Clarify trace byte_order
[ctf.git] / common-trace-format-specification.txt
CommitLineData
f5cb0e57 1Common Trace Format (CTF) Specification (v1.8.2)
5ba9f198
MD
2
3Mathieu Desnoyers, EfficiOS Inc.
4
339a7dde 5The goal of the present document is to specify a trace format that suits the
cc089c3a 6needs of the embedded, telecom, high-performance and kernel communities. It is
5ba9f198 7based on the Common Trace Format Requirements (v1.4) document. It is designed to
cc089c3a 8allow traces to be natively generated by the Linux kernel, Linux user-space
6c7226e9
MD
9applications written in C/C++, and hardware components. One major element of
10CTF is the Trace Stream Description Language (TSDL) which flexibility
11enables description of various binary trace stream layouts.
cc089c3a
MD
12
13The latest version of this document can be found at:
14
15 git tree: git://git.efficios.com/ctf.git
16 gitweb: http://git.efficios.com/?p=ctf.git
5ba9f198
MD
17
18A reference implementation of a library to read and write this trace format is
19being implemented within the BabelTrace project, a converter between trace
20formats. The development tree is available at:
21
22 git tree: git://git.efficios.com/babeltrace.git
23 gitweb: http://git.efficios.com/?p=babeltrace.git
24
339a7dde
MD
25The CE Workgroup of the Linux Foundation, Ericsson, and EfficiOS have
26sponsored this work.
27
5ba9f198 28
beabf088
MD
29Table of Contents
30
311. Preliminary definitions
322. High-level representation of a trace
333. Event stream
344. Types
35 4.1 Basic types
36 4.1.1 Type inheritance
37 4.1.2 Alignment
38 4.1.3 Byte order
39 4.1.4 Size
40 4.1.5 Integers
41 4.1.6 GNU/C bitfields
42 4.1.7 Floating point
43 4.1.8 Enumerations
444.2 Compound types
45 4.2.1 Structures
46 4.2.2 Variants (Discriminated/Tagged Unions)
47 4.2.3 Arrays
48 4.2.4 Sequences
49 4.2.5 Strings
505. Event Packet Header
51 5.1 Event Packet Header Description
52 5.2 Event Packet Context Description
536. Event Structure
54 6.1 Event Header
55 6.1.1 Type 1 - Few event IDs
56 6.1.2 Type 2 - Many event IDs
57 6.2 Event Context
58 6.3 Event Payload
59 6.3.1 Padding
60 6.3.2 Alignment
617. Trace Stream Description Language (TSDL)
62 7.1 Meta-data
63 7.2 Declaration vs Definition
64 7.3 TSDL Scopes
65 7.3.1 Lexical Scope
37ab95c3 66 7.3.2 Static and Dynamic Scopes
beabf088 67 7.4 TSDL Examples
2fa70eba 688. Clocks
beabf088
MD
69
70
5ba9f198
MD
711. Preliminary definitions
72
3bf79539
MD
73 - Event Trace: An ordered sequence of events.
74 - Event Stream: An ordered sequence of events, containing a subset of the
75 trace event types.
76 - Event Packet: A sequence of physically contiguous events within an event
77 stream.
5ba9f198
MD
78 - Event: This is the basic entry in a trace. (aka: a trace record).
79 - An event identifier (ID) relates to the class (a type) of event within
3bf79539
MD
80 an event stream.
81 e.g. event: irq_entry.
5ba9f198
MD
82 - An event (or event record) relates to a specific instance of an event
83 class.
3bf79539
MD
84 e.g. event: irq_entry, at time X, on CPU Y
85 - Source Architecture: Architecture writing the trace.
86 - Reader Architecture: Architecture reading the trace.
5ba9f198
MD
87
88
892. High-level representation of a trace
90
3bf79539
MD
91A trace is divided into multiple event streams. Each event stream contains a
92subset of the trace event types.
5ba9f198 93
3bf79539
MD
94The final output of the trace, after its generation and optional transport over
95the network, is expected to be either on permanent or temporary storage in a
96virtual file system. Because each event stream is appended to while a trace is
32254806
MD
97being recorded, each is associated with a distinct set of files for
98output. Therefore, a stored trace can be represented as a directory
99containing zero, one or more files per stream.
5ba9f198 100
6672e9e1
MD
101Meta-data description associated with the trace contains information on
102trace event types expressed in the Trace Stream Description Language
103(TSDL). This language describes:
5ba9f198
MD
104
105- Trace version.
106- Types available.
6672e9e1 107- Per-trace event header description.
3bf79539 108- Per-stream event header description.
6672e9e1 109- Per-stream event context description.
5ba9f198 110- Per-event
3bf79539 111 - Event type to stream mapping.
5ba9f198
MD
112 - Event type to name mapping.
113 - Event type to ID mapping.
6672e9e1 114 - Event context description.
5ba9f198
MD
115 - Event fields description.
116
117
3bf79539 1183. Event stream
5ba9f198 119
6672e9e1 120An event stream can be divided into contiguous event packets of variable
12c9069f
MD
121size. An event packet can contain a certain amount of padding at the
122end. The stream header is repeated at the beginning of each event
123packet. The rationale for the event stream design choices is explained
124in Appendix B. Stream Header Rationale.
5ba9f198 125
3bf79539
MD
126The event stream header will therefore be referred to as the "event packet
127header" throughout the rest of this document.
5ba9f198
MD
128
129
1304. Types
131
1fad7a85
MD
132Types are organized as type classes. Each type class belong to either of two
133kind of types: basic types or compound types.
134
5ba9f198
MD
1354.1 Basic types
136
1fad7a85
MD
137A basic type is a scalar type, as described in this section. It includes
138integers, GNU/C bitfields, enumerations, and floating point values.
5ba9f198
MD
139
1404.1.1 Type inheritance
141
80fd2569
MD
142Type specifications can be inherited to allow deriving types from a
143type class. For example, see the uint32_t named type derived from the "integer"
144type class below ("Integers" section). Types have a precise binary
145representation in the trace. A type class has methods to read and write these
146types, but must be derived into a type to be usable in an event field.
5ba9f198
MD
147
1484.1.2 Alignment
149
150We define "byte-packed" types as aligned on the byte size, namely 8-bit.
151We define "bit-packed" types as following on the next bit, as defined by the
370eae99 152"Integers" section.
5ba9f198 153
6672e9e1 154Each basic type must specify its alignment, in bits. Examples of
ec4404a7
MD
155possible alignments are: bit-packed (align = 1), byte-packed (align =
1568), or word-aligned (e.g. align = 32 or align = 64). The choice depends
157on the architecture preference and compactness vs performance trade-offs
158of the implementation. Architectures providing fast unaligned write
159byte-packed basic types to save space, aligning each type on byte
160boundaries (8-bit). Architectures with slow unaligned writes align types
161on specific alignment values. If no specific alignment is declared for a
162type, it is assumed to be bit-packed for integers with size not multiple
163of 8 bits and for gcc bitfields. All other basic types are byte-packed
164by default. It is however recommended to always specify the alignment
165explicitly. Alignment values must be power of two. Compound types are
166aligned as specified in their individual specification.
5ba9f198 167
6672e9e1 168TSDL meta-data attribute representation of a specific alignment:
5ba9f198
MD
169
170 align = value; /* value in bits */
171
1724.1.3 Byte order
173
9f296be4 174By default, the native endianness of the source architecture is used.
3bf79539
MD
175Byte order can be overridden for a basic type by specifying a "byte_order"
176attribute. Typical use-case is to specify the network byte order (big endian:
177"be") to save data captured from the network into the trace without conversion.
178If not specified, the byte order is native.
5ba9f198 179
6672e9e1 180TSDL meta-data representation:
5ba9f198
MD
181
182 byte_order = native OR network OR be OR le; /* network and be are aliases */
183
7e58b8bf
MD
184Even though the trace description section is not per se a type, for sake
185of clarity, it should be noted that native and network byte orders are
186only allowed within type declaration. The byte_order specified in the
187trace description section only accepts be OR le values.
188
5ba9f198
MD
1894.1.4 Size
190
191Type size, in bits, for integers and floats is that returned by "sizeof()" in C
192multiplied by CHAR_BIT.
193We require the size of "char" and "unsigned char" types (CHAR_BIT) to be fixed
194to 8 bits for cross-endianness compatibility.
195
6672e9e1 196TSDL meta-data representation:
5ba9f198
MD
197
198 size = value; (value is in bits)
199
2004.1.5 Integers
201
6672e9e1
MD
202Signed integers are represented in two-complement. Integer alignment,
203size, signedness and byte ordering are defined in the TSDL meta-data.
204Integers aligned on byte size (8-bit) and with length multiple of byte
205size (8-bit) correspond to the C99 standard integers. In addition,
206integers with alignment and/or size that are _not_ a multiple of the
207byte size are permitted; these correspond to the C99 standard bitfields,
208with the added specification that the CTF integer bitfields have a fixed
209binary representation. A MIT-licensed reference implementation of the
210CTF portable bitfields is available at:
5ba9f198
MD
211
212 http://git.efficios.com/?p=babeltrace.git;a=blob;f=include/babeltrace/bitfield.h
213
214Binary representation of integers:
215
216- On little and big endian:
217 - Within a byte, high bits correspond to an integer high bits, and low bits
218 correspond to low bits.
219- On little endian:
220 - Integer across multiple bytes are placed from the less significant to the
221 most significant.
222 - Consecutive integers are placed from lower bits to higher bits (even within
223 a byte).
224- On big endian:
225 - Integer across multiple bytes are placed from the most significant to the
226 less significant.
227 - Consecutive integers are placed from higher bits to lower bits (even within
228 a byte).
229
230This binary representation is derived from the bitfield implementation in GCC
231for little and big endian. However, contrary to what GCC does, integers can
6672e9e1 232cross units boundaries (no padding is required). Padding can be explicitly
5ba9f198
MD
233added (see 4.1.6 GNU/C bitfields) to follow the GCC layout if needed.
234
6672e9e1 235TSDL meta-data representation:
5ba9f198 236
80fd2569 237 integer {
5ba9f198
MD
238 signed = true OR false; /* default false */
239 byte_order = native OR network OR be OR le; /* default native */
240 size = value; /* value in bits, no default */
241 align = value; /* value in bits */
56b77c06
MD
242 /* based used for pretty-printing output, default: decimal. */
243 base = decimal OR dec OR OR d OR i OR u OR 10 OR hexadecimal OR hex OR x OR X OR p OR 16
244 OR octal OR oct OR o OR 8 OR binary OR b OR 2;
13441bf5
MD
245 /* character encoding, default: none */
246 encoding = none or UTF8 or ASCII;
2152348f 247 }
5ba9f198 248
80fd2569 249Example of type inheritance (creation of a uint32_t named type):
5ba9f198 250
359894ac 251typealias integer {
9e4e34e9 252 size = 32;
5ba9f198
MD
253 signed = false;
254 align = 32;
38b8da21 255} := uint32_t;
5ba9f198 256
80fd2569 257Definition of a named 5-bit signed bitfield:
5ba9f198 258
359894ac 259typealias integer {
5ba9f198
MD
260 size = 5;
261 signed = true;
262 align = 1;
38b8da21 263} := int5_t;
5ba9f198 264
13441bf5
MD
265The character encoding field can be used to specify that the integer
266must be printed as a text character when read. e.g.:
267
268typealias integer {
269 size = 8;
270 align = 8;
271 signed = false;
272 encoding = UTF8;
273} := utf_char;
274
275
5ba9f198
MD
2764.1.6 GNU/C bitfields
277
278The GNU/C bitfields follow closely the integer representation, with a
279particularity on alignment: if a bitfield cannot fit in the current unit, the
80fd2569
MD
280unit is padded and the bitfield starts at the following unit. The unit size is
281defined by the size of the type "unit_type".
5ba9f198 282
6672e9e1 283TSDL meta-data representation:
80fd2569 284
d674f4b8 285 unit_type name:size;
80fd2569 286
5ba9f198
MD
287As an example, the following structure declared in C compiled by GCC:
288
289struct example {
290 short a:12;
291 short b:5;
292};
293
2152348f
MD
294The example structure is aligned on the largest element (short). The second
295bitfield would be aligned on the next unit boundary, because it would not fit in
296the current unit.
5ba9f198
MD
297
2984.1.7 Floating point
299
6672e9e1 300The floating point values byte ordering is defined in the TSDL meta-data.
5ba9f198
MD
301
302Floating point values follow the IEEE 754-2008 standard interchange formats.
303Description of the floating point values include the exponent and mantissa size
304in bits. Some requirements are imposed on the floating point values:
305
306- FLT_RADIX must be 2.
307- mant_dig is the number of digits represented in the mantissa. It is specified
308 by the ISO C99 standard, section 5.2.4, as FLT_MANT_DIG, DBL_MANT_DIG and
309 LDBL_MANT_DIG as defined by <float.h>.
310- exp_dig is the number of digits represented in the exponent. Given that
311 mant_dig is one bit more than its actual size in bits (leading 1 is not
312 needed) and also given that the sign bit always takes one bit, exp_dig can be
313 specified as:
314
315 - sizeof(float) * CHAR_BIT - FLT_MANT_DIG
316 - sizeof(double) * CHAR_BIT - DBL_MANT_DIG
317 - sizeof(long double) * CHAR_BIT - LDBL_MANT_DIG
318
6672e9e1 319TSDL meta-data representation:
5ba9f198 320
80fd2569 321floating_point {
ec4404a7
MD
322 exp_dig = value;
323 mant_dig = value;
324 byte_order = native OR network OR be OR le;
325 align = value;
2152348f 326}
5ba9f198
MD
327
328Example of type inheritance:
329
359894ac 330typealias floating_point {
5ba9f198
MD
331 exp_dig = 8; /* sizeof(float) * CHAR_BIT - FLT_MANT_DIG */
332 mant_dig = 24; /* FLT_MANT_DIG */
333 byte_order = native;
ec4404a7 334 align = 32;
38b8da21 335} := float;
5ba9f198
MD
336
337TODO: define NaN, +inf, -inf behavior.
338
ec4404a7
MD
339Bit-packed, byte-packed or larger alignments can be used for floating
340point values, similarly to integers.
341
5ba9f198
MD
3424.1.8 Enumerations
343
344Enumerations are a mapping between an integer type and a table of strings. The
345numerical representation of the enumeration follows the integer type specified
6672e9e1
MD
346by the meta-data. The enumeration mapping table is detailed in the enumeration
347description within the meta-data. The mapping table maps inclusive value
348ranges (or single values) to strings. Instead of being limited to simple
3bf79539 349"value -> string" mappings, these enumerations map
80fd2569 350"[ start_value ... end_value ] -> string", which map inclusive ranges of
3bf79539
MD
351values to strings. An enumeration from the C language can be represented in
352this format by having the same start_value and end_value for each element, which
353is in fact a range of size 1. This single-value range is supported without
4767a9e7 354repeating the start and end values with the value = string declaration.
80fd2569 355
a9b83695 356enum name : integer_type {
359894ac 357 somestring = start_value1 ... end_value1,
80fd2569
MD
358 "other string" = start_value2 ... end_value2,
359 yet_another_string, /* will be assigned to end_value2 + 1 */
360 "some other string" = value,
361 ...
362};
363
364If the values are omitted, the enumeration starts at 0 and increment of 1 for
0cedf6e5
MD
365each entry. An entry with omitted value that follows a range entry takes
366as value the end_value of the previous range + 1:
80fd2569 367
a9b83695 368enum name : unsigned int {
80fd2569
MD
369 ZERO,
370 ONE,
371 TWO,
372 TEN = 10,
373 ELEVEN,
3bf79539 374};
5ba9f198 375
80fd2569 376Overlapping ranges within a single enumeration are implementation defined.
5ba9f198 377
2152348f
MD
378A nameless enumeration can be declared as a field type or as part of a typedef:
379
a9b83695 380enum : integer_type {
2152348f
MD
381 ...
382}
383
c2742c56
MD
384Enumerations omitting the container type ": integer_type" use the "int"
385type (for compatibility with C99). The "int" type must be previously
386declared. E.g.:
387
388typealias integer { size = 32; align = 32; signed = true } := int;
389
390enum {
391 ...
392}
393
1fad7a85 394
5ba9f198
MD
3954.2 Compound types
396
1fad7a85
MD
397Compound are aggregation of type declarations. Compound types include
398structures, variant, arrays, sequences, and strings.
399
5ba9f198
MD
4004.2.1 Structures
401
402Structures are aligned on the largest alignment required by basic types
403contained within the structure. (This follows the ISO/C standard for structures)
404
6672e9e1 405TSDL meta-data representation of a named structure:
5ba9f198 406
80fd2569
MD
407struct name {
408 field_type field_name;
409 field_type field_name;
410 ...
411};
5ba9f198
MD
412
413Example:
414
80fd2569
MD
415struct example {
416 integer { /* Nameless type */
417 size = 16;
418 signed = true;
419 align = 16;
420 } first_field_name;
6672e9e1 421 uint64_t second_field_name; /* Named type declared in the meta-data */
3bf79539 422};
5ba9f198 423
4cac83ee
MD
424The fields are placed in a sequence next to each other. They each
425possess a field name, which is a unique identifier within the structure.
426The identifier is not allowed to use any reserved keyword
427(see Section C.1.2). Replacing reserved keywords with
70375f92 428underscore-prefixed field names is recommended. Fields starting with an
92250c71
MD
429underscore should have their leading underscore removed by the CTF trace
430readers.
5ba9f198 431
2152348f 432A nameless structure can be declared as a field type or as part of a typedef:
80fd2569
MD
433
434struct {
435 ...
2152348f 436}
80fd2569 437
ec4404a7
MD
438Alignment for a structure compound type can be forced to a minimum value
439by adding an "align" specifier after the declaration of a structure
440body. This attribute is read as: align(value). The value is specified in
441bits. The structure will be aligned on the maximum value between this
442attribute and the alignment required by the basic types contained within
443the structure. e.g.
444
445struct {
446 ...
447} align(32)
448
77a98c82 4494.2.2 Variants (Discriminated/Tagged Unions)
fcba70d4 450
fdf2bb05
MD
451A CTF variant is a selection between different types. A CTF variant must
452always be defined within the scope of a structure or within fields
453contained within a structure (defined recursively). A "tag" enumeration
37ab95c3
MD
454field must appear in either the same static scope, prior to the variant
455field (in field declaration order), in an upper static scope , or in an
456upper dynamic scope (see Section 7.3.2). The type selection is indicated
457by the mapping from the enumeration value to the string used as variant
458type selector. The field to use as tag is specified by the "tag_field",
459specified between "< >" after the "variant" keyword for unnamed
460variants, and after "variant name" for named variants.
fcba70d4 461
0bed08fc
MD
462The alignment of the variant is the alignment of the type as selected by
463the tag value for the specific instance of the variant. The size of the
464variant is the size as selected by the tag value for the specific
465instance of the variant.
466
467The alignment of the type containing the variant is independent of the
468variant alignment. For instance, if a structure contains two fields, a
46932-bit integer, aligned on 32 bits, and a variant, which contains two
470choices: either a 32-bit field, aligned on 32 bits, or a 64-bit field,
471aligned on 64 bits, the alignment of the outmost structure will be
47232-bit (the alignment of its largest field, disregarding the alignment
473of the variant). The alignment of the variant will depend on the
474selector: if the variant's 32-bit field is selected, its alignment will
475be 32-bit, or 64-bit otherwise. It is important to note that variants
476are specifically tailored for compactness in a stream. Therefore, the
477relative offsets of compound type fields can vary depending on
478the offset at which the compound type starts if it contains a variant
479that itself contains a type with alignment larger than the largest field
480contained within the compound type. This is caused by the fact that the
481compound type may contain the enumeration that select the variant's
482choice, and therefore the alignment to be applied to the compound type
483cannot be determined before encountering the enumeration.
fcba70d4 484
4cac83ee
MD
485Each variant type selector possess a field name, which is a unique
486identifier within the variant. The identifier is not allowed to use any
487reserved keyword (see Section C.1.2). Replacing reserved keywords with
70375f92 488underscore-prefixed field names is recommended. Fields starting with an
92250c71
MD
489underscore should have their leading underscore removed by the CTF trace
490readers.
70375f92 491
4cac83ee 492
fcba70d4
MD
493A named variant declaration followed by its definition within a structure
494declaration:
495
496variant name {
497 field_type sel1;
498 field_type sel2;
499 field_type sel3;
500 ...
501};
502
503struct {
a9b83695 504 enum : integer_type { sel1, sel2, sel3, ... } tag_field;
fcba70d4
MD
505 ...
506 variant name <tag_field> v;
507}
508
509An unnamed variant definition within a structure is expressed by the following
6672e9e1 510TSDL meta-data:
fcba70d4
MD
511
512struct {
a9b83695 513 enum : integer_type { sel1, sel2, sel3, ... } tag_field;
fcba70d4
MD
514 ...
515 variant <tag_field> {
516 field_type sel1;
517 field_type sel2;
518 field_type sel3;
519 ...
520 } v;
521}
522
523Example of a named variant within a sequence that refers to a single tag field:
524
525variant example {
526 uint32_t a;
527 uint64_t b;
528 short c;
529};
530
531struct {
a9b83695 532 enum : uint2_t { a, b, c } choice;
1ab22b2a
MD
533 unsigned int seqlen;
534 variant example <choice> v[seqlen];
fcba70d4
MD
535}
536
537Example of an unnamed variant:
538
539struct {
a9b83695 540 enum : uint2_t { a, b, c, d } choice;
fcba70d4
MD
541 /* Unrelated fields can be added between the variant and its tag */
542 int32_t somevalue;
543 variant <choice> {
544 uint32_t a;
545 uint64_t b;
546 short c;
547 struct {
548 unsigned int field1;
549 uint64_t field2;
550 } d;
551 } s;
552}
553
554Example of an unnamed variant within an array:
555
556struct {
a9b83695 557 enum : uint2_t { a, b, c } choice;
fcba70d4
MD
558 variant <choice> {
559 uint32_t a;
560 uint64_t b;
561 short c;
15850440 562 } v[10];
fcba70d4
MD
563}
564
565Example of a variant type definition within a structure, where the defined type
566is then declared within an array of structures. This variant refers to a tag
37ab95c3 567located in an upper static scope. This example clearly shows that a variant
fcba70d4 568type definition referring to the tag "x" uses the closest preceding field from
37ab95c3 569the static scope of the type definition.
fcba70d4
MD
570
571struct {
a9b83695 572 enum : uint2_t { a, b, c, d } x;
fcba70d4
MD
573
574 typedef variant <x> { /*
575 * "x" refers to the preceding "x" enumeration in the
37ab95c3 576 * static scope of the type definition.
fcba70d4
MD
577 */
578 uint32_t a;
579 uint64_t b;
580 short c;
581 } example_variant;
582
583 struct {
a9b83695 584 enum : int { x, y, z } x; /* This enumeration is not used by "v". */
fcba70d4 585 example_variant v; /*
a9b83695 586 * "v" uses the "enum : uint2_t { a, b, c, d }"
fcba70d4
MD
587 * tag.
588 */
589 } a[10];
590}
591
5924.2.3 Arrays
5ba9f198 593
6672e9e1
MD
594Arrays are fixed-length. Their length is declared in the type
595declaration within the meta-data. They contain an array of "inner type"
596elements, which can refer to any type not containing the type of the
597array being declared (no circular dependency). The length is the number
598of elements in an array.
5ba9f198 599
6672e9e1 600TSDL meta-data representation of a named array:
80fd2569
MD
601
602typedef elem_type name[length];
5ba9f198 603
2152348f 604A nameless array can be declared as a field type within a structure, e.g.:
5ba9f198 605
2152348f 606 uint8_t field_name[10];
80fd2569 607
ec4404a7 608Arrays are always aligned on their element alignment requirement.
5ba9f198 609
fcba70d4 6104.2.4 Sequences
5ba9f198 611
f3afd1c7 612Sequences are dynamically-sized arrays. They refer to a "length"
37ab95c3 613unsigned integer field, which must appear in either the same static scope,
1ab22b2a 614prior to the sequence field (in field declaration order), in an upper
37ab95c3
MD
615static scope, or in an upper dynamic scope (see Section 7.3.2). This
616length field represents the number of elements in the sequence. The
617sequence per se is an array of "inner type" elements.
5ba9f198 618
1ab22b2a 619TSDL meta-data representation for a sequence type definition:
80fd2569 620
1ab22b2a
MD
621struct {
622 unsigned int length_field;
623 typedef elem_type typename[length_field];
624 typename seq_field_name;
625}
626
627A sequence can also be declared as a field type, e.g.:
80fd2569 628
1ab22b2a
MD
629struct {
630 unsigned int length_field;
631 long seq_field_name[length_field];
632}
80fd2569 633
1ab22b2a
MD
634Multiple sequences can refer to the same length field, and these length
635fields can be in a different upper dynamic scope:
636
637e.g., assuming the stream.event.header defines:
638
639stream {
640 ...
641 id = 1;
642 event.header := struct {
643 uint16_t seq_len;
644 };
645};
646
647event {
648 ...
649 stream_id = 1;
650 fields := struct {
651 long seq_a[stream.event.header.seq_len];
652 char seq_b[stream.event.header.seq_len];
653 };
654};
80fd2569 655
1ab22b2a 656The sequence elements follow the "array" specifications.
5ba9f198 657
fcba70d4 6584.2.5 Strings
5ba9f198
MD
659
660Strings are an array of bytes of variable size and are terminated by a '\0'
6672e9e1
MD
661"NULL" character. Their encoding is described in the TSDL meta-data. In
662absence of encoding attribute information, the default encoding is
663UTF-8.
5ba9f198 664
6672e9e1 665TSDL meta-data representation of a named string type:
80fd2569 666
359894ac 667typealias string {
5ba9f198 668 encoding = UTF8 OR ASCII;
38b8da21 669} := name;
5ba9f198 670
80fd2569
MD
671A nameless string type can be declared as a field type:
672
673string field_name; /* Use default UTF8 encoding */
5ba9f198 674
ec4404a7
MD
675Strings are always aligned on byte size.
676
3bf79539
MD
6775. Event Packet Header
678
6672e9e1
MD
679The event packet header consists of two parts: the "event packet header"
680is the same for all streams of a trace. The second part, the "event
681packet context", is described on a per-stream basis. Both are described
682in the TSDL meta-data. The packets are aligned on architecture-page-sized
683addresses.
3bf79539 684
6672e9e1 685Event packet header (all fields are optional, specified by TSDL meta-data):
3bf79539 686
6672e9e1
MD
687- Magic number (CTF magic number: 0xC1FC1FC1) specifies that this is a
688 CTF packet. This magic number is optional, but when present, it should
689 come at the very beginning of the packet.
690- Trace UUID, used to ensure the event packet match the meta-data used.
691 (note: we cannot use a meta-data checksum in every cases instead of a
692 UUID because meta-data can be appended to while tracing is active)
693 This field is optional.
694- Stream ID, used as reference to stream description in meta-data.
695 This field is optional if there is only one stream description in the
696 meta-data, but becomes required if there are more than one stream in
697 the TSDL meta-data description.
3bf79539 698
6672e9e1 699Event packet context (all fields are optional, specified by TSDL meta-data):
3bf79539 700
4dfca05b
MD
701- Event packet content size (in bits).
702- Event packet size (in bits, includes padding).
cda89682 703- Event packet content checksum. Checksum excludes the event packet
3bf79539
MD
704 header.
705- Per-stream event packet sequence count (to deal with UDP packet loss). The
706 number of significant sequence counter bits should also be present, so
b11853af 707 wrap-arounds are dealt with correctly.
6672e9e1 708- Time-stamp at the beginning and time-stamp at the end of the event packet.
3bf79539
MD
709 Both timestamps are written in the packet header, but sampled respectively
710 while (or before) writing the first event and while (or after) writing the
711 last event in the packet. The inclusive range between these timestamps should
712 include all event timestamps assigned to events contained within the packet.
5ba9f198 713- Events discarded count
3bf79539 714 - Snapshot of a per-stream free-running counter, counting the number of
8f34dd4f
MD
715 events discarded that were supposed to be written in the stream after
716 the last event in the event packet.
717 * Note: producer-consumer buffer full condition can fill the current
3bf79539 718 event packet with padding so we know exactly where events have been
8f34dd4f
MD
719 discarded. However, if the buffer full condition chooses not
720 to fill the current event packet with padding, all we know
721 about the timestamp range in which the events have been
722 discarded is that it is somewhere between the beginning and
723 the end of the packet.
3bf79539
MD
724- Lossless compression scheme used for the event packet content. Applied
725 directly to raw data. New types of compression can be added in following
726 versions of the format.
5ba9f198
MD
727 0: no compression scheme
728 1: bzip2
729 2: gzip
3bf79539
MD
730 3: xz
731- Cypher used for the event packet content. Applied after compression.
5ba9f198
MD
732 0: no encryption
733 1: AES
3bf79539 734- Checksum scheme used for the event packet content. Applied after encryption.
5ba9f198
MD
735 0: no checksum
736 1: md5
737 2: sha1
738 3: crc32
739
6672e9e1 7405.1 Event Packet Header Description
3bf79539 741
fc5425db 742The event packet header layout is indicated by the trace packet.header
6672e9e1
MD
743field. Here is a recommended structure type for the packet header with
744the fields typically expected (although these fields are each optional):
fc5425db 745
80fd2569
MD
746struct event_packet_header {
747 uint32_t magic;
3fde5da1 748 uint8_t uuid[16];
3bf79539 749 uint32_t stream_id;
80fd2569 750};
5ba9f198 751
fc5425db
MD
752trace {
753 ...
754 packet.header := struct event_packet_header;
755};
756
6672e9e1
MD
757If the magic number is not present, tools such as "file" will have no
758mean to discover the file type.
759
3fde5da1
MD
760If the uuid is not present, no validation that the meta-data actually
761corresponds to the stream is performed.
37b39e47
MD
762
763If the stream_id packet header field is missing, the trace can only
764contain a single stream. Its "id" field can be left out, and its events
765don't need to declare a "stream_id" field.
766
767
3bf79539
MD
7685.2 Event Packet Context Description
769
770Event packet context example. These are declared within the stream declaration
6672e9e1 771in the meta-data. All these fields are optional. If the packet size field is
6a7c61df 772missing, the whole stream only contains a single packet. If the content
f6b81726
MD
773size field is missing, the packet is filled (no padding). The content
774and packet sizes include all headers.
3bf79539
MD
775
776An example event packet context type:
777
80fd2569 778struct event_packet_context {
3bf79539
MD
779 uint64_t timestamp_begin;
780 uint64_t timestamp_end;
781 uint32_t checksum;
782 uint32_t stream_packet_count;
783 uint32_t events_discarded;
784 uint32_t cpu_id;
47911ef2
MD
785 uint64_t/uint32_t/uint16_t content_size;
786 uint64_t/uint32_t/uint16_t packet_size;
3bf79539
MD
787 uint8_t compression_scheme;
788 uint8_t encryption_scheme;
3b0f8e4d 789 uint8_t checksum_scheme;
3bf79539 790};
5ba9f198 791
fcba70d4 792
5ba9f198
MD
7936. Event Structure
794
795The overall structure of an event is:
796
6672e9e1
MD
7971 - Stream Packet Context (as specified by the stream meta-data)
798 2 - Event Header (as specified by the stream meta-data)
799 3 - Stream Event Context (as specified by the stream meta-data)
800 4 - Event Context (as specified by the event meta-data)
801 5 - Event Payload (as specified by the event meta-data)
5ba9f198 802
fdf2bb05 803This structure defines an implicit dynamic scoping, where variants
7d9d7e92
MD
804located in inner structures (those with a higher number in the listing
805above) can refer to the fields of outer structures (with lower number in
6c7226e9 806the listing above). See Section 7.3 TSDL Scopes for more detail.
5ba9f198 807
fdf2bb05 8086.1 Event Header
fcba70d4 809
6672e9e1 810Event headers can be described within the meta-data. We hereby propose, as an
fcba70d4
MD
811example, two types of events headers. Type 1 accommodates streams with less than
81231 event IDs. Type 2 accommodates streams with 31 or more event IDs.
5ba9f198 813
3bf79539
MD
814One major factor can vary between streams: the number of event IDs assigned to
815a stream. Luckily, this information tends to stay relatively constant (modulo
5ba9f198 816event registration while trace is being recorded), so we can specify different
3bf79539 817representations for streams containing few event IDs and streams containing
6672e9e1
MD
818many event IDs, so we end up representing the event ID and time-stamp as
819densely as possible in each case.
5ba9f198 820
fcba70d4
MD
821The header is extended in the rare occasions where the information cannot be
822represented in the ranges available in the standard event header. They are also
3bf79539
MD
823used in the rare occasions where the data required for a field could not be
824collected: the flag corresponding to the missing field within the missing_fields
825array is then set to 1.
5ba9f198 826
6672e9e1
MD
827Types uintX_t represent an X-bit unsigned integer, as declared with
828either:
5ba9f198 829
6672e9e1
MD
830 typealias integer { size = X; align = X; signed = false } := uintX_t;
831
832 or
833
834 typealias integer { size = X; align = 1; signed = false } := uintX_t;
5ba9f198 835
fdf2bb05 8366.1.1 Type 1 - Few event IDs
5ba9f198
MD
837
838 - Aligned on 32-bit (or 8-bit if byte-packed, depending on the architecture
839 preference).
5ba9f198 840 - Native architecture byte ordering.
fcba70d4
MD
841 - For "compact" selection
842 - Fixed size: 32 bits.
843 - For "extended" selection
844 - Size depends on the architecture and variant alignment.
5ba9f198 845
80fd2569 846struct event_header_1 {
fcba70d4
MD
847 /*
848 * id: range: 0 - 30.
849 * id 31 is reserved to indicate an extended header.
850 */
a9b83695 851 enum : uint5_t { compact = 0 ... 30, extended = 31 } id;
fcba70d4
MD
852 variant <id> {
853 struct {
854 uint27_t timestamp;
855 } compact;
856 struct {
857 uint32_t id; /* 32-bit event IDs */
858 uint64_t timestamp; /* 64-bit timestamps */
859 } extended;
860 } v;
cb108fea 861} align(32); /* or align(8) */
5ba9f198 862
5ba9f198 863
fdf2bb05 8646.1.2 Type 2 - Many event IDs
5ba9f198 865
fcba70d4 866 - Aligned on 16-bit (or 8-bit if byte-packed, depending on the architecture
5ba9f198 867 preference).
5ba9f198 868 - Native architecture byte ordering.
fcba70d4
MD
869 - For "compact" selection
870 - Size depends on the architecture and variant alignment.
871 - For "extended" selection
872 - Size depends on the architecture and variant alignment.
5ba9f198 873
80fd2569 874struct event_header_2 {
fcba70d4
MD
875 /*
876 * id: range: 0 - 65534.
877 * id 65535 is reserved to indicate an extended header.
878 */
a9b83695 879 enum : uint16_t { compact = 0 ... 65534, extended = 65535 } id;
fcba70d4
MD
880 variant <id> {
881 struct {
882 uint32_t timestamp;
883 } compact;
884 struct {
885 uint32_t id; /* 32-bit event IDs */
886 uint64_t timestamp; /* 64-bit timestamps */
887 } extended;
888 } v;
cb108fea 889} align(16); /* or align(8) */
5ba9f198 890
5ba9f198
MD
891
8926.2 Event Context
893
6672e9e1
MD
894The event context contains information relative to the current event.
895The choice and meaning of this information is specified by the TSDL
896stream and event meta-data descriptions. The stream context is applied
897to all events within the stream. The stream context structure follows
898the event header. The event context is applied to specific events. Its
899structure follows the stream context structure.
5ba9f198 900
fcba70d4
MD
901An example of stream-level event context is to save the event payload size with
902each event, or to save the current PID with each event. These are declared
6672e9e1 903within the stream declaration within the meta-data:
5ba9f198 904
fcba70d4
MD
905 stream {
906 ...
6672e9e1 907 event.context := struct {
80fd2569
MD
908 uint pid;
909 uint16_t payload_size;
6672e9e1 910 };
fcba70d4
MD
911 };
912
913An example of event-specific event context is to declare a bitmap of missing
914fields, only appended after the stream event context if the extended event
915header is selected. NR_FIELDS is the number of fields within the event (a
916numeric value).
5ba9f198 917
fcba70d4
MD
918 event {
919 context = struct {
920 variant <id> {
921 struct { } compact;
922 struct {
923 uint1_t missing_fields[NR_FIELDS]; /* missing event fields bitmap */
924 } extended;
925 } v;
926 };
927 ...
928 }
5ba9f198
MD
929
9306.3 Event Payload
931
932An event payload contains fields specific to a given event type. The fields
6672e9e1 933belonging to an event type are described in the event-specific meta-data
5ba9f198
MD
934within a structure type.
935
9366.3.1 Padding
937
938No padding at the end of the event payload. This differs from the ISO/C standard
939for structures, but follows the CTF standard for structures. In a trace, even
940though it makes sense to align the beginning of a structure, it really makes no
941sense to add padding at the end of the structure, because structures are usually
942not followed by a structure of the same type.
943
944This trick can be done by adding a zero-length "end" field at the end of the C
945structures, and by using the offset of this field rather than using sizeof()
3bf79539 946when calculating the size of a structure (see Appendix "A. Helper macros").
5ba9f198
MD
947
9486.3.2 Alignment
949
950The event payload is aligned on the largest alignment required by types
951contained within the payload. (This follows the ISO/C standard for structures)
952
953
6c7226e9
MD
9547. Trace Stream Description Language (TSDL)
955
956The Trace Stream Description Language (TSDL) allows expression of the
957binary trace streams layout in a C99-like Domain Specific Language
958(DSL).
959
960
6672e9e1 9617.1 Meta-data
6c7226e9
MD
962
963The trace stream layout description is located in the trace meta-data.
964The meta-data is itself located in a stream identified by its name:
965"metadata".
5ba9f198 966
6672e9e1
MD
967The meta-data description can be expressed in two different formats:
968text-only and packet-based. The text-only description facilitates
969generation of meta-data and provides a convenient way to enter the
970meta-data information by hand. The packet-based meta-data provides the
971CTF stream packet facilities (checksumming, compression, encryption,
972network-readiness) for meta-data stream generated and transported by a
973tracer.
974
1b4d35eb
MD
975The text-only meta-data file is a plain-text TSDL description. This file
976must begin with the following characters to identify the file as a CTF
9486a18c 977TSDL text-based metadata file (without the double-quotes) :
1b4d35eb 978
ec2b4db8 979"/* CTF"
1b4d35eb 980
ec2b4db8
MD
981It must be followed by a space, and the version of the specification
982followed by the CTF trace, e.g.:
983
984" 1.8"
985
986These characters allow automated discovery of file type and CTF
987specification version. They are interpreted as a the beginning of a
988comment by the TSDL metadata parser. The comment can be continued to
989contain extra commented characters before it is closed.
6672e9e1
MD
990
991The packet-based meta-data is made of "meta-data packets", which each
992start with a meta-data packet header. The packet-based meta-data
993description is detected by reading the magic number "0x75D11D57" at the
994beginning of the file. This magic number is also used to detect the
995endianness of the architecture by trying to read the CTF magic number
996and its counterpart in reversed endianness. The events within the
997meta-data stream have no event header nor event context. Each event only
46400bde
MD
998contains a special "sequence" payload, which is a sequence of bits which
999length is implicitly calculated by using the
1000"trace.packet.header.content_size" field, minus the packet header size.
1001The formatting of this sequence of bits is a plain-text representation
1002of the TSDL description. Each meta-data packet start with a special
1003packet header, specific to the meta-data stream, which contains,
1004exactly:
6672e9e1
MD
1005
1006struct metadata_packet_header {
2daeaa3a 1007 uint32_t magic; /* 0x75D11D57 */
3fde5da1 1008 uint8_t uuid[16]; /* Unique Universal Identifier */
6672e9e1
MD
1009 uint32_t checksum; /* 0 if unused */
1010 uint32_t content_size; /* in bits */
1011 uint32_t packet_size; /* in bits */
1012 uint8_t compression_scheme; /* 0 if unused */
1013 uint8_t encryption_scheme; /* 0 if unused */
1014 uint8_t checksum_scheme; /* 0 if unused */
ec2b4db8
MD
1015 uint8_t major; /* CTF spec version major number */
1016 uint8_t minor; /* CTF spec version minor number */
6672e9e1
MD
1017};
1018
1019The packet-based meta-data can be converted to a text-only meta-data by
3568031f 1020concatenating all the strings it contains.
4fafe1ad 1021
6672e9e1
MD
1022In the textual representation of the meta-data, the text contained
1023within "/*" and "*/", as well as within "//" and end of line, are
1024treated as comments. Boolean values can be represented as true, TRUE,
1025or 1 for true, and false, FALSE, or 0 for false. Within the string-based
1026meta-data description, the trace UUID is represented as a string of
1027hexadecimal digits and dashes "-". In the event packet header, the trace
1028UUID is represented as an array of bytes.
fcba70d4 1029
fdf2bb05 1030
6c7226e9 10317.2 Declaration vs Definition
fdf2bb05
MD
1032
1033A declaration associates a layout to a type, without specifying where
1034this type is located in the event structure hierarchy (see Section 6).
1035This therefore includes typedef, typealias, as well as all type
1036specifiers. In certain circumstances (typedef, structure field and
1037variant field), a declaration is followed by a declarator, which specify
1038the newly defined type name (for typedef), or the field name (for
1039declarations located within structure and variants). Array and sequence,
1040declared with square brackets ("[" "]"), are part of the declarator,
a9b83695 1041similarly to C99. The enumeration base type is specified by
6c7226e9 1042": enum_base", which is part of the type specifier. The variant tag
a9b83695 1043name, specified between "<" ">", is also part of the type specifier.
fdf2bb05
MD
1044
1045A definition associates a type to a location in the event structure
b9606a77
MD
1046hierarchy (see Section 6). This association is denoted by ":=", as shown
1047in Section 7.3.
fdf2bb05
MD
1048
1049
6c7226e9 10507.3 TSDL Scopes
fdf2bb05 1051
37ab95c3
MD
1052TSDL uses three different types of scoping: a lexical scope is used for
1053declarations and type definitions, and static and dynamic scopes are
1054used for variants references to tag fields (with relative and absolute
1055path lookups) and for sequence references to length fields.
fdf2bb05 1056
6c7226e9 10577.3.1 Lexical Scope
fdf2bb05 1058
570ecabe
MD
1059Each of "trace", "env", "stream", "event", "struct" and "variant" have
1060their own nestable declaration scope, within which types can be declared
1061using "typedef" and "typealias". A root declaration scope also contains
1062all declarations located outside of any of the aforementioned
1063declarations. An inner declaration scope can refer to type declared
1064within its container lexical scope prior to the inner declaration scope.
1065Redefinition of a typedef or typealias is not valid, although hiding an
1066upper scope typedef or typealias is allowed within a sub-scope.
fdf2bb05 1067
37ab95c3 10687.3.2 Static and Dynamic Scopes
fdf2bb05 1069
37ab95c3
MD
1070A local static scope consists in the scope generated by the declaration
1071of fields within a compound type. A static scope is a local static scope
1072augmented with the nested sub-static-scopes it contains.
1073
1074A dynamic scope consists in the static scope augmented with the
7d9d7e92 1075implicit event structure definition hierarchy presented at Section 6.
fdf2bb05 1076
37ab95c3
MD
1077Multiple declarations of the same field name within a local static scope
1078is not valid. It is however valid to re-use the same field name in
1079different local scopes.
1080
1081Nested static and dynamic scopes form lookup paths. These are used for
1082variant tag and sequence length references. They are used at the variant
1083and sequence definition site to look up the location of the tag field
1084associated with a variant, and to lookup up the location of the length
1085field associated with a sequence.
1086
1087Variants and sequences can refer to a tag field either using a relative
4cac83ee
MD
1088path or an absolute path. The relative path is relative to the scope in
1089which the variant or sequence performing the lookup is located.
37ab95c3
MD
1090Relative paths are only allowed to lookup within the same static scope,
1091which includes its nested static scopes. Lookups targeting parent static
1092scopes need to be performed with an absolute path.
1093
1094Absolute path lookups use the full path including the dynamic scope
1095followed by a "." and then the static scope. Therefore, variants (or
1096sequences) in lower levels in the dynamic scope (e.g. event context) can
1097refer to a tag (or length) field located in upper levels (e.g. in the
1098event header) by specifying, in this case, the associated tag with
1099<stream.event.header.field_name>. This allows, for instance, the event
1100context to define a variant referring to the "id" field of the event
1101header as selector.
1102
284724ae 1103The dynamic scope prefixes are thus:
fdf2bb05 1104
570ecabe 1105 - Trace Environment: <env. >,
e0d9e2c7 1106 - Trace Packet Header: <trace.packet.header. >,
7d9d7e92
MD
1107 - Stream Packet Context: <stream.packet.context. >,
1108 - Event Header: <stream.event.header. >,
1109 - Stream Event Context: <stream.event.context. >,
1110 - Event Context: <event.context. >,
1111 - Event Payload: <event.fields. >.
fdf2bb05 1112
37ab95c3
MD
1113
1114The target dynamic scope must be specified explicitly when referring to
4cac83ee
MD
1115a field outside of the static scope (absolute scope reference). No
1116conflict can occur between relative and dynamic paths, because the
1117keywords "trace", "stream", and "event" are reserved, and thus
1118not permitted as field names. It is recommended that field names
1119clashing with CTF and C99 reserved keywords use an underscore prefix to
1120eliminate the risk of generating a description containing an invalid
70375f92 1121field name. Consequently, fields starting with an underscore should have
92250c71 1122their leading underscore removed by the CTF trace readers.
70375f92 1123
fdf2bb05 1124
457d8b0a
MD
1125The information available in the dynamic scopes can be thought of as the
1126current tracing context. At trace production, information about the
1127current context is saved into the specified scope field levels. At trace
1128consumption, for each event, the current trace context is therefore
1129readable by accessing the upper dynamic scopes.
1130
fdf2bb05 1131
6c7226e9 11327.4 TSDL Examples
d285084f 1133
6672e9e1 1134The grammar representing the TSDL meta-data is presented in Appendix C.
7df6b93a 1135TSDL Grammar. This section presents a rather lighter reading that
6672e9e1 1136consists in examples of TSDL meta-data, with template values.
969f30c0 1137
350bb633
MD
1138The stream "id" can be left out if there is only one stream in the
1139trace. The event "id" field can be left out if there is only one event
1140in a stream.
1141
5ba9f198 1142trace {
1e9de2d1
MD
1143 major = value; /* CTF spec version major number */
1144 minor = value; /* CTF spec version minor number */
fdf2bb05 1145 uuid = "aaaaaaaa-aaaa-aaaa-aaaa-aaaaaaaaaaaa"; /* Trace UUID */
58997e9e 1146 byte_order = be OR le; /* Endianness (required) */
fc5425db
MD
1147 packet.header := struct {
1148 uint32_t magic;
3fde5da1 1149 uint8_t uuid[16];
fc5425db
MD
1150 uint32_t stream_id;
1151 };
3bf79539 1152};
5ba9f198 1153
570ecabe
MD
1154/*
1155 * The "env" (environment) scope contains assignment expressions. The
1156 * field names and content are implementation-defined.
1157 */
1158env {
1159 pid = value; /* example */
1160 proc_name = "name"; /* example */
1161 ...
1162};
1163
3bf79539
MD
1164stream {
1165 id = stream_id;
fdf2bb05 1166 /* Type 1 - Few event IDs; Type 2 - Many event IDs. See section 6.1. */
4fa992a5
MD
1167 event.header := event_header_1 OR event_header_2;
1168 event.context := struct {
77a98c82 1169 ...
3bf79539 1170 };
4fa992a5 1171 packet.context := struct {
77a98c82 1172 ...
3bf79539
MD
1173 };
1174};
5ba9f198
MD
1175
1176event {
980015f9 1177 name = "event_name";
3bf79539 1178 id = value; /* Numeric identifier within the stream */
67f02e24 1179 stream_id = stream_id;
dc56f167 1180 loglevel = value;
8e9060f2 1181 model.emf.uri = "string";
4fa992a5 1182 context := struct {
fcba70d4
MD
1183 ...
1184 };
4fa992a5 1185 fields := struct {
80fd2569
MD
1186 ...
1187 };
3bf79539 1188};
5ba9f198 1189
555f54e6
MD
1190callsite {
1191 name = "event_name";
1192 func = "func_name";
1193 file = "myfile.c";
1194 line = 39;
5a6b4ee1 1195 ip = 0x40096c;
555f54e6
MD
1196};
1197
5ba9f198
MD
1198/* More detail on types in section 4. Types */
1199
3d13ef1a
MD
1200/*
1201 * Named types:
1202 *
4fa992a5 1203 * Type declarations behave similarly to the C standard.
3d13ef1a
MD
1204 */
1205
80af8ac6 1206typedef aliased_type_specifiers new_type_declarators;
2152348f 1207
3d13ef1a 1208/* e.g.: typedef struct example new_type_name[10]; */
80fd2569 1209
4fa992a5
MD
1210/*
1211 * typealias
1212 *
1213 * The "typealias" declaration can be used to give a name (including
80af8ac6
MD
1214 * pointer declarator specifier) to a type. It should also be used to
1215 * map basic C types (float, int, unsigned long, ...) to a CTF type.
1216 * Typealias is a superset of "typedef": it also allows assignment of a
38b8da21 1217 * simple variable identifier to a type.
4fa992a5
MD
1218 */
1219
1220typealias type_class {
80fd2569 1221 ...
38b8da21 1222} := type_specifiers type_declarator;
2152348f 1223
3d13ef1a
MD
1224/*
1225 * e.g.:
4fa992a5 1226 * typealias integer {
3d13ef1a
MD
1227 * size = 32;
1228 * align = 32;
1229 * signed = false;
38b8da21 1230 * } := struct page *;
359894ac
MD
1231 *
1232 * typealias integer {
1233 * size = 32;
1234 * align = 32;
1235 * signed = true;
38b8da21 1236 * } := int;
3d13ef1a 1237 */
80fd2569
MD
1238
1239struct name {
3bf79539
MD
1240 ...
1241};
5ba9f198 1242
fcba70d4
MD
1243variant name {
1244 ...
1245};
1246
a9b83695 1247enum name : integer_type {
3bf79539
MD
1248 ...
1249};
1250
2152348f 1251
4fa992a5
MD
1252/*
1253 * Unnamed types, contained within compound type fields, typedef or typealias.
1254 */
2152348f 1255
80fd2569
MD
1256struct {
1257 ...
2152348f 1258}
5ba9f198 1259
ec4404a7
MD
1260struct {
1261 ...
1262} align(value)
1263
fcba70d4
MD
1264variant {
1265 ...
1266}
1267
a9b83695 1268enum : integer_type {
80fd2569 1269 ...
2152348f
MD
1270}
1271
1272typedef type new_type[length];
3bf79539 1273
2152348f
MD
1274struct {
1275 type field_name[length];
1276}
1277
1278typedef type new_type[length_type];
1279
1280struct {
1281 type field_name[length_type];
1282}
1283
1284integer {
80fd2569 1285 ...
2152348f 1286}
3bf79539 1287
2152348f 1288floating_point {
80fd2569 1289 ...
2152348f
MD
1290}
1291
1292struct {
1293 integer_type field_name:size; /* GNU/C bitfield */
1294}
1295
1296struct {
1297 string field_name;
1298}
3bf79539 1299
fcba70d4 1300
2fa70eba
MD
13018. Clocks
1302
1303Clock metadata allows to describe the clock topology of the system, as
1304well as to detail each clock parameter. In absence of clock description,
1305it is assumed that all fields named "timestamp" use the same clock
aed18b5e 1306source, which increments once per nanosecond.
2fa70eba
MD
1307
1308Describing a clock and how it is used by streams is threefold: first,
1309the clock and clock topology should be described in a "clock"
1310description block, e.g.:
1311
d803bfcb 1312clock {
58262d97 1313 name = cycle_counter_sync;
2fa70eba 1314 uuid = "62189bee-96dc-11e0-91a8-cfa3d89f3923";
58262d97 1315 description = "Cycle counter synchronized across CPUs";
2fa70eba
MD
1316 freq = 1000000000; /* frequency, in Hz */
1317 /* precision in seconds is: 1000 * (1/freq) */
1318 precision = 1000;
a40cccda
MD
1319 /*
1320 * clock value offset from Epoch is:
1321 * offset_s + (offset * (1/freq))
1322 */
1323 offset_s = 1326476837;
1324 offset = 897235420;
ce0fadbd 1325 absolute = FALSE;
2fa70eba
MD
1326};
1327
d803bfcb
MD
1328The mandatory "name" field specifies the name of the clock identifier,
1329which can later be used as a reference. The optional field "uuid" is the
1330unique identifier of the clock. It can be used to correlate different
1331traces that use the same clock. An optional textual description string
1332can be added with the "description" field. The "freq" field is the
1333initial frequency of the clock, in Hz. If the "freq" field is not
1334present, the frequency is assumed to be 1000000000 (providing clock
1335increment of 1 ns). The optional "precision" field details the
a40cccda
MD
1336uncertainty on the clock measurements, in (1/freq) units. The "offset_s"
1337and "offset" fields indicate the offset from POSIX.1 Epoch, 1970-01-01
133800:00:00 +0000 (UTC), to the zero of value of the clock. The "offset_s"
1339field is in seconds. The "offset" field is in (1/freq) units. If any of
1340the "offset_s" or "offset" field is not present, it is assigned the 0
ce0fadbd
MD
1341value. The field "absolute" is TRUE if the clock is a global reference
1342across different clock uuid (e.g. NTP time). Otherwise, "absolute" is
1343FALSE, and the clock can be considered as synchronized only with other
1344clocks that have the same uuid.
1345
2fa70eba
MD
1346
1347Secondly, a reference to this clock should be added within an integer
1348type:
1349
1350typealias integer {
1351 size = 64; align = 1; signed = false;
58262d97 1352 map = clock.cycle_counter_sync.value;
2fa70eba
MD
1353} := uint64_ccnt_t;
1354
1355Thirdly, stream declarations can reference the clock they use as a
1356time-stamp source:
1357
1358struct packet_context {
1359 uint64_ccnt_t ccnt_begin;
1360 uint64_ccnt_t ccnt_end;
1361 /* ... */
1362};
1363
1364stream {
1365 /* ... */
1366 event.header := struct {
1367 uint64_ccnt_t timestamp;
1368 /* ... */
1369 }
1370 packet.context := struct packet_context;
1371};
1372
1373For a N-bit integer type referring to a clock, if the integer overflows
1374compared to the N low order bits of the clock prior value, then it is
1375assumed that one, and only one, overflow occurred. It is therefore
1376important that events encoding time on a small number of bits happen
1377frequently enough to detect when more than one N-bit overflow occurs.
1378
1379In a packet context, clock field names ending with "_begin" and "_end"
1380have a special meaning: this refers to the time-stamps at, respectively,
1381the beginning and the end of each packet.
1382
1383
3bf79539 1384A. Helper macros
5ba9f198
MD
1385
1386The two following macros keep track of the size of a GNU/C structure without
1387padding at the end by placing HEADER_END as the last field. A one byte end field
1388is used for C90 compatibility (C99 flexible arrays could be used here). Note
1389that this does not affect the effective structure size, which should always be
1390calculated with the header_sizeof() helper.
1391
1392#define HEADER_END char end_field
1393#define header_sizeof(type) offsetof(typeof(type), end_field)
3bf79539
MD
1394
1395
1396B. Stream Header Rationale
1397
1398An event stream is divided in contiguous event packets of variable size. These
1399subdivisions allow the trace analyzer to perform a fast binary search by time
1400within the stream (typically requiring to index only the event packet headers)
1401without reading the whole stream. These subdivisions have a variable size to
1402eliminate the need to transfer the event packet padding when partially filled
1403event packets must be sent when streaming a trace for live viewing/analysis.
1404An event packet can contain a certain amount of padding at the end. Dividing
1405streams into event packets is also useful for network streaming over UDP and
1406flight recorder mode tracing (a whole event packet can be swapped out of the
1407buffer atomically for reading).
1408
1409The stream header is repeated at the beginning of each event packet to allow
1410flexibility in terms of:
1411
1412 - streaming support,
1413 - allowing arbitrary buffers to be discarded without making the trace
1414 unreadable,
1415 - allow UDP packet loss handling by either dealing with missing event packet
1416 or asking for re-transmission.
1417 - transparently support flight recorder mode,
1418 - transparently support crash dump.
1419
6c7226e9
MD
1420
1421C. TSDL Grammar
fcba70d4 1422
4fa992a5 1423/*
6c7226e9 1424 * Common Trace Format (CTF) Trace Stream Description Language (TSDL) Grammar.
4fa992a5
MD
1425 *
1426 * Inspired from the C99 grammar:
1427 * http://www.open-std.org/jtc1/sc22/wg14/www/docs/n1124.pdf (Annex A)
6c7226e9
MD
1428 * and c++1x grammar (draft)
1429 * http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2011/n3291.pdf (Annex A)
4fa992a5
MD
1430 *
1431 * Specialized for CTF needs by including only constant and declarations from
1432 * C99 (excluding function declarations), and by adding support for variants,
6c7226e9
MD
1433 * sequences and CTF-specific specifiers. Enumeration container types
1434 * semantic is inspired from c++1x enum-base.
4fa992a5
MD
1435 */
1436
14371) Lexical grammar
1438
14391.1) Lexical elements
1440
1441token:
1442 keyword
1443 identifier
1444 constant
1445 string-literal
1446 punctuator
1447
14481.2) Keywords
1449
1450keyword: is one of
1451
ec4404a7 1452align
dbb8b280 1453callsite
4fa992a5
MD
1454const
1455char
2fa70eba 1456clock
4fa992a5
MD
1457double
1458enum
570ecabe 1459env
4fa992a5
MD
1460event
1461floating_point
1462float
1463integer
1464int
1465long
1466short
1467signed
1468stream
1469string
1470struct
1471trace
3e1e1a78 1472typealias
4fa992a5
MD
1473typedef
1474unsigned
1475variant
1476void
1477_Bool
1478_Complex
1479_Imaginary
1480
1481
14821.3) Identifiers
1483
1484identifier:
1485 identifier-nondigit
1486 identifier identifier-nondigit
1487 identifier digit
1488
1489identifier-nondigit:
1490 nondigit
1491 universal-character-name
1492 any other implementation-defined characters
1493
1494nondigit:
1495 _
1496 [a-zA-Z] /* regular expression */
1497
1498digit:
1499 [0-9] /* regular expression */
1500
15011.4) Universal character names
1502
1503universal-character-name:
1504 \u hex-quad
1505 \U hex-quad hex-quad
1506
1507hex-quad:
1508 hexadecimal-digit hexadecimal-digit hexadecimal-digit hexadecimal-digit
1509
15101.5) Constants
1511
1512constant:
1513 integer-constant
1514 enumeration-constant
1515 character-constant
1516
1517integer-constant:
1518 decimal-constant integer-suffix-opt
1519 octal-constant integer-suffix-opt
1520 hexadecimal-constant integer-suffix-opt
1521
1522decimal-constant:
1523 nonzero-digit
1524 decimal-constant digit
1525
1526octal-constant:
1527 0
1528 octal-constant octal-digit
1529
1530hexadecimal-constant:
1531 hexadecimal-prefix hexadecimal-digit
1532 hexadecimal-constant hexadecimal-digit
1533
1534hexadecimal-prefix:
1535 0x
1536 0X
1537
1538nonzero-digit:
1539 [1-9]
1540
1541integer-suffix:
1542 unsigned-suffix long-suffix-opt
1543 unsigned-suffix long-long-suffix
1544 long-suffix unsigned-suffix-opt
1545 long-long-suffix unsigned-suffix-opt
1546
1547unsigned-suffix:
1548 u
1549 U
1550
1551long-suffix:
1552 l
1553 L
1554
1555long-long-suffix:
1556 ll
1557 LL
1558
4fa992a5
MD
1559enumeration-constant:
1560 identifier
1561 string-literal
1562
1563character-constant:
1564 ' c-char-sequence '
1565 L' c-char-sequence '
1566
1567c-char-sequence:
1568 c-char
1569 c-char-sequence c-char
1570
1571c-char:
1572 any member of source charset except single-quote ('), backslash
1573 (\), or new-line character.
1574 escape-sequence
1575
1576escape-sequence:
1577 simple-escape-sequence
1578 octal-escape-sequence
1579 hexadecimal-escape-sequence
1580 universal-character-name
1581
1582simple-escape-sequence: one of
1583 \' \" \? \\ \a \b \f \n \r \t \v
1584
1585octal-escape-sequence:
1586 \ octal-digit
1587 \ octal-digit octal-digit
1588 \ octal-digit octal-digit octal-digit
1589
1590hexadecimal-escape-sequence:
1591 \x hexadecimal-digit
1592 hexadecimal-escape-sequence hexadecimal-digit
1593
15941.6) String literals
1595
1596string-literal:
1597 " s-char-sequence-opt "
1598 L" s-char-sequence-opt "
1599
1600s-char-sequence:
1601 s-char
1602 s-char-sequence s-char
1603
1604s-char:
1605 any member of source charset except double-quote ("), backslash
1606 (\), or new-line character.
1607 escape-sequence
1608
16091.7) Punctuators
1610
1611punctuator: one of
1612 [ ] ( ) { } . -> * + - < > : ; ... = ,
1613
1614
16152) Phrase structure grammar
1616
1617primary-expression:
1618 identifier
1619 constant
1620 string-literal
1621 ( unary-expression )
1622
1623postfix-expression:
1624 primary-expression
1625 postfix-expression [ unary-expression ]
1626 postfix-expression . identifier
1627 postfix-expressoin -> identifier
1628
1629unary-expression:
1630 postfix-expression
1631 unary-operator postfix-expression
1632
1633unary-operator: one of
1634 + -
1635
4fa992a5
MD
1636assignment-operator:
1637 =
1638
b9606a77
MD
1639type-assignment-operator:
1640 :=
1641
4fa992a5 1642constant-expression-range:
73d61ac3 1643 unary-expression ... unary-expression
4fa992a5
MD
1644
16452.2) Declarations:
1646
1647declaration:
689e04b4 1648 declaration-specifiers declarator-list-opt ;
4fa992a5
MD
1649 ctf-specifier ;
1650
1651declaration-specifiers:
689e04b4 1652 storage-class-specifier declaration-specifiers-opt
4fa992a5
MD
1653 type-specifier declaration-specifiers-opt
1654 type-qualifier declaration-specifiers-opt
1655
1656declarator-list:
1657 declarator
1658 declarator-list , declarator
1659
d285084f
MD
1660abstract-declarator-list:
1661 abstract-declarator
1662 abstract-declarator-list , abstract-declarator
1663
4fa992a5
MD
1664storage-class-specifier:
1665 typedef
1666
1667type-specifier:
1668 void
1669 char
1670 short
1671 int
1672 long
1673 float
1674 double
1675 signed
1676 unsigned
1677 _Bool
1678 _Complex
cfdd51ec 1679 _Imaginary
9dfcfc0f
MD
1680 struct-specifier
1681 variant-specifier
4fa992a5
MD
1682 enum-specifier
1683 typedef-name
1684 ctf-type-specifier
1685
ec4404a7 1686align-attribute:
73d61ac3 1687 align ( unary-expression )
ec4404a7 1688
4fa992a5 1689struct-specifier:
ec4404a7
MD
1690 struct identifier-opt { struct-or-variant-declaration-list-opt } align-attribute-opt
1691 struct identifier align-attribute-opt
4fa992a5
MD
1692
1693struct-or-variant-declaration-list:
1694 struct-or-variant-declaration
1695 struct-or-variant-declaration-list struct-or-variant-declaration
1696
1697struct-or-variant-declaration:
1698 specifier-qualifier-list struct-or-variant-declarator-list ;
eacb16d1 1699 declaration-specifiers-opt storage-class-specifier declaration-specifiers-opt declarator-list ;
6143bab7
MD
1700 typealias declaration-specifiers abstract-declarator-list type-assignment-operator declaration-specifiers abstract-declarator-list ;
1701 typealias declaration-specifiers abstract-declarator-list type-assignment-operator declarator-list ;
4fa992a5
MD
1702
1703specifier-qualifier-list:
1704 type-specifier specifier-qualifier-list-opt
1705 type-qualifier specifier-qualifier-list-opt
1706
1707struct-or-variant-declarator-list:
1708 struct-or-variant-declarator
1709 struct-or-variant-declarator-list , struct-or-variant-declarator
1710
1711struct-or-variant-declarator:
1712 declarator
73d61ac3 1713 declarator-opt : unary-expression
4fa992a5
MD
1714
1715variant-specifier:
1716 variant identifier-opt variant-tag-opt { struct-or-variant-declaration-list }
1717 variant identifier variant-tag
1718
1719variant-tag:
37ab95c3 1720 < unary-expression >
4fa992a5
MD
1721
1722enum-specifier:
1723 enum identifier-opt { enumerator-list }
1724 enum identifier-opt { enumerator-list , }
1725 enum identifier
a9b83695
MD
1726 enum identifier-opt : declaration-specifiers { enumerator-list }
1727 enum identifier-opt : declaration-specifiers { enumerator-list , }
4fa992a5
MD
1728
1729enumerator-list:
1730 enumerator
1731 enumerator-list , enumerator
1732
1733enumerator:
1734 enumeration-constant
8d2d41f7
MD
1735 enumeration-constant assignment-operator unary-expression
1736 enumeration-constant assignment-operator constant-expression-range
4fa992a5
MD
1737
1738type-qualifier:
1739 const
1740
1741declarator:
1742 pointer-opt direct-declarator
1743
1744direct-declarator:
1745 identifier
1746 ( declarator )
1ab22b2a 1747 direct-declarator [ unary-expression ]
4fa992a5 1748
d285084f
MD
1749abstract-declarator:
1750 pointer-opt direct-abstract-declarator
1751
1752direct-abstract-declarator:
1753 identifier-opt
1754 ( abstract-declarator )
1ab22b2a 1755 direct-abstract-declarator [ unary-expression ]
d285084f
MD
1756 direct-abstract-declarator [ ]
1757
4fa992a5 1758pointer:
3b0f8e4d
MD
1759 * type-qualifier-list-opt
1760 * type-qualifier-list-opt pointer
4fa992a5
MD
1761
1762type-qualifier-list:
1763 type-qualifier
1764 type-qualifier-list type-qualifier
1765
4fa992a5
MD
1766typedef-name:
1767 identifier
1768
17692.3) CTF-specific declarations
1770
1771ctf-specifier:
662baf84 1772 clock { ctf-assignment-expression-list-opt }
4fa992a5
MD
1773 event { ctf-assignment-expression-list-opt }
1774 stream { ctf-assignment-expression-list-opt }
570ecabe 1775 env { ctf-assignment-expression-list-opt }
4fa992a5 1776 trace { ctf-assignment-expression-list-opt }
555f54e6 1777 callsite { ctf-assignment-expression-list-opt }
b12919a5
MD
1778 typealias declaration-specifiers abstract-declarator-list type-assignment-operator declaration-specifiers abstract-declarator-list
1779 typealias declaration-specifiers abstract-declarator-list type-assignment-operator declarator-list
4fa992a5
MD
1780
1781ctf-type-specifier:
1782 floating_point { ctf-assignment-expression-list-opt }
1783 integer { ctf-assignment-expression-list-opt }
1784 string { ctf-assignment-expression-list-opt }
7609d3c7 1785 string
4fa992a5
MD
1786
1787ctf-assignment-expression-list:
13441bf5
MD
1788 ctf-assignment-expression ;
1789 ctf-assignment-expression-list ctf-assignment-expression ;
4fa992a5
MD
1790
1791ctf-assignment-expression:
1792 unary-expression assignment-operator unary-expression
1793 unary-expression type-assignment-operator type-specifier
eacb16d1 1794 declaration-specifiers-opt storage-class-specifier declaration-specifiers-opt declarator-list
6143bab7
MD
1795 typealias declaration-specifiers abstract-declarator-list type-assignment-operator declaration-specifiers abstract-declarator-list
1796 typealias declaration-specifiers abstract-declarator-list type-assignment-operator declarator-list
This page took 0.217282 seconds and 4 git commands to generate.