NetBSD/share/man/man4/bpf.4

.\" -*- nroff -*-
.\"
.\"	$NetBSD: bpf.4,v 1.10 1997/07/11 07:41:32 mikel Exp $
.\"
.\" Copyright (c) 1990, 1991, 1992, 1993, 1994
.\"	The Regents of the University of California.  All rights reserved.
.\"
.\" Redistribution and use in source and binary forms, with or without
.\" modification, are permitted provided that: (1) source code distributions
.\" retain the above copyright notice and this paragraph in its entirety, (2)
.\" distributions including binary code include the above copyright notice and
.\" this paragraph in its entirety in the documentation or other materials
.\" provided with the distribution, and (3) all advertising materials mentioning
.\" features or use of this software display the following acknowledgement:
.\" ``This product includes software developed by the University of California,
.\" Lawrence Berkeley Laboratory and its contributors.'' Neither the name of
.\" the University 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 ``AS IS'' AND WITHOUT ANY EXPRESS OR IMPLIED
.\" WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED WARRANTIES OF
.\" MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.
.\"
.\" This document is derived in part from the enet man page (enet.4)
.\" distributed with 4.3BSD Unix.
.\"
.TH BPF 4  "28 June 1994"
.SH NAME
bpf \- Berkeley Packet Filter
.SH SYNOPSIS
.B "pseudo-device bpfilter 16"
.SH DESCRIPTION
The Berkeley Packet Filter
provides a raw interface to data link layers in a protocol
independent fashion.
All packets on the network, even those destined for other hosts,
are accessible through this mechanism.
.PP
The packet filter appears as a character special device,
.I /dev/bpf0, /dev/bpf1,
etc.
After opening the device, the file descriptor must be bound to a
specific network interface with the BIOSETIF ioctl.
A given interface can be shared be multiple listeners, and the filter
underlying each descriptor will see an identical packet stream.
The total number of open
files is limited to the value given in the kernel configuration; the
example given in the SYNOPSIS above sets the limit to 16.
.PP
A separate device file is required for each minor device.
If a file is in use, the open will fail and
.I errno
will be set to EBUSY.
.PP
Associated with each open instance of a
.I bpf
file is a user-settable packet filter.
Whenever a packet is received by an interface,
all file descriptors listening on that interface apply their filter.
Each descriptor that accepts the packet receives its own copy.
.PP
Reads from these files return the next group of packets
that have matched the filter.
To improve performance, the buffer passed to read must be
the same size as the buffers used internally by
.I bpf.
This size is returned by the BIOCGBLEN ioctl (see below), and under
BSD, can be set with BIOCSBLEN.
Note that an individual packet larger than this size is necessarily
truncated.
.PP
The packet filter will support any link level protocol that has fixed length
headers.  Currently, only Ethernet, SLIP and PPP drivers have been
modified to interact with
.I bpf.
.PP
Since packet data is in network byte order, applications should use the
.I byteorder(3n)
macros to extract multi-byte values.
.PP
A packet can be sent out on the network by writing to a
.I bpf
file descriptor.  The writes are unbuffered, meaning only one
packet can be processed per write.
Currently, only writes to Ethernets and SLIP links are supported.
.SH IOCTLS
The
.I ioctl
command codes below are defined in <net/bpf.h>.  All commands require
these includes:
.ft B
.nf

	#include <sys/types.h>
	#include <sys/time.h>
	#include <sys/ioctl.h>
	#include <net/bpf.h>

.fi
.ft R
Additionally, BIOCGETIF and BIOCSETIF require \fB<net/if.h>\fR.

The (third) argument to the
.I ioctl
should be a pointer to the type indicated.
.TP 10
.B BIOCGBLEN (u_int)
Returns the required buffer length for reads on
.I bpf
files.
.TP 10
.B BIOCSBLEN (u_int)
Sets the buffer length for reads on
.I bpf
files.  The buffer must be set before the file is attached to an interface
with BIOCSETIF.
If the requested buffer size cannot be accommodated, the closest
allowable size will be set and returned in the argument.
A read call will result in EIO if it is passed a buffer that is not this size.
.TP 10
.B BIOCGDLT (u_int)
Returns the type of the data link layer underlying the attached interface.
EINVAL is returned if no interface has been specified.
The device types, prefixed with ``DLT_'', are defined in <net/bpf.h>.
.TP 10
.B BIOCPROMISC
Forces the interface into promiscuous mode.
All packets, not just those destined for the local host, are processed.
Since more than one file can be listening on a given interface,
a listener that opened its interface non-promiscuously may receive
packets promiscuously.  This problem can be remedied with an
appropriate filter.
.IP
The interface remains in promiscuous mode until all files listening
promiscuously are closed.
.TP 10
.B BIOCFLUSH
Flushes the buffer of incoming packets,
and resets the statistics that are returned by BIOCGSTATS.
.TP 10
.B BIOCGETIF (struct ifreq)
Returns the name of the hardware interface that the file is listening on.
The name is returned in the ifr_name field of
.I ifr.
All other fields are undefined.
.TP 10
.B BIOCSETIF (struct ifreq)
Sets the hardware interface associate with the file.  This
command must be performed before any packets can be read.
The device is indicated by name using the
.I ifr_name
field of the
.I ifreq.
Additionally, performs the actions of BIOCFLUSH.
.TP 10
.B BIOCSRTIMEOUT, BIOCGRTIMEOUT (struct timeval)
Set or get the read timeout parameter.
The
.I timeval
specifies the length of time to wait before timing
out on a read request.
This parameter is initialized to zero by
.IR open(2),
indicating no timeout.
.TP 10
.B BIOCGSTATS (struct bpf_stat)
Returns the following structure of packet statistics:
.ft B
.nf

struct bpf_stat {
	u_int bs_recv;
	u_int bs_drop;
};
.fi
.ft R
.IP
The fields are:
.RS
.TP 15
.I bs_recv
the number of packets received by the descriptor since opened or reset
(including any buffered since the last read call);
and
.TP
.I bs_drop
the number of packets which were accepted by the filter but dropped by the
kernel because of buffer overflows
(i.e., the application's reads aren't keeping up with the packet traffic).
.RE
.TP 10
.B BIOCIMMEDIATE (u_int)
Enable or disable ``immediate mode'', based on the truth value of the argument.
When immediate mode is enabled, reads return immediately upon packet
reception.  Otherwise, a read will block until either the kernel buffer
becomes full or a timeout occurs.
This is useful for programs like
.I rarpd(8c),
which must respond to messages in real time.
The default for a new file is off.
.TP 10
.B BIOCSETF (struct bpf_program)
Sets the filter program used by the kernel to discard uninteresting
packets.  An array of instructions and its length is passed in using
the following structure:
.ft B
.nf

struct bpf_program {
	int bf_len;
	struct bpf_insn *bf_insns;
};
.fi
.ft R
.IP
The filter program is pointed to by the
.I bf_insns
field while its length in units of `struct bpf_insn' is given by the
.I bf_len
field.
Also, the actions of BIOCFLUSH are performed.
.IP
See section \fBFILTER MACHINE\fP for an explanation of the filter language.
.TP 10
.B BIOCVERSION (struct bpf_version)
Returns the major and minor version numbers of the filter language currently
recognized by the kernel.  Before installing a filter, applications must check
that the current version is compatible with the running kernel.  Version
numbers are compatible if the major numbers match and the application minor
is less than or equal to the kernel minor.  The kernel version number is
returned in the following structure:
.ft B
.nf

struct bpf_version {
	u_short bv_major;
	u_short bv_minor;
};
.fi
.ft R
.IP
The current version numbers are given by
.B BPF_MAJOR_VERSION
and
.B BPF_MINOR_VERSION
from <net/bpf.h>.
An incompatible filter
may result in undefined behavior (most likely, an error returned by
.I ioctl()
or haphazard packet matching).
.TP 10
.B BIOCSRSIG BIOCGRSIG (u_int signal)
Set or get the receive signal.  This signal will be sent to the process or process group
specified by FIOSETOWN.  It defaults to SIGIO.
.SH STANDARD IOCTLS
.I bpf
now supports several standard
.I ioctls
which allow the user to do async and/or non-blocking I/O to an open
.I bpf
file descriptor.
.TP 10
.B FIONREAD (int)
Returns the number of bytes that are immediately available for reading.
.TP 10
.B SIOCGIFADDR (struct ifreq)
Returns the address associated with the interface.
.TP 10
.B FIONBIO (int)
Set or clear non-blocking I/O.  If arg is non-zero, then doing a
.I read
when no data is available will return -1 and
.I errno
will be set to EWOULDBLOCK.
If arg is zero, non-blocking I/O is disabled.  Note:  setting this
overrides the timeout set by BIOCSRTIMEOUT.
.TP 10
.B FIOASYNC (int)
Enable or disable async I/O.  When enabled (arg is non-zero), the process or
process group specified by FIOSETOWN will start receiving SIGIO's when packets
arrive.  Note that you must do an FIOSETOWN in order for this to take affect, as
the system will not default this for you.
The signal may be changed via BIOCSRSIG.
.TP 10
.B FIOSETOWN FIOGETOWN (int)
Set or get the process or process group (if negative) that should receive SIGIO
when packets are available.  The signal may be changed using BIOCSRSIG (see above).
.SH BPF HEADER
The following structure is prepended to each packet returned by
.I read(2):
.in 15
.ft B
.nf

struct bpf_hdr {
	struct timeval bh_tstamp;
	u_long bh_caplen;
	u_long bh_datalen;
	u_short bh_hdrlen;
};
.fi
.ft R
.in -15
.PP
The fields, whose values are stored in host order, and are:
.TP 15
.I bh_tstamp
The time at which the packet was processed by the packet filter.
.TP
.I bh_caplen
The length of the captured portion of the packet.  This is the minimum of
the truncation amount specified by the filter and the length of the packet.
.TP
.I bh_datalen
The length of the packet off the wire.
This value is independent of the truncation amount specified by the filter.
.TP
.I bh_hdrlen
The length of the BPF header, which may not be equal to
.I sizeof(struct bpf_hdr).
.RE
.PP
The
.I bh_hdrlen
field exists to account for
padding between the header and the link level protocol.
The purpose here is to guarantee proper alignment of the packet
data structures, which is required on alignment sensitive
architectures and and improves performance on many other architectures.
The packet filter insures that the
.I bpf_hdr
and the \fInetwork layer\fR header will be word aligned.  Suitable precautions
must be taken when accessing the link layer protocol fields on alignment
restricted machines.  (This isn't a problem on an Ethernet, since
the type field is a short falling on an even offset,
and the addresses are probably accessed in a bytewise fashion).
.PP
Additionally, individual packets are padded so that each starts
on a word boundary.  This requires that an application
has some knowledge of how to get from packet to packet.
The macro BPF_WORDALIGN is defined in <net/bpf.h> to facilitate
this process.  It rounds up its argument
to the nearest word aligned value (where a word is BPF_ALIGNMENT bytes wide).
.PP
For example, if `p' points to the start of a packet, this expression
will advance it to the next packet:
.sp
.RS
.ce 1
.nf
p = (char *)p + BPF_WORDALIGN(p->bh_hdrlen + p->bh_caplen)
.fi
.RE
.PP
For the alignment mechanisms to work properly, the
buffer passed to
.I read(2)
must itself be word aligned.
.I malloc(3)
will always return an aligned buffer.
.ft R
.SH FILTER MACHINE
A filter program is an array of instructions, with all branches forwardly
directed, terminated by a \fBreturn\fP instruction.
Each instruction performs some action on the pseudo-machine state,
which consists of an accumulator, index register, scratch memory store,
and implicit program counter.

The following structure defines the instruction format:
.RS
.ft B
.nf

struct bpf_insn {
	u_short	code;
	u_char 	jt;
	u_char 	jf;
	long k;
};
.fi
.ft R
.RE

The \fIk\fP field is used in different ways by different instructions,
and the \fIjt\fP and \fIjf\fP fields are used as offsets
by the branch instructions.
The opcodes are encoded in a semi-hierarchical fashion.
There are eight classes of instructions: BPF_LD, BPF_LDX, BPF_ST, BPF_STX,
BPF_ALU, BPF_JMP, BPF_RET, and BPF_MISC.  Various other mode and
operator bits are or'd into the class to give the actual instructions.
The classes and modes are defined in <net/bpf.h>.

Below are the semantics for each defined BPF instruction.
We use the convention that A is the accumulator, X is the index register,
P[] packet data, and M[] scratch memory store.
P[i:n] gives the data at byte offset ``i'' in the packet,
interpreted as a word (n=4),
unsigned halfword (n=2), or unsigned byte (n=1).
M[i] gives the i'th word in the scratch memory store, which is only
addressed in word units.  The memory store is indexed from 0 to BPF_MEMWORDS-1.
\fIk\fP, \fIjt\fP, and \fIjf\fP are the corresponding fields in the
instruction definition.  ``len'' refers to the length of the packet.

.TP 10
.B BPF_LD
These instructions copy a value into the accumulator.  The type of the
source operand is specified by an ``addressing mode'' and can be
a constant (\fBBPF_IMM\fP), packet data at a fixed offset (\fBBPF_ABS\fP),
packet data at a variable offset (\fBBPF_IND\fP), the packet length
(\fBBPF_LEN\fP),
or a word in the scratch memory store (\fBBPF_MEM\fP).
For \fBBPF_IND\fP and \fBBPF_ABS\fP, the data size must be specified as a word
(\fBBPF_W\fP), halfword (\fBBPF_H\fP), or byte (\fBBPF_B\fP).
The semantics of all the recognized BPF_LD instructions follow.

.RS
.TP 30
.B BPF_LD+BPF_W+BPF_ABS
A <- P[k:4]
.TP
.B BPF_LD+BPF_H+BPF_ABS
A <- P[k:2]
.TP
.B BPF_LD+BPF_B+BPF_ABS
A <- P[k:1]
.TP
.B BPF_LD+BPF_W+BPF_IND
A <- P[X+k:4]
.TP
.B BPF_LD+BPF_H+BPF_IND
A <- P[X+k:2]
.TP
.B BPF_LD+BPF_B+BPF_IND
A <- P[X+k:1]
.TP
.B BPF_LD+BPF_W+BPF_LEN
A <- len
.TP
.B BPF_LD+BPF_IMM
A <- k
.TP
.B BPF_LD+BPF_MEM
A <- M[k]
.RE

.TP 10
.B BPF_LDX
These instructions load a value into the index register.  Note that
the addressing modes are more restricted than those of the accumulator loads,
but they include
.B BPF_MSH,
a hack for efficiently loading the IP header length.
.RS
.TP 30
.B BPF_LDX+BPF_W+BPF_IMM
X <- k
.TP
.B BPF_LDX+BPF_W+BPF_MEM
X <- M[k]
.TP
.B BPF_LDX+BPF_W+BPF_LEN
X <- len
.TP
.B BPF_LDX+BPF_B+BPF_MSH
X <- 4*(P[k:1]&0xf)
.RE

.TP 10
.B BPF_ST
This instruction stores the accumulator into the scratch memory.
We do not need an addressing mode since there is only one possibility
for the destination.
.RS
.TP 30
.B BPF_ST
M[k] <- A
.RE

.TP 10
.B BPF_STX
This instruction stores the index register in the scratch memory store.
.RS
.TP 30
.B BPF_STX
M[k] <- X
.RE

.TP 10
.B BPF_ALU
The alu instructions perform operations between the accumulator and
index register or constant, and store the result back in the accumulator.
For binary operations, a source mode is required (\fBBPF_K\fP or
\fBBPF_X\fP).
.RS
.TP 30
.B BPF_ALU+BPF_ADD+BPF_K
A <- A + k
.TP
.B BPF_ALU+BPF_SUB+BPF_K
A <- A - k
.TP
.B BPF_ALU+BPF_MUL+BPF_K
A <- A * k
.TP
.B BPF_ALU+BPF_DIV+BPF_K
A <- A / k
.TP
.B BPF_ALU+BPF_AND+BPF_K
A <- A & k
.TP
.B BPF_ALU+BPF_OR+BPF_K
A <- A | k
.TP
.B BPF_ALU+BPF_LSH+BPF_K
A <- A << k
.TP
.B BPF_ALU+BPF_RSH+BPF_K
A <- A >> k
.TP
.B BPF_ALU+BPF_ADD+BPF_X
A <- A + X
.TP
.B BPF_ALU+BPF_SUB+BPF_X
A <- A - X
.TP
.B BPF_ALU+BPF_MUL+BPF_X
A <- A * X
.TP
.B BPF_ALU+BPF_DIV+BPF_X
A <- A / X
.TP
.B BPF_ALU+BPF_AND+BPF_X
A <- A & X
.TP
.B BPF_ALU+BPF_OR+BPF_X
A <- A | X
.TP
.B BPF_ALU+BPF_LSH+BPF_X
A <- A << X
.TP
.B BPF_ALU+BPF_RSH+BPF_X
A <- A >> X
.TP
.B BPF_ALU+BPF_NEG
A <- -A
.RE

.TP 10
.B BPF_JMP
The jump instructions alter flow of control.  Conditional jumps
compare the accumulator against a constant (\fBBPF_K\fP) or
the index register (\fBBPF_X\fP).  If the result is true (or non-zero),
the true branch is taken, otherwise the false branch is taken.
Jump offsets are encoded in 8 bits so the longest jump is 256 instructions.
However, the jump always (\fBBPF_JA\fP) opcode uses the 32 bit \fIk\fP
field as the offset, allowing arbitrarily distant destinations.
All conditionals use unsigned comparison conventions.
.RS
.TP 30
.B BPF_JMP+BPF_JA
pc += k
.TP
.B BPF_JMP+BPF_JGT+BPF_K
pc += (A > k) ? jt : jf
.TP
.B BPF_JMP+BPF_JGE+BPF_K
pc += (A >= k) ? jt : jf
.TP
.B BPF_JMP+BPF_JEQ+BPF_K
pc += (A == k) ? jt : jf
.TP
.B BPF_JMP+BPF_JSET+BPF_K
pc += (A & k) ? jt : jf
.TP
.B BPF_JMP+BPF_JGT+BPF_X
pc += (A > X) ? jt : jf
.TP
.B BPF_JMP+BPF_JGE+BPF_X
pc += (A >= X) ? jt : jf
.TP
.B BPF_JMP+BPF_JEQ+BPF_X
pc += (A == X) ? jt : jf
.TP
.B BPF_JMP+BPF_JSET+BPF_X
pc += (A & X) ? jt : jf
.RE
.TP 10
.B BPF_RET
The return instructions terminate the filter program and specify the amount
of packet to accept (i.e., they return the truncation amount).  A return
value of zero indicates that the packet should be ignored.
The return value is either a constant (\fBBPF_K\fP) or the accumulator
(\fBBPF_A\fP).
.RS
.TP 30
.B BPF_RET+BPF_A
accept A bytes
.TP
.B BPF_RET+BPF_K
accept k bytes
.RE
.TP 10
.B BPF_MISC
The miscellaneous category was created for anything that doesn't
fit into the above classes, and for any new instructions that might need to
be added.  Currently, these are the register transfer instructions
that copy the index register to the accumulator or vice versa.
.RS
.TP 30
.B BPF_MISC+BPF_TAX
X <- A
.TP
.B BPF_MISC+BPF_TXA
A <- X
.RE
.PP
The BPF interface provides the following macros to facilitate
array initializers:
.RS
\fBBPF_STMT\fI(opcode, operand)\fR
.br
and
.br
\fBBPF_JUMP\fI(opcode, operand, true_offset, false_offset)\fR
.RE
.PP
.SH EXAMPLES
The following filter is taken from the Reverse ARP Daemon.  It accepts
only Reverse ARP requests.
.RS
.nf

struct bpf_insn insns[] = {
	BPF_STMT(BPF_LD+BPF_H+BPF_ABS, 12),
	BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, ETHERTYPE_REVARP, 0, 3),
	BPF_STMT(BPF_LD+BPF_H+BPF_ABS, 20),
	BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, REVARP_REQUEST, 0, 1),
	BPF_STMT(BPF_RET+BPF_K, sizeof(struct ether_arp) +
		 sizeof(struct ether_header)),
	BPF_STMT(BPF_RET+BPF_K, 0),
};
.fi
.RE
.PP
This filter accepts only IP packets between host 128.3.112.15 and
128.3.112.35.
.RS
.nf

struct bpf_insn insns[] = {
	BPF_STMT(BPF_LD+BPF_H+BPF_ABS, 12),
	BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, ETHERTYPE_IP, 0, 8),
	BPF_STMT(BPF_LD+BPF_W+BPF_ABS, 26),
	BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, 0x8003700f, 0, 2),
	BPF_STMT(BPF_LD+BPF_W+BPF_ABS, 30),
	BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, 0x80037023, 3, 4),
	BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, 0x80037023, 0, 3),
	BPF_STMT(BPF_LD+BPF_W+BPF_ABS, 30),
	BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, 0x8003700f, 0, 1),
	BPF_STMT(BPF_RET+BPF_K, (u_int)-1),
	BPF_STMT(BPF_RET+BPF_K, 0),
};
.fi
.RE
.PP
Finally, this filter returns only TCP finger packets.  We must parse
the IP header to reach the TCP header.  The \fBBPF_JSET\fP instruction
checks that the IP fragment offset is 0 so we are sure
that we have a TCP header.
.RS
.nf

struct bpf_insn insns[] = {
	BPF_STMT(BPF_LD+BPF_H+BPF_ABS, 12),
	BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, ETHERTYPE_IP, 0, 10),
	BPF_STMT(BPF_LD+BPF_B+BPF_ABS, 23),
	BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, IPPROTO_TCP, 0, 8),
	BPF_STMT(BPF_LD+BPF_H+BPF_ABS, 20),
	BPF_JUMP(BPF_JMP+BPF_JSET+BPF_K, 0x1fff, 6, 0),
	BPF_STMT(BPF_LDX+BPF_B+BPF_MSH, 14),
	BPF_STMT(BPF_LD+BPF_H+BPF_IND, 14),
	BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, 79, 2, 0),
	BPF_STMT(BPF_LD+BPF_H+BPF_IND, 16),
	BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, 79, 0, 1),
	BPF_STMT(BPF_RET+BPF_K, (u_int)-1),
	BPF_STMT(BPF_RET+BPF_K, 0),
};
.fi
.RE
.SH SEE ALSO
tcpdump(8), signal(3), ioctl(2), read(2), select(2)
.LP
McCanne, S. and Jacobson V.
.RI "The BSD Packet Filter: A New Architecture for User-level Packet Capture".
Proceedings of the 1993 Winter USENIX
Technical Conference, San Diego, CA.
.SH FILES
/dev/bpf0, /dev/bpf1, ...
.SH BUGS
The read buffer must be of a fixed size (returned by the BIOCGBLEN ioctl).
.PP
A file that does not request promiscuous mode may receive promiscuously
received packets as a side effect of another file requesting this
mode on the same hardware interface.  This could be fixed in the kernel
with additional processing overhead.  However, we favor the model where
all files must assume that the interface is promiscuous, and if
so desired, must utilize a filter to reject foreign packets.
.PP
Data link protocols with variable length headers are not currently supported.
.PP
Under SunOS, if a BPF application reads more than 2^31 bytes of
data, read will fail in EINVAL.  You can either fix the bug in SunOS,
or lseek to 0 when read fails for this reason.
.PP
"Immediate mode" and the "read timeout" are misguided features.
This functionality can be emulated with non-blocking mode and
.I select(2).
.SH HISTORY
.PP
The Enet packet filter was created in 1980 by Mike Accetta and
Rick Rashid at Carnegie-Mellon University.  Jeffrey Mogul, at
Stanford, ported the code to BSD and continued its development from
1983 on.  Since then, it has evolved into the Ultrix Packet Filter
at DEC, a STREAMS NIT module under SunOS 4.1, and BPF.
.SH AUTHORS
.PP
Steven McCanne, of Lawrence Berkeley Laboratory, implemented BPF in
Summer 1990.  The design was in collaboration with Van Jacobson,
also of Lawrence Berkeley Laboratory.