8 BTF (BPF Type Format) is the metadata format which encodes the debug info
9 related to BPF program/map. The name BTF was used initially to describe data
10 types. The BTF was later extended to include function info for defined
11 subroutines, and line info for source/line information.
13 The debug info is used for map pretty print, function signature, etc. The
14 function signature enables better bpf program/function kernel symbol. The line
15 info helps generate source annotated translated byte code, jited code and
18 The BTF specification contains two parts,
22 The kernel API is the contract between user space and kernel. The kernel
23 verifies the BTF info before using it. The ELF file format is a user space
24 contract between ELF file and libbpf loader.
26 The type and string sections are part of the BTF kernel API, describing the
27 debug info (mostly types related) referenced by the bpf program. These two
28 sections are discussed in details in :ref:`BTF_Type_String`.
32 2. BTF Type and String Encoding
33 ===============================
35 The file ``include/uapi/linux/btf.h`` provides high-level definition of how
36 types/strings are encoded.
38 The beginning of data blob must be::
46 /* All offsets are in bytes relative to the end of this header */
47 __u32 type_off; /* offset of type section */
48 __u32 type_len; /* length of type section */
49 __u32 str_off; /* offset of string section */
50 __u32 str_len; /* length of string section */
53 The magic is ``0xeB9F``, which has different encoding for big and little
54 endian systems, and can be used to test whether BTF is generated for big- or
55 little-endian target. The ``btf_header`` is designed to be extensible with
56 ``hdr_len`` equal to ``sizeof(struct btf_header)`` when a data blob is
62 The first string in the string section must be a null string. The rest of
63 string table is a concatenation of other null-terminated strings.
68 The type id ``0`` is reserved for ``void`` type. The type section is parsed
69 sequentially and type id is assigned to each recognized type starting from id
70 ``1``. Currently, the following types are supported::
72 #define BTF_KIND_INT 1 /* Integer */
73 #define BTF_KIND_PTR 2 /* Pointer */
74 #define BTF_KIND_ARRAY 3 /* Array */
75 #define BTF_KIND_STRUCT 4 /* Struct */
76 #define BTF_KIND_UNION 5 /* Union */
77 #define BTF_KIND_ENUM 6 /* Enumeration */
78 #define BTF_KIND_FWD 7 /* Forward */
79 #define BTF_KIND_TYPEDEF 8 /* Typedef */
80 #define BTF_KIND_VOLATILE 9 /* Volatile */
81 #define BTF_KIND_CONST 10 /* Const */
82 #define BTF_KIND_RESTRICT 11 /* Restrict */
83 #define BTF_KIND_FUNC 12 /* Function */
84 #define BTF_KIND_FUNC_PROTO 13 /* Function Proto */
85 #define BTF_KIND_VAR 14 /* Variable */
86 #define BTF_KIND_DATASEC 15 /* Section */
87 #define BTF_KIND_FLOAT 16 /* Floating point */
88 #define BTF_KIND_DECL_TAG 17 /* Decl Tag */
89 #define BTF_KIND_TYPE_TAG 18 /* Type Tag */
91 Note that the type section encodes debug info, not just pure types.
92 ``BTF_KIND_FUNC`` is not a type, and it represents a defined subprogram.
94 Each type contains the following common data::
98 /* "info" bits arrangement
99 * bits 0-15: vlen (e.g. # of struct's members)
101 * bits 24-28: kind (e.g. int, ptr, array...etc)
103 * bit 31: kind_flag, currently used by
104 * struct, union and fwd
107 /* "size" is used by INT, ENUM, STRUCT and UNION.
108 * "size" tells the size of the type it is describing.
110 * "type" is used by PTR, TYPEDEF, VOLATILE, CONST, RESTRICT,
111 * FUNC, FUNC_PROTO, DECL_TAG and TYPE_TAG.
112 * "type" is a type_id referring to another type.
120 For certain kinds, the common data are followed by kind-specific data. The
121 ``name_off`` in ``struct btf_type`` specifies the offset in the string table.
122 The following sections detail encoding of each kind.
127 ``struct btf_type`` encoding requirement:
128 * ``name_off``: any valid offset
129 * ``info.kind_flag``: 0
130 * ``info.kind``: BTF_KIND_INT
132 * ``size``: the size of the int type in bytes.
134 ``btf_type`` is followed by a ``u32`` with the following bits arrangement::
136 #define BTF_INT_ENCODING(VAL) (((VAL) & 0x0f000000) >> 24)
137 #define BTF_INT_OFFSET(VAL) (((VAL) & 0x00ff0000) >> 16)
138 #define BTF_INT_BITS(VAL) ((VAL) & 0x000000ff)
140 The ``BTF_INT_ENCODING`` has the following attributes::
142 #define BTF_INT_SIGNED (1 << 0)
143 #define BTF_INT_CHAR (1 << 1)
144 #define BTF_INT_BOOL (1 << 2)
146 The ``BTF_INT_ENCODING()`` provides extra information: signedness, char, or
147 bool, for the int type. The char and bool encoding are mostly useful for
148 pretty print. At most one encoding can be specified for the int type.
150 The ``BTF_INT_BITS()`` specifies the number of actual bits held by this int
151 type. For example, a 4-bit bitfield encodes ``BTF_INT_BITS()`` equals to 4.
152 The ``btf_type.size * 8`` must be equal to or greater than ``BTF_INT_BITS()``
153 for the type. The maximum value of ``BTF_INT_BITS()`` is 128.
155 The ``BTF_INT_OFFSET()`` specifies the starting bit offset to calculate values
156 for this int. For example, a bitfield struct member has:
158 * btf member bit offset 100 from the start of the structure,
159 * btf member pointing to an int type,
160 * the int type has ``BTF_INT_OFFSET() = 2`` and ``BTF_INT_BITS() = 4``
162 Then in the struct memory layout, this member will occupy ``4`` bits starting
163 from bits ``100 + 2 = 102``.
165 Alternatively, the bitfield struct member can be the following to access the
166 same bits as the above:
168 * btf member bit offset 102,
169 * btf member pointing to an int type,
170 * the int type has ``BTF_INT_OFFSET() = 0`` and ``BTF_INT_BITS() = 4``
172 The original intention of ``BTF_INT_OFFSET()`` is to provide flexibility of
173 bitfield encoding. Currently, both llvm and pahole generate
174 ``BTF_INT_OFFSET() = 0`` for all int types.
179 ``struct btf_type`` encoding requirement:
181 * ``info.kind_flag``: 0
182 * ``info.kind``: BTF_KIND_PTR
184 * ``type``: the pointee type of the pointer
186 No additional type data follow ``btf_type``.
191 ``struct btf_type`` encoding requirement:
193 * ``info.kind_flag``: 0
194 * ``info.kind``: BTF_KIND_ARRAY
196 * ``size/type``: 0, not used
198 ``btf_type`` is followed by one ``struct btf_array``::
206 The ``struct btf_array`` encoding:
207 * ``type``: the element type
208 * ``index_type``: the index type
209 * ``nelems``: the number of elements for this array (``0`` is also allowed).
211 The ``index_type`` can be any regular int type (``u8``, ``u16``, ``u32``,
212 ``u64``, ``unsigned __int128``). The original design of including
213 ``index_type`` follows DWARF, which has an ``index_type`` for its array type.
214 Currently in BTF, beyond type verification, the ``index_type`` is not used.
216 The ``struct btf_array`` allows chaining through element type to represent
217 multidimensional arrays. For example, for ``int a[5][6]``, the following type
218 information illustrates the chaining:
221 * [2]: array, ``btf_array.type = [1]``, ``btf_array.nelems = 6``
222 * [3]: array, ``btf_array.type = [2]``, ``btf_array.nelems = 5``
224 Currently, both pahole and llvm collapse multidimensional array into
225 one-dimensional array, e.g., for ``a[5][6]``, the ``btf_array.nelems`` is
226 equal to ``30``. This is because the original use case is map pretty print
227 where the whole array is dumped out so one-dimensional array is enough. As
228 more BTF usage is explored, pahole and llvm can be changed to generate proper
229 chained representation for multidimensional arrays.
231 2.2.4 BTF_KIND_STRUCT
232 ~~~~~~~~~~~~~~~~~~~~~
236 ``struct btf_type`` encoding requirement:
237 * ``name_off``: 0 or offset to a valid C identifier
238 * ``info.kind_flag``: 0 or 1
239 * ``info.kind``: BTF_KIND_STRUCT or BTF_KIND_UNION
240 * ``info.vlen``: the number of struct/union members
241 * ``info.size``: the size of the struct/union in bytes
243 ``btf_type`` is followed by ``info.vlen`` number of ``struct btf_member``.::
251 ``struct btf_member`` encoding:
252 * ``name_off``: offset to a valid C identifier
253 * ``type``: the member type
254 * ``offset``: <see below>
256 If the type info ``kind_flag`` is not set, the offset contains only bit offset
257 of the member. Note that the base type of the bitfield can only be int or enum
258 type. If the bitfield size is 32, the base type can be either int or enum
259 type. If the bitfield size is not 32, the base type must be int, and int type
260 ``BTF_INT_BITS()`` encodes the bitfield size.
262 If the ``kind_flag`` is set, the ``btf_member.offset`` contains both member
263 bitfield size and bit offset. The bitfield size and bit offset are calculated
266 #define BTF_MEMBER_BITFIELD_SIZE(val) ((val) >> 24)
267 #define BTF_MEMBER_BIT_OFFSET(val) ((val) & 0xffffff)
269 In this case, if the base type is an int type, it must be a regular int type:
271 * ``BTF_INT_OFFSET()`` must be 0.
272 * ``BTF_INT_BITS()`` must be equal to ``{1,2,4,8,16} * 8``.
274 The following kernel patch introduced ``kind_flag`` and explained why both
277 https://github.com/torvalds/linux/commit/9d5f9f701b1891466fb3dbb1806ad97716f95cc3#diff-fa650a64fdd3968396883d2fe8215ff3
282 ``struct btf_type`` encoding requirement:
283 * ``name_off``: 0 or offset to a valid C identifier
284 * ``info.kind_flag``: 0
285 * ``info.kind``: BTF_KIND_ENUM
286 * ``info.vlen``: number of enum values
289 ``btf_type`` is followed by ``info.vlen`` number of ``struct btf_enum``.::
296 The ``btf_enum`` encoding:
297 * ``name_off``: offset to a valid C identifier
303 ``struct btf_type`` encoding requirement:
304 * ``name_off``: offset to a valid C identifier
305 * ``info.kind_flag``: 0 for struct, 1 for union
306 * ``info.kind``: BTF_KIND_FWD
310 No additional type data follow ``btf_type``.
312 2.2.8 BTF_KIND_TYPEDEF
313 ~~~~~~~~~~~~~~~~~~~~~~
315 ``struct btf_type`` encoding requirement:
316 * ``name_off``: offset to a valid C identifier
317 * ``info.kind_flag``: 0
318 * ``info.kind``: BTF_KIND_TYPEDEF
320 * ``type``: the type which can be referred by name at ``name_off``
322 No additional type data follow ``btf_type``.
324 2.2.9 BTF_KIND_VOLATILE
325 ~~~~~~~~~~~~~~~~~~~~~~~
327 ``struct btf_type`` encoding requirement:
329 * ``info.kind_flag``: 0
330 * ``info.kind``: BTF_KIND_VOLATILE
332 * ``type``: the type with ``volatile`` qualifier
334 No additional type data follow ``btf_type``.
336 2.2.10 BTF_KIND_CONST
337 ~~~~~~~~~~~~~~~~~~~~~
339 ``struct btf_type`` encoding requirement:
341 * ``info.kind_flag``: 0
342 * ``info.kind``: BTF_KIND_CONST
344 * ``type``: the type with ``const`` qualifier
346 No additional type data follow ``btf_type``.
348 2.2.11 BTF_KIND_RESTRICT
349 ~~~~~~~~~~~~~~~~~~~~~~~~
351 ``struct btf_type`` encoding requirement:
353 * ``info.kind_flag``: 0
354 * ``info.kind``: BTF_KIND_RESTRICT
356 * ``type``: the type with ``restrict`` qualifier
358 No additional type data follow ``btf_type``.
363 ``struct btf_type`` encoding requirement:
364 * ``name_off``: offset to a valid C identifier
365 * ``info.kind_flag``: 0
366 * ``info.kind``: BTF_KIND_FUNC
368 * ``type``: a BTF_KIND_FUNC_PROTO type
370 No additional type data follow ``btf_type``.
372 A BTF_KIND_FUNC defines not a type, but a subprogram (function) whose
373 signature is defined by ``type``. The subprogram is thus an instance of that
374 type. The BTF_KIND_FUNC may in turn be referenced by a func_info in the
375 :ref:`BTF_Ext_Section` (ELF) or in the arguments to :ref:`BPF_Prog_Load`
378 2.2.13 BTF_KIND_FUNC_PROTO
379 ~~~~~~~~~~~~~~~~~~~~~~~~~~
381 ``struct btf_type`` encoding requirement:
383 * ``info.kind_flag``: 0
384 * ``info.kind``: BTF_KIND_FUNC_PROTO
385 * ``info.vlen``: # of parameters
386 * ``type``: the return type
388 ``btf_type`` is followed by ``info.vlen`` number of ``struct btf_param``.::
395 If a BTF_KIND_FUNC_PROTO type is referred by a BTF_KIND_FUNC type, then
396 ``btf_param.name_off`` must point to a valid C identifier except for the
397 possible last argument representing the variable argument. The btf_param.type
398 refers to parameter type.
400 If the function has variable arguments, the last parameter is encoded with
401 ``name_off = 0`` and ``type = 0``.
406 ``struct btf_type`` encoding requirement:
407 * ``name_off``: offset to a valid C identifier
408 * ``info.kind_flag``: 0
409 * ``info.kind``: BTF_KIND_VAR
411 * ``type``: the type of the variable
413 ``btf_type`` is followed by a single ``struct btf_variable`` with the
420 ``struct btf_var`` encoding:
421 * ``linkage``: currently only static variable 0, or globally allocated
422 variable in ELF sections 1
424 Not all type of global variables are supported by LLVM at this point.
425 The following is currently available:
427 * static variables with or without section attributes
428 * global variables with section attributes
430 The latter is for future extraction of map key/value type id's from a
433 2.2.15 BTF_KIND_DATASEC
434 ~~~~~~~~~~~~~~~~~~~~~~~
436 ``struct btf_type`` encoding requirement:
437 * ``name_off``: offset to a valid name associated with a variable or
438 one of .data/.bss/.rodata
439 * ``info.kind_flag``: 0
440 * ``info.kind``: BTF_KIND_DATASEC
441 * ``info.vlen``: # of variables
442 * ``size``: total section size in bytes (0 at compilation time, patched
443 to actual size by BPF loaders such as libbpf)
445 ``btf_type`` is followed by ``info.vlen`` number of ``struct btf_var_secinfo``.::
447 struct btf_var_secinfo {
453 ``struct btf_var_secinfo`` encoding:
454 * ``type``: the type of the BTF_KIND_VAR variable
455 * ``offset``: the in-section offset of the variable
456 * ``size``: the size of the variable in bytes
458 2.2.16 BTF_KIND_FLOAT
459 ~~~~~~~~~~~~~~~~~~~~~
461 ``struct btf_type`` encoding requirement:
462 * ``name_off``: any valid offset
463 * ``info.kind_flag``: 0
464 * ``info.kind``: BTF_KIND_FLOAT
466 * ``size``: the size of the float type in bytes: 2, 4, 8, 12 or 16.
468 No additional type data follow ``btf_type``.
470 2.2.17 BTF_KIND_DECL_TAG
471 ~~~~~~~~~~~~~~~~~~~~~~~~
473 ``struct btf_type`` encoding requirement:
474 * ``name_off``: offset to a non-empty string
475 * ``info.kind_flag``: 0
476 * ``info.kind``: BTF_KIND_DECL_TAG
478 * ``type``: ``struct``, ``union``, ``func``, ``var`` or ``typedef``
480 ``btf_type`` is followed by ``struct btf_decl_tag``.::
482 struct btf_decl_tag {
486 The ``name_off`` encodes btf_decl_tag attribute string.
487 The ``type`` should be ``struct``, ``union``, ``func``, ``var`` or ``typedef``.
488 For ``var`` or ``typedef`` type, ``btf_decl_tag.component_idx`` must be ``-1``.
489 For the other three types, if the btf_decl_tag attribute is
490 applied to the ``struct``, ``union`` or ``func`` itself,
491 ``btf_decl_tag.component_idx`` must be ``-1``. Otherwise,
492 the attribute is applied to a ``struct``/``union`` member or
493 a ``func`` argument, and ``btf_decl_tag.component_idx`` should be a
494 valid index (starting from 0) pointing to a member or an argument.
496 2.2.17 BTF_KIND_TYPE_TAG
497 ~~~~~~~~~~~~~~~~~~~~~~~~
499 ``struct btf_type`` encoding requirement:
500 * ``name_off``: offset to a non-empty string
501 * ``info.kind_flag``: 0
502 * ``info.kind``: BTF_KIND_TYPE_TAG
504 * ``type``: the type with ``btf_type_tag`` attribute
509 The following bpf syscall command involves BTF:
510 * BPF_BTF_LOAD: load a blob of BTF data into kernel
511 * BPF_MAP_CREATE: map creation with btf key and value type info.
512 * BPF_PROG_LOAD: prog load with btf function and line info.
513 * BPF_BTF_GET_FD_BY_ID: get a btf fd
514 * BPF_OBJ_GET_INFO_BY_FD: btf, func_info, line_info
515 and other btf related info are returned.
517 The workflow typically looks like:
524 BPF_MAP_CREATE and BPF_PROG_LOAD
531 BPF_{PROG,MAP}_GET_NEXT_ID (get prog/map id's)
534 BPF_{PROG,MAP}_GET_FD_BY_ID (get a prog/map fd)
537 BPF_OBJ_GET_INFO_BY_FD (get bpf_prog_info/bpf_map_info with btf_id)
540 BPF_BTF_GET_FD_BY_ID (get btf_fd) |
543 BPF_OBJ_GET_INFO_BY_FD (get btf) |
546 pretty print types, dump func signatures and line info, etc.
552 Load a blob of BTF data into kernel. A blob of data, described in
553 :ref:`BTF_Type_String`, can be directly loaded into the kernel. A ``btf_fd``
554 is returned to a userspace.
559 A map can be created with ``btf_fd`` and specified key/value type id.::
561 __u32 btf_fd; /* fd pointing to a BTF type data */
562 __u32 btf_key_type_id; /* BTF type_id of the key */
563 __u32 btf_value_type_id; /* BTF type_id of the value */
565 In libbpf, the map can be defined with extra annotation like below:
568 struct bpf_map_def SEC("maps") btf_map = {
569 .type = BPF_MAP_TYPE_ARRAY,
570 .key_size = sizeof(int),
571 .value_size = sizeof(struct ipv_counts),
574 BPF_ANNOTATE_KV_PAIR(btf_map, int, struct ipv_counts);
576 Here, the parameters for macro BPF_ANNOTATE_KV_PAIR are map name, key and
577 value types for the map. During ELF parsing, libbpf is able to extract
578 key/value type_id's and assign them to BPF_MAP_CREATE attributes
586 During prog_load, func_info and line_info can be passed to kernel with proper
587 values for the following attributes:
593 __u32 prog_btf_fd; /* fd pointing to BTF type data */
594 __u32 func_info_rec_size; /* userspace bpf_func_info size */
595 __aligned_u64 func_info; /* func info */
596 __u32 func_info_cnt; /* number of bpf_func_info records */
597 __u32 line_info_rec_size; /* userspace bpf_line_info size */
598 __aligned_u64 line_info; /* line info */
599 __u32 line_info_cnt; /* number of bpf_line_info records */
601 The func_info and line_info are an array of below, respectively.::
603 struct bpf_func_info {
604 __u32 insn_off; /* [0, insn_cnt - 1] */
605 __u32 type_id; /* pointing to a BTF_KIND_FUNC type */
607 struct bpf_line_info {
608 __u32 insn_off; /* [0, insn_cnt - 1] */
609 __u32 file_name_off; /* offset to string table for the filename */
610 __u32 line_off; /* offset to string table for the source line */
611 __u32 line_col; /* line number and column number */
614 func_info_rec_size is the size of each func_info record, and
615 line_info_rec_size is the size of each line_info record. Passing the record
616 size to kernel make it possible to extend the record itself in the future.
618 Below are requirements for func_info:
619 * func_info[0].insn_off must be 0.
620 * the func_info insn_off is in strictly increasing order and matches
623 Below are requirements for line_info:
624 * the first insn in each func must have a line_info record pointing to it.
625 * the line_info insn_off is in strictly increasing order.
627 For line_info, the line number and column number are defined as below:
630 #define BPF_LINE_INFO_LINE_NUM(line_col) ((line_col) >> 10)
631 #define BPF_LINE_INFO_LINE_COL(line_col) ((line_col) & 0x3ff)
633 3.4 BPF_{PROG,MAP}_GET_NEXT_ID
634 ------------------------------
636 In kernel, every loaded program, map or btf has a unique id. The id won't
637 change during the lifetime of a program, map, or btf.
639 The bpf syscall command BPF_{PROG,MAP}_GET_NEXT_ID returns all id's, one for
640 each command, to user space, for bpf program or maps, respectively, so an
641 inspection tool can inspect all programs and maps.
643 3.5 BPF_{PROG,MAP}_GET_FD_BY_ID
644 -------------------------------
646 An introspection tool cannot use id to get details about program or maps.
647 A file descriptor needs to be obtained first for reference-counting purpose.
649 3.6 BPF_OBJ_GET_INFO_BY_FD
650 --------------------------
652 Once a program/map fd is acquired, an introspection tool can get the detailed
653 information from kernel about this fd, some of which are BTF-related. For
654 example, ``bpf_map_info`` returns ``btf_id`` and key/value type ids.
655 ``bpf_prog_info`` returns ``btf_id``, func_info, and line info for translated
656 bpf byte codes, and jited_line_info.
658 3.7 BPF_BTF_GET_FD_BY_ID
659 ------------------------
661 With ``btf_id`` obtained in ``bpf_map_info`` and ``bpf_prog_info``, bpf
662 syscall command BPF_BTF_GET_FD_BY_ID can retrieve a btf fd. Then, with
663 command BPF_OBJ_GET_INFO_BY_FD, the btf blob, originally loaded into the
664 kernel with BPF_BTF_LOAD, can be retrieved.
666 With the btf blob, ``bpf_map_info``, and ``bpf_prog_info``, an introspection
667 tool has full btf knowledge and is able to pretty print map key/values, dump
668 func signatures and line info, along with byte/jit codes.
670 4. ELF File Format Interface
671 ============================
676 The .BTF section contains type and string data. The format of this section is
677 same as the one describe in :ref:`BTF_Type_String`.
684 The .BTF.ext section encodes func_info and line_info which needs loader
685 manipulation before loading into the kernel.
687 The specification for .BTF.ext section is defined at ``tools/lib/bpf/btf.h``
688 and ``tools/lib/bpf/btf.c``.
690 The current header of .BTF.ext section::
692 struct btf_ext_header {
698 /* All offsets are in bytes relative to the end of this header */
705 It is very similar to .BTF section. Instead of type/string section, it
706 contains func_info and line_info section. See :ref:`BPF_Prog_Load` for details
707 about func_info and line_info record format.
709 The func_info is organized as below.::
712 btf_ext_info_sec for section #1 /* func_info for section #1 */
713 btf_ext_info_sec for section #2 /* func_info for section #2 */
716 ``func_info_rec_size`` specifies the size of ``bpf_func_info`` structure when
717 .BTF.ext is generated. ``btf_ext_info_sec``, defined below, is a collection of
718 func_info for each specific ELF section.::
720 struct btf_ext_info_sec {
721 __u32 sec_name_off; /* offset to section name */
723 /* Followed by num_info * record_size number of bytes */
727 Here, num_info must be greater than 0.
729 The line_info is organized as below.::
732 btf_ext_info_sec for section #1 /* line_info for section #1 */
733 btf_ext_info_sec for section #2 /* line_info for section #2 */
736 ``line_info_rec_size`` specifies the size of ``bpf_line_info`` structure when
737 .BTF.ext is generated.
739 The interpretation of ``bpf_func_info->insn_off`` and
740 ``bpf_line_info->insn_off`` is different between kernel API and ELF API. For
741 kernel API, the ``insn_off`` is the instruction offset in the unit of ``struct
742 bpf_insn``. For ELF API, the ``insn_off`` is the byte offset from the
743 beginning of section (``btf_ext_info_sec->sec_name_off``).
748 The .BTF_ids section encodes BTF ID values that are used within the kernel.
750 This section is created during the kernel compilation with the help of
751 macros defined in ``include/linux/btf_ids.h`` header file. Kernel code can
752 use them to create lists and sets (sorted lists) of BTF ID values.
754 The ``BTF_ID_LIST`` and ``BTF_ID`` macros define unsorted list of BTF ID values,
755 with following syntax::
761 resulting in following layout in .BTF_ids section::
763 __BTF_ID__type1__name1__1:
765 __BTF_ID__type2__name2__2:
768 The ``u32 list[];`` variable is defined to access the list.
770 The ``BTF_ID_UNUSED`` macro defines 4 zero bytes. It's used when we
771 want to define unused entry in BTF_ID_LIST, like::
773 BTF_ID_LIST(bpf_skb_output_btf_ids)
774 BTF_ID(struct, sk_buff)
776 BTF_ID(struct, task_struct)
778 The ``BTF_SET_START/END`` macros pair defines sorted list of BTF ID values
779 and their count, with following syntax::
786 resulting in following layout in .BTF_ids section::
790 __BTF_ID__type1__name1__3:
792 __BTF_ID__type2__name2__4:
795 The ``struct btf_id_set set;`` variable is defined to access the list.
797 The ``typeX`` name can be one of following::
799 struct, union, typedef, func
801 and is used as a filter when resolving the BTF ID value.
803 All the BTF ID lists and sets are compiled in the .BTF_ids section and
804 resolved during the linking phase of kernel build by ``resolve_btfids`` tool.
809 5.1 bpftool map pretty print
810 ----------------------------
812 With BTF, the map key/value can be printed based on fields rather than simply
813 raw bytes. This is especially valuable for large structure or if your data
814 structure has bitfields. For example, for the following map,::
816 enum A { A1, A2, A3, A4, A5 };
827 struct bpf_map_def SEC("maps") tmpmap = {
828 .type = BPF_MAP_TYPE_ARRAY,
829 .key_size = sizeof(__u32),
830 .value_size = sizeof(struct tmp_t),
833 BPF_ANNOTATE_KV_PAIR(tmpmap, int, struct tmp_t);
835 bpftool is able to pretty print like below:
851 5.2 bpftool prog dump
852 ---------------------
854 The following is an example showing how func_info and line_info can help prog
855 dump with better kernel symbol names, function prototypes and line
858 $ bpftool prog dump jited pinned /sys/fs/bpf/test_btf_haskv
860 int test_long_fname_2(struct dummy_tracepoint_args * arg):
861 bpf_prog_44a040bf25481309_test_long_fname_2:
862 ; static int test_long_fname_2(struct dummy_tracepoint_args *arg)
867 f: mov %rbx,0x0(%rbp)
868 13: mov %r13,0x8(%rbp)
869 17: mov %r14,0x10(%rbp)
870 1b: mov %r15,0x18(%rbp)
872 21: mov %rax,0x20(%rbp)
875 27: mov %esi,-0x4(%rbp)
877 2a: mov 0x8(%rdi),%rdi
880 32: je 0x0000000000000070
882 ; counts = bpf_map_lookup_elem(&btf_map, &key);
888 The following is an example of how line_info can help debugging verification
891 /* The code at tools/testing/selftests/bpf/test_xdp_noinline.c
892 * is modified as below.
894 data = (void *)(long)xdp->data;
895 data_end = (void *)(long)xdp->data_end;
897 if (data + 4 > data_end)
900 *(u32 *)data = dst->dst;
902 $ bpftool prog load ./test_xdp_noinline.o /sys/fs/bpf/test_xdp_noinline type xdp
903 ; data = (void *)(long)xdp->data;
904 224: (79) r2 = *(u64 *)(r10 -112)
905 225: (61) r2 = *(u32 *)(r2 +0)
906 ; *(u32 *)data = dst->dst;
907 226: (63) *(u32 *)(r2 +0) = r1
908 invalid access to packet, off=0 size=4, R2(id=0,off=0,r=0)
909 R2 offset is outside of the packet
914 You need latest pahole
916 https://git.kernel.org/pub/scm/devel/pahole/pahole.git/
918 or llvm (8.0 or later). The pahole acts as a dwarf2btf converter. It doesn't
919 support .BTF.ext and btf BTF_KIND_FUNC type yet. For example,::
927 -bash-4.4$ gcc -c -O2 -g t.c
928 -bash-4.4$ pahole -JV t.o
930 [1] STRUCT t kind_flag=1 size=4 vlen=3
931 a type_id=2 bitfield_size=2 bits_offset=0
932 b type_id=2 bitfield_size=3 bits_offset=2
933 c type_id=2 bitfield_size=2 bits_offset=5
934 [2] INT int size=4 bit_offset=0 nr_bits=32 encoding=SIGNED
936 The llvm is able to generate .BTF and .BTF.ext directly with -g for bpf target
937 only. The assembly code (-S) is able to show the BTF encoding in assembly
944 int (*f2)(char q1, __int32 q2, ...);
947 int main() { return 0; }
948 int test() { return 0; }
949 -bash-4.4$ clang -c -g -O2 -target bpf t2.c
950 -bash-4.4$ readelf -S t2.o
952 [ 8] .BTF PROGBITS 0000000000000000 00000247
953 000000000000016e 0000000000000000 0 0 1
954 [ 9] .BTF.ext PROGBITS 0000000000000000 000003b5
955 0000000000000060 0000000000000000 0 0 1
956 [10] .rel.BTF.ext REL 0000000000000000 000007e0
957 0000000000000040 0000000000000010 16 9 8
959 -bash-4.4$ clang -S -g -O2 -target bpf t2.c
962 .section .BTF,"",@progbits
963 .short 60319 # 0xeb9f
971 .long 0 # BTF_KIND_FUNC_PROTO(id = 1)
972 .long 218103808 # 0xd000000
974 .long 83 # BTF_KIND_INT(id = 2)
975 .long 16777216 # 0x1000000
977 .long 16777248 # 0x1000020
979 .byte 0 # string offset=0
980 .ascii ".text" # string offset=1
982 .ascii "/home/yhs/tmp-pahole/t2.c" # string offset=7
984 .ascii "int main() { return 0; }" # string offset=33
986 .ascii "int test() { return 0; }" # string offset=58
988 .ascii "int" # string offset=83
990 .section .BTF.ext,"",@progbits
991 .short 60319 # 0xeb9f
1000 .long 1 # FuncInfo section string offset=1
1007 .long 1 # LineInfo section string offset=1
1012 .long 7182 # Line 7 Col 14
1016 .long 8206 # Line 8 Col 14
1021 Kernel bpf selftest `test_btf.c` provides extensive set of BTF-related tests.