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 */
89 Note that the type section encodes debug info, not just pure types.
90 ``BTF_KIND_FUNC`` is not a type, and it represents a defined subprogram.
92 Each type contains the following common data::
96 /* "info" bits arrangement
97 * bits 0-15: vlen (e.g. # of struct's members)
99 * bits 24-28: kind (e.g. int, ptr, array...etc)
101 * bit 31: kind_flag, currently used by
102 * struct, union and fwd
105 /* "size" is used by INT, ENUM, STRUCT and UNION.
106 * "size" tells the size of the type it is describing.
108 * "type" is used by PTR, TYPEDEF, VOLATILE, CONST, RESTRICT,
109 * FUNC and FUNC_PROTO.
110 * "type" is a type_id referring to another type.
118 For certain kinds, the common data are followed by kind-specific data. The
119 ``name_off`` in ``struct btf_type`` specifies the offset in the string table.
120 The following sections detail encoding of each kind.
125 ``struct btf_type`` encoding requirement:
126 * ``name_off``: any valid offset
127 * ``info.kind_flag``: 0
128 * ``info.kind``: BTF_KIND_INT
130 * ``size``: the size of the int type in bytes.
132 ``btf_type`` is followed by a ``u32`` with the following bits arrangement::
134 #define BTF_INT_ENCODING(VAL) (((VAL) & 0x0f000000) >> 24)
135 #define BTF_INT_OFFSET(VAL) (((VAL) & 0x00ff0000) >> 16)
136 #define BTF_INT_BITS(VAL) ((VAL) & 0x000000ff)
138 The ``BTF_INT_ENCODING`` has the following attributes::
140 #define BTF_INT_SIGNED (1 << 0)
141 #define BTF_INT_CHAR (1 << 1)
142 #define BTF_INT_BOOL (1 << 2)
144 The ``BTF_INT_ENCODING()`` provides extra information: signedness, char, or
145 bool, for the int type. The char and bool encoding are mostly useful for
146 pretty print. At most one encoding can be specified for the int type.
148 The ``BTF_INT_BITS()`` specifies the number of actual bits held by this int
149 type. For example, a 4-bit bitfield encodes ``BTF_INT_BITS()`` equals to 4.
150 The ``btf_type.size * 8`` must be equal to or greater than ``BTF_INT_BITS()``
151 for the type. The maximum value of ``BTF_INT_BITS()`` is 128.
153 The ``BTF_INT_OFFSET()`` specifies the starting bit offset to calculate values
154 for this int. For example, a bitfield struct member has:
156 * btf member bit offset 100 from the start of the structure,
157 * btf member pointing to an int type,
158 * the int type has ``BTF_INT_OFFSET() = 2`` and ``BTF_INT_BITS() = 4``
160 Then in the struct memory layout, this member will occupy ``4`` bits starting
161 from bits ``100 + 2 = 102``.
163 Alternatively, the bitfield struct member can be the following to access the
164 same bits as the above:
166 * btf member bit offset 102,
167 * btf member pointing to an int type,
168 * the int type has ``BTF_INT_OFFSET() = 0`` and ``BTF_INT_BITS() = 4``
170 The original intention of ``BTF_INT_OFFSET()`` is to provide flexibility of
171 bitfield encoding. Currently, both llvm and pahole generate
172 ``BTF_INT_OFFSET() = 0`` for all int types.
177 ``struct btf_type`` encoding requirement:
179 * ``info.kind_flag``: 0
180 * ``info.kind``: BTF_KIND_PTR
182 * ``type``: the pointee type of the pointer
184 No additional type data follow ``btf_type``.
189 ``struct btf_type`` encoding requirement:
191 * ``info.kind_flag``: 0
192 * ``info.kind``: BTF_KIND_ARRAY
194 * ``size/type``: 0, not used
196 ``btf_type`` is followed by one ``struct btf_array``::
204 The ``struct btf_array`` encoding:
205 * ``type``: the element type
206 * ``index_type``: the index type
207 * ``nelems``: the number of elements for this array (``0`` is also allowed).
209 The ``index_type`` can be any regular int type (``u8``, ``u16``, ``u32``,
210 ``u64``, ``unsigned __int128``). The original design of including
211 ``index_type`` follows DWARF, which has an ``index_type`` for its array type.
212 Currently in BTF, beyond type verification, the ``index_type`` is not used.
214 The ``struct btf_array`` allows chaining through element type to represent
215 multidimensional arrays. For example, for ``int a[5][6]``, the following type
216 information illustrates the chaining:
219 * [2]: array, ``btf_array.type = [1]``, ``btf_array.nelems = 6``
220 * [3]: array, ``btf_array.type = [2]``, ``btf_array.nelems = 5``
222 Currently, both pahole and llvm collapse multidimensional array into
223 one-dimensional array, e.g., for ``a[5][6]``, the ``btf_array.nelems`` is
224 equal to ``30``. This is because the original use case is map pretty print
225 where the whole array is dumped out so one-dimensional array is enough. As
226 more BTF usage is explored, pahole and llvm can be changed to generate proper
227 chained representation for multidimensional arrays.
229 2.2.4 BTF_KIND_STRUCT
230 ~~~~~~~~~~~~~~~~~~~~~
234 ``struct btf_type`` encoding requirement:
235 * ``name_off``: 0 or offset to a valid C identifier
236 * ``info.kind_flag``: 0 or 1
237 * ``info.kind``: BTF_KIND_STRUCT or BTF_KIND_UNION
238 * ``info.vlen``: the number of struct/union members
239 * ``info.size``: the size of the struct/union in bytes
241 ``btf_type`` is followed by ``info.vlen`` number of ``struct btf_member``.::
249 ``struct btf_member`` encoding:
250 * ``name_off``: offset to a valid C identifier
251 * ``type``: the member type
252 * ``offset``: <see below>
254 If the type info ``kind_flag`` is not set, the offset contains only bit offset
255 of the member. Note that the base type of the bitfield can only be int or enum
256 type. If the bitfield size is 32, the base type can be either int or enum
257 type. If the bitfield size is not 32, the base type must be int, and int type
258 ``BTF_INT_BITS()`` encodes the bitfield size.
260 If the ``kind_flag`` is set, the ``btf_member.offset`` contains both member
261 bitfield size and bit offset. The bitfield size and bit offset are calculated
264 #define BTF_MEMBER_BITFIELD_SIZE(val) ((val) >> 24)
265 #define BTF_MEMBER_BIT_OFFSET(val) ((val) & 0xffffff)
267 In this case, if the base type is an int type, it must be a regular int type:
269 * ``BTF_INT_OFFSET()`` must be 0.
270 * ``BTF_INT_BITS()`` must be equal to ``{1,2,4,8,16} * 8``.
272 The following kernel patch introduced ``kind_flag`` and explained why both
275 https://github.com/torvalds/linux/commit/9d5f9f701b1891466fb3dbb1806ad97716f95cc3#diff-fa650a64fdd3968396883d2fe8215ff3
280 ``struct btf_type`` encoding requirement:
281 * ``name_off``: 0 or offset to a valid C identifier
282 * ``info.kind_flag``: 0
283 * ``info.kind``: BTF_KIND_ENUM
284 * ``info.vlen``: number of enum values
287 ``btf_type`` is followed by ``info.vlen`` number of ``struct btf_enum``.::
294 The ``btf_enum`` encoding:
295 * ``name_off``: offset to a valid C identifier
301 ``struct btf_type`` encoding requirement:
302 * ``name_off``: offset to a valid C identifier
303 * ``info.kind_flag``: 0 for struct, 1 for union
304 * ``info.kind``: BTF_KIND_FWD
308 No additional type data follow ``btf_type``.
310 2.2.8 BTF_KIND_TYPEDEF
311 ~~~~~~~~~~~~~~~~~~~~~~
313 ``struct btf_type`` encoding requirement:
314 * ``name_off``: offset to a valid C identifier
315 * ``info.kind_flag``: 0
316 * ``info.kind``: BTF_KIND_TYPEDEF
318 * ``type``: the type which can be referred by name at ``name_off``
320 No additional type data follow ``btf_type``.
322 2.2.9 BTF_KIND_VOLATILE
323 ~~~~~~~~~~~~~~~~~~~~~~~
325 ``struct btf_type`` encoding requirement:
327 * ``info.kind_flag``: 0
328 * ``info.kind``: BTF_KIND_VOLATILE
330 * ``type``: the type with ``volatile`` qualifier
332 No additional type data follow ``btf_type``.
334 2.2.10 BTF_KIND_CONST
335 ~~~~~~~~~~~~~~~~~~~~~
337 ``struct btf_type`` encoding requirement:
339 * ``info.kind_flag``: 0
340 * ``info.kind``: BTF_KIND_CONST
342 * ``type``: the type with ``const`` qualifier
344 No additional type data follow ``btf_type``.
346 2.2.11 BTF_KIND_RESTRICT
347 ~~~~~~~~~~~~~~~~~~~~~~~~
349 ``struct btf_type`` encoding requirement:
351 * ``info.kind_flag``: 0
352 * ``info.kind``: BTF_KIND_RESTRICT
354 * ``type``: the type with ``restrict`` qualifier
356 No additional type data follow ``btf_type``.
361 ``struct btf_type`` encoding requirement:
362 * ``name_off``: offset to a valid C identifier
363 * ``info.kind_flag``: 0
364 * ``info.kind``: BTF_KIND_FUNC
366 * ``type``: a BTF_KIND_FUNC_PROTO type
368 No additional type data follow ``btf_type``.
370 A BTF_KIND_FUNC defines not a type, but a subprogram (function) whose
371 signature is defined by ``type``. The subprogram is thus an instance of that
372 type. The BTF_KIND_FUNC may in turn be referenced by a func_info in the
373 :ref:`BTF_Ext_Section` (ELF) or in the arguments to :ref:`BPF_Prog_Load`
376 2.2.13 BTF_KIND_FUNC_PROTO
377 ~~~~~~~~~~~~~~~~~~~~~~~~~~
379 ``struct btf_type`` encoding requirement:
381 * ``info.kind_flag``: 0
382 * ``info.kind``: BTF_KIND_FUNC_PROTO
383 * ``info.vlen``: # of parameters
384 * ``type``: the return type
386 ``btf_type`` is followed by ``info.vlen`` number of ``struct btf_param``.::
393 If a BTF_KIND_FUNC_PROTO type is referred by a BTF_KIND_FUNC type, then
394 ``btf_param.name_off`` must point to a valid C identifier except for the
395 possible last argument representing the variable argument. The btf_param.type
396 refers to parameter type.
398 If the function has variable arguments, the last parameter is encoded with
399 ``name_off = 0`` and ``type = 0``.
404 ``struct btf_type`` encoding requirement:
405 * ``name_off``: offset to a valid C identifier
406 * ``info.kind_flag``: 0
407 * ``info.kind``: BTF_KIND_VAR
409 * ``type``: the type of the variable
411 ``btf_type`` is followed by a single ``struct btf_variable`` with the
418 ``struct btf_var`` encoding:
419 * ``linkage``: currently only static variable 0, or globally allocated
420 variable in ELF sections 1
422 Not all type of global variables are supported by LLVM at this point.
423 The following is currently available:
425 * static variables with or without section attributes
426 * global variables with section attributes
428 The latter is for future extraction of map key/value type id's from a
431 2.2.15 BTF_KIND_DATASEC
432 ~~~~~~~~~~~~~~~~~~~~~~~
434 ``struct btf_type`` encoding requirement:
435 * ``name_off``: offset to a valid name associated with a variable or
436 one of .data/.bss/.rodata
437 * ``info.kind_flag``: 0
438 * ``info.kind``: BTF_KIND_DATASEC
439 * ``info.vlen``: # of variables
440 * ``size``: total section size in bytes (0 at compilation time, patched
441 to actual size by BPF loaders such as libbpf)
443 ``btf_type`` is followed by ``info.vlen`` number of ``struct btf_var_secinfo``.::
445 struct btf_var_secinfo {
451 ``struct btf_var_secinfo`` encoding:
452 * ``type``: the type of the BTF_KIND_VAR variable
453 * ``offset``: the in-section offset of the variable
454 * ``size``: the size of the variable in bytes
456 2.2.16 BTF_KIND_FLOAT
457 ~~~~~~~~~~~~~~~~~~~~~
459 ``struct btf_type`` encoding requirement:
460 * ``name_off``: any valid offset
461 * ``info.kind_flag``: 0
462 * ``info.kind``: BTF_KIND_FLOAT
464 * ``size``: the size of the float type in bytes: 2, 4, 8, 12 or 16.
466 No additional type data follow ``btf_type``.
471 The following bpf syscall command involves BTF:
472 * BPF_BTF_LOAD: load a blob of BTF data into kernel
473 * BPF_MAP_CREATE: map creation with btf key and value type info.
474 * BPF_PROG_LOAD: prog load with btf function and line info.
475 * BPF_BTF_GET_FD_BY_ID: get a btf fd
476 * BPF_OBJ_GET_INFO_BY_FD: btf, func_info, line_info
477 and other btf related info are returned.
479 The workflow typically looks like:
486 BPF_MAP_CREATE and BPF_PROG_LOAD
493 BPF_{PROG,MAP}_GET_NEXT_ID (get prog/map id's)
496 BPF_{PROG,MAP}_GET_FD_BY_ID (get a prog/map fd)
499 BPF_OBJ_GET_INFO_BY_FD (get bpf_prog_info/bpf_map_info with btf_id)
502 BPF_BTF_GET_FD_BY_ID (get btf_fd) |
505 BPF_OBJ_GET_INFO_BY_FD (get btf) |
508 pretty print types, dump func signatures and line info, etc.
514 Load a blob of BTF data into kernel. A blob of data, described in
515 :ref:`BTF_Type_String`, can be directly loaded into the kernel. A ``btf_fd``
516 is returned to a userspace.
521 A map can be created with ``btf_fd`` and specified key/value type id.::
523 __u32 btf_fd; /* fd pointing to a BTF type data */
524 __u32 btf_key_type_id; /* BTF type_id of the key */
525 __u32 btf_value_type_id; /* BTF type_id of the value */
527 In libbpf, the map can be defined with extra annotation like below:
530 struct bpf_map_def SEC("maps") btf_map = {
531 .type = BPF_MAP_TYPE_ARRAY,
532 .key_size = sizeof(int),
533 .value_size = sizeof(struct ipv_counts),
536 BPF_ANNOTATE_KV_PAIR(btf_map, int, struct ipv_counts);
538 Here, the parameters for macro BPF_ANNOTATE_KV_PAIR are map name, key and
539 value types for the map. During ELF parsing, libbpf is able to extract
540 key/value type_id's and assign them to BPF_MAP_CREATE attributes
548 During prog_load, func_info and line_info can be passed to kernel with proper
549 values for the following attributes:
555 __u32 prog_btf_fd; /* fd pointing to BTF type data */
556 __u32 func_info_rec_size; /* userspace bpf_func_info size */
557 __aligned_u64 func_info; /* func info */
558 __u32 func_info_cnt; /* number of bpf_func_info records */
559 __u32 line_info_rec_size; /* userspace bpf_line_info size */
560 __aligned_u64 line_info; /* line info */
561 __u32 line_info_cnt; /* number of bpf_line_info records */
563 The func_info and line_info are an array of below, respectively.::
565 struct bpf_func_info {
566 __u32 insn_off; /* [0, insn_cnt - 1] */
567 __u32 type_id; /* pointing to a BTF_KIND_FUNC type */
569 struct bpf_line_info {
570 __u32 insn_off; /* [0, insn_cnt - 1] */
571 __u32 file_name_off; /* offset to string table for the filename */
572 __u32 line_off; /* offset to string table for the source line */
573 __u32 line_col; /* line number and column number */
576 func_info_rec_size is the size of each func_info record, and
577 line_info_rec_size is the size of each line_info record. Passing the record
578 size to kernel make it possible to extend the record itself in the future.
580 Below are requirements for func_info:
581 * func_info[0].insn_off must be 0.
582 * the func_info insn_off is in strictly increasing order and matches
585 Below are requirements for line_info:
586 * the first insn in each func must have a line_info record pointing to it.
587 * the line_info insn_off is in strictly increasing order.
589 For line_info, the line number and column number are defined as below:
592 #define BPF_LINE_INFO_LINE_NUM(line_col) ((line_col) >> 10)
593 #define BPF_LINE_INFO_LINE_COL(line_col) ((line_col) & 0x3ff)
595 3.4 BPF_{PROG,MAP}_GET_NEXT_ID
596 ==============================
598 In kernel, every loaded program, map or btf has a unique id. The id won't
599 change during the lifetime of a program, map, or btf.
601 The bpf syscall command BPF_{PROG,MAP}_GET_NEXT_ID returns all id's, one for
602 each command, to user space, for bpf program or maps, respectively, so an
603 inspection tool can inspect all programs and maps.
605 3.5 BPF_{PROG,MAP}_GET_FD_BY_ID
606 ===============================
608 An introspection tool cannot use id to get details about program or maps.
609 A file descriptor needs to be obtained first for reference-counting purpose.
611 3.6 BPF_OBJ_GET_INFO_BY_FD
612 ==========================
614 Once a program/map fd is acquired, an introspection tool can get the detailed
615 information from kernel about this fd, some of which are BTF-related. For
616 example, ``bpf_map_info`` returns ``btf_id`` and key/value type ids.
617 ``bpf_prog_info`` returns ``btf_id``, func_info, and line info for translated
618 bpf byte codes, and jited_line_info.
620 3.7 BPF_BTF_GET_FD_BY_ID
621 ========================
623 With ``btf_id`` obtained in ``bpf_map_info`` and ``bpf_prog_info``, bpf
624 syscall command BPF_BTF_GET_FD_BY_ID can retrieve a btf fd. Then, with
625 command BPF_OBJ_GET_INFO_BY_FD, the btf blob, originally loaded into the
626 kernel with BPF_BTF_LOAD, can be retrieved.
628 With the btf blob, ``bpf_map_info``, and ``bpf_prog_info``, an introspection
629 tool has full btf knowledge and is able to pretty print map key/values, dump
630 func signatures and line info, along with byte/jit codes.
632 4. ELF File Format Interface
633 ****************************
638 The .BTF section contains type and string data. The format of this section is
639 same as the one describe in :ref:`BTF_Type_String`.
646 The .BTF.ext section encodes func_info and line_info which needs loader
647 manipulation before loading into the kernel.
649 The specification for .BTF.ext section is defined at ``tools/lib/bpf/btf.h``
650 and ``tools/lib/bpf/btf.c``.
652 The current header of .BTF.ext section::
654 struct btf_ext_header {
660 /* All offsets are in bytes relative to the end of this header */
667 It is very similar to .BTF section. Instead of type/string section, it
668 contains func_info and line_info section. See :ref:`BPF_Prog_Load` for details
669 about func_info and line_info record format.
671 The func_info is organized as below.::
674 btf_ext_info_sec for section #1 /* func_info for section #1 */
675 btf_ext_info_sec for section #2 /* func_info for section #2 */
678 ``func_info_rec_size`` specifies the size of ``bpf_func_info`` structure when
679 .BTF.ext is generated. ``btf_ext_info_sec``, defined below, is a collection of
680 func_info for each specific ELF section.::
682 struct btf_ext_info_sec {
683 __u32 sec_name_off; /* offset to section name */
685 /* Followed by num_info * record_size number of bytes */
689 Here, num_info must be greater than 0.
691 The line_info is organized as below.::
694 btf_ext_info_sec for section #1 /* line_info for section #1 */
695 btf_ext_info_sec for section #2 /* line_info for section #2 */
698 ``line_info_rec_size`` specifies the size of ``bpf_line_info`` structure when
699 .BTF.ext is generated.
701 The interpretation of ``bpf_func_info->insn_off`` and
702 ``bpf_line_info->insn_off`` is different between kernel API and ELF API. For
703 kernel API, the ``insn_off`` is the instruction offset in the unit of ``struct
704 bpf_insn``. For ELF API, the ``insn_off`` is the byte offset from the
705 beginning of section (``btf_ext_info_sec->sec_name_off``).
710 The .BTF_ids section encodes BTF ID values that are used within the kernel.
712 This section is created during the kernel compilation with the help of
713 macros defined in ``include/linux/btf_ids.h`` header file. Kernel code can
714 use them to create lists and sets (sorted lists) of BTF ID values.
716 The ``BTF_ID_LIST`` and ``BTF_ID`` macros define unsorted list of BTF ID values,
717 with following syntax::
723 resulting in following layout in .BTF_ids section::
725 __BTF_ID__type1__name1__1:
727 __BTF_ID__type2__name2__2:
730 The ``u32 list[];`` variable is defined to access the list.
732 The ``BTF_ID_UNUSED`` macro defines 4 zero bytes. It's used when we
733 want to define unused entry in BTF_ID_LIST, like::
735 BTF_ID_LIST(bpf_skb_output_btf_ids)
736 BTF_ID(struct, sk_buff)
738 BTF_ID(struct, task_struct)
740 The ``BTF_SET_START/END`` macros pair defines sorted list of BTF ID values
741 and their count, with following syntax::
748 resulting in following layout in .BTF_ids section::
752 __BTF_ID__type1__name1__3:
754 __BTF_ID__type2__name2__4:
757 The ``struct btf_id_set set;`` variable is defined to access the list.
759 The ``typeX`` name can be one of following::
761 struct, union, typedef, func
763 and is used as a filter when resolving the BTF ID value.
765 All the BTF ID lists and sets are compiled in the .BTF_ids section and
766 resolved during the linking phase of kernel build by ``resolve_btfids`` tool.
771 5.1 bpftool map pretty print
772 ============================
774 With BTF, the map key/value can be printed based on fields rather than simply
775 raw bytes. This is especially valuable for large structure or if your data
776 structure has bitfields. For example, for the following map,::
778 enum A { A1, A2, A3, A4, A5 };
789 struct bpf_map_def SEC("maps") tmpmap = {
790 .type = BPF_MAP_TYPE_ARRAY,
791 .key_size = sizeof(__u32),
792 .value_size = sizeof(struct tmp_t),
795 BPF_ANNOTATE_KV_PAIR(tmpmap, int, struct tmp_t);
797 bpftool is able to pretty print like below:
813 5.2 bpftool prog dump
814 =====================
816 The following is an example showing how func_info and line_info can help prog
817 dump with better kernel symbol names, function prototypes and line
820 $ bpftool prog dump jited pinned /sys/fs/bpf/test_btf_haskv
822 int test_long_fname_2(struct dummy_tracepoint_args * arg):
823 bpf_prog_44a040bf25481309_test_long_fname_2:
824 ; static int test_long_fname_2(struct dummy_tracepoint_args *arg)
829 f: mov %rbx,0x0(%rbp)
830 13: mov %r13,0x8(%rbp)
831 17: mov %r14,0x10(%rbp)
832 1b: mov %r15,0x18(%rbp)
834 21: mov %rax,0x20(%rbp)
837 27: mov %esi,-0x4(%rbp)
839 2a: mov 0x8(%rdi),%rdi
842 32: je 0x0000000000000070
844 ; counts = bpf_map_lookup_elem(&btf_map, &key);
850 The following is an example of how line_info can help debugging verification
853 /* The code at tools/testing/selftests/bpf/test_xdp_noinline.c
854 * is modified as below.
856 data = (void *)(long)xdp->data;
857 data_end = (void *)(long)xdp->data_end;
859 if (data + 4 > data_end)
862 *(u32 *)data = dst->dst;
864 $ bpftool prog load ./test_xdp_noinline.o /sys/fs/bpf/test_xdp_noinline type xdp
865 ; data = (void *)(long)xdp->data;
866 224: (79) r2 = *(u64 *)(r10 -112)
867 225: (61) r2 = *(u32 *)(r2 +0)
868 ; *(u32 *)data = dst->dst;
869 226: (63) *(u32 *)(r2 +0) = r1
870 invalid access to packet, off=0 size=4, R2(id=0,off=0,r=0)
871 R2 offset is outside of the packet
876 You need latest pahole
878 https://git.kernel.org/pub/scm/devel/pahole/pahole.git/
880 or llvm (8.0 or later). The pahole acts as a dwarf2btf converter. It doesn't
881 support .BTF.ext and btf BTF_KIND_FUNC type yet. For example,::
889 -bash-4.4$ gcc -c -O2 -g t.c
890 -bash-4.4$ pahole -JV t.o
892 [1] STRUCT t kind_flag=1 size=4 vlen=3
893 a type_id=2 bitfield_size=2 bits_offset=0
894 b type_id=2 bitfield_size=3 bits_offset=2
895 c type_id=2 bitfield_size=2 bits_offset=5
896 [2] INT int size=4 bit_offset=0 nr_bits=32 encoding=SIGNED
898 The llvm is able to generate .BTF and .BTF.ext directly with -g for bpf target
899 only. The assembly code (-S) is able to show the BTF encoding in assembly
906 int (*f2)(char q1, __int32 q2, ...);
909 int main() { return 0; }
910 int test() { return 0; }
911 -bash-4.4$ clang -c -g -O2 -target bpf t2.c
912 -bash-4.4$ readelf -S t2.o
914 [ 8] .BTF PROGBITS 0000000000000000 00000247
915 000000000000016e 0000000000000000 0 0 1
916 [ 9] .BTF.ext PROGBITS 0000000000000000 000003b5
917 0000000000000060 0000000000000000 0 0 1
918 [10] .rel.BTF.ext REL 0000000000000000 000007e0
919 0000000000000040 0000000000000010 16 9 8
921 -bash-4.4$ clang -S -g -O2 -target bpf t2.c
924 .section .BTF,"",@progbits
925 .short 60319 # 0xeb9f
933 .long 0 # BTF_KIND_FUNC_PROTO(id = 1)
934 .long 218103808 # 0xd000000
936 .long 83 # BTF_KIND_INT(id = 2)
937 .long 16777216 # 0x1000000
939 .long 16777248 # 0x1000020
941 .byte 0 # string offset=0
942 .ascii ".text" # string offset=1
944 .ascii "/home/yhs/tmp-pahole/t2.c" # string offset=7
946 .ascii "int main() { return 0; }" # string offset=33
948 .ascii "int test() { return 0; }" # string offset=58
950 .ascii "int" # string offset=83
952 .section .BTF.ext,"",@progbits
953 .short 60319 # 0xeb9f
962 .long 1 # FuncInfo section string offset=1
969 .long 1 # LineInfo section string offset=1
974 .long 7182 # Line 7 Col 14
978 .long 8206 # Line 8 Col 14
983 Kernel bpf selftest `test_btf.c` provides extensive set of BTF-related tests.