Update alignment section
[ctf.git] / common-trace-format-proposal.txt
CommitLineData
5ba9f198 1
4767a9e7 2RFC: Common Trace Format (CTF) Proposal (pre-v1.7)
5ba9f198
MD
3
4Mathieu Desnoyers, EfficiOS Inc.
5
6The goal of the present document is to propose a trace format that suits the
cc089c3a 7needs of the embedded, telecom, high-performance and kernel communities. It is
5ba9f198 8based on the Common Trace Format Requirements (v1.4) document. It is designed to
cc089c3a
MD
9allow traces to be natively generated by the Linux kernel, Linux user-space
10applications written in C/C++, and hardware components.
11
12The latest version of this document can be found at:
13
14 git tree: git://git.efficios.com/ctf.git
15 gitweb: http://git.efficios.com/?p=ctf.git
5ba9f198
MD
16
17A reference implementation of a library to read and write this trace format is
18being implemented within the BabelTrace project, a converter between trace
19formats. The development tree is available at:
20
21 git tree: git://git.efficios.com/babeltrace.git
22 gitweb: http://git.efficios.com/?p=babeltrace.git
23
24
251. Preliminary definitions
26
3bf79539
MD
27 - Event Trace: An ordered sequence of events.
28 - Event Stream: An ordered sequence of events, containing a subset of the
29 trace event types.
30 - Event Packet: A sequence of physically contiguous events within an event
31 stream.
5ba9f198
MD
32 - Event: This is the basic entry in a trace. (aka: a trace record).
33 - An event identifier (ID) relates to the class (a type) of event within
3bf79539
MD
34 an event stream.
35 e.g. event: irq_entry.
5ba9f198
MD
36 - An event (or event record) relates to a specific instance of an event
37 class.
3bf79539
MD
38 e.g. event: irq_entry, at time X, on CPU Y
39 - Source Architecture: Architecture writing the trace.
40 - Reader Architecture: Architecture reading the trace.
5ba9f198
MD
41
42
432. High-level representation of a trace
44
3bf79539
MD
45A trace is divided into multiple event streams. Each event stream contains a
46subset of the trace event types.
5ba9f198 47
3bf79539
MD
48The final output of the trace, after its generation and optional transport over
49the network, is expected to be either on permanent or temporary storage in a
50virtual file system. Because each event stream is appended to while a trace is
51being recorded, each is associated with a separate file for output. Therefore,
52a stored trace can be represented as a directory containing one file per stream.
5ba9f198 53
3bf79539 54A metadata event stream contains information on trace event types. It describes:
5ba9f198
MD
55
56- Trace version.
57- Types available.
3bf79539
MD
58- Per-stream event header description.
59- Per-stream event header selection.
60- Per-stream event context fields.
5ba9f198 61- Per-event
3bf79539 62 - Event type to stream mapping.
5ba9f198
MD
63 - Event type to name mapping.
64 - Event type to ID mapping.
65 - Event fields description.
66
67
3bf79539 683. Event stream
5ba9f198 69
3bf79539 70An event stream is divided in contiguous event packets of variable size. These
ae8c075a
MD
71subdivisions have a variable size. An event packet can contain a certain
72amount of padding at the end. The stream header is repeated at the
73beginning of each event packet. The rationale for the event stream
74design choices is explained in Appendix B. Stream Header Rationale.
5ba9f198 75
3bf79539
MD
76The event stream header will therefore be referred to as the "event packet
77header" throughout the rest of this document.
5ba9f198
MD
78
79
804. Types
81
1fad7a85
MD
82Types are organized as type classes. Each type class belong to either of two
83kind of types: basic types or compound types.
84
5ba9f198
MD
854.1 Basic types
86
1fad7a85
MD
87A basic type is a scalar type, as described in this section. It includes
88integers, GNU/C bitfields, enumerations, and floating point values.
5ba9f198
MD
89
904.1.1 Type inheritance
91
80fd2569
MD
92Type specifications can be inherited to allow deriving types from a
93type class. For example, see the uint32_t named type derived from the "integer"
94type class below ("Integers" section). Types have a precise binary
95representation in the trace. A type class has methods to read and write these
96types, but must be derived into a type to be usable in an event field.
5ba9f198
MD
97
984.1.2 Alignment
99
100We define "byte-packed" types as aligned on the byte size, namely 8-bit.
101We define "bit-packed" types as following on the next bit, as defined by the
370eae99 102"Integers" section.
5ba9f198 103
3bf79539
MD
104All basic types, except bitfields, are either aligned on an architecture-defined
105specific alignment or byte-packed, depending on the architecture preference.
106Architectures providing fast unaligned write byte-packed basic types to save
5ba9f198 107space, aligning each type on byte boundaries (8-bit). Architectures with slow
3bf79539 108unaligned writes align types on specific alignment values. If no specific
370eae99
MD
109alignment is declared for a type, it is assumed to be bit-packed for
110integers with size not multiple of 8 bits and for gcc bitfields. All
111other types are byte-packed.
5ba9f198 112
3bf79539 113Metadata attribute representation of a specific alignment:
5ba9f198
MD
114
115 align = value; /* value in bits */
116
1174.1.3 Byte order
118
3bf79539
MD
119By default, the native endianness of the source architecture the trace is used.
120Byte order can be overridden for a basic type by specifying a "byte_order"
121attribute. Typical use-case is to specify the network byte order (big endian:
122"be") to save data captured from the network into the trace without conversion.
123If not specified, the byte order is native.
5ba9f198
MD
124
125Metadata representation:
126
127 byte_order = native OR network OR be OR le; /* network and be are aliases */
128
1294.1.4 Size
130
131Type size, in bits, for integers and floats is that returned by "sizeof()" in C
132multiplied by CHAR_BIT.
133We require the size of "char" and "unsigned char" types (CHAR_BIT) to be fixed
134to 8 bits for cross-endianness compatibility.
135
136Metadata representation:
137
138 size = value; (value is in bits)
139
1404.1.5 Integers
141
142Signed integers are represented in two-complement. Integer alignment, size,
143signedness and byte ordering are defined in the metadata. Integers aligned on
144byte size (8-bit) and with length multiple of byte size (8-bit) correspond to
145the C99 standard integers. In addition, integers with alignment and/or size that
146are _not_ a multiple of the byte size are permitted; these correspond to the C99
147standard bitfields, with the added specification that the CTF integer bitfields
148have a fixed binary representation. A MIT-licensed reference implementation of
149the CTF portable bitfields is available at:
150
151 http://git.efficios.com/?p=babeltrace.git;a=blob;f=include/babeltrace/bitfield.h
152
153Binary representation of integers:
154
155- On little and big endian:
156 - Within a byte, high bits correspond to an integer high bits, and low bits
157 correspond to low bits.
158- On little endian:
159 - Integer across multiple bytes are placed from the less significant to the
160 most significant.
161 - Consecutive integers are placed from lower bits to higher bits (even within
162 a byte).
163- On big endian:
164 - Integer across multiple bytes are placed from the most significant to the
165 less significant.
166 - Consecutive integers are placed from higher bits to lower bits (even within
167 a byte).
168
169This binary representation is derived from the bitfield implementation in GCC
170for little and big endian. However, contrary to what GCC does, integers can
171cross units boundaries (no padding is required). Padding can be explicitely
172added (see 4.1.6 GNU/C bitfields) to follow the GCC layout if needed.
173
174Metadata representation:
175
80fd2569 176 integer {
5ba9f198
MD
177 signed = true OR false; /* default false */
178 byte_order = native OR network OR be OR le; /* default native */
179 size = value; /* value in bits, no default */
180 align = value; /* value in bits */
2152348f 181 }
5ba9f198 182
80fd2569 183Example of type inheritance (creation of a uint32_t named type):
5ba9f198 184
359894ac 185typealias integer {
9e4e34e9 186 size = 32;
5ba9f198
MD
187 signed = false;
188 align = 32;
359894ac 189} : uint32_t;
5ba9f198 190
80fd2569 191Definition of a named 5-bit signed bitfield:
5ba9f198 192
359894ac 193typealias integer {
5ba9f198
MD
194 size = 5;
195 signed = true;
196 align = 1;
359894ac 197} : int5_t;
5ba9f198
MD
198
1994.1.6 GNU/C bitfields
200
201The GNU/C bitfields follow closely the integer representation, with a
202particularity on alignment: if a bitfield cannot fit in the current unit, the
80fd2569
MD
203unit is padded and the bitfield starts at the following unit. The unit size is
204defined by the size of the type "unit_type".
5ba9f198 205
2152348f 206Metadata representation:
80fd2569
MD
207
208 unit_type name:size:
209
5ba9f198
MD
210As an example, the following structure declared in C compiled by GCC:
211
212struct example {
213 short a:12;
214 short b:5;
215};
216
2152348f
MD
217The example structure is aligned on the largest element (short). The second
218bitfield would be aligned on the next unit boundary, because it would not fit in
219the current unit.
5ba9f198
MD
220
2214.1.7 Floating point
222
223The floating point values byte ordering is defined in the metadata.
224
225Floating point values follow the IEEE 754-2008 standard interchange formats.
226Description of the floating point values include the exponent and mantissa size
227in bits. Some requirements are imposed on the floating point values:
228
229- FLT_RADIX must be 2.
230- mant_dig is the number of digits represented in the mantissa. It is specified
231 by the ISO C99 standard, section 5.2.4, as FLT_MANT_DIG, DBL_MANT_DIG and
232 LDBL_MANT_DIG as defined by <float.h>.
233- exp_dig is the number of digits represented in the exponent. Given that
234 mant_dig is one bit more than its actual size in bits (leading 1 is not
235 needed) and also given that the sign bit always takes one bit, exp_dig can be
236 specified as:
237
238 - sizeof(float) * CHAR_BIT - FLT_MANT_DIG
239 - sizeof(double) * CHAR_BIT - DBL_MANT_DIG
240 - sizeof(long double) * CHAR_BIT - LDBL_MANT_DIG
241
242Metadata representation:
243
80fd2569 244floating_point {
5ba9f198
MD
245 exp_dig = value;
246 mant_dig = value;
247 byte_order = native OR network OR be OR le;
2152348f 248}
5ba9f198
MD
249
250Example of type inheritance:
251
359894ac 252typealias floating_point {
5ba9f198
MD
253 exp_dig = 8; /* sizeof(float) * CHAR_BIT - FLT_MANT_DIG */
254 mant_dig = 24; /* FLT_MANT_DIG */
255 byte_order = native;
359894ac 256} : float;
5ba9f198
MD
257
258TODO: define NaN, +inf, -inf behavior.
259
2604.1.8 Enumerations
261
262Enumerations are a mapping between an integer type and a table of strings. The
263numerical representation of the enumeration follows the integer type specified
264by the metadata. The enumeration mapping table is detailed in the enumeration
3bf79539
MD
265description within the metadata. The mapping table maps inclusive value ranges
266(or single values) to strings. Instead of being limited to simple
267"value -> string" mappings, these enumerations map
80fd2569 268"[ start_value ... end_value ] -> string", which map inclusive ranges of
3bf79539
MD
269values to strings. An enumeration from the C language can be represented in
270this format by having the same start_value and end_value for each element, which
271is in fact a range of size 1. This single-value range is supported without
4767a9e7 272repeating the start and end values with the value = string declaration.
80fd2569 273
4767a9e7
MD
274If a numeric value is encountered between < >, it represents the integer type
275size used to hold the enumeration, in bits.
276
cfc73fdc 277enum name <integer_type OR size> {
359894ac 278 somestring = start_value1 ... end_value1,
80fd2569
MD
279 "other string" = start_value2 ... end_value2,
280 yet_another_string, /* will be assigned to end_value2 + 1 */
281 "some other string" = value,
282 ...
283};
284
285If the values are omitted, the enumeration starts at 0 and increment of 1 for
286each entry:
287
cfc73fdc 288enum name <32> {
80fd2569
MD
289 ZERO,
290 ONE,
291 TWO,
292 TEN = 10,
293 ELEVEN,
3bf79539 294};
5ba9f198 295
80fd2569 296Overlapping ranges within a single enumeration are implementation defined.
5ba9f198 297
2152348f
MD
298A nameless enumeration can be declared as a field type or as part of a typedef:
299
300enum <integer_type> {
301 ...
302}
303
1fad7a85 304
5ba9f198
MD
3054.2 Compound types
306
1fad7a85
MD
307Compound are aggregation of type declarations. Compound types include
308structures, variant, arrays, sequences, and strings.
309
5ba9f198
MD
3104.2.1 Structures
311
312Structures are aligned on the largest alignment required by basic types
313contained within the structure. (This follows the ISO/C standard for structures)
314
80fd2569 315Metadata representation of a named structure:
5ba9f198 316
80fd2569
MD
317struct name {
318 field_type field_name;
319 field_type field_name;
320 ...
321};
5ba9f198
MD
322
323Example:
324
80fd2569
MD
325struct example {
326 integer { /* Nameless type */
327 size = 16;
328 signed = true;
329 align = 16;
330 } first_field_name;
331 uint64_t second_field_name; /* Named type declared in the metadata */
3bf79539 332};
5ba9f198
MD
333
334The fields are placed in a sequence next to each other. They each possess a
335field name, which is a unique identifier within the structure.
336
2152348f 337A nameless structure can be declared as a field type or as part of a typedef:
80fd2569
MD
338
339struct {
340 ...
2152348f 341}
80fd2569 342
77a98c82 3434.2.2 Variants (Discriminated/Tagged Unions)
fcba70d4 344
fdf2bb05
MD
345A CTF variant is a selection between different types. A CTF variant must
346always be defined within the scope of a structure or within fields
347contained within a structure (defined recursively). A "tag" enumeration
348field must appear in either the same lexical scope, prior to the variant
349field (in field declaration order), in an uppermost lexical scope (see
350Section 7.2.1), or in an uppermost dynamic scope (see Section 7.2.2).
351The type selection is indicated by the mapping from the enumeration
352value to the string used as variant type selector. The field to use as
353tag is specified by the "tag_field", specified between "< >" after the
354"variant" keyword for unnamed variants, and after "variant name" for
355named variants.
fcba70d4
MD
356
357The alignment of the variant is the alignment of the type as selected by the tag
358value for the specific instance of the variant. The alignment of the type
359containing the variant is independent of the variant alignment. The size of the
360variant is the size as selected by the tag value for the specific instance of
361the variant.
362
363A named variant declaration followed by its definition within a structure
364declaration:
365
366variant name {
367 field_type sel1;
368 field_type sel2;
369 field_type sel3;
370 ...
371};
372
373struct {
374 enum <integer_type or size> { sel1, sel2, sel3, ... } tag_field;
375 ...
376 variant name <tag_field> v;
377}
378
379An unnamed variant definition within a structure is expressed by the following
380metadata:
381
382struct {
383 enum <integer_type or size> { sel1, sel2, sel3, ... } tag_field;
384 ...
385 variant <tag_field> {
386 field_type sel1;
387 field_type sel2;
388 field_type sel3;
389 ...
390 } v;
391}
392
393Example of a named variant within a sequence that refers to a single tag field:
394
395variant example {
396 uint32_t a;
397 uint64_t b;
398 short c;
399};
400
401struct {
402 enum <uint2_t> { a, b, c } choice;
15850440 403 variant example <choice> v[unsigned int];
fcba70d4
MD
404}
405
406Example of an unnamed variant:
407
408struct {
409 enum <uint2_t> { a, b, c, d } choice;
410 /* Unrelated fields can be added between the variant and its tag */
411 int32_t somevalue;
412 variant <choice> {
413 uint32_t a;
414 uint64_t b;
415 short c;
416 struct {
417 unsigned int field1;
418 uint64_t field2;
419 } d;
420 } s;
421}
422
423Example of an unnamed variant within an array:
424
425struct {
426 enum <uint2_t> { a, b, c } choice;
427 variant <choice> {
428 uint32_t a;
429 uint64_t b;
430 short c;
15850440 431 } v[10];
fcba70d4
MD
432}
433
434Example of a variant type definition within a structure, where the defined type
435is then declared within an array of structures. This variant refers to a tag
436located in an upper lexical scope. This example clearly shows that a variant
437type definition referring to the tag "x" uses the closest preceding field from
438the lexical scope of the type definition.
439
440struct {
441 enum <uint2_t> { a, b, c, d } x;
442
443 typedef variant <x> { /*
444 * "x" refers to the preceding "x" enumeration in the
445 * lexical scope of the type definition.
446 */
447 uint32_t a;
448 uint64_t b;
449 short c;
450 } example_variant;
451
452 struct {
453 enum <int> { x, y, z } x; /* This enumeration is not used by "v". */
454 example_variant v; /*
455 * "v" uses the "enum <uint2_t> { a, b, c, d }"
456 * tag.
457 */
458 } a[10];
459}
460
4614.2.3 Arrays
5ba9f198
MD
462
463Arrays are fixed-length. Their length is declared in the type declaration within
464the metadata. They contain an array of "inner type" elements, which can refer to
465any type not containing the type of the array being declared (no circular
3bf79539 466dependency). The length is the number of elements in an array.
5ba9f198 467
2152348f 468Metadata representation of a named array:
80fd2569
MD
469
470typedef elem_type name[length];
5ba9f198 471
2152348f 472A nameless array can be declared as a field type within a structure, e.g.:
5ba9f198 473
2152348f 474 uint8_t field_name[10];
80fd2569 475
5ba9f198 476
fcba70d4 4774.2.4 Sequences
5ba9f198
MD
478
479Sequences are dynamically-sized arrays. They start with an integer that specify
480the length of the sequence, followed by an array of "inner type" elements.
3bf79539 481The length is the number of elements in the sequence.
5ba9f198 482
2152348f 483Metadata representation for a named sequence:
80fd2569
MD
484
485typedef elem_type name[length_type];
486
487A nameless sequence can be declared as a field type, e.g.:
488
80fd2569
MD
489long field_name[int];
490
491The length type follows the integer types specifications, and the sequence
5ba9f198
MD
492elements follow the "array" specifications.
493
fcba70d4 4944.2.5 Strings
5ba9f198
MD
495
496Strings are an array of bytes of variable size and are terminated by a '\0'
497"NULL" character. Their encoding is described in the metadata. In absence of
498encoding attribute information, the default encoding is UTF-8.
499
80fd2569
MD
500Metadata representation of a named string type:
501
359894ac 502typealias string {
5ba9f198 503 encoding = UTF8 OR ASCII;
359894ac 504} : name;
5ba9f198 505
80fd2569
MD
506A nameless string type can be declared as a field type:
507
508string field_name; /* Use default UTF8 encoding */
5ba9f198 509
3bf79539
MD
5105. Event Packet Header
511
512The event packet header consists of two part: one is mandatory and have a fixed
513layout. The second part, the "event packet context", has its layout described in
514the metadata.
5ba9f198 515
3bf79539
MD
516- Aligned on page size. Fixed size. Fields either aligned or packed (depending
517 on the architecture preference).
518 No padding at the end of the event packet header. Native architecture byte
5ba9f198 519 ordering.
3bf79539
MD
520
521Fixed layout (event packet header):
522
5ba9f198
MD
523- Magic number (CTF magic numbers: 0xC1FC1FC1 and its reverse endianness
524 representation: 0xC11FFCC1) It needs to have a non-symmetric bytewise
525 representation. Used to distinguish between big and little endian traces (this
526 information is determined by knowing the endianness of the architecture
527 reading the trace and comparing the magic number against its value and the
528 reverse, 0xC11FFCC1). This magic number specifies that we use the CTF metadata
529 description language described in this document. Different magic numbers
530 should be used for other metadata description languages.
3bf79539 531- Trace UUID, used to ensure the event packet match the metadata used.
5ba9f198
MD
532 (note: we cannot use a metadata checksum because metadata can be appended to
533 while tracing is active)
3bf79539
MD
534- Stream ID, used as reference to stream description in metadata.
535
536Metadata-defined layout (event packet context):
537
538- Event packet content size (in bytes).
539- Event packet size (in bytes, includes padding).
540- Event packet content checksum (optional). Checksum excludes the event packet
541 header.
542- Per-stream event packet sequence count (to deal with UDP packet loss). The
543 number of significant sequence counter bits should also be present, so
544 wrap-arounds are deal with correctly.
545- Timestamp at the beginning and timestamp at the end of the event packet.
546 Both timestamps are written in the packet header, but sampled respectively
547 while (or before) writing the first event and while (or after) writing the
548 last event in the packet. The inclusive range between these timestamps should
549 include all event timestamps assigned to events contained within the packet.
5ba9f198 550- Events discarded count
3bf79539
MD
551 - Snapshot of a per-stream free-running counter, counting the number of
552 events discarded that were supposed to be written in the stream prior to
553 the first event in the event packet.
5ba9f198 554 * Note: producer-consumer buffer full condition should fill the current
3bf79539 555 event packet with padding so we know exactly where events have been
5ba9f198 556 discarded.
3bf79539
MD
557- Lossless compression scheme used for the event packet content. Applied
558 directly to raw data. New types of compression can be added in following
559 versions of the format.
5ba9f198
MD
560 0: no compression scheme
561 1: bzip2
562 2: gzip
3bf79539
MD
563 3: xz
564- Cypher used for the event packet content. Applied after compression.
5ba9f198
MD
565 0: no encryption
566 1: AES
3bf79539 567- Checksum scheme used for the event packet content. Applied after encryption.
5ba9f198
MD
568 0: no checksum
569 1: md5
570 2: sha1
571 3: crc32
572
3bf79539
MD
5735.1 Event Packet Header Fixed Layout Description
574
80fd2569
MD
575struct event_packet_header {
576 uint32_t magic;
577 uint8_t trace_uuid[16];
3bf79539 578 uint32_t stream_id;
80fd2569 579};
5ba9f198 580
3bf79539
MD
5815.2 Event Packet Context Description
582
583Event packet context example. These are declared within the stream declaration
584in the metadata. All these fields are optional except for "content_size" and
585"packet_size", which must be present in the context.
586
587An example event packet context type:
588
80fd2569 589struct event_packet_context {
3bf79539
MD
590 uint64_t timestamp_begin;
591 uint64_t timestamp_end;
592 uint32_t checksum;
593 uint32_t stream_packet_count;
594 uint32_t events_discarded;
595 uint32_t cpu_id;
596 uint32_t/uint16_t content_size;
597 uint32_t/uint16_t packet_size;
598 uint8_t stream_packet_count_bits; /* Significant counter bits */
599 uint8_t compression_scheme;
600 uint8_t encryption_scheme;
3b0f8e4d 601 uint8_t checksum_scheme;
3bf79539 602};
5ba9f198 603
fcba70d4 604
5ba9f198
MD
6056. Event Structure
606
607The overall structure of an event is:
608
fcba70d4 6091 - Stream Packet Context (as specified by the stream metadata)
fdf2bb05
MD
610 2 - Event Header (as specified by the stream metadata)
611 3 - Stream Event Context (as specified by the stream metadata)
612 4 - Event Context (as specified by the event metadata)
613 5 - Event Payload (as specified by the event metadata)
5ba9f198 614
fdf2bb05 615This structure defines an implicit dynamic scoping, where variants
7d9d7e92
MD
616located in inner structures (those with a higher number in the listing
617above) can refer to the fields of outer structures (with lower number in
618the listing above). See Section 7.2 Metadata Scopes for more detail.
5ba9f198 619
fdf2bb05 6206.1 Event Header
fcba70d4
MD
621
622Event headers can be described within the metadata. We hereby propose, as an
623example, two types of events headers. Type 1 accommodates streams with less than
62431 event IDs. Type 2 accommodates streams with 31 or more event IDs.
5ba9f198 625
3bf79539
MD
626One major factor can vary between streams: the number of event IDs assigned to
627a stream. Luckily, this information tends to stay relatively constant (modulo
5ba9f198 628event registration while trace is being recorded), so we can specify different
3bf79539 629representations for streams containing few event IDs and streams containing
5ba9f198
MD
630many event IDs, so we end up representing the event ID and timestamp as densely
631as possible in each case.
632
fcba70d4
MD
633The header is extended in the rare occasions where the information cannot be
634represented in the ranges available in the standard event header. They are also
3bf79539
MD
635used in the rare occasions where the data required for a field could not be
636collected: the flag corresponding to the missing field within the missing_fields
637array is then set to 1.
5ba9f198
MD
638
639Types uintX_t represent an X-bit unsigned integer.
640
641
fdf2bb05 6426.1.1 Type 1 - Few event IDs
5ba9f198
MD
643
644 - Aligned on 32-bit (or 8-bit if byte-packed, depending on the architecture
645 preference).
5ba9f198 646 - Native architecture byte ordering.
fcba70d4
MD
647 - For "compact" selection
648 - Fixed size: 32 bits.
649 - For "extended" selection
650 - Size depends on the architecture and variant alignment.
5ba9f198 651
80fd2569 652struct event_header_1 {
fcba70d4
MD
653 /*
654 * id: range: 0 - 30.
655 * id 31 is reserved to indicate an extended header.
656 */
657 enum <uint5_t> { compact = 0 ... 30, extended = 31 } id;
658 variant <id> {
659 struct {
660 uint27_t timestamp;
661 } compact;
662 struct {
663 uint32_t id; /* 32-bit event IDs */
664 uint64_t timestamp; /* 64-bit timestamps */
665 } extended;
666 } v;
5ba9f198
MD
667};
668
5ba9f198 669
fdf2bb05 6706.1.2 Type 2 - Many event IDs
5ba9f198 671
fcba70d4 672 - Aligned on 16-bit (or 8-bit if byte-packed, depending on the architecture
5ba9f198 673 preference).
5ba9f198 674 - Native architecture byte ordering.
fcba70d4
MD
675 - For "compact" selection
676 - Size depends on the architecture and variant alignment.
677 - For "extended" selection
678 - Size depends on the architecture and variant alignment.
5ba9f198 679
80fd2569 680struct event_header_2 {
fcba70d4
MD
681 /*
682 * id: range: 0 - 65534.
683 * id 65535 is reserved to indicate an extended header.
684 */
685 enum <uint16_t> { compact = 0 ... 65534, extended = 65535 } id;
686 variant <id> {
687 struct {
688 uint32_t timestamp;
689 } compact;
690 struct {
691 uint32_t id; /* 32-bit event IDs */
692 uint64_t timestamp; /* 64-bit timestamps */
693 } extended;
694 } v;
5ba9f198
MD
695};
696
5ba9f198
MD
697
6986.2 Event Context
699
700The event context contains information relative to the current event. The choice
fcba70d4
MD
701and meaning of this information is specified by the metadata "stream" and
702"event" information. The "stream" context is applied to all events within the
703stream. The "stream" context structure follows the event header. The "event"
704context is applied to specific events. Its structure follows the "stream"
705context stucture.
5ba9f198 706
fcba70d4
MD
707An example of stream-level event context is to save the event payload size with
708each event, or to save the current PID with each event. These are declared
709within the stream declaration within the metadata:
5ba9f198 710
fcba70d4
MD
711 stream {
712 ...
713 event {
714 ...
4fa992a5 715 context := struct {
80fd2569
MD
716 uint pid;
717 uint16_t payload_size;
3bf79539 718 };
fcba70d4
MD
719 }
720 };
721
722An example of event-specific event context is to declare a bitmap of missing
723fields, only appended after the stream event context if the extended event
724header is selected. NR_FIELDS is the number of fields within the event (a
725numeric value).
5ba9f198 726
fcba70d4
MD
727 event {
728 context = struct {
729 variant <id> {
730 struct { } compact;
731 struct {
732 uint1_t missing_fields[NR_FIELDS]; /* missing event fields bitmap */
733 } extended;
734 } v;
735 };
736 ...
737 }
5ba9f198
MD
738
7396.3 Event Payload
740
741An event payload contains fields specific to a given event type. The fields
742belonging to an event type are described in the event-specific metadata
743within a structure type.
744
7456.3.1 Padding
746
747No padding at the end of the event payload. This differs from the ISO/C standard
748for structures, but follows the CTF standard for structures. In a trace, even
749though it makes sense to align the beginning of a structure, it really makes no
750sense to add padding at the end of the structure, because structures are usually
751not followed by a structure of the same type.
752
753This trick can be done by adding a zero-length "end" field at the end of the C
754structures, and by using the offset of this field rather than using sizeof()
3bf79539 755when calculating the size of a structure (see Appendix "A. Helper macros").
5ba9f198
MD
756
7576.3.2 Alignment
758
759The event payload is aligned on the largest alignment required by types
760contained within the payload. (This follows the ISO/C standard for structures)
761
762
5ba9f198
MD
7637. Metadata
764
4fafe1ad
MD
765The meta-data is located in a stream identified by its name: "metadata".
766It is made of "event packets", which each start with an event packet
767header. The event type within the metadata stream have no event header
768nor event context. Each event only contains a null-terminated "string"
769payload, which is a metadata description entry. The events are packed
770one next to another. Each event packet start with an event packet
771header, which contains, amongst other fields, the magic number and trace
772UUID. In the event packet header, the trace UUID is represented as an
773array of bytes. Within the string-based metadata description, the trace
774UUID is represented as a string of hexadecimal digits and dashes "-".
775
776The metadata can be parsed by reading through the metadata strings,
777skipping null-characters. Type names are made of a single identifier,
778and can be surrounded by prefix/postfix. Text contained within "/*" and
779"*/", as well as within "//" and end of line, are treated as comments.
780Boolean values can be represented as true, TRUE, or 1 for true, and
781false, FALSE, or 0 for false.
fcba70d4 782
fdf2bb05
MD
783
7847.1 Declaration vs Definition
785
786A declaration associates a layout to a type, without specifying where
787this type is located in the event structure hierarchy (see Section 6).
788This therefore includes typedef, typealias, as well as all type
789specifiers. In certain circumstances (typedef, structure field and
790variant field), a declaration is followed by a declarator, which specify
791the newly defined type name (for typedef), or the field name (for
792declarations located within structure and variants). Array and sequence,
793declared with square brackets ("[" "]"), are part of the declarator,
457d8b0a
MD
794similarly to C99. The enumeration type specifier and variant tag name
795(both specified with "<" ">") are part of the type specifier.
fdf2bb05
MD
796
797A definition associates a type to a location in the event structure
b9606a77
MD
798hierarchy (see Section 6). This association is denoted by ":=", as shown
799in Section 7.3.
fdf2bb05
MD
800
801
8027.2 Metadata Scopes
803
804CTF metadata uses two different types of scoping: a lexical scope is
805used for declarations and type definitions, and a dynamic scope is used
806for variants references to tag fields.
807
8087.2.1 Lexical Scope
809
d285084f
MD
810Each of "trace", "stream", "event", "struct" and "variant" have their own
811nestable declaration scope, within which types can be declared using "typedef"
fdf2bb05 812and "typealias". A root declaration scope also contains all declarations
7d9d7e92 813located outside of any of the aforementioned declarations. An inner
fdf2bb05 814declaration scope can refer to type declared within its container
7d9d7e92
MD
815lexical scope prior to the inner declaration scope. Redefinition of a
816typedef or typealias is not valid, although hiding an upper scope
fdf2bb05
MD
817typedef or typealias is allowed within a sub-scope.
818
8197.2.2 Dynamic Scope
820
7d9d7e92
MD
821A dynamic scope consists in the lexical scope augmented with the
822implicit event structure definition hierarchy presented at Section 6.
823The dynamic scope is only used for variant tag definitions. It is used
824at definition time to look up the location of the tag field associated
825with a variant.
826
827Therefore, variants in lower levels in the dynamic scope (e.g. event
828context) can refer to a tag field located in upper levels (e.g. in the
829event header) by specifying, in this case, the associated tag with
830<header.field_name>. This allows, for instance, the event context to
831define a variant referring to the "id" field of the event header as
832selector.
fdf2bb05
MD
833
834The target dynamic scope must be specified explicitly when referring to
835a field outside of the local static scope. The dynamic scope prefixes
836are thus:
837
7d9d7e92
MD
838 - Stream Packet Context: <stream.packet.context. >,
839 - Event Header: <stream.event.header. >,
840 - Stream Event Context: <stream.event.context. >,
841 - Event Context: <event.context. >,
842 - Event Payload: <event.fields. >.
fdf2bb05
MD
843
844Multiple declarations of the same field name within a single scope is
845not valid. It is however valid to re-use the same field name in
846different scopes. There is no possible conflict, because the dynamic
847scope must be specified when a variant refers to a tag field located in
848a different dynamic scope.
849
457d8b0a
MD
850The information available in the dynamic scopes can be thought of as the
851current tracing context. At trace production, information about the
852current context is saved into the specified scope field levels. At trace
853consumption, for each event, the current trace context is therefore
854readable by accessing the upper dynamic scopes.
855
fdf2bb05 856
b9606a77 8577.3 Metadata Examples
d285084f 858
fcba70d4 859The grammar representing the CTF metadata is presented in
fdf2bb05
MD
860Appendix C. CTF Metadata Grammar. This section presents a rather ligher
861reading that consists in examples of CTF metadata, with template values:
5ba9f198
MD
862
863trace {
fdf2bb05 864 major = value; /* Trace format version */
5ba9f198 865 minor = value;
fdf2bb05 866 uuid = "aaaaaaaa-aaaa-aaaa-aaaa-aaaaaaaaaaaa"; /* Trace UUID */
3bf79539
MD
867 word_size = value;
868};
5ba9f198 869
3bf79539
MD
870stream {
871 id = stream_id;
fdf2bb05 872 /* Type 1 - Few event IDs; Type 2 - Many event IDs. See section 6.1. */
4fa992a5
MD
873 event.header := event_header_1 OR event_header_2;
874 event.context := struct {
77a98c82 875 ...
3bf79539 876 };
4fa992a5 877 packet.context := struct {
77a98c82 878 ...
3bf79539
MD
879 };
880};
5ba9f198
MD
881
882event {
3d13ef1a 883 name = event_name;
3bf79539
MD
884 id = value; /* Numeric identifier within the stream */
885 stream = stream_id;
4fa992a5 886 context := struct {
fcba70d4
MD
887 ...
888 };
4fa992a5 889 fields := struct {
80fd2569
MD
890 ...
891 };
3bf79539 892};
5ba9f198
MD
893
894/* More detail on types in section 4. Types */
895
3d13ef1a
MD
896/*
897 * Named types:
898 *
4fa992a5 899 * Type declarations behave similarly to the C standard.
3d13ef1a
MD
900 */
901
902typedef aliased_type_prefix aliased_type new_type aliased_type_postfix;
2152348f 903
3d13ef1a 904/* e.g.: typedef struct example new_type_name[10]; */
80fd2569 905
4fa992a5
MD
906/*
907 * typealias
908 *
909 * The "typealias" declaration can be used to give a name (including
359894ac
MD
910 * prefix/postfix) to a type. It should also be used to map basic C types
911 * (float, int, unsigned long, ...) to a CTF type. Typealias is a superset of
912 * "typedef": it also allows assignment of a simple variable identifier to a
913 * type.
4fa992a5
MD
914 */
915
916typealias type_class {
80fd2569 917 ...
fcba70d4 918} : new_type_prefix new_type new_type_postfix;
2152348f 919
3d13ef1a
MD
920/*
921 * e.g.:
4fa992a5 922 * typealias integer {
3d13ef1a
MD
923 * size = 32;
924 * align = 32;
925 * signed = false;
fcba70d4 926 * } : struct page *;
359894ac
MD
927 *
928 * typealias integer {
929 * size = 32;
930 * align = 32;
931 * signed = true;
932 * } : int;
3d13ef1a 933 */
80fd2569
MD
934
935struct name {
3bf79539
MD
936 ...
937};
5ba9f198 938
fcba70d4
MD
939variant name {
940 ...
941};
942
cfc73fdc 943enum name <integer_type or size> {
3bf79539
MD
944 ...
945};
946
2152348f 947
4fa992a5
MD
948/*
949 * Unnamed types, contained within compound type fields, typedef or typealias.
950 */
2152348f 951
80fd2569
MD
952struct {
953 ...
2152348f 954}
5ba9f198 955
fcba70d4
MD
956variant {
957 ...
958}
959
4767a9e7 960enum <integer_type or size> {
80fd2569 961 ...
2152348f
MD
962}
963
964typedef type new_type[length];
3bf79539 965
2152348f
MD
966struct {
967 type field_name[length];
968}
969
970typedef type new_type[length_type];
971
972struct {
973 type field_name[length_type];
974}
975
976integer {
80fd2569 977 ...
2152348f 978}
3bf79539 979
2152348f 980floating_point {
80fd2569 981 ...
2152348f
MD
982}
983
984struct {
985 integer_type field_name:size; /* GNU/C bitfield */
986}
987
988struct {
989 string field_name;
990}
3bf79539 991
fcba70d4 992
3bf79539 993A. Helper macros
5ba9f198
MD
994
995The two following macros keep track of the size of a GNU/C structure without
996padding at the end by placing HEADER_END as the last field. A one byte end field
997is used for C90 compatibility (C99 flexible arrays could be used here). Note
998that this does not affect the effective structure size, which should always be
999calculated with the header_sizeof() helper.
1000
1001#define HEADER_END char end_field
1002#define header_sizeof(type) offsetof(typeof(type), end_field)
3bf79539
MD
1003
1004
1005B. Stream Header Rationale
1006
1007An event stream is divided in contiguous event packets of variable size. These
1008subdivisions allow the trace analyzer to perform a fast binary search by time
1009within the stream (typically requiring to index only the event packet headers)
1010without reading the whole stream. These subdivisions have a variable size to
1011eliminate the need to transfer the event packet padding when partially filled
1012event packets must be sent when streaming a trace for live viewing/analysis.
1013An event packet can contain a certain amount of padding at the end. Dividing
1014streams into event packets is also useful for network streaming over UDP and
1015flight recorder mode tracing (a whole event packet can be swapped out of the
1016buffer atomically for reading).
1017
1018The stream header is repeated at the beginning of each event packet to allow
1019flexibility in terms of:
1020
1021 - streaming support,
1022 - allowing arbitrary buffers to be discarded without making the trace
1023 unreadable,
1024 - allow UDP packet loss handling by either dealing with missing event packet
1025 or asking for re-transmission.
1026 - transparently support flight recorder mode,
1027 - transparently support crash dump.
1028
1029The event stream header will therefore be referred to as the "event packet
1030header" throughout the rest of this document.
fcba70d4
MD
1031
1032C. CTF Metadata Grammar
1033
4fa992a5
MD
1034/*
1035 * Common Trace Format (CTF) Metadata Grammar.
1036 *
1037 * Inspired from the C99 grammar:
1038 * http://www.open-std.org/jtc1/sc22/wg14/www/docs/n1124.pdf (Annex A)
1039 *
1040 * Specialized for CTF needs by including only constant and declarations from
1041 * C99 (excluding function declarations), and by adding support for variants,
1042 * sequences and CTF-specific specifiers.
1043 */
1044
10451) Lexical grammar
1046
10471.1) Lexical elements
1048
1049token:
1050 keyword
1051 identifier
1052 constant
1053 string-literal
1054 punctuator
1055
10561.2) Keywords
1057
1058keyword: is one of
1059
1060const
1061char
1062double
1063enum
1064event
1065floating_point
1066float
1067integer
1068int
1069long
1070short
1071signed
1072stream
1073string
1074struct
1075trace
3e1e1a78 1076typealias
4fa992a5
MD
1077typedef
1078unsigned
1079variant
1080void
1081_Bool
1082_Complex
1083_Imaginary
1084
1085
10861.3) Identifiers
1087
1088identifier:
1089 identifier-nondigit
1090 identifier identifier-nondigit
1091 identifier digit
1092
1093identifier-nondigit:
1094 nondigit
1095 universal-character-name
1096 any other implementation-defined characters
1097
1098nondigit:
1099 _
1100 [a-zA-Z] /* regular expression */
1101
1102digit:
1103 [0-9] /* regular expression */
1104
11051.4) Universal character names
1106
1107universal-character-name:
1108 \u hex-quad
1109 \U hex-quad hex-quad
1110
1111hex-quad:
1112 hexadecimal-digit hexadecimal-digit hexadecimal-digit hexadecimal-digit
1113
11141.5) Constants
1115
1116constant:
1117 integer-constant
1118 enumeration-constant
1119 character-constant
1120
1121integer-constant:
1122 decimal-constant integer-suffix-opt
1123 octal-constant integer-suffix-opt
1124 hexadecimal-constant integer-suffix-opt
1125
1126decimal-constant:
1127 nonzero-digit
1128 decimal-constant digit
1129
1130octal-constant:
1131 0
1132 octal-constant octal-digit
1133
1134hexadecimal-constant:
1135 hexadecimal-prefix hexadecimal-digit
1136 hexadecimal-constant hexadecimal-digit
1137
1138hexadecimal-prefix:
1139 0x
1140 0X
1141
1142nonzero-digit:
1143 [1-9]
1144
1145integer-suffix:
1146 unsigned-suffix long-suffix-opt
1147 unsigned-suffix long-long-suffix
1148 long-suffix unsigned-suffix-opt
1149 long-long-suffix unsigned-suffix-opt
1150
1151unsigned-suffix:
1152 u
1153 U
1154
1155long-suffix:
1156 l
1157 L
1158
1159long-long-suffix:
1160 ll
1161 LL
1162
1163digit-sequence:
1164 digit
1165 digit-sequence digit
1166
1167hexadecimal-digit-sequence:
1168 hexadecimal-digit
1169 hexadecimal-digit-sequence hexadecimal-digit
1170
1171enumeration-constant:
1172 identifier
1173 string-literal
1174
1175character-constant:
1176 ' c-char-sequence '
1177 L' c-char-sequence '
1178
1179c-char-sequence:
1180 c-char
1181 c-char-sequence c-char
1182
1183c-char:
1184 any member of source charset except single-quote ('), backslash
1185 (\), or new-line character.
1186 escape-sequence
1187
1188escape-sequence:
1189 simple-escape-sequence
1190 octal-escape-sequence
1191 hexadecimal-escape-sequence
1192 universal-character-name
1193
1194simple-escape-sequence: one of
1195 \' \" \? \\ \a \b \f \n \r \t \v
1196
1197octal-escape-sequence:
1198 \ octal-digit
1199 \ octal-digit octal-digit
1200 \ octal-digit octal-digit octal-digit
1201
1202hexadecimal-escape-sequence:
1203 \x hexadecimal-digit
1204 hexadecimal-escape-sequence hexadecimal-digit
1205
12061.6) String literals
1207
1208string-literal:
1209 " s-char-sequence-opt "
1210 L" s-char-sequence-opt "
1211
1212s-char-sequence:
1213 s-char
1214 s-char-sequence s-char
1215
1216s-char:
1217 any member of source charset except double-quote ("), backslash
1218 (\), or new-line character.
1219 escape-sequence
1220
12211.7) Punctuators
1222
1223punctuator: one of
1224 [ ] ( ) { } . -> * + - < > : ; ... = ,
1225
1226
12272) Phrase structure grammar
1228
1229primary-expression:
1230 identifier
1231 constant
1232 string-literal
1233 ( unary-expression )
1234
1235postfix-expression:
1236 primary-expression
1237 postfix-expression [ unary-expression ]
1238 postfix-expression . identifier
1239 postfix-expressoin -> identifier
1240
1241unary-expression:
1242 postfix-expression
1243 unary-operator postfix-expression
1244
1245unary-operator: one of
1246 + -
1247
4fa992a5
MD
1248assignment-operator:
1249 =
1250
b9606a77
MD
1251type-assignment-operator:
1252 :=
1253
4fa992a5
MD
1254constant-expression:
1255 unary-expression
1256
1257constant-expression-range:
1258 constant-expression ... constant-expression
1259
12602.2) Declarations:
1261
1262declaration:
689e04b4 1263 declaration-specifiers declarator-list-opt ;
4fa992a5
MD
1264 ctf-specifier ;
1265
1266declaration-specifiers:
689e04b4 1267 storage-class-specifier declaration-specifiers-opt
4fa992a5
MD
1268 type-specifier declaration-specifiers-opt
1269 type-qualifier declaration-specifiers-opt
1270
1271declarator-list:
1272 declarator
1273 declarator-list , declarator
1274
d285084f
MD
1275abstract-declarator-list:
1276 abstract-declarator
1277 abstract-declarator-list , abstract-declarator
1278
4fa992a5
MD
1279storage-class-specifier:
1280 typedef
1281
1282type-specifier:
1283 void
1284 char
1285 short
1286 int
1287 long
1288 float
1289 double
1290 signed
1291 unsigned
1292 _Bool
1293 _Complex
cfdd51ec 1294 _Imaginary
9dfcfc0f
MD
1295 struct-specifier
1296 variant-specifier
4fa992a5
MD
1297 enum-specifier
1298 typedef-name
1299 ctf-type-specifier
1300
1301struct-specifier:
3b0f8e4d 1302 struct identifier-opt { struct-or-variant-declaration-list-opt }
4fa992a5
MD
1303 struct identifier
1304
1305struct-or-variant-declaration-list:
1306 struct-or-variant-declaration
1307 struct-or-variant-declaration-list struct-or-variant-declaration
1308
1309struct-or-variant-declaration:
1310 specifier-qualifier-list struct-or-variant-declarator-list ;
550aca33 1311 declaration-specifiers storage-class-specifier declaration-specifiers declarator-list ;
d285084f
MD
1312 typealias declaration-specifiers abstract-declarator-list : declaration-specifiers abstract-declarator-list ;
1313 typealias declaration-specifiers abstract-declarator-list : declarator-list ;
4fa992a5
MD
1314
1315specifier-qualifier-list:
1316 type-specifier specifier-qualifier-list-opt
1317 type-qualifier specifier-qualifier-list-opt
1318
1319struct-or-variant-declarator-list:
1320 struct-or-variant-declarator
1321 struct-or-variant-declarator-list , struct-or-variant-declarator
1322
1323struct-or-variant-declarator:
1324 declarator
1325 declarator-opt : constant-expression
1326
1327variant-specifier:
1328 variant identifier-opt variant-tag-opt { struct-or-variant-declaration-list }
1329 variant identifier variant-tag
1330
1331variant-tag:
1332 < identifier >
1333
1334enum-specifier:
1335 enum identifier-opt { enumerator-list }
1336 enum identifier-opt { enumerator-list , }
1337 enum identifier
359894ac
MD
1338 enum identifier-opt < declaration-specifiers > { enumerator-list }
1339 enum identifier-opt < declaration-specifiers > { enumerator-list , }
1340 enum identifier < declaration-specifiers >
4fa992a5
MD
1341 enum identifier-opt < integer-constant > { enumerator-list }
1342 enum identifier-opt < integer-constant > { enumerator-list , }
1343 enum identifier < integer-constant >
1344
1345enumerator-list:
1346 enumerator
1347 enumerator-list , enumerator
1348
1349enumerator:
1350 enumeration-constant
1351 enumeration-constant = constant-expression
1352 enumeration-constant = constant-expression-range
1353
1354type-qualifier:
1355 const
1356
1357declarator:
1358 pointer-opt direct-declarator
1359
1360direct-declarator:
1361 identifier
1362 ( declarator )
1363 direct-declarator [ type-specifier ]
1364 direct-declarator [ constant-expression ]
1365
d285084f
MD
1366abstract-declarator:
1367 pointer-opt direct-abstract-declarator
1368
1369direct-abstract-declarator:
1370 identifier-opt
1371 ( abstract-declarator )
1372 direct-abstract-declarator [ type-specifier ]
1373 direct-abstract-declarator [ constant-expression ]
1374 direct-abstract-declarator [ ]
1375
4fa992a5 1376pointer:
3b0f8e4d
MD
1377 * type-qualifier-list-opt
1378 * type-qualifier-list-opt pointer
4fa992a5
MD
1379
1380type-qualifier-list:
1381 type-qualifier
1382 type-qualifier-list type-qualifier
1383
4fa992a5
MD
1384typedef-name:
1385 identifier
1386
13872.3) CTF-specific declarations
1388
1389ctf-specifier:
1390 event { ctf-assignment-expression-list-opt }
1391 stream { ctf-assignment-expression-list-opt }
1392 trace { ctf-assignment-expression-list-opt }
d285084f
MD
1393 typealias declaration-specifiers abstract-declarator-list : declaration-specifiers abstract-declarator-list ;
1394 typealias declaration-specifiers abstract-declarator-list : declarator-list ;
4fa992a5
MD
1395
1396ctf-type-specifier:
1397 floating_point { ctf-assignment-expression-list-opt }
1398 integer { ctf-assignment-expression-list-opt }
1399 string { ctf-assignment-expression-list-opt }
1400
1401ctf-assignment-expression-list:
1402 ctf-assignment-expression
1403 ctf-assignment-expression-list ; ctf-assignment-expression
1404
1405ctf-assignment-expression:
1406 unary-expression assignment-operator unary-expression
1407 unary-expression type-assignment-operator type-specifier
550aca33 1408 declaration-specifiers storage-class-specifier declaration-specifiers declarator-list
d285084f
MD
1409 typealias declaration-specifiers abstract-declarator-list : declaration-specifiers abstract-declarator-list
1410 typealias declaration-specifiers abstract-declarator-list : declarator-list
This page took 0.146131 seconds and 4 git commands to generate.