1 .\" Copyright (c) 2007 Seccuris Inc.
2 .\" All rights reserved.
4 .\" This software was developed by Robert N. M. Watson under contract to
7 .\" Redistribution and use in source and binary forms, with or without
8 .\" modification, are permitted provided that the following conditions
10 .\" 1. Redistributions of source code must retain the above copyright
11 .\" notice, this list of conditions and the following disclaimer.
12 .\" 2. Redistributions in binary form must reproduce the above copyright
13 .\" notice, this list of conditions and the following disclaimer in the
14 .\" documentation and/or other materials provided with the distribution.
16 .\" THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
17 .\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
18 .\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
19 .\" ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
20 .\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
21 .\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
22 .\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
23 .\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
24 .\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
25 .\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
28 .\" Copyright (c) 1990 The Regents of the University of California.
29 .\" All rights reserved.
31 .\" Redistribution and use in source and binary forms, with or without
32 .\" modification, are permitted provided that: (1) source code distributions
33 .\" retain the above copyright notice and this paragraph in its entirety, (2)
34 .\" distributions including binary code include the above copyright notice and
35 .\" this paragraph in its entirety in the documentation or other materials
36 .\" provided with the distribution, and (3) all advertising materials mentioning
37 .\" features or use of this software display the following acknowledgement:
38 .\" ``This product includes software developed by the University of California,
39 .\" Lawrence Berkeley Laboratory and its contributors.'' Neither the name of
40 .\" the University nor the names of its contributors may be used to endorse
41 .\" or promote products derived from this software without specific prior
42 .\" written permission.
43 .\" THIS SOFTWARE IS PROVIDED ``AS IS'' AND WITHOUT ANY EXPRESS OR IMPLIED
44 .\" WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED WARRANTIES OF
45 .\" MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.
47 .\" This document is derived in part from the enet man page (enet.4)
48 .\" distributed with 4.3BSD Unix.
57 .Nd Berkeley Packet Filter
61 The Berkeley Packet Filter
62 provides a raw interface to data link layers in a protocol
64 All packets on the network, even those destined for other hosts,
65 are accessible through this mechanism.
67 The packet filter appears as a character special device,
69 After opening the device, the file descriptor must be bound to a
70 specific network interface with the
73 A given interface can be shared by multiple listeners, and the filter
74 underlying each descriptor will see an identical packet stream.
76 A separate device file is required for each minor device.
77 If a file is in use, the open will fail and
82 Associated with each open instance of a
84 file is a user-settable packet filter.
85 Whenever a packet is received by an interface,
86 all file descriptors listening on that interface apply their filter.
87 Each descriptor that accepts the packet receives its own copy.
89 The packet filter will support any link level protocol that has fixed length
91 Currently, only Ethernet,
95 drivers have been modified to interact with
98 Since packet data is in network byte order, applications should use the
100 macros to extract multi-byte values.
102 A packet can be sent out on the network by writing to a
105 The writes are unbuffered, meaning only one packet can be processed per write.
106 Currently, only writes to Ethernets and
111 devices deliver packet data to the application via memory buffers provided by
113 The buffer mode is set using the
115 ioctl, and read using the
118 .Ss Buffered read mode
121 devices operate in the
122 .Dv BPF_BUFMODE_BUFFER
123 mode, in which packet data is copied explicitly from kernel to user memory
127 The user process will declare a fixed buffer size that will be used both for
128 sizing internal buffers and for all
130 operations on the file.
131 This size is queried using the
133 ioctl, and is set using the
136 Note that an individual packet larger than the buffer size is necessarily
138 .Ss Zero-copy buffer mode
140 devices may also operate in the
141 .Dv BPF_BUFMODE_ZEROCOPY
142 mode, in which packet data is written directly into two user memory buffers
143 by the kernel, avoiding both system call and copying overhead.
144 Buffers are of fixed (and equal) size, page-aligned, and an even multiple of
146 The maximum zero-copy buffer size is returned by the
149 Note that an individual packet larger than the buffer size is necessarily
152 The user process registers two memory buffers using the
154 ioctl, which accepts a
156 pointer as an argument:
166 is a pointer to the userspace address of the first buffer that will be
169 is a pointer to the second buffer.
171 will then cycle between the two buffers as they fill and are acknowledged.
173 Each buffer begins with a fixed-length header to hold synchronization and
174 data length information for the buffer:
176 struct bpf_zbuf_header {
177 volatile u_int bzh_kernel_gen; /* Kernel generation number. */
178 volatile u_int bzh_kernel_len; /* Length of data in the buffer. */
179 volatile u_int bzh_user_gen; /* User generation number. */
180 /* ...padding for future use... */
184 The header structure of each buffer, including all padding, should be zeroed
185 before it is configured using
187 Remaining space in the buffer will be used by the kernel to store packet
188 data, laid out in the same format as with buffered read mode.
190 The kernel and the user process follow a simple acknowledgement protocol via
191 the buffer header to synchronize access to the buffer: when the header
196 hold the same value, the kernel owns the buffer, and when they differ,
197 userspace owns the buffer.
199 While the kernel owns the buffer, the contents are unstable and may change
200 asynchronously; while the user process owns the buffer, its contents are
201 stable and will not be changed until the buffer has been acknowledged.
203 Initializing the buffer headers to all 0's before registering the buffer has
204 the effect of assigning initial ownership of both buffers to the kernel.
205 The kernel signals that a buffer has been assigned to userspace by modifying
207 and userspace acknowledges the buffer and returns it to the kernel by setting
213 In order to avoid caching and memory re-ordering effects, the user process
214 must use atomic operations and memory barriers when checking for and
215 acknowledging buffers:
217 #include <machine/atomic.h>
220 * Return ownership of a buffer to the kernel for reuse.
223 buffer_acknowledge(struct bpf_zbuf_header *bzh)
226 atomic_store_rel_int(&bzh->bzh_user_gen, bzh->bzh_kernel_gen);
230 * Check whether a buffer has been assigned to userspace by the kernel.
231 * Return true if userspace owns the buffer, and false otherwise.
234 buffer_check(struct bpf_zbuf_header *bzh)
237 return (bzh->bzh_user_gen !=
238 atomic_load_acq_int(&bzh->bzh_kernel_gen));
242 The user process may force the assignment of the next buffer, if any data
243 is pending, to userspace using the
246 This allows the user process to retrieve data in a partially filled buffer
247 before the buffer is full, such as following a timeout; the process must
248 recheck for buffer ownership using the header generation numbers, as the
249 buffer will not be assigned to userspace if no data was present.
251 As in the buffered read mode,
256 may be used to sleep awaiting the availability of a completed buffer.
257 They will return a readable file descriptor when ownership of the next buffer
258 is assigned to user space.
260 In the current implementation, the kernel may assign zero, one, or both
261 buffers to the user process; however, an earlier implementation maintained
262 the invariant that at most one buffer could be assigned to the user process
264 In order to both ensure progress and high performance, user processes should
265 acknowledge a completely processed buffer as quickly as possible, returning
266 it for reuse, and not block waiting on a second buffer while holding another
271 command codes below are defined in
276 #include <sys/types.h>
277 #include <sys/time.h>
278 #include <sys/ioctl.h>
293 the following commands may be applied to any open
296 The (third) argument to
298 should be a pointer to the type indicated.
299 .Bl -tag -width BIOCGETBUFMODE
302 Returns the required buffer length for reads on
307 Sets the buffer length for reads on
310 The buffer must be set before the file is attached to an interface
313 If the requested buffer size cannot be accommodated, the closest
314 allowable size will be set and returned in the argument.
315 A read call will result in
317 if it is passed a buffer that is not this size.
320 Returns the type of the data link layer underlying the attached interface.
322 is returned if no interface has been specified.
323 The device types, prefixed with
328 Forces the interface into promiscuous mode.
329 All packets, not just those destined for the local host, are processed.
330 Since more than one file can be listening on a given interface,
331 a listener that opened its interface non-promiscuously may receive
332 packets promiscuously.
333 This problem can be remedied with an appropriate filter.
335 Flushes the buffer of incoming packets,
336 and resets the statistics that are returned by BIOCGSTATS.
338 .Pq Li "struct ifreq"
339 Returns the name of the hardware interface that the file is listening on.
340 The name is returned in the ifr_name field of
344 All other fields are undefined.
346 .Pq Li "struct ifreq"
347 Sets the hardware interface associate with the file.
349 command must be performed before any packets can be read.
350 The device is indicated by name using the
355 Additionally, performs the actions of
359 .Pq Li "struct timeval"
360 Set or get the read timeout parameter.
362 specifies the length of time to wait before timing
363 out on a read request.
364 This parameter is initialized to zero by
366 indicating no timeout.
368 .Pq Li "struct bpf_stat"
369 Returns the following structure of packet statistics:
372 u_int bs_recv; /* number of packets received */
373 u_int bs_drop; /* number of packets dropped */
378 .Bl -hang -offset indent
380 the number of packets received by the descriptor since opened or reset
381 (including any buffered since the last read call);
384 the number of packets which were accepted by the filter but dropped by the
385 kernel because of buffer overflows
386 (i.e., the application's reads are not keeping up with the packet traffic).
392 based on the truth value of the argument.
393 When immediate mode is enabled, reads return immediately upon packet
395 Otherwise, a read will block until either the kernel buffer
396 becomes full or a timeout occurs.
397 This is useful for programs like
399 which must respond to messages in real time.
400 The default for a new file is off.
403 .Pq Li "struct bpf_program"
404 Sets the read filter program used by the kernel to discard uninteresting
406 An array of instructions and its length is passed in using
407 the following structure:
411 struct bpf_insn *bf_insns;
415 The filter program is pointed to by the
417 field while its length in units of
418 .Sq Li struct bpf_insn
424 for an explanation of the filter language.
425 The only difference between
431 performs the actions of
437 .Pq Li "struct bpf_program"
438 Sets the write filter program used by the kernel to control what type of
439 packets can be written to the interface.
447 .Pq Li "struct bpf_version"
448 Returns the major and minor version numbers of the filter language currently
449 recognized by the kernel.
450 Before installing a filter, applications must check
451 that the current version is compatible with the running kernel.
452 Version numbers are compatible if the major numbers match and the application minor
453 is less than or equal to the kernel minor.
454 The kernel version number is returned in the following structure:
462 The current version numbers are given by
463 .Dv BPF_MAJOR_VERSION
465 .Dv BPF_MINOR_VERSION
468 An incompatible filter
469 may result in undefined behavior (most likely, an error returned by
471 or haphazard packet matching).
475 Set or get the status of the
478 Set to zero if the link level source address should be filled in automatically
479 by the interface output routine.
480 Set to one if the link level source
481 address will be written, as provided, to the wire.
482 This flag is initialized to zero by default.
486 These commands are obsolete but left for compatibility.
492 Set or get the flag determining whether locally generated packets on the
493 interface should be returned by BPF.
494 Set to zero to see only incoming packets on the interface.
495 Set to one to see packets originating locally and remotely on the interface.
496 This flag is initialized to one by default.
497 .It Dv BIOCSDIRECTION
498 .It Dv BIOCGDIRECTION
500 Set or get the setting determining whether incoming, outgoing, or all packets
501 on the interface should be returned by BPF.
504 to see only incoming packets on the interface.
507 to see packets originating locally and remotely on the interface.
510 to see only outgoing packets on the interface.
511 This setting is initialized to
517 Set or get format and resolution of the time stamps returned by BPF.
519 .Dv BPF_T_MICROTIME ,
520 .Dv BPF_T_MICROTIME_FAST ,
521 .Dv BPF_T_MICROTIME_MONOTONIC ,
523 .Dv BPF_T_MICROTIME_MONOTONIC_FAST
524 to get time stamps in 64-bit
529 .Dv BPF_T_NANOTIME_FAST ,
530 .Dv BPF_T_NANOTIME_MONOTONIC ,
532 .Dv BPF_T_NANOTIME_MONOTONIC_FAST
533 to get time stamps in 64-bit
538 .Dv BPF_T_BINTIME_FAST ,
539 .Dv BPF_T_NANOTIME_MONOTONIC ,
541 .Dv BPF_T_BINTIME_MONOTONIC_FAST
542 to get time stamps in 64-bit
547 to ignore time stamp.
548 All 64-bit time stamp formats are wrapped in
551 .Dv BPF_T_MICROTIME_FAST ,
552 .Dv BPF_T_NANOTIME_FAST ,
553 .Dv BPF_T_BINTIME_FAST ,
554 .Dv BPF_T_MICROTIME_MONOTONIC_FAST ,
555 .Dv BPF_T_NANOTIME_MONOTONIC_FAST ,
557 .Dv BPF_T_BINTIME_MONOTONIC_FAST
558 are analogs of corresponding formats without _FAST suffix but do not perform
559 a full time counter query, so their accuracy is one timer tick.
561 .Dv BPF_T_MICROTIME_MONOTONIC ,
562 .Dv BPF_T_NANOTIME_MONOTONIC ,
563 .Dv BPF_T_BINTIME_MONOTONIC ,
564 .Dv BPF_T_MICROTIME_MONOTONIC_FAST ,
565 .Dv BPF_T_NANOTIME_MONOTONIC_FAST ,
567 .Dv BPF_T_BINTIME_MONOTONIC_FAST
568 store the time elapsed since kernel boot.
569 This setting is initialized to
574 Set packet feedback mode.
575 This allows injected packets to be fed back as input to the interface when
576 output via the interface is successful.
579 direction is set, injected outgoing packet is not returned by BPF to avoid
580 duplication. This flag is initialized to zero by default.
582 Set the locked flag on the
585 This prevents the execution of
586 ioctl commands which could change the underlying operating parameters of
588 .It Dv BIOCGETBUFMODE
589 .It Dv BIOCSETBUFMODE
591 Get or set the current
593 buffering mode; possible values are
594 .Dv BPF_BUFMODE_BUFFER ,
595 buffered read mode, and
596 .Dv BPF_BUFMODE_ZBUF ,
597 zero-copy buffer mode.
599 .Pq Li struct bpf_zbuf
600 Set the current zero-copy buffer locations; buffer locations may be
601 set only once zero-copy buffer mode has been selected, and prior to attaching
603 Buffers must be of identical size, page-aligned, and an integer multiple of
611 If buffers have already been set for this device, the ioctl will fail.
614 Get the largest individual zero-copy buffer size allowed.
615 As two buffers are used in zero-copy buffer mode, the limit (in practice) is
616 twice the returned size.
617 As zero-copy buffers consume kernel address space, conservative selection of
618 buffer size is suggested, especially when there are multiple
620 descriptors in use on 32-bit systems.
622 Force ownership of the next buffer to be assigned to userspace, if any data
623 present in the buffer.
624 If no data is present, the buffer will remain owned by the kernel.
625 This allows consumers of zero-copy buffering to implement timeouts and
626 retrieve partially filled buffers.
627 In order to handle the case where no data is present in the buffer and
628 therefore ownership is not assigned, the user process must check
634 One of the following structures is prepended to each packet returned by
636 or via a zero-copy buffer:
639 struct bpf_ts bh_tstamp; /* time stamp */
640 uint32_t bh_caplen; /* length of captured portion */
641 uint32_t bh_datalen; /* original length of packet */
642 u_short bh_hdrlen; /* length of bpf header (this struct
643 plus alignment padding) */
647 struct timeval bh_tstamp; /* time stamp */
648 uint32_t bh_caplen; /* length of captured portion */
649 uint32_t bh_datalen; /* original length of packet */
650 u_short bh_hdrlen; /* length of bpf header (this struct
651 plus alignment padding) */
655 The fields, whose values are stored in host order, and are:
657 .Bl -tag -compact -width bh_datalen
659 The time at which the packet was processed by the packet filter.
661 The length of the captured portion of the packet.
662 This is the minimum of
663 the truncation amount specified by the filter and the length of the packet.
665 The length of the packet off the wire.
666 This value is independent of the truncation amount specified by the filter.
670 header, which may not be equal to
671 .\" XXX - not really a function call
672 .Fn sizeof "struct bpf_xhdr"
674 .Fn sizeof "struct bpf_hdr" .
679 field exists to account for
680 padding between the header and the link level protocol.
681 The purpose here is to guarantee proper alignment of the packet
682 data structures, which is required on alignment sensitive
683 architectures and improves performance on many other architectures.
684 The packet filter ensures that the
687 and the network layer
688 header will be word aligned.
691 is used when the time stamp is set to
692 .Dv BPF_T_MICROTIME ,
693 .Dv BPF_T_MICROTIME_FAST ,
694 .Dv BPF_T_MICROTIME_MONOTONIC ,
695 .Dv BPF_T_MICROTIME_MONOTONIC_FAST ,
698 for backward compatibility reasons.
704 may be deprecated in the near future.
706 must be taken when accessing the link layer protocol fields on alignment
708 (This is not a problem on an Ethernet, since
709 the type field is a short falling on an even offset,
710 and the addresses are probably accessed in a bytewise fashion).
712 Additionally, individual packets are padded so that each starts
714 This requires that an application
715 has some knowledge of how to get from packet to packet.
722 It rounds up its argument to the nearest word aligned value (where a word is
728 points to the start of a packet, this expression
729 will advance it to the next packet:
730 .Dl p = (char *)p + BPF_WORDALIGN(p->bh_hdrlen + p->bh_caplen)
732 For the alignment mechanisms to work properly, the
735 must itself be word aligned.
739 will always return an aligned buffer.
741 A filter program is an array of instructions, with all branches forwardly
742 directed, terminated by a
745 Each instruction performs some action on the pseudo-machine state,
746 which consists of an accumulator, index register, scratch memory store,
747 and implicit program counter.
749 The following structure defines the instruction format:
761 field is used in different ways by different instructions,
766 fields are used as offsets
767 by the branch instructions.
768 The opcodes are encoded in a semi-hierarchical fashion.
769 There are eight classes of instructions:
779 Various other mode and
780 operator bits are or'd into the class to give the actual instructions.
781 The classes and modes are defined in
784 Below are the semantics for each defined
787 We use the convention that A is the accumulator, X is the index register,
788 P[] packet data, and M[] scratch memory store.
789 P[i:n] gives the data at byte offset
792 interpreted as a word (n=4),
793 unsigned halfword (n=2), or unsigned byte (n=1).
794 M[i] gives the i'th word in the scratch memory store, which is only
795 addressed in word units.
796 The memory store is indexed from 0 to
803 are the corresponding fields in the
804 instruction definition.
806 refers to the length of the packet.
807 .Bl -tag -width BPF_STXx
809 These instructions copy a value into the accumulator.
810 The type of the source operand is specified by an
812 and can be a constant
814 packet data at a fixed offset
816 packet data at a variable offset
820 or a word in the scratch memory store
826 the data size must be specified as a word
832 The semantics of all the recognized
836 BPF_LD+BPF_W+BPF_ABS A <- P[k:4]
837 BPF_LD+BPF_H+BPF_ABS A <- P[k:2]
838 BPF_LD+BPF_B+BPF_ABS A <- P[k:1]
839 BPF_LD+BPF_W+BPF_IND A <- P[X+k:4]
840 BPF_LD+BPF_H+BPF_IND A <- P[X+k:2]
841 BPF_LD+BPF_B+BPF_IND A <- P[X+k:1]
842 BPF_LD+BPF_W+BPF_LEN A <- len
843 BPF_LD+BPF_IMM A <- k
844 BPF_LD+BPF_MEM A <- M[k]
847 These instructions load a value into the index register.
849 the addressing modes are more restrictive than those of the accumulator loads,
852 a hack for efficiently loading the IP header length.
854 BPF_LDX+BPF_W+BPF_IMM X <- k
855 BPF_LDX+BPF_W+BPF_MEM X <- M[k]
856 BPF_LDX+BPF_W+BPF_LEN X <- len
857 BPF_LDX+BPF_B+BPF_MSH X <- 4*(P[k:1]&0xf)
860 This instruction stores the accumulator into the scratch memory.
861 We do not need an addressing mode since there is only one possibility
867 This instruction stores the index register in the scratch memory store.
872 The alu instructions perform operations between the accumulator and
873 index register or constant, and store the result back in the accumulator.
874 For binary operations, a source mode is required
879 BPF_ALU+BPF_ADD+BPF_K A <- A + k
880 BPF_ALU+BPF_SUB+BPF_K A <- A - k
881 BPF_ALU+BPF_MUL+BPF_K A <- A * k
882 BPF_ALU+BPF_DIV+BPF_K A <- A / k
883 BPF_ALU+BPF_AND+BPF_K A <- A & k
884 BPF_ALU+BPF_OR+BPF_K A <- A | k
885 BPF_ALU+BPF_LSH+BPF_K A <- A << k
886 BPF_ALU+BPF_RSH+BPF_K A <- A >> k
887 BPF_ALU+BPF_ADD+BPF_X A <- A + X
888 BPF_ALU+BPF_SUB+BPF_X A <- A - X
889 BPF_ALU+BPF_MUL+BPF_X A <- A * X
890 BPF_ALU+BPF_DIV+BPF_X A <- A / X
891 BPF_ALU+BPF_AND+BPF_X A <- A & X
892 BPF_ALU+BPF_OR+BPF_X A <- A | X
893 BPF_ALU+BPF_LSH+BPF_X A <- A << X
894 BPF_ALU+BPF_RSH+BPF_X A <- A >> X
895 BPF_ALU+BPF_NEG A <- -A
898 The jump instructions alter flow of control.
900 compare the accumulator against a constant
902 or the index register
904 If the result is true (or non-zero),
905 the true branch is taken, otherwise the false branch is taken.
906 Jump offsets are encoded in 8 bits so the longest jump is 256 instructions.
907 However, the jump always
909 opcode uses the 32 bit
911 field as the offset, allowing arbitrarily distant destinations.
912 All conditionals use unsigned comparison conventions.
914 BPF_JMP+BPF_JA pc += k
915 BPF_JMP+BPF_JGT+BPF_K pc += (A > k) ? jt : jf
916 BPF_JMP+BPF_JGE+BPF_K pc += (A >= k) ? jt : jf
917 BPF_JMP+BPF_JEQ+BPF_K pc += (A == k) ? jt : jf
918 BPF_JMP+BPF_JSET+BPF_K pc += (A & k) ? jt : jf
919 BPF_JMP+BPF_JGT+BPF_X pc += (A > X) ? jt : jf
920 BPF_JMP+BPF_JGE+BPF_X pc += (A >= X) ? jt : jf
921 BPF_JMP+BPF_JEQ+BPF_X pc += (A == X) ? jt : jf
922 BPF_JMP+BPF_JSET+BPF_X pc += (A & X) ? jt : jf
925 The return instructions terminate the filter program and specify the amount
926 of packet to accept (i.e., they return the truncation amount).
927 A return value of zero indicates that the packet should be ignored.
928 The return value is either a constant
933 BPF_RET+BPF_A accept A bytes
934 BPF_RET+BPF_K accept k bytes
937 The miscellaneous category was created for anything that does not
938 fit into the above classes, and for any new instructions that might need to
940 Currently, these are the register transfer instructions
941 that copy the index register to the accumulator or vice versa.
943 BPF_MISC+BPF_TAX X <- A
944 BPF_MISC+BPF_TXA A <- X
950 interface provides the following macros to facilitate
952 .Fn BPF_STMT opcode operand
954 .Fn BPF_JUMP opcode operand true_offset false_offset .
958 variables controls the behaviour of the
961 .Bl -tag -width indent
962 .It Va net.bpf.optimize_writers: No 0
963 Various programs use BPF to send (but not receive) raw packets
964 (cdpd, lldpd, dhcpd, dhcp relays, etc. are good examples of such programs).
965 They do not need incoming packets to be send to them.
966 Turning this option on
967 makes new BPF users to be attached to write-only interface list until program
968 explicitly specifies read filter via
969 .Fn pcap_set_filter .
970 This removes any performance degradation for high-speed interfaces.
971 .It Va net.bpf.stats:
972 Binary interface for retrieving general statistics.
973 .It Va net.bpf.zerocopy_enable: No 0
974 Permits zero-copy to be used with net BPF readers.
976 .It Va net.bpf.maxinsns: No 512
977 Maximum number of instructions that BPF program can contain.
981 option to determine approximate number of instruction for any filter.
982 .It Va net.bpf.maxbufsize: No 524288
983 Maximum buffer size to allocate for packets buffer.
984 .It Va net.bpf.bufsize: No 4096
985 Default buffer size to allocate for packets buffer.
988 The following filter is taken from the Reverse ARP Daemon.
989 It accepts only Reverse ARP requests.
991 struct bpf_insn insns[] = {
992 BPF_STMT(BPF_LD+BPF_H+BPF_ABS, 12),
993 BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, ETHERTYPE_REVARP, 0, 3),
994 BPF_STMT(BPF_LD+BPF_H+BPF_ABS, 20),
995 BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, REVARP_REQUEST, 0, 1),
996 BPF_STMT(BPF_RET+BPF_K, sizeof(struct ether_arp) +
997 sizeof(struct ether_header)),
998 BPF_STMT(BPF_RET+BPF_K, 0),
1002 This filter accepts only IP packets between host 128.3.112.15 and
1005 struct bpf_insn insns[] = {
1006 BPF_STMT(BPF_LD+BPF_H+BPF_ABS, 12),
1007 BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, ETHERTYPE_IP, 0, 8),
1008 BPF_STMT(BPF_LD+BPF_W+BPF_ABS, 26),
1009 BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, 0x8003700f, 0, 2),
1010 BPF_STMT(BPF_LD+BPF_W+BPF_ABS, 30),
1011 BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, 0x80037023, 3, 4),
1012 BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, 0x80037023, 0, 3),
1013 BPF_STMT(BPF_LD+BPF_W+BPF_ABS, 30),
1014 BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, 0x8003700f, 0, 1),
1015 BPF_STMT(BPF_RET+BPF_K, (u_int)-1),
1016 BPF_STMT(BPF_RET+BPF_K, 0),
1020 Finally, this filter returns only TCP finger packets.
1021 We must parse the IP header to reach the TCP header.
1025 checks that the IP fragment offset is 0 so we are sure
1026 that we have a TCP header.
1028 struct bpf_insn insns[] = {
1029 BPF_STMT(BPF_LD+BPF_H+BPF_ABS, 12),
1030 BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, ETHERTYPE_IP, 0, 10),
1031 BPF_STMT(BPF_LD+BPF_B+BPF_ABS, 23),
1032 BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, IPPROTO_TCP, 0, 8),
1033 BPF_STMT(BPF_LD+BPF_H+BPF_ABS, 20),
1034 BPF_JUMP(BPF_JMP+BPF_JSET+BPF_K, 0x1fff, 6, 0),
1035 BPF_STMT(BPF_LDX+BPF_B+BPF_MSH, 14),
1036 BPF_STMT(BPF_LD+BPF_H+BPF_IND, 14),
1037 BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, 79, 2, 0),
1038 BPF_STMT(BPF_LD+BPF_H+BPF_IND, 16),
1039 BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, 79, 0, 1),
1040 BPF_STMT(BPF_RET+BPF_K, (u_int)-1),
1041 BPF_STMT(BPF_RET+BPF_K, 0),
1056 .%T "An efficient, extensible, and portable network monitor"
1059 The Enet packet filter was created in 1980 by Mike Accetta and
1060 Rick Rashid at Carnegie-Mellon University.
1062 Stanford, ported the code to
1064 and continued its development from
1066 Since then, it has evolved into the Ultrix Packet Filter at
1077 .An Steven McCanne ,
1078 of Lawrence Berkeley Laboratory, implemented BPF in
1080 Much of the design is due to
1083 Support for zero-copy buffers was added by
1084 .An Robert N. M. Watson
1085 under contract to Seccuris Inc.
1087 The read buffer must be of a fixed size (returned by the
1091 A file that does not request promiscuous mode may receive promiscuously
1092 received packets as a side effect of another file requesting this
1093 mode on the same hardware interface.
1094 This could be fixed in the kernel with additional processing overhead.
1095 However, we favor the model where
1096 all files must assume that the interface is promiscuous, and if
1097 so desired, must utilize a filter to reject foreign packets.
1099 Data link protocols with variable length headers are not currently supported.
1106 settings have been observed to work incorrectly on some interface
1107 types, including those with hardware loopback rather than software loopback,
1108 and point-to-point interfaces.
1109 They appear to function correctly on a
1110 broad range of Ethernet-style interfaces.