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 If a numeric value is encountered between < >, it represents the integer type
275 size used to hold the enumeration, in bits.
277 enum name <integer_type OR size> {
278 somestring = start_value1 ... end_value1,
279 "other string" = start_value2 ... end_value2,
280 yet_another_string, /* will be assigned to end_value2 + 1 */
281 "some other string" = value,
285 If the values are omitted, the enumeration starts at 0 and increment of 1 for
296 Overlapping ranges within a single enumeration are implementation defined.
298 A nameless enumeration can be declared as a field type or as part of a typedef:
300 enum <integer_type> {
307 Compound are aggregation of type declarations. Compound types include
308 structures, variant, arrays, sequences, and strings.
312 Structures are aligned on the largest alignment required by basic types
313 contained within the structure. (This follows the ISO/C standard for structures)
315 Metadata representation of a named structure:
318 field_type field_name;
319 field_type field_name;
326 integer { /* Nameless type */
331 uint64_t second_field_name; /* Named type declared in the metadata */
334 The fields are placed in a sequence next to each other. They each possess a
335 field name, which is a unique identifier within the structure.
337 A nameless structure can be declared as a field type or as part of a typedef:
343 4.2.2 Variants (Discriminated/Tagged Unions)
345 A CTF variant is a selection between different types. A CTF variant must
346 always be defined within the scope of a structure or within fields
347 contained within a structure (defined recursively). A "tag" enumeration
348 field must appear in either the same lexical scope, prior to the variant
349 field (in field declaration order), in an uppermost lexical scope (see
350 Section 7.2.1), or in an uppermost dynamic scope (see Section 7.2.2).
351 The type selection is indicated by the mapping from the enumeration
352 value to the string used as variant type selector. The field to use as
353 tag is specified by the "tag_field", specified between "< >" after the
354 "variant" keyword for unnamed variants, and after "variant name" for
357 The alignment of the variant is the alignment of the type as selected by the tag
358 value for the specific instance of the variant. The alignment of the type
359 containing the variant is independent of the variant alignment. The size of the
360 variant is the size as selected by the tag value for the specific instance of
363 A named variant declaration followed by its definition within a structure
374 enum <integer_type or size> { sel1, sel2, sel3, ... } tag_field;
376 variant name <tag_field> v;
379 An unnamed variant definition within a structure is expressed by the following
383 enum <integer_type or size> { sel1, sel2, sel3, ... } tag_field;
385 variant <tag_field> {
393 Example of a named variant within a sequence that refers to a single tag field:
402 enum <uint2_t> { a, b, c } choice;
403 variant example <choice> v[unsigned int];
406 Example of an unnamed variant:
409 enum <uint2_t> { a, b, c, d } choice;
410 /* Unrelated fields can be added between the variant and its tag */
423 Example of an unnamed variant within an array:
426 enum <uint2_t> { a, b, c } choice;
434 Example of a variant type definition within a structure, where the defined type
435 is then declared within an array of structures. This variant refers to a tag
436 located in an upper lexical scope. This example clearly shows that a variant
437 type definition referring to the tag "x" uses the closest preceding field from
438 the lexical scope of the type definition.
441 enum <uint2_t> { a, b, c, d } x;
443 typedef variant <x> { /*
444 * "x" refers to the preceding "x" enumeration in the
445 * lexical scope of the type definition.
453 enum <int> { x, y, z } x; /* This enumeration is not used by "v". */
454 example_variant v; /*
455 * "v" uses the "enum <uint2_t> { a, b, c, d }"
463 Arrays are fixed-length. Their length is declared in the type declaration within
464 the metadata. They contain an array of "inner type" elements, which can refer to
465 any type not containing the type of the array being declared (no circular
466 dependency). The length is the number of elements in an array.
468 Metadata representation of a named array:
470 typedef elem_type name[length];
472 A nameless array can be declared as a field type within a structure, e.g.:
474 uint8_t field_name[10];
479 Sequences are dynamically-sized arrays. They start with an integer that specify
480 the length of the sequence, followed by an array of "inner type" elements.
481 The length is the number of elements in the sequence.
483 Metadata representation for a named sequence:
485 typedef elem_type name[length_type];
487 A nameless sequence can be declared as a field type, e.g.:
489 long field_name[int];
491 The length type follows the integer types specifications, and the sequence
492 elements follow the "array" specifications.
496 Strings are an array of bytes of variable size and are terminated by a '\0'
497 "NULL" character. Their encoding is described in the metadata. In absence of
498 encoding attribute information, the default encoding is UTF-8.
500 Metadata representation of a named string type:
503 encoding = UTF8 OR ASCII;
506 A nameless string type can be declared as a field type:
508 string field_name; /* Use default UTF8 encoding */
510 5. Event Packet Header
512 The event packet header consists of two part: one is mandatory and have a fixed
513 layout. The second part, the "event packet context", has its layout described in
516 - Aligned on page size. Fixed size. Fields either aligned or packed (depending
517 on the architecture preference).
518 No padding at the end of the event packet header. Native architecture byte
521 Fixed layout (event packet header):
523 - Magic number (CTF magic numbers: 0xC1FC1FC1 and its reverse endianness
524 representation: 0xC11FFCC1) It needs to have a non-symmetric bytewise
525 representation. Used to distinguish between big and little endian traces (this
526 information is determined by knowing the endianness of the architecture
527 reading the trace and comparing the magic number against its value and the
528 reverse, 0xC11FFCC1). This magic number specifies that we use the CTF metadata
529 description language described in this document. Different magic numbers
530 should be used for other metadata description languages.
531 - Trace UUID, used to ensure the event packet match the metadata used.
532 (note: we cannot use a metadata checksum because metadata can be appended to
533 while tracing is active)
534 - Stream ID, used as reference to stream description in metadata.
536 Metadata-defined layout (event packet context):
538 - Event packet content size (in bytes).
539 - Event packet size (in bytes, includes padding).
540 - Event packet content checksum (optional). Checksum excludes the event packet
542 - Per-stream event packet sequence count (to deal with UDP packet loss). The
543 number of significant sequence counter bits should also be present, so
544 wrap-arounds are deal with correctly.
545 - Timestamp at the beginning and timestamp at the end of the event packet.
546 Both timestamps are written in the packet header, but sampled respectively
547 while (or before) writing the first event and while (or after) writing the
548 last event in the packet. The inclusive range between these timestamps should
549 include all event timestamps assigned to events contained within the packet.
550 - Events discarded count
551 - Snapshot of a per-stream free-running counter, counting the number of
552 events discarded that were supposed to be written in the stream prior to
553 the first event in the event packet.
554 * Note: producer-consumer buffer full condition should fill the current
555 event packet with padding so we know exactly where events have been
557 - Lossless compression scheme used for the event packet content. Applied
558 directly to raw data. New types of compression can be added in following
559 versions of the format.
560 0: no compression scheme
564 - Cypher used for the event packet content. Applied after compression.
567 - Checksum scheme used for the event packet content. Applied after encryption.
573 5.1 Event Packet Header Fixed Layout Description
575 struct event_packet_header {
577 uint8_t trace_uuid[16];
581 5.2 Event Packet Context Description
583 Event packet context example. These are declared within the stream declaration
584 in the metadata. All these fields are optional except for "content_size" and
585 "packet_size", which must be present in the context.
587 An example event packet context type:
589 struct event_packet_context {
590 uint64_t timestamp_begin;
591 uint64_t timestamp_end;
593 uint32_t stream_packet_count;
594 uint32_t events_discarded;
596 uint32_t/uint16_t content_size;
597 uint32_t/uint16_t packet_size;
598 uint8_t stream_packet_count_bits; /* Significant counter bits */
599 uint8_t compression_scheme;
600 uint8_t encryption_scheme;
601 uint8_t checksum_scheme;
607 The overall structure of an event is:
609 1 - Stream Packet Context (as specified by the stream metadata)
610 2 - Event Header (as specified by the stream metadata)
611 3 - Stream Event Context (as specified by the stream metadata)
612 4 - Event Context (as specified by the event metadata)
613 5 - Event Payload (as specified by the event metadata)
615 This structure defines an implicit dynamic scoping, where variants
616 located in inner structures (those with a higher number in the listing
617 above) can refer to the fields of outer structures (with lower number in
618 the listing above). See Section 7.2 Metadata Scopes for more detail.
622 Event headers can be described within the metadata. We hereby propose, as an
623 example, two types of events headers. Type 1 accommodates streams with less than
624 31 event IDs. Type 2 accommodates streams with 31 or more event IDs.
626 One major factor can vary between streams: the number of event IDs assigned to
627 a stream. Luckily, this information tends to stay relatively constant (modulo
628 event registration while trace is being recorded), so we can specify different
629 representations for streams containing few event IDs and streams containing
630 many event IDs, so we end up representing the event ID and timestamp as densely
631 as possible in each case.
633 The header is extended in the rare occasions where the information cannot be
634 represented in the ranges available in the standard event header. They are also
635 used in the rare occasions where the data required for a field could not be
636 collected: the flag corresponding to the missing field within the missing_fields
637 array is then set to 1.
639 Types uintX_t represent an X-bit unsigned integer.
642 6.1.1 Type 1 - Few event IDs
644 - Aligned on 32-bit (or 8-bit if byte-packed, depending on the architecture
646 - Native architecture byte ordering.
647 - For "compact" selection
648 - Fixed size: 32 bits.
649 - For "extended" selection
650 - Size depends on the architecture and variant alignment.
652 struct event_header_1 {
655 * id 31 is reserved to indicate an extended header.
657 enum <uint5_t> { compact = 0 ... 30, extended = 31 } id;
663 uint32_t id; /* 32-bit event IDs */
664 uint64_t timestamp; /* 64-bit timestamps */
670 6.1.2 Type 2 - Many event IDs
672 - Aligned on 16-bit (or 8-bit if byte-packed, depending on the architecture
674 - Native architecture byte ordering.
675 - For "compact" selection
676 - Size depends on the architecture and variant alignment.
677 - For "extended" selection
678 - Size depends on the architecture and variant alignment.
680 struct event_header_2 {
682 * id: range: 0 - 65534.
683 * id 65535 is reserved to indicate an extended header.
685 enum <uint16_t> { compact = 0 ... 65534, extended = 65535 } id;
691 uint32_t id; /* 32-bit event IDs */
692 uint64_t timestamp; /* 64-bit timestamps */
700 The event context contains information relative to the current event. The choice
701 and meaning of this information is specified by the metadata "stream" and
702 "event" information. The "stream" context is applied to all events within the
703 stream. The "stream" context structure follows the event header. The "event"
704 context is applied to specific events. Its structure follows the "stream"
707 An example of stream-level event context is to save the event payload size with
708 each event, or to save the current PID with each event. These are declared
709 within the stream declaration within the metadata:
717 uint16_t payload_size;
722 An example of event-specific event context is to declare a bitmap of missing
723 fields, only appended after the stream event context if the extended event
724 header is selected. NR_FIELDS is the number of fields within the event (a
732 uint1_t missing_fields[NR_FIELDS]; /* missing event fields bitmap */
741 An event payload contains fields specific to a given event type. The fields
742 belonging to an event type are described in the event-specific metadata
743 within a structure type.
747 No padding at the end of the event payload. This differs from the ISO/C standard
748 for structures, but follows the CTF standard for structures. In a trace, even
749 though it makes sense to align the beginning of a structure, it really makes no
750 sense to add padding at the end of the structure, because structures are usually
751 not followed by a structure of the same type.
753 This trick can be done by adding a zero-length "end" field at the end of the C
754 structures, and by using the offset of this field rather than using sizeof()
755 when calculating the size of a structure (see Appendix "A. Helper macros").
759 The event payload is aligned on the largest alignment required by types
760 contained within the payload. (This follows the ISO/C standard for structures)
765 The meta-data is located in a stream identified by its name: "metadata".
766 It is made of "event packets", which each start with an event packet
767 header. The event type within the metadata stream have no event header
768 nor event context. Each event only contains a null-terminated "string"
769 payload, which is a metadata description entry. The events are packed
770 one next to another. Each event packet start with an event packet
771 header, which contains, amongst other fields, the magic number and trace
772 UUID. In the event packet header, the trace UUID is represented as an
773 array of bytes. Within the string-based metadata description, the trace
774 UUID is represented as a string of hexadecimal digits and dashes "-".
776 The metadata can be parsed by reading through the metadata strings,
777 skipping null-characters. Type names are made of a single identifier,
778 and can be surrounded by prefix/postfix. Text contained within "/*" and
779 "*/", as well as within "//" and end of line, are treated as comments.
780 Boolean values can be represented as true, TRUE, or 1 for true, and
781 false, FALSE, or 0 for false.
784 7.1 Declaration vs Definition
786 A declaration associates a layout to a type, without specifying where
787 this type is located in the event structure hierarchy (see Section 6).
788 This therefore includes typedef, typealias, as well as all type
789 specifiers. In certain circumstances (typedef, structure field and
790 variant field), a declaration is followed by a declarator, which specify
791 the newly defined type name (for typedef), or the field name (for
792 declarations located within structure and variants). Array and sequence,
793 declared with square brackets ("[" "]"), are part of the declarator,
794 similarly to C99. The enumeration type specifier and variant tag name
795 (both specified with "<" ">") are part of the type specifier.
797 A definition associates a type to a location in the event structure
798 hierarchy (see Section 6). This association is denoted by ":=", as shown
804 CTF metadata uses two different types of scoping: a lexical scope is
805 used for declarations and type definitions, and a dynamic scope is used
806 for variants references to tag fields.
810 Each of "trace", "stream", "event", "struct" and "variant" have their own
811 nestable declaration scope, within which types can be declared using "typedef"
812 and "typealias". A root declaration scope also contains all declarations
813 located outside of any of the aforementioned declarations. An inner
814 declaration scope can refer to type declared within its container
815 lexical scope prior to the inner declaration scope. Redefinition of a
816 typedef or typealias is not valid, although hiding an upper scope
817 typedef or typealias is allowed within a sub-scope.
821 A dynamic scope consists in the lexical scope augmented with the
822 implicit event structure definition hierarchy presented at Section 6.
823 The dynamic scope is only used for variant tag definitions. It is used
824 at definition time to look up the location of the tag field associated
827 Therefore, variants in lower levels in the dynamic scope (e.g. event
828 context) can refer to a tag field located in upper levels (e.g. in the
829 event header) by specifying, in this case, the associated tag with
830 <header.field_name>. This allows, for instance, the event context to
831 define a variant referring to the "id" field of the event header as
834 The target dynamic scope must be specified explicitly when referring to
835 a field outside of the local static scope. The dynamic scope prefixes
838 - Stream Packet Context: <stream.packet.context. >,
839 - Event Header: <stream.event.header. >,
840 - Stream Event Context: <stream.event.context. >,
841 - Event Context: <event.context. >,
842 - Event Payload: <event.fields. >.
844 Multiple declarations of the same field name within a single scope is
845 not valid. It is however valid to re-use the same field name in
846 different scopes. There is no possible conflict, because the dynamic
847 scope must be specified when a variant refers to a tag field located in
848 a different dynamic scope.
850 The information available in the dynamic scopes can be thought of as the
851 current tracing context. At trace production, information about the
852 current context is saved into the specified scope field levels. At trace
853 consumption, for each event, the current trace context is therefore
854 readable by accessing the upper dynamic scopes.
857 7.3 Metadata Examples
859 The grammar representing the CTF metadata is presented in
860 Appendix C. CTF Metadata Grammar. This section presents a rather ligher
861 reading that consists in examples of CTF metadata, with template values:
864 major = value; /* Trace format version */
866 uuid = "aaaaaaaa-aaaa-aaaa-aaaa-aaaaaaaaaaaa"; /* Trace UUID */
872 /* Type 1 - Few event IDs; Type 2 - Many event IDs. See section 6.1. */
873 event.header := event_header_1 OR event_header_2;
874 event.context := struct {
877 packet.context := struct {
884 id = value; /* Numeric identifier within the stream */
894 /* More detail on types in section 4. Types */
899 * Type declarations behave similarly to the C standard.
902 typedef aliased_type_prefix aliased_type new_type aliased_type_postfix;
904 /* e.g.: typedef struct example new_type_name[10]; */
909 * The "typealias" declaration can be used to give a name (including
910 * prefix/postfix) to a type. It should also be used to map basic C types
911 * (float, int, unsigned long, ...) to a CTF type. Typealias is a superset of
912 * "typedef": it also allows assignment of a simple variable identifier to a
916 typealias type_class {
918 } : new_type_prefix new_type new_type_postfix;
922 * typealias integer {
928 * typealias integer {
943 enum name <integer_type or size> {
949 * Unnamed types, contained within compound type fields, typedef or typealias.
960 enum <integer_type or size> {
964 typedef type new_type[length];
967 type field_name[length];
970 typedef type new_type[length_type];
973 type field_name[length_type];
985 integer_type field_name:size; /* GNU/C bitfield */
995 The two following macros keep track of the size of a GNU/C structure without
996 padding at the end by placing HEADER_END as the last field. A one byte end field
997 is used for C90 compatibility (C99 flexible arrays could be used here). Note
998 that this does not affect the effective structure size, which should always be
999 calculated with the header_sizeof() helper.
1001 #define HEADER_END char end_field
1002 #define header_sizeof(type) offsetof(typeof(type), end_field)
1005 B. Stream Header Rationale
1007 An event stream is divided in contiguous event packets of variable size. These
1008 subdivisions allow the trace analyzer to perform a fast binary search by time
1009 within the stream (typically requiring to index only the event packet headers)
1010 without reading the whole stream. These subdivisions have a variable size to
1011 eliminate the need to transfer the event packet padding when partially filled
1012 event packets must be sent when streaming a trace for live viewing/analysis.
1013 An event packet can contain a certain amount of padding at the end. Dividing
1014 streams into event packets is also useful for network streaming over UDP and
1015 flight recorder mode tracing (a whole event packet can be swapped out of the
1016 buffer atomically for reading).
1018 The stream header is repeated at the beginning of each event packet to allow
1019 flexibility in terms of:
1021 - streaming support,
1022 - allowing arbitrary buffers to be discarded without making the trace
1024 - allow UDP packet loss handling by either dealing with missing event packet
1025 or asking for re-transmission.
1026 - transparently support flight recorder mode,
1027 - transparently support crash dump.
1029 The event stream header will therefore be referred to as the "event packet
1030 header" throughout the rest of this document.
1032 C. CTF Metadata Grammar
1035 * Common Trace Format (CTF) Metadata Grammar.
1037 * Inspired from the C99 grammar:
1038 * http://www.open-std.org/jtc1/sc22/wg14/www/docs/n1124.pdf (Annex A)
1040 * Specialized for CTF needs by including only constant and declarations from
1041 * C99 (excluding function declarations), and by adding support for variants,
1042 * sequences and CTF-specific specifiers.
1047 1.1) Lexical elements
1090 identifier identifier-nondigit
1093 identifier-nondigit:
1095 universal-character-name
1096 any other implementation-defined characters
1100 [a-zA-Z] /* regular expression */
1103 [0-9] /* regular expression */
1105 1.4) Universal character names
1107 universal-character-name:
1109 \U hex-quad hex-quad
1112 hexadecimal-digit hexadecimal-digit hexadecimal-digit hexadecimal-digit
1118 enumeration-constant
1122 decimal-constant integer-suffix-opt
1123 octal-constant integer-suffix-opt
1124 hexadecimal-constant integer-suffix-opt
1128 decimal-constant digit
1132 octal-constant octal-digit
1134 hexadecimal-constant:
1135 hexadecimal-prefix hexadecimal-digit
1136 hexadecimal-constant hexadecimal-digit
1146 unsigned-suffix long-suffix-opt
1147 unsigned-suffix long-long-suffix
1148 long-suffix unsigned-suffix-opt
1149 long-long-suffix unsigned-suffix-opt
1165 digit-sequence digit
1167 hexadecimal-digit-sequence:
1169 hexadecimal-digit-sequence hexadecimal-digit
1171 enumeration-constant:
1177 L' c-char-sequence '
1181 c-char-sequence c-char
1184 any member of source charset except single-quote ('), backslash
1185 (\), or new-line character.
1189 simple-escape-sequence
1190 octal-escape-sequence
1191 hexadecimal-escape-sequence
1192 universal-character-name
1194 simple-escape-sequence: one of
1195 \' \" \? \\ \a \b \f \n \r \t \v
1197 octal-escape-sequence:
1199 \ octal-digit octal-digit
1200 \ octal-digit octal-digit octal-digit
1202 hexadecimal-escape-sequence:
1203 \x hexadecimal-digit
1204 hexadecimal-escape-sequence hexadecimal-digit
1206 1.6) String literals
1209 " s-char-sequence-opt "
1210 L" s-char-sequence-opt "
1214 s-char-sequence s-char
1217 any member of source charset except double-quote ("), backslash
1218 (\), or new-line character.
1224 [ ] ( ) { } . -> * + - < > : ; ... = ,
1227 2) Phrase structure grammar
1233 ( unary-expression )
1237 postfix-expression [ unary-expression ]
1238 postfix-expression . identifier
1239 postfix-expressoin -> identifier
1243 unary-operator postfix-expression
1245 unary-operator: one of
1248 assignment-operator:
1251 type-assignment-operator:
1254 constant-expression:
1257 constant-expression-range:
1258 constant-expression ... constant-expression
1263 declaration-specifiers declarator-list-opt ;
1266 declaration-specifiers:
1267 storage-class-specifier declaration-specifiers-opt
1268 type-specifier declaration-specifiers-opt
1269 type-qualifier declaration-specifiers-opt
1273 declarator-list , declarator
1275 abstract-declarator-list:
1277 abstract-declarator-list , abstract-declarator
1279 storage-class-specifier:
1302 struct identifier-opt { struct-or-variant-declaration-list-opt }
1305 struct-or-variant-declaration-list:
1306 struct-or-variant-declaration
1307 struct-or-variant-declaration-list struct-or-variant-declaration
1309 struct-or-variant-declaration:
1310 specifier-qualifier-list struct-or-variant-declarator-list ;
1311 declaration-specifiers storage-class-specifier declaration-specifiers declarator-list ;
1312 typealias declaration-specifiers abstract-declarator-list : declaration-specifiers abstract-declarator-list ;
1313 typealias declaration-specifiers abstract-declarator-list : declarator-list ;
1315 specifier-qualifier-list:
1316 type-specifier specifier-qualifier-list-opt
1317 type-qualifier specifier-qualifier-list-opt
1319 struct-or-variant-declarator-list:
1320 struct-or-variant-declarator
1321 struct-or-variant-declarator-list , struct-or-variant-declarator
1323 struct-or-variant-declarator:
1325 declarator-opt : constant-expression
1328 variant identifier-opt variant-tag-opt { struct-or-variant-declaration-list }
1329 variant identifier variant-tag
1335 enum identifier-opt { enumerator-list }
1336 enum identifier-opt { enumerator-list , }
1338 enum identifier-opt < declaration-specifiers > { enumerator-list }
1339 enum identifier-opt < declaration-specifiers > { enumerator-list , }
1340 enum identifier < declaration-specifiers >
1341 enum identifier-opt < integer-constant > { enumerator-list }
1342 enum identifier-opt < integer-constant > { enumerator-list , }
1343 enum identifier < integer-constant >
1347 enumerator-list , enumerator
1350 enumeration-constant
1351 enumeration-constant = constant-expression
1352 enumeration-constant = constant-expression-range
1358 pointer-opt direct-declarator
1363 direct-declarator [ type-specifier ]
1364 direct-declarator [ constant-expression ]
1366 abstract-declarator:
1367 pointer-opt direct-abstract-declarator
1369 direct-abstract-declarator:
1371 ( abstract-declarator )
1372 direct-abstract-declarator [ type-specifier ]
1373 direct-abstract-declarator [ constant-expression ]
1374 direct-abstract-declarator [ ]
1377 * type-qualifier-list-opt
1378 * type-qualifier-list-opt pointer
1380 type-qualifier-list:
1382 type-qualifier-list type-qualifier
1387 2.3) CTF-specific declarations
1390 event { ctf-assignment-expression-list-opt }
1391 stream { ctf-assignment-expression-list-opt }
1392 trace { ctf-assignment-expression-list-opt }
1393 typealias declaration-specifiers abstract-declarator-list : declaration-specifiers abstract-declarator-list ;
1394 typealias declaration-specifiers abstract-declarator-list : declarator-list ;
1397 floating_point { ctf-assignment-expression-list-opt }
1398 integer { ctf-assignment-expression-list-opt }
1399 string { ctf-assignment-expression-list-opt }
1401 ctf-assignment-expression-list:
1402 ctf-assignment-expression
1403 ctf-assignment-expression-list ; ctf-assignment-expression
1405 ctf-assignment-expression:
1406 unary-expression assignment-operator unary-expression
1407 unary-expression type-assignment-operator type-specifier
1408 declaration-specifiers storage-class-specifier declaration-specifiers declarator-list
1409 typealias declaration-specifiers abstract-declarator-list : declaration-specifiers abstract-declarator-list
1410 typealias declaration-specifiers abstract-declarator-list : declarator-list