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>
295 the following commands may be applied to any open
298 The (third) argument to
300 should be a pointer to the type indicated.
301 .Bl -tag -width BIOCGETBUFMODE
304 Returns the required buffer length for reads on
309 Sets the buffer length for reads on
312 The buffer must be set before the file is attached to an interface
315 If the requested buffer size cannot be accommodated, the closest
316 allowable size will be set and returned in the argument.
317 A read call will result in
319 if it is passed a buffer that is not this size.
322 Returns the type of the data link layer underlying the attached interface.
324 is returned if no interface has been specified.
325 The device types, prefixed with
330 Forces the interface into promiscuous mode.
331 All packets, not just those destined for the local host, are processed.
332 Since more than one file can be listening on a given interface,
333 a listener that opened its interface non-promiscuously may receive
334 packets promiscuously.
335 This problem can be remedied with an appropriate filter.
337 Flushes the buffer of incoming packets,
338 and resets the statistics that are returned by BIOCGSTATS.
340 .Pq Li "struct ifreq"
341 Returns the name of the hardware interface that the file is listening on.
342 The name is returned in the ifr_name field of
346 All other fields are undefined.
348 .Pq Li "struct ifreq"
349 Sets the hardware interface associate with the file.
351 command must be performed before any packets can be read.
352 The device is indicated by name using the
357 Additionally, performs the actions of
361 .Pq Li "struct timeval"
362 Set or get the read timeout parameter.
364 specifies the length of time to wait before timing
365 out on a read request.
366 This parameter is initialized to zero by
368 indicating no timeout.
370 .Pq Li "struct bpf_stat"
371 Returns the following structure of packet statistics:
374 u_int bs_recv; /* number of packets received */
375 u_int bs_drop; /* number of packets dropped */
380 .Bl -hang -offset indent
382 the number of packets received by the descriptor since opened or reset
383 (including any buffered since the last read call);
386 the number of packets which were accepted by the filter but dropped by the
387 kernel because of buffer overflows
388 (i.e., the application's reads are not keeping up with the packet traffic).
394 based on the truth value of the argument.
395 When immediate mode is enabled, reads return immediately upon packet
397 Otherwise, a read will block until either the kernel buffer
398 becomes full or a timeout occurs.
399 This is useful for programs like
401 which must respond to messages in real time.
402 The default for a new file is off.
405 .Pq Li "struct bpf_program"
406 Sets the read filter program used by the kernel to discard uninteresting
408 An array of instructions and its length is passed in using
409 the following structure:
413 struct bpf_insn *bf_insns;
417 The filter program is pointed to by the
419 field while its length in units of
420 .Sq Li struct bpf_insn
426 for an explanation of the filter language.
427 The only difference between
433 performs the actions of
439 .Pq Li "struct bpf_program"
440 Sets the write filter program used by the kernel to control what type of
441 packets can be written to the interface.
449 .Pq Li "struct bpf_version"
450 Returns the major and minor version numbers of the filter language currently
451 recognized by the kernel.
452 Before installing a filter, applications must check
453 that the current version is compatible with the running kernel.
454 Version numbers are compatible if the major numbers match and the application minor
455 is less than or equal to the kernel minor.
456 The kernel version number is returned in the following structure:
464 The current version numbers are given by
465 .Dv BPF_MAJOR_VERSION
467 .Dv BPF_MINOR_VERSION
470 An incompatible filter
471 may result in undefined behavior (most likely, an error returned by
473 or haphazard packet matching).
477 Set or get the status of the
480 Set to zero if the link level source address should be filled in automatically
481 by the interface output routine.
482 Set to one if the link level source
483 address will be written, as provided, to the wire.
484 This flag is initialized to zero by default.
488 These commands are obsolete but left for compatibility.
494 Set or get the flag determining whether locally generated packets on the
495 interface should be returned by BPF.
496 Set to zero to see only incoming packets on the interface.
497 Set to one to see packets originating locally and remotely on the interface.
498 This flag is initialized to one by default.
499 .It Dv BIOCSDIRECTION
500 .It Dv BIOCGDIRECTION
502 Set or get the setting determining whether incoming, outgoing, or all packets
503 on the interface should be returned by BPF.
506 to see only incoming packets on the interface.
509 to see packets originating locally and remotely on the interface.
512 to see only outgoing packets on the interface.
513 This setting is initialized to
519 Set or get format and resolution of the time stamps returned by BPF.
521 .Dv BPF_T_MICROTIME ,
522 .Dv BPF_T_MICROTIME_FAST ,
523 .Dv BPF_T_MICROTIME_MONOTONIC ,
525 .Dv BPF_T_MICROTIME_MONOTONIC_FAST
526 to get time stamps in 64-bit
531 .Dv BPF_T_NANOTIME_FAST ,
532 .Dv BPF_T_NANOTIME_MONOTONIC ,
534 .Dv BPF_T_NANOTIME_MONOTONIC_FAST
535 to get time stamps in 64-bit
540 .Dv BPF_T_BINTIME_FAST ,
541 .Dv BPF_T_NANOTIME_MONOTONIC ,
543 .Dv BPF_T_BINTIME_MONOTONIC_FAST
544 to get time stamps in 64-bit
549 to ignore time stamp.
550 All 64-bit time stamp formats are wrapped in
553 .Dv BPF_T_MICROTIME_FAST ,
554 .Dv BPF_T_NANOTIME_FAST ,
555 .Dv BPF_T_BINTIME_FAST ,
556 .Dv BPF_T_MICROTIME_MONOTONIC_FAST ,
557 .Dv BPF_T_NANOTIME_MONOTONIC_FAST ,
559 .Dv BPF_T_BINTIME_MONOTONIC_FAST
560 are analogs of corresponding formats without _FAST suffix but do not perform
561 a full time counter query, so their accuracy is one timer tick.
563 .Dv BPF_T_MICROTIME_MONOTONIC ,
564 .Dv BPF_T_NANOTIME_MONOTONIC ,
565 .Dv BPF_T_BINTIME_MONOTONIC ,
566 .Dv BPF_T_MICROTIME_MONOTONIC_FAST ,
567 .Dv BPF_T_NANOTIME_MONOTONIC_FAST ,
569 .Dv BPF_T_BINTIME_MONOTONIC_FAST
570 store the time elapsed since kernel boot.
571 This setting is initialized to
576 Set packet feedback mode.
577 This allows injected packets to be fed back as input to the interface when
578 output via the interface is successful.
581 direction is set, injected outgoing packet is not returned by BPF to avoid
582 duplication. This flag is initialized to zero by default.
584 Set the locked flag on the
587 This prevents the execution of
588 ioctl commands which could change the underlying operating parameters of
590 .It Dv BIOCGETBUFMODE
591 .It Dv BIOCSETBUFMODE
593 Get or set the current
595 buffering mode; possible values are
596 .Dv BPF_BUFMODE_BUFFER ,
597 buffered read mode, and
598 .Dv BPF_BUFMODE_ZBUF ,
599 zero-copy buffer mode.
601 .Pq Li struct bpf_zbuf
602 Set the current zero-copy buffer locations; buffer locations may be
603 set only once zero-copy buffer mode has been selected, and prior to attaching
605 Buffers must be of identical size, page-aligned, and an integer multiple of
613 If buffers have already been set for this device, the ioctl will fail.
616 Get the largest individual zero-copy buffer size allowed.
617 As two buffers are used in zero-copy buffer mode, the limit (in practice) is
618 twice the returned size.
619 As zero-copy buffers consume kernel address space, conservative selection of
620 buffer size is suggested, especially when there are multiple
622 descriptors in use on 32-bit systems.
624 Force ownership of the next buffer to be assigned to userspace, if any data
625 present in the buffer.
626 If no data is present, the buffer will remain owned by the kernel.
627 This allows consumers of zero-copy buffering to implement timeouts and
628 retrieve partially filled buffers.
629 In order to handle the case where no data is present in the buffer and
630 therefore ownership is not assigned, the user process must check
636 One of the following structures is prepended to each packet returned by
638 or via a zero-copy buffer:
641 struct bpf_ts bh_tstamp; /* time stamp */
642 uint32_t bh_caplen; /* length of captured portion */
643 uint32_t bh_datalen; /* original length of packet */
644 u_short bh_hdrlen; /* length of bpf header (this struct
645 plus alignment padding) */
649 struct timeval bh_tstamp; /* time stamp */
650 uint32_t bh_caplen; /* length of captured portion */
651 uint32_t bh_datalen; /* original length of packet */
652 u_short bh_hdrlen; /* length of bpf header (this struct
653 plus alignment padding) */
657 The fields, whose values are stored in host order, and are:
659 .Bl -tag -compact -width bh_datalen
661 The time at which the packet was processed by the packet filter.
663 The length of the captured portion of the packet.
664 This is the minimum of
665 the truncation amount specified by the filter and the length of the packet.
667 The length of the packet off the wire.
668 This value is independent of the truncation amount specified by the filter.
672 header, which may not be equal to
673 .\" XXX - not really a function call
674 .Fn sizeof "struct bpf_xhdr"
676 .Fn sizeof "struct bpf_hdr" .
681 field exists to account for
682 padding between the header and the link level protocol.
683 The purpose here is to guarantee proper alignment of the packet
684 data structures, which is required on alignment sensitive
685 architectures and improves performance on many other architectures.
686 The packet filter ensures that the
689 and the network layer
690 header will be word aligned.
693 is used when the time stamp is set to
694 .Dv BPF_T_MICROTIME ,
695 .Dv BPF_T_MICROTIME_FAST ,
696 .Dv BPF_T_MICROTIME_MONOTONIC ,
697 .Dv BPF_T_MICROTIME_MONOTONIC_FAST ,
700 for backward compatibility reasons.
706 may be deprecated in the near future.
708 must be taken when accessing the link layer protocol fields on alignment
710 (This is not a problem on an Ethernet, since
711 the type field is a short falling on an even offset,
712 and the addresses are probably accessed in a bytewise fashion).
714 Additionally, individual packets are padded so that each starts
716 This requires that an application
717 has some knowledge of how to get from packet to packet.
724 It rounds up its argument to the nearest word aligned value (where a word is
730 points to the start of a packet, this expression
731 will advance it to the next packet:
732 .Dl p = (char *)p + BPF_WORDALIGN(p->bh_hdrlen + p->bh_caplen)
734 For the alignment mechanisms to work properly, the
737 must itself be word aligned.
741 will always return an aligned buffer.
743 A filter program is an array of instructions, with all branches forwardly
744 directed, terminated by a
747 Each instruction performs some action on the pseudo-machine state,
748 which consists of an accumulator, index register, scratch memory store,
749 and implicit program counter.
751 The following structure defines the instruction format:
763 field is used in different ways by different instructions,
768 fields are used as offsets
769 by the branch instructions.
770 The opcodes are encoded in a semi-hierarchical fashion.
771 There are eight classes of instructions:
781 Various other mode and
782 operator bits are or'd into the class to give the actual instructions.
783 The classes and modes are defined in
786 Below are the semantics for each defined
789 We use the convention that A is the accumulator, X is the index register,
790 P[] packet data, and M[] scratch memory store.
791 P[i:n] gives the data at byte offset
794 interpreted as a word (n=4),
795 unsigned halfword (n=2), or unsigned byte (n=1).
796 M[i] gives the i'th word in the scratch memory store, which is only
797 addressed in word units.
798 The memory store is indexed from 0 to
805 are the corresponding fields in the
806 instruction definition.
808 refers to the length of the packet.
809 .Bl -tag -width BPF_STXx
811 These instructions copy a value into the accumulator.
812 The type of the source operand is specified by an
814 and can be a constant
816 packet data at a fixed offset
818 packet data at a variable offset
822 or a word in the scratch memory store
828 the data size must be specified as a word
834 The semantics of all the recognized
838 BPF_LD+BPF_W+BPF_ABS A <- P[k:4]
839 BPF_LD+BPF_H+BPF_ABS A <- P[k:2]
840 BPF_LD+BPF_B+BPF_ABS A <- P[k:1]
841 BPF_LD+BPF_W+BPF_IND A <- P[X+k:4]
842 BPF_LD+BPF_H+BPF_IND A <- P[X+k:2]
843 BPF_LD+BPF_B+BPF_IND A <- P[X+k:1]
844 BPF_LD+BPF_W+BPF_LEN A <- len
845 BPF_LD+BPF_IMM A <- k
846 BPF_LD+BPF_MEM A <- M[k]
849 These instructions load a value into the index register.
851 the addressing modes are more restrictive than those of the accumulator loads,
854 a hack for efficiently loading the IP header length.
856 BPF_LDX+BPF_W+BPF_IMM X <- k
857 BPF_LDX+BPF_W+BPF_MEM X <- M[k]
858 BPF_LDX+BPF_W+BPF_LEN X <- len
859 BPF_LDX+BPF_B+BPF_MSH X <- 4*(P[k:1]&0xf)
862 This instruction stores the accumulator into the scratch memory.
863 We do not need an addressing mode since there is only one possibility
869 This instruction stores the index register in the scratch memory store.
874 The alu instructions perform operations between the accumulator and
875 index register or constant, and store the result back in the accumulator.
876 For binary operations, a source mode is required
881 BPF_ALU+BPF_ADD+BPF_K A <- A + k
882 BPF_ALU+BPF_SUB+BPF_K A <- A - k
883 BPF_ALU+BPF_MUL+BPF_K A <- A * k
884 BPF_ALU+BPF_DIV+BPF_K A <- A / k
885 BPF_ALU+BPF_AND+BPF_K A <- A & k
886 BPF_ALU+BPF_OR+BPF_K A <- A | k
887 BPF_ALU+BPF_LSH+BPF_K A <- A << k
888 BPF_ALU+BPF_RSH+BPF_K A <- A >> k
889 BPF_ALU+BPF_ADD+BPF_X A <- A + X
890 BPF_ALU+BPF_SUB+BPF_X A <- A - X
891 BPF_ALU+BPF_MUL+BPF_X A <- A * X
892 BPF_ALU+BPF_DIV+BPF_X A <- A / X
893 BPF_ALU+BPF_AND+BPF_X A <- A & X
894 BPF_ALU+BPF_OR+BPF_X A <- A | X
895 BPF_ALU+BPF_LSH+BPF_X A <- A << X
896 BPF_ALU+BPF_RSH+BPF_X A <- A >> X
897 BPF_ALU+BPF_NEG A <- -A
900 The jump instructions alter flow of control.
902 compare the accumulator against a constant
904 or the index register
906 If the result is true (or non-zero),
907 the true branch is taken, otherwise the false branch is taken.
908 Jump offsets are encoded in 8 bits so the longest jump is 256 instructions.
909 However, the jump always
911 opcode uses the 32 bit
913 field as the offset, allowing arbitrarily distant destinations.
914 All conditionals use unsigned comparison conventions.
916 BPF_JMP+BPF_JA pc += k
917 BPF_JMP+BPF_JGT+BPF_K pc += (A > k) ? jt : jf
918 BPF_JMP+BPF_JGE+BPF_K pc += (A >= k) ? jt : jf
919 BPF_JMP+BPF_JEQ+BPF_K pc += (A == k) ? jt : jf
920 BPF_JMP+BPF_JSET+BPF_K pc += (A & k) ? jt : jf
921 BPF_JMP+BPF_JGT+BPF_X pc += (A > X) ? jt : jf
922 BPF_JMP+BPF_JGE+BPF_X pc += (A >= X) ? jt : jf
923 BPF_JMP+BPF_JEQ+BPF_X pc += (A == X) ? jt : jf
924 BPF_JMP+BPF_JSET+BPF_X pc += (A & X) ? jt : jf
927 The return instructions terminate the filter program and specify the amount
928 of packet to accept (i.e., they return the truncation amount).
929 A return value of zero indicates that the packet should be ignored.
930 The return value is either a constant
935 BPF_RET+BPF_A accept A bytes
936 BPF_RET+BPF_K accept k bytes
939 The miscellaneous category was created for anything that does not
940 fit into the above classes, and for any new instructions that might need to
942 Currently, these are the register transfer instructions
943 that copy the index register to the accumulator or vice versa.
945 BPF_MISC+BPF_TAX X <- A
946 BPF_MISC+BPF_TXA A <- X
952 interface provides the following macros to facilitate
954 .Fn BPF_STMT opcode operand
956 .Fn BPF_JUMP opcode operand true_offset false_offset .
960 variables controls the behaviour of the
963 .Bl -tag -width indent
964 .It Va net.bpf.optimize_writers: No 0
965 Various programs use BPF to send (but not receive) raw packets
966 (cdpd, lldpd, dhcpd, dhcp relays, etc. are good examples of such programs).
967 They do not need incoming packets to be send to them.
968 Turning this option on
969 makes new BPF users to be attached to write-only interface list until program
970 explicitly specifies read filter via
971 .Fn pcap_set_filter .
972 This removes any performance degradation for high-speed interfaces.
973 .It Va net.bpf.stats:
974 Binary interface for retrieving general statistics.
975 .It Va net.bpf.zerocopy_enable: No 0
976 Permits zero-copy to be used with net BPF readers.
978 .It Va net.bpf.maxinsns: No 512
979 Maximum number of instructions that BPF program can contain.
983 option to determine approximate number of instruction for any filter.
984 .It Va net.bpf.maxbufsize: No 524288
985 Maximum buffer size to allocate for packets buffer.
986 .It Va net.bpf.bufsize: No 4096
987 Default buffer size to allocate for packets buffer.
990 The following filter is taken from the Reverse ARP Daemon.
991 It accepts only Reverse ARP requests.
993 struct bpf_insn insns[] = {
994 BPF_STMT(BPF_LD+BPF_H+BPF_ABS, 12),
995 BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, ETHERTYPE_REVARP, 0, 3),
996 BPF_STMT(BPF_LD+BPF_H+BPF_ABS, 20),
997 BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, REVARP_REQUEST, 0, 1),
998 BPF_STMT(BPF_RET+BPF_K, sizeof(struct ether_arp) +
999 sizeof(struct ether_header)),
1000 BPF_STMT(BPF_RET+BPF_K, 0),
1004 This filter accepts only IP packets between host 128.3.112.15 and
1007 struct bpf_insn insns[] = {
1008 BPF_STMT(BPF_LD+BPF_H+BPF_ABS, 12),
1009 BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, ETHERTYPE_IP, 0, 8),
1010 BPF_STMT(BPF_LD+BPF_W+BPF_ABS, 26),
1011 BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, 0x8003700f, 0, 2),
1012 BPF_STMT(BPF_LD+BPF_W+BPF_ABS, 30),
1013 BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, 0x80037023, 3, 4),
1014 BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, 0x80037023, 0, 3),
1015 BPF_STMT(BPF_LD+BPF_W+BPF_ABS, 30),
1016 BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, 0x8003700f, 0, 1),
1017 BPF_STMT(BPF_RET+BPF_K, (u_int)-1),
1018 BPF_STMT(BPF_RET+BPF_K, 0),
1022 Finally, this filter returns only TCP finger packets.
1023 We must parse the IP header to reach the TCP header.
1027 checks that the IP fragment offset is 0 so we are sure
1028 that we have a TCP header.
1030 struct bpf_insn insns[] = {
1031 BPF_STMT(BPF_LD+BPF_H+BPF_ABS, 12),
1032 BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, ETHERTYPE_IP, 0, 10),
1033 BPF_STMT(BPF_LD+BPF_B+BPF_ABS, 23),
1034 BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, IPPROTO_TCP, 0, 8),
1035 BPF_STMT(BPF_LD+BPF_H+BPF_ABS, 20),
1036 BPF_JUMP(BPF_JMP+BPF_JSET+BPF_K, 0x1fff, 6, 0),
1037 BPF_STMT(BPF_LDX+BPF_B+BPF_MSH, 14),
1038 BPF_STMT(BPF_LD+BPF_H+BPF_IND, 14),
1039 BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, 79, 2, 0),
1040 BPF_STMT(BPF_LD+BPF_H+BPF_IND, 16),
1041 BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, 79, 0, 1),
1042 BPF_STMT(BPF_RET+BPF_K, (u_int)-1),
1043 BPF_STMT(BPF_RET+BPF_K, 0),
1058 .%T "An efficient, extensible, and portable network monitor"
1061 The Enet packet filter was created in 1980 by Mike Accetta and
1062 Rick Rashid at Carnegie-Mellon University.
1064 Stanford, ported the code to
1066 and continued its development from
1068 Since then, it has evolved into the Ultrix Packet Filter at
1079 .An Steven McCanne ,
1080 of Lawrence Berkeley Laboratory, implemented BPF in
1082 Much of the design is due to
1085 Support for zero-copy buffers was added by
1086 .An Robert N. M. Watson
1087 under contract to Seccuris Inc.
1089 The read buffer must be of a fixed size (returned by the
1093 A file that does not request promiscuous mode may receive promiscuously
1094 received packets as a side effect of another file requesting this
1095 mode on the same hardware interface.
1096 This could be fixed in the kernel with additional processing overhead.
1097 However, we favor the model where
1098 all files must assume that the interface is promiscuous, and if
1099 so desired, must utilize a filter to reject foreign packets.
1101 Data link protocols with variable length headers are not currently supported.
1108 settings have been observed to work incorrectly on some interface
1109 types, including those with hardware loopback rather than software loopback,
1110 and point-to-point interfaces.
1111 They appear to function correctly on a
1112 broad range of Ethernet-style interfaces.