xref: /linux/Documentation/bpf/standardization/instruction-set.rst (revision a1ff5a7d78a036d6c2178ee5acd6ba4946243800)
1.. contents::
2.. sectnum::
3
4======================================
5BPF Instruction Set Architecture (ISA)
6======================================
7
8eBPF, also commonly
9referred to as BPF, is a technology with origins in the Linux kernel
10that can run untrusted programs in a privileged context such as an
11operating system kernel. This document specifies the BPF instruction
12set architecture (ISA).
13
14As a historical note, BPF originally stood for Berkeley Packet Filter,
15but now that it can do so much more than packet filtering, the acronym
16no longer makes sense. BPF is now considered a standalone term that
17does not stand for anything.  The original BPF is sometimes referred to
18as cBPF (classic BPF) to distinguish it from the now widely deployed
19eBPF (extended BPF).
20
21Documentation conventions
22=========================
23
24The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
25"SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and
26"OPTIONAL" in this document are to be interpreted as described in
27BCP 14 `<https://www.rfc-editor.org/info/rfc2119>`_
28`<https://www.rfc-editor.org/info/rfc8174>`_
29when, and only when, they appear in all capitals, as shown here.
30
31For brevity and consistency, this document refers to families
32of types using a shorthand syntax and refers to several expository,
33mnemonic functions when describing the semantics of instructions.
34The range of valid values for those types and the semantics of those
35functions are defined in the following subsections.
36
37Types
38-----
39This document refers to integer types with the notation `SN` to specify
40a type's signedness (`S`) and bit width (`N`), respectively.
41
42.. table:: Meaning of signedness notation
43
44  ==== =========
45  S    Meaning
46  ==== =========
47  u    unsigned
48  s    signed
49  ==== =========
50
51.. table:: Meaning of bit-width notation
52
53  ===== =========
54  N     Bit width
55  ===== =========
56  8     8 bits
57  16    16 bits
58  32    32 bits
59  64    64 bits
60  128   128 bits
61  ===== =========
62
63For example, `u32` is a type whose valid values are all the 32-bit unsigned
64numbers and `s16` is a type whose valid values are all the 16-bit signed
65numbers.
66
67Functions
68---------
69
70The following byteswap functions are direction-agnostic.  That is,
71the same function is used for conversion in either direction discussed
72below.
73
74* be16: Takes an unsigned 16-bit number and converts it between
75  host byte order and big-endian
76  (`IEN137 <https://www.rfc-editor.org/ien/ien137.txt>`_) byte order.
77* be32: Takes an unsigned 32-bit number and converts it between
78  host byte order and big-endian byte order.
79* be64: Takes an unsigned 64-bit number and converts it between
80  host byte order and big-endian byte order.
81* bswap16: Takes an unsigned 16-bit number in either big- or little-endian
82  format and returns the equivalent number with the same bit width but
83  opposite endianness.
84* bswap32: Takes an unsigned 32-bit number in either big- or little-endian
85  format and returns the equivalent number with the same bit width but
86  opposite endianness.
87* bswap64: Takes an unsigned 64-bit number in either big- or little-endian
88  format and returns the equivalent number with the same bit width but
89  opposite endianness.
90* le16: Takes an unsigned 16-bit number and converts it between
91  host byte order and little-endian byte order.
92* le32: Takes an unsigned 32-bit number and converts it between
93  host byte order and little-endian byte order.
94* le64: Takes an unsigned 64-bit number and converts it between
95  host byte order and little-endian byte order.
96
97Definitions
98-----------
99
100.. glossary::
101
102  Sign Extend
103    To `sign extend an` ``X`` `-bit number, A, to a` ``Y`` `-bit number, B  ,` means to
104
105    #. Copy all ``X`` bits from `A` to the lower ``X`` bits of `B`.
106    #. Set the value of the remaining ``Y`` - ``X`` bits of `B` to the value of
107       the  most-significant bit of `A`.
108
109.. admonition:: Example
110
111  Sign extend an 8-bit number ``A`` to a 16-bit number ``B`` on a big-endian platform:
112  ::
113
114    A:          10000110
115    B: 11111111 10000110
116
117Conformance groups
118------------------
119
120An implementation does not need to support all instructions specified in this
121document (e.g., deprecated instructions).  Instead, a number of conformance
122groups are specified.  An implementation MUST support the base32 conformance
123group and MAY support additional conformance groups, where supporting a
124conformance group means it MUST support all instructions in that conformance
125group.
126
127The use of named conformance groups enables interoperability between a runtime
128that executes instructions, and tools such as compilers that generate
129instructions for the runtime.  Thus, capability discovery in terms of
130conformance groups might be done manually by users or automatically by tools.
131
132Each conformance group has a short ASCII label (e.g., "base32") that
133corresponds to a set of instructions that are mandatory.  That is, each
134instruction has one or more conformance groups of which it is a member.
135
136This document defines the following conformance groups:
137
138* base32: includes all instructions defined in this
139  specification unless otherwise noted.
140* base64: includes base32, plus instructions explicitly noted
141  as being in the base64 conformance group.
142* atomic32: includes 32-bit atomic operation instructions (see `Atomic operations`_).
143* atomic64: includes atomic32, plus 64-bit atomic operation instructions.
144* divmul32: includes 32-bit division, multiplication, and modulo instructions.
145* divmul64: includes divmul32, plus 64-bit division, multiplication,
146  and modulo instructions.
147* packet: deprecated packet access instructions.
148
149Instruction encoding
150====================
151
152BPF has two instruction encodings:
153
154* the basic instruction encoding, which uses 64 bits to encode an instruction
155* the wide instruction encoding, which appends a second 64 bits
156  after the basic instruction for a total of 128 bits.
157
158Basic instruction encoding
159--------------------------
160
161A basic instruction is encoded as follows::
162
163  +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
164  |    opcode     |     regs      |            offset             |
165  +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
166  |                              imm                              |
167  +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
168
169**opcode**
170  operation to perform, encoded as follows::
171
172    +-+-+-+-+-+-+-+-+
173    |specific |class|
174    +-+-+-+-+-+-+-+-+
175
176  **specific**
177    The format of these bits varies by instruction class
178
179  **class**
180    The instruction class (see `Instruction classes`_)
181
182**regs**
183  The source and destination register numbers, encoded as follows
184  on a little-endian host::
185
186    +-+-+-+-+-+-+-+-+
187    |src_reg|dst_reg|
188    +-+-+-+-+-+-+-+-+
189
190  and as follows on a big-endian host::
191
192    +-+-+-+-+-+-+-+-+
193    |dst_reg|src_reg|
194    +-+-+-+-+-+-+-+-+
195
196  **src_reg**
197    the source register number (0-10), except where otherwise specified
198    (`64-bit immediate instructions`_ reuse this field for other purposes)
199
200  **dst_reg**
201    destination register number (0-10), unless otherwise specified
202    (future instructions might reuse this field for other purposes)
203
204**offset**
205  signed integer offset used with pointer arithmetic, except where
206  otherwise specified (some arithmetic instructions reuse this field
207  for other purposes)
208
209**imm**
210  signed integer immediate value
211
212Note that the contents of multi-byte fields ('offset' and 'imm') are
213stored using big-endian byte ordering on big-endian hosts and
214little-endian byte ordering on little-endian hosts.
215
216For example::
217
218  opcode                  offset imm          assembly
219         src_reg dst_reg
220  07     0       1        00 00  44 33 22 11  r1 += 0x11223344 // little
221         dst_reg src_reg
222  07     1       0        00 00  11 22 33 44  r1 += 0x11223344 // big
223
224Note that most instructions do not use all of the fields.
225Unused fields SHALL be cleared to zero.
226
227Wide instruction encoding
228--------------------------
229
230Some instructions are defined to use the wide instruction encoding,
231which uses two 32-bit immediate values.  The 64 bits following
232the basic instruction format contain a pseudo instruction
233with 'opcode', 'dst_reg', 'src_reg', and 'offset' all set to zero.
234
235This is depicted in the following figure::
236
237  +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
238  |    opcode     |     regs      |            offset             |
239  +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
240  |                              imm                              |
241  +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
242  |                           reserved                            |
243  +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
244  |                           next_imm                            |
245  +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
246
247**opcode**
248  operation to perform, encoded as explained above
249
250**regs**
251  The source and destination register numbers (unless otherwise
252  specified), encoded as explained above
253
254**offset**
255  signed integer offset used with pointer arithmetic, unless
256  otherwise specified
257
258**imm**
259  signed integer immediate value
260
261**reserved**
262  unused, set to zero
263
264**next_imm**
265  second signed integer immediate value
266
267Instruction classes
268-------------------
269
270The three least significant bits of the 'opcode' field store the instruction class:
271
272.. table:: Instruction class
273
274  =====  =====  ===============================  ===================================
275  class  value  description                      reference
276  =====  =====  ===============================  ===================================
277  LD     0x0    non-standard load operations     `Load and store instructions`_
278  LDX    0x1    load into register operations    `Load and store instructions`_
279  ST     0x2    store from immediate operations  `Load and store instructions`_
280  STX    0x3    store from register operations   `Load and store instructions`_
281  ALU    0x4    32-bit arithmetic operations     `Arithmetic and jump instructions`_
282  JMP    0x5    64-bit jump operations           `Arithmetic and jump instructions`_
283  JMP32  0x6    32-bit jump operations           `Arithmetic and jump instructions`_
284  ALU64  0x7    64-bit arithmetic operations     `Arithmetic and jump instructions`_
285  =====  =====  ===============================  ===================================
286
287Arithmetic and jump instructions
288================================
289
290For arithmetic and jump instructions (``ALU``, ``ALU64``, ``JMP`` and
291``JMP32``), the 8-bit 'opcode' field is divided into three parts::
292
293  +-+-+-+-+-+-+-+-+
294  |  code |s|class|
295  +-+-+-+-+-+-+-+-+
296
297**code**
298  the operation code, whose meaning varies by instruction class
299
300**s (source)**
301  the source operand location, which unless otherwise specified is one of:
302
303  .. table:: Source operand location
304
305    ======  =====  ==============================================
306    source  value  description
307    ======  =====  ==============================================
308    K       0      use 32-bit 'imm' value as source operand
309    X       1      use 'src_reg' register value as source operand
310    ======  =====  ==============================================
311
312**instruction class**
313  the instruction class (see `Instruction classes`_)
314
315Arithmetic instructions
316-----------------------
317
318``ALU`` uses 32-bit wide operands while ``ALU64`` uses 64-bit wide operands for
319otherwise identical operations. ``ALU64`` instructions belong to the
320base64 conformance group unless noted otherwise.
321The 'code' field encodes the operation as below, where 'src' refers to the
322the source operand and 'dst' refers to the value of the destination
323register.
324
325.. table:: Arithmetic instructions
326
327  =====  =====  =======  ==========================================================
328  name   code   offset   description
329  =====  =====  =======  ==========================================================
330  ADD    0x0    0        dst += src
331  SUB    0x1    0        dst -= src
332  MUL    0x2    0        dst \*= src
333  DIV    0x3    0        dst = (src != 0) ? (dst / src) : 0
334  SDIV   0x3    1        dst = (src != 0) ? (dst s/ src) : 0
335  OR     0x4    0        dst \|= src
336  AND    0x5    0        dst &= src
337  LSH    0x6    0        dst <<= (src & mask)
338  RSH    0x7    0        dst >>= (src & mask)
339  NEG    0x8    0        dst = -dst
340  MOD    0x9    0        dst = (src != 0) ? (dst % src) : dst
341  SMOD   0x9    1        dst = (src != 0) ? (dst s% src) : dst
342  XOR    0xa    0        dst ^= src
343  MOV    0xb    0        dst = src
344  MOVSX  0xb    8/16/32  dst = (s8,s16,s32)src
345  ARSH   0xc    0        :term:`sign extending<Sign Extend>` dst >>= (src & mask)
346  END    0xd    0        byte swap operations (see `Byte swap instructions`_ below)
347  =====  =====  =======  ==========================================================
348
349Underflow and overflow are allowed during arithmetic operations, meaning
350the 64-bit or 32-bit value will wrap. If BPF program execution would
351result in division by zero, the destination register is instead set to zero.
352If execution would result in modulo by zero, for ``ALU64`` the value of
353the destination register is unchanged whereas for ``ALU`` the upper
35432 bits of the destination register are zeroed.
355
356``{ADD, X, ALU}``, where 'code' = ``ADD``, 'source' = ``X``, and 'class' = ``ALU``, means::
357
358  dst = (u32) ((u32) dst + (u32) src)
359
360where '(u32)' indicates that the upper 32 bits are zeroed.
361
362``{ADD, X, ALU64}`` means::
363
364  dst = dst + src
365
366``{XOR, K, ALU}`` means::
367
368  dst = (u32) dst ^ (u32) imm
369
370``{XOR, K, ALU64}`` means::
371
372  dst = dst ^ imm
373
374Note that most arithmetic instructions have 'offset' set to 0. Only three instructions
375(``SDIV``, ``SMOD``, ``MOVSX``) have a non-zero 'offset'.
376
377Division, multiplication, and modulo operations for ``ALU`` are part
378of the "divmul32" conformance group, and division, multiplication, and
379modulo operations for ``ALU64`` are part of the "divmul64" conformance
380group.
381The division and modulo operations support both unsigned and signed flavors.
382
383For unsigned operations (``DIV`` and ``MOD``), for ``ALU``,
384'imm' is interpreted as a 32-bit unsigned value. For ``ALU64``,
385'imm' is first :term:`sign extended<Sign Extend>` from 32 to 64 bits, and then
386interpreted as a 64-bit unsigned value.
387
388For signed operations (``SDIV`` and ``SMOD``), for ``ALU``,
389'imm' is interpreted as a 32-bit signed value. For ``ALU64``, 'imm'
390is first :term:`sign extended<Sign Extend>` from 32 to 64 bits, and then
391interpreted as a 64-bit signed value.
392
393Note that there are varying definitions of the signed modulo operation
394when the dividend or divisor are negative, where implementations often
395vary by language such that Python, Ruby, etc.  differ from C, Go, Java,
396etc. This specification requires that signed modulo MUST use truncated division
397(where -13 % 3 == -1) as implemented in C, Go, etc.::
398
399   a % n = a - n * trunc(a / n)
400
401The ``MOVSX`` instruction does a move operation with sign extension.
402``{MOVSX, X, ALU}`` :term:`sign extends<Sign Extend>` 8-bit and 16-bit operands into
40332-bit operands, and zeroes the remaining upper 32 bits.
404``{MOVSX, X, ALU64}`` :term:`sign extends<Sign Extend>` 8-bit, 16-bit, and 32-bit
405operands into 64-bit operands.  Unlike other arithmetic instructions,
406``MOVSX`` is only defined for register source operands (``X``).
407
408``{MOV, K, ALU64}`` means::
409
410  dst = (s64)imm
411
412``{MOV, X, ALU}`` means::
413
414  dst = (u32)src
415
416``{MOVSX, X, ALU}`` with 'offset' 8 means::
417
418  dst = (u32)(s32)(s8)src
419
420
421The ``NEG`` instruction is only defined when the source bit is clear
422(``K``).
423
424Shift operations use a mask of 0x3F (63) for 64-bit operations and 0x1F (31)
425for 32-bit operations.
426
427Byte swap instructions
428----------------------
429
430The byte swap instructions use instruction classes of ``ALU`` and ``ALU64``
431and a 4-bit 'code' field of ``END``.
432
433The byte swap instructions operate on the destination register
434only and do not use a separate source register or immediate value.
435
436For ``ALU``, the 1-bit source operand field in the opcode is used to
437select what byte order the operation converts from or to. For
438``ALU64``, the 1-bit source operand field in the opcode is reserved
439and MUST be set to 0.
440
441.. table:: Byte swap instructions
442
443  =====  ========  =====  =================================================
444  class  source    value  description
445  =====  ========  =====  =================================================
446  ALU    LE        0      convert between host byte order and little endian
447  ALU    BE        1      convert between host byte order and big endian
448  ALU64  Reserved  0      do byte swap unconditionally
449  =====  ========  =====  =================================================
450
451The 'imm' field encodes the width of the swap operations.  The following widths
452are supported: 16, 32 and 64.  Width 64 operations belong to the base64
453conformance group and other swap operations belong to the base32
454conformance group.
455
456Examples:
457
458``{END, LE, ALU}`` with 'imm' = 16/32/64 means::
459
460  dst = le16(dst)
461  dst = le32(dst)
462  dst = le64(dst)
463
464``{END, BE, ALU}`` with 'imm' = 16/32/64 means::
465
466  dst = be16(dst)
467  dst = be32(dst)
468  dst = be64(dst)
469
470``{END, TO, ALU64}`` with 'imm' = 16/32/64 means::
471
472  dst = bswap16(dst)
473  dst = bswap32(dst)
474  dst = bswap64(dst)
475
476Jump instructions
477-----------------
478
479``JMP32`` uses 32-bit wide operands and indicates the base32
480conformance group, while ``JMP`` uses 64-bit wide operands for
481otherwise identical operations, and indicates the base64 conformance
482group unless otherwise specified.
483The 'code' field encodes the operation as below:
484
485.. table:: Jump instructions
486
487  ========  =====  =======  =================================  ===================================================
488  code      value  src_reg  description                        notes
489  ========  =====  =======  =================================  ===================================================
490  JA        0x0    0x0      PC += offset                       {JA, K, JMP} only
491  JA        0x0    0x0      PC += imm                          {JA, K, JMP32} only
492  JEQ       0x1    any      PC += offset if dst == src
493  JGT       0x2    any      PC += offset if dst > src          unsigned
494  JGE       0x3    any      PC += offset if dst >= src         unsigned
495  JSET      0x4    any      PC += offset if dst & src
496  JNE       0x5    any      PC += offset if dst != src
497  JSGT      0x6    any      PC += offset if dst > src          signed
498  JSGE      0x7    any      PC += offset if dst >= src         signed
499  CALL      0x8    0x0      call helper function by static ID  {CALL, K, JMP} only, see `Helper functions`_
500  CALL      0x8    0x1      call PC += imm                     {CALL, K, JMP} only, see `Program-local functions`_
501  CALL      0x8    0x2      call helper function by BTF ID     {CALL, K, JMP} only, see `Helper functions`_
502  EXIT      0x9    0x0      return                             {CALL, K, JMP} only
503  JLT       0xa    any      PC += offset if dst < src          unsigned
504  JLE       0xb    any      PC += offset if dst <= src         unsigned
505  JSLT      0xc    any      PC += offset if dst < src          signed
506  JSLE      0xd    any      PC += offset if dst <= src         signed
507  ========  =====  =======  =================================  ===================================================
508
509where 'PC' denotes the program counter, and the offset to increment by
510is in units of 64-bit instructions relative to the instruction following
511the jump instruction.  Thus 'PC += 1' skips execution of the next
512instruction if it's a basic instruction or results in undefined behavior
513if the next instruction is a 128-bit wide instruction.
514
515Example:
516
517``{JSGE, X, JMP32}`` means::
518
519  if (s32)dst s>= (s32)src goto +offset
520
521where 's>=' indicates a signed '>=' comparison.
522
523``{JLE, K, JMP}`` means::
524
525  if dst <= (u64)(s64)imm goto +offset
526
527``{JA, K, JMP32}`` means::
528
529  gotol +imm
530
531where 'imm' means the branch offset comes from the 'imm' field.
532
533Note that there are two flavors of ``JA`` instructions. The
534``JMP`` class permits a 16-bit jump offset specified by the 'offset'
535field, whereas the ``JMP32`` class permits a 32-bit jump offset
536specified by the 'imm' field. A > 16-bit conditional jump may be
537converted to a < 16-bit conditional jump plus a 32-bit unconditional
538jump.
539
540All ``CALL`` and ``JA`` instructions belong to the
541base32 conformance group.
542
543Helper functions
544~~~~~~~~~~~~~~~~
545
546Helper functions are a concept whereby BPF programs can call into a
547set of function calls exposed by the underlying platform.
548
549Historically, each helper function was identified by a static ID
550encoded in the 'imm' field.  Further documentation of helper functions
551is outside the scope of this document and standardization is left for
552future work, but use is widely deployed and more information can be
553found in platform-specific documentation (e.g., Linux kernel documentation).
554
555Platforms that support the BPF Type Format (BTF) support identifying
556a helper function by a BTF ID encoded in the 'imm' field, where the BTF ID
557identifies the helper name and type.  Further documentation of BTF
558is outside the scope of this document and standardization is left for
559future work, but use is widely deployed and more information can be
560found in platform-specific documentation (e.g., Linux kernel documentation).
561
562Program-local functions
563~~~~~~~~~~~~~~~~~~~~~~~
564Program-local functions are functions exposed by the same BPF program as the
565caller, and are referenced by offset from the instruction following the call
566instruction, similar to ``JA``.  The offset is encoded in the 'imm' field of
567the call instruction. An ``EXIT`` within the program-local function will
568return to the caller.
569
570Load and store instructions
571===========================
572
573For load and store instructions (``LD``, ``LDX``, ``ST``, and ``STX``), the
5748-bit 'opcode' field is divided as follows::
575
576  +-+-+-+-+-+-+-+-+
577  |mode |sz |class|
578  +-+-+-+-+-+-+-+-+
579
580**mode**
581  The mode modifier is one of:
582
583  .. table:: Mode modifier
584
585    =============  =====  ====================================  =============
586    mode modifier  value  description                           reference
587    =============  =====  ====================================  =============
588    IMM            0      64-bit immediate instructions         `64-bit immediate instructions`_
589    ABS            1      legacy BPF packet access (absolute)   `Legacy BPF Packet access instructions`_
590    IND            2      legacy BPF packet access (indirect)   `Legacy BPF Packet access instructions`_
591    MEM            3      regular load and store operations     `Regular load and store operations`_
592    MEMSX          4      sign-extension load operations        `Sign-extension load operations`_
593    ATOMIC         6      atomic operations                     `Atomic operations`_
594    =============  =====  ====================================  =============
595
596**sz (size)**
597  The size modifier is one of:
598
599  .. table:: Size modifier
600
601    ====  =====  =====================
602    size  value  description
603    ====  =====  =====================
604    W     0      word        (4 bytes)
605    H     1      half word   (2 bytes)
606    B     2      byte
607    DW    3      double word (8 bytes)
608    ====  =====  =====================
609
610  Instructions using ``DW`` belong to the base64 conformance group.
611
612**class**
613  The instruction class (see `Instruction classes`_)
614
615Regular load and store operations
616---------------------------------
617
618The ``MEM`` mode modifier is used to encode regular load and store
619instructions that transfer data between a register and memory.
620
621``{MEM, <size>, STX}`` means::
622
623  *(size *) (dst + offset) = src
624
625``{MEM, <size>, ST}`` means::
626
627  *(size *) (dst + offset) = imm
628
629``{MEM, <size>, LDX}`` means::
630
631  dst = *(unsigned size *) (src + offset)
632
633Where '<size>' is one of: ``B``, ``H``, ``W``, or ``DW``, and
634'unsigned size' is one of: u8, u16, u32, or u64.
635
636Sign-extension load operations
637------------------------------
638
639The ``MEMSX`` mode modifier is used to encode :term:`sign-extension<Sign Extend>` load
640instructions that transfer data between a register and memory.
641
642``{MEMSX, <size>, LDX}`` means::
643
644  dst = *(signed size *) (src + offset)
645
646Where '<size>' is one of: ``B``, ``H``, or ``W``, and
647'signed size' is one of: s8, s16, or s32.
648
649Atomic operations
650-----------------
651
652Atomic operations are operations that operate on memory and can not be
653interrupted or corrupted by other access to the same memory region
654by other BPF programs or means outside of this specification.
655
656All atomic operations supported by BPF are encoded as store operations
657that use the ``ATOMIC`` mode modifier as follows:
658
659* ``{ATOMIC, W, STX}`` for 32-bit operations, which are
660  part of the "atomic32" conformance group.
661* ``{ATOMIC, DW, STX}`` for 64-bit operations, which are
662  part of the "atomic64" conformance group.
663* 8-bit and 16-bit wide atomic operations are not supported.
664
665The 'imm' field is used to encode the actual atomic operation.
666Simple atomic operation use a subset of the values defined to encode
667arithmetic operations in the 'imm' field to encode the atomic operation:
668
669.. table:: Simple atomic operations
670
671  ========  =====  ===========
672  imm       value  description
673  ========  =====  ===========
674  ADD       0x00   atomic add
675  OR        0x40   atomic or
676  AND       0x50   atomic and
677  XOR       0xa0   atomic xor
678  ========  =====  ===========
679
680
681``{ATOMIC, W, STX}`` with 'imm' = ADD means::
682
683  *(u32 *)(dst + offset) += src
684
685``{ATOMIC, DW, STX}`` with 'imm' = ADD means::
686
687  *(u64 *)(dst + offset) += src
688
689In addition to the simple atomic operations, there also is a modifier and
690two complex atomic operations:
691
692.. table:: Complex atomic operations
693
694  ===========  ================  ===========================
695  imm          value             description
696  ===========  ================  ===========================
697  FETCH        0x01              modifier: return old value
698  XCHG         0xe0 | FETCH      atomic exchange
699  CMPXCHG      0xf0 | FETCH      atomic compare and exchange
700  ===========  ================  ===========================
701
702The ``FETCH`` modifier is optional for simple atomic operations, and
703always set for the complex atomic operations.  If the ``FETCH`` flag
704is set, then the operation also overwrites ``src`` with the value that
705was in memory before it was modified.
706
707The ``XCHG`` operation atomically exchanges ``src`` with the value
708addressed by ``dst + offset``.
709
710The ``CMPXCHG`` operation atomically compares the value addressed by
711``dst + offset`` with ``R0``. If they match, the value addressed by
712``dst + offset`` is replaced with ``src``. In either case, the
713value that was at ``dst + offset`` before the operation is zero-extended
714and loaded back to ``R0``.
715
71664-bit immediate instructions
717-----------------------------
718
719Instructions with the ``IMM`` 'mode' modifier use the wide instruction
720encoding defined in `Instruction encoding`_, and use the 'src_reg' field of the
721basic instruction to hold an opcode subtype.
722
723The following table defines a set of ``{IMM, DW, LD}`` instructions
724with opcode subtypes in the 'src_reg' field, using new terms such as "map"
725defined further below:
726
727.. table:: 64-bit immediate instructions
728
729  =======  =========================================  ===========  ==============
730  src_reg  pseudocode                                 imm type     dst type
731  =======  =========================================  ===========  ==============
732  0x0      dst = (next_imm << 32) | imm               integer      integer
733  0x1      dst = map_by_fd(imm)                       map fd       map
734  0x2      dst = map_val(map_by_fd(imm)) + next_imm   map fd       data address
735  0x3      dst = var_addr(imm)                        variable id  data address
736  0x4      dst = code_addr(imm)                       integer      code address
737  0x5      dst = map_by_idx(imm)                      map index    map
738  0x6      dst = map_val(map_by_idx(imm)) + next_imm  map index    data address
739  =======  =========================================  ===========  ==============
740
741where
742
743* map_by_fd(imm) means to convert a 32-bit file descriptor into an address of a map (see `Maps`_)
744* map_by_idx(imm) means to convert a 32-bit index into an address of a map
745* map_val(map) gets the address of the first value in a given map
746* var_addr(imm) gets the address of a platform variable (see `Platform Variables`_) with a given id
747* code_addr(imm) gets the address of the instruction at a specified relative offset in number of (64-bit) instructions
748* the 'imm type' can be used by disassemblers for display
749* the 'dst type' can be used for verification and JIT compilation purposes
750
751Maps
752~~~~
753
754Maps are shared memory regions accessible by BPF programs on some platforms.
755A map can have various semantics as defined in a separate document, and may or
756may not have a single contiguous memory region, but the 'map_val(map)' is
757currently only defined for maps that do have a single contiguous memory region.
758
759Each map can have a file descriptor (fd) if supported by the platform, where
760'map_by_fd(imm)' means to get the map with the specified file descriptor. Each
761BPF program can also be defined to use a set of maps associated with the
762program at load time, and 'map_by_idx(imm)' means to get the map with the given
763index in the set associated with the BPF program containing the instruction.
764
765Platform Variables
766~~~~~~~~~~~~~~~~~~
767
768Platform variables are memory regions, identified by integer ids, exposed by
769the runtime and accessible by BPF programs on some platforms.  The
770'var_addr(imm)' operation means to get the address of the memory region
771identified by the given id.
772
773Legacy BPF Packet access instructions
774-------------------------------------
775
776BPF previously introduced special instructions for access to packet data that were
777carried over from classic BPF. These instructions used an instruction
778class of ``LD``, a size modifier of ``W``, ``H``, or ``B``, and a
779mode modifier of ``ABS`` or ``IND``.  The 'dst_reg' and 'offset' fields were
780set to zero, and 'src_reg' was set to zero for ``ABS``.  However, these
781instructions are deprecated and SHOULD no longer be used.  All legacy packet
782access instructions belong to the "packet" conformance group.
783