to an array of .Vt struct pfr_table containing at least .Vt pfrio_size elements. .Vt pfrio_esize must be the size of .Vt struct pfr_table . On exit, .Va pfrio_nadd contains the number of tables effectively created. .Bd -literal struct pfr_table { char pfrt_anchor[MAXPATHLEN]; char pfrt_name[PF_TABLE_NAME_SIZE]; u_int32_t pfrt_flags; u_int8_t pfrt_fback; }; .Ed .It Dv DIOCRDELTABLES Fa "struct pfioc_table *io" Delete one or more tables. On entry, .Va pfrio_buffer must point to an array of .Vt struct pfr_table containing at least .Vt pfrio_size elements. .Vt pfrio_esize must be the size of .Vt struct pfr_table . On exit, .Va pfrio_ndel contains the number of tables effectively deleted. .It Dv DIOCRGETTABLES Fa "struct pfioc_table *io" Get the list of all tables. On entry, .Va pfrio_buffer[pfrio_size] contains a valid writeable buffer for .Vt pfr_table structures. On exit, .Va pfrio_size contains the number of tables written into the buffer. If the buffer is too small, the kernel does not store anything but just returns the required buffer size, without error. .It Dv DIOCRGETTSTATS Fa "struct pfioc_table *io" This call is like .Dv DIOCRGETTABLES but is used to get an array of .Vt pfr_tstats structures. .Bd -literal struct pfr_tstats { struct pfr_table pfrts_t; u_int64_t pfrts_packets [PFR_DIR_MAX][PFR_OP_TABLE_MAX]; u_int64_t pfrts_bytes [PFR_DIR_MAX][PFR_OP_TABLE_MAX]; u_int64_t pfrts_match; u_int64_t pfrts_nomatch; long pfrts_tzero; int pfrts_cnt; int pfrts_refcnt[PFR_REFCNT_MAX]; }; #define pfrts_name pfrts_t.pfrt_name #define pfrts_flags pfrts_t.pfrt_flags .Ed .It Dv DIOCRCLRTSTATS Fa "struct pfioc_table *io" Clear the statistics of one or more tables. On entry, .Va pfrio_buffer must point to an array of .Vt struct pfr_table containing at least .Vt pfrio_size elements. .Vt pfrio_esize must be the size of .Vt struct pfr_table . On exit, .Va pfrio_nzero contains the number of tables effectively cleared. .It Dv DIOCRCLRADDRS Fa "struct pfioc_table *io" Clear all addresses in a table. On entry, .Va pfrio_table contains the table to clear. On exit, .Va pfrio_ndel contains the number of addresses removed. .It Dv DIOCRADDADDRS Fa "struct pfioc_table *io" Add one or more addresses to a table. On entry, .Va pfrio_table contains the table ID and .Va pfrio_buffer must point to an array of .Vt struct pfr_addr containing at least .Vt pfrio_size elements to add to the table. .Vt pfrio_esize must be the size of .Vt struct pfr_addr . On exit, .Va pfrio_nadd contains the number of addresses effectively added. .Bd -literal struct pfr_addr { union { struct in_addr _pfra_ip4addr; struct in6_addr _pfra_ip6addr; } pfra_u; u_int8_t pfra_af; u_int8_t pfra_net; u_int8_t pfra_not; u_int8_t pfra_fback; }; #define pfra_ip4addr pfra_u._pfra_ip4addr #define pfra_ip6addr pfra_u._pfra_ip6addr .Ed .It Dv DIOCRDELADDRS Fa "struct pfioc_table *io" Delete one or more addresses from a table. On entry, .Va pfrio_table contains the table ID and .Va pfrio_buffer must point to an array of .Vt struct pfr_addr containing at least .Vt pfrio_size elements to delete from the table. .Vt pfrio_esize must be the size of .Vt struct pfr_addr . On exit, .Va pfrio_ndel contains the number of addresses effectively deleted. .It Dv DIOCRSETADDRS Fa "struct pfioc_table *io" Replace the content of a table by a new address list. This is the most complicated command, which uses all the structure members. .Pp On entry, .Va pfrio_table contains the table ID and .Va pfrio_buffer must point to an array of .Vt struct pfr_addr containing at least .Vt pfrio_size elements which become the new contents of the table. .Vt pfrio_esize must be the size of .Vt struct pfr_addr . Additionally, if .Va pfrio_size2 is non-zero, .Va pfrio_buffer[pfrio_size..pfrio_size2] must be a writeable buffer, into which the kernel can copy the addresses that have been deleted during the replace operation. On exit, .Va pfrio_ndel , .Va pfrio_nadd , and .Va pfrio_nchange contain the number of addresses deleted, added, and changed by the kernel. If .Va pfrio_size2 was set on entry, .Va pfrio_size2 will point to the size of the buffer used, exactly like .Dv DIOCRGETADDRS . .It Dv DIOCRGETADDRS Fa "struct pfioc_table *io" Get all the addresses of a table. On entry, .Va pfrio_table contains the table ID and .Va pfrio_buffer[pfrio_size] contains a valid writeable buffer for .Vt pfr_addr structures. On exit, .Va pfrio_size contains the number of addresses written into the buffer. If the buffer was too small, the kernel does not store anything but just returns the required buffer size, without returning an error. .It Dv DIOCRGETASTATS Fa "struct pfioc_table *io" This call is like .Dv DIOCRGETADDRS but is used to get an array of .Vt pfr_astats structures. .Bd -literal struct pfr_astats { struct pfr_addr pfras_a; u_int64_t pfras_packets [PFR_DIR_MAX][PFR_OP_ADDR_MAX]; u_int64_t pfras_bytes [PFR_DIR_MAX][PFR_OP_ADDR_MAX]; long pfras_tzero; }; .Ed .It Dv DIOCRCLRASTATS Fa "struct pfioc_table *io" Clear the statistics of one or more addresses. On entry, .Va pfrio_table contains the table ID and .Va pfrio_buffer must point to an array of .Vt struct pfr_addr containing at least .Vt pfrio_size elements to be cleared from the table. .Vt pfrio_esize must be the size of .Vt struct pfr_addr . On exit, .Va pfrio_nzero contains the number of addresses effectively cleared. .It Dv DIOCRTSTADDRS Fa "struct pfioc_table *io" Test if the given addresses match a table. On entry, .Va pfrio_table contains the table ID and .Va pfrio_buffer must point to an array of .Vt struct pfr_addr containing at least .Vt pfrio_size elements, each of which will be tested for a match in the table. .Vt pfrio_esize must be the size of .Vt struct pfr_addr . On exit, the kernel updates the .Vt pfr_addr array by setting the .Va pfra_fback member appropriately. .It Dv DIOCRSETTFLAGS Fa "struct pfioc_table *io" Change the .Dv PFR_TFLAG_CONST or .Dv PFR_TFLAG_PERSIST flags of a table. On entry, .Va pfrio_buffer must point to an array of .Vt struct pfr_table containing at least .Vt pfrio_size elements. .Va pfrio_esize must be the size of .Vt struct pfr_table . .Va pfrio_setflag must contain the flags to add, while .Va pfrio_clrflag must contain the flags to remove. On exit, .Va pfrio_nchange and .Va pfrio_ndel contain the number of tables altered or deleted by the kernel. Yes, tables can be deleted if one removes the .Dv PFR_TFLAG_PERSIST flag of an unreferenced table. .It Dv DIOCRINADEFINE Fa "struct pfioc_table *io" Defines a table in the inactive set. On entry, .Va pfrio_table contains the table ID and .Va pfrio_buffer[pfrio_size] contains an array of .Vt pfr_addr structures to put in the table. A valid ticket must also be supplied to .Va pfrio_ticket . On exit, .Va pfrio_nadd contains 0 if the table was already defined in the inactive list or 1 if a new table has been created. .Va pfrio_naddr contains the number of addresses effectively put in the table. .It Dv DIOCXBEGIN Fa "struct pfioc_trans *io" .Bd -literal struct pfioc_trans { int size; /* number of elements */ int esize; /* size of each element in bytes */ struct pfioc_trans_e { int rs_num; char anchor[MAXPATHLEN]; u_int32_t ticket; } *array; }; .Ed .Pp Clear all the inactive rulesets specified in the .Vt pfioc_trans_e array. For each ruleset, a ticket is returned for subsequent "add rule" ioctls, as well as for the .Dv DIOCXCOMMIT and .Dv DIOCXROLLBACK calls. .Pp Ruleset types, identified by .Va rs_num , include the following: .Pp .Bl -tag -width PF_RULESET_FILTER -offset ind -compact .It Dv PF_RULESET_SCRUB Scrub (packet normalization) rules. .It Dv PF_RULESET_FILTER Filter rules. .It Dv PF_RULESET_NAT NAT (Network Address Translation) rules. .It Dv PF_RULESET_BINAT Bidirectional NAT rules. .It Dv PF_RULESET_RDR Redirect rules. .It Dv PF_RULESET_ALTQ ALTQ disciplines. .It Dv PF_RULESET_TABLE Address tables. .El .It Dv DIOCXCOMMIT Fa "struct pfioc_trans *io" Atomically switch a vector of inactive rulesets to the active rulesets. This call is implemented as a standard two-phase commit, which will either fail for all rulesets or completely succeed. All tickets need to be valid. This ioctl returns .Er EBUSY if another process is concurrently updating some of the same rulesets. .It Dv DIOCXROLLBACK Fa "struct pfioc_trans *io" Clean up the kernel by undoing all changes that have taken place on the inactive rulesets since the last .Dv DIOCXBEGIN . .Dv DIOCXROLLBACK will silently ignore rulesets for which the ticket is invalid. .It Dv DIOCSETHOSTID Fa "u_int32_t *hostid" Set the host ID, which is used by .Xr pfsync 4 to identify which host created state table entries. .It Dv DIOCOSFPFLUSH Flush the passive OS fingerprint table. .It Dv DIOCOSFPADD Fa "struct pf_osfp_ioctl *io" .Bd -literal struct pf_osfp_ioctl { struct pf_osfp_entry { SLIST_ENTRY(pf_osfp_entry) fp_entry; pf_osfp_t fp_os; char fp_class_nm[PF_OSFP_LEN]; char fp_version_nm[PF_OSFP_LEN]; char fp_subtype_nm[PF_OSFP_LEN]; } fp_os; pf_tcpopts_t fp_tcpopts; u_int16_t fp_wsize; u_int16_t fp_psize; u_int16_t fp_mss; u_int16_t fp_flags; u_int8_t fp_optcnt; u_int8_t fp_wscale; u_int8_t fp_ttl; int fp_getnum; }; .Ed .Pp Add a passive OS fingerprint to the table. Set .Va fp_os.fp_os to the packed fingerprint, .Va fp_os.fp_class_nm to the name of the class (Linux, Windows, etc), .Va fp_os.fp_version_nm to the name of the version (NT, 95, 98), and .Va fp_os.fp_subtype_nm to the name of the subtype or patchlevel. The members .Va fp_mss , .Va fp_wsize , .Va fp_psize , .Va fp_ttl , .Va fp_optcnt , and .Va fp_wscale are set to the TCP MSS, the TCP window size, the IP length, the IP TTL, the number of TCP options, and the TCP window scaling constant of the TCP SYN packet, respectively. .Pp The .Va fp_flags member is filled according to the .Aq Pa net/pfvar.h include file .Dv PF_OSFP_* defines. The .Va fp_tcpopts member contains packed TCP options. Each option uses .Dv PF_OSFP_TCPOPT_BITS bits in the packed value. Options include any of .Dv PF_OSFP_TCPOPT_NOP , .Dv PF_OSFP_TCPOPT_SACK , .Dv PF_OSFP_TCPOPT_WSCALE , .Dv PF_OSFP_TCPOPT_MSS , or .Dv PF_OSFP_TCPOPT_TS . .Pp The .Va fp_getnum member is not used with this ioctl. .Pp The structure's slack space must be zeroed for correct operation; .Xr memset 3 the whole structure to zero before filling and sending to the kernel. .It Dv DIOCOSFPGET Fa "struct pf_osfp_ioctl *io" Get the passive OS fingerprint number .Va fp_getnum from the kernel's fingerprint list. The rest of the structure members will come back filled. Get the whole list by repeatedly incrementing the .Va fp_getnum number until the ioctl returns .Er EBUSY . .It Dv DIOCGETSRCNODES Fa "struct pfioc_src_nodes *psn" .Bd -literal struct pfioc_src_nodes { int psn_len; union { caddr_t psu_buf; struct pf_src_node *psu_src_nodes; } psn_u; #define psn_buf psn_u.psu_buf #define psn_src_nodes psn_u.psu_src_nodes }; .Ed .Pp Get the list of source nodes kept by sticky addresses and source tracking. The ioctl must be called once with .Va psn_len set to 0. If the ioctl returns without error, .Va psn_len will be set to the size of the buffer required to hold all the .Va pf_src_node structures held in the table. A buffer of this size should then be allocated, and a pointer to this buffer placed in .Va psn_buf . The ioctl must then be called again to fill this buffer with the actual source node data. After that call, .Va psn_len will be set to the length of the buffer actually used. .It Dv DIOCCLRSRCNODES Clear the tree of source tracking nodes. .It Dv DIOCIGETIFACES Fa "struct pfioc_iface *io" Get the list of interfaces and interface drivers known to .Nm . All the ioctls that manipulate interfaces use the same structure described below: .Bd -literal struct pfioc_iface { char pfiio_name[IFNAMSIZ]; void *pfiio_buffer; int pfiio_esize; int pfiio_size; int pfiio_nzero; int pfiio_flags; }; .Ed .Pp If not empty, .Va pfiio_name can be used to restrict the search to a specific interface or driver. .Va pfiio_buffer[pfiio_size] is the user-supplied buffer for returning the data. On entry, .Va pfiio_size contains the number of .Vt pfi_kif entries that can fit into the buffer. The kernel will replace this value by the real number of entries it wants to return. .Va pfiio_esize should be set to .Li sizeof(struct pfi_kif) . .Pp The data is returned in the .Vt pfi_kif structure described below: .Bd -literal struct pfi_kif { char pfik_name[IFNAMSIZ]; union { RB_ENTRY(pfi_kif) pfik_tree; LIST_ENTRY(pfi_kif) pfik_list; }; u_int64_t pfik_packets[2][2][2]; u_int64_t pfik_bytes[2][2][2]; u_int32_t pfik_tzero; u_int pfik_flags; struct ifnet *pfik_ifp; struct ifg_group *pfik_group; u_int pfik_rulerefs; TAILQ_HEAD(, pfi_dynaddr) pfik_dynaddrs; }; .Ed .It Dv DIOCSETIFFLAG Fa "struct pfioc_iface *io" Set the user settable flags (described above) of the .Nm internal interface description. The filtering process is the same as for .Dv DIOCIGETIFACES . .Bd -literal #define PFI_IFLAG_SKIP 0x0100 /* skip filtering on interface */ .Ed .It Dv DIOCCLRIFFLAG Fa "struct pfioc_iface *io" Works as .Dv DIOCSETIFFLAG above but clears the flags. .It Dv DIOCKILLSRCNODES Fa "struct pfioc_iface *io" Explicitly remove source tracking nodes. .El .Sh FILES .Bl -tag -width /dev/pf -compact .It Pa /dev/pf packet filtering device. .El .Sh EXAMPLES The following example demonstrates how to use the .Dv DIOCNATLOOK command to find the internal host/port of a NATed connection: .Bd -literal #include #include #include #include #include #include #include #include #include #include u_int32_t read_address(const char *s) { int a, b, c, d; sscanf(s, "%i.%i.%i.%i", &a, &b, &c, &d); return htonl(a << 24 | b << 16 | c << 8 | d); } void print_address(u_int32_t a) { a = ntohl(a); printf("%d.%d.%d.%d", a >> 24 & 255, a >> 16 & 255, a >> 8 & 255, a & 255); } int main(int argc, char *argv[]) { struct pfioc_natlook nl; int dev; if (argc != 5) { printf("%s \\n", argv[0]); return 1; } dev = open("/dev/pf", O_RDWR); if (dev == -1) err(1, "open(\\"/dev/pf\\") failed"); memset(&nl, 0, sizeof(struct pfioc_natlook)); nl.saddr.v4.s_addr = read_address(argv[1]); nl.sport = htons(atoi(argv[2])); nl.daddr.v4.s_addr = read_address(argv[3]); nl.dport = htons(atoi(argv[4])); nl.af = AF_INET; nl.proto = IPPROTO_TCP; nl.direction = PF_IN; if (ioctl(dev, DIOCNATLOOK, &nl)) err(1, "DIOCNATLOOK"); printf("internal host "); print_address(nl.rsaddr.v4.s_addr); printf(":%u\\n", ntohs(nl.rsport)); return 0; } .Ed .Sh SEE ALSO .Xr ioctl 2 , .Xr altq 4 , .Xr if_bridge 4 , .Xr pflog 4 , .Xr pfsync 4 , .Xr pfctl 8 , .Xr altq 9 .Sh HISTORY The .Nm packet filtering mechanism first appeared in .Ox 3.0 and then .Fx 5.2 . .Pp This implementation is derived from .Ox 4.5 . A number of individual features, improvements, bug fixes and security fixes have been ported from later versions of .Ox . It has been heavily modified to be capable of running in multithreaded .Fx kernel and scale its performance on multiple CPUs.