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, it is assumed to be bit-packed for
110 integers with size not multiple of 8 bits and for gcc bitfields. All
111 other types are byte-packed.
113 Metadata attribute representation of a specific alignment:
115 align = value; /* value in bits */
119 By default, the native endianness of the source architecture the trace is used.
120 Byte order can be overridden for a basic type by specifying a "byte_order"
121 attribute. 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.
123 If not specified, the byte order is native.
125 Metadata representation:
127 byte_order = native OR network OR be OR le; /* network and be are aliases */
131 Type size, in bits, for integers and floats is that returned by "sizeof()" in C
132 multiplied by CHAR_BIT.
133 We require the size of "char" and "unsigned char" types (CHAR_BIT) to be fixed
134 to 8 bits for cross-endianness compatibility.
136 Metadata representation:
138 size = value; (value is in bits)
142 Signed integers are represented in two-complement. Integer alignment, size,
143 signedness and byte ordering are defined in the metadata. Integers aligned on
144 byte size (8-bit) and with length multiple of byte size (8-bit) correspond to
145 the C99 standard integers. In addition, integers with alignment and/or size that
146 are _not_ a multiple of the byte size are permitted; these correspond to the C99
147 standard bitfields, with the added specification that the CTF integer bitfields
148 have a fixed binary representation. A MIT-licensed reference implementation of
149 the CTF portable bitfields is available at:
151 http://git.efficios.com/?p=babeltrace.git;a=blob;f=include/babeltrace/bitfield.h
153 Binary representation of integers:
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.
159 - Integer across multiple bytes are placed from the less significant to the
161 - Consecutive integers are placed from lower bits to higher bits (even within
164 - Integer across multiple bytes are placed from the most significant to the
166 - Consecutive integers are placed from higher bits to lower bits (even within
169 This binary representation is derived from the bitfield implementation in GCC
170 for little and big endian. However, contrary to what GCC does, integers can
171 cross units boundaries (no padding is required). Padding can be explicitely
172 added (see 4.1.6 GNU/C bitfields) to follow the GCC layout if needed.
174 Metadata representation:
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 */
183 Example of type inheritance (creation of a uint32_t named type):
191 Definition of a named 5-bit signed bitfield:
199 4.1.6 GNU/C bitfields
201 The GNU/C bitfields follow closely the integer representation, with a
202 particularity on alignment: if a bitfield cannot fit in the current unit, the
203 unit is padded and the bitfield starts at the following unit. The unit size is
204 defined by the size of the type "unit_type".
206 Metadata representation:
210 As an example, the following structure declared in C compiled by GCC:
217 The example structure is aligned on the largest element (short). The second
218 bitfield would be aligned on the next unit boundary, because it would not fit in
223 The floating point values byte ordering is defined in the metadata.
225 Floating point values follow the IEEE 754-2008 standard interchange formats.
226 Description of the floating point values include the exponent and mantissa size
227 in bits. Some requirements are imposed on the floating point values:
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
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
242 Metadata representation:
247 byte_order = native OR network OR be OR le;
250 Example of type inheritance:
252 typealias floating_point {
253 exp_dig = 8; /* sizeof(float) * CHAR_BIT - FLT_MANT_DIG */
254 mant_dig = 24; /* FLT_MANT_DIG */
258 TODO: define NaN, +inf, -inf behavior.
262 Enumerations are a mapping between an integer type and a table of strings. The
263 numerical representation of the enumeration follows the integer type specified
264 by the metadata. The enumeration mapping table is detailed in the enumeration
265 description 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
268 "[ start_value ... end_value ] -> string", which map inclusive ranges of
269 values to strings. An enumeration from the C language can be represented in
270 this format by having the same start_value and end_value for each element, which
271 is in fact a range of size 1. This single-value range is supported without
272 repeating the start and end values with the value = string declaration.
274 enum name <integer_type> {
275 somestring = start_value1 ... end_value1,
276 "other string" = start_value2 ... end_value2,
277 yet_another_string, /* will be assigned to end_value2 + 1 */
278 "some other string" = value,
282 If the values are omitted, the enumeration starts at 0 and increment of 1 for
285 enum name <unsigned int> {
293 Overlapping ranges within a single enumeration are implementation defined.
295 A nameless enumeration can be declared as a field type or as part of a typedef:
297 enum <integer_type> {
304 Compound are aggregation of type declarations. Compound types include
305 structures, variant, arrays, sequences, and strings.
309 Structures are aligned on the largest alignment required by basic types
310 contained within the structure. (This follows the ISO/C standard for structures)
312 Metadata representation of a named structure:
315 field_type field_name;
316 field_type field_name;
323 integer { /* Nameless type */
328 uint64_t second_field_name; /* Named type declared in the metadata */
331 The fields are placed in a sequence next to each other. They each possess a
332 field name, which is a unique identifier within the structure.
334 A nameless structure can be declared as a field type or as part of a typedef:
340 4.2.2 Variants (Discriminated/Tagged Unions)
342 A CTF variant is a selection between different types. A CTF variant must
343 always be defined within the scope of a structure or within fields
344 contained within a structure (defined recursively). A "tag" enumeration
345 field must appear in either the same lexical scope, prior to the variant
346 field (in field declaration order), in an uppermost lexical scope (see
347 Section 7.2.1), or in an uppermost dynamic scope (see Section 7.2.2).
348 The type selection is indicated by the mapping from the enumeration
349 value to the string used as variant type selector. The field to use as
350 tag is specified by the "tag_field", specified between "< >" after the
351 "variant" keyword for unnamed variants, and after "variant name" for
354 The alignment of the variant is the alignment of the type as selected by the tag
355 value for the specific instance of the variant. The alignment of the type
356 containing the variant is independent of the variant alignment. The size of the
357 variant is the size as selected by the tag value for the specific instance of
360 A named variant declaration followed by its definition within a structure
371 enum <integer_type> { sel1, sel2, sel3, ... } tag_field;
373 variant name <tag_field> v;
376 An unnamed variant definition within a structure is expressed by the following
380 enum <integer_type> { sel1, sel2, sel3, ... } tag_field;
382 variant <tag_field> {
390 Example of a named variant within a sequence that refers to a single tag field:
399 enum <uint2_t> { a, b, c } choice;
400 variant example <choice> v[unsigned int];
403 Example of an unnamed variant:
406 enum <uint2_t> { a, b, c, d } choice;
407 /* Unrelated fields can be added between the variant and its tag */
420 Example of an unnamed variant within an array:
423 enum <uint2_t> { a, b, c } choice;
431 Example of a variant type definition within a structure, where the defined type
432 is then declared within an array of structures. This variant refers to a tag
433 located in an upper lexical scope. This example clearly shows that a variant
434 type definition referring to the tag "x" uses the closest preceding field from
435 the lexical scope of the type definition.
438 enum <uint2_t> { a, b, c, d } x;
440 typedef variant <x> { /*
441 * "x" refers to the preceding "x" enumeration in the
442 * lexical scope of the type definition.
450 enum <int> { x, y, z } x; /* This enumeration is not used by "v". */
451 example_variant v; /*
452 * "v" uses the "enum <uint2_t> { a, b, c, d }"
460 Arrays are fixed-length. Their length is declared in the type declaration within
461 the metadata. They contain an array of "inner type" elements, which can refer to
462 any type not containing the type of the array being declared (no circular
463 dependency). The length is the number of elements in an array.
465 Metadata representation of a named array:
467 typedef elem_type name[length];
469 A nameless array can be declared as a field type within a structure, e.g.:
471 uint8_t field_name[10];
476 Sequences are dynamically-sized arrays. They start with an integer that specify
477 the length of the sequence, followed by an array of "inner type" elements.
478 The length is the number of elements in the sequence.
480 Metadata representation for a named sequence:
482 typedef elem_type name[length_type];
484 A nameless sequence can be declared as a field type, e.g.:
486 long field_name[int];
488 The length type follows the integer types specifications, and the sequence
489 elements follow the "array" specifications.
493 Strings are an array of bytes of variable size and are terminated by a '\0'
494 "NULL" character. Their encoding is described in the metadata. In absence of
495 encoding attribute information, the default encoding is UTF-8.
497 Metadata representation of a named string type:
500 encoding = UTF8 OR ASCII;
503 A nameless string type can be declared as a field type:
505 string field_name; /* Use default UTF8 encoding */
507 5. Event Packet Header
509 The event packet header consists of two part: one is mandatory and have a fixed
510 layout. The second part, the "event packet context", has its layout described in
513 - Aligned on page size. Fixed size. Fields either aligned or packed (depending
514 on the architecture preference).
515 No padding at the end of the event packet header. Native architecture byte
518 Fixed layout (event packet header):
520 - Magic number (CTF magic numbers: 0xC1FC1FC1 and its reverse endianness
521 representation: 0xC11FFCC1) It needs to have a non-symmetric bytewise
522 representation. Used to distinguish between big and little endian traces (this
523 information is determined by knowing the endianness of the architecture
524 reading the trace and comparing the magic number against its value and the
525 reverse, 0xC11FFCC1). This magic number specifies that we use the CTF metadata
526 description language described in this document. Different magic numbers
527 should be used for other metadata description languages.
528 - Trace UUID, used to ensure the event packet match the metadata used.
529 (note: we cannot use a metadata checksum because metadata can be appended to
530 while tracing is active)
531 - Stream ID, used as reference to stream description in metadata.
533 Metadata-defined layout (event packet context):
535 - Event packet content size (in bytes).
536 - Event packet size (in bytes, includes padding).
537 - Event packet content checksum (optional). Checksum excludes the event packet
539 - Per-stream event packet sequence count (to deal with UDP packet loss). The
540 number of significant sequence counter bits should also be present, so
541 wrap-arounds are deal with correctly.
542 - Timestamp at the beginning and timestamp at the end of the event packet.
543 Both timestamps are written in the packet header, but sampled respectively
544 while (or before) writing the first event and while (or after) writing the
545 last event in the packet. The inclusive range between these timestamps should
546 include all event timestamps assigned to events contained within the packet.
547 - Events discarded count
548 - Snapshot of a per-stream free-running counter, counting the number of
549 events discarded that were supposed to be written in the stream prior to
550 the first event in the event packet.
551 * Note: producer-consumer buffer full condition should fill the current
552 event packet with padding so we know exactly where events have been
554 - Lossless compression scheme used for the event packet content. Applied
555 directly to raw data. New types of compression can be added in following
556 versions of the format.
557 0: no compression scheme
561 - Cypher used for the event packet content. Applied after compression.
564 - Checksum scheme used for the event packet content. Applied after encryption.
570 5.1 Event Packet Header Fixed Layout Description
572 struct event_packet_header {
574 uint8_t trace_uuid[16];
578 5.2 Event Packet Context Description
580 Event packet context example. These are declared within the stream declaration
581 in the metadata. All these fields are optional except for "content_size" and
582 "packet_size", which must be present in the context.
584 An example event packet context type:
586 struct event_packet_context {
587 uint64_t timestamp_begin;
588 uint64_t timestamp_end;
590 uint32_t stream_packet_count;
591 uint32_t events_discarded;
593 uint32_t/uint16_t content_size;
594 uint32_t/uint16_t packet_size;
595 uint8_t stream_packet_count_bits; /* Significant counter bits */
596 uint8_t compression_scheme;
597 uint8_t encryption_scheme;
598 uint8_t checksum_scheme;
604 The overall structure of an event is:
606 1 - Stream Packet Context (as specified by the stream metadata)
607 2 - Event Header (as specified by the stream metadata)
608 3 - Stream Event Context (as specified by the stream metadata)
609 4 - Event Context (as specified by the event metadata)
610 5 - Event Payload (as specified by the event metadata)
612 This structure defines an implicit dynamic scoping, where variants
613 located in inner structures (those with a higher number in the listing
614 above) can refer to the fields of outer structures (with lower number in
615 the listing above). See Section 7.2 Metadata Scopes for more detail.
619 Event headers can be described within the metadata. We hereby propose, as an
620 example, two types of events headers. Type 1 accommodates streams with less than
621 31 event IDs. Type 2 accommodates streams with 31 or more event IDs.
623 One major factor can vary between streams: the number of event IDs assigned to
624 a stream. Luckily, this information tends to stay relatively constant (modulo
625 event registration while trace is being recorded), so we can specify different
626 representations for streams containing few event IDs and streams containing
627 many event IDs, so we end up representing the event ID and timestamp as densely
628 as possible in each case.
630 The header is extended in the rare occasions where the information cannot be
631 represented in the ranges available in the standard event header. They are also
632 used in the rare occasions where the data required for a field could not be
633 collected: the flag corresponding to the missing field within the missing_fields
634 array is then set to 1.
636 Types uintX_t represent an X-bit unsigned integer.
639 6.1.1 Type 1 - Few event IDs
641 - Aligned on 32-bit (or 8-bit if byte-packed, depending on the architecture
643 - Native architecture byte ordering.
644 - For "compact" selection
645 - Fixed size: 32 bits.
646 - For "extended" selection
647 - Size depends on the architecture and variant alignment.
649 struct event_header_1 {
652 * id 31 is reserved to indicate an extended header.
654 enum <uint5_t> { compact = 0 ... 30, extended = 31 } id;
660 uint32_t id; /* 32-bit event IDs */
661 uint64_t timestamp; /* 64-bit timestamps */
667 6.1.2 Type 2 - Many event IDs
669 - Aligned on 16-bit (or 8-bit if byte-packed, depending on the architecture
671 - Native architecture byte ordering.
672 - For "compact" selection
673 - Size depends on the architecture and variant alignment.
674 - For "extended" selection
675 - Size depends on the architecture and variant alignment.
677 struct event_header_2 {
679 * id: range: 0 - 65534.
680 * id 65535 is reserved to indicate an extended header.
682 enum <uint16_t> { compact = 0 ... 65534, extended = 65535 } id;
688 uint32_t id; /* 32-bit event IDs */
689 uint64_t timestamp; /* 64-bit timestamps */
697 The event context contains information relative to the current event. The choice
698 and meaning of this information is specified by the metadata "stream" and
699 "event" information. The "stream" context is applied to all events within the
700 stream. The "stream" context structure follows the event header. The "event"
701 context is applied to specific events. Its structure follows the "stream"
704 An example of stream-level event context is to save the event payload size with
705 each event, or to save the current PID with each event. These are declared
706 within the stream declaration within the metadata:
714 uint16_t payload_size;
719 An example of event-specific event context is to declare a bitmap of missing
720 fields, only appended after the stream event context if the extended event
721 header is selected. NR_FIELDS is the number of fields within the event (a
729 uint1_t missing_fields[NR_FIELDS]; /* missing event fields bitmap */
738 An event payload contains fields specific to a given event type. The fields
739 belonging to an event type are described in the event-specific metadata
740 within a structure type.
744 No padding at the end of the event payload. This differs from the ISO/C standard
745 for structures, but follows the CTF standard for structures. In a trace, even
746 though it makes sense to align the beginning of a structure, it really makes no
747 sense to add padding at the end of the structure, because structures are usually
748 not followed by a structure of the same type.
750 This trick can be done by adding a zero-length "end" field at the end of the C
751 structures, and by using the offset of this field rather than using sizeof()
752 when calculating the size of a structure (see Appendix "A. Helper macros").
756 The event payload is aligned on the largest alignment required by types
757 contained within the payload. (This follows the ISO/C standard for structures)
762 The meta-data is located in a stream identified by its name: "metadata".
763 It is made of "event packets", which each start with an event packet
764 header. The event type within the metadata stream have no event header
765 nor event context. Each event only contains a null-terminated "string"
766 payload, which is a metadata description entry. The events are packed
767 one next to another. Each event packet start with an event packet
768 header, which contains, amongst other fields, the magic number and trace
769 UUID. In the event packet header, the trace UUID is represented as an
770 array of bytes. Within the string-based metadata description, the trace
771 UUID is represented as a string of hexadecimal digits and dashes "-".
773 The metadata can be parsed by reading through the metadata strings,
774 skipping null-characters. Type names are made of a single identifier,
775 and can be surrounded by prefix/postfix. Text contained within "/*" and
776 "*/", as well as within "//" and end of line, are treated as comments.
777 Boolean values can be represented as true, TRUE, or 1 for true, and
778 false, FALSE, or 0 for false.
781 7.1 Declaration vs Definition
783 A declaration associates a layout to a type, without specifying where
784 this type is located in the event structure hierarchy (see Section 6).
785 This therefore includes typedef, typealias, as well as all type
786 specifiers. In certain circumstances (typedef, structure field and
787 variant field), a declaration is followed by a declarator, which specify
788 the newly defined type name (for typedef), or the field name (for
789 declarations located within structure and variants). Array and sequence,
790 declared with square brackets ("[" "]"), are part of the declarator,
791 similarly to C99. The enumeration type specifier and variant tag name
792 (both specified with "<" ">") are part of the type specifier.
794 A definition associates a type to a location in the event structure
795 hierarchy (see Section 6). This association is denoted by ":=", as shown
801 CTF metadata uses two different types of scoping: a lexical scope is
802 used for declarations and type definitions, and a dynamic scope is used
803 for variants references to tag fields.
807 Each of "trace", "stream", "event", "struct" and "variant" have their own
808 nestable declaration scope, within which types can be declared using "typedef"
809 and "typealias". A root declaration scope also contains all declarations
810 located outside of any of the aforementioned declarations. An inner
811 declaration scope can refer to type declared within its container
812 lexical scope prior to the inner declaration scope. Redefinition of a
813 typedef or typealias is not valid, although hiding an upper scope
814 typedef or typealias is allowed within a sub-scope.
818 A dynamic scope consists in the lexical scope augmented with the
819 implicit event structure definition hierarchy presented at Section 6.
820 The dynamic scope is only used for variant tag definitions. It is used
821 at definition time to look up the location of the tag field associated
824 Therefore, variants in lower levels in the dynamic scope (e.g. event
825 context) can refer to a tag field located in upper levels (e.g. in the
826 event header) by specifying, in this case, the associated tag with
827 <header.field_name>. This allows, for instance, the event context to
828 define a variant referring to the "id" field of the event header as
831 The target dynamic scope must be specified explicitly when referring to
832 a field outside of the local static scope. The dynamic scope prefixes
835 - Stream Packet Context: <stream.packet.context. >,
836 - Event Header: <stream.event.header. >,
837 - Stream Event Context: <stream.event.context. >,
838 - Event Context: <event.context. >,
839 - Event Payload: <event.fields. >.
841 Multiple declarations of the same field name within a single scope is
842 not valid. It is however valid to re-use the same field name in
843 different scopes. There is no possible conflict, because the dynamic
844 scope must be specified when a variant refers to a tag field located in
845 a different dynamic scope.
847 The information available in the dynamic scopes can be thought of as the
848 current tracing context. At trace production, information about the
849 current context is saved into the specified scope field levels. At trace
850 consumption, for each event, the current trace context is therefore
851 readable by accessing the upper dynamic scopes.
854 7.3 Metadata Examples
856 The grammar representing the CTF metadata is presented in
857 Appendix C. CTF Metadata Grammar. This section presents a rather ligher
858 reading that consists in examples of CTF metadata, with template values:
861 major = value; /* Trace format version */
863 uuid = "aaaaaaaa-aaaa-aaaa-aaaa-aaaaaaaaaaaa"; /* Trace UUID */
869 /* Type 1 - Few event IDs; Type 2 - Many event IDs. See section 6.1. */
870 event.header := event_header_1 OR event_header_2;
871 event.context := struct {
874 packet.context := struct {
881 id = value; /* Numeric identifier within the stream */
891 /* More detail on types in section 4. Types */
896 * Type declarations behave similarly to the C standard.
899 typedef aliased_type_specifiers new_type_declarators;
901 /* e.g.: typedef struct example new_type_name[10]; */
906 * The "typealias" declaration can be used to give a name (including
907 * pointer declarator specifier) to a type. It should also be used to
908 * map basic C types (float, int, unsigned long, ...) to a CTF type.
909 * Typealias is a superset of "typedef": it also allows assignment of a
910 * simple variable identifier to a type.
913 typealias type_class {
915 } := type_specifiers type_declarator;
919 * typealias integer {
923 * } := struct page *;
925 * typealias integer {
940 enum name <integer_type> {
946 * Unnamed types, contained within compound type fields, typedef or typealias.
957 enum <integer_type> {
961 typedef type new_type[length];
964 type field_name[length];
967 typedef type new_type[length_type];
970 type field_name[length_type];
982 integer_type field_name:size; /* GNU/C bitfield */
992 The two following macros keep track of the size of a GNU/C structure without
993 padding at the end by placing HEADER_END as the last field. A one byte end field
994 is used for C90 compatibility (C99 flexible arrays could be used here). Note
995 that this does not affect the effective structure size, which should always be
996 calculated with the header_sizeof() helper.
998 #define HEADER_END char end_field
999 #define header_sizeof(type) offsetof(typeof(type), end_field)
1002 B. Stream Header Rationale
1004 An event stream is divided in contiguous event packets of variable size. These
1005 subdivisions allow the trace analyzer to perform a fast binary search by time
1006 within the stream (typically requiring to index only the event packet headers)
1007 without reading the whole stream. These subdivisions have a variable size to
1008 eliminate the need to transfer the event packet padding when partially filled
1009 event packets must be sent when streaming a trace for live viewing/analysis.
1010 An event packet can contain a certain amount of padding at the end. Dividing
1011 streams into event packets is also useful for network streaming over UDP and
1012 flight recorder mode tracing (a whole event packet can be swapped out of the
1013 buffer atomically for reading).
1015 The stream header is repeated at the beginning of each event packet to allow
1016 flexibility in terms of:
1018 - streaming support,
1019 - allowing arbitrary buffers to be discarded without making the trace
1021 - allow UDP packet loss handling by either dealing with missing event packet
1022 or asking for re-transmission.
1023 - transparently support flight recorder mode,
1024 - transparently support crash dump.
1026 The event stream header will therefore be referred to as the "event packet
1027 header" throughout the rest of this document.
1029 C. CTF Metadata Grammar
1032 * Common Trace Format (CTF) Metadata Grammar.
1034 * Inspired from the C99 grammar:
1035 * http://www.open-std.org/jtc1/sc22/wg14/www/docs/n1124.pdf (Annex A)
1037 * Specialized for CTF needs by including only constant and declarations from
1038 * C99 (excluding function declarations), and by adding support for variants,
1039 * sequences and CTF-specific specifiers.
1044 1.1) Lexical elements
1087 identifier identifier-nondigit
1090 identifier-nondigit:
1092 universal-character-name
1093 any other implementation-defined characters
1097 [a-zA-Z] /* regular expression */
1100 [0-9] /* regular expression */
1102 1.4) Universal character names
1104 universal-character-name:
1106 \U hex-quad hex-quad
1109 hexadecimal-digit hexadecimal-digit hexadecimal-digit hexadecimal-digit
1115 enumeration-constant
1119 decimal-constant integer-suffix-opt
1120 octal-constant integer-suffix-opt
1121 hexadecimal-constant integer-suffix-opt
1125 decimal-constant digit
1129 octal-constant octal-digit
1131 hexadecimal-constant:
1132 hexadecimal-prefix hexadecimal-digit
1133 hexadecimal-constant hexadecimal-digit
1143 unsigned-suffix long-suffix-opt
1144 unsigned-suffix long-long-suffix
1145 long-suffix unsigned-suffix-opt
1146 long-long-suffix unsigned-suffix-opt
1162 digit-sequence digit
1164 hexadecimal-digit-sequence:
1166 hexadecimal-digit-sequence hexadecimal-digit
1168 enumeration-constant:
1174 L' c-char-sequence '
1178 c-char-sequence c-char
1181 any member of source charset except single-quote ('), backslash
1182 (\), or new-line character.
1186 simple-escape-sequence
1187 octal-escape-sequence
1188 hexadecimal-escape-sequence
1189 universal-character-name
1191 simple-escape-sequence: one of
1192 \' \" \? \\ \a \b \f \n \r \t \v
1194 octal-escape-sequence:
1196 \ octal-digit octal-digit
1197 \ octal-digit octal-digit octal-digit
1199 hexadecimal-escape-sequence:
1200 \x hexadecimal-digit
1201 hexadecimal-escape-sequence hexadecimal-digit
1203 1.6) String literals
1206 " s-char-sequence-opt "
1207 L" s-char-sequence-opt "
1211 s-char-sequence s-char
1214 any member of source charset except double-quote ("), backslash
1215 (\), or new-line character.
1221 [ ] ( ) { } . -> * + - < > : ; ... = ,
1224 2) Phrase structure grammar
1230 ( unary-expression )
1234 postfix-expression [ unary-expression ]
1235 postfix-expression . identifier
1236 postfix-expressoin -> identifier
1240 unary-operator postfix-expression
1242 unary-operator: one of
1245 assignment-operator:
1248 type-assignment-operator:
1251 constant-expression:
1254 constant-expression-range:
1255 constant-expression ... constant-expression
1260 declaration-specifiers declarator-list-opt ;
1263 declaration-specifiers:
1264 storage-class-specifier declaration-specifiers-opt
1265 type-specifier declaration-specifiers-opt
1266 type-qualifier declaration-specifiers-opt
1270 declarator-list , declarator
1272 abstract-declarator-list:
1274 abstract-declarator-list , abstract-declarator
1276 storage-class-specifier:
1299 struct identifier-opt { struct-or-variant-declaration-list-opt }
1302 struct-or-variant-declaration-list:
1303 struct-or-variant-declaration
1304 struct-or-variant-declaration-list struct-or-variant-declaration
1306 struct-or-variant-declaration:
1307 specifier-qualifier-list struct-or-variant-declarator-list ;
1308 declaration-specifiers storage-class-specifier declaration-specifiers declarator-list ;
1309 typealias declaration-specifiers abstract-declarator-list := declaration-specifiers abstract-declarator-list ;
1310 typealias declaration-specifiers abstract-declarator-list := declarator-list ;
1312 specifier-qualifier-list:
1313 type-specifier specifier-qualifier-list-opt
1314 type-qualifier specifier-qualifier-list-opt
1316 struct-or-variant-declarator-list:
1317 struct-or-variant-declarator
1318 struct-or-variant-declarator-list , struct-or-variant-declarator
1320 struct-or-variant-declarator:
1322 declarator-opt : constant-expression
1325 variant identifier-opt variant-tag-opt { struct-or-variant-declaration-list }
1326 variant identifier variant-tag
1332 enum identifier-opt { enumerator-list }
1333 enum identifier-opt { enumerator-list , }
1335 enum identifier-opt < declaration-specifiers > { enumerator-list }
1336 enum identifier-opt < declaration-specifiers > { enumerator-list , }
1337 enum identifier < declaration-specifiers >
1341 enumerator-list , enumerator
1344 enumeration-constant
1345 enumeration-constant = constant-expression
1346 enumeration-constant = constant-expression-range
1352 pointer-opt direct-declarator
1357 direct-declarator [ type-specifier ]
1358 direct-declarator [ constant-expression ]
1360 abstract-declarator:
1361 pointer-opt direct-abstract-declarator
1363 direct-abstract-declarator:
1365 ( abstract-declarator )
1366 direct-abstract-declarator [ type-specifier ]
1367 direct-abstract-declarator [ constant-expression ]
1368 direct-abstract-declarator [ ]
1371 * type-qualifier-list-opt
1372 * type-qualifier-list-opt pointer
1374 type-qualifier-list:
1376 type-qualifier-list type-qualifier
1381 2.3) CTF-specific declarations
1384 event { ctf-assignment-expression-list-opt }
1385 stream { ctf-assignment-expression-list-opt }
1386 trace { ctf-assignment-expression-list-opt }
1387 typealias declaration-specifiers abstract-declarator-list := declaration-specifiers abstract-declarator-list ;
1388 typealias declaration-specifiers abstract-declarator-list := declarator-list ;
1391 floating_point { ctf-assignment-expression-list-opt }
1392 integer { ctf-assignment-expression-list-opt }
1393 string { ctf-assignment-expression-list-opt }
1395 ctf-assignment-expression-list:
1396 ctf-assignment-expression
1397 ctf-assignment-expression-list ; ctf-assignment-expression
1399 ctf-assignment-expression:
1400 unary-expression assignment-operator unary-expression
1401 unary-expression type-assignment-operator type-specifier
1402 declaration-specifiers storage-class-specifier declaration-specifiers declarator-list
1403 typealias declaration-specifiers abstract-declarator-list := declaration-specifiers abstract-declarator-list
1404 typealias declaration-specifiers abstract-declarator-list := declarator-list