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