Add base integer attribute for pretty-print
[ctf.git] / common-trace-format-proposal.txt
1
2 RFC: Common Trace Format (CTF) Proposal (pre-v1.7)
3
4 Mathieu Desnoyers, EfficiOS Inc.
5
6 The goal of the present document is to propose a trace format that suits the
7 needs of the embedded, telecom, high-performance and kernel communities. It is
8 based on the Common Trace Format Requirements (v1.4) document. It is designed to
9 allow traces to be natively generated by the Linux kernel, Linux user-space
10 applications written in C/C++, and hardware components. One major element of
11 CTF is the Trace Stream Description Language (TSDL) which flexibility
12 enables description of various binary trace stream layouts.
13
14 The latest version of this document can be found at:
15
16 git tree: git://git.efficios.com/ctf.git
17 gitweb: http://git.efficios.com/?p=ctf.git
18
19 A reference implementation of a library to read and write this trace format is
20 being implemented within the BabelTrace project, a converter between trace
21 formats. The development tree is available at:
22
23 git tree: git://git.efficios.com/babeltrace.git
24 gitweb: http://git.efficios.com/?p=babeltrace.git
25
26
27 1. Preliminary definitions
28
29 - Event Trace: An ordered sequence of events.
30 - Event Stream: An ordered sequence of events, containing a subset of the
31 trace event types.
32 - Event Packet: A sequence of physically contiguous events within an event
33 stream.
34 - Event: This is the basic entry in a trace. (aka: a trace record).
35 - An event identifier (ID) relates to the class (a type) of event within
36 an event stream.
37 e.g. event: irq_entry.
38 - An event (or event record) relates to a specific instance of an event
39 class.
40 e.g. event: irq_entry, at time X, on CPU Y
41 - Source Architecture: Architecture writing the trace.
42 - Reader Architecture: Architecture reading the trace.
43
44
45 2. High-level representation of a trace
46
47 A trace is divided into multiple event streams. Each event stream contains a
48 subset of the trace event types.
49
50 The final output of the trace, after its generation and optional transport over
51 the network, is expected to be either on permanent or temporary storage in a
52 virtual file system. Because each event stream is appended to while a trace is
53 being recorded, each is associated with a separate file for output. Therefore,
54 a stored trace can be represented as a directory containing one file per stream.
55
56 Meta-data description associated with the trace contains information on
57 trace event types expressed in the Trace Stream Description Language
58 (TSDL). This language describes:
59
60 - Trace version.
61 - Types available.
62 - Per-trace event header description.
63 - Per-stream event header description.
64 - Per-stream event context description.
65 - Per-event
66 - Event type to stream mapping.
67 - Event type to name mapping.
68 - Event type to ID mapping.
69 - Event context description.
70 - Event fields description.
71
72
73 3. Event stream
74
75 An event stream can be divided into contiguous event packets of variable
76 size. These subdivisions have a variable size. An event packet can
77 contain a certain amount of padding at the end. The stream header is
78 repeated at the beginning of each event packet. The rationale for the
79 event stream design choices is explained in Appendix B. Stream Header
80 Rationale.
81
82 The event stream header will therefore be referred to as the "event packet
83 header" throughout the rest of this document.
84
85
86 4. Types
87
88 Types are organized as type classes. Each type class belong to either of two
89 kind of types: basic types or compound types.
90
91 4.1 Basic types
92
93 A basic type is a scalar type, as described in this section. It includes
94 integers, GNU/C bitfields, enumerations, and floating point values.
95
96 4.1.1 Type inheritance
97
98 Type specifications can be inherited to allow deriving types from a
99 type class. For example, see the uint32_t named type derived from the "integer"
100 type class below ("Integers" section). Types have a precise binary
101 representation in the trace. A type class has methods to read and write these
102 types, but must be derived into a type to be usable in an event field.
103
104 4.1.2 Alignment
105
106 We define "byte-packed" types as aligned on the byte size, namely 8-bit.
107 We define "bit-packed" types as following on the next bit, as defined by the
108 "Integers" section.
109
110 Each basic type must specify its alignment, in bits. Examples of
111 possible alignments are: bit-packed (align = 1), byte-packed (align =
112 8), or word-aligned (e.g. align = 32 or align = 64). The choice depends
113 on the architecture preference and compactness vs performance trade-offs
114 of the implementation. Architectures providing fast unaligned write
115 byte-packed basic types to save space, aligning each type on byte
116 boundaries (8-bit). Architectures with slow unaligned writes align types
117 on specific alignment values. If no specific alignment is declared for a
118 type, it is assumed to be bit-packed for integers with size not multiple
119 of 8 bits and for gcc bitfields. All other basic types are byte-packed
120 by default. It is however recommended to always specify the alignment
121 explicitly. Alignment values must be power of two. Compound types are
122 aligned as specified in their individual specification.
123
124 TSDL meta-data attribute representation of a specific alignment:
125
126 align = value; /* value in bits */
127
128 4.1.3 Byte order
129
130 By default, the native endianness of the source architecture the trace is used.
131 Byte order can be overridden for a basic type by specifying a "byte_order"
132 attribute. Typical use-case is to specify the network byte order (big endian:
133 "be") to save data captured from the network into the trace without conversion.
134 If not specified, the byte order is native.
135
136 TSDL meta-data representation:
137
138 byte_order = native OR network OR be OR le; /* network and be are aliases */
139
140 4.1.4 Size
141
142 Type size, in bits, for integers and floats is that returned by "sizeof()" in C
143 multiplied by CHAR_BIT.
144 We require the size of "char" and "unsigned char" types (CHAR_BIT) to be fixed
145 to 8 bits for cross-endianness compatibility.
146
147 TSDL meta-data representation:
148
149 size = value; (value is in bits)
150
151 4.1.5 Integers
152
153 Signed integers are represented in two-complement. Integer alignment,
154 size, signedness and byte ordering are defined in the TSDL meta-data.
155 Integers aligned on byte size (8-bit) and with length multiple of byte
156 size (8-bit) correspond to the C99 standard integers. In addition,
157 integers with alignment and/or size that are _not_ a multiple of the
158 byte size are permitted; these correspond to the C99 standard bitfields,
159 with the added specification that the CTF integer bitfields have a fixed
160 binary representation. A MIT-licensed reference implementation of the
161 CTF portable bitfields is available at:
162
163 http://git.efficios.com/?p=babeltrace.git;a=blob;f=include/babeltrace/bitfield.h
164
165 Binary representation of integers:
166
167 - On little and big endian:
168 - Within a byte, high bits correspond to an integer high bits, and low bits
169 correspond to low bits.
170 - On little endian:
171 - Integer across multiple bytes are placed from the less significant to the
172 most significant.
173 - Consecutive integers are placed from lower bits to higher bits (even within
174 a byte).
175 - On big endian:
176 - Integer across multiple bytes are placed from the most significant to the
177 less significant.
178 - Consecutive integers are placed from higher bits to lower bits (even within
179 a byte).
180
181 This binary representation is derived from the bitfield implementation in GCC
182 for little and big endian. However, contrary to what GCC does, integers can
183 cross units boundaries (no padding is required). Padding can be explicitly
184 added (see 4.1.6 GNU/C bitfields) to follow the GCC layout if needed.
185
186 TSDL meta-data representation:
187
188 integer {
189 signed = true OR false; /* default false */
190 byte_order = native OR network OR be OR le; /* default native */
191 size = value; /* value in bits, no default */
192 align = value; /* value in bits */
193 /* based used for pretty-printing output, default: decimal. */
194 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
195 OR octal OR oct OR o OR 8 OR binary OR b OR 2;
196 }
197
198 Example of type inheritance (creation of a uint32_t named type):
199
200 typealias integer {
201 size = 32;
202 signed = false;
203 align = 32;
204 } := uint32_t;
205
206 Definition of a named 5-bit signed bitfield:
207
208 typealias integer {
209 size = 5;
210 signed = true;
211 align = 1;
212 } := int5_t;
213
214 4.1.6 GNU/C bitfields
215
216 The GNU/C bitfields follow closely the integer representation, with a
217 particularity on alignment: if a bitfield cannot fit in the current unit, the
218 unit is padded and the bitfield starts at the following unit. The unit size is
219 defined by the size of the type "unit_type".
220
221 TSDL meta-data representation:
222
223 unit_type name:size:
224
225 As an example, the following structure declared in C compiled by GCC:
226
227 struct example {
228 short a:12;
229 short b:5;
230 };
231
232 The example structure is aligned on the largest element (short). The second
233 bitfield would be aligned on the next unit boundary, because it would not fit in
234 the current unit.
235
236 4.1.7 Floating point
237
238 The floating point values byte ordering is defined in the TSDL meta-data.
239
240 Floating point values follow the IEEE 754-2008 standard interchange formats.
241 Description of the floating point values include the exponent and mantissa size
242 in bits. Some requirements are imposed on the floating point values:
243
244 - FLT_RADIX must be 2.
245 - mant_dig is the number of digits represented in the mantissa. It is specified
246 by the ISO C99 standard, section 5.2.4, as FLT_MANT_DIG, DBL_MANT_DIG and
247 LDBL_MANT_DIG as defined by <float.h>.
248 - exp_dig is the number of digits represented in the exponent. Given that
249 mant_dig is one bit more than its actual size in bits (leading 1 is not
250 needed) and also given that the sign bit always takes one bit, exp_dig can be
251 specified as:
252
253 - sizeof(float) * CHAR_BIT - FLT_MANT_DIG
254 - sizeof(double) * CHAR_BIT - DBL_MANT_DIG
255 - sizeof(long double) * CHAR_BIT - LDBL_MANT_DIG
256
257 TSDL meta-data representation:
258
259 floating_point {
260 exp_dig = value;
261 mant_dig = value;
262 byte_order = native OR network OR be OR le;
263 align = value;
264 }
265
266 Example of type inheritance:
267
268 typealias floating_point {
269 exp_dig = 8; /* sizeof(float) * CHAR_BIT - FLT_MANT_DIG */
270 mant_dig = 24; /* FLT_MANT_DIG */
271 byte_order = native;
272 align = 32;
273 } := float;
274
275 TODO: define NaN, +inf, -inf behavior.
276
277 Bit-packed, byte-packed or larger alignments can be used for floating
278 point values, similarly to integers.
279
280 4.1.8 Enumerations
281
282 Enumerations are a mapping between an integer type and a table of strings. The
283 numerical representation of the enumeration follows the integer type specified
284 by the meta-data. The enumeration mapping table is detailed in the enumeration
285 description within the meta-data. The mapping table maps inclusive value
286 ranges (or single values) to strings. Instead of being limited to simple
287 "value -> string" mappings, these enumerations map
288 "[ start_value ... end_value ] -> string", which map inclusive ranges of
289 values to strings. An enumeration from the C language can be represented in
290 this format by having the same start_value and end_value for each element, which
291 is in fact a range of size 1. This single-value range is supported without
292 repeating the start and end values with the value = string declaration.
293
294 enum name : integer_type {
295 somestring = start_value1 ... end_value1,
296 "other string" = start_value2 ... end_value2,
297 yet_another_string, /* will be assigned to end_value2 + 1 */
298 "some other string" = value,
299 ...
300 };
301
302 If the values are omitted, the enumeration starts at 0 and increment of 1 for
303 each entry:
304
305 enum name : unsigned int {
306 ZERO,
307 ONE,
308 TWO,
309 TEN = 10,
310 ELEVEN,
311 };
312
313 Overlapping ranges within a single enumeration are implementation defined.
314
315 A nameless enumeration can be declared as a field type or as part of a typedef:
316
317 enum : integer_type {
318 ...
319 }
320
321 Enumerations omitting the container type ": integer_type" use the "int"
322 type (for compatibility with C99). The "int" type must be previously
323 declared. E.g.:
324
325 typealias integer { size = 32; align = 32; signed = true } := int;
326
327 enum {
328 ...
329 }
330
331
332 4.2 Compound types
333
334 Compound are aggregation of type declarations. Compound types include
335 structures, variant, arrays, sequences, and strings.
336
337 4.2.1 Structures
338
339 Structures are aligned on the largest alignment required by basic types
340 contained within the structure. (This follows the ISO/C standard for structures)
341
342 TSDL meta-data representation of a named structure:
343
344 struct name {
345 field_type field_name;
346 field_type field_name;
347 ...
348 };
349
350 Example:
351
352 struct example {
353 integer { /* Nameless type */
354 size = 16;
355 signed = true;
356 align = 16;
357 } first_field_name;
358 uint64_t second_field_name; /* Named type declared in the meta-data */
359 };
360
361 The fields are placed in a sequence next to each other. They each possess a
362 field name, which is a unique identifier within the structure.
363
364 A nameless structure can be declared as a field type or as part of a typedef:
365
366 struct {
367 ...
368 }
369
370 Alignment for a structure compound type can be forced to a minimum value
371 by adding an "align" specifier after the declaration of a structure
372 body. This attribute is read as: align(value). The value is specified in
373 bits. The structure will be aligned on the maximum value between this
374 attribute and the alignment required by the basic types contained within
375 the structure. e.g.
376
377 struct {
378 ...
379 } align(32)
380
381 4.2.2 Variants (Discriminated/Tagged Unions)
382
383 A CTF variant is a selection between different types. A CTF variant must
384 always be defined within the scope of a structure or within fields
385 contained within a structure (defined recursively). A "tag" enumeration
386 field must appear in either the same lexical scope, prior to the variant
387 field (in field declaration order), in an uppermost lexical scope (see
388 Section 7.3.1), or in an uppermost dynamic scope (see Section 7.3.2).
389 The type selection is indicated by the mapping from the enumeration
390 value to the string used as variant type selector. The field to use as
391 tag is specified by the "tag_field", specified between "< >" after the
392 "variant" keyword for unnamed variants, and after "variant name" for
393 named variants.
394
395 The alignment of the variant is the alignment of the type as selected by the tag
396 value for the specific instance of the variant. The alignment of the type
397 containing the variant is independent of the variant alignment. The size of the
398 variant is the size as selected by the tag value for the specific instance of
399 the variant.
400
401 A named variant declaration followed by its definition within a structure
402 declaration:
403
404 variant name {
405 field_type sel1;
406 field_type sel2;
407 field_type sel3;
408 ...
409 };
410
411 struct {
412 enum : integer_type { sel1, sel2, sel3, ... } tag_field;
413 ...
414 variant name <tag_field> v;
415 }
416
417 An unnamed variant definition within a structure is expressed by the following
418 TSDL meta-data:
419
420 struct {
421 enum : integer_type { sel1, sel2, sel3, ... } tag_field;
422 ...
423 variant <tag_field> {
424 field_type sel1;
425 field_type sel2;
426 field_type sel3;
427 ...
428 } v;
429 }
430
431 Example of a named variant within a sequence that refers to a single tag field:
432
433 variant example {
434 uint32_t a;
435 uint64_t b;
436 short c;
437 };
438
439 struct {
440 enum : uint2_t { a, b, c } choice;
441 variant example <choice> v[unsigned int];
442 }
443
444 Example of an unnamed variant:
445
446 struct {
447 enum : uint2_t { a, b, c, d } choice;
448 /* Unrelated fields can be added between the variant and its tag */
449 int32_t somevalue;
450 variant <choice> {
451 uint32_t a;
452 uint64_t b;
453 short c;
454 struct {
455 unsigned int field1;
456 uint64_t field2;
457 } d;
458 } s;
459 }
460
461 Example of an unnamed variant within an array:
462
463 struct {
464 enum : uint2_t { a, b, c } choice;
465 variant <choice> {
466 uint32_t a;
467 uint64_t b;
468 short c;
469 } v[10];
470 }
471
472 Example of a variant type definition within a structure, where the defined type
473 is then declared within an array of structures. This variant refers to a tag
474 located in an upper lexical scope. This example clearly shows that a variant
475 type definition referring to the tag "x" uses the closest preceding field from
476 the lexical scope of the type definition.
477
478 struct {
479 enum : uint2_t { a, b, c, d } x;
480
481 typedef variant <x> { /*
482 * "x" refers to the preceding "x" enumeration in the
483 * lexical scope of the type definition.
484 */
485 uint32_t a;
486 uint64_t b;
487 short c;
488 } example_variant;
489
490 struct {
491 enum : int { x, y, z } x; /* This enumeration is not used by "v". */
492 example_variant v; /*
493 * "v" uses the "enum : uint2_t { a, b, c, d }"
494 * tag.
495 */
496 } a[10];
497 }
498
499 4.2.3 Arrays
500
501 Arrays are fixed-length. Their length is declared in the type
502 declaration within the meta-data. They contain an array of "inner type"
503 elements, which can refer to any type not containing the type of the
504 array being declared (no circular dependency). The length is the number
505 of elements in an array.
506
507 TSDL meta-data representation of a named array:
508
509 typedef elem_type name[length];
510
511 A nameless array can be declared as a field type within a structure, e.g.:
512
513 uint8_t field_name[10];
514
515 Arrays are always aligned on their element alignment requirement.
516
517 4.2.4 Sequences
518
519 Sequences are dynamically-sized arrays. They start with an integer that specify
520 the length of the sequence, followed by an array of "inner type" elements.
521 The length is the number of elements in the sequence.
522
523 TSDL meta-data representation for a named sequence:
524
525 typedef elem_type name[length_type];
526
527 A nameless sequence can be declared as a field type, e.g.:
528
529 long field_name[int];
530
531 The length type follows the integer types specifications, and the sequence
532 elements follow the "array" specifications.
533
534 4.2.5 Strings
535
536 Strings are an array of bytes of variable size and are terminated by a '\0'
537 "NULL" character. Their encoding is described in the TSDL meta-data. In
538 absence of encoding attribute information, the default encoding is
539 UTF-8.
540
541 TSDL meta-data representation of a named string type:
542
543 typealias string {
544 encoding = UTF8 OR ASCII;
545 } := name;
546
547 A nameless string type can be declared as a field type:
548
549 string field_name; /* Use default UTF8 encoding */
550
551 Strings are always aligned on byte size.
552
553 5. Event Packet Header
554
555 The event packet header consists of two parts: the "event packet header"
556 is the same for all streams of a trace. The second part, the "event
557 packet context", is described on a per-stream basis. Both are described
558 in the TSDL meta-data. The packets are aligned on architecture-page-sized
559 addresses.
560
561 Event packet header (all fields are optional, specified by TSDL meta-data):
562
563 - Magic number (CTF magic number: 0xC1FC1FC1) specifies that this is a
564 CTF packet. This magic number is optional, but when present, it should
565 come at the very beginning of the packet.
566 - Trace UUID, used to ensure the event packet match the meta-data used.
567 (note: we cannot use a meta-data checksum in every cases instead of a
568 UUID because meta-data can be appended to while tracing is active)
569 This field is optional.
570 - Stream ID, used as reference to stream description in meta-data.
571 This field is optional if there is only one stream description in the
572 meta-data, but becomes required if there are more than one stream in
573 the TSDL meta-data description.
574
575 Event packet context (all fields are optional, specified by TSDL meta-data):
576
577 - Event packet content size (in bytes).
578 - Event packet size (in bytes, includes padding).
579 - Event packet content checksum (optional). Checksum excludes the event packet
580 header.
581 - Per-stream event packet sequence count (to deal with UDP packet loss). The
582 number of significant sequence counter bits should also be present, so
583 wrap-arounds are dealt with correctly.
584 - Time-stamp at the beginning and time-stamp at the end of the event packet.
585 Both timestamps are written in the packet header, but sampled respectively
586 while (or before) writing the first event and while (or after) writing the
587 last event in the packet. The inclusive range between these timestamps should
588 include all event timestamps assigned to events contained within the packet.
589 - Events discarded count
590 - Snapshot of a per-stream free-running counter, counting the number of
591 events discarded that were supposed to be written in the stream prior to
592 the first event in the event packet.
593 * Note: producer-consumer buffer full condition should fill the current
594 event packet with padding so we know exactly where events have been
595 discarded.
596 - Lossless compression scheme used for the event packet content. Applied
597 directly to raw data. New types of compression can be added in following
598 versions of the format.
599 0: no compression scheme
600 1: bzip2
601 2: gzip
602 3: xz
603 - Cypher used for the event packet content. Applied after compression.
604 0: no encryption
605 1: AES
606 - Checksum scheme used for the event packet content. Applied after encryption.
607 0: no checksum
608 1: md5
609 2: sha1
610 3: crc32
611
612 5.1 Event Packet Header Description
613
614 The event packet header layout is indicated by the trace packet.header
615 field. Here is a recommended structure type for the packet header with
616 the fields typically expected (although these fields are each optional):
617
618 struct event_packet_header {
619 uint32_t magic;
620 uint8_t trace_uuid[16];
621 uint32_t stream_id;
622 };
623
624 trace {
625 ...
626 packet.header := struct event_packet_header;
627 };
628
629 If the magic number is not present, tools such as "file" will have no
630 mean to discover the file type.
631
632 If the trace_uuid is not present, no validation that the meta-data
633 actually corresponds to the stream is performed.
634
635 If the stream_id packet header field is missing, the trace can only
636 contain a single stream. Its "id" field can be left out, and its events
637 don't need to declare a "stream_id" field.
638
639
640 5.2 Event Packet Context Description
641
642 Event packet context example. These are declared within the stream declaration
643 in the meta-data. All these fields are optional. If the packet size field is
644 missing, the whole stream only contains a single packet. If the content
645 size field is missing, the packet is filled (no padding). The content
646 and packet sizes include all headers.
647
648 An example event packet context type:
649
650 struct event_packet_context {
651 uint64_t timestamp_begin;
652 uint64_t timestamp_end;
653 uint32_t checksum;
654 uint32_t stream_packet_count;
655 uint32_t events_discarded;
656 uint32_t cpu_id;
657 uint32_t/uint16_t content_size;
658 uint32_t/uint16_t packet_size;
659 uint8_t stream_packet_count_bits; /* Significant counter bits */
660 uint8_t compression_scheme;
661 uint8_t encryption_scheme;
662 uint8_t checksum_scheme;
663 };
664
665
666 6. Event Structure
667
668 The overall structure of an event is:
669
670 1 - Stream Packet Context (as specified by the stream meta-data)
671 2 - Event Header (as specified by the stream meta-data)
672 3 - Stream Event Context (as specified by the stream meta-data)
673 4 - Event Context (as specified by the event meta-data)
674 5 - Event Payload (as specified by the event meta-data)
675
676 This structure defines an implicit dynamic scoping, where variants
677 located in inner structures (those with a higher number in the listing
678 above) can refer to the fields of outer structures (with lower number in
679 the listing above). See Section 7.3 TSDL Scopes for more detail.
680
681 6.1 Event Header
682
683 Event headers can be described within the meta-data. We hereby propose, as an
684 example, two types of events headers. Type 1 accommodates streams with less than
685 31 event IDs. Type 2 accommodates streams with 31 or more event IDs.
686
687 One major factor can vary between streams: the number of event IDs assigned to
688 a stream. Luckily, this information tends to stay relatively constant (modulo
689 event registration while trace is being recorded), so we can specify different
690 representations for streams containing few event IDs and streams containing
691 many event IDs, so we end up representing the event ID and time-stamp as
692 densely as possible in each case.
693
694 The header is extended in the rare occasions where the information cannot be
695 represented in the ranges available in the standard event header. They are also
696 used in the rare occasions where the data required for a field could not be
697 collected: the flag corresponding to the missing field within the missing_fields
698 array is then set to 1.
699
700 Types uintX_t represent an X-bit unsigned integer, as declared with
701 either:
702
703 typealias integer { size = X; align = X; signed = false } := uintX_t;
704
705 or
706
707 typealias integer { size = X; align = 1; signed = false } := uintX_t;
708
709 6.1.1 Type 1 - Few event IDs
710
711 - Aligned on 32-bit (or 8-bit if byte-packed, depending on the architecture
712 preference).
713 - Native architecture byte ordering.
714 - For "compact" selection
715 - Fixed size: 32 bits.
716 - For "extended" selection
717 - Size depends on the architecture and variant alignment.
718
719 struct event_header_1 {
720 /*
721 * id: range: 0 - 30.
722 * id 31 is reserved to indicate an extended header.
723 */
724 enum : uint5_t { compact = 0 ... 30, extended = 31 } id;
725 variant <id> {
726 struct {
727 uint27_t timestamp;
728 } compact;
729 struct {
730 uint32_t id; /* 32-bit event IDs */
731 uint64_t timestamp; /* 64-bit timestamps */
732 } extended;
733 } v;
734 };
735
736
737 6.1.2 Type 2 - Many event IDs
738
739 - Aligned on 16-bit (or 8-bit if byte-packed, depending on the architecture
740 preference).
741 - Native architecture byte ordering.
742 - For "compact" selection
743 - Size depends on the architecture and variant alignment.
744 - For "extended" selection
745 - Size depends on the architecture and variant alignment.
746
747 struct event_header_2 {
748 /*
749 * id: range: 0 - 65534.
750 * id 65535 is reserved to indicate an extended header.
751 */
752 enum : uint16_t { compact = 0 ... 65534, extended = 65535 } id;
753 variant <id> {
754 struct {
755 uint32_t timestamp;
756 } compact;
757 struct {
758 uint32_t id; /* 32-bit event IDs */
759 uint64_t timestamp; /* 64-bit timestamps */
760 } extended;
761 } v;
762 };
763
764
765 6.2 Event Context
766
767 The event context contains information relative to the current event.
768 The choice and meaning of this information is specified by the TSDL
769 stream and event meta-data descriptions. The stream context is applied
770 to all events within the stream. The stream context structure follows
771 the event header. The event context is applied to specific events. Its
772 structure follows the stream context structure.
773
774 An example of stream-level event context is to save the event payload size with
775 each event, or to save the current PID with each event. These are declared
776 within the stream declaration within the meta-data:
777
778 stream {
779 ...
780 event.context := struct {
781 uint pid;
782 uint16_t payload_size;
783 };
784 };
785
786 An example of event-specific event context is to declare a bitmap of missing
787 fields, only appended after the stream event context if the extended event
788 header is selected. NR_FIELDS is the number of fields within the event (a
789 numeric value).
790
791 event {
792 context = struct {
793 variant <id> {
794 struct { } compact;
795 struct {
796 uint1_t missing_fields[NR_FIELDS]; /* missing event fields bitmap */
797 } extended;
798 } v;
799 };
800 ...
801 }
802
803 6.3 Event Payload
804
805 An event payload contains fields specific to a given event type. The fields
806 belonging to an event type are described in the event-specific meta-data
807 within a structure type.
808
809 6.3.1 Padding
810
811 No padding at the end of the event payload. This differs from the ISO/C standard
812 for structures, but follows the CTF standard for structures. In a trace, even
813 though it makes sense to align the beginning of a structure, it really makes no
814 sense to add padding at the end of the structure, because structures are usually
815 not followed by a structure of the same type.
816
817 This trick can be done by adding a zero-length "end" field at the end of the C
818 structures, and by using the offset of this field rather than using sizeof()
819 when calculating the size of a structure (see Appendix "A. Helper macros").
820
821 6.3.2 Alignment
822
823 The event payload is aligned on the largest alignment required by types
824 contained within the payload. (This follows the ISO/C standard for structures)
825
826
827 7. Trace Stream Description Language (TSDL)
828
829 The Trace Stream Description Language (TSDL) allows expression of the
830 binary trace streams layout in a C99-like Domain Specific Language
831 (DSL).
832
833
834 7.1 Meta-data
835
836 The trace stream layout description is located in the trace meta-data.
837 The meta-data is itself located in a stream identified by its name:
838 "metadata".
839
840 The meta-data description can be expressed in two different formats:
841 text-only and packet-based. The text-only description facilitates
842 generation of meta-data and provides a convenient way to enter the
843 meta-data information by hand. The packet-based meta-data provides the
844 CTF stream packet facilities (checksumming, compression, encryption,
845 network-readiness) for meta-data stream generated and transported by a
846 tracer.
847
848 The text-only meta-data file is a plain text TSDL description.
849
850 The packet-based meta-data is made of "meta-data packets", which each
851 start with a meta-data packet header. The packet-based meta-data
852 description is detected by reading the magic number "0x75D11D57" at the
853 beginning of the file. This magic number is also used to detect the
854 endianness of the architecture by trying to read the CTF magic number
855 and its counterpart in reversed endianness. The events within the
856 meta-data stream have no event header nor event context. Each event only
857 contains a "string" payload. Each meta-data packet start with a special
858 packet header, specific to the meta-data stream, which contains,
859 exactly:
860
861 struct metadata_packet_header {
862 uint32_t magic; /* 0x75D11D57 */
863 uint8_t trace_uuid[16]; /* Unique Universal Identifier */
864 uint32_t checksum; /* 0 if unused */
865 uint32_t content_size; /* in bits */
866 uint32_t packet_size; /* in bits */
867 uint8_t compression_scheme; /* 0 if unused */
868 uint8_t encryption_scheme; /* 0 if unused */
869 uint8_t checksum_scheme; /* 0 if unused */
870 };
871
872 The packet-based meta-data can be converted to a text-only meta-data by
873 concatenating all the strings in contains.
874
875 In the textual representation of the meta-data, the text contained
876 within "/*" and "*/", as well as within "//" and end of line, are
877 treated as comments. Boolean values can be represented as true, TRUE,
878 or 1 for true, and false, FALSE, or 0 for false. Within the string-based
879 meta-data description, the trace UUID is represented as a string of
880 hexadecimal digits and dashes "-". In the event packet header, the trace
881 UUID is represented as an array of bytes.
882
883
884 7.2 Declaration vs Definition
885
886 A declaration associates a layout to a type, without specifying where
887 this type is located in the event structure hierarchy (see Section 6).
888 This therefore includes typedef, typealias, as well as all type
889 specifiers. In certain circumstances (typedef, structure field and
890 variant field), a declaration is followed by a declarator, which specify
891 the newly defined type name (for typedef), or the field name (for
892 declarations located within structure and variants). Array and sequence,
893 declared with square brackets ("[" "]"), are part of the declarator,
894 similarly to C99. The enumeration base type is specified by
895 ": enum_base", which is part of the type specifier. The variant tag
896 name, specified between "<" ">", is also part of the type specifier.
897
898 A definition associates a type to a location in the event structure
899 hierarchy (see Section 6). This association is denoted by ":=", as shown
900 in Section 7.3.
901
902
903 7.3 TSDL Scopes
904
905 TSDL uses two different types of scoping: a lexical scope is used for
906 declarations and type definitions, and a dynamic scope is used for
907 variants references to tag fields.
908
909 7.3.1 Lexical Scope
910
911 Each of "trace", "stream", "event", "struct" and "variant" have their own
912 nestable declaration scope, within which types can be declared using "typedef"
913 and "typealias". A root declaration scope also contains all declarations
914 located outside of any of the aforementioned declarations. An inner
915 declaration scope can refer to type declared within its container
916 lexical scope prior to the inner declaration scope. Redefinition of a
917 typedef or typealias is not valid, although hiding an upper scope
918 typedef or typealias is allowed within a sub-scope.
919
920 7.3.2 Dynamic Scope
921
922 A dynamic scope consists in the lexical scope augmented with the
923 implicit event structure definition hierarchy presented at Section 6.
924 The dynamic scope is only used for variant tag definitions. It is used
925 at definition time to look up the location of the tag field associated
926 with a variant.
927
928 Therefore, variants in lower levels in the dynamic scope (e.g. event
929 context) can refer to a tag field located in upper levels (e.g. in the
930 event header) by specifying, in this case, the associated tag with
931 <header.field_name>. This allows, for instance, the event context to
932 define a variant referring to the "id" field of the event header as
933 selector.
934
935 The target dynamic scope must be specified explicitly when referring to
936 a field outside of the local static scope. The dynamic scope prefixes
937 are thus:
938
939 - Trace Packet Header: <trace.packet.header. >,
940 - Stream Packet Context: <stream.packet.context. >,
941 - Event Header: <stream.event.header. >,
942 - Stream Event Context: <stream.event.context. >,
943 - Event Context: <event.context. >,
944 - Event Payload: <event.fields. >.
945
946 Multiple declarations of the same field name within a single scope is
947 not valid. It is however valid to re-use the same field name in
948 different scopes. There is no possible conflict, because the dynamic
949 scope must be specified when a variant refers to a tag field located in
950 a different dynamic scope.
951
952 The information available in the dynamic scopes can be thought of as the
953 current tracing context. At trace production, information about the
954 current context is saved into the specified scope field levels. At trace
955 consumption, for each event, the current trace context is therefore
956 readable by accessing the upper dynamic scopes.
957
958
959 7.4 TSDL Examples
960
961 The grammar representing the TSDL meta-data is presented in Appendix C.
962 TSDL Grammar. This section presents a rather lighter reading that
963 consists in examples of TSDL meta-data, with template values.
964
965 The stream "id" can be left out if there is only one stream in the
966 trace. The event "id" field can be left out if there is only one event
967 in a stream.
968
969 trace {
970 major = value; /* Trace format version */
971 minor = value;
972 uuid = "aaaaaaaa-aaaa-aaaa-aaaa-aaaaaaaaaaaa"; /* Trace UUID */
973 byte_order = be OR le; /* Endianness (required) */
974 packet.header := struct {
975 uint32_t magic;
976 uint8_t trace_uuid[16];
977 uint32_t stream_id;
978 };
979 };
980
981 stream {
982 id = stream_id;
983 /* Type 1 - Few event IDs; Type 2 - Many event IDs. See section 6.1. */
984 event.header := event_header_1 OR event_header_2;
985 event.context := struct {
986 ...
987 };
988 packet.context := struct {
989 ...
990 };
991 };
992
993 event {
994 name = event_name;
995 id = value; /* Numeric identifier within the stream */
996 stream_id = stream_id;
997 context := struct {
998 ...
999 };
1000 fields := struct {
1001 ...
1002 };
1003 };
1004
1005 /* More detail on types in section 4. Types */
1006
1007 /*
1008 * Named types:
1009 *
1010 * Type declarations behave similarly to the C standard.
1011 */
1012
1013 typedef aliased_type_specifiers new_type_declarators;
1014
1015 /* e.g.: typedef struct example new_type_name[10]; */
1016
1017 /*
1018 * typealias
1019 *
1020 * The "typealias" declaration can be used to give a name (including
1021 * pointer declarator specifier) to a type. It should also be used to
1022 * map basic C types (float, int, unsigned long, ...) to a CTF type.
1023 * Typealias is a superset of "typedef": it also allows assignment of a
1024 * simple variable identifier to a type.
1025 */
1026
1027 typealias type_class {
1028 ...
1029 } := type_specifiers type_declarator;
1030
1031 /*
1032 * e.g.:
1033 * typealias integer {
1034 * size = 32;
1035 * align = 32;
1036 * signed = false;
1037 * } := struct page *;
1038 *
1039 * typealias integer {
1040 * size = 32;
1041 * align = 32;
1042 * signed = true;
1043 * } := int;
1044 */
1045
1046 struct name {
1047 ...
1048 };
1049
1050 variant name {
1051 ...
1052 };
1053
1054 enum name : integer_type {
1055 ...
1056 };
1057
1058
1059 /*
1060 * Unnamed types, contained within compound type fields, typedef or typealias.
1061 */
1062
1063 struct {
1064 ...
1065 }
1066
1067 struct {
1068 ...
1069 } align(value)
1070
1071 variant {
1072 ...
1073 }
1074
1075 enum : integer_type {
1076 ...
1077 }
1078
1079 typedef type new_type[length];
1080
1081 struct {
1082 type field_name[length];
1083 }
1084
1085 typedef type new_type[length_type];
1086
1087 struct {
1088 type field_name[length_type];
1089 }
1090
1091 integer {
1092 ...
1093 }
1094
1095 floating_point {
1096 ...
1097 }
1098
1099 struct {
1100 integer_type field_name:size; /* GNU/C bitfield */
1101 }
1102
1103 struct {
1104 string field_name;
1105 }
1106
1107
1108 A. Helper macros
1109
1110 The two following macros keep track of the size of a GNU/C structure without
1111 padding at the end by placing HEADER_END as the last field. A one byte end field
1112 is used for C90 compatibility (C99 flexible arrays could be used here). Note
1113 that this does not affect the effective structure size, which should always be
1114 calculated with the header_sizeof() helper.
1115
1116 #define HEADER_END char end_field
1117 #define header_sizeof(type) offsetof(typeof(type), end_field)
1118
1119
1120 B. Stream Header Rationale
1121
1122 An event stream is divided in contiguous event packets of variable size. These
1123 subdivisions allow the trace analyzer to perform a fast binary search by time
1124 within the stream (typically requiring to index only the event packet headers)
1125 without reading the whole stream. These subdivisions have a variable size to
1126 eliminate the need to transfer the event packet padding when partially filled
1127 event packets must be sent when streaming a trace for live viewing/analysis.
1128 An event packet can contain a certain amount of padding at the end. Dividing
1129 streams into event packets is also useful for network streaming over UDP and
1130 flight recorder mode tracing (a whole event packet can be swapped out of the
1131 buffer atomically for reading).
1132
1133 The stream header is repeated at the beginning of each event packet to allow
1134 flexibility in terms of:
1135
1136 - streaming support,
1137 - allowing arbitrary buffers to be discarded without making the trace
1138 unreadable,
1139 - allow UDP packet loss handling by either dealing with missing event packet
1140 or asking for re-transmission.
1141 - transparently support flight recorder mode,
1142 - transparently support crash dump.
1143
1144
1145 C. TSDL Grammar
1146
1147 /*
1148 * Common Trace Format (CTF) Trace Stream Description Language (TSDL) Grammar.
1149 *
1150 * Inspired from the C99 grammar:
1151 * http://www.open-std.org/jtc1/sc22/wg14/www/docs/n1124.pdf (Annex A)
1152 * and c++1x grammar (draft)
1153 * http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2011/n3291.pdf (Annex A)
1154 *
1155 * Specialized for CTF needs by including only constant and declarations from
1156 * C99 (excluding function declarations), and by adding support for variants,
1157 * sequences and CTF-specific specifiers. Enumeration container types
1158 * semantic is inspired from c++1x enum-base.
1159 */
1160
1161 1) Lexical grammar
1162
1163 1.1) Lexical elements
1164
1165 token:
1166 keyword
1167 identifier
1168 constant
1169 string-literal
1170 punctuator
1171
1172 1.2) Keywords
1173
1174 keyword: is one of
1175
1176 align
1177 const
1178 char
1179 double
1180 enum
1181 event
1182 floating_point
1183 float
1184 integer
1185 int
1186 long
1187 short
1188 signed
1189 stream
1190 string
1191 struct
1192 trace
1193 typealias
1194 typedef
1195 unsigned
1196 variant
1197 void
1198 _Bool
1199 _Complex
1200 _Imaginary
1201
1202
1203 1.3) Identifiers
1204
1205 identifier:
1206 identifier-nondigit
1207 identifier identifier-nondigit
1208 identifier digit
1209
1210 identifier-nondigit:
1211 nondigit
1212 universal-character-name
1213 any other implementation-defined characters
1214
1215 nondigit:
1216 _
1217 [a-zA-Z] /* regular expression */
1218
1219 digit:
1220 [0-9] /* regular expression */
1221
1222 1.4) Universal character names
1223
1224 universal-character-name:
1225 \u hex-quad
1226 \U hex-quad hex-quad
1227
1228 hex-quad:
1229 hexadecimal-digit hexadecimal-digit hexadecimal-digit hexadecimal-digit
1230
1231 1.5) Constants
1232
1233 constant:
1234 integer-constant
1235 enumeration-constant
1236 character-constant
1237
1238 integer-constant:
1239 decimal-constant integer-suffix-opt
1240 octal-constant integer-suffix-opt
1241 hexadecimal-constant integer-suffix-opt
1242
1243 decimal-constant:
1244 nonzero-digit
1245 decimal-constant digit
1246
1247 octal-constant:
1248 0
1249 octal-constant octal-digit
1250
1251 hexadecimal-constant:
1252 hexadecimal-prefix hexadecimal-digit
1253 hexadecimal-constant hexadecimal-digit
1254
1255 hexadecimal-prefix:
1256 0x
1257 0X
1258
1259 nonzero-digit:
1260 [1-9]
1261
1262 integer-suffix:
1263 unsigned-suffix long-suffix-opt
1264 unsigned-suffix long-long-suffix
1265 long-suffix unsigned-suffix-opt
1266 long-long-suffix unsigned-suffix-opt
1267
1268 unsigned-suffix:
1269 u
1270 U
1271
1272 long-suffix:
1273 l
1274 L
1275
1276 long-long-suffix:
1277 ll
1278 LL
1279
1280 enumeration-constant:
1281 identifier
1282 string-literal
1283
1284 character-constant:
1285 ' c-char-sequence '
1286 L' c-char-sequence '
1287
1288 c-char-sequence:
1289 c-char
1290 c-char-sequence c-char
1291
1292 c-char:
1293 any member of source charset except single-quote ('), backslash
1294 (\), or new-line character.
1295 escape-sequence
1296
1297 escape-sequence:
1298 simple-escape-sequence
1299 octal-escape-sequence
1300 hexadecimal-escape-sequence
1301 universal-character-name
1302
1303 simple-escape-sequence: one of
1304 \' \" \? \\ \a \b \f \n \r \t \v
1305
1306 octal-escape-sequence:
1307 \ octal-digit
1308 \ octal-digit octal-digit
1309 \ octal-digit octal-digit octal-digit
1310
1311 hexadecimal-escape-sequence:
1312 \x hexadecimal-digit
1313 hexadecimal-escape-sequence hexadecimal-digit
1314
1315 1.6) String literals
1316
1317 string-literal:
1318 " s-char-sequence-opt "
1319 L" s-char-sequence-opt "
1320
1321 s-char-sequence:
1322 s-char
1323 s-char-sequence s-char
1324
1325 s-char:
1326 any member of source charset except double-quote ("), backslash
1327 (\), or new-line character.
1328 escape-sequence
1329
1330 1.7) Punctuators
1331
1332 punctuator: one of
1333 [ ] ( ) { } . -> * + - < > : ; ... = ,
1334
1335
1336 2) Phrase structure grammar
1337
1338 primary-expression:
1339 identifier
1340 constant
1341 string-literal
1342 ( unary-expression )
1343
1344 postfix-expression:
1345 primary-expression
1346 postfix-expression [ unary-expression ]
1347 postfix-expression . identifier
1348 postfix-expressoin -> identifier
1349
1350 unary-expression:
1351 postfix-expression
1352 unary-operator postfix-expression
1353
1354 unary-operator: one of
1355 + -
1356
1357 assignment-operator:
1358 =
1359
1360 type-assignment-operator:
1361 :=
1362
1363 constant-expression:
1364 unary-expression
1365
1366 constant-expression-range:
1367 constant-expression ... constant-expression
1368
1369 2.2) Declarations:
1370
1371 declaration:
1372 declaration-specifiers declarator-list-opt ;
1373 ctf-specifier ;
1374
1375 declaration-specifiers:
1376 storage-class-specifier declaration-specifiers-opt
1377 type-specifier declaration-specifiers-opt
1378 type-qualifier declaration-specifiers-opt
1379
1380 declarator-list:
1381 declarator
1382 declarator-list , declarator
1383
1384 abstract-declarator-list:
1385 abstract-declarator
1386 abstract-declarator-list , abstract-declarator
1387
1388 storage-class-specifier:
1389 typedef
1390
1391 type-specifier:
1392 void
1393 char
1394 short
1395 int
1396 long
1397 float
1398 double
1399 signed
1400 unsigned
1401 _Bool
1402 _Complex
1403 _Imaginary
1404 struct-specifier
1405 variant-specifier
1406 enum-specifier
1407 typedef-name
1408 ctf-type-specifier
1409
1410 align-attribute:
1411 align ( constant-expression )
1412
1413 struct-specifier:
1414 struct identifier-opt { struct-or-variant-declaration-list-opt } align-attribute-opt
1415 struct identifier align-attribute-opt
1416
1417 struct-or-variant-declaration-list:
1418 struct-or-variant-declaration
1419 struct-or-variant-declaration-list struct-or-variant-declaration
1420
1421 struct-or-variant-declaration:
1422 specifier-qualifier-list struct-or-variant-declarator-list ;
1423 declaration-specifiers storage-class-specifier declaration-specifiers declarator-list ;
1424 typealias declaration-specifiers abstract-declarator-list := declaration-specifiers abstract-declarator-list ;
1425 typealias declaration-specifiers abstract-declarator-list := declarator-list ;
1426
1427 specifier-qualifier-list:
1428 type-specifier specifier-qualifier-list-opt
1429 type-qualifier specifier-qualifier-list-opt
1430
1431 struct-or-variant-declarator-list:
1432 struct-or-variant-declarator
1433 struct-or-variant-declarator-list , struct-or-variant-declarator
1434
1435 struct-or-variant-declarator:
1436 declarator
1437 declarator-opt : constant-expression
1438
1439 variant-specifier:
1440 variant identifier-opt variant-tag-opt { struct-or-variant-declaration-list }
1441 variant identifier variant-tag
1442
1443 variant-tag:
1444 < identifier >
1445
1446 enum-specifier:
1447 enum identifier-opt { enumerator-list }
1448 enum identifier-opt { enumerator-list , }
1449 enum identifier
1450 enum identifier-opt : declaration-specifiers { enumerator-list }
1451 enum identifier-opt : declaration-specifiers { enumerator-list , }
1452
1453 enumerator-list:
1454 enumerator
1455 enumerator-list , enumerator
1456
1457 enumerator:
1458 enumeration-constant
1459 enumeration-constant = constant-expression
1460 enumeration-constant = constant-expression-range
1461
1462 type-qualifier:
1463 const
1464
1465 declarator:
1466 pointer-opt direct-declarator
1467
1468 direct-declarator:
1469 identifier
1470 ( declarator )
1471 direct-declarator [ type-specifier ]
1472 direct-declarator [ constant-expression ]
1473
1474 abstract-declarator:
1475 pointer-opt direct-abstract-declarator
1476
1477 direct-abstract-declarator:
1478 identifier-opt
1479 ( abstract-declarator )
1480 direct-abstract-declarator [ type-specifier ]
1481 direct-abstract-declarator [ constant-expression ]
1482 direct-abstract-declarator [ ]
1483
1484 pointer:
1485 * type-qualifier-list-opt
1486 * type-qualifier-list-opt pointer
1487
1488 type-qualifier-list:
1489 type-qualifier
1490 type-qualifier-list type-qualifier
1491
1492 typedef-name:
1493 identifier
1494
1495 2.3) CTF-specific declarations
1496
1497 ctf-specifier:
1498 event { ctf-assignment-expression-list-opt }
1499 stream { ctf-assignment-expression-list-opt }
1500 trace { ctf-assignment-expression-list-opt }
1501 typealias declaration-specifiers abstract-declarator-list := declaration-specifiers abstract-declarator-list ;
1502 typealias declaration-specifiers abstract-declarator-list := declarator-list ;
1503
1504 ctf-type-specifier:
1505 floating_point { ctf-assignment-expression-list-opt }
1506 integer { ctf-assignment-expression-list-opt }
1507 string { ctf-assignment-expression-list-opt }
1508
1509 ctf-assignment-expression-list:
1510 ctf-assignment-expression
1511 ctf-assignment-expression-list ; ctf-assignment-expression
1512
1513 ctf-assignment-expression:
1514 unary-expression assignment-operator unary-expression
1515 unary-expression type-assignment-operator type-specifier
1516 declaration-specifiers storage-class-specifier declaration-specifiers declarator-list
1517 typealias declaration-specifiers abstract-declarator-list := declaration-specifiers abstract-declarator-list
1518 typealias declaration-specifiers abstract-declarator-list := declarator-list
This page took 0.103415 seconds and 4 git commands to generate.