Linux Socket Filtering aka Berkeley Packet Filter (BPF)

Notice

This file used to document the eBPF format and mechanisms even when not related to socket filtering. The BPF Documentation has more details on eBPF.

Introduction

Linux Socket Filtering (LSF) is derived from the Berkeley Packet Filter. Though there are some distinct differences between the BSD and Linux Kernel filtering, but when we speak of BPF or LSF in Linux context, we mean the very same mechanism of filtering in the Linux kernel.

BPF allows a user-space program to attach a filter onto any socket and allow or disallow certain types of data to come through the socket. LSF follows exactly the same filter code structure as BSD’s BPF, so referring to the BSD bpf.4 manpage is very helpful in creating filters.

On Linux, BPF is much simpler than on BSD. One does not have to worry about devices or anything like that. You simply create your filter code, send it to the kernel via the SO_ATTACH_FILTER option and if your filter code passes the kernel check on it, you then immediately begin filtering data on that socket.

You can also detach filters from your socket via the SO_DETACH_FILTER option. This will probably not be used much since when you close a socket that has a filter on it the filter is automagically removed. The other less common case may be adding a different filter on the same socket where you had another filter that is still running: the kernel takes care of removing the old one and placing your new one in its place, assuming your filter has passed the checks, otherwise if it fails the old filter will remain on that socket.

SO_LOCK_FILTER option allows to lock the filter attached to a socket. Once set, a filter cannot be removed or changed. This allows one process to setup a socket, attach a filter, lock it then drop privileges and be assured that the filter will be kept until the socket is closed.

The biggest user of this construct might be libpcap. Issuing a high-level filter command like tcpdump -i em1 port 22 passes through the libpcap internal compiler that generates a structure that can eventually be loaded via SO_ATTACH_FILTER to the kernel. tcpdump -i em1 port 22 -ddd displays what is being placed into this structure.

Although we were only speaking about sockets here, BPF in Linux is used in many more places. There’s xt_bpf for netfilter, cls_bpf in the kernel qdisc layer, SECCOMP-BPF (SECure COMPuting [1]), and lots of other places such as team driver, PTP code, etc where BPF is being used.

Original BPF paper:

Steven McCanne and Van Jacobson. 1993. The BSD packet filter: a new architecture for user-level packet capture. In Proceedings of the USENIX Winter 1993 Conference Proceedings on USENIX Winter 1993 Conference Proceedings (USENIX’93). USENIX Association, Berkeley, CA, USA, 2-2. [http://www.tcpdump.org/papers/bpf-usenix93.pdf]

Structure

User space applications include <linux/filter.h> which contains the following relevant structures:

struct sock_filter {    /* Filter block */
        __u16   code;   /* Actual filter code */
        __u8    jt;     /* Jump true */
        __u8    jf;     /* Jump false */
        __u32   k;      /* Generic multiuse field */
};

Such a structure is assembled as an array of 4-tuples, that contains a code, jt, jf and k value. jt and jf are jump offsets and k a generic value to be used for a provided code:

struct sock_fprog {                     /* Required for SO_ATTACH_FILTER. */
        unsigned short             len; /* Number of filter blocks */
        struct sock_filter __user *filter;
};

For socket filtering, a pointer to this structure (as shown in follow-up example) is being passed to the kernel through setsockopt(2).

Example

#include <sys/socket.h>
#include <sys/types.h>
#include <arpa/inet.h>
#include <linux/if_ether.h>
/* ... */

/* From the example above: tcpdump -i em1 port 22 -dd */
struct sock_filter code[] = {
        { 0x28,  0,  0, 0x0000000c },
        { 0x15,  0,  8, 0x000086dd },
        { 0x30,  0,  0, 0x00000014 },
        { 0x15,  2,  0, 0x00000084 },
        { 0x15,  1,  0, 0x00000006 },
        { 0x15,  0, 17, 0x00000011 },
        { 0x28,  0,  0, 0x00000036 },
        { 0x15, 14,  0, 0x00000016 },
        { 0x28,  0,  0, 0x00000038 },
        { 0x15, 12, 13, 0x00000016 },
        { 0x15,  0, 12, 0x00000800 },
        { 0x30,  0,  0, 0x00000017 },
        { 0x15,  2,  0, 0x00000084 },
        { 0x15,  1,  0, 0x00000006 },
        { 0x15,  0,  8, 0x00000011 },
        { 0x28,  0,  0, 0x00000014 },
        { 0x45,  6,  0, 0x00001fff },
        { 0xb1,  0,  0, 0x0000000e },
        { 0x48,  0,  0, 0x0000000e },
        { 0x15,  2,  0, 0x00000016 },
        { 0x48,  0,  0, 0x00000010 },
        { 0x15,  0,  1, 0x00000016 },
        { 0x06,  0,  0, 0x0000ffff },
        { 0x06,  0,  0, 0x00000000 },
};

struct sock_fprog bpf = {
        .len = ARRAY_SIZE(code),
        .filter = code,
};

sock = socket(PF_PACKET, SOCK_RAW, htons(ETH_P_ALL));
if (sock < 0)
        /* ... bail out ... */

ret = setsockopt(sock, SOL_SOCKET, SO_ATTACH_FILTER, &bpf, sizeof(bpf));
if (ret < 0)
        /* ... bail out ... */

/* ... */
close(sock);

The above example code attaches a socket filter for a PF_PACKET socket in order to let all IPv4/IPv6 packets with port 22 pass. The rest will be dropped for this socket.

The setsockopt(2) call to SO_DETACH_FILTER doesn’t need any arguments and SO_LOCK_FILTER for preventing the filter to be detached, takes an integer value with 0 or 1.

Note that socket filters are not restricted to PF_PACKET sockets only, but can also be used on other socket families.

Summary of system calls:

  • setsockopt(sockfd, SOL_SOCKET, SO_ATTACH_FILTER, &val, sizeof(val));

  • setsockopt(sockfd, SOL_SOCKET, SO_DETACH_FILTER, &val, sizeof(val));

  • setsockopt(sockfd, SOL_SOCKET, SO_LOCK_FILTER, &val, sizeof(val));

Normally, most use cases for socket filtering on packet sockets will be covered by libpcap in high-level syntax, so as an application developer you should stick to that. libpcap wraps its own layer around all that.

Unless i) using/linking to libpcap is not an option, ii) the required BPF filters use Linux extensions that are not supported by libpcap’s compiler, iii) a filter might be more complex and not cleanly implementable with libpcap’s compiler, or iv) particular filter codes should be optimized differently than libpcap’s internal compiler does; then in such cases writing such a filter “by hand” can be of an alternative. For example, xt_bpf and cls_bpf users might have requirements that could result in more complex filter code, or one that cannot be expressed with libpcap (e.g. different return codes for various code paths). Moreover, BPF JIT implementors may wish to manually write test cases and thus need low-level access to BPF code as well.

BPF engine and instruction set

The examples below use an assembly notation instead of less readable opcodes. The notation is closely modelled after Steven McCanne’s and Van Jacobson’s BPF paper.

The BPF architecture consists of the following basic elements:

Element

Description

A

32 bit wide accumulator

X

32 bit wide X register

M[]

16 x 32 bit wide misc registers aka “scratch memory store”, addressable from 0 to 15

A BPF program is an array of instructions with the following elements (as already mentioned):

op:16, jt:8, jf:8, k:32

The element op is a 16 bit wide opcode that has a particular instruction encoded. jt and jf are two 8 bit wide jump targets, one for condition “jump if true”, the other one “jump if false”. Eventually, element k contains a miscellaneous argument that can be interpreted in different ways depending on the given instruction in op.

The instruction set consists of load, store, branch, alu, miscellaneous and return instructions. This table lists the assembly notation and what the corresponding opcodes defined in linux/filter.h do:

Instruction

Addressing mode

Description

ld

1, 2, 3, 4, 12

Load word into A

ldi

4

Load word into A

ldh

1, 2

Load half-word into A

ldb

1, 2

Load byte into A

ldx

3, 4, 5, 12

Load word into X

ldxi

4

Load word into X

ldxb

5

Load byte into X

st

3

Store A into M[]

stx

3

Store X into M[]

jmp

6

Jump to label

ja

6

Jump to label

jeq

7, 8, 9, 10

Jump on A == <x>

jneq

9, 10

Jump on A != <x>

jne

9, 10

Jump on A != <x>

jlt

9, 10

Jump on A < <x>

jle

9, 10

Jump on A <= <x>

jgt

7, 8, 9, 10

Jump on A > <x>

jge

7, 8, 9, 10

Jump on A >= <x>

jset

7, 8, 9, 10

Jump on A & <x>

add

0, 4

A + <x>

sub

0, 4

A - <x>

mul

0, 4

A * <x>

div

0, 4

A / <x>

mod

0, 4

A % <x>

neg

!A

and

0, 4

A & <x>

or

0, 4

A | <x>

xor

0, 4

A ^ <x>

lsh

0, 4

A << <x>

rsh

0, 4

A >> <x>

tax

Copy A into X

txa

Copy X into A

ret

4, 11

Return

The next table shows addressing formats from the 2nd column:

Addressing mode

Syntax

Description

0

x/%x

Register X

1

[k]

BHW at byte offset k in the packet

2

[x + k]

BHW at the offset X + k in the packet

3

M[k]

Word at offset k in M[]

4

#k

Literal value stored in k

5

4*([k]&0xf)

Lower nibble * 4 at byte offset k in the packet

6

L

Jump label L

7

#k,Lt,Lf

Jump to Lt if true, otherwise jump to Lf

8

x/%x,Lt,Lf

Jump to Lt if true, otherwise jump to Lf

9

#k,Lt

Jump to Lt if predicate is true

10

x/%x,Lt

Jump to Lt if predicate is true

11

a/%a

Accumulator A

12

extension

BPF extension

The Linux kernel also has a couple of BPF extensions that are used along with the class of load instructions by “overloading” the k argument with a negative offset + a particular extension offset. The result of such BPF extensions are loaded into A.

Possible BPF extensions are shown in the following table:

Extension

Description

len

skb->len

proto

skb->protocol

type

skb->pkt_type

poff

Payload start offset

ifidx

skb->dev->ifindex

nla

Netlink attribute of type X with offset A

nlan

Nested Netlink attribute of type X with offset A

mark

skb->mark

queue

skb->queue_mapping

hatype

skb->dev->type

rxhash

skb->hash

cpu

raw_smp_processor_id()

vlan_tci

skb_vlan_tag_get(skb)

vlan_avail

skb_vlan_tag_present(skb)

vlan_tpid

skb->vlan_proto

rand

get_random_u32()

These extensions can also be prefixed with ‘#’. Examples for low-level BPF:

ARP packets:

ldh [12]
jne #0x806, drop
ret #-1
drop: ret #0

IPv4 TCP packets:

ldh [12]
jne #0x800, drop
ldb [23]
jneq #6, drop
ret #-1
drop: ret #0

icmp random packet sampling, 1 in 4:

ldh [12]
jne #0x800, drop
ldb [23]
jneq #1, drop
# get a random uint32 number
ld rand
mod #4
jneq #1, drop
ret #-1
drop: ret #0

SECCOMP filter example:

ld [4]                  /* offsetof(struct seccomp_data, arch) */
jne #0xc000003e, bad    /* AUDIT_ARCH_X86_64 */
ld [0]                  /* offsetof(struct seccomp_data, nr) */
jeq #15, good           /* __NR_rt_sigreturn */
jeq #231, good          /* __NR_exit_group */
jeq #60, good           /* __NR_exit */
jeq #0, good            /* __NR_read */
jeq #1, good            /* __NR_write */
jeq #5, good            /* __NR_fstat */
jeq #9, good            /* __NR_mmap */
jeq #14, good           /* __NR_rt_sigprocmask */
jeq #13, good           /* __NR_rt_sigaction */
jeq #35, good           /* __NR_nanosleep */
bad: ret #0             /* SECCOMP_RET_KILL_THREAD */
good: ret #0x7fff0000   /* SECCOMP_RET_ALLOW */

Examples for low-level BPF extension:

Packet for interface index 13:

ld ifidx
jneq #13, drop
ret #-1
drop: ret #0

(Accelerated) VLAN w/ id 10:

ld vlan_tci
jneq #10, drop
ret #-1
drop: ret #0

The ARP filter shown above corresponds to the following sock_filter entries:

{ 0x28,  0,  0, 0x0000000c },
{ 0x15,  0,  1, 0x00000806 },
{ 0x06,  0,  0, 0xffffffff },
{ 0x06,  0,  0, 0000000000 },

JIT compiler

The Linux kernel has a built-in BPF JIT compiler for x86_64, SPARC, PowerPC, ARM, ARM64, MIPS, RISC-V, s390, and ARC and can be enabled through CONFIG_BPF_JIT. The JIT compiler is transparently invoked for each attached filter from user space or for internal kernel users if it has been previously enabled by root:

echo 1 > /proc/sys/net/core/bpf_jit_enable

For JIT developers, doing audits etc, each compile run can output the generated opcode image into the kernel log via:

echo 2 > /proc/sys/net/core/bpf_jit_enable

Example output from dmesg:

[ 3389.935842] flen=6 proglen=70 pass=3 image=ffffffffa0069c8f
[ 3389.935847] JIT code: 00000000: 55 48 89 e5 48 83 ec 60 48 89 5d f8 44 8b 4f 68
[ 3389.935849] JIT code: 00000010: 44 2b 4f 6c 4c 8b 87 d8 00 00 00 be 0c 00 00 00
[ 3389.935850] JIT code: 00000020: e8 1d 94 ff e0 3d 00 08 00 00 75 16 be 17 00 00
[ 3389.935851] JIT code: 00000030: 00 e8 28 94 ff e0 83 f8 01 75 07 b8 ff ff 00 00
[ 3389.935852] JIT code: 00000040: eb 02 31 c0 c9 c3

When CONFIG_BPF_JIT_ALWAYS_ON is enabled, bpf_jit_enable is permanently set to 1 and setting any other value than that will return in failure. This is even the case for setting bpf_jit_enable to 2, since dumping the final JIT image into the kernel log is discouraged and introspection through bpftool (under tools/bpf/bpftool/) is the generally recommended approach instead.

BPF kernel internals

Internally, for the kernel interpreter, a different instruction set format with similar underlying principles from BPF described in previous paragraphs is being used. However, the instruction set format is modelled closer to the underlying architecture to mimic native instruction sets, so that a better performance can be achieved (more details later). This new ISA is called eBPF. See the BPF Documentation for details. (Note: eBPF which originates from [e]xtended BPF is not the same as BPF extensions! While eBPF is an ISA, BPF extensions date back to classic BPF’s ‘overloading’ of BPF_LD | BPF_{B,H,W} | BPF_ABS instruction.)

The new instruction set was originally designed with the possible goal in mind to write programs in “restricted C” and compile into eBPF with a optional GCC/LLVM backend, so that it can just-in-time map to modern 64-bit CPUs with minimal performance overhead over two steps, that is, C -> eBPF -> native code.

Currently, the new format is being used for running user BPF programs, which includes seccomp BPF, classic socket filters, cls_bpf traffic classifier, team driver’s classifier for its load-balancing mode, netfilter’s xt_bpf extension, PTP dissector/classifier, and much more. They are all internally converted by the kernel into the new instruction set representation and run in the eBPF interpreter. For in-kernel handlers, this all works transparently by using bpf_prog_create() for setting up the filter, resp. bpf_prog_destroy() for destroying it. The function bpf_prog_run(filter, ctx) transparently invokes eBPF interpreter or JITed code to run the filter. ‘filter’ is a pointer to struct bpf_prog that we got from bpf_prog_create(), and ‘ctx’ the given context (e.g. skb pointer). All constraints and restrictions from bpf_check_classic() apply before a conversion to the new layout is being done behind the scenes!

Currently, the classic BPF format is being used for JITing on most 32-bit architectures, whereas x86-64, aarch64, s390x, powerpc64, sparc64, arm32, riscv64, riscv32, loongarch64, arc perform JIT compilation from eBPF instruction set.

Testing

The BPF selftests under tools/testing/selftests/bpf ship instruction tests in test_progs, see jit_insns, which runs the eBPF instructions over a range of operand magnitudes and signs, register combinations, memory offsets and jump offsets against the BPF interpreter or the JIT compiler, depending on the net.core.bpf_jit_enable setting.

Misc

Also trinity, the Linux syscall fuzzer, has built-in support for BPF and SECCOMP-BPF kernel fuzzing.

Written by

The document was written in the hope that it is found useful and in order to give potential BPF hackers or security auditors a better overview of the underlying architecture.