2 RFC: Common Trace Format (CTF) Proposal (pre-v1.7)
4 Mathieu Desnoyers, EfficiOS Inc.
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.
12 The latest version of this document can be found at:
14 git tree: git://git.efficios.com/ctf.git
15 gitweb: http://git.efficios.com/?p=ctf.git
17 A reference implementation of a library to read and write this trace format is
18 being implemented within the BabelTrace project, a converter between trace
19 formats. The development tree is available at:
21 git tree: git://git.efficios.com/babeltrace.git
22 gitweb: http://git.efficios.com/?p=babeltrace.git
25 1. Preliminary definitions
27 - Event Trace: An ordered sequence of events.
28 - Event Stream: An ordered sequence of events, containing a subset of the
30 - Event Packet: A sequence of physically contiguous events within an event
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
35 e.g. event: irq_entry.
36 - An event (or event record) relates to a specific instance of an event
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.
43 2. High-level representation of a trace
45 A trace is divided into multiple event streams. Each event stream contains a
46 subset of the trace event types.
48 The final output of the trace, after its generation and optional transport over
49 the network, is expected to be either on permanent or temporary storage in a
50 virtual file system. Because each event stream is appended to while a trace is
51 being recorded, each is associated with a separate file for output. Therefore,
52 a stored trace can be represented as a directory containing one file per stream.
54 A metadata event stream contains information on trace event types. It describes:
58 - Per-stream event header description.
59 - Per-stream event header selection.
60 - Per-stream event context fields.
62 - Event type to stream mapping.
63 - Event type to name mapping.
64 - Event type to ID mapping.
65 - Event fields description.
70 An event stream is divided in contiguous event packets of variable size. These
71 subdivisions have a variable size. An event packet can contain a certain
72 amount of padding at the end. The stream header is repeated at the
73 beginning of each event packet. The rationale for the event stream
74 design choices is explained in Appendix B. Stream Header Rationale.
76 The event stream header will therefore be referred to as the "event packet
77 header" throughout the rest of this document.
82 Types are organized as type classes. Each type class belong to either of two
83 kind of types: basic types or compound types.
87 A basic type is a scalar type, as described in this section. It includes
88 integers, GNU/C bitfields, enumerations, and floating point values.
90 4.1.1 Type inheritance
92 Type specifications can be inherited to allow deriving types from a
93 type class. For example, see the uint32_t named type derived from the "integer"
94 type class below ("Integers" section). Types have a precise binary
95 representation in the trace. A type class has methods to read and write these
96 types, but must be derived into a type to be usable in an event field.
100 We define "byte-packed" types as aligned on the byte size, namely 8-bit.
101 We define "bit-packed" types as following on the next bit, as defined by the
104 All basic types, except bitfields, are either aligned on an architecture-defined
105 specific alignment or byte-packed, depending on the architecture preference.
106 Architectures providing fast unaligned write byte-packed basic types to save
107 space, aligning each type on byte boundaries (8-bit). Architectures with slow
108 unaligned writes align types on specific alignment values. If no specific
109 alignment is declared for a type nor its parents, it is assumed to be bit-packed
110 for bitfields and byte-packed for other types.
112 Metadata attribute representation of a specific alignment:
114 align = value; /* value in bits */
118 By default, the native endianness of the source architecture the trace is used.
119 Byte order can be overridden for a basic type by specifying a "byte_order"
120 attribute. Typical use-case is to specify the network byte order (big endian:
121 "be") to save data captured from the network into the trace without conversion.
122 If not specified, the byte order is native.
124 Metadata representation:
126 byte_order = native OR network OR be OR le; /* network and be are aliases */
130 Type size, in bits, for integers and floats is that returned by "sizeof()" in C
131 multiplied by CHAR_BIT.
132 We require the size of "char" and "unsigned char" types (CHAR_BIT) to be fixed
133 to 8 bits for cross-endianness compatibility.
135 Metadata representation:
137 size = value; (value is in bits)
141 Signed integers are represented in two-complement. Integer alignment, size,
142 signedness and byte ordering are defined in the metadata. Integers aligned on
143 byte size (8-bit) and with length multiple of byte size (8-bit) correspond to
144 the C99 standard integers. In addition, integers with alignment and/or size that
145 are _not_ a multiple of the byte size are permitted; these correspond to the C99
146 standard bitfields, with the added specification that the CTF integer bitfields
147 have a fixed binary representation. A MIT-licensed reference implementation of
148 the CTF portable bitfields is available at:
150 http://git.efficios.com/?p=babeltrace.git;a=blob;f=include/babeltrace/bitfield.h
152 Binary representation of integers:
154 - On little and big endian:
155 - Within a byte, high bits correspond to an integer high bits, and low bits
156 correspond to low bits.
158 - Integer across multiple bytes are placed from the less significant to the
160 - Consecutive integers are placed from lower bits to higher bits (even within
163 - Integer across multiple bytes are placed from the most significant to the
165 - Consecutive integers are placed from higher bits to lower bits (even within
168 This binary representation is derived from the bitfield implementation in GCC
169 for little and big endian. However, contrary to what GCC does, integers can
170 cross units boundaries (no padding is required). Padding can be explicitely
171 added (see 4.1.6 GNU/C bitfields) to follow the GCC layout if needed.
173 Metadata representation:
176 signed = true OR false; /* default false */
177 byte_order = native OR network OR be OR le; /* default native */
178 size = value; /* value in bits, no default */
179 align = value; /* value in bits */
182 Example of type inheritance (creation of a uint32_t named type):
190 Definition of a named 5-bit signed bitfield:
198 4.1.6 GNU/C bitfields
200 The GNU/C bitfields follow closely the integer representation, with a
201 particularity on alignment: if a bitfield cannot fit in the current unit, the
202 unit is padded and the bitfield starts at the following unit. The unit size is
203 defined by the size of the type "unit_type".
205 Metadata representation:
209 As an example, the following structure declared in C compiled by GCC:
216 The example structure is aligned on the largest element (short). The second
217 bitfield would be aligned on the next unit boundary, because it would not fit in
222 The floating point values byte ordering is defined in the metadata.
224 Floating point values follow the IEEE 754-2008 standard interchange formats.
225 Description of the floating point values include the exponent and mantissa size
226 in bits. Some requirements are imposed on the floating point values:
228 - FLT_RADIX must be 2.
229 - mant_dig is the number of digits represented in the mantissa. It is specified
230 by the ISO C99 standard, section 5.2.4, as FLT_MANT_DIG, DBL_MANT_DIG and
231 LDBL_MANT_DIG as defined by <float.h>.
232 - exp_dig is the number of digits represented in the exponent. Given that
233 mant_dig is one bit more than its actual size in bits (leading 1 is not
234 needed) and also given that the sign bit always takes one bit, exp_dig can be
237 - sizeof(float) * CHAR_BIT - FLT_MANT_DIG
238 - sizeof(double) * CHAR_BIT - DBL_MANT_DIG
239 - sizeof(long double) * CHAR_BIT - LDBL_MANT_DIG
241 Metadata representation:
246 byte_order = native OR network OR be OR le;
249 Example of type inheritance:
251 typealias floating_point {
252 exp_dig = 8; /* sizeof(float) * CHAR_BIT - FLT_MANT_DIG */
253 mant_dig = 24; /* FLT_MANT_DIG */
257 TODO: define NaN, +inf, -inf behavior.
261 Enumerations are a mapping between an integer type and a table of strings. The
262 numerical representation of the enumeration follows the integer type specified
263 by the metadata. The enumeration mapping table is detailed in the enumeration
264 description within the metadata. The mapping table maps inclusive value ranges
265 (or single values) to strings. Instead of being limited to simple
266 "value -> string" mappings, these enumerations map
267 "[ start_value ... end_value ] -> string", which map inclusive ranges of
268 values to strings. An enumeration from the C language can be represented in
269 this format by having the same start_value and end_value for each element, which
270 is in fact a range of size 1. This single-value range is supported without
271 repeating the start and end values with the value = string declaration.
273 If a numeric value is encountered between < >, it represents the integer type
274 size used to hold the enumeration, in bits.
276 enum name <integer_type OR size> {
277 somestring = start_value1 ... end_value1,
278 "other string" = start_value2 ... end_value2,
279 yet_another_string, /* will be assigned to end_value2 + 1 */
280 "some other string" = value,
284 If the values are omitted, the enumeration starts at 0 and increment of 1 for
295 Overlapping ranges within a single enumeration are implementation defined.
297 A nameless enumeration can be declared as a field type or as part of a typedef:
299 enum <integer_type> {
306 Compound are aggregation of type declarations. Compound types include
307 structures, variant, arrays, sequences, and strings.
311 Structures are aligned on the largest alignment required by basic types
312 contained within the structure. (This follows the ISO/C standard for structures)
314 Metadata representation of a named structure:
317 field_type field_name;
318 field_type field_name;
325 integer { /* Nameless type */
330 uint64_t second_field_name; /* Named type declared in the metadata */
333 The fields are placed in a sequence next to each other. They each possess a
334 field name, which is a unique identifier within the structure.
336 A nameless structure can be declared as a field type or as part of a typedef:
342 4.2.2 Variants (Discriminated/Tagged Unions)
344 A CTF variant is a selection between different types. A CTF variant must
345 always be defined within the scope of a structure or within fields
346 contained within a structure (defined recursively). A "tag" enumeration
347 field must appear in either the same lexical scope, prior to the variant
348 field (in field declaration order), in an uppermost lexical scope (see
349 Section 7.2.1), or in an uppermost dynamic scope (see Section 7.2.2).
350 The type selection is indicated by the mapping from the enumeration
351 value to the string used as variant type selector. The field to use as
352 tag is specified by the "tag_field", specified between "< >" after the
353 "variant" keyword for unnamed variants, and after "variant name" for
356 The alignment of the variant is the alignment of the type as selected by the tag
357 value for the specific instance of the variant. The alignment of the type
358 containing the variant is independent of the variant alignment. The size of the
359 variant is the size as selected by the tag value for the specific instance of
362 A named variant declaration followed by its definition within a structure
373 enum <integer_type or size> { sel1, sel2, sel3, ... } tag_field;
375 variant name <tag_field> v;
378 An unnamed variant definition within a structure is expressed by the following
382 enum <integer_type or size> { sel1, sel2, sel3, ... } tag_field;
384 variant <tag_field> {
392 Example of a named variant within a sequence that refers to a single tag field:
401 enum <uint2_t> { a, b, c } choice;
402 variant example <choice> v[unsigned int];
405 Example of an unnamed variant:
408 enum <uint2_t> { a, b, c, d } choice;
409 /* Unrelated fields can be added between the variant and its tag */
422 Example of an unnamed variant within an array:
425 enum <uint2_t> { a, b, c } choice;
433 Example of a variant type definition within a structure, where the defined type
434 is then declared within an array of structures. This variant refers to a tag
435 located in an upper lexical scope. This example clearly shows that a variant
436 type definition referring to the tag "x" uses the closest preceding field from
437 the lexical scope of the type definition.
440 enum <uint2_t> { a, b, c, d } x;
442 typedef variant <x> { /*
443 * "x" refers to the preceding "x" enumeration in the
444 * lexical scope of the type definition.
452 enum <int> { x, y, z } x; /* This enumeration is not used by "v". */
453 example_variant v; /*
454 * "v" uses the "enum <uint2_t> { a, b, c, d }"
462 Arrays are fixed-length. Their length is declared in the type declaration within
463 the metadata. They contain an array of "inner type" elements, which can refer to
464 any type not containing the type of the array being declared (no circular
465 dependency). The length is the number of elements in an array.
467 Metadata representation of a named array:
469 typedef elem_type name[length];
471 A nameless array can be declared as a field type within a structure, e.g.:
473 uint8_t field_name[10];
478 Sequences are dynamically-sized arrays. They start with an integer that specify
479 the length of the sequence, followed by an array of "inner type" elements.
480 The length is the number of elements in the sequence.
482 Metadata representation for a named sequence:
484 typedef elem_type name[length_type];
486 A nameless sequence can be declared as a field type, e.g.:
488 long field_name[int];
490 The length type follows the integer types specifications, and the sequence
491 elements follow the "array" specifications.
495 Strings are an array of bytes of variable size and are terminated by a '\0'
496 "NULL" character. Their encoding is described in the metadata. In absence of
497 encoding attribute information, the default encoding is UTF-8.
499 Metadata representation of a named string type:
502 encoding = UTF8 OR ASCII;
505 A nameless string type can be declared as a field type:
507 string field_name; /* Use default UTF8 encoding */
509 5. Event Packet Header
511 The event packet header consists of two part: one is mandatory and have a fixed
512 layout. The second part, the "event packet context", has its layout described in
515 - Aligned on page size. Fixed size. Fields either aligned or packed (depending
516 on the architecture preference).
517 No padding at the end of the event packet header. Native architecture byte
520 Fixed layout (event packet header):
522 - Magic number (CTF magic numbers: 0xC1FC1FC1 and its reverse endianness
523 representation: 0xC11FFCC1) It needs to have a non-symmetric bytewise
524 representation. Used to distinguish between big and little endian traces (this
525 information is determined by knowing the endianness of the architecture
526 reading the trace and comparing the magic number against its value and the
527 reverse, 0xC11FFCC1). This magic number specifies that we use the CTF metadata
528 description language described in this document. Different magic numbers
529 should be used for other metadata description languages.
530 - Trace UUID, used to ensure the event packet match the metadata used.
531 (note: we cannot use a metadata checksum because metadata can be appended to
532 while tracing is active)
533 - Stream ID, used as reference to stream description in metadata.
535 Metadata-defined layout (event packet context):
537 - Event packet content size (in bytes).
538 - Event packet size (in bytes, includes padding).
539 - Event packet content checksum (optional). Checksum excludes the event packet
541 - Per-stream event packet sequence count (to deal with UDP packet loss). The
542 number of significant sequence counter bits should also be present, so
543 wrap-arounds are deal with correctly.
544 - Timestamp at the beginning and timestamp at the end of the event packet.
545 Both timestamps are written in the packet header, but sampled respectively
546 while (or before) writing the first event and while (or after) writing the
547 last event in the packet. The inclusive range between these timestamps should
548 include all event timestamps assigned to events contained within the packet.
549 - Events discarded count
550 - Snapshot of a per-stream free-running counter, counting the number of
551 events discarded that were supposed to be written in the stream prior to
552 the first event in the event packet.
553 * Note: producer-consumer buffer full condition should fill the current
554 event packet with padding so we know exactly where events have been
556 - Lossless compression scheme used for the event packet content. Applied
557 directly to raw data. New types of compression can be added in following
558 versions of the format.
559 0: no compression scheme
563 - Cypher used for the event packet content. Applied after compression.
566 - Checksum scheme used for the event packet content. Applied after encryption.
572 5.1 Event Packet Header Fixed Layout Description
574 struct event_packet_header {
576 uint8_t trace_uuid[16];
580 5.2 Event Packet Context Description
582 Event packet context example. These are declared within the stream declaration
583 in the metadata. All these fields are optional except for "content_size" and
584 "packet_size", which must be present in the context.
586 An example event packet context type:
588 struct event_packet_context {
589 uint64_t timestamp_begin;
590 uint64_t timestamp_end;
592 uint32_t stream_packet_count;
593 uint32_t events_discarded;
595 uint32_t/uint16_t content_size;
596 uint32_t/uint16_t packet_size;
597 uint8_t stream_packet_count_bits; /* Significant counter bits */
598 uint8_t compression_scheme;
599 uint8_t encryption_scheme;
600 uint8_t checksum_scheme;
606 The overall structure of an event is:
608 1 - Stream Packet Context (as specified by the stream metadata)
609 2 - Event Header (as specified by the stream metadata)
610 3 - Stream Event Context (as specified by the stream metadata)
611 4 - Event Context (as specified by the event metadata)
612 5 - Event Payload (as specified by the event metadata)
614 This structure defines an implicit dynamic scoping, where variants
615 located in inner structures (those with a higher number in the listing
616 above) can refer to the fields of outer structures (with lower number in
617 the listing above). See Section 7.2 Metadata Scopes for more detail.
621 Event headers can be described within the metadata. We hereby propose, as an
622 example, two types of events headers. Type 1 accommodates streams with less than
623 31 event IDs. Type 2 accommodates streams with 31 or more event IDs.
625 One major factor can vary between streams: the number of event IDs assigned to
626 a stream. Luckily, this information tends to stay relatively constant (modulo
627 event registration while trace is being recorded), so we can specify different
628 representations for streams containing few event IDs and streams containing
629 many event IDs, so we end up representing the event ID and timestamp as densely
630 as possible in each case.
632 The header is extended in the rare occasions where the information cannot be
633 represented in the ranges available in the standard event header. They are also
634 used in the rare occasions where the data required for a field could not be
635 collected: the flag corresponding to the missing field within the missing_fields
636 array is then set to 1.
638 Types uintX_t represent an X-bit unsigned integer.
641 6.1.1 Type 1 - Few event IDs
643 - Aligned on 32-bit (or 8-bit if byte-packed, depending on the architecture
645 - Native architecture byte ordering.
646 - For "compact" selection
647 - Fixed size: 32 bits.
648 - For "extended" selection
649 - Size depends on the architecture and variant alignment.
651 struct event_header_1 {
654 * id 31 is reserved to indicate an extended header.
656 enum <uint5_t> { compact = 0 ... 30, extended = 31 } id;
662 uint32_t id; /* 32-bit event IDs */
663 uint64_t timestamp; /* 64-bit timestamps */
669 6.1.2 Type 2 - Many event IDs
671 - Aligned on 16-bit (or 8-bit if byte-packed, depending on the architecture
673 - Native architecture byte ordering.
674 - For "compact" selection
675 - Size depends on the architecture and variant alignment.
676 - For "extended" selection
677 - Size depends on the architecture and variant alignment.
679 struct event_header_2 {
681 * id: range: 0 - 65534.
682 * id 65535 is reserved to indicate an extended header.
684 enum <uint16_t> { compact = 0 ... 65534, extended = 65535 } id;
690 uint32_t id; /* 32-bit event IDs */
691 uint64_t timestamp; /* 64-bit timestamps */
699 The event context contains information relative to the current event. The choice
700 and meaning of this information is specified by the metadata "stream" and
701 "event" information. The "stream" context is applied to all events within the
702 stream. The "stream" context structure follows the event header. The "event"
703 context is applied to specific events. Its structure follows the "stream"
706 An example of stream-level event context is to save the event payload size with
707 each event, or to save the current PID with each event. These are declared
708 within the stream declaration within the metadata:
716 uint16_t payload_size;
721 An example of event-specific event context is to declare a bitmap of missing
722 fields, only appended after the stream event context if the extended event
723 header is selected. NR_FIELDS is the number of fields within the event (a
731 uint1_t missing_fields[NR_FIELDS]; /* missing event fields bitmap */
740 An event payload contains fields specific to a given event type. The fields
741 belonging to an event type are described in the event-specific metadata
742 within a structure type.
746 No padding at the end of the event payload. This differs from the ISO/C standard
747 for structures, but follows the CTF standard for structures. In a trace, even
748 though it makes sense to align the beginning of a structure, it really makes no
749 sense to add padding at the end of the structure, because structures are usually
750 not followed by a structure of the same type.
752 This trick can be done by adding a zero-length "end" field at the end of the C
753 structures, and by using the offset of this field rather than using sizeof()
754 when calculating the size of a structure (see Appendix "A. Helper macros").
758 The event payload is aligned on the largest alignment required by types
759 contained within the payload. (This follows the ISO/C standard for structures)
764 The meta-data is located in a stream named "metadata". It is made of "event
765 packets", which each start with an event packet header. The event type within
766 the metadata stream have no event header nor event context. Each event only
767 contains a null-terminated "string" payload, which is a metadata description
768 entry. The events are packed one next to another. Each event packet start with
769 an event packet header, which contains, amongst other fields, the magic number
770 and trace UUID. The trace UUID is represented as a string of hexadecimal digits
773 The metadata can be parsed by reading through the metadata strings, skipping
774 newlines and null-characters. Type names are made of a single identifier, and
775 can be surrounded by prefix/postfix. Text contained within "/*" and "*/", as
776 well as within "//" and end of line, are treated as comments. Boolean values can
777 be represented as true, TRUE, or 1 for true, and false, FALSE, or 0 for false.
780 7.1 Declaration vs Definition
782 A declaration associates a layout to a type, without specifying where
783 this type is located in the event structure hierarchy (see Section 6).
784 This therefore includes typedef, typealias, as well as all type
785 specifiers. In certain circumstances (typedef, structure field and
786 variant field), a declaration is followed by a declarator, which specify
787 the newly defined type name (for typedef), or the field name (for
788 declarations located within structure and variants). Array and sequence,
789 declared with square brackets ("[" "]"), are part of the declarator,
790 similarly to C99. The enumeration type specifier and variant tag name
791 (both specified with "<" ">") are part of the type specifier.
793 A definition associates a type to a location in the event structure
794 hierarchy (see Section 6). This association is denoted by ":=", as shown
800 CTF metadata uses two different types of scoping: a lexical scope is
801 used for declarations and type definitions, and a dynamic scope is used
802 for variants references to tag fields.
806 Each of "trace", "stream", "event", "struct" and "variant" have their own
807 nestable declaration scope, within which types can be declared using "typedef"
808 and "typealias". A root declaration scope also contains all declarations
809 located outside of any of the aforementioned declarations. An inner
810 declaration scope can refer to type declared within its container
811 lexical scope prior to the inner declaration scope. Redefinition of a
812 typedef or typealias is not valid, although hiding an upper scope
813 typedef or typealias is allowed within a sub-scope.
817 A dynamic scope consists in the lexical scope augmented with the
818 implicit event structure definition hierarchy presented at Section 6.
819 The dynamic scope is only used for variant tag definitions. It is used
820 at definition time to look up the location of the tag field associated
823 Therefore, variants in lower levels in the dynamic scope (e.g. event
824 context) can refer to a tag field located in upper levels (e.g. in the
825 event header) by specifying, in this case, the associated tag with
826 <header.field_name>. This allows, for instance, the event context to
827 define a variant referring to the "id" field of the event header as
830 The target dynamic scope must be specified explicitly when referring to
831 a field outside of the local static scope. The dynamic scope prefixes
834 - Stream Packet Context: <stream.packet.context. >,
835 - Event Header: <stream.event.header. >,
836 - Stream Event Context: <stream.event.context. >,
837 - Event Context: <event.context. >,
838 - Event Payload: <event.fields. >.
840 Multiple declarations of the same field name within a single scope is
841 not valid. It is however valid to re-use the same field name in
842 different scopes. There is no possible conflict, because the dynamic
843 scope must be specified when a variant refers to a tag field located in
844 a different dynamic scope.
846 The information available in the dynamic scopes can be thought of as the
847 current tracing context. At trace production, information about the
848 current context is saved into the specified scope field levels. At trace
849 consumption, for each event, the current trace context is therefore
850 readable by accessing the upper dynamic scopes.
853 7.3 Metadata Examples
855 The grammar representing the CTF metadata is presented in
856 Appendix C. CTF Metadata Grammar. This section presents a rather ligher
857 reading that consists in examples of CTF metadata, with template values:
860 major = value; /* Trace format version */
862 uuid = "aaaaaaaa-aaaa-aaaa-aaaa-aaaaaaaaaaaa"; /* Trace UUID */
868 /* Type 1 - Few event IDs; Type 2 - Many event IDs. See section 6.1. */
869 event.header := event_header_1 OR event_header_2;
870 event.context := struct {
873 packet.context := struct {
880 id = value; /* Numeric identifier within the stream */
890 /* More detail on types in section 4. Types */
895 * Type declarations behave similarly to the C standard.
898 typedef aliased_type_prefix aliased_type new_type aliased_type_postfix;
900 /* e.g.: typedef struct example new_type_name[10]; */
905 * The "typealias" declaration can be used to give a name (including
906 * prefix/postfix) to a type. It should also be used to map basic C types
907 * (float, int, unsigned long, ...) to a CTF type. Typealias is a superset of
908 * "typedef": it also allows assignment of a simple variable identifier to a
912 typealias type_class {
914 } : new_type_prefix new_type new_type_postfix;
918 * typealias integer {
924 * typealias integer {
939 enum name <integer_type or size> {
945 * Unnamed types, contained within compound type fields, typedef or typealias.
956 enum <integer_type or size> {
960 typedef type new_type[length];
963 type field_name[length];
966 typedef type new_type[length_type];
969 type field_name[length_type];
981 integer_type field_name:size; /* GNU/C bitfield */
991 The two following macros keep track of the size of a GNU/C structure without
992 padding at the end by placing HEADER_END as the last field. A one byte end field
993 is used for C90 compatibility (C99 flexible arrays could be used here). Note
994 that this does not affect the effective structure size, which should always be
995 calculated with the header_sizeof() helper.
997 #define HEADER_END char end_field
998 #define header_sizeof(type) offsetof(typeof(type), end_field)
1001 B. Stream Header Rationale
1003 An event stream is divided in contiguous event packets of variable size. These
1004 subdivisions allow the trace analyzer to perform a fast binary search by time
1005 within the stream (typically requiring to index only the event packet headers)
1006 without reading the whole stream. These subdivisions have a variable size to
1007 eliminate the need to transfer the event packet padding when partially filled
1008 event packets must be sent when streaming a trace for live viewing/analysis.
1009 An event packet can contain a certain amount of padding at the end. Dividing
1010 streams into event packets is also useful for network streaming over UDP and
1011 flight recorder mode tracing (a whole event packet can be swapped out of the
1012 buffer atomically for reading).
1014 The stream header is repeated at the beginning of each event packet to allow
1015 flexibility in terms of:
1017 - streaming support,
1018 - allowing arbitrary buffers to be discarded without making the trace
1020 - allow UDP packet loss handling by either dealing with missing event packet
1021 or asking for re-transmission.
1022 - transparently support flight recorder mode,
1023 - transparently support crash dump.
1025 The event stream header will therefore be referred to as the "event packet
1026 header" throughout the rest of this document.
1028 C. CTF Metadata Grammar
1031 * Common Trace Format (CTF) Metadata Grammar.
1033 * Inspired from the C99 grammar:
1034 * http://www.open-std.org/jtc1/sc22/wg14/www/docs/n1124.pdf (Annex A)
1036 * Specialized for CTF needs by including only constant and declarations from
1037 * C99 (excluding function declarations), and by adding support for variants,
1038 * sequences and CTF-specific specifiers.
1043 1.1) Lexical elements
1086 identifier identifier-nondigit
1089 identifier-nondigit:
1091 universal-character-name
1092 any other implementation-defined characters
1096 [a-zA-Z] /* regular expression */
1099 [0-9] /* regular expression */
1101 1.4) Universal character names
1103 universal-character-name:
1105 \U hex-quad hex-quad
1108 hexadecimal-digit hexadecimal-digit hexadecimal-digit hexadecimal-digit
1114 enumeration-constant
1118 decimal-constant integer-suffix-opt
1119 octal-constant integer-suffix-opt
1120 hexadecimal-constant integer-suffix-opt
1124 decimal-constant digit
1128 octal-constant octal-digit
1130 hexadecimal-constant:
1131 hexadecimal-prefix hexadecimal-digit
1132 hexadecimal-constant hexadecimal-digit
1142 unsigned-suffix long-suffix-opt
1143 unsigned-suffix long-long-suffix
1144 long-suffix unsigned-suffix-opt
1145 long-long-suffix unsigned-suffix-opt
1161 digit-sequence digit
1163 hexadecimal-digit-sequence:
1165 hexadecimal-digit-sequence hexadecimal-digit
1167 enumeration-constant:
1173 L' c-char-sequence '
1177 c-char-sequence c-char
1180 any member of source charset except single-quote ('), backslash
1181 (\), or new-line character.
1185 simple-escape-sequence
1186 octal-escape-sequence
1187 hexadecimal-escape-sequence
1188 universal-character-name
1190 simple-escape-sequence: one of
1191 \' \" \? \\ \a \b \f \n \r \t \v
1193 octal-escape-sequence:
1195 \ octal-digit octal-digit
1196 \ octal-digit octal-digit octal-digit
1198 hexadecimal-escape-sequence:
1199 \x hexadecimal-digit
1200 hexadecimal-escape-sequence hexadecimal-digit
1202 1.6) String literals
1205 " s-char-sequence-opt "
1206 L" s-char-sequence-opt "
1210 s-char-sequence s-char
1213 any member of source charset except double-quote ("), backslash
1214 (\), or new-line character.
1220 [ ] ( ) { } . -> * + - < > : ; ... = ,
1223 2) Phrase structure grammar
1229 ( unary-expression )
1233 postfix-expression [ unary-expression ]
1234 postfix-expression . identifier
1235 postfix-expressoin -> identifier
1239 unary-operator postfix-expression
1241 unary-operator: one of
1244 assignment-operator:
1247 type-assignment-operator:
1250 constant-expression:
1253 constant-expression-range:
1254 constant-expression ... constant-expression
1259 declaration-specifiers declarator-list-opt ;
1262 declaration-specifiers:
1263 storage-class-specifier declaration-specifiers-opt
1264 type-specifier declaration-specifiers-opt
1265 type-qualifier declaration-specifiers-opt
1269 declarator-list , declarator
1271 abstract-declarator-list:
1273 abstract-declarator-list , abstract-declarator
1275 storage-class-specifier:
1298 struct identifier-opt { struct-or-variant-declaration-list-opt }
1301 struct-or-variant-declaration-list:
1302 struct-or-variant-declaration
1303 struct-or-variant-declaration-list struct-or-variant-declaration
1305 struct-or-variant-declaration:
1306 specifier-qualifier-list struct-or-variant-declarator-list ;
1307 declaration-specifiers storage-class-specifier declaration-specifiers declarator-list ;
1308 typealias declaration-specifiers abstract-declarator-list : declaration-specifiers abstract-declarator-list ;
1309 typealias declaration-specifiers abstract-declarator-list : declarator-list ;
1311 specifier-qualifier-list:
1312 type-specifier specifier-qualifier-list-opt
1313 type-qualifier specifier-qualifier-list-opt
1315 struct-or-variant-declarator-list:
1316 struct-or-variant-declarator
1317 struct-or-variant-declarator-list , struct-or-variant-declarator
1319 struct-or-variant-declarator:
1321 declarator-opt : constant-expression
1324 variant identifier-opt variant-tag-opt { struct-or-variant-declaration-list }
1325 variant identifier variant-tag
1331 enum identifier-opt { enumerator-list }
1332 enum identifier-opt { enumerator-list , }
1334 enum identifier-opt < declaration-specifiers > { enumerator-list }
1335 enum identifier-opt < declaration-specifiers > { enumerator-list , }
1336 enum identifier < declaration-specifiers >
1337 enum identifier-opt < integer-constant > { enumerator-list }
1338 enum identifier-opt < integer-constant > { enumerator-list , }
1339 enum identifier < integer-constant >
1343 enumerator-list , enumerator
1346 enumeration-constant
1347 enumeration-constant = constant-expression
1348 enumeration-constant = constant-expression-range
1354 pointer-opt direct-declarator
1359 direct-declarator [ type-specifier ]
1360 direct-declarator [ constant-expression ]
1362 abstract-declarator:
1363 pointer-opt direct-abstract-declarator
1365 direct-abstract-declarator:
1367 ( abstract-declarator )
1368 direct-abstract-declarator [ type-specifier ]
1369 direct-abstract-declarator [ constant-expression ]
1370 direct-abstract-declarator [ ]
1373 * type-qualifier-list-opt
1374 * type-qualifier-list-opt pointer
1376 type-qualifier-list:
1378 type-qualifier-list type-qualifier
1383 2.3) CTF-specific declarations
1386 event { ctf-assignment-expression-list-opt }
1387 stream { ctf-assignment-expression-list-opt }
1388 trace { ctf-assignment-expression-list-opt }
1389 typealias declaration-specifiers abstract-declarator-list : declaration-specifiers abstract-declarator-list ;
1390 typealias declaration-specifiers abstract-declarator-list : declarator-list ;
1393 floating_point { ctf-assignment-expression-list-opt }
1394 integer { ctf-assignment-expression-list-opt }
1395 string { ctf-assignment-expression-list-opt }
1397 ctf-assignment-expression-list:
1398 ctf-assignment-expression
1399 ctf-assignment-expression-list ; ctf-assignment-expression
1401 ctf-assignment-expression:
1402 unary-expression assignment-operator unary-expression
1403 unary-expression type-assignment-operator type-specifier
1404 declaration-specifiers storage-class-specifier declaration-specifiers declarator-list
1405 typealias declaration-specifiers abstract-declarator-list : declaration-specifiers abstract-declarator-list
1406 typealias declaration-specifiers abstract-declarator-list : declarator-list