Clarify use of underscores for fields
[ctf.git] / common-trace-format-specification.txt
CommitLineData
4dfca05b 1Common Trace Format (CTF) Specification (pre-v1.8)
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
MD
120An event stream can be divided into contiguous event packets of variable
121size. These subdivisions have a variable size. An event packet can
122contain a certain amount of padding at the end. The stream header is
123repeated at the beginning of each event packet. The rationale for the
124event stream design choices is explained in Appendix B. Stream Header
125Rationale.
5ba9f198 126
3bf79539
MD
127The event stream header will therefore be referred to as the "event packet
128header" throughout the rest of this document.
5ba9f198
MD
129
130
1314. Types
132
1fad7a85
MD
133Types are organized as type classes. Each type class belong to either of two
134kind of types: basic types or compound types.
135
5ba9f198
MD
1364.1 Basic types
137
1fad7a85
MD
138A basic type is a scalar type, as described in this section. It includes
139integers, GNU/C bitfields, enumerations, and floating point values.
5ba9f198
MD
140
1414.1.1 Type inheritance
142
80fd2569
MD
143Type specifications can be inherited to allow deriving types from a
144type class. For example, see the uint32_t named type derived from the "integer"
145type class below ("Integers" section). Types have a precise binary
146representation in the trace. A type class has methods to read and write these
147types, but must be derived into a type to be usable in an event field.
5ba9f198
MD
148
1494.1.2 Alignment
150
151We define "byte-packed" types as aligned on the byte size, namely 8-bit.
152We define "bit-packed" types as following on the next bit, as defined by the
370eae99 153"Integers" section.
5ba9f198 154
6672e9e1 155Each basic type must specify its alignment, in bits. Examples of
ec4404a7
MD
156possible alignments are: bit-packed (align = 1), byte-packed (align =
1578), or word-aligned (e.g. align = 32 or align = 64). The choice depends
158on the architecture preference and compactness vs performance trade-offs
159of the implementation. Architectures providing fast unaligned write
160byte-packed basic types to save space, aligning each type on byte
161boundaries (8-bit). Architectures with slow unaligned writes align types
162on specific alignment values. If no specific alignment is declared for a
163type, it is assumed to be bit-packed for integers with size not multiple
164of 8 bits and for gcc bitfields. All other basic types are byte-packed
165by default. It is however recommended to always specify the alignment
166explicitly. Alignment values must be power of two. Compound types are
167aligned as specified in their individual specification.
5ba9f198 168
6672e9e1 169TSDL meta-data attribute representation of a specific alignment:
5ba9f198
MD
170
171 align = value; /* value in bits */
172
1734.1.3 Byte order
174
3bf79539
MD
175By default, the native endianness of the source architecture the trace is used.
176Byte order can be overridden for a basic type by specifying a "byte_order"
177attribute. Typical use-case is to specify the network byte order (big endian:
178"be") to save data captured from the network into the trace without conversion.
179If not specified, the byte order is native.
5ba9f198 180
6672e9e1 181TSDL meta-data representation:
5ba9f198
MD
182
183 byte_order = native OR network OR be OR le; /* network and be are aliases */
184
1854.1.4 Size
186
187Type size, in bits, for integers and floats is that returned by "sizeof()" in C
188multiplied by CHAR_BIT.
189We require the size of "char" and "unsigned char" types (CHAR_BIT) to be fixed
190to 8 bits for cross-endianness compatibility.
191
6672e9e1 192TSDL meta-data representation:
5ba9f198
MD
193
194 size = value; (value is in bits)
195
1964.1.5 Integers
197
6672e9e1
MD
198Signed integers are represented in two-complement. Integer alignment,
199size, signedness and byte ordering are defined in the TSDL meta-data.
200Integers aligned on byte size (8-bit) and with length multiple of byte
201size (8-bit) correspond to the C99 standard integers. In addition,
202integers with alignment and/or size that are _not_ a multiple of the
203byte size are permitted; these correspond to the C99 standard bitfields,
204with the added specification that the CTF integer bitfields have a fixed
205binary representation. A MIT-licensed reference implementation of the
206CTF portable bitfields is available at:
5ba9f198
MD
207
208 http://git.efficios.com/?p=babeltrace.git;a=blob;f=include/babeltrace/bitfield.h
209
210Binary representation of integers:
211
212- On little and big endian:
213 - Within a byte, high bits correspond to an integer high bits, and low bits
214 correspond to low bits.
215- On little endian:
216 - Integer across multiple bytes are placed from the less significant to the
217 most significant.
218 - Consecutive integers are placed from lower bits to higher bits (even within
219 a byte).
220- On big endian:
221 - Integer across multiple bytes are placed from the most significant to the
222 less significant.
223 - Consecutive integers are placed from higher bits to lower bits (even within
224 a byte).
225
226This binary representation is derived from the bitfield implementation in GCC
227for little and big endian. However, contrary to what GCC does, integers can
6672e9e1 228cross units boundaries (no padding is required). Padding can be explicitly
5ba9f198
MD
229added (see 4.1.6 GNU/C bitfields) to follow the GCC layout if needed.
230
6672e9e1 231TSDL meta-data representation:
5ba9f198 232
80fd2569 233 integer {
5ba9f198
MD
234 signed = true OR false; /* default false */
235 byte_order = native OR network OR be OR le; /* default native */
236 size = value; /* value in bits, no default */
237 align = value; /* value in bits */
56b77c06
MD
238 /* based used for pretty-printing output, default: decimal. */
239 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
240 OR octal OR oct OR o OR 8 OR binary OR b OR 2;
13441bf5
MD
241 /* character encoding, default: none */
242 encoding = none or UTF8 or ASCII;
2152348f 243 }
5ba9f198 244
80fd2569 245Example of type inheritance (creation of a uint32_t named type):
5ba9f198 246
359894ac 247typealias integer {
9e4e34e9 248 size = 32;
5ba9f198
MD
249 signed = false;
250 align = 32;
38b8da21 251} := uint32_t;
5ba9f198 252
80fd2569 253Definition of a named 5-bit signed bitfield:
5ba9f198 254
359894ac 255typealias integer {
5ba9f198
MD
256 size = 5;
257 signed = true;
258 align = 1;
38b8da21 259} := int5_t;
5ba9f198 260
13441bf5
MD
261The character encoding field can be used to specify that the integer
262must be printed as a text character when read. e.g.:
263
264typealias integer {
265 size = 8;
266 align = 8;
267 signed = false;
268 encoding = UTF8;
269} := utf_char;
270
271
5ba9f198
MD
2724.1.6 GNU/C bitfields
273
274The GNU/C bitfields follow closely the integer representation, with a
275particularity on alignment: if a bitfield cannot fit in the current unit, the
80fd2569
MD
276unit is padded and the bitfield starts at the following unit. The unit size is
277defined by the size of the type "unit_type".
5ba9f198 278
6672e9e1 279TSDL meta-data representation:
80fd2569 280
d674f4b8 281 unit_type name:size;
80fd2569 282
5ba9f198
MD
283As an example, the following structure declared in C compiled by GCC:
284
285struct example {
286 short a:12;
287 short b:5;
288};
289
2152348f
MD
290The example structure is aligned on the largest element (short). The second
291bitfield would be aligned on the next unit boundary, because it would not fit in
292the current unit.
5ba9f198
MD
293
2944.1.7 Floating point
295
6672e9e1 296The floating point values byte ordering is defined in the TSDL meta-data.
5ba9f198
MD
297
298Floating point values follow the IEEE 754-2008 standard interchange formats.
299Description of the floating point values include the exponent and mantissa size
300in bits. Some requirements are imposed on the floating point values:
301
302- FLT_RADIX must be 2.
303- mant_dig is the number of digits represented in the mantissa. It is specified
304 by the ISO C99 standard, section 5.2.4, as FLT_MANT_DIG, DBL_MANT_DIG and
305 LDBL_MANT_DIG as defined by <float.h>.
306- exp_dig is the number of digits represented in the exponent. Given that
307 mant_dig is one bit more than its actual size in bits (leading 1 is not
308 needed) and also given that the sign bit always takes one bit, exp_dig can be
309 specified as:
310
311 - sizeof(float) * CHAR_BIT - FLT_MANT_DIG
312 - sizeof(double) * CHAR_BIT - DBL_MANT_DIG
313 - sizeof(long double) * CHAR_BIT - LDBL_MANT_DIG
314
6672e9e1 315TSDL meta-data representation:
5ba9f198 316
80fd2569 317floating_point {
ec4404a7
MD
318 exp_dig = value;
319 mant_dig = value;
320 byte_order = native OR network OR be OR le;
321 align = value;
2152348f 322}
5ba9f198
MD
323
324Example of type inheritance:
325
359894ac 326typealias floating_point {
5ba9f198
MD
327 exp_dig = 8; /* sizeof(float) * CHAR_BIT - FLT_MANT_DIG */
328 mant_dig = 24; /* FLT_MANT_DIG */
329 byte_order = native;
ec4404a7 330 align = 32;
38b8da21 331} := float;
5ba9f198
MD
332
333TODO: define NaN, +inf, -inf behavior.
334
ec4404a7
MD
335Bit-packed, byte-packed or larger alignments can be used for floating
336point values, similarly to integers.
337
5ba9f198
MD
3384.1.8 Enumerations
339
340Enumerations are a mapping between an integer type and a table of strings. The
341numerical representation of the enumeration follows the integer type specified
6672e9e1
MD
342by the meta-data. The enumeration mapping table is detailed in the enumeration
343description within the meta-data. The mapping table maps inclusive value
344ranges (or single values) to strings. Instead of being limited to simple
3bf79539 345"value -> string" mappings, these enumerations map
80fd2569 346"[ start_value ... end_value ] -> string", which map inclusive ranges of
3bf79539
MD
347values to strings. An enumeration from the C language can be represented in
348this format by having the same start_value and end_value for each element, which
349is in fact a range of size 1. This single-value range is supported without
4767a9e7 350repeating the start and end values with the value = string declaration.
80fd2569 351
a9b83695 352enum name : integer_type {
359894ac 353 somestring = start_value1 ... end_value1,
80fd2569
MD
354 "other string" = start_value2 ... end_value2,
355 yet_another_string, /* will be assigned to end_value2 + 1 */
356 "some other string" = value,
357 ...
358};
359
360If the values are omitted, the enumeration starts at 0 and increment of 1 for
361each entry:
362
a9b83695 363enum name : unsigned int {
80fd2569
MD
364 ZERO,
365 ONE,
366 TWO,
367 TEN = 10,
368 ELEVEN,
3bf79539 369};
5ba9f198 370
80fd2569 371Overlapping ranges within a single enumeration are implementation defined.
5ba9f198 372
2152348f
MD
373A nameless enumeration can be declared as a field type or as part of a typedef:
374
a9b83695 375enum : integer_type {
2152348f
MD
376 ...
377}
378
c2742c56
MD
379Enumerations omitting the container type ": integer_type" use the "int"
380type (for compatibility with C99). The "int" type must be previously
381declared. E.g.:
382
383typealias integer { size = 32; align = 32; signed = true } := int;
384
385enum {
386 ...
387}
388
1fad7a85 389
5ba9f198
MD
3904.2 Compound types
391
1fad7a85
MD
392Compound are aggregation of type declarations. Compound types include
393structures, variant, arrays, sequences, and strings.
394
5ba9f198
MD
3954.2.1 Structures
396
397Structures are aligned on the largest alignment required by basic types
398contained within the structure. (This follows the ISO/C standard for structures)
399
6672e9e1 400TSDL meta-data representation of a named structure:
5ba9f198 401
80fd2569
MD
402struct name {
403 field_type field_name;
404 field_type field_name;
405 ...
406};
5ba9f198
MD
407
408Example:
409
80fd2569
MD
410struct example {
411 integer { /* Nameless type */
412 size = 16;
413 signed = true;
414 align = 16;
415 } first_field_name;
6672e9e1 416 uint64_t second_field_name; /* Named type declared in the meta-data */
3bf79539 417};
5ba9f198 418
4cac83ee
MD
419The fields are placed in a sequence next to each other. They each
420possess a field name, which is a unique identifier within the structure.
421The identifier is not allowed to use any reserved keyword
422(see Section C.1.2). Replacing reserved keywords with
70375f92
MD
423underscore-prefixed field names is recommended. Fields starting with an
424underscore should have their leading underscore removed by the CTF parser.
5ba9f198 425
2152348f 426A nameless structure can be declared as a field type or as part of a typedef:
80fd2569
MD
427
428struct {
429 ...
2152348f 430}
80fd2569 431
ec4404a7
MD
432Alignment for a structure compound type can be forced to a minimum value
433by adding an "align" specifier after the declaration of a structure
434body. This attribute is read as: align(value). The value is specified in
435bits. The structure will be aligned on the maximum value between this
436attribute and the alignment required by the basic types contained within
437the structure. e.g.
438
439struct {
440 ...
441} align(32)
442
77a98c82 4434.2.2 Variants (Discriminated/Tagged Unions)
fcba70d4 444
fdf2bb05
MD
445A CTF variant is a selection between different types. A CTF variant must
446always be defined within the scope of a structure or within fields
447contained within a structure (defined recursively). A "tag" enumeration
37ab95c3
MD
448field must appear in either the same static scope, prior to the variant
449field (in field declaration order), in an upper static scope , or in an
450upper dynamic scope (see Section 7.3.2). The type selection is indicated
451by the mapping from the enumeration value to the string used as variant
452type selector. The field to use as tag is specified by the "tag_field",
453specified between "< >" after the "variant" keyword for unnamed
454variants, and after "variant name" for named variants.
fcba70d4
MD
455
456The alignment of the variant is the alignment of the type as selected by the tag
457value for the specific instance of the variant. The alignment of the type
458containing the variant is independent of the variant alignment. The size of the
459variant is the size as selected by the tag value for the specific instance of
460the variant.
461
4cac83ee
MD
462Each variant type selector possess a field name, which is a unique
463identifier within the variant. The identifier is not allowed to use any
464reserved keyword (see Section C.1.2). Replacing reserved keywords with
70375f92
MD
465underscore-prefixed field names is recommended. Fields starting with an
466underscore should have their leading underscore removed by the CTF parser.
467
4cac83ee 468
fcba70d4
MD
469A named variant declaration followed by its definition within a structure
470declaration:
471
472variant name {
473 field_type sel1;
474 field_type sel2;
475 field_type sel3;
476 ...
477};
478
479struct {
a9b83695 480 enum : integer_type { sel1, sel2, sel3, ... } tag_field;
fcba70d4
MD
481 ...
482 variant name <tag_field> v;
483}
484
485An unnamed variant definition within a structure is expressed by the following
6672e9e1 486TSDL meta-data:
fcba70d4
MD
487
488struct {
a9b83695 489 enum : integer_type { sel1, sel2, sel3, ... } tag_field;
fcba70d4
MD
490 ...
491 variant <tag_field> {
492 field_type sel1;
493 field_type sel2;
494 field_type sel3;
495 ...
496 } v;
497}
498
499Example of a named variant within a sequence that refers to a single tag field:
500
501variant example {
502 uint32_t a;
503 uint64_t b;
504 short c;
505};
506
507struct {
a9b83695 508 enum : uint2_t { a, b, c } choice;
1ab22b2a
MD
509 unsigned int seqlen;
510 variant example <choice> v[seqlen];
fcba70d4
MD
511}
512
513Example of an unnamed variant:
514
515struct {
a9b83695 516 enum : uint2_t { a, b, c, d } choice;
fcba70d4
MD
517 /* Unrelated fields can be added between the variant and its tag */
518 int32_t somevalue;
519 variant <choice> {
520 uint32_t a;
521 uint64_t b;
522 short c;
523 struct {
524 unsigned int field1;
525 uint64_t field2;
526 } d;
527 } s;
528}
529
530Example of an unnamed variant within an array:
531
532struct {
a9b83695 533 enum : uint2_t { a, b, c } choice;
fcba70d4
MD
534 variant <choice> {
535 uint32_t a;
536 uint64_t b;
537 short c;
15850440 538 } v[10];
fcba70d4
MD
539}
540
541Example of a variant type definition within a structure, where the defined type
542is then declared within an array of structures. This variant refers to a tag
37ab95c3 543located in an upper static scope. This example clearly shows that a variant
fcba70d4 544type definition referring to the tag "x" uses the closest preceding field from
37ab95c3 545the static scope of the type definition.
fcba70d4
MD
546
547struct {
a9b83695 548 enum : uint2_t { a, b, c, d } x;
fcba70d4
MD
549
550 typedef variant <x> { /*
551 * "x" refers to the preceding "x" enumeration in the
37ab95c3 552 * static scope of the type definition.
fcba70d4
MD
553 */
554 uint32_t a;
555 uint64_t b;
556 short c;
557 } example_variant;
558
559 struct {
a9b83695 560 enum : int { x, y, z } x; /* This enumeration is not used by "v". */
fcba70d4 561 example_variant v; /*
a9b83695 562 * "v" uses the "enum : uint2_t { a, b, c, d }"
fcba70d4
MD
563 * tag.
564 */
565 } a[10];
566}
567
5684.2.3 Arrays
5ba9f198 569
6672e9e1
MD
570Arrays are fixed-length. Their length is declared in the type
571declaration within the meta-data. They contain an array of "inner type"
572elements, which can refer to any type not containing the type of the
573array being declared (no circular dependency). The length is the number
574of elements in an array.
5ba9f198 575
6672e9e1 576TSDL meta-data representation of a named array:
80fd2569
MD
577
578typedef elem_type name[length];
5ba9f198 579
2152348f 580A nameless array can be declared as a field type within a structure, e.g.:
5ba9f198 581
2152348f 582 uint8_t field_name[10];
80fd2569 583
ec4404a7 584Arrays are always aligned on their element alignment requirement.
5ba9f198 585
fcba70d4 5864.2.4 Sequences
5ba9f198 587
1ab22b2a 588Sequences are dynamically-sized arrays. They refer to a a "length"
37ab95c3 589unsigned integer field, which must appear in either the same static scope,
1ab22b2a 590prior to the sequence field (in field declaration order), in an upper
37ab95c3
MD
591static scope, or in an upper dynamic scope (see Section 7.3.2). This
592length field represents the number of elements in the sequence. The
593sequence per se is an array of "inner type" elements.
5ba9f198 594
1ab22b2a 595TSDL meta-data representation for a sequence type definition:
80fd2569 596
1ab22b2a
MD
597struct {
598 unsigned int length_field;
599 typedef elem_type typename[length_field];
600 typename seq_field_name;
601}
602
603A sequence can also be declared as a field type, e.g.:
80fd2569 604
1ab22b2a
MD
605struct {
606 unsigned int length_field;
607 long seq_field_name[length_field];
608}
80fd2569 609
1ab22b2a
MD
610Multiple sequences can refer to the same length field, and these length
611fields can be in a different upper dynamic scope:
612
613e.g., assuming the stream.event.header defines:
614
615stream {
616 ...
617 id = 1;
618 event.header := struct {
619 uint16_t seq_len;
620 };
621};
622
623event {
624 ...
625 stream_id = 1;
626 fields := struct {
627 long seq_a[stream.event.header.seq_len];
628 char seq_b[stream.event.header.seq_len];
629 };
630};
80fd2569 631
1ab22b2a 632The sequence elements follow the "array" specifications.
5ba9f198 633
fcba70d4 6344.2.5 Strings
5ba9f198
MD
635
636Strings are an array of bytes of variable size and are terminated by a '\0'
6672e9e1
MD
637"NULL" character. Their encoding is described in the TSDL meta-data. In
638absence of encoding attribute information, the default encoding is
639UTF-8.
5ba9f198 640
6672e9e1 641TSDL meta-data representation of a named string type:
80fd2569 642
359894ac 643typealias string {
5ba9f198 644 encoding = UTF8 OR ASCII;
38b8da21 645} := name;
5ba9f198 646
80fd2569
MD
647A nameless string type can be declared as a field type:
648
649string field_name; /* Use default UTF8 encoding */
5ba9f198 650
ec4404a7
MD
651Strings are always aligned on byte size.
652
3bf79539
MD
6535. Event Packet Header
654
6672e9e1
MD
655The event packet header consists of two parts: the "event packet header"
656is the same for all streams of a trace. The second part, the "event
657packet context", is described on a per-stream basis. Both are described
658in the TSDL meta-data. The packets are aligned on architecture-page-sized
659addresses.
3bf79539 660
6672e9e1 661Event packet header (all fields are optional, specified by TSDL meta-data):
3bf79539 662
6672e9e1
MD
663- Magic number (CTF magic number: 0xC1FC1FC1) specifies that this is a
664 CTF packet. This magic number is optional, but when present, it should
665 come at the very beginning of the packet.
666- Trace UUID, used to ensure the event packet match the meta-data used.
667 (note: we cannot use a meta-data checksum in every cases instead of a
668 UUID because meta-data can be appended to while tracing is active)
669 This field is optional.
670- Stream ID, used as reference to stream description in meta-data.
671 This field is optional if there is only one stream description in the
672 meta-data, but becomes required if there are more than one stream in
673 the TSDL meta-data description.
3bf79539 674
6672e9e1 675Event packet context (all fields are optional, specified by TSDL meta-data):
3bf79539 676
4dfca05b
MD
677- Event packet content size (in bits).
678- Event packet size (in bits, includes padding).
cda89682 679- Event packet content checksum. Checksum excludes the event packet
3bf79539
MD
680 header.
681- Per-stream event packet sequence count (to deal with UDP packet loss). The
682 number of significant sequence counter bits should also be present, so
b11853af 683 wrap-arounds are dealt with correctly.
6672e9e1 684- Time-stamp at the beginning and time-stamp at the end of the event packet.
3bf79539
MD
685 Both timestamps are written in the packet header, but sampled respectively
686 while (or before) writing the first event and while (or after) writing the
687 last event in the packet. The inclusive range between these timestamps should
688 include all event timestamps assigned to events contained within the packet.
5ba9f198 689- Events discarded count
3bf79539
MD
690 - Snapshot of a per-stream free-running counter, counting the number of
691 events discarded that were supposed to be written in the stream prior to
692 the first event in the event packet.
5ba9f198 693 * Note: producer-consumer buffer full condition should fill the current
3bf79539 694 event packet with padding so we know exactly where events have been
5ba9f198 695 discarded.
3bf79539
MD
696- Lossless compression scheme used for the event packet content. Applied
697 directly to raw data. New types of compression can be added in following
698 versions of the format.
5ba9f198
MD
699 0: no compression scheme
700 1: bzip2
701 2: gzip
3bf79539
MD
702 3: xz
703- Cypher used for the event packet content. Applied after compression.
5ba9f198
MD
704 0: no encryption
705 1: AES
3bf79539 706- Checksum scheme used for the event packet content. Applied after encryption.
5ba9f198
MD
707 0: no checksum
708 1: md5
709 2: sha1
710 3: crc32
711
6672e9e1 7125.1 Event Packet Header Description
3bf79539 713
fc5425db 714The event packet header layout is indicated by the trace packet.header
6672e9e1
MD
715field. Here is a recommended structure type for the packet header with
716the fields typically expected (although these fields are each optional):
fc5425db 717
80fd2569
MD
718struct event_packet_header {
719 uint32_t magic;
3fde5da1 720 uint8_t uuid[16];
3bf79539 721 uint32_t stream_id;
80fd2569 722};
5ba9f198 723
fc5425db
MD
724trace {
725 ...
726 packet.header := struct event_packet_header;
727};
728
6672e9e1
MD
729If the magic number is not present, tools such as "file" will have no
730mean to discover the file type.
731
3fde5da1
MD
732If the uuid is not present, no validation that the meta-data actually
733corresponds to the stream is performed.
37b39e47
MD
734
735If the stream_id packet header field is missing, the trace can only
736contain a single stream. Its "id" field can be left out, and its events
737don't need to declare a "stream_id" field.
738
739
3bf79539
MD
7405.2 Event Packet Context Description
741
742Event packet context example. These are declared within the stream declaration
6672e9e1 743in the meta-data. All these fields are optional. If the packet size field is
6a7c61df 744missing, the whole stream only contains a single packet. If the content
f6b81726
MD
745size field is missing, the packet is filled (no padding). The content
746and packet sizes include all headers.
3bf79539
MD
747
748An example event packet context type:
749
80fd2569 750struct event_packet_context {
3bf79539
MD
751 uint64_t timestamp_begin;
752 uint64_t timestamp_end;
753 uint32_t checksum;
754 uint32_t stream_packet_count;
755 uint32_t events_discarded;
756 uint32_t cpu_id;
757 uint32_t/uint16_t content_size;
758 uint32_t/uint16_t packet_size;
3bf79539
MD
759 uint8_t compression_scheme;
760 uint8_t encryption_scheme;
3b0f8e4d 761 uint8_t checksum_scheme;
3bf79539 762};
5ba9f198 763
fcba70d4 764
5ba9f198
MD
7656. Event Structure
766
767The overall structure of an event is:
768
6672e9e1
MD
7691 - Stream Packet Context (as specified by the stream meta-data)
770 2 - Event Header (as specified by the stream meta-data)
771 3 - Stream Event Context (as specified by the stream meta-data)
772 4 - Event Context (as specified by the event meta-data)
773 5 - Event Payload (as specified by the event meta-data)
5ba9f198 774
fdf2bb05 775This structure defines an implicit dynamic scoping, where variants
7d9d7e92
MD
776located in inner structures (those with a higher number in the listing
777above) can refer to the fields of outer structures (with lower number in
6c7226e9 778the listing above). See Section 7.3 TSDL Scopes for more detail.
5ba9f198 779
fdf2bb05 7806.1 Event Header
fcba70d4 781
6672e9e1 782Event headers can be described within the meta-data. We hereby propose, as an
fcba70d4
MD
783example, two types of events headers. Type 1 accommodates streams with less than
78431 event IDs. Type 2 accommodates streams with 31 or more event IDs.
5ba9f198 785
3bf79539
MD
786One major factor can vary between streams: the number of event IDs assigned to
787a stream. Luckily, this information tends to stay relatively constant (modulo
5ba9f198 788event registration while trace is being recorded), so we can specify different
3bf79539 789representations for streams containing few event IDs and streams containing
6672e9e1
MD
790many event IDs, so we end up representing the event ID and time-stamp as
791densely as possible in each case.
5ba9f198 792
fcba70d4
MD
793The header is extended in the rare occasions where the information cannot be
794represented in the ranges available in the standard event header. They are also
3bf79539
MD
795used in the rare occasions where the data required for a field could not be
796collected: the flag corresponding to the missing field within the missing_fields
797array is then set to 1.
5ba9f198 798
6672e9e1
MD
799Types uintX_t represent an X-bit unsigned integer, as declared with
800either:
5ba9f198 801
6672e9e1
MD
802 typealias integer { size = X; align = X; signed = false } := uintX_t;
803
804 or
805
806 typealias integer { size = X; align = 1; signed = false } := uintX_t;
5ba9f198 807
fdf2bb05 8086.1.1 Type 1 - Few event IDs
5ba9f198
MD
809
810 - Aligned on 32-bit (or 8-bit if byte-packed, depending on the architecture
811 preference).
5ba9f198 812 - Native architecture byte ordering.
fcba70d4
MD
813 - For "compact" selection
814 - Fixed size: 32 bits.
815 - For "extended" selection
816 - Size depends on the architecture and variant alignment.
5ba9f198 817
80fd2569 818struct event_header_1 {
fcba70d4
MD
819 /*
820 * id: range: 0 - 30.
821 * id 31 is reserved to indicate an extended header.
822 */
a9b83695 823 enum : uint5_t { compact = 0 ... 30, extended = 31 } id;
fcba70d4
MD
824 variant <id> {
825 struct {
826 uint27_t timestamp;
827 } compact;
828 struct {
829 uint32_t id; /* 32-bit event IDs */
830 uint64_t timestamp; /* 64-bit timestamps */
831 } extended;
832 } v;
cb108fea 833} align(32); /* or align(8) */
5ba9f198 834
5ba9f198 835
fdf2bb05 8366.1.2 Type 2 - Many event IDs
5ba9f198 837
fcba70d4 838 - Aligned on 16-bit (or 8-bit if byte-packed, depending on the architecture
5ba9f198 839 preference).
5ba9f198 840 - Native architecture byte ordering.
fcba70d4
MD
841 - For "compact" selection
842 - Size depends on the architecture and variant alignment.
843 - For "extended" selection
844 - Size depends on the architecture and variant alignment.
5ba9f198 845
80fd2569 846struct event_header_2 {
fcba70d4
MD
847 /*
848 * id: range: 0 - 65534.
849 * id 65535 is reserved to indicate an extended header.
850 */
a9b83695 851 enum : uint16_t { compact = 0 ... 65534, extended = 65535 } id;
fcba70d4
MD
852 variant <id> {
853 struct {
854 uint32_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(16); /* or align(8) */
5ba9f198 862
5ba9f198
MD
863
8646.2 Event Context
865
6672e9e1
MD
866The event context contains information relative to the current event.
867The choice and meaning of this information is specified by the TSDL
868stream and event meta-data descriptions. The stream context is applied
869to all events within the stream. The stream context structure follows
870the event header. The event context is applied to specific events. Its
871structure follows the stream context structure.
5ba9f198 872
fcba70d4
MD
873An example of stream-level event context is to save the event payload size with
874each event, or to save the current PID with each event. These are declared
6672e9e1 875within the stream declaration within the meta-data:
5ba9f198 876
fcba70d4
MD
877 stream {
878 ...
6672e9e1 879 event.context := struct {
80fd2569
MD
880 uint pid;
881 uint16_t payload_size;
6672e9e1 882 };
fcba70d4
MD
883 };
884
885An example of event-specific event context is to declare a bitmap of missing
886fields, only appended after the stream event context if the extended event
887header is selected. NR_FIELDS is the number of fields within the event (a
888numeric value).
5ba9f198 889
fcba70d4
MD
890 event {
891 context = struct {
892 variant <id> {
893 struct { } compact;
894 struct {
895 uint1_t missing_fields[NR_FIELDS]; /* missing event fields bitmap */
896 } extended;
897 } v;
898 };
899 ...
900 }
5ba9f198
MD
901
9026.3 Event Payload
903
904An event payload contains fields specific to a given event type. The fields
6672e9e1 905belonging to an event type are described in the event-specific meta-data
5ba9f198
MD
906within a structure type.
907
9086.3.1 Padding
909
910No padding at the end of the event payload. This differs from the ISO/C standard
911for structures, but follows the CTF standard for structures. In a trace, even
912though it makes sense to align the beginning of a structure, it really makes no
913sense to add padding at the end of the structure, because structures are usually
914not followed by a structure of the same type.
915
916This trick can be done by adding a zero-length "end" field at the end of the C
917structures, and by using the offset of this field rather than using sizeof()
3bf79539 918when calculating the size of a structure (see Appendix "A. Helper macros").
5ba9f198
MD
919
9206.3.2 Alignment
921
922The event payload is aligned on the largest alignment required by types
923contained within the payload. (This follows the ISO/C standard for structures)
924
925
6c7226e9
MD
9267. Trace Stream Description Language (TSDL)
927
928The Trace Stream Description Language (TSDL) allows expression of the
929binary trace streams layout in a C99-like Domain Specific Language
930(DSL).
931
932
6672e9e1 9337.1 Meta-data
6c7226e9
MD
934
935The trace stream layout description is located in the trace meta-data.
936The meta-data is itself located in a stream identified by its name:
937"metadata".
5ba9f198 938
6672e9e1
MD
939The meta-data description can be expressed in two different formats:
940text-only and packet-based. The text-only description facilitates
941generation of meta-data and provides a convenient way to enter the
942meta-data information by hand. The packet-based meta-data provides the
943CTF stream packet facilities (checksumming, compression, encryption,
944network-readiness) for meta-data stream generated and transported by a
945tracer.
946
1b4d35eb
MD
947The text-only meta-data file is a plain-text TSDL description. This file
948must begin with the following characters to identify the file as a CTF
9486a18c 949TSDL text-based metadata file (without the double-quotes) :
1b4d35eb 950
ec2b4db8 951"/* CTF"
1b4d35eb 952
ec2b4db8
MD
953It must be followed by a space, and the version of the specification
954followed by the CTF trace, e.g.:
955
956" 1.8"
957
958These characters allow automated discovery of file type and CTF
959specification version. They are interpreted as a the beginning of a
960comment by the TSDL metadata parser. The comment can be continued to
961contain extra commented characters before it is closed.
6672e9e1
MD
962
963The packet-based meta-data is made of "meta-data packets", which each
964start with a meta-data packet header. The packet-based meta-data
965description is detected by reading the magic number "0x75D11D57" at the
966beginning of the file. This magic number is also used to detect the
967endianness of the architecture by trying to read the CTF magic number
968and its counterpart in reversed endianness. The events within the
969meta-data stream have no event header nor event context. Each event only
5ae2352d 970contains a "sequence" payload, which is a sequence of bits using the
ce2228da
MD
971"trace.packet.header.content_size" field as a placeholder for its length
972(the packet header size should be substracted). The formatting of this
973sequence of bits is a plain-text representation of the TSDL description.
974Each meta-data packet start with a special packet header, specific to
975the meta-data stream, which contains, exactly:
6672e9e1
MD
976
977struct metadata_packet_header {
2daeaa3a 978 uint32_t magic; /* 0x75D11D57 */
3fde5da1 979 uint8_t uuid[16]; /* Unique Universal Identifier */
6672e9e1
MD
980 uint32_t checksum; /* 0 if unused */
981 uint32_t content_size; /* in bits */
982 uint32_t packet_size; /* in bits */
983 uint8_t compression_scheme; /* 0 if unused */
984 uint8_t encryption_scheme; /* 0 if unused */
985 uint8_t checksum_scheme; /* 0 if unused */
ec2b4db8
MD
986 uint8_t major; /* CTF spec version major number */
987 uint8_t minor; /* CTF spec version minor number */
6672e9e1
MD
988};
989
990The packet-based meta-data can be converted to a text-only meta-data by
991concatenating all the strings in contains.
4fafe1ad 992
6672e9e1
MD
993In the textual representation of the meta-data, the text contained
994within "/*" and "*/", as well as within "//" and end of line, are
995treated as comments. Boolean values can be represented as true, TRUE,
996or 1 for true, and false, FALSE, or 0 for false. Within the string-based
997meta-data description, the trace UUID is represented as a string of
998hexadecimal digits and dashes "-". In the event packet header, the trace
999UUID is represented as an array of bytes.
fcba70d4 1000
fdf2bb05 1001
6c7226e9 10027.2 Declaration vs Definition
fdf2bb05
MD
1003
1004A declaration associates a layout to a type, without specifying where
1005this type is located in the event structure hierarchy (see Section 6).
1006This therefore includes typedef, typealias, as well as all type
1007specifiers. In certain circumstances (typedef, structure field and
1008variant field), a declaration is followed by a declarator, which specify
1009the newly defined type name (for typedef), or the field name (for
1010declarations located within structure and variants). Array and sequence,
1011declared with square brackets ("[" "]"), are part of the declarator,
a9b83695 1012similarly to C99. The enumeration base type is specified by
6c7226e9 1013": enum_base", which is part of the type specifier. The variant tag
a9b83695 1014name, specified between "<" ">", is also part of the type specifier.
fdf2bb05
MD
1015
1016A definition associates a type to a location in the event structure
b9606a77
MD
1017hierarchy (see Section 6). This association is denoted by ":=", as shown
1018in Section 7.3.
fdf2bb05
MD
1019
1020
6c7226e9 10217.3 TSDL Scopes
fdf2bb05 1022
37ab95c3
MD
1023TSDL uses three different types of scoping: a lexical scope is used for
1024declarations and type definitions, and static and dynamic scopes are
1025used for variants references to tag fields (with relative and absolute
1026path lookups) and for sequence references to length fields.
fdf2bb05 1027
6c7226e9 10287.3.1 Lexical Scope
fdf2bb05 1029
d285084f
MD
1030Each of "trace", "stream", "event", "struct" and "variant" have their own
1031nestable declaration scope, within which types can be declared using "typedef"
fdf2bb05 1032and "typealias". A root declaration scope also contains all declarations
7d9d7e92 1033located outside of any of the aforementioned declarations. An inner
fdf2bb05 1034declaration scope can refer to type declared within its container
7d9d7e92
MD
1035lexical scope prior to the inner declaration scope. Redefinition of a
1036typedef or typealias is not valid, although hiding an upper scope
fdf2bb05
MD
1037typedef or typealias is allowed within a sub-scope.
1038
37ab95c3 10397.3.2 Static and Dynamic Scopes
fdf2bb05 1040
37ab95c3
MD
1041A local static scope consists in the scope generated by the declaration
1042of fields within a compound type. A static scope is a local static scope
1043augmented with the nested sub-static-scopes it contains.
1044
1045A dynamic scope consists in the static scope augmented with the
7d9d7e92 1046implicit event structure definition hierarchy presented at Section 6.
fdf2bb05 1047
37ab95c3
MD
1048Multiple declarations of the same field name within a local static scope
1049is not valid. It is however valid to re-use the same field name in
1050different local scopes.
1051
1052Nested static and dynamic scopes form lookup paths. These are used for
1053variant tag and sequence length references. They are used at the variant
1054and sequence definition site to look up the location of the tag field
1055associated with a variant, and to lookup up the location of the length
1056field associated with a sequence.
1057
1058Variants and sequences can refer to a tag field either using a relative
4cac83ee
MD
1059path or an absolute path. The relative path is relative to the scope in
1060which the variant or sequence performing the lookup is located.
37ab95c3
MD
1061Relative paths are only allowed to lookup within the same static scope,
1062which includes its nested static scopes. Lookups targeting parent static
1063scopes need to be performed with an absolute path.
1064
1065Absolute path lookups use the full path including the dynamic scope
1066followed by a "." and then the static scope. Therefore, variants (or
1067sequences) in lower levels in the dynamic scope (e.g. event context) can
1068refer to a tag (or length) field located in upper levels (e.g. in the
1069event header) by specifying, in this case, the associated tag with
1070<stream.event.header.field_name>. This allows, for instance, the event
1071context to define a variant referring to the "id" field of the event
1072header as selector.
1073
284724ae 1074The dynamic scope prefixes are thus:
fdf2bb05 1075
e0d9e2c7 1076 - Trace Packet Header: <trace.packet.header. >,
7d9d7e92
MD
1077 - Stream Packet Context: <stream.packet.context. >,
1078 - Event Header: <stream.event.header. >,
1079 - Stream Event Context: <stream.event.context. >,
1080 - Event Context: <event.context. >,
1081 - Event Payload: <event.fields. >.
fdf2bb05 1082
37ab95c3
MD
1083
1084The target dynamic scope must be specified explicitly when referring to
4cac83ee
MD
1085a field outside of the static scope (absolute scope reference). No
1086conflict can occur between relative and dynamic paths, because the
1087keywords "trace", "stream", and "event" are reserved, and thus
1088not permitted as field names. It is recommended that field names
1089clashing with CTF and C99 reserved keywords use an underscore prefix to
1090eliminate the risk of generating a description containing an invalid
70375f92
MD
1091field name. Consequently, fields starting with an underscore should have
1092their leading underscore removed by the CTF parser.
1093
fdf2bb05 1094
457d8b0a
MD
1095The information available in the dynamic scopes can be thought of as the
1096current tracing context. At trace production, information about the
1097current context is saved into the specified scope field levels. At trace
1098consumption, for each event, the current trace context is therefore
1099readable by accessing the upper dynamic scopes.
1100
fdf2bb05 1101
6c7226e9 11027.4 TSDL Examples
d285084f 1103
6672e9e1 1104The grammar representing the TSDL meta-data is presented in Appendix C.
7df6b93a 1105TSDL Grammar. This section presents a rather lighter reading that
6672e9e1 1106consists in examples of TSDL meta-data, with template values.
969f30c0 1107
350bb633
MD
1108The stream "id" can be left out if there is only one stream in the
1109trace. The event "id" field can be left out if there is only one event
1110in a stream.
1111
5ba9f198 1112trace {
fdf2bb05 1113 major = value; /* Trace format version */
5ba9f198 1114 minor = value;
fdf2bb05 1115 uuid = "aaaaaaaa-aaaa-aaaa-aaaa-aaaaaaaaaaaa"; /* Trace UUID */
58997e9e 1116 byte_order = be OR le; /* Endianness (required) */
fc5425db
MD
1117 packet.header := struct {
1118 uint32_t magic;
3fde5da1 1119 uint8_t uuid[16];
fc5425db
MD
1120 uint32_t stream_id;
1121 };
3bf79539 1122};
5ba9f198 1123
3bf79539
MD
1124stream {
1125 id = stream_id;
fdf2bb05 1126 /* Type 1 - Few event IDs; Type 2 - Many event IDs. See section 6.1. */
4fa992a5
MD
1127 event.header := event_header_1 OR event_header_2;
1128 event.context := struct {
77a98c82 1129 ...
3bf79539 1130 };
4fa992a5 1131 packet.context := struct {
77a98c82 1132 ...
3bf79539
MD
1133 };
1134};
5ba9f198
MD
1135
1136event {
980015f9 1137 name = "event_name";
3bf79539 1138 id = value; /* Numeric identifier within the stream */
67f02e24 1139 stream_id = stream_id;
980015f9
MD
1140 loglevel.identifier = "loglevel_identifier";
1141 loglevel.value = value;
4fa992a5 1142 context := struct {
fcba70d4
MD
1143 ...
1144 };
4fa992a5 1145 fields := struct {
80fd2569
MD
1146 ...
1147 };
3bf79539 1148};
5ba9f198
MD
1149
1150/* More detail on types in section 4. Types */
1151
3d13ef1a
MD
1152/*
1153 * Named types:
1154 *
4fa992a5 1155 * Type declarations behave similarly to the C standard.
3d13ef1a
MD
1156 */
1157
80af8ac6 1158typedef aliased_type_specifiers new_type_declarators;
2152348f 1159
3d13ef1a 1160/* e.g.: typedef struct example new_type_name[10]; */
80fd2569 1161
4fa992a5
MD
1162/*
1163 * typealias
1164 *
1165 * The "typealias" declaration can be used to give a name (including
80af8ac6
MD
1166 * pointer declarator specifier) to a type. It should also be used to
1167 * map basic C types (float, int, unsigned long, ...) to a CTF type.
1168 * Typealias is a superset of "typedef": it also allows assignment of a
38b8da21 1169 * simple variable identifier to a type.
4fa992a5
MD
1170 */
1171
1172typealias type_class {
80fd2569 1173 ...
38b8da21 1174} := type_specifiers type_declarator;
2152348f 1175
3d13ef1a
MD
1176/*
1177 * e.g.:
4fa992a5 1178 * typealias integer {
3d13ef1a
MD
1179 * size = 32;
1180 * align = 32;
1181 * signed = false;
38b8da21 1182 * } := struct page *;
359894ac
MD
1183 *
1184 * typealias integer {
1185 * size = 32;
1186 * align = 32;
1187 * signed = true;
38b8da21 1188 * } := int;
3d13ef1a 1189 */
80fd2569
MD
1190
1191struct name {
3bf79539
MD
1192 ...
1193};
5ba9f198 1194
fcba70d4
MD
1195variant name {
1196 ...
1197};
1198
a9b83695 1199enum name : integer_type {
3bf79539
MD
1200 ...
1201};
1202
2152348f 1203
4fa992a5
MD
1204/*
1205 * Unnamed types, contained within compound type fields, typedef or typealias.
1206 */
2152348f 1207
80fd2569
MD
1208struct {
1209 ...
2152348f 1210}
5ba9f198 1211
ec4404a7
MD
1212struct {
1213 ...
1214} align(value)
1215
fcba70d4
MD
1216variant {
1217 ...
1218}
1219
a9b83695 1220enum : integer_type {
80fd2569 1221 ...
2152348f
MD
1222}
1223
1224typedef type new_type[length];
3bf79539 1225
2152348f
MD
1226struct {
1227 type field_name[length];
1228}
1229
1230typedef type new_type[length_type];
1231
1232struct {
1233 type field_name[length_type];
1234}
1235
1236integer {
80fd2569 1237 ...
2152348f 1238}
3bf79539 1239
2152348f 1240floating_point {
80fd2569 1241 ...
2152348f
MD
1242}
1243
1244struct {
1245 integer_type field_name:size; /* GNU/C bitfield */
1246}
1247
1248struct {
1249 string field_name;
1250}
3bf79539 1251
fcba70d4 1252
2fa70eba
MD
12538. Clocks
1254
1255Clock metadata allows to describe the clock topology of the system, as
1256well as to detail each clock parameter. In absence of clock description,
1257it is assumed that all fields named "timestamp" use the same clock
aed18b5e 1258source, which increments once per nanosecond.
2fa70eba
MD
1259
1260Describing a clock and how it is used by streams is threefold: first,
1261the clock and clock topology should be described in a "clock"
1262description block, e.g.:
1263
d803bfcb 1264clock {
58262d97 1265 name = cycle_counter_sync;
2fa70eba 1266 uuid = "62189bee-96dc-11e0-91a8-cfa3d89f3923";
58262d97 1267 description = "Cycle counter synchronized across CPUs";
2fa70eba
MD
1268 freq = 1000000000; /* frequency, in Hz */
1269 /* precision in seconds is: 1000 * (1/freq) */
1270 precision = 1000;
a40cccda
MD
1271 /*
1272 * clock value offset from Epoch is:
1273 * offset_s + (offset * (1/freq))
1274 */
1275 offset_s = 1326476837;
1276 offset = 897235420;
ce0fadbd 1277 absolute = FALSE;
2fa70eba
MD
1278};
1279
d803bfcb
MD
1280The mandatory "name" field specifies the name of the clock identifier,
1281which can later be used as a reference. The optional field "uuid" is the
1282unique identifier of the clock. It can be used to correlate different
1283traces that use the same clock. An optional textual description string
1284can be added with the "description" field. The "freq" field is the
1285initial frequency of the clock, in Hz. If the "freq" field is not
1286present, the frequency is assumed to be 1000000000 (providing clock
1287increment of 1 ns). The optional "precision" field details the
a40cccda
MD
1288uncertainty on the clock measurements, in (1/freq) units. The "offset_s"
1289and "offset" fields indicate the offset from POSIX.1 Epoch, 1970-01-01
129000:00:00 +0000 (UTC), to the zero of value of the clock. The "offset_s"
1291field is in seconds. The "offset" field is in (1/freq) units. If any of
1292the "offset_s" or "offset" field is not present, it is assigned the 0
ce0fadbd
MD
1293value. The field "absolute" is TRUE if the clock is a global reference
1294across different clock uuid (e.g. NTP time). Otherwise, "absolute" is
1295FALSE, and the clock can be considered as synchronized only with other
1296clocks that have the same uuid.
1297
2fa70eba
MD
1298
1299Secondly, a reference to this clock should be added within an integer
1300type:
1301
1302typealias integer {
1303 size = 64; align = 1; signed = false;
58262d97 1304 map = clock.cycle_counter_sync.value;
2fa70eba
MD
1305} := uint64_ccnt_t;
1306
1307Thirdly, stream declarations can reference the clock they use as a
1308time-stamp source:
1309
1310struct packet_context {
1311 uint64_ccnt_t ccnt_begin;
1312 uint64_ccnt_t ccnt_end;
1313 /* ... */
1314};
1315
1316stream {
1317 /* ... */
1318 event.header := struct {
1319 uint64_ccnt_t timestamp;
1320 /* ... */
1321 }
1322 packet.context := struct packet_context;
1323};
1324
1325For a N-bit integer type referring to a clock, if the integer overflows
1326compared to the N low order bits of the clock prior value, then it is
1327assumed that one, and only one, overflow occurred. It is therefore
1328important that events encoding time on a small number of bits happen
1329frequently enough to detect when more than one N-bit overflow occurs.
1330
1331In a packet context, clock field names ending with "_begin" and "_end"
1332have a special meaning: this refers to the time-stamps at, respectively,
1333the beginning and the end of each packet.
1334
1335
3bf79539 1336A. Helper macros
5ba9f198
MD
1337
1338The two following macros keep track of the size of a GNU/C structure without
1339padding at the end by placing HEADER_END as the last field. A one byte end field
1340is used for C90 compatibility (C99 flexible arrays could be used here). Note
1341that this does not affect the effective structure size, which should always be
1342calculated with the header_sizeof() helper.
1343
1344#define HEADER_END char end_field
1345#define header_sizeof(type) offsetof(typeof(type), end_field)
3bf79539
MD
1346
1347
1348B. Stream Header Rationale
1349
1350An event stream is divided in contiguous event packets of variable size. These
1351subdivisions allow the trace analyzer to perform a fast binary search by time
1352within the stream (typically requiring to index only the event packet headers)
1353without reading the whole stream. These subdivisions have a variable size to
1354eliminate the need to transfer the event packet padding when partially filled
1355event packets must be sent when streaming a trace for live viewing/analysis.
1356An event packet can contain a certain amount of padding at the end. Dividing
1357streams into event packets is also useful for network streaming over UDP and
1358flight recorder mode tracing (a whole event packet can be swapped out of the
1359buffer atomically for reading).
1360
1361The stream header is repeated at the beginning of each event packet to allow
1362flexibility in terms of:
1363
1364 - streaming support,
1365 - allowing arbitrary buffers to be discarded without making the trace
1366 unreadable,
1367 - allow UDP packet loss handling by either dealing with missing event packet
1368 or asking for re-transmission.
1369 - transparently support flight recorder mode,
1370 - transparently support crash dump.
1371
6c7226e9
MD
1372
1373C. TSDL Grammar
fcba70d4 1374
4fa992a5 1375/*
6c7226e9 1376 * Common Trace Format (CTF) Trace Stream Description Language (TSDL) Grammar.
4fa992a5
MD
1377 *
1378 * Inspired from the C99 grammar:
1379 * http://www.open-std.org/jtc1/sc22/wg14/www/docs/n1124.pdf (Annex A)
6c7226e9
MD
1380 * and c++1x grammar (draft)
1381 * http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2011/n3291.pdf (Annex A)
4fa992a5
MD
1382 *
1383 * Specialized for CTF needs by including only constant and declarations from
1384 * C99 (excluding function declarations), and by adding support for variants,
6c7226e9
MD
1385 * sequences and CTF-specific specifiers. Enumeration container types
1386 * semantic is inspired from c++1x enum-base.
4fa992a5
MD
1387 */
1388
13891) Lexical grammar
1390
13911.1) Lexical elements
1392
1393token:
1394 keyword
1395 identifier
1396 constant
1397 string-literal
1398 punctuator
1399
14001.2) Keywords
1401
1402keyword: is one of
1403
ec4404a7 1404align
4fa992a5
MD
1405const
1406char
2fa70eba 1407clock
4fa992a5
MD
1408double
1409enum
1410event
1411floating_point
1412float
1413integer
1414int
1415long
1416short
1417signed
1418stream
1419string
1420struct
1421trace
3e1e1a78 1422typealias
4fa992a5
MD
1423typedef
1424unsigned
1425variant
1426void
1427_Bool
1428_Complex
1429_Imaginary
1430
1431
14321.3) Identifiers
1433
1434identifier:
1435 identifier-nondigit
1436 identifier identifier-nondigit
1437 identifier digit
1438
1439identifier-nondigit:
1440 nondigit
1441 universal-character-name
1442 any other implementation-defined characters
1443
1444nondigit:
1445 _
1446 [a-zA-Z] /* regular expression */
1447
1448digit:
1449 [0-9] /* regular expression */
1450
14511.4) Universal character names
1452
1453universal-character-name:
1454 \u hex-quad
1455 \U hex-quad hex-quad
1456
1457hex-quad:
1458 hexadecimal-digit hexadecimal-digit hexadecimal-digit hexadecimal-digit
1459
14601.5) Constants
1461
1462constant:
1463 integer-constant
1464 enumeration-constant
1465 character-constant
1466
1467integer-constant:
1468 decimal-constant integer-suffix-opt
1469 octal-constant integer-suffix-opt
1470 hexadecimal-constant integer-suffix-opt
1471
1472decimal-constant:
1473 nonzero-digit
1474 decimal-constant digit
1475
1476octal-constant:
1477 0
1478 octal-constant octal-digit
1479
1480hexadecimal-constant:
1481 hexadecimal-prefix hexadecimal-digit
1482 hexadecimal-constant hexadecimal-digit
1483
1484hexadecimal-prefix:
1485 0x
1486 0X
1487
1488nonzero-digit:
1489 [1-9]
1490
1491integer-suffix:
1492 unsigned-suffix long-suffix-opt
1493 unsigned-suffix long-long-suffix
1494 long-suffix unsigned-suffix-opt
1495 long-long-suffix unsigned-suffix-opt
1496
1497unsigned-suffix:
1498 u
1499 U
1500
1501long-suffix:
1502 l
1503 L
1504
1505long-long-suffix:
1506 ll
1507 LL
1508
4fa992a5
MD
1509enumeration-constant:
1510 identifier
1511 string-literal
1512
1513character-constant:
1514 ' c-char-sequence '
1515 L' c-char-sequence '
1516
1517c-char-sequence:
1518 c-char
1519 c-char-sequence c-char
1520
1521c-char:
1522 any member of source charset except single-quote ('), backslash
1523 (\), or new-line character.
1524 escape-sequence
1525
1526escape-sequence:
1527 simple-escape-sequence
1528 octal-escape-sequence
1529 hexadecimal-escape-sequence
1530 universal-character-name
1531
1532simple-escape-sequence: one of
1533 \' \" \? \\ \a \b \f \n \r \t \v
1534
1535octal-escape-sequence:
1536 \ octal-digit
1537 \ octal-digit octal-digit
1538 \ octal-digit octal-digit octal-digit
1539
1540hexadecimal-escape-sequence:
1541 \x hexadecimal-digit
1542 hexadecimal-escape-sequence hexadecimal-digit
1543
15441.6) String literals
1545
1546string-literal:
1547 " s-char-sequence-opt "
1548 L" s-char-sequence-opt "
1549
1550s-char-sequence:
1551 s-char
1552 s-char-sequence s-char
1553
1554s-char:
1555 any member of source charset except double-quote ("), backslash
1556 (\), or new-line character.
1557 escape-sequence
1558
15591.7) Punctuators
1560
1561punctuator: one of
1562 [ ] ( ) { } . -> * + - < > : ; ... = ,
1563
1564
15652) Phrase structure grammar
1566
1567primary-expression:
1568 identifier
1569 constant
1570 string-literal
1571 ( unary-expression )
1572
1573postfix-expression:
1574 primary-expression
1575 postfix-expression [ unary-expression ]
1576 postfix-expression . identifier
1577 postfix-expressoin -> identifier
1578
1579unary-expression:
1580 postfix-expression
1581 unary-operator postfix-expression
1582
1583unary-operator: one of
1584 + -
1585
4fa992a5
MD
1586assignment-operator:
1587 =
1588
b9606a77
MD
1589type-assignment-operator:
1590 :=
1591
4fa992a5 1592constant-expression-range:
73d61ac3 1593 unary-expression ... unary-expression
4fa992a5
MD
1594
15952.2) Declarations:
1596
1597declaration:
689e04b4 1598 declaration-specifiers declarator-list-opt ;
4fa992a5
MD
1599 ctf-specifier ;
1600
1601declaration-specifiers:
689e04b4 1602 storage-class-specifier declaration-specifiers-opt
4fa992a5
MD
1603 type-specifier declaration-specifiers-opt
1604 type-qualifier declaration-specifiers-opt
1605
1606declarator-list:
1607 declarator
1608 declarator-list , declarator
1609
d285084f
MD
1610abstract-declarator-list:
1611 abstract-declarator
1612 abstract-declarator-list , abstract-declarator
1613
4fa992a5
MD
1614storage-class-specifier:
1615 typedef
1616
1617type-specifier:
1618 void
1619 char
1620 short
1621 int
1622 long
1623 float
1624 double
1625 signed
1626 unsigned
1627 _Bool
1628 _Complex
cfdd51ec 1629 _Imaginary
9dfcfc0f
MD
1630 struct-specifier
1631 variant-specifier
4fa992a5
MD
1632 enum-specifier
1633 typedef-name
1634 ctf-type-specifier
1635
ec4404a7 1636align-attribute:
73d61ac3 1637 align ( unary-expression )
ec4404a7 1638
4fa992a5 1639struct-specifier:
ec4404a7
MD
1640 struct identifier-opt { struct-or-variant-declaration-list-opt } align-attribute-opt
1641 struct identifier align-attribute-opt
4fa992a5
MD
1642
1643struct-or-variant-declaration-list:
1644 struct-or-variant-declaration
1645 struct-or-variant-declaration-list struct-or-variant-declaration
1646
1647struct-or-variant-declaration:
1648 specifier-qualifier-list struct-or-variant-declarator-list ;
eacb16d1 1649 declaration-specifiers-opt storage-class-specifier declaration-specifiers-opt declarator-list ;
6143bab7
MD
1650 typealias declaration-specifiers abstract-declarator-list type-assignment-operator declaration-specifiers abstract-declarator-list ;
1651 typealias declaration-specifiers abstract-declarator-list type-assignment-operator declarator-list ;
4fa992a5
MD
1652
1653specifier-qualifier-list:
1654 type-specifier specifier-qualifier-list-opt
1655 type-qualifier specifier-qualifier-list-opt
1656
1657struct-or-variant-declarator-list:
1658 struct-or-variant-declarator
1659 struct-or-variant-declarator-list , struct-or-variant-declarator
1660
1661struct-or-variant-declarator:
1662 declarator
73d61ac3 1663 declarator-opt : unary-expression
4fa992a5
MD
1664
1665variant-specifier:
1666 variant identifier-opt variant-tag-opt { struct-or-variant-declaration-list }
1667 variant identifier variant-tag
1668
1669variant-tag:
37ab95c3 1670 < unary-expression >
4fa992a5
MD
1671
1672enum-specifier:
1673 enum identifier-opt { enumerator-list }
1674 enum identifier-opt { enumerator-list , }
1675 enum identifier
a9b83695
MD
1676 enum identifier-opt : declaration-specifiers { enumerator-list }
1677 enum identifier-opt : declaration-specifiers { enumerator-list , }
4fa992a5
MD
1678
1679enumerator-list:
1680 enumerator
1681 enumerator-list , enumerator
1682
1683enumerator:
1684 enumeration-constant
8d2d41f7
MD
1685 enumeration-constant assignment-operator unary-expression
1686 enumeration-constant assignment-operator constant-expression-range
4fa992a5
MD
1687
1688type-qualifier:
1689 const
1690
1691declarator:
1692 pointer-opt direct-declarator
1693
1694direct-declarator:
1695 identifier
1696 ( declarator )
1ab22b2a 1697 direct-declarator [ unary-expression ]
4fa992a5 1698
d285084f
MD
1699abstract-declarator:
1700 pointer-opt direct-abstract-declarator
1701
1702direct-abstract-declarator:
1703 identifier-opt
1704 ( abstract-declarator )
1ab22b2a 1705 direct-abstract-declarator [ unary-expression ]
d285084f
MD
1706 direct-abstract-declarator [ ]
1707
4fa992a5 1708pointer:
3b0f8e4d
MD
1709 * type-qualifier-list-opt
1710 * type-qualifier-list-opt pointer
4fa992a5
MD
1711
1712type-qualifier-list:
1713 type-qualifier
1714 type-qualifier-list type-qualifier
1715
4fa992a5
MD
1716typedef-name:
1717 identifier
1718
17192.3) CTF-specific declarations
1720
1721ctf-specifier:
1722 event { ctf-assignment-expression-list-opt }
1723 stream { ctf-assignment-expression-list-opt }
1724 trace { ctf-assignment-expression-list-opt }
b12919a5
MD
1725 typealias declaration-specifiers abstract-declarator-list type-assignment-operator declaration-specifiers abstract-declarator-list
1726 typealias declaration-specifiers abstract-declarator-list type-assignment-operator declarator-list
4fa992a5
MD
1727
1728ctf-type-specifier:
1729 floating_point { ctf-assignment-expression-list-opt }
1730 integer { ctf-assignment-expression-list-opt }
1731 string { ctf-assignment-expression-list-opt }
7609d3c7 1732 string
4fa992a5
MD
1733
1734ctf-assignment-expression-list:
13441bf5
MD
1735 ctf-assignment-expression ;
1736 ctf-assignment-expression-list ctf-assignment-expression ;
4fa992a5
MD
1737
1738ctf-assignment-expression:
1739 unary-expression assignment-operator unary-expression
1740 unary-expression type-assignment-operator type-specifier
eacb16d1 1741 declaration-specifiers-opt storage-class-specifier declaration-specifiers-opt declarator-list
6143bab7
MD
1742 typealias declaration-specifiers abstract-declarator-list type-assignment-operator declaration-specifiers abstract-declarator-list
1743 typealias declaration-specifiers abstract-declarator-list type-assignment-operator declarator-list
This page took 0.201763 seconds and 4 git commands to generate.