1 Common Trace Format (CTF) Specification (v1.8.2)
3 Mathieu Desnoyers, EfficiOS Inc.
5 The goal of the present document is to specify a trace format that suits the
6 needs of the embedded, telecom, high-performance and kernel communities. It is
7 based on the Common Trace Format Requirements (v1.4) document. It is designed to
8 allow traces to be natively generated by the Linux kernel, Linux user-space
9 applications written in C/C++, and hardware components. One major element of
10 CTF is the Trace Stream Description Language (TSDL) which flexibility
11 enables description of various binary trace stream layouts.
13 The latest version of this document can be found at:
15 git tree: git://git.efficios.com/ctf.git
16 gitweb: http://git.efficios.com/?p=ctf.git
18 A reference implementation of a library to read and write this trace format is
19 being implemented within the BabelTrace project, a converter between trace
20 formats. The development tree is available at:
22 git tree: git://git.efficios.com/babeltrace.git
23 gitweb: http://git.efficios.com/?p=babeltrace.git
25 The CE Workgroup of the Linux Foundation, Ericsson, and EfficiOS have
31 1. Preliminary definitions
32 2. High-level representation of a trace
36 4.1.1 Type inheritance
46 4.2.2 Variants (Discriminated/Tagged Unions)
50 5. Event Packet Header
51 5.1 Event Packet Header Description
52 5.2 Event Packet Context Description
55 6.1.1 Type 1 - Few event IDs
56 6.1.2 Type 2 - Many event IDs
57 6.2 Stream Event Context and Event Context
61 7. Trace Stream Description Language (TSDL)
63 7.2 Declaration vs Definition
66 7.3.2 Static and Dynamic Scopes
71 1. Preliminary definitions
73 - Event Trace: An ordered sequence of events.
74 - Event Stream: An ordered sequence of events, containing a subset of the
76 - Event Packet: A sequence of physically contiguous events within an event
78 - Event: This is the basic entry in a trace. (aka: a trace record).
79 - An event identifier (ID) relates to the class (a type) of event within
81 e.g. event: irq_entry.
82 - An event (or event record) relates to a specific instance of an event
84 e.g. event: irq_entry, at time X, on CPU Y
85 - Source Architecture: Architecture writing the trace.
86 - Reader Architecture: Architecture reading the trace.
89 2. High-level representation of a trace
91 A trace is divided into multiple event streams. Each event stream contains a
92 subset of the trace event types.
94 The final output of the trace, after its generation and optional transport over
95 the network, is expected to be either on permanent or temporary storage in a
96 virtual file system. Because each event stream is appended to while a trace is
97 being recorded, each is associated with a distinct set of files for
98 output. Therefore, a stored trace can be represented as a directory
99 containing zero, one or more files per stream.
101 Meta-data description associated with the trace contains information on
102 trace event types expressed in the Trace Stream Description Language
103 (TSDL). This language describes:
107 - Per-trace event header description.
108 - Per-stream event header description.
109 - Per-stream event context description.
111 - Event type to stream mapping.
112 - Event type to name mapping.
113 - Event type to ID mapping.
114 - Event context description.
115 - Event fields description.
120 An event stream can be divided into contiguous event packets of variable
121 size. An event packet can contain a certain amount of padding at the
122 end. The stream header is repeated at the beginning of each event
123 packet. The rationale for the event stream design choices is explained
124 in Appendix B. Stream Header Rationale.
126 The event stream header will therefore be referred to as the "event packet
127 header" throughout the rest of this document.
132 Types are organized as type classes. Each type class belong to either of two
133 kind of types: basic types or compound types.
137 A basic type is a scalar type, as described in this section. It includes
138 integers, GNU/C bitfields, enumerations, and floating point values.
140 4.1.1 Type inheritance
142 Type specifications can be inherited to allow deriving types from a
143 type class. For example, see the uint32_t named type derived from the "integer"
144 type class below ("Integers" section). Types have a precise binary
145 representation in the trace. A type class has methods to read and write these
146 types, but must be derived into a type to be usable in an event field.
150 We define "byte-packed" types as aligned on the byte size, namely 8-bit.
151 We define "bit-packed" types as following on the next bit, as defined by the
154 Each basic type must specify its alignment, in bits. Examples of
155 possible alignments are: bit-packed (align = 1), byte-packed (align =
156 8), or word-aligned (e.g. align = 32 or align = 64). The choice depends
157 on the architecture preference and compactness vs performance trade-offs
158 of the implementation. Architectures providing fast unaligned write
159 byte-packed basic types to save space, aligning each type on byte
160 boundaries (8-bit). Architectures with slow unaligned writes align types
161 on specific alignment values. If no specific alignment is declared for a
162 type, it is assumed to be bit-packed for integers with size not multiple
163 of 8 bits and for gcc bitfields. All other basic types are byte-packed
164 by default. It is however recommended to always specify the alignment
165 explicitly. Alignment values must be power of two. Compound types are
166 aligned as specified in their individual specification.
168 The base offset used for field alignment is the start of the packet
169 containing the field. For instance, a field aligned on 32-bit needs to
170 be at an offset multiple of 32-bit from the start of the packet that
173 TSDL meta-data attribute representation of a specific alignment:
175 align = value; /* value in bits */
179 By default, byte order of a basic type is the byte order described in
180 the trace description. It can be overridden by specifying a
181 "byte_order" attribute for a basic type. Typical use-case is to specify
182 the network byte order (big endian: "be") to save data captured from the
183 network into the trace without conversion.
185 TSDL meta-data representation:
187 byte_order = native OR network OR be OR le; /* network and be are aliases */
189 The "native" keyword selects the byte order described in the trace
190 description. The "network" byte order is an alias for big endian.
192 Even though the trace description section is not per se a type, for sake
193 of clarity, it should be noted that "native" and "network" byte orders
194 are only allowed within type declaration. The byte_order specified in
195 the trace description section only accepts "be" or "le" values.
199 Type size, in bits, for integers and floats is that returned by "sizeof()" in C
200 multiplied by CHAR_BIT.
201 We require the size of "char" and "unsigned char" types (CHAR_BIT) to be fixed
202 to 8 bits for cross-endianness compatibility.
204 TSDL meta-data representation:
206 size = value; (value is in bits)
210 Signed integers are represented in two-complement. Integer alignment,
211 size, signedness and byte ordering are defined in the TSDL meta-data.
212 Integers aligned on byte size (8-bit) and with length multiple of byte
213 size (8-bit) correspond to the C99 standard integers. In addition,
214 integers with alignment and/or size that are _not_ a multiple of the
215 byte size are permitted; these correspond to the C99 standard bitfields,
216 with the added specification that the CTF integer bitfields have a fixed
217 binary representation. A MIT-licensed reference implementation of the
218 CTF portable bitfields is available at:
220 http://git.efficios.com/?p=babeltrace.git;a=blob;f=include/babeltrace/bitfield.h
222 Binary representation of integers:
224 - On little and big endian:
225 - Within a byte, high bits correspond to an integer high bits, and low bits
226 correspond to low bits.
228 - Integer across multiple bytes are placed from the less significant to the
230 - Consecutive integers are placed from lower bits to higher bits (even within
233 - Integer across multiple bytes are placed from the most significant to the
235 - Consecutive integers are placed from higher bits to lower bits (even within
238 This binary representation is derived from the bitfield implementation in GCC
239 for little and big endian. However, contrary to what GCC does, integers can
240 cross units boundaries (no padding is required). Padding can be explicitly
241 added (see 4.1.6 GNU/C bitfields) to follow the GCC layout if needed.
243 TSDL meta-data representation:
246 signed = true OR false; /* default false */
247 byte_order = native OR network OR be OR le; /* default native */
248 size = value; /* value in bits, no default */
249 align = value; /* value in bits */
250 /* based used for pretty-printing output, default: decimal. */
251 base = decimal OR dec OR d OR i OR u OR 10 OR hexadecimal OR hex OR x OR X OR p OR 16
252 OR octal OR oct OR o OR 8 OR binary OR b OR 2;
253 /* character encoding, default: none */
254 encoding = none or UTF8 or ASCII;
257 Example of type inheritance (creation of a uint32_t named type):
265 Definition of a named 5-bit signed bitfield:
273 The character encoding field can be used to specify that the integer
274 must be printed as a text character when read. e.g.:
284 4.1.6 GNU/C bitfields
286 The GNU/C bitfields follow closely the integer representation, with a
287 particularity on alignment: if a bitfield cannot fit in the current unit, the
288 unit is padded and the bitfield starts at the following unit. The unit size is
289 defined by the size of the type "unit_type".
291 TSDL meta-data representation:
295 As an example, the following structure declared in C compiled by GCC:
302 The example structure is aligned on the largest element (short). The second
303 bitfield would be aligned on the next unit boundary, because it would not fit in
308 The floating point values byte ordering is defined in the TSDL meta-data.
310 Floating point values follow the IEEE 754-2008 standard interchange formats.
311 Description of the floating point values include the exponent and mantissa size
312 in bits. Some requirements are imposed on the floating point values:
314 - FLT_RADIX must be 2.
315 - mant_dig is the number of digits represented in the mantissa. It is specified
316 by the ISO C99 standard, section 5.2.4, as FLT_MANT_DIG, DBL_MANT_DIG and
317 LDBL_MANT_DIG as defined by <float.h>.
318 - exp_dig is the number of digits represented in the exponent. Given that
319 mant_dig is one bit more than its actual size in bits (leading 1 is not
320 needed) and also given that the sign bit always takes one bit, exp_dig can be
323 - sizeof(float) * CHAR_BIT - FLT_MANT_DIG
324 - sizeof(double) * CHAR_BIT - DBL_MANT_DIG
325 - sizeof(long double) * CHAR_BIT - LDBL_MANT_DIG
327 TSDL meta-data representation:
332 byte_order = native OR network OR be OR le;
336 Example of type inheritance:
338 typealias floating_point {
339 exp_dig = 8; /* sizeof(float) * CHAR_BIT - FLT_MANT_DIG */
340 mant_dig = 24; /* FLT_MANT_DIG */
345 TODO: define NaN, +inf, -inf behavior.
347 Bit-packed, byte-packed or larger alignments can be used for floating
348 point values, similarly to integers.
352 Enumerations are a mapping between an integer type and a table of strings. The
353 numerical representation of the enumeration follows the integer type specified
354 by the meta-data. The enumeration mapping table is detailed in the enumeration
355 description within the meta-data. The mapping table maps inclusive value
356 ranges (or single values) to strings. Instead of being limited to simple
357 "value -> string" mappings, these enumerations map
358 "[ start_value ... end_value ] -> string", which map inclusive ranges of
359 values to strings. An enumeration from the C language can be represented in
360 this format by having the same start_value and end_value for each element, which
361 is in fact a range of size 1. This single-value range is supported without
362 repeating the start and end values with the value = string declaration.
363 Enumerations need to contain at least one entry.
365 enum name : integer_type {
366 somestring = start_value1 ... end_value1,
367 "other string" = start_value2 ... end_value2,
368 yet_another_string, /* will be assigned to end_value2 + 1 */
369 "some other string" = value,
373 If the values are omitted, the enumeration starts at 0 and increment of 1 for
374 each entry. An entry with omitted value that follows a range entry takes
375 as value the end_value of the previous range + 1:
377 enum name : unsigned int {
385 Overlapping ranges within a single enumeration are implementation defined.
387 A nameless enumeration can be declared as a field type or as part of a typedef:
389 enum : integer_type {
393 Enumerations omitting the container type ": integer_type" use the "int"
394 type (for compatibility with C99). The "int" type must be previously
397 typealias integer { size = 32; align = 32; signed = true; } := int;
406 Compound are aggregation of type declarations. Compound types include
407 structures, variant, arrays, sequences, and strings.
411 Structures are aligned on the largest alignment required by basic types
412 contained within the structure. (This follows the ISO/C standard for structures)
414 TSDL meta-data representation of a named structure:
417 field_type field_name;
418 field_type field_name;
425 integer { /* Nameless type */
430 uint64_t second_field_name; /* Named type declared in the meta-data */
433 The fields are placed in a sequence next to each other. They each
434 possess a field name, which is a unique identifier within the structure.
435 The identifier is not allowed to use any reserved keyword
436 (see Section C.1.2). Replacing reserved keywords with
437 underscore-prefixed field names is recommended. Fields starting with an
438 underscore should have their leading underscore removed by the CTF trace
441 A nameless structure can be declared as a field type or as part of a typedef:
447 Alignment for a structure compound type can be forced to a minimum value
448 by adding an "align" specifier after the declaration of a structure
449 body. This attribute is read as: align(value). The value is specified in
450 bits. The structure will be aligned on the maximum value between this
451 attribute and the alignment required by the basic types contained within
458 4.2.2 Variants (Discriminated/Tagged Unions)
460 A CTF variant is a selection between different types. A CTF variant must
461 always be defined within the scope of a structure or within fields
462 contained within a structure (defined recursively). A "tag" enumeration
463 field must appear in either the same static scope, prior to the variant
464 field (in field declaration order), in an upper static scope , or in an
465 upper dynamic scope (see Section 7.3.2). The type selection is indicated
466 by the mapping from the enumeration value to the string used as variant
467 type selector. The field to use as tag is specified by the "tag_field",
468 specified between "< >" after the "variant" keyword for unnamed
469 variants, and after "variant name" for named variants.
471 The alignment of the variant is the alignment of the type as selected by
472 the tag value for the specific instance of the variant. The size of the
473 variant is the size as selected by the tag value for the specific
474 instance of the variant.
476 The alignment of the type containing the variant is independent of the
477 variant alignment. For instance, if a structure contains two fields, a
478 32-bit integer, aligned on 32 bits, and a variant, which contains two
479 choices: either a 32-bit field, aligned on 32 bits, or a 64-bit field,
480 aligned on 64 bits, the alignment of the outmost structure will be
481 32-bit (the alignment of its largest field, disregarding the alignment
482 of the variant). The alignment of the variant will depend on the
483 selector: if the variant's 32-bit field is selected, its alignment will
484 be 32-bit, or 64-bit otherwise. It is important to note that variants
485 are specifically tailored for compactness in a stream. Therefore, the
486 relative offsets of compound type fields can vary depending on
487 the offset at which the compound type starts if it contains a variant
488 that itself contains a type with alignment larger than the largest field
489 contained within the compound type. This is caused by the fact that the
490 compound type may contain the enumeration that select the variant's
491 choice, and therefore the alignment to be applied to the compound type
492 cannot be determined before encountering the enumeration.
494 Each variant type selector possess a field name, which is a unique
495 identifier within the variant. The identifier is not allowed to use any
496 reserved keyword (see Section C.1.2). Replacing reserved keywords with
497 underscore-prefixed field names is recommended. Fields starting with an
498 underscore should have their leading underscore removed by the CTF trace
502 A named variant declaration followed by its definition within a structure
513 enum : integer_type { sel1, sel2, sel3, ... } tag_field;
515 variant name <tag_field> v;
518 An unnamed variant definition within a structure is expressed by the following
522 enum : integer_type { sel1, sel2, sel3, ... } tag_field;
524 variant <tag_field> {
532 Example of a named variant within a sequence that refers to a single tag field:
541 enum : uint2_t { a, b, c } choice;
543 variant example <choice> v[seqlen];
546 Example of an unnamed variant:
549 enum : uint2_t { a, b, c, d } choice;
550 /* Unrelated fields can be added between the variant and its tag */
563 Example of an unnamed variant within an array:
566 enum : uint2_t { a, b, c } choice;
574 Example of a variant type definition within a structure, where the defined type
575 is then declared within an array of structures. This variant refers to a tag
576 located in an upper static scope. This example clearly shows that a variant
577 type definition referring to the tag "x" uses the closest preceding field from
578 the static scope of the type definition.
581 enum : uint2_t { a, b, c, d } x;
583 typedef variant <x> { /*
584 * "x" refers to the preceding "x" enumeration in the
585 * static scope of the type definition.
593 enum : int { x, y, z } x; /* This enumeration is not used by "v". */
594 example_variant v; /*
595 * "v" uses the "enum : uint2_t { a, b, c, d }"
603 Arrays are fixed-length. Their length is declared in the type
604 declaration within the meta-data. They contain an array of "inner type"
605 elements, which can refer to any type not containing the type of the
606 array being declared (no circular dependency). The length is the number
607 of elements in an array.
609 TSDL meta-data representation of a named array:
611 typedef elem_type name[length];
613 A nameless array can be declared as a field type within a structure, e.g.:
615 uint8_t field_name[10];
617 Arrays are always aligned on their element alignment requirement.
621 Sequences are dynamically-sized arrays. They refer to a "length"
622 unsigned integer field, which must appear in either the same static scope,
623 prior to the sequence field (in field declaration order), in an upper
624 static scope, or in an upper dynamic scope (see Section 7.3.2). This
625 length field represents the number of elements in the sequence. The
626 sequence per se is an array of "inner type" elements.
628 TSDL meta-data representation for a sequence type definition:
631 unsigned int length_field;
632 typedef elem_type typename[length_field];
633 typename seq_field_name;
636 A sequence can also be declared as a field type, e.g.:
639 unsigned int length_field;
640 long seq_field_name[length_field];
643 Multiple sequences can refer to the same length field, and these length
644 fields can be in a different upper dynamic scope:
646 e.g., assuming the stream.event.header defines:
651 event.header := struct {
660 long seq_a[stream.event.header.seq_len];
661 char seq_b[stream.event.header.seq_len];
665 The sequence elements follow the "array" specifications.
669 Strings are an array of bytes of variable size and are terminated by a '\0'
670 "NULL" character. Their encoding is described in the TSDL meta-data. In
671 absence of encoding attribute information, the default encoding is
674 TSDL meta-data representation of a named string type:
677 encoding = UTF8 OR ASCII;
680 A nameless string type can be declared as a field type:
682 string field_name; /* Use default UTF8 encoding */
684 Strings are always aligned on byte size.
686 5. Event Packet Header
688 The event packet header consists of two parts: the "event packet header"
689 is the same for all streams of a trace. The second part, the "event
690 packet context", is described on a per-stream basis. Both are described
691 in the TSDL meta-data.
693 Event packet header (all fields are optional, specified by TSDL meta-data):
695 - Magic number (CTF magic number: 0xC1FC1FC1) specifies that this is a
696 CTF packet. This magic number is optional, but when present, it should
697 come at the very beginning of the packet.
698 - Trace UUID, used to ensure the event packet match the meta-data used.
699 (note: we cannot use a meta-data checksum in every cases instead of a
700 UUID because meta-data can be appended to while tracing is active)
701 This field is optional.
702 - Stream ID, used as reference to stream description in meta-data.
703 This field is optional if there is only one stream description in the
704 meta-data, but becomes required if there are more than one stream in
705 the TSDL meta-data description.
707 Event packet context (all fields are optional, specified by TSDL meta-data):
709 - Event packet content size (in bits).
710 - Event packet size (in bits, includes padding).
711 - Event packet content checksum. Checksum excludes the event packet
713 - Per-stream event packet sequence count (to deal with UDP packet loss). The
714 number of significant sequence counter bits should also be present, so
715 wrap-arounds are dealt with correctly.
716 - Time-stamp at the beginning and time-stamp at the end of the event packet.
717 Both timestamps are written in the packet header, but sampled respectively
718 while (or before) writing the first event and while (or after) writing the
719 last event in the packet. The inclusive range between these timestamps should
720 include all event timestamps assigned to events contained within the packet.
721 The timestamp at the beginning of an event packet is guaranteed to be
722 below or equal the timestamp at the end of that event packet.
723 The timestamp at the end of an event packet is guaranteed to be below
724 or equal the timestamps at the end of any following packet within the
725 same stream. See Section 8. Clocks for more detail.
726 - Events discarded count
727 - Snapshot of a per-stream free-running counter, counting the number of
728 events discarded that were supposed to be written in the stream after
729 the last event in the event packet.
730 * Note: producer-consumer buffer full condition can fill the current
731 event packet with padding so we know exactly where events have been
732 discarded. However, if the buffer full condition chooses not
733 to fill the current event packet with padding, all we know
734 about the timestamp range in which the events have been
735 discarded is that it is somewhere between the beginning and
736 the end of the packet.
737 - Lossless compression scheme used for the event packet content. Applied
738 directly to raw data. New types of compression can be added in following
739 versions of the format.
740 0: no compression scheme
744 - Cypher used for the event packet content. Applied after compression.
747 - Checksum scheme used for the event packet content. Applied after encryption.
753 5.1 Event Packet Header Description
755 The event packet header layout is indicated by the trace packet.header
756 field. Here is a recommended structure type for the packet header with
757 the fields typically expected (although these fields are each optional):
759 struct event_packet_header {
767 packet.header := struct event_packet_header;
770 If the magic number is not present, tools such as "file" will have no
771 mean to discover the file type.
773 If the uuid is not present, no validation that the meta-data actually
774 corresponds to the stream is performed.
776 If the stream_id packet header field is missing, the trace can only
777 contain a single stream. Its "id" field can be left out, and its events
778 don't need to declare a "stream_id" field.
781 5.2 Event Packet Context Description
783 Event packet context example. These are declared within the stream declaration
784 in the meta-data. All these fields are optional. If the packet size field is
785 missing, the whole stream only contains a single packet. If the content
786 size field is missing, the packet is filled (no padding). The content
787 and packet sizes include all headers.
789 An example event packet context type:
791 struct event_packet_context {
792 uint64_t timestamp_begin;
793 uint64_t timestamp_end;
795 uint32_t stream_packet_count;
796 uint32_t events_discarded;
798 uint64_t/uint32_t/uint16_t content_size;
799 uint64_t/uint32_t/uint16_t packet_size;
800 uint8_t compression_scheme;
801 uint8_t encryption_scheme;
802 uint8_t checksum_scheme;
808 The overall structure of an event is:
810 1 - Event Header (as specified by the stream meta-data)
811 2 - Stream Event Context (as specified by the stream meta-data)
812 3 - Event Context (as specified by the event meta-data)
813 4 - Event Payload (as specified by the event meta-data)
815 This structure defines an implicit dynamic scoping, where variants
816 located in inner structures (those with a higher number in the listing
817 above) can refer to the fields of outer structures (with lower number in
818 the listing above). See Section 7.3 TSDL Scopes for more detail.
820 The total length of an event is defined as the difference between the
821 end of its Event Payload and the end of the previous event's Event
822 Payload. Therefore, it includes the event header alignment padding, and
823 all its fields and their respective alignment padding. Events of length
828 Event headers can be described within the meta-data. We hereby propose, as an
829 example, two types of events headers. Type 1 accommodates streams with less than
830 31 event IDs. Type 2 accommodates streams with 31 or more event IDs.
832 One major factor can vary between streams: the number of event IDs assigned to
833 a stream. Luckily, this information tends to stay relatively constant (modulo
834 event registration while trace is being recorded), so we can specify different
835 representations for streams containing few event IDs and streams containing
836 many event IDs, so we end up representing the event ID and time-stamp as
837 densely as possible in each case.
839 The header is extended in the rare occasions where the information cannot be
840 represented in the ranges available in the standard event header. They are also
841 used in the rare occasions where the data required for a field could not be
842 collected: the flag corresponding to the missing field within the missing_fields
843 array is then set to 1.
845 Types uintX_t represent an X-bit unsigned integer, as declared with
848 typealias integer { size = X; align = X; signed = false; } := uintX_t;
852 typealias integer { size = X; align = 1; signed = false; } := uintX_t;
854 For more information about timestamp fields, see Section 8. Clocks.
856 6.1.1 Type 1 - Few event IDs
858 - Aligned on 32-bit (or 8-bit if byte-packed, depending on the architecture
860 - Native architecture byte ordering.
861 - For "compact" selection
862 - Fixed size: 32 bits.
863 - For "extended" selection
864 - Size depends on the architecture and variant alignment.
866 struct event_header_1 {
869 * id 31 is reserved to indicate an extended header.
871 enum : uint5_t { compact = 0 ... 30, extended = 31 } id;
877 uint32_t id; /* 32-bit event IDs */
878 uint64_t timestamp; /* 64-bit timestamps */
881 } align(32); /* or align(8) */
884 6.1.2 Type 2 - Many event IDs
886 - Aligned on 16-bit (or 8-bit if byte-packed, depending on the architecture
888 - Native architecture byte ordering.
889 - For "compact" selection
890 - Size depends on the architecture and variant alignment.
891 - For "extended" selection
892 - Size depends on the architecture and variant alignment.
894 struct event_header_2 {
896 * id: range: 0 - 65534.
897 * id 65535 is reserved to indicate an extended header.
899 enum : uint16_t { compact = 0 ... 65534, extended = 65535 } id;
905 uint32_t id; /* 32-bit event IDs */
906 uint64_t timestamp; /* 64-bit timestamps */
909 } align(16); /* or align(8) */
912 6.2 Stream Event Context and Event Context
914 The event context contains information relative to the current event.
915 The choice and meaning of this information is specified by the TSDL
916 stream and event meta-data descriptions. The stream context is applied
917 to all events within the stream. The stream context structure follows
918 the event header. The event context is applied to specific events. Its
919 structure follows the stream context structure.
921 An example of stream-level event context is to save the event payload size with
922 each event, or to save the current PID with each event. These are declared
923 within the stream declaration within the meta-data:
927 event.context := struct {
929 uint16_t payload_size;
933 An example of event-specific event context is to declare a bitmap of missing
934 fields, only appended after the stream event context if the extended event
935 header is selected. NR_FIELDS is the number of fields within the event (a
943 uint1_t missing_fields[NR_FIELDS]; /* missing event fields bitmap */
952 An event payload contains fields specific to a given event type. The fields
953 belonging to an event type are described in the event-specific meta-data
954 within a structure type.
958 No padding at the end of the event payload. This differs from the ISO/C standard
959 for structures, but follows the CTF standard for structures. In a trace, even
960 though it makes sense to align the beginning of a structure, it really makes no
961 sense to add padding at the end of the structure, because structures are usually
962 not followed by a structure of the same type.
964 This trick can be done by adding a zero-length "end" field at the end of the C
965 structures, and by using the offset of this field rather than using sizeof()
966 when calculating the size of a structure (see Appendix "A. Helper macros").
970 The event payload is aligned on the largest alignment required by types
971 contained within the payload. (This follows the ISO/C standard for structures)
974 7. Trace Stream Description Language (TSDL)
976 The Trace Stream Description Language (TSDL) allows expression of the
977 binary trace streams layout in a C99-like Domain Specific Language
983 The trace stream layout description is located in the trace meta-data.
984 The meta-data is itself located in a stream identified by its name:
987 The meta-data description can be expressed in two different formats:
988 text-only and packet-based. The text-only description facilitates
989 generation of meta-data and provides a convenient way to enter the
990 meta-data information by hand. The packet-based meta-data provides the
991 CTF stream packet facilities (checksumming, compression, encryption,
992 network-readiness) for meta-data stream generated and transported by a
995 The text-only meta-data file is a plain-text TSDL description. This file
996 must begin with the following characters to identify the file as a CTF
997 TSDL text-based metadata file (without the double-quotes) :
1001 It must be followed by a space, and the version of the specification
1002 followed by the CTF trace, e.g.:
1006 These characters allow automated discovery of file type and CTF
1007 specification version. They are interpreted as a the beginning of a
1008 comment by the TSDL metadata parser. The comment can be continued to
1009 contain extra commented characters before it is closed.
1011 The packet-based meta-data is made of "meta-data packets", which each
1012 start with a meta-data packet header. The packet-based meta-data
1013 description is detected by reading the magic number "0x75D11D57" at the
1014 beginning of the file. This magic number is also used to detect the
1015 endianness of the architecture by trying to read the CTF magic number
1016 and its counterpart in reversed endianness. The events within the
1017 meta-data stream have no event header nor event context. Each event only
1018 contains a special "sequence" payload, which is a sequence of bits which
1019 length is implicitly calculated by using the
1020 "trace.packet.header.content_size" field, minus the packet header size.
1021 The formatting of this sequence of bits is a plain-text representation
1022 of the TSDL description. Each meta-data packet start with a special
1023 packet header, specific to the meta-data stream, which contains,
1026 struct metadata_packet_header {
1027 uint32_t magic; /* 0x75D11D57 */
1028 uint8_t uuid[16]; /* Unique Universal Identifier */
1029 uint32_t checksum; /* 0 if unused */
1030 uint32_t content_size; /* in bits */
1031 uint32_t packet_size; /* in bits */
1032 uint8_t compression_scheme; /* 0 if unused */
1033 uint8_t encryption_scheme; /* 0 if unused */
1034 uint8_t checksum_scheme; /* 0 if unused */
1035 uint8_t major; /* CTF spec version major number */
1036 uint8_t minor; /* CTF spec version minor number */
1039 The packet-based meta-data can be converted to a text-only meta-data by
1040 concatenating all the strings it contains.
1042 In the textual representation of the meta-data, the text contained
1043 within "/*" and "*/", as well as within "//" and end of line, are
1044 treated as comments. Boolean values can be represented as true, TRUE,
1045 or 1 for true, and false, FALSE, or 0 for false. Within the string-based
1046 meta-data description, the trace UUID is represented as a string of
1047 hexadecimal digits and dashes "-". In the event packet header, the trace
1048 UUID is represented as an array of bytes.
1051 7.2 Declaration vs Definition
1053 A declaration associates a layout to a type, without specifying where
1054 this type is located in the event structure hierarchy (see Section 6).
1055 This therefore includes typedef, typealias, as well as all type
1056 specifiers. In certain circumstances (typedef, structure field and
1057 variant field), a declaration is followed by a declarator, which specify
1058 the newly defined type name (for typedef), or the field name (for
1059 declarations located within structure and variants). Array and sequence,
1060 declared with square brackets ("[" "]"), are part of the declarator,
1061 similarly to C99. The enumeration base type is specified by
1062 ": enum_base", which is part of the type specifier. The variant tag
1063 name, specified between "<" ">", is also part of the type specifier.
1065 A definition associates a type to a location in the event structure
1066 hierarchy (see Section 6). This association is denoted by ":=", as shown
1072 TSDL uses three different types of scoping: a lexical scope is used for
1073 declarations and type definitions, and static and dynamic scopes are
1074 used for variants references to tag fields (with relative and absolute
1075 path lookups) and for sequence references to length fields.
1079 Each of "trace", "env", "stream", "event", "struct" and "variant" have
1080 their own nestable declaration scope, within which types can be declared
1081 using "typedef" and "typealias". A root declaration scope also contains
1082 all declarations located outside of any of the aforementioned
1083 declarations. An inner declaration scope can refer to type declared
1084 within its container lexical scope prior to the inner declaration scope.
1085 Redefinition of a typedef or typealias is not valid, although hiding an
1086 upper scope typedef or typealias is allowed within a sub-scope.
1088 7.3.2 Static and Dynamic Scopes
1090 A local static scope consists in the scope generated by the declaration
1091 of fields within a compound type. A static scope is a local static scope
1092 augmented with the nested sub-static-scopes it contains.
1094 A dynamic scope consists in the static scope augmented with the
1095 implicit event structure definition hierarchy presented at Section 6.
1097 Multiple declarations of the same field name within a local static scope
1098 is not valid. It is however valid to re-use the same field name in
1099 different local scopes.
1101 Nested static and dynamic scopes form lookup paths. These are used for
1102 variant tag and sequence length references. They are used at the variant
1103 and sequence definition site to look up the location of the tag field
1104 associated with a variant, and to lookup up the location of the length
1105 field associated with a sequence.
1107 Variants and sequences can refer to a tag field either using a relative
1108 path or an absolute path. The relative path is relative to the scope in
1109 which the variant or sequence performing the lookup is located.
1110 Relative paths are only allowed to lookup within the same static scope,
1111 which includes its nested static scopes. Lookups targeting parent static
1112 scopes need to be performed with an absolute path.
1114 Absolute path lookups use the full path including the dynamic scope
1115 followed by a "." and then the static scope. Therefore, variants (or
1116 sequences) in lower levels in the dynamic scope (e.g. event context) can
1117 refer to a tag (or length) field located in upper levels (e.g. in the
1118 event header) by specifying, in this case, the associated tag with
1119 <stream.event.header.field_name>. This allows, for instance, the event
1120 context to define a variant referring to the "id" field of the event
1123 The dynamic scope prefixes are thus:
1125 - Trace Environment: <env. >,
1126 - Trace Packet Header: <trace.packet.header. >,
1127 - Stream Packet Context: <stream.packet.context. >,
1128 - Event Header: <stream.event.header. >,
1129 - Stream Event Context: <stream.event.context. >,
1130 - Event Context: <event.context. >,
1131 - Event Payload: <event.fields. >.
1134 The target dynamic scope must be specified explicitly when referring to
1135 a field outside of the static scope (absolute scope reference). No
1136 conflict can occur between relative and dynamic paths, because the
1137 keywords "trace", "stream", and "event" are reserved, and thus
1138 not permitted as field names. It is recommended that field names
1139 clashing with CTF and C99 reserved keywords use an underscore prefix to
1140 eliminate the risk of generating a description containing an invalid
1141 field name. Consequently, fields starting with an underscore should have
1142 their leading underscore removed by the CTF trace readers.
1145 The information available in the dynamic scopes can be thought of as the
1146 current tracing context. At trace production, information about the
1147 current context is saved into the specified scope field levels. At trace
1148 consumption, for each event, the current trace context is therefore
1149 readable by accessing the upper dynamic scopes.
1154 The grammar representing the TSDL meta-data is presented in Appendix C.
1155 TSDL Grammar. This section presents a rather lighter reading that
1156 consists in examples of TSDL meta-data, with template values.
1158 The stream "id" can be left out if there is only one stream in the
1159 trace. The event "id" field can be left out if there is only one event
1163 major = value; /* CTF spec version major number */
1164 minor = value; /* CTF spec version minor number */
1165 uuid = "aaaaaaaa-aaaa-aaaa-aaaa-aaaaaaaaaaaa"; /* Trace UUID */
1166 byte_order = be OR le; /* Endianness (required) */
1167 packet.header := struct {
1175 * The "env" (environment) scope contains assignment expressions. The
1176 * field names and content are implementation-defined.
1179 pid = value; /* example */
1180 proc_name = "name"; /* example */
1186 /* Type 1 - Few event IDs; Type 2 - Many event IDs. See section 6.1. */
1187 event.header := event_header_1 OR event_header_2;
1188 event.context := struct {
1191 packet.context := struct {
1197 name = "event_name";
1198 id = value; /* Numeric identifier within the stream */
1199 stream_id = stream_id;
1201 model.emf.uri = "string";
1211 name = "event_name";
1218 /* More detail on types in section 4. Types */
1223 * Type declarations behave similarly to the C standard.
1226 typedef aliased_type_specifiers new_type_declarators;
1228 /* e.g.: typedef struct example new_type_name[10]; */
1233 * The "typealias" declaration can be used to give a name (including
1234 * pointer declarator specifier) to a type. It should also be used to
1235 * map basic C types (float, int, unsigned long, ...) to a CTF type.
1236 * Typealias is a superset of "typedef": it also allows assignment of a
1237 * simple variable identifier to a type.
1240 typealias type_class {
1242 } := type_specifiers type_declarator;
1246 * typealias integer {
1250 * } := struct page *;
1252 * typealias integer {
1267 enum name : integer_type {
1273 * Unnamed types, contained within compound type fields, typedef or typealias.
1288 enum : integer_type {
1292 typedef type new_type[length];
1295 type field_name[length];
1298 typedef type new_type[length_type];
1301 type field_name[length_type];
1313 integer_type field_name:size; /* GNU/C bitfield */
1323 Clock metadata allows to describe the clock topology of the system, as
1324 well as to detail each clock parameter. In absence of clock description,
1325 it is assumed that all fields named "timestamp" use the same clock
1326 source, which increments once per nanosecond.
1328 Describing a clock and how it is used by streams is threefold: first,
1329 the clock and clock topology should be described in a "clock"
1330 description block, e.g.:
1333 name = cycle_counter_sync;
1334 uuid = "62189bee-96dc-11e0-91a8-cfa3d89f3923";
1335 description = "Cycle counter synchronized across CPUs";
1336 freq = 1000000000; /* frequency, in Hz */
1337 /* precision in seconds is: 1000 * (1/freq) */
1340 * clock value offset from Epoch is:
1341 * offset_s + (offset * (1/freq))
1343 offset_s = 1326476837;
1348 The mandatory "name" field specifies the name of the clock identifier,
1349 which can later be used as a reference. The optional field "uuid" is the
1350 unique identifier of the clock. It can be used to correlate different
1351 traces that use the same clock. An optional textual description string
1352 can be added with the "description" field. The "freq" field is the
1353 initial frequency of the clock, in Hz. If the "freq" field is not
1354 present, the frequency is assumed to be 1000000000 (providing clock
1355 increment of 1 ns). The optional "precision" field details the
1356 uncertainty on the clock measurements, in (1/freq) units. The "offset_s"
1357 and "offset" fields indicate the offset from POSIX.1 Epoch, 1970-01-01
1358 00:00:00 +0000 (UTC), to the zero of value of the clock. The "offset_s"
1359 field is in seconds. The "offset" field is in (1/freq) units. If any of
1360 the "offset_s" or "offset" field is not present, it is assigned the 0
1361 value. The field "absolute" is TRUE if the clock is a global reference
1362 across different clock uuid (e.g. NTP time). Otherwise, "absolute" is
1363 FALSE, and the clock can be considered as synchronized only with other
1364 clocks that have the same uuid.
1367 Secondly, a reference to this clock should be added within an integer
1371 size = 64; align = 1; signed = false;
1372 map = clock.cycle_counter_sync.value;
1375 Thirdly, stream declarations can reference the clock they use as a
1378 struct packet_context {
1379 uint64_ccnt_t ccnt_begin;
1380 uint64_ccnt_t ccnt_end;
1386 event.header := struct {
1387 uint64_ccnt_t timestamp;
1390 packet.context := struct packet_context;
1393 For a N-bit integer type referring to a clock, if the integer overflows
1394 compared to the N low order bits of the clock prior value found in the
1395 same stream, then it is assumed that one, and only one, overflow
1396 occurred. It is therefore important that events encoding time on a small
1397 number of bits happen frequently enough to detect when more than one
1398 N-bit overflow occurs.
1400 In a packet context, clock field names ending with "_begin" and "_end"
1401 have a special meaning: this refers to the time-stamps at, respectively,
1402 the beginning and the end of each packet.
1407 The two following macros keep track of the size of a GNU/C structure without
1408 padding at the end by placing HEADER_END as the last field. A one byte end field
1409 is used for C90 compatibility (C99 flexible arrays could be used here). Note
1410 that this does not affect the effective structure size, which should always be
1411 calculated with the header_sizeof() helper.
1413 #define HEADER_END char end_field
1414 #define header_sizeof(type) offsetof(typeof(type), end_field)
1417 B. Stream Header Rationale
1419 An event stream is divided in contiguous event packets of variable size. These
1420 subdivisions allow the trace analyzer to perform a fast binary search by time
1421 within the stream (typically requiring to index only the event packet headers)
1422 without reading the whole stream. These subdivisions have a variable size to
1423 eliminate the need to transfer the event packet padding when partially filled
1424 event packets must be sent when streaming a trace for live viewing/analysis.
1425 An event packet can contain a certain amount of padding at the end. Dividing
1426 streams into event packets is also useful for network streaming over UDP and
1427 flight recorder mode tracing (a whole event packet can be swapped out of the
1428 buffer atomically for reading).
1430 The stream header is repeated at the beginning of each event packet to allow
1431 flexibility in terms of:
1433 - streaming support,
1434 - allowing arbitrary buffers to be discarded without making the trace
1436 - allow UDP packet loss handling by either dealing with missing event packet
1437 or asking for re-transmission.
1438 - transparently support flight recorder mode,
1439 - transparently support crash dump.
1445 * Common Trace Format (CTF) Trace Stream Description Language (TSDL) Grammar.
1447 * Inspired from the C99 grammar:
1448 * http://www.open-std.org/jtc1/sc22/wg14/www/docs/n1124.pdf (Annex A)
1449 * and c++1x grammar (draft)
1450 * http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2011/n3291.pdf (Annex A)
1452 * Specialized for CTF needs by including only constant and declarations from
1453 * C99 (excluding function declarations), and by adding support for variants,
1454 * sequences and CTF-specific specifiers. Enumeration container types
1455 * semantic is inspired from c++1x enum-base.
1460 1.1) Lexical elements
1507 identifier identifier-nondigit
1510 identifier-nondigit:
1512 universal-character-name
1513 any other implementation-defined characters
1517 [a-zA-Z] /* regular expression */
1520 [0-9] /* regular expression */
1522 1.4) Universal character names
1524 universal-character-name:
1526 \U hex-quad hex-quad
1529 hexadecimal-digit hexadecimal-digit hexadecimal-digit hexadecimal-digit
1535 enumeration-constant
1539 decimal-constant integer-suffix-opt
1540 octal-constant integer-suffix-opt
1541 hexadecimal-constant integer-suffix-opt
1545 decimal-constant digit
1549 octal-constant octal-digit
1551 hexadecimal-constant:
1552 hexadecimal-prefix hexadecimal-digit
1553 hexadecimal-constant hexadecimal-digit
1563 unsigned-suffix long-suffix-opt
1564 unsigned-suffix long-long-suffix
1565 long-suffix unsigned-suffix-opt
1566 long-long-suffix unsigned-suffix-opt
1580 enumeration-constant:
1586 L' c-char-sequence '
1590 c-char-sequence c-char
1593 any member of source charset except single-quote ('), backslash
1594 (\), or new-line character.
1598 simple-escape-sequence
1599 octal-escape-sequence
1600 hexadecimal-escape-sequence
1601 universal-character-name
1603 simple-escape-sequence: one of
1604 \' \" \? \\ \a \b \f \n \r \t \v
1606 octal-escape-sequence:
1608 \ octal-digit octal-digit
1609 \ octal-digit octal-digit octal-digit
1611 hexadecimal-escape-sequence:
1612 \x hexadecimal-digit
1613 hexadecimal-escape-sequence hexadecimal-digit
1615 1.6) String literals
1618 " s-char-sequence-opt "
1619 L" s-char-sequence-opt "
1623 s-char-sequence s-char
1626 any member of source charset except double-quote ("), backslash
1627 (\), or new-line character.
1633 [ ] ( ) { } . -> * + - < > : ; ... = ,
1636 2) Phrase structure grammar
1642 ( unary-expression )
1646 postfix-expression [ unary-expression ]
1647 postfix-expression . identifier
1648 postfix-expressoin -> identifier
1652 unary-operator postfix-expression
1654 unary-operator: one of
1657 assignment-operator:
1660 type-assignment-operator:
1663 constant-expression-range:
1664 unary-expression ... unary-expression
1669 declaration-specifiers declarator-list-opt ;
1672 declaration-specifiers:
1673 storage-class-specifier declaration-specifiers-opt
1674 type-specifier declaration-specifiers-opt
1675 type-qualifier declaration-specifiers-opt
1679 declarator-list , declarator
1681 abstract-declarator-list:
1683 abstract-declarator-list , abstract-declarator
1685 storage-class-specifier:
1708 align ( unary-expression )
1711 struct identifier-opt { struct-or-variant-declaration-list-opt } align-attribute-opt
1712 struct identifier align-attribute-opt
1714 struct-or-variant-declaration-list:
1715 struct-or-variant-declaration
1716 struct-or-variant-declaration-list struct-or-variant-declaration
1718 struct-or-variant-declaration:
1719 specifier-qualifier-list struct-or-variant-declarator-list ;
1720 declaration-specifiers-opt storage-class-specifier declaration-specifiers-opt declarator-list ;
1721 typealias declaration-specifiers abstract-declarator-list type-assignment-operator declaration-specifiers abstract-declarator-list ;
1722 typealias declaration-specifiers abstract-declarator-list type-assignment-operator declarator-list ;
1724 specifier-qualifier-list:
1725 type-specifier specifier-qualifier-list-opt
1726 type-qualifier specifier-qualifier-list-opt
1728 struct-or-variant-declarator-list:
1729 struct-or-variant-declarator
1730 struct-or-variant-declarator-list , struct-or-variant-declarator
1732 struct-or-variant-declarator:
1734 declarator-opt : unary-expression
1737 variant identifier-opt variant-tag-opt { struct-or-variant-declaration-list }
1738 variant identifier variant-tag
1741 < unary-expression >
1744 enum identifier-opt { enumerator-list }
1745 enum identifier-opt { enumerator-list , }
1747 enum identifier-opt : declaration-specifiers { enumerator-list }
1748 enum identifier-opt : declaration-specifiers { enumerator-list , }
1752 enumerator-list , enumerator
1755 enumeration-constant
1756 enumeration-constant assignment-operator unary-expression
1757 enumeration-constant assignment-operator constant-expression-range
1763 pointer-opt direct-declarator
1768 direct-declarator [ unary-expression ]
1770 abstract-declarator:
1771 pointer-opt direct-abstract-declarator
1773 direct-abstract-declarator:
1775 ( abstract-declarator )
1776 direct-abstract-declarator [ unary-expression ]
1777 direct-abstract-declarator [ ]
1780 * type-qualifier-list-opt
1781 * type-qualifier-list-opt pointer
1783 type-qualifier-list:
1785 type-qualifier-list type-qualifier
1790 2.3) CTF-specific declarations
1793 clock { ctf-assignment-expression-list-opt }
1794 event { ctf-assignment-expression-list-opt }
1795 stream { ctf-assignment-expression-list-opt }
1796 env { ctf-assignment-expression-list-opt }
1797 trace { ctf-assignment-expression-list-opt }
1798 callsite { ctf-assignment-expression-list-opt }
1799 typealias declaration-specifiers abstract-declarator-list type-assignment-operator declaration-specifiers abstract-declarator-list
1800 typealias declaration-specifiers abstract-declarator-list type-assignment-operator declarator-list
1803 floating_point { ctf-assignment-expression-list-opt }
1804 integer { ctf-assignment-expression-list-opt }
1805 string { ctf-assignment-expression-list-opt }
1808 ctf-assignment-expression-list:
1809 ctf-assignment-expression ;
1810 ctf-assignment-expression-list ctf-assignment-expression ;
1812 ctf-assignment-expression:
1813 unary-expression assignment-operator unary-expression
1814 unary-expression type-assignment-operator type-specifier
1815 declaration-specifiers-opt storage-class-specifier declaration-specifiers-opt declarator-list
1816 typealias declaration-specifiers abstract-declarator-list type-assignment-operator declaration-specifiers abstract-declarator-list
1817 typealias declaration-specifiers abstract-declarator-list type-assignment-operator declarator-list