1 ===================== 1 =====================
2 BPF Type Format (BTF) 2 BPF Type Format (BTF)
3 ===================== 3 =====================
4 4
5 1. Introduction 5 1. Introduction
6 =============== 6 ===============
7 7
8 BTF (BPF Type Format) is the metadata format w 8 BTF (BPF Type Format) is the metadata format which encodes the debug info
9 related to BPF program/map. The name BTF was u 9 related to BPF program/map. The name BTF was used initially to describe data
10 types. The BTF was later extended to include f 10 types. The BTF was later extended to include function info for defined
11 subroutines, and line info for source/line inf 11 subroutines, and line info for source/line information.
12 12
13 The debug info is used for map pretty print, f 13 The debug info is used for map pretty print, function signature, etc. The
14 function signature enables better bpf program/ 14 function signature enables better bpf program/function kernel symbol. The line
15 info helps generate source annotated translate 15 info helps generate source annotated translated byte code, jited code and
16 verifier log. 16 verifier log.
17 17
18 The BTF specification contains two parts, 18 The BTF specification contains two parts,
19 * BTF kernel API 19 * BTF kernel API
20 * BTF ELF file format 20 * BTF ELF file format
21 21
22 The kernel API is the contract between user sp 22 The kernel API is the contract between user space and kernel. The kernel
23 verifies the BTF info before using it. The ELF 23 verifies the BTF info before using it. The ELF file format is a user space
24 contract between ELF file and libbpf loader. 24 contract between ELF file and libbpf loader.
25 25
26 The type and string sections are part of the B 26 The type and string sections are part of the BTF kernel API, describing the
27 debug info (mostly types related) referenced b 27 debug info (mostly types related) referenced by the bpf program. These two
28 sections are discussed in details in :ref:`BTF 28 sections are discussed in details in :ref:`BTF_Type_String`.
29 29
30 .. _BTF_Type_String: 30 .. _BTF_Type_String:
31 31
32 2. BTF Type and String Encoding 32 2. BTF Type and String Encoding
33 =============================== 33 ===============================
34 34
35 The file ``include/uapi/linux/btf.h`` provides 35 The file ``include/uapi/linux/btf.h`` provides high-level definition of how
36 types/strings are encoded. 36 types/strings are encoded.
37 37
38 The beginning of data blob must be:: 38 The beginning of data blob must be::
39 39
40 struct btf_header { 40 struct btf_header {
41 __u16 magic; 41 __u16 magic;
42 __u8 version; 42 __u8 version;
43 __u8 flags; 43 __u8 flags;
44 __u32 hdr_len; 44 __u32 hdr_len;
45 45
46 /* All offsets are in bytes relative t 46 /* All offsets are in bytes relative to the end of this header */
47 __u32 type_off; /* offset of t 47 __u32 type_off; /* offset of type section */
48 __u32 type_len; /* length of t 48 __u32 type_len; /* length of type section */
49 __u32 str_off; /* offset of s 49 __u32 str_off; /* offset of string section */
50 __u32 str_len; /* length of s 50 __u32 str_len; /* length of string section */
51 }; 51 };
52 52
53 The magic is ``0xeB9F``, which has different e 53 The magic is ``0xeB9F``, which has different encoding for big and little
54 endian systems, and can be used to test whethe 54 endian systems, and can be used to test whether BTF is generated for big- or
55 little-endian target. The ``btf_header`` is de 55 little-endian target. The ``btf_header`` is designed to be extensible with
56 ``hdr_len`` equal to ``sizeof(struct btf_heade 56 ``hdr_len`` equal to ``sizeof(struct btf_header)`` when a data blob is
57 generated. 57 generated.
58 58
59 2.1 String Encoding 59 2.1 String Encoding
60 ------------------- 60 -------------------
61 61
62 The first string in the string section must be 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- 63 string table is a concatenation of other null-terminated strings.
64 64
65 2.2 Type Encoding 65 2.2 Type Encoding
66 ----------------- 66 -----------------
67 67
68 The type id ``0`` is reserved for ``void`` typ 68 The type id ``0`` is reserved for ``void`` type. The type section is parsed
69 sequentially and type id is assigned to each r 69 sequentially and type id is assigned to each recognized type starting from id
70 ``1``. Currently, the following types are supp 70 ``1``. Currently, the following types are supported::
71 71
72 #define BTF_KIND_INT 1 /* 72 #define BTF_KIND_INT 1 /* Integer */
73 #define BTF_KIND_PTR 2 /* 73 #define BTF_KIND_PTR 2 /* Pointer */
74 #define BTF_KIND_ARRAY 3 /* 74 #define BTF_KIND_ARRAY 3 /* Array */
75 #define BTF_KIND_STRUCT 4 /* 75 #define BTF_KIND_STRUCT 4 /* Struct */
76 #define BTF_KIND_UNION 5 /* 76 #define BTF_KIND_UNION 5 /* Union */
77 #define BTF_KIND_ENUM 6 /* !! 77 #define BTF_KIND_ENUM 6 /* Enumeration */
78 #define BTF_KIND_FWD 7 /* 78 #define BTF_KIND_FWD 7 /* Forward */
79 #define BTF_KIND_TYPEDEF 8 /* 79 #define BTF_KIND_TYPEDEF 8 /* Typedef */
80 #define BTF_KIND_VOLATILE 9 /* 80 #define BTF_KIND_VOLATILE 9 /* Volatile */
81 #define BTF_KIND_CONST 10 /* 81 #define BTF_KIND_CONST 10 /* Const */
82 #define BTF_KIND_RESTRICT 11 /* 82 #define BTF_KIND_RESTRICT 11 /* Restrict */
83 #define BTF_KIND_FUNC 12 /* 83 #define BTF_KIND_FUNC 12 /* Function */
84 #define BTF_KIND_FUNC_PROTO 13 /* 84 #define BTF_KIND_FUNC_PROTO 13 /* Function Proto */
85 #define BTF_KIND_VAR 14 /* 85 #define BTF_KIND_VAR 14 /* Variable */
86 #define BTF_KIND_DATASEC 15 /* 86 #define BTF_KIND_DATASEC 15 /* Section */
87 #define BTF_KIND_FLOAT 16 /* 87 #define BTF_KIND_FLOAT 16 /* Floating point */
88 #define BTF_KIND_DECL_TAG 17 /* 88 #define BTF_KIND_DECL_TAG 17 /* Decl Tag */
89 #define BTF_KIND_TYPE_TAG 18 /* 89 #define BTF_KIND_TYPE_TAG 18 /* Type Tag */
90 #define BTF_KIND_ENUM64 19 /* <<
91 90
92 Note that the type section encodes debug info, 91 Note that the type section encodes debug info, not just pure types.
93 ``BTF_KIND_FUNC`` is not a type, and it repres 92 ``BTF_KIND_FUNC`` is not a type, and it represents a defined subprogram.
94 93
95 Each type contains the following common data:: 94 Each type contains the following common data::
96 95
97 struct btf_type { 96 struct btf_type {
98 __u32 name_off; 97 __u32 name_off;
99 /* "info" bits arrangement 98 /* "info" bits arrangement
100 * bits 0-15: vlen (e.g. # of struct' 99 * bits 0-15: vlen (e.g. # of struct's members)
101 * bits 16-23: unused 100 * bits 16-23: unused
102 * bits 24-28: kind (e.g. int, ptr, ar 101 * bits 24-28: kind (e.g. int, ptr, array...etc)
103 * bits 29-30: unused 102 * bits 29-30: unused
104 * bit 31: kind_flag, currently us 103 * bit 31: kind_flag, currently used by
105 * struct, union, fwd, enu !! 104 * struct, union and fwd
106 */ 105 */
107 __u32 info; 106 __u32 info;
108 /* "size" is used by INT, ENUM, STRUCT !! 107 /* "size" is used by INT, ENUM, STRUCT and UNION.
109 * "size" tells the size of the type i 108 * "size" tells the size of the type it is describing.
110 * 109 *
111 * "type" is used by PTR, TYPEDEF, VOL 110 * "type" is used by PTR, TYPEDEF, VOLATILE, CONST, RESTRICT,
112 * FUNC, FUNC_PROTO, DECL_TAG and TYPE 111 * FUNC, FUNC_PROTO, DECL_TAG and TYPE_TAG.
113 * "type" is a type_id referring to an 112 * "type" is a type_id referring to another type.
114 */ 113 */
115 union { 114 union {
116 __u32 size; 115 __u32 size;
117 __u32 type; 116 __u32 type;
118 }; 117 };
119 }; 118 };
120 119
121 For certain kinds, the common data are followe 120 For certain kinds, the common data are followed by kind-specific data. The
122 ``name_off`` in ``struct btf_type`` specifies 121 ``name_off`` in ``struct btf_type`` specifies the offset in the string table.
123 The following sections detail encoding of each 122 The following sections detail encoding of each kind.
124 123
125 2.2.1 BTF_KIND_INT 124 2.2.1 BTF_KIND_INT
126 ~~~~~~~~~~~~~~~~~~ 125 ~~~~~~~~~~~~~~~~~~
127 126
128 ``struct btf_type`` encoding requirement: 127 ``struct btf_type`` encoding requirement:
129 * ``name_off``: any valid offset 128 * ``name_off``: any valid offset
130 * ``info.kind_flag``: 0 129 * ``info.kind_flag``: 0
131 * ``info.kind``: BTF_KIND_INT 130 * ``info.kind``: BTF_KIND_INT
132 * ``info.vlen``: 0 131 * ``info.vlen``: 0
133 * ``size``: the size of the int type in bytes 132 * ``size``: the size of the int type in bytes.
134 133
135 ``btf_type`` is followed by a ``u32`` with the 134 ``btf_type`` is followed by a ``u32`` with the following bits arrangement::
136 135
137 #define BTF_INT_ENCODING(VAL) (((VAL) & 0x 136 #define BTF_INT_ENCODING(VAL) (((VAL) & 0x0f000000) >> 24)
138 #define BTF_INT_OFFSET(VAL) (((VAL) & 0x 137 #define BTF_INT_OFFSET(VAL) (((VAL) & 0x00ff0000) >> 16)
139 #define BTF_INT_BITS(VAL) ((VAL) & 0x 138 #define BTF_INT_BITS(VAL) ((VAL) & 0x000000ff)
140 139
141 The ``BTF_INT_ENCODING`` has the following att 140 The ``BTF_INT_ENCODING`` has the following attributes::
142 141
143 #define BTF_INT_SIGNED (1 << 0) 142 #define BTF_INT_SIGNED (1 << 0)
144 #define BTF_INT_CHAR (1 << 1) 143 #define BTF_INT_CHAR (1 << 1)
145 #define BTF_INT_BOOL (1 << 2) 144 #define BTF_INT_BOOL (1 << 2)
146 145
147 The ``BTF_INT_ENCODING()`` provides extra info 146 The ``BTF_INT_ENCODING()`` provides extra information: signedness, char, or
148 bool, for the int type. The char and bool enco 147 bool, for the int type. The char and bool encoding are mostly useful for
149 pretty print. At most one encoding can be spec 148 pretty print. At most one encoding can be specified for the int type.
150 149
151 The ``BTF_INT_BITS()`` specifies the number of 150 The ``BTF_INT_BITS()`` specifies the number of actual bits held by this int
152 type. For example, a 4-bit bitfield encodes `` 151 type. For example, a 4-bit bitfield encodes ``BTF_INT_BITS()`` equals to 4.
153 The ``btf_type.size * 8`` must be equal to or 152 The ``btf_type.size * 8`` must be equal to or greater than ``BTF_INT_BITS()``
154 for the type. The maximum value of ``BTF_INT_B 153 for the type. The maximum value of ``BTF_INT_BITS()`` is 128.
155 154
156 The ``BTF_INT_OFFSET()`` specifies the startin 155 The ``BTF_INT_OFFSET()`` specifies the starting bit offset to calculate values
157 for this int. For example, a bitfield struct m 156 for this int. For example, a bitfield struct member has:
158 157
159 * btf member bit offset 100 from the start of 158 * btf member bit offset 100 from the start of the structure,
160 * btf member pointing to an int type, 159 * btf member pointing to an int type,
161 * the int type has ``BTF_INT_OFFSET() = 2`` a 160 * the int type has ``BTF_INT_OFFSET() = 2`` and ``BTF_INT_BITS() = 4``
162 161
163 Then in the struct memory layout, this member 162 Then in the struct memory layout, this member will occupy ``4`` bits starting
164 from bits ``100 + 2 = 102``. 163 from bits ``100 + 2 = 102``.
165 164
166 Alternatively, the bitfield struct member can 165 Alternatively, the bitfield struct member can be the following to access the
167 same bits as the above: 166 same bits as the above:
168 167
169 * btf member bit offset 102, 168 * btf member bit offset 102,
170 * btf member pointing to an int type, 169 * btf member pointing to an int type,
171 * the int type has ``BTF_INT_OFFSET() = 0`` a 170 * the int type has ``BTF_INT_OFFSET() = 0`` and ``BTF_INT_BITS() = 4``
172 171
173 The original intention of ``BTF_INT_OFFSET()`` 172 The original intention of ``BTF_INT_OFFSET()`` is to provide flexibility of
174 bitfield encoding. Currently, both llvm and pa 173 bitfield encoding. Currently, both llvm and pahole generate
175 ``BTF_INT_OFFSET() = 0`` for all int types. 174 ``BTF_INT_OFFSET() = 0`` for all int types.
176 175
177 2.2.2 BTF_KIND_PTR 176 2.2.2 BTF_KIND_PTR
178 ~~~~~~~~~~~~~~~~~~ 177 ~~~~~~~~~~~~~~~~~~
179 178
180 ``struct btf_type`` encoding requirement: 179 ``struct btf_type`` encoding requirement:
181 * ``name_off``: 0 180 * ``name_off``: 0
182 * ``info.kind_flag``: 0 181 * ``info.kind_flag``: 0
183 * ``info.kind``: BTF_KIND_PTR 182 * ``info.kind``: BTF_KIND_PTR
184 * ``info.vlen``: 0 183 * ``info.vlen``: 0
185 * ``type``: the pointee type of the pointer 184 * ``type``: the pointee type of the pointer
186 185
187 No additional type data follow ``btf_type``. 186 No additional type data follow ``btf_type``.
188 187
189 2.2.3 BTF_KIND_ARRAY 188 2.2.3 BTF_KIND_ARRAY
190 ~~~~~~~~~~~~~~~~~~~~ 189 ~~~~~~~~~~~~~~~~~~~~
191 190
192 ``struct btf_type`` encoding requirement: 191 ``struct btf_type`` encoding requirement:
193 * ``name_off``: 0 192 * ``name_off``: 0
194 * ``info.kind_flag``: 0 193 * ``info.kind_flag``: 0
195 * ``info.kind``: BTF_KIND_ARRAY 194 * ``info.kind``: BTF_KIND_ARRAY
196 * ``info.vlen``: 0 195 * ``info.vlen``: 0
197 * ``size/type``: 0, not used 196 * ``size/type``: 0, not used
198 197
199 ``btf_type`` is followed by one ``struct btf_a 198 ``btf_type`` is followed by one ``struct btf_array``::
200 199
201 struct btf_array { 200 struct btf_array {
202 __u32 type; 201 __u32 type;
203 __u32 index_type; 202 __u32 index_type;
204 __u32 nelems; 203 __u32 nelems;
205 }; 204 };
206 205
207 The ``struct btf_array`` encoding: 206 The ``struct btf_array`` encoding:
208 * ``type``: the element type 207 * ``type``: the element type
209 * ``index_type``: the index type 208 * ``index_type``: the index type
210 * ``nelems``: the number of elements for thi 209 * ``nelems``: the number of elements for this array (``0`` is also allowed).
211 210
212 The ``index_type`` can be any regular int type 211 The ``index_type`` can be any regular int type (``u8``, ``u16``, ``u32``,
213 ``u64``, ``unsigned __int128``). The original 212 ``u64``, ``unsigned __int128``). The original design of including
214 ``index_type`` follows DWARF, which has an ``i 213 ``index_type`` follows DWARF, which has an ``index_type`` for its array type.
215 Currently in BTF, beyond type verification, th 214 Currently in BTF, beyond type verification, the ``index_type`` is not used.
216 215
217 The ``struct btf_array`` allows chaining throu 216 The ``struct btf_array`` allows chaining through element type to represent
218 multidimensional arrays. For example, for ``in 217 multidimensional arrays. For example, for ``int a[5][6]``, the following type
219 information illustrates the chaining: 218 information illustrates the chaining:
220 219
221 * [1]: int 220 * [1]: int
222 * [2]: array, ``btf_array.type = [1]``, ``bt 221 * [2]: array, ``btf_array.type = [1]``, ``btf_array.nelems = 6``
223 * [3]: array, ``btf_array.type = [2]``, ``bt 222 * [3]: array, ``btf_array.type = [2]``, ``btf_array.nelems = 5``
224 223
225 Currently, both pahole and llvm collapse multi 224 Currently, both pahole and llvm collapse multidimensional array into
226 one-dimensional array, e.g., for ``a[5][6]``, 225 one-dimensional array, e.g., for ``a[5][6]``, the ``btf_array.nelems`` is
227 equal to ``30``. This is because the original 226 equal to ``30``. This is because the original use case is map pretty print
228 where the whole array is dumped out so one-dim 227 where the whole array is dumped out so one-dimensional array is enough. As
229 more BTF usage is explored, pahole and llvm ca 228 more BTF usage is explored, pahole and llvm can be changed to generate proper
230 chained representation for multidimensional ar 229 chained representation for multidimensional arrays.
231 230
232 2.2.4 BTF_KIND_STRUCT 231 2.2.4 BTF_KIND_STRUCT
233 ~~~~~~~~~~~~~~~~~~~~~ 232 ~~~~~~~~~~~~~~~~~~~~~
234 2.2.5 BTF_KIND_UNION 233 2.2.5 BTF_KIND_UNION
235 ~~~~~~~~~~~~~~~~~~~~ 234 ~~~~~~~~~~~~~~~~~~~~
236 235
237 ``struct btf_type`` encoding requirement: 236 ``struct btf_type`` encoding requirement:
238 * ``name_off``: 0 or offset to a valid C ide 237 * ``name_off``: 0 or offset to a valid C identifier
239 * ``info.kind_flag``: 0 or 1 238 * ``info.kind_flag``: 0 or 1
240 * ``info.kind``: BTF_KIND_STRUCT or BTF_KIND 239 * ``info.kind``: BTF_KIND_STRUCT or BTF_KIND_UNION
241 * ``info.vlen``: the number of struct/union 240 * ``info.vlen``: the number of struct/union members
242 * ``info.size``: the size of the struct/unio 241 * ``info.size``: the size of the struct/union in bytes
243 242
244 ``btf_type`` is followed by ``info.vlen`` numb 243 ``btf_type`` is followed by ``info.vlen`` number of ``struct btf_member``.::
245 244
246 struct btf_member { 245 struct btf_member {
247 __u32 name_off; 246 __u32 name_off;
248 __u32 type; 247 __u32 type;
249 __u32 offset; 248 __u32 offset;
250 }; 249 };
251 250
252 ``struct btf_member`` encoding: 251 ``struct btf_member`` encoding:
253 * ``name_off``: offset to a valid C identifi 252 * ``name_off``: offset to a valid C identifier
254 * ``type``: the member type 253 * ``type``: the member type
255 * ``offset``: <see below> 254 * ``offset``: <see below>
256 255
257 If the type info ``kind_flag`` is not set, the 256 If the type info ``kind_flag`` is not set, the offset contains only bit offset
258 of the member. Note that the base type of the 257 of the member. Note that the base type of the bitfield can only be int or enum
259 type. If the bitfield size is 32, the base typ 258 type. If the bitfield size is 32, the base type can be either int or enum
260 type. If the bitfield size is not 32, the base 259 type. If the bitfield size is not 32, the base type must be int, and int type
261 ``BTF_INT_BITS()`` encodes the bitfield size. 260 ``BTF_INT_BITS()`` encodes the bitfield size.
262 261
263 If the ``kind_flag`` is set, the ``btf_member. 262 If the ``kind_flag`` is set, the ``btf_member.offset`` contains both member
264 bitfield size and bit offset. The bitfield siz 263 bitfield size and bit offset. The bitfield size and bit offset are calculated
265 as below.:: 264 as below.::
266 265
267 #define BTF_MEMBER_BITFIELD_SIZE(val) ((va 266 #define BTF_MEMBER_BITFIELD_SIZE(val) ((val) >> 24)
268 #define BTF_MEMBER_BIT_OFFSET(val) ((va 267 #define BTF_MEMBER_BIT_OFFSET(val) ((val) & 0xffffff)
269 268
270 In this case, if the base type is an int type, 269 In this case, if the base type is an int type, it must be a regular int type:
271 270
272 * ``BTF_INT_OFFSET()`` must be 0. 271 * ``BTF_INT_OFFSET()`` must be 0.
273 * ``BTF_INT_BITS()`` must be equal to ``{1,2 272 * ``BTF_INT_BITS()`` must be equal to ``{1,2,4,8,16} * 8``.
274 273
275 Commit 9d5f9f701b18 introduced ``kind_flag`` a !! 274 The following kernel patch introduced ``kind_flag`` and explained why both
276 exist. !! 275 modes exist:
>> 276
>> 277 https://github.com/torvalds/linux/commit/9d5f9f701b1891466fb3dbb1806ad97716f95cc3#diff-fa650a64fdd3968396883d2fe8215ff3
277 278
278 2.2.6 BTF_KIND_ENUM 279 2.2.6 BTF_KIND_ENUM
279 ~~~~~~~~~~~~~~~~~~~ 280 ~~~~~~~~~~~~~~~~~~~
280 281
281 ``struct btf_type`` encoding requirement: 282 ``struct btf_type`` encoding requirement:
282 * ``name_off``: 0 or offset to a valid C ide 283 * ``name_off``: 0 or offset to a valid C identifier
283 * ``info.kind_flag``: 0 for unsigned, 1 for !! 284 * ``info.kind_flag``: 0
284 * ``info.kind``: BTF_KIND_ENUM 285 * ``info.kind``: BTF_KIND_ENUM
285 * ``info.vlen``: number of enum values 286 * ``info.vlen``: number of enum values
286 * ``size``: 1/2/4/8 !! 287 * ``size``: 4
287 288
288 ``btf_type`` is followed by ``info.vlen`` numb 289 ``btf_type`` is followed by ``info.vlen`` number of ``struct btf_enum``.::
289 290
290 struct btf_enum { 291 struct btf_enum {
291 __u32 name_off; 292 __u32 name_off;
292 __s32 val; 293 __s32 val;
293 }; 294 };
294 295
295 The ``btf_enum`` encoding: 296 The ``btf_enum`` encoding:
296 * ``name_off``: offset to a valid C identifi 297 * ``name_off``: offset to a valid C identifier
297 * ``val``: any value 298 * ``val``: any value
298 299
299 If the original enum value is signed and the s <<
300 that value will be sign extended into 4 bytes. <<
301 the value will be truncated into 4 bytes. <<
302 <<
303 2.2.7 BTF_KIND_FWD 300 2.2.7 BTF_KIND_FWD
304 ~~~~~~~~~~~~~~~~~~ 301 ~~~~~~~~~~~~~~~~~~
305 302
306 ``struct btf_type`` encoding requirement: 303 ``struct btf_type`` encoding requirement:
307 * ``name_off``: offset to a valid C identifi 304 * ``name_off``: offset to a valid C identifier
308 * ``info.kind_flag``: 0 for struct, 1 for un 305 * ``info.kind_flag``: 0 for struct, 1 for union
309 * ``info.kind``: BTF_KIND_FWD 306 * ``info.kind``: BTF_KIND_FWD
310 * ``info.vlen``: 0 307 * ``info.vlen``: 0
311 * ``type``: 0 308 * ``type``: 0
312 309
313 No additional type data follow ``btf_type``. 310 No additional type data follow ``btf_type``.
314 311
315 2.2.8 BTF_KIND_TYPEDEF 312 2.2.8 BTF_KIND_TYPEDEF
316 ~~~~~~~~~~~~~~~~~~~~~~ 313 ~~~~~~~~~~~~~~~~~~~~~~
317 314
318 ``struct btf_type`` encoding requirement: 315 ``struct btf_type`` encoding requirement:
319 * ``name_off``: offset to a valid C identifi 316 * ``name_off``: offset to a valid C identifier
320 * ``info.kind_flag``: 0 317 * ``info.kind_flag``: 0
321 * ``info.kind``: BTF_KIND_TYPEDEF 318 * ``info.kind``: BTF_KIND_TYPEDEF
322 * ``info.vlen``: 0 319 * ``info.vlen``: 0
323 * ``type``: the type which can be referred b 320 * ``type``: the type which can be referred by name at ``name_off``
324 321
325 No additional type data follow ``btf_type``. 322 No additional type data follow ``btf_type``.
326 323
327 2.2.9 BTF_KIND_VOLATILE 324 2.2.9 BTF_KIND_VOLATILE
328 ~~~~~~~~~~~~~~~~~~~~~~~ 325 ~~~~~~~~~~~~~~~~~~~~~~~
329 326
330 ``struct btf_type`` encoding requirement: 327 ``struct btf_type`` encoding requirement:
331 * ``name_off``: 0 328 * ``name_off``: 0
332 * ``info.kind_flag``: 0 329 * ``info.kind_flag``: 0
333 * ``info.kind``: BTF_KIND_VOLATILE 330 * ``info.kind``: BTF_KIND_VOLATILE
334 * ``info.vlen``: 0 331 * ``info.vlen``: 0
335 * ``type``: the type with ``volatile`` quali 332 * ``type``: the type with ``volatile`` qualifier
336 333
337 No additional type data follow ``btf_type``. 334 No additional type data follow ``btf_type``.
338 335
339 2.2.10 BTF_KIND_CONST 336 2.2.10 BTF_KIND_CONST
340 ~~~~~~~~~~~~~~~~~~~~~ 337 ~~~~~~~~~~~~~~~~~~~~~
341 338
342 ``struct btf_type`` encoding requirement: 339 ``struct btf_type`` encoding requirement:
343 * ``name_off``: 0 340 * ``name_off``: 0
344 * ``info.kind_flag``: 0 341 * ``info.kind_flag``: 0
345 * ``info.kind``: BTF_KIND_CONST 342 * ``info.kind``: BTF_KIND_CONST
346 * ``info.vlen``: 0 343 * ``info.vlen``: 0
347 * ``type``: the type with ``const`` qualifie 344 * ``type``: the type with ``const`` qualifier
348 345
349 No additional type data follow ``btf_type``. 346 No additional type data follow ``btf_type``.
350 347
351 2.2.11 BTF_KIND_RESTRICT 348 2.2.11 BTF_KIND_RESTRICT
352 ~~~~~~~~~~~~~~~~~~~~~~~~ 349 ~~~~~~~~~~~~~~~~~~~~~~~~
353 350
354 ``struct btf_type`` encoding requirement: 351 ``struct btf_type`` encoding requirement:
355 * ``name_off``: 0 352 * ``name_off``: 0
356 * ``info.kind_flag``: 0 353 * ``info.kind_flag``: 0
357 * ``info.kind``: BTF_KIND_RESTRICT 354 * ``info.kind``: BTF_KIND_RESTRICT
358 * ``info.vlen``: 0 355 * ``info.vlen``: 0
359 * ``type``: the type with ``restrict`` quali 356 * ``type``: the type with ``restrict`` qualifier
360 357
361 No additional type data follow ``btf_type``. 358 No additional type data follow ``btf_type``.
362 359
363 2.2.12 BTF_KIND_FUNC 360 2.2.12 BTF_KIND_FUNC
364 ~~~~~~~~~~~~~~~~~~~~ 361 ~~~~~~~~~~~~~~~~~~~~
365 362
366 ``struct btf_type`` encoding requirement: 363 ``struct btf_type`` encoding requirement:
367 * ``name_off``: offset to a valid C identifi 364 * ``name_off``: offset to a valid C identifier
368 * ``info.kind_flag``: 0 365 * ``info.kind_flag``: 0
369 * ``info.kind``: BTF_KIND_FUNC 366 * ``info.kind``: BTF_KIND_FUNC
370 * ``info.vlen``: linkage information (BTF_FU !! 367 * ``info.vlen``: 0
371 or BTF_FUNC_EXTERN - see :r <<
372 * ``type``: a BTF_KIND_FUNC_PROTO type 368 * ``type``: a BTF_KIND_FUNC_PROTO type
373 369
374 No additional type data follow ``btf_type``. 370 No additional type data follow ``btf_type``.
375 371
376 A BTF_KIND_FUNC defines not a type, but a subp 372 A BTF_KIND_FUNC defines not a type, but a subprogram (function) whose
377 signature is defined by ``type``. The subprogr 373 signature is defined by ``type``. The subprogram is thus an instance of that
378 type. The BTF_KIND_FUNC may in turn be referen 374 type. The BTF_KIND_FUNC may in turn be referenced by a func_info in the
379 :ref:`BTF_Ext_Section` (ELF) or in the argumen 375 :ref:`BTF_Ext_Section` (ELF) or in the arguments to :ref:`BPF_Prog_Load`
380 (ABI). 376 (ABI).
381 377
382 Currently, only linkage values of BTF_FUNC_STA <<
383 supported in the kernel. <<
384 <<
385 2.2.13 BTF_KIND_FUNC_PROTO 378 2.2.13 BTF_KIND_FUNC_PROTO
386 ~~~~~~~~~~~~~~~~~~~~~~~~~~ 379 ~~~~~~~~~~~~~~~~~~~~~~~~~~
387 380
388 ``struct btf_type`` encoding requirement: 381 ``struct btf_type`` encoding requirement:
389 * ``name_off``: 0 382 * ``name_off``: 0
390 * ``info.kind_flag``: 0 383 * ``info.kind_flag``: 0
391 * ``info.kind``: BTF_KIND_FUNC_PROTO 384 * ``info.kind``: BTF_KIND_FUNC_PROTO
392 * ``info.vlen``: # of parameters 385 * ``info.vlen``: # of parameters
393 * ``type``: the return type 386 * ``type``: the return type
394 387
395 ``btf_type`` is followed by ``info.vlen`` numb 388 ``btf_type`` is followed by ``info.vlen`` number of ``struct btf_param``.::
396 389
397 struct btf_param { 390 struct btf_param {
398 __u32 name_off; 391 __u32 name_off;
399 __u32 type; 392 __u32 type;
400 }; 393 };
401 394
402 If a BTF_KIND_FUNC_PROTO type is referred by a 395 If a BTF_KIND_FUNC_PROTO type is referred by a BTF_KIND_FUNC type, then
403 ``btf_param.name_off`` must point to a valid C 396 ``btf_param.name_off`` must point to a valid C identifier except for the
404 possible last argument representing the variab 397 possible last argument representing the variable argument. The btf_param.type
405 refers to parameter type. 398 refers to parameter type.
406 399
407 If the function has variable arguments, the la 400 If the function has variable arguments, the last parameter is encoded with
408 ``name_off = 0`` and ``type = 0``. 401 ``name_off = 0`` and ``type = 0``.
409 402
410 2.2.14 BTF_KIND_VAR 403 2.2.14 BTF_KIND_VAR
411 ~~~~~~~~~~~~~~~~~~~ 404 ~~~~~~~~~~~~~~~~~~~
412 405
413 ``struct btf_type`` encoding requirement: 406 ``struct btf_type`` encoding requirement:
414 * ``name_off``: offset to a valid C identifi 407 * ``name_off``: offset to a valid C identifier
415 * ``info.kind_flag``: 0 408 * ``info.kind_flag``: 0
416 * ``info.kind``: BTF_KIND_VAR 409 * ``info.kind``: BTF_KIND_VAR
417 * ``info.vlen``: 0 410 * ``info.vlen``: 0
418 * ``type``: the type of the variable 411 * ``type``: the type of the variable
419 412
420 ``btf_type`` is followed by a single ``struct 413 ``btf_type`` is followed by a single ``struct btf_variable`` with the
421 following data:: 414 following data::
422 415
423 struct btf_var { 416 struct btf_var {
424 __u32 linkage; 417 __u32 linkage;
425 }; 418 };
426 419
427 ``btf_var.linkage`` may take the values: BTF_V !! 420 ``struct btf_var`` encoding:
428 see :ref:`BTF_Var_Linkage_Constants`. !! 421 * ``linkage``: currently only static variable 0, or globally allocated
>> 422 variable in ELF sections 1
429 423
430 Not all type of global variables are supported 424 Not all type of global variables are supported by LLVM at this point.
431 The following is currently available: 425 The following is currently available:
432 426
433 * static variables with or without section a 427 * static variables with or without section attributes
434 * global variables with section attributes 428 * global variables with section attributes
435 429
436 The latter is for future extraction of map key 430 The latter is for future extraction of map key/value type id's from a
437 map definition. 431 map definition.
438 432
439 2.2.15 BTF_KIND_DATASEC 433 2.2.15 BTF_KIND_DATASEC
440 ~~~~~~~~~~~~~~~~~~~~~~~ 434 ~~~~~~~~~~~~~~~~~~~~~~~
441 435
442 ``struct btf_type`` encoding requirement: 436 ``struct btf_type`` encoding requirement:
443 * ``name_off``: offset to a valid name assoc 437 * ``name_off``: offset to a valid name associated with a variable or
444 one of .data/.bss/.rodata 438 one of .data/.bss/.rodata
445 * ``info.kind_flag``: 0 439 * ``info.kind_flag``: 0
446 * ``info.kind``: BTF_KIND_DATASEC 440 * ``info.kind``: BTF_KIND_DATASEC
447 * ``info.vlen``: # of variables 441 * ``info.vlen``: # of variables
448 * ``size``: total section size in bytes (0 a 442 * ``size``: total section size in bytes (0 at compilation time, patched
449 to actual size by BPF loaders su 443 to actual size by BPF loaders such as libbpf)
450 444
451 ``btf_type`` is followed by ``info.vlen`` numb 445 ``btf_type`` is followed by ``info.vlen`` number of ``struct btf_var_secinfo``.::
452 446
453 struct btf_var_secinfo { 447 struct btf_var_secinfo {
454 __u32 type; 448 __u32 type;
455 __u32 offset; 449 __u32 offset;
456 __u32 size; 450 __u32 size;
457 }; 451 };
458 452
459 ``struct btf_var_secinfo`` encoding: 453 ``struct btf_var_secinfo`` encoding:
460 * ``type``: the type of the BTF_KIND_VAR var 454 * ``type``: the type of the BTF_KIND_VAR variable
461 * ``offset``: the in-section offset of the v 455 * ``offset``: the in-section offset of the variable
462 * ``size``: the size of the variable in byte 456 * ``size``: the size of the variable in bytes
463 457
464 2.2.16 BTF_KIND_FLOAT 458 2.2.16 BTF_KIND_FLOAT
465 ~~~~~~~~~~~~~~~~~~~~~ 459 ~~~~~~~~~~~~~~~~~~~~~
466 460
467 ``struct btf_type`` encoding requirement: 461 ``struct btf_type`` encoding requirement:
468 * ``name_off``: any valid offset 462 * ``name_off``: any valid offset
469 * ``info.kind_flag``: 0 463 * ``info.kind_flag``: 0
470 * ``info.kind``: BTF_KIND_FLOAT 464 * ``info.kind``: BTF_KIND_FLOAT
471 * ``info.vlen``: 0 465 * ``info.vlen``: 0
472 * ``size``: the size of the float type in byt 466 * ``size``: the size of the float type in bytes: 2, 4, 8, 12 or 16.
473 467
474 No additional type data follow ``btf_type``. 468 No additional type data follow ``btf_type``.
475 469
476 2.2.17 BTF_KIND_DECL_TAG 470 2.2.17 BTF_KIND_DECL_TAG
477 ~~~~~~~~~~~~~~~~~~~~~~~~ 471 ~~~~~~~~~~~~~~~~~~~~~~~~
478 472
479 ``struct btf_type`` encoding requirement: 473 ``struct btf_type`` encoding requirement:
480 * ``name_off``: offset to a non-empty string 474 * ``name_off``: offset to a non-empty string
481 * ``info.kind_flag``: 0 475 * ``info.kind_flag``: 0
482 * ``info.kind``: BTF_KIND_DECL_TAG 476 * ``info.kind``: BTF_KIND_DECL_TAG
483 * ``info.vlen``: 0 477 * ``info.vlen``: 0
484 * ``type``: ``struct``, ``union``, ``func``, 478 * ``type``: ``struct``, ``union``, ``func``, ``var`` or ``typedef``
485 479
486 ``btf_type`` is followed by ``struct btf_decl_ 480 ``btf_type`` is followed by ``struct btf_decl_tag``.::
487 481
488 struct btf_decl_tag { 482 struct btf_decl_tag {
489 __u32 component_idx; 483 __u32 component_idx;
490 }; 484 };
491 485
492 The ``name_off`` encodes btf_decl_tag attribut 486 The ``name_off`` encodes btf_decl_tag attribute string.
493 The ``type`` should be ``struct``, ``union``, 487 The ``type`` should be ``struct``, ``union``, ``func``, ``var`` or ``typedef``.
494 For ``var`` or ``typedef`` type, ``btf_decl_ta 488 For ``var`` or ``typedef`` type, ``btf_decl_tag.component_idx`` must be ``-1``.
495 For the other three types, if the btf_decl_tag 489 For the other three types, if the btf_decl_tag attribute is
496 applied to the ``struct``, ``union`` or ``func 490 applied to the ``struct``, ``union`` or ``func`` itself,
497 ``btf_decl_tag.component_idx`` must be ``-1``. 491 ``btf_decl_tag.component_idx`` must be ``-1``. Otherwise,
498 the attribute is applied to a ``struct``/``uni 492 the attribute is applied to a ``struct``/``union`` member or
499 a ``func`` argument, and ``btf_decl_tag.compon 493 a ``func`` argument, and ``btf_decl_tag.component_idx`` should be a
500 valid index (starting from 0) pointing to a me 494 valid index (starting from 0) pointing to a member or an argument.
501 495
502 2.2.18 BTF_KIND_TYPE_TAG !! 496 2.2.17 BTF_KIND_TYPE_TAG
503 ~~~~~~~~~~~~~~~~~~~~~~~~ 497 ~~~~~~~~~~~~~~~~~~~~~~~~
504 498
505 ``struct btf_type`` encoding requirement: 499 ``struct btf_type`` encoding requirement:
506 * ``name_off``: offset to a non-empty string 500 * ``name_off``: offset to a non-empty string
507 * ``info.kind_flag``: 0 501 * ``info.kind_flag``: 0
508 * ``info.kind``: BTF_KIND_TYPE_TAG 502 * ``info.kind``: BTF_KIND_TYPE_TAG
509 * ``info.vlen``: 0 503 * ``info.vlen``: 0
510 * ``type``: the type with ``btf_type_tag`` at 504 * ``type``: the type with ``btf_type_tag`` attribute
511 505
512 Currently, ``BTF_KIND_TYPE_TAG`` is only emitt 506 Currently, ``BTF_KIND_TYPE_TAG`` is only emitted for pointer types.
513 It has the following btf type chain: 507 It has the following btf type chain:
514 :: 508 ::
515 509
516 ptr -> [type_tag]* 510 ptr -> [type_tag]*
517 -> [const | volatile | restrict | typede 511 -> [const | volatile | restrict | typedef]*
518 -> base_type 512 -> base_type
519 513
520 Basically, a pointer type points to zero or mo 514 Basically, a pointer type points to zero or more
521 type_tag, then zero or more const/volatile/res 515 type_tag, then zero or more const/volatile/restrict/typedef
522 and finally the base type. The base type is on 516 and finally the base type. The base type is one of
523 int, ptr, array, struct, union, enum, func_pro 517 int, ptr, array, struct, union, enum, func_proto and float types.
524 518
525 2.2.19 BTF_KIND_ENUM64 <<
526 ~~~~~~~~~~~~~~~~~~~~~~ <<
527 <<
528 ``struct btf_type`` encoding requirement: <<
529 * ``name_off``: 0 or offset to a valid C ide <<
530 * ``info.kind_flag``: 0 for unsigned, 1 for <<
531 * ``info.kind``: BTF_KIND_ENUM64 <<
532 * ``info.vlen``: number of enum values <<
533 * ``size``: 1/2/4/8 <<
534 <<
535 ``btf_type`` is followed by ``info.vlen`` numb <<
536 <<
537 struct btf_enum64 { <<
538 __u32 name_off; <<
539 __u32 val_lo32; <<
540 __u32 val_hi32; <<
541 }; <<
542 <<
543 The ``btf_enum64`` encoding: <<
544 * ``name_off``: offset to a valid C identifi <<
545 * ``val_lo32``: lower 32-bit value for a 64- <<
546 * ``val_hi32``: high 32-bit value for a 64-b <<
547 <<
548 If the original enum value is signed and the s <<
549 that value will be sign extended into 8 bytes. <<
550 <<
551 2.3 Constant Values <<
552 ------------------- <<
553 <<
554 .. _BTF_Function_Linkage_Constants: <<
555 <<
556 2.3.1 Function Linkage Constant Values <<
557 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ <<
558 .. table:: Function Linkage Values and Meaning <<
559 <<
560 =================== ===== =========== <<
561 kind value description <<
562 =================== ===== =========== <<
563 ``BTF_FUNC_STATIC`` 0x0 definition of su <<
564 ``BTF_FUNC_GLOBAL`` 0x1 definition of su <<
565 ``BTF_FUNC_EXTERN`` 0x2 declaration of a <<
566 =================== ===== =========== <<
567 <<
568 <<
569 .. _BTF_Var_Linkage_Constants: <<
570 <<
571 2.3.2 Variable Linkage Constant Values <<
572 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ <<
573 .. table:: Variable Linkage Values and Meaning <<
574 <<
575 ============================ ===== ======= <<
576 kind value descrip <<
577 ============================ ===== ======= <<
578 ``BTF_VAR_STATIC`` 0x0 definit <<
579 ``BTF_VAR_GLOBAL_ALLOCATED`` 0x1 definit <<
580 ``BTF_VAR_GLOBAL_EXTERN`` 0x2 declara <<
581 ============================ ===== ======= <<
582 <<
583 3. BTF Kernel API 519 3. BTF Kernel API
584 ================= 520 =================
585 521
586 The following bpf syscall command involves BTF 522 The following bpf syscall command involves BTF:
587 * BPF_BTF_LOAD: load a blob of BTF data int 523 * BPF_BTF_LOAD: load a blob of BTF data into kernel
588 * BPF_MAP_CREATE: map creation with btf key 524 * BPF_MAP_CREATE: map creation with btf key and value type info.
589 * BPF_PROG_LOAD: prog load with btf functio 525 * BPF_PROG_LOAD: prog load with btf function and line info.
590 * BPF_BTF_GET_FD_BY_ID: get a btf fd 526 * BPF_BTF_GET_FD_BY_ID: get a btf fd
591 * BPF_OBJ_GET_INFO_BY_FD: btf, func_info, l 527 * BPF_OBJ_GET_INFO_BY_FD: btf, func_info, line_info
592 and other btf related info are returned. 528 and other btf related info are returned.
593 529
594 The workflow typically looks like: 530 The workflow typically looks like:
595 :: 531 ::
596 532
597 Application: 533 Application:
598 BPF_BTF_LOAD 534 BPF_BTF_LOAD
599 | 535 |
600 v 536 v
601 BPF_MAP_CREATE and BPF_PROG_LOAD 537 BPF_MAP_CREATE and BPF_PROG_LOAD
602 | 538 |
603 V 539 V
604 ...... 540 ......
605 541
606 Introspection tool: 542 Introspection tool:
607 ...... 543 ......
608 BPF_{PROG,MAP}_GET_NEXT_ID (get prog/map 544 BPF_{PROG,MAP}_GET_NEXT_ID (get prog/map id's)
609 | 545 |
610 V 546 V
611 BPF_{PROG,MAP}_GET_FD_BY_ID (get a prog/ 547 BPF_{PROG,MAP}_GET_FD_BY_ID (get a prog/map fd)
612 | 548 |
613 V 549 V
614 BPF_OBJ_GET_INFO_BY_FD (get bpf_prog_inf 550 BPF_OBJ_GET_INFO_BY_FD (get bpf_prog_info/bpf_map_info with btf_id)
615 | 551 | |
616 V 552 V |
617 BPF_BTF_GET_FD_BY_ID (get btf_fd) 553 BPF_BTF_GET_FD_BY_ID (get btf_fd) |
618 | 554 | |
619 V 555 V |
620 BPF_OBJ_GET_INFO_BY_FD (get btf) 556 BPF_OBJ_GET_INFO_BY_FD (get btf) |
621 | 557 | |
622 V 558 V V
623 pretty print types, dump func signatures 559 pretty print types, dump func signatures and line info, etc.
624 560
625 561
626 3.1 BPF_BTF_LOAD 562 3.1 BPF_BTF_LOAD
627 ---------------- 563 ----------------
628 564
629 Load a blob of BTF data into kernel. A blob of 565 Load a blob of BTF data into kernel. A blob of data, described in
630 :ref:`BTF_Type_String`, can be directly loaded 566 :ref:`BTF_Type_String`, can be directly loaded into the kernel. A ``btf_fd``
631 is returned to a userspace. 567 is returned to a userspace.
632 568
633 3.2 BPF_MAP_CREATE 569 3.2 BPF_MAP_CREATE
634 ------------------ 570 ------------------
635 571
636 A map can be created with ``btf_fd`` and speci 572 A map can be created with ``btf_fd`` and specified key/value type id.::
637 573
638 __u32 btf_fd; /* fd pointing to 574 __u32 btf_fd; /* fd pointing to a BTF type data */
639 __u32 btf_key_type_id; /* BTF typ 575 __u32 btf_key_type_id; /* BTF type_id of the key */
640 __u32 btf_value_type_id; /* BTF typ 576 __u32 btf_value_type_id; /* BTF type_id of the value */
641 577
642 In libbpf, the map can be defined with extra a 578 In libbpf, the map can be defined with extra annotation like below:
643 :: 579 ::
644 580
645 struct { 581 struct {
646 __uint(type, BPF_MAP_TYPE_ARRAY); 582 __uint(type, BPF_MAP_TYPE_ARRAY);
647 __type(key, int); 583 __type(key, int);
648 __type(value, struct ipv_counts); 584 __type(value, struct ipv_counts);
649 __uint(max_entries, 4); 585 __uint(max_entries, 4);
650 } btf_map SEC(".maps"); 586 } btf_map SEC(".maps");
651 587
652 During ELF parsing, libbpf is able to extract 588 During ELF parsing, libbpf is able to extract key/value type_id's and assign
653 them to BPF_MAP_CREATE attributes automaticall 589 them to BPF_MAP_CREATE attributes automatically.
654 590
655 .. _BPF_Prog_Load: 591 .. _BPF_Prog_Load:
656 592
657 3.3 BPF_PROG_LOAD 593 3.3 BPF_PROG_LOAD
658 ----------------- 594 -----------------
659 595
660 During prog_load, func_info and line_info can 596 During prog_load, func_info and line_info can be passed to kernel with proper
661 values for the following attributes: 597 values for the following attributes:
662 :: 598 ::
663 599
664 __u32 insn_cnt; 600 __u32 insn_cnt;
665 __aligned_u64 insns; 601 __aligned_u64 insns;
666 ...... 602 ......
667 __u32 prog_btf_fd; /* fd poin 603 __u32 prog_btf_fd; /* fd pointing to BTF type data */
668 __u32 func_info_rec_size; /* 604 __u32 func_info_rec_size; /* userspace bpf_func_info size */
669 __aligned_u64 func_info; /* func in 605 __aligned_u64 func_info; /* func info */
670 __u32 func_info_cnt; /* number 606 __u32 func_info_cnt; /* number of bpf_func_info records */
671 __u32 line_info_rec_size; /* 607 __u32 line_info_rec_size; /* userspace bpf_line_info size */
672 __aligned_u64 line_info; /* line in 608 __aligned_u64 line_info; /* line info */
673 __u32 line_info_cnt; /* number 609 __u32 line_info_cnt; /* number of bpf_line_info records */
674 610
675 The func_info and line_info are an array of be 611 The func_info and line_info are an array of below, respectively.::
676 612
677 struct bpf_func_info { 613 struct bpf_func_info {
678 __u32 insn_off; /* [0, insn_cnt - 1] 614 __u32 insn_off; /* [0, insn_cnt - 1] */
679 __u32 type_id; /* pointing to a BTF 615 __u32 type_id; /* pointing to a BTF_KIND_FUNC type */
680 }; 616 };
681 struct bpf_line_info { 617 struct bpf_line_info {
682 __u32 insn_off; /* [0, insn_cnt - 1] 618 __u32 insn_off; /* [0, insn_cnt - 1] */
683 __u32 file_name_off; /* offset to st 619 __u32 file_name_off; /* offset to string table for the filename */
684 __u32 line_off; /* offset to string 620 __u32 line_off; /* offset to string table for the source line */
685 __u32 line_col; /* line number and c 621 __u32 line_col; /* line number and column number */
686 }; 622 };
687 623
688 func_info_rec_size is the size of each func_in 624 func_info_rec_size is the size of each func_info record, and
689 line_info_rec_size is the size of each line_in 625 line_info_rec_size is the size of each line_info record. Passing the record
690 size to kernel make it possible to extend the 626 size to kernel make it possible to extend the record itself in the future.
691 627
692 Below are requirements for func_info: 628 Below are requirements for func_info:
693 * func_info[0].insn_off must be 0. 629 * func_info[0].insn_off must be 0.
694 * the func_info insn_off is in strictly incr 630 * the func_info insn_off is in strictly increasing order and matches
695 bpf func boundaries. 631 bpf func boundaries.
696 632
697 Below are requirements for line_info: 633 Below are requirements for line_info:
698 * the first insn in each func must have a li 634 * the first insn in each func must have a line_info record pointing to it.
699 * the line_info insn_off is in strictly incr 635 * the line_info insn_off is in strictly increasing order.
700 636
701 For line_info, the line number and column numb 637 For line_info, the line number and column number are defined as below:
702 :: 638 ::
703 639
704 #define BPF_LINE_INFO_LINE_NUM(line_col) 640 #define BPF_LINE_INFO_LINE_NUM(line_col) ((line_col) >> 10)
705 #define BPF_LINE_INFO_LINE_COL(line_col) 641 #define BPF_LINE_INFO_LINE_COL(line_col) ((line_col) & 0x3ff)
706 642
707 3.4 BPF_{PROG,MAP}_GET_NEXT_ID 643 3.4 BPF_{PROG,MAP}_GET_NEXT_ID
708 ------------------------------ 644 ------------------------------
709 645
710 In kernel, every loaded program, map or btf ha 646 In kernel, every loaded program, map or btf has a unique id. The id won't
711 change during the lifetime of a program, map, 647 change during the lifetime of a program, map, or btf.
712 648
713 The bpf syscall command BPF_{PROG,MAP}_GET_NEX 649 The bpf syscall command BPF_{PROG,MAP}_GET_NEXT_ID returns all id's, one for
714 each command, to user space, for bpf program o 650 each command, to user space, for bpf program or maps, respectively, so an
715 inspection tool can inspect all programs and m 651 inspection tool can inspect all programs and maps.
716 652
717 3.5 BPF_{PROG,MAP}_GET_FD_BY_ID 653 3.5 BPF_{PROG,MAP}_GET_FD_BY_ID
718 ------------------------------- 654 -------------------------------
719 655
720 An introspection tool cannot use id to get det 656 An introspection tool cannot use id to get details about program or maps.
721 A file descriptor needs to be obtained first f 657 A file descriptor needs to be obtained first for reference-counting purpose.
722 658
723 3.6 BPF_OBJ_GET_INFO_BY_FD 659 3.6 BPF_OBJ_GET_INFO_BY_FD
724 -------------------------- 660 --------------------------
725 661
726 Once a program/map fd is acquired, an introspe 662 Once a program/map fd is acquired, an introspection tool can get the detailed
727 information from kernel about this fd, some of 663 information from kernel about this fd, some of which are BTF-related. For
728 example, ``bpf_map_info`` returns ``btf_id`` a 664 example, ``bpf_map_info`` returns ``btf_id`` and key/value type ids.
729 ``bpf_prog_info`` returns ``btf_id``, func_inf 665 ``bpf_prog_info`` returns ``btf_id``, func_info, and line info for translated
730 bpf byte codes, and jited_line_info. 666 bpf byte codes, and jited_line_info.
731 667
732 3.7 BPF_BTF_GET_FD_BY_ID 668 3.7 BPF_BTF_GET_FD_BY_ID
733 ------------------------ 669 ------------------------
734 670
735 With ``btf_id`` obtained in ``bpf_map_info`` a 671 With ``btf_id`` obtained in ``bpf_map_info`` and ``bpf_prog_info``, bpf
736 syscall command BPF_BTF_GET_FD_BY_ID can retri 672 syscall command BPF_BTF_GET_FD_BY_ID can retrieve a btf fd. Then, with
737 command BPF_OBJ_GET_INFO_BY_FD, the btf blob, 673 command BPF_OBJ_GET_INFO_BY_FD, the btf blob, originally loaded into the
738 kernel with BPF_BTF_LOAD, can be retrieved. 674 kernel with BPF_BTF_LOAD, can be retrieved.
739 675
740 With the btf blob, ``bpf_map_info``, and ``bpf 676 With the btf blob, ``bpf_map_info``, and ``bpf_prog_info``, an introspection
741 tool has full btf knowledge and is able to pre 677 tool has full btf knowledge and is able to pretty print map key/values, dump
742 func signatures and line info, along with byte 678 func signatures and line info, along with byte/jit codes.
743 679
744 4. ELF File Format Interface 680 4. ELF File Format Interface
745 ============================ 681 ============================
746 682
747 4.1 .BTF section 683 4.1 .BTF section
748 ---------------- 684 ----------------
749 685
750 The .BTF section contains type and string data 686 The .BTF section contains type and string data. The format of this section is
751 same as the one describe in :ref:`BTF_Type_Str 687 same as the one describe in :ref:`BTF_Type_String`.
752 688
753 .. _BTF_Ext_Section: 689 .. _BTF_Ext_Section:
754 690
755 4.2 .BTF.ext section 691 4.2 .BTF.ext section
756 -------------------- 692 --------------------
757 693
758 The .BTF.ext section encodes func_info, line_i !! 694 The .BTF.ext section encodes func_info and line_info which needs loader
759 which needs loader manipulation before loading !! 695 manipulation before loading into the kernel.
760 696
761 The specification for .BTF.ext section is defi 697 The specification for .BTF.ext section is defined at ``tools/lib/bpf/btf.h``
762 and ``tools/lib/bpf/btf.c``. 698 and ``tools/lib/bpf/btf.c``.
763 699
764 The current header of .BTF.ext section:: 700 The current header of .BTF.ext section::
765 701
766 struct btf_ext_header { 702 struct btf_ext_header {
767 __u16 magic; 703 __u16 magic;
768 __u8 version; 704 __u8 version;
769 __u8 flags; 705 __u8 flags;
770 __u32 hdr_len; 706 __u32 hdr_len;
771 707
772 /* All offsets are in bytes relative t 708 /* All offsets are in bytes relative to the end of this header */
773 __u32 func_info_off; 709 __u32 func_info_off;
774 __u32 func_info_len; 710 __u32 func_info_len;
775 __u32 line_info_off; 711 __u32 line_info_off;
776 __u32 line_info_len; 712 __u32 line_info_len;
777 <<
778 /* optional part of .BTF.ext header */ <<
779 __u32 core_relo_off; <<
780 __u32 core_relo_len; <<
781 }; 713 };
782 714
783 It is very similar to .BTF section. Instead of 715 It is very similar to .BTF section. Instead of type/string section, it
784 contains func_info, line_info and core_relo su !! 716 contains func_info and line_info section. See :ref:`BPF_Prog_Load` for details
785 See :ref:`BPF_Prog_Load` for details about fun !! 717 about func_info and line_info record format.
786 record format. <<
787 718
788 The func_info is organized as below.:: 719 The func_info is organized as below.::
789 720
790 func_info_rec_size /* __u32 !! 721 func_info_rec_size
791 btf_ext_info_sec for section #1 /* func_i 722 btf_ext_info_sec for section #1 /* func_info for section #1 */
792 btf_ext_info_sec for section #2 /* func_i 723 btf_ext_info_sec for section #2 /* func_info for section #2 */
793 ... 724 ...
794 725
795 ``func_info_rec_size`` specifies the size of ` 726 ``func_info_rec_size`` specifies the size of ``bpf_func_info`` structure when
796 .BTF.ext is generated. ``btf_ext_info_sec``, d 727 .BTF.ext is generated. ``btf_ext_info_sec``, defined below, is a collection of
797 func_info for each specific ELF section.:: 728 func_info for each specific ELF section.::
798 729
799 struct btf_ext_info_sec { 730 struct btf_ext_info_sec {
800 __u32 sec_name_off; /* offset to sec 731 __u32 sec_name_off; /* offset to section name */
801 __u32 num_info; 732 __u32 num_info;
802 /* Followed by num_info * record_size 733 /* Followed by num_info * record_size number of bytes */
803 __u8 data[0]; 734 __u8 data[0];
804 }; 735 };
805 736
806 Here, num_info must be greater than 0. 737 Here, num_info must be greater than 0.
807 738
808 The line_info is organized as below.:: 739 The line_info is organized as below.::
809 740
810 line_info_rec_size /* __u32 !! 741 line_info_rec_size
811 btf_ext_info_sec for section #1 /* line_i 742 btf_ext_info_sec for section #1 /* line_info for section #1 */
812 btf_ext_info_sec for section #2 /* line_i 743 btf_ext_info_sec for section #2 /* line_info for section #2 */
813 ... 744 ...
814 745
815 ``line_info_rec_size`` specifies the size of ` 746 ``line_info_rec_size`` specifies the size of ``bpf_line_info`` structure when
816 .BTF.ext is generated. 747 .BTF.ext is generated.
817 748
818 The interpretation of ``bpf_func_info->insn_of 749 The interpretation of ``bpf_func_info->insn_off`` and
819 ``bpf_line_info->insn_off`` is different betwe 750 ``bpf_line_info->insn_off`` is different between kernel API and ELF API. For
820 kernel API, the ``insn_off`` is the instructio 751 kernel API, the ``insn_off`` is the instruction offset in the unit of ``struct
821 bpf_insn``. For ELF API, the ``insn_off`` is t 752 bpf_insn``. For ELF API, the ``insn_off`` is the byte offset from the
822 beginning of section (``btf_ext_info_sec->sec_ 753 beginning of section (``btf_ext_info_sec->sec_name_off``).
823 754
824 The core_relo is organized as below.:: <<
825 <<
826 core_relo_rec_size /* __u32 <<
827 btf_ext_info_sec for section #1 /* core_r <<
828 btf_ext_info_sec for section #2 /* core_r <<
829 <<
830 ``core_relo_rec_size`` specifies the size of ` <<
831 structure when .BTF.ext is generated. All ``bp <<
832 within a single ``btf_ext_info_sec`` describe <<
833 section named by ``btf_ext_info_sec->sec_name_ <<
834 <<
835 See :ref:`Documentation/bpf/llvm_reloc.rst <bt <<
836 for more information on CO-RE relocations. <<
837 <<
838 4.2 .BTF_ids section 755 4.2 .BTF_ids section
839 -------------------- 756 --------------------
840 757
841 The .BTF_ids section encodes BTF ID values tha 758 The .BTF_ids section encodes BTF ID values that are used within the kernel.
842 759
843 This section is created during the kernel comp 760 This section is created during the kernel compilation with the help of
844 macros defined in ``include/linux/btf_ids.h`` 761 macros defined in ``include/linux/btf_ids.h`` header file. Kernel code can
845 use them to create lists and sets (sorted list 762 use them to create lists and sets (sorted lists) of BTF ID values.
846 763
847 The ``BTF_ID_LIST`` and ``BTF_ID`` macros defi 764 The ``BTF_ID_LIST`` and ``BTF_ID`` macros define unsorted list of BTF ID values,
848 with following syntax:: 765 with following syntax::
849 766
850 BTF_ID_LIST(list) 767 BTF_ID_LIST(list)
851 BTF_ID(type1, name1) 768 BTF_ID(type1, name1)
852 BTF_ID(type2, name2) 769 BTF_ID(type2, name2)
853 770
854 resulting in following layout in .BTF_ids sect 771 resulting in following layout in .BTF_ids section::
855 772
856 __BTF_ID__type1__name1__1: 773 __BTF_ID__type1__name1__1:
857 .zero 4 774 .zero 4
858 __BTF_ID__type2__name2__2: 775 __BTF_ID__type2__name2__2:
859 .zero 4 776 .zero 4
860 777
861 The ``u32 list[];`` variable is defined to acc 778 The ``u32 list[];`` variable is defined to access the list.
862 779
863 The ``BTF_ID_UNUSED`` macro defines 4 zero byt 780 The ``BTF_ID_UNUSED`` macro defines 4 zero bytes. It's used when we
864 want to define unused entry in BTF_ID_LIST, li 781 want to define unused entry in BTF_ID_LIST, like::
865 782
866 BTF_ID_LIST(bpf_skb_output_btf_ids) 783 BTF_ID_LIST(bpf_skb_output_btf_ids)
867 BTF_ID(struct, sk_buff) 784 BTF_ID(struct, sk_buff)
868 BTF_ID_UNUSED 785 BTF_ID_UNUSED
869 BTF_ID(struct, task_struct) 786 BTF_ID(struct, task_struct)
870 787
871 The ``BTF_SET_START/END`` macros pair defines 788 The ``BTF_SET_START/END`` macros pair defines sorted list of BTF ID values
872 and their count, with following syntax:: 789 and their count, with following syntax::
873 790
874 BTF_SET_START(set) 791 BTF_SET_START(set)
875 BTF_ID(type1, name1) 792 BTF_ID(type1, name1)
876 BTF_ID(type2, name2) 793 BTF_ID(type2, name2)
877 BTF_SET_END(set) 794 BTF_SET_END(set)
878 795
879 resulting in following layout in .BTF_ids sect 796 resulting in following layout in .BTF_ids section::
880 797
881 __BTF_ID__set__set: 798 __BTF_ID__set__set:
882 .zero 4 799 .zero 4
883 __BTF_ID__type1__name1__3: 800 __BTF_ID__type1__name1__3:
884 .zero 4 801 .zero 4
885 __BTF_ID__type2__name2__4: 802 __BTF_ID__type2__name2__4:
886 .zero 4 803 .zero 4
887 804
888 The ``struct btf_id_set set;`` variable is def 805 The ``struct btf_id_set set;`` variable is defined to access the list.
889 806
890 The ``typeX`` name can be one of following:: 807 The ``typeX`` name can be one of following::
891 808
892 struct, union, typedef, func 809 struct, union, typedef, func
893 810
894 and is used as a filter when resolving the BTF 811 and is used as a filter when resolving the BTF ID value.
895 812
896 All the BTF ID lists and sets are compiled in 813 All the BTF ID lists and sets are compiled in the .BTF_ids section and
897 resolved during the linking phase of kernel bu 814 resolved during the linking phase of kernel build by ``resolve_btfids`` tool.
898 815
899 5. Using BTF 816 5. Using BTF
900 ============ 817 ============
901 818
902 5.1 bpftool map pretty print 819 5.1 bpftool map pretty print
903 ---------------------------- 820 ----------------------------
904 821
905 With BTF, the map key/value can be printed bas 822 With BTF, the map key/value can be printed based on fields rather than simply
906 raw bytes. This is especially valuable for lar 823 raw bytes. This is especially valuable for large structure or if your data
907 structure has bitfields. For example, for the 824 structure has bitfields. For example, for the following map,::
908 825
909 enum A { A1, A2, A3, A4, A5 }; 826 enum A { A1, A2, A3, A4, A5 };
910 typedef enum A ___A; 827 typedef enum A ___A;
911 struct tmp_t { 828 struct tmp_t {
912 char a1:4; 829 char a1:4;
913 int a2:4; 830 int a2:4;
914 int :4; 831 int :4;
915 __u32 a3:4; 832 __u32 a3:4;
916 int b; 833 int b;
917 ___A b1:4; 834 ___A b1:4;
918 enum A b2:4; 835 enum A b2:4;
919 }; 836 };
920 struct { 837 struct {
921 __uint(type, BPF_MAP_TYPE_ARRAY); 838 __uint(type, BPF_MAP_TYPE_ARRAY);
922 __type(key, int); 839 __type(key, int);
923 __type(value, struct tmp_t); 840 __type(value, struct tmp_t);
924 __uint(max_entries, 1); 841 __uint(max_entries, 1);
925 } tmpmap SEC(".maps"); 842 } tmpmap SEC(".maps");
926 843
927 bpftool is able to pretty print like below: 844 bpftool is able to pretty print like below:
928 :: 845 ::
929 846
930 [{ 847 [{
931 "key": 0, 848 "key": 0,
932 "value": { 849 "value": {
933 "a1": 0x2, 850 "a1": 0x2,
934 "a2": 0x4, 851 "a2": 0x4,
935 "a3": 0x6, 852 "a3": 0x6,
936 "b": 7, 853 "b": 7,
937 "b1": 0x8, 854 "b1": 0x8,
938 "b2": 0xa 855 "b2": 0xa
939 } 856 }
940 } 857 }
941 ] 858 ]
942 859
943 5.2 bpftool prog dump 860 5.2 bpftool prog dump
944 --------------------- 861 ---------------------
945 862
946 The following is an example showing how func_i 863 The following is an example showing how func_info and line_info can help prog
947 dump with better kernel symbol names, function 864 dump with better kernel symbol names, function prototypes and line
948 information.:: 865 information.::
949 866
950 $ bpftool prog dump jited pinned /sys/fs/b 867 $ bpftool prog dump jited pinned /sys/fs/bpf/test_btf_haskv
951 [...] 868 [...]
952 int test_long_fname_2(struct dummy_tracepo 869 int test_long_fname_2(struct dummy_tracepoint_args * arg):
953 bpf_prog_44a040bf25481309_test_long_fname_ 870 bpf_prog_44a040bf25481309_test_long_fname_2:
954 ; static int test_long_fname_2(struct dumm 871 ; static int test_long_fname_2(struct dummy_tracepoint_args *arg)
955 0: push %rbp 872 0: push %rbp
956 1: mov %rsp,%rbp 873 1: mov %rsp,%rbp
957 4: sub $0x30,%rsp 874 4: sub $0x30,%rsp
958 b: sub $0x28,%rbp 875 b: sub $0x28,%rbp
959 f: mov %rbx,0x0(%rbp) 876 f: mov %rbx,0x0(%rbp)
960 13: mov %r13,0x8(%rbp) 877 13: mov %r13,0x8(%rbp)
961 17: mov %r14,0x10(%rbp) 878 17: mov %r14,0x10(%rbp)
962 1b: mov %r15,0x18(%rbp) 879 1b: mov %r15,0x18(%rbp)
963 1f: xor %eax,%eax 880 1f: xor %eax,%eax
964 21: mov %rax,0x20(%rbp) 881 21: mov %rax,0x20(%rbp)
965 25: xor %esi,%esi 882 25: xor %esi,%esi
966 ; int key = 0; 883 ; int key = 0;
967 27: mov %esi,-0x4(%rbp) 884 27: mov %esi,-0x4(%rbp)
968 ; if (!arg->sock) 885 ; if (!arg->sock)
969 2a: mov 0x8(%rdi),%rdi 886 2a: mov 0x8(%rdi),%rdi
970 ; if (!arg->sock) 887 ; if (!arg->sock)
971 2e: cmp $0x0,%rdi 888 2e: cmp $0x0,%rdi
972 32: je 0x0000000000000070 889 32: je 0x0000000000000070
973 34: mov %rbp,%rsi 890 34: mov %rbp,%rsi
974 ; counts = bpf_map_lookup_elem(&btf_map, & 891 ; counts = bpf_map_lookup_elem(&btf_map, &key);
975 [...] 892 [...]
976 893
977 5.3 Verifier Log 894 5.3 Verifier Log
978 ---------------- 895 ----------------
979 896
980 The following is an example of how line_info c 897 The following is an example of how line_info can help debugging verification
981 failure.:: 898 failure.::
982 899
983 /* The code at tools/testing/selftests/ 900 /* The code at tools/testing/selftests/bpf/test_xdp_noinline.c
984 * is modified as below. 901 * is modified as below.
985 */ 902 */
986 data = (void *)(long)xdp->data; 903 data = (void *)(long)xdp->data;
987 data_end = (void *)(long)xdp->data_end; 904 data_end = (void *)(long)xdp->data_end;
988 /* 905 /*
989 if (data + 4 > data_end) 906 if (data + 4 > data_end)
990 return XDP_DROP; 907 return XDP_DROP;
991 */ 908 */
992 *(u32 *)data = dst->dst; 909 *(u32 *)data = dst->dst;
993 910
994 $ bpftool prog load ./test_xdp_noinline.o 911 $ bpftool prog load ./test_xdp_noinline.o /sys/fs/bpf/test_xdp_noinline type xdp
995 ; data = (void *)(long)xdp->data; 912 ; data = (void *)(long)xdp->data;
996 224: (79) r2 = *(u64 *)(r10 -112) 913 224: (79) r2 = *(u64 *)(r10 -112)
997 225: (61) r2 = *(u32 *)(r2 +0) 914 225: (61) r2 = *(u32 *)(r2 +0)
998 ; *(u32 *)data = dst->dst; 915 ; *(u32 *)data = dst->dst;
999 226: (63) *(u32 *)(r2 +0) = r1 916 226: (63) *(u32 *)(r2 +0) = r1
1000 invalid access to packet, off=0 size= 917 invalid access to packet, off=0 size=4, R2(id=0,off=0,r=0)
1001 R2 offset is outside of the packet 918 R2 offset is outside of the packet
1002 919
1003 6. BTF Generation 920 6. BTF Generation
1004 ================= 921 =================
1005 922
1006 You need latest pahole 923 You need latest pahole
1007 924
1008 https://git.kernel.org/pub/scm/devel/pahole 925 https://git.kernel.org/pub/scm/devel/pahole/pahole.git/
1009 926
1010 or llvm (8.0 or later). The pahole acts as a 927 or llvm (8.0 or later). The pahole acts as a dwarf2btf converter. It doesn't
1011 support .BTF.ext and btf BTF_KIND_FUNC type y 928 support .BTF.ext and btf BTF_KIND_FUNC type yet. For example,::
1012 929
1013 -bash-4.4$ cat t.c 930 -bash-4.4$ cat t.c
1014 struct t { 931 struct t {
1015 int a:2; 932 int a:2;
1016 int b:3; 933 int b:3;
1017 int c:2; 934 int c:2;
1018 } g; 935 } g;
1019 -bash-4.4$ gcc -c -O2 -g t.c 936 -bash-4.4$ gcc -c -O2 -g t.c
1020 -bash-4.4$ pahole -JV t.o 937 -bash-4.4$ pahole -JV t.o
1021 File t.o: 938 File t.o:
1022 [1] STRUCT t kind_flag=1 size=4 vlen=3 939 [1] STRUCT t kind_flag=1 size=4 vlen=3
1023 a type_id=2 bitfield_size=2 bit 940 a type_id=2 bitfield_size=2 bits_offset=0
1024 b type_id=2 bitfield_size=3 bit 941 b type_id=2 bitfield_size=3 bits_offset=2
1025 c type_id=2 bitfield_size=2 bit 942 c type_id=2 bitfield_size=2 bits_offset=5
1026 [2] INT int size=4 bit_offset=0 nr_bits 943 [2] INT int size=4 bit_offset=0 nr_bits=32 encoding=SIGNED
1027 944
1028 The llvm is able to generate .BTF and .BTF.ex 945 The llvm is able to generate .BTF and .BTF.ext directly with -g for bpf target
1029 only. The assembly code (-S) is able to show 946 only. The assembly code (-S) is able to show the BTF encoding in assembly
1030 format.:: 947 format.::
1031 948
1032 -bash-4.4$ cat t2.c 949 -bash-4.4$ cat t2.c
1033 typedef int __int32; 950 typedef int __int32;
1034 struct t2 { 951 struct t2 {
1035 int a2; 952 int a2;
1036 int (*f2)(char q1, __int32 q2, ...); 953 int (*f2)(char q1, __int32 q2, ...);
1037 int (*f3)(); 954 int (*f3)();
1038 } g2; 955 } g2;
1039 int main() { return 0; } 956 int main() { return 0; }
1040 int test() { return 0; } 957 int test() { return 0; }
1041 -bash-4.4$ clang -c -g -O2 --target=bpf t !! 958 -bash-4.4$ clang -c -g -O2 -target bpf t2.c
1042 -bash-4.4$ readelf -S t2.o 959 -bash-4.4$ readelf -S t2.o
1043 ...... 960 ......
1044 [ 8] .BTF PROGBITS 961 [ 8] .BTF PROGBITS 0000000000000000 00000247
1045 000000000000016e 0000000000000000 962 000000000000016e 0000000000000000 0 0 1
1046 [ 9] .BTF.ext PROGBITS 963 [ 9] .BTF.ext PROGBITS 0000000000000000 000003b5
1047 0000000000000060 0000000000000000 964 0000000000000060 0000000000000000 0 0 1
1048 [10] .rel.BTF.ext REL 965 [10] .rel.BTF.ext REL 0000000000000000 000007e0
1049 0000000000000040 0000000000000010 966 0000000000000040 0000000000000010 16 9 8
1050 ...... 967 ......
1051 -bash-4.4$ clang -S -g -O2 --target=bpf t !! 968 -bash-4.4$ clang -S -g -O2 -target bpf t2.c
1052 -bash-4.4$ cat t2.s 969 -bash-4.4$ cat t2.s
1053 ...... 970 ......
1054 .section .BTF,"",@progbits 971 .section .BTF,"",@progbits
1055 .short 60319 # 972 .short 60319 # 0xeb9f
1056 .byte 1 973 .byte 1
1057 .byte 0 974 .byte 0
1058 .long 24 975 .long 24
1059 .long 0 976 .long 0
1060 .long 220 977 .long 220
1061 .long 220 978 .long 220
1062 .long 122 979 .long 122
1063 .long 0 # 980 .long 0 # BTF_KIND_FUNC_PROTO(id = 1)
1064 .long 218103808 # 981 .long 218103808 # 0xd000000
1065 .long 2 982 .long 2
1066 .long 83 # 983 .long 83 # BTF_KIND_INT(id = 2)
1067 .long 16777216 # 984 .long 16777216 # 0x1000000
1068 .long 4 985 .long 4
1069 .long 16777248 # 986 .long 16777248 # 0x1000020
1070 ...... 987 ......
1071 .byte 0 # 988 .byte 0 # string offset=0
1072 .ascii ".text" # 989 .ascii ".text" # string offset=1
1073 .byte 0 990 .byte 0
1074 .ascii "/home/yhs/tmp-pahole/t2. 991 .ascii "/home/yhs/tmp-pahole/t2.c" # string offset=7
1075 .byte 0 992 .byte 0
1076 .ascii "int main() { return 0; } 993 .ascii "int main() { return 0; }" # string offset=33
1077 .byte 0 994 .byte 0
1078 .ascii "int test() { return 0; } 995 .ascii "int test() { return 0; }" # string offset=58
1079 .byte 0 996 .byte 0
1080 .ascii "int" # 997 .ascii "int" # string offset=83
1081 ...... 998 ......
1082 .section .BTF.ext,"",@prog 999 .section .BTF.ext,"",@progbits
1083 .short 60319 # 1000 .short 60319 # 0xeb9f
1084 .byte 1 1001 .byte 1
1085 .byte 0 1002 .byte 0
1086 .long 24 1003 .long 24
1087 .long 0 1004 .long 0
1088 .long 28 1005 .long 28
1089 .long 28 1006 .long 28
1090 .long 44 1007 .long 44
1091 .long 8 # 1008 .long 8 # FuncInfo
1092 .long 1 # 1009 .long 1 # FuncInfo section string offset=1
1093 .long 2 1010 .long 2
1094 .long .Lfunc_begin0 1011 .long .Lfunc_begin0
1095 .long 3 1012 .long 3
1096 .long .Lfunc_begin1 1013 .long .Lfunc_begin1
1097 .long 5 1014 .long 5
1098 .long 16 # 1015 .long 16 # LineInfo
1099 .long 1 # 1016 .long 1 # LineInfo section string offset=1
1100 .long 2 1017 .long 2
1101 .long .Ltmp0 1018 .long .Ltmp0
1102 .long 7 1019 .long 7
1103 .long 33 1020 .long 33
1104 .long 7182 # 1021 .long 7182 # Line 7 Col 14
1105 .long .Ltmp3 1022 .long .Ltmp3
1106 .long 7 1023 .long 7
1107 .long 58 1024 .long 58
1108 .long 8206 # 1025 .long 8206 # Line 8 Col 14
1109 1026
1110 7. Testing 1027 7. Testing
1111 ========== 1028 ==========
1112 1029
1113 The kernel BPF selftest `tools/testing/selfte !! 1030 Kernel bpf selftest `test_btf.c` provides extensive set of BTF-related tests.
1114 provides an extensive set of BTF-related test <<
1115 <<
1116 .. Links <<
1117 .. _tools/testing/selftests/bpf/prog_tests/bt <<
1118 https://git.kernel.org/pub/scm/linux/kerne <<
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.