1 Common Trace Format (CTF) Specification (v1.8.1)
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
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. These subdivisions have a variable size. An event packet can
122 contain a certain amount of padding at the end. The stream header is
123 repeated at the beginning of each event packet. The rationale for the
124 event stream design choices is explained in Appendix B. Stream Header
127 The event stream header will therefore be referred to as the "event packet
128 header" throughout the rest of this document.
133 Types are organized as type classes. Each type class belong to either of two
134 kind of types: basic types or compound types.
138 A basic type is a scalar type, as described in this section. It includes
139 integers, GNU/C bitfields, enumerations, and floating point values.
141 4.1.1 Type inheritance
143 Type specifications can be inherited to allow deriving types from a
144 type class. For example, see the uint32_t named type derived from the "integer"
145 type class below ("Integers" section). Types have a precise binary
146 representation in the trace. A type class has methods to read and write these
147 types, but must be derived into a type to be usable in an event field.
151 We define "byte-packed" types as aligned on the byte size, namely 8-bit.
152 We define "bit-packed" types as following on the next bit, as defined by the
155 Each basic type must specify its alignment, in bits. Examples of
156 possible alignments are: bit-packed (align = 1), byte-packed (align =
157 8), or word-aligned (e.g. align = 32 or align = 64). The choice depends
158 on the architecture preference and compactness vs performance trade-offs
159 of the implementation. Architectures providing fast unaligned write
160 byte-packed basic types to save space, aligning each type on byte
161 boundaries (8-bit). Architectures with slow unaligned writes align types
162 on specific alignment values. If no specific alignment is declared for a
163 type, it is assumed to be bit-packed for integers with size not multiple
164 of 8 bits and for gcc bitfields. All other basic types are byte-packed
165 by default. It is however recommended to always specify the alignment
166 explicitly. Alignment values must be power of two. Compound types are
167 aligned as specified in their individual specification.
169 TSDL meta-data attribute representation of a specific alignment:
171 align = value; /* value in bits */
175 By default, the native endianness of the source architecture is used.
176 Byte order can be overridden for a basic type by specifying a "byte_order"
177 attribute. Typical use-case is to specify the network byte order (big endian:
178 "be") to save data captured from the network into the trace without conversion.
179 If not specified, the byte order is native.
181 TSDL meta-data representation:
183 byte_order = native OR network OR be OR le; /* network and be are aliases */
187 Type size, in bits, for integers and floats is that returned by "sizeof()" in C
188 multiplied by CHAR_BIT.
189 We require the size of "char" and "unsigned char" types (CHAR_BIT) to be fixed
190 to 8 bits for cross-endianness compatibility.
192 TSDL meta-data representation:
194 size = value; (value is in bits)
198 Signed integers are represented in two-complement. Integer alignment,
199 size, signedness and byte ordering are defined in the TSDL meta-data.
200 Integers aligned on byte size (8-bit) and with length multiple of byte
201 size (8-bit) correspond to the C99 standard integers. In addition,
202 integers with alignment and/or size that are _not_ a multiple of the
203 byte size are permitted; these correspond to the C99 standard bitfields,
204 with the added specification that the CTF integer bitfields have a fixed
205 binary representation. A MIT-licensed reference implementation of the
206 CTF portable bitfields is available at:
208 http://git.efficios.com/?p=babeltrace.git;a=blob;f=include/babeltrace/bitfield.h
210 Binary representation of integers:
212 - On little and big endian:
213 - Within a byte, high bits correspond to an integer high bits, and low bits
214 correspond to low bits.
216 - Integer across multiple bytes are placed from the less significant to the
218 - Consecutive integers are placed from lower bits to higher bits (even within
221 - Integer across multiple bytes are placed from the most significant to the
223 - Consecutive integers are placed from higher bits to lower bits (even within
226 This binary representation is derived from the bitfield implementation in GCC
227 for little and big endian. However, contrary to what GCC does, integers can
228 cross units boundaries (no padding is required). Padding can be explicitly
229 added (see 4.1.6 GNU/C bitfields) to follow the GCC layout if needed.
231 TSDL meta-data representation:
234 signed = true OR false; /* default false */
235 byte_order = native OR network OR be OR le; /* default native */
236 size = value; /* value in bits, no default */
237 align = value; /* value in bits */
238 /* based used for pretty-printing output, default: decimal. */
239 base = decimal OR dec OR OR d OR i OR u OR 10 OR hexadecimal OR hex OR x OR X OR p OR 16
240 OR octal OR oct OR o OR 8 OR binary OR b OR 2;
241 /* character encoding, default: none */
242 encoding = none or UTF8 or ASCII;
245 Example of type inheritance (creation of a uint32_t named type):
253 Definition of a named 5-bit signed bitfield:
261 The character encoding field can be used to specify that the integer
262 must be printed as a text character when read. e.g.:
272 4.1.6 GNU/C bitfields
274 The GNU/C bitfields follow closely the integer representation, with a
275 particularity on alignment: if a bitfield cannot fit in the current unit, the
276 unit is padded and the bitfield starts at the following unit. The unit size is
277 defined by the size of the type "unit_type".
279 TSDL meta-data representation:
283 As an example, the following structure declared in C compiled by GCC:
290 The example structure is aligned on the largest element (short). The second
291 bitfield would be aligned on the next unit boundary, because it would not fit in
296 The floating point values byte ordering is defined in the TSDL meta-data.
298 Floating point values follow the IEEE 754-2008 standard interchange formats.
299 Description of the floating point values include the exponent and mantissa size
300 in bits. Some requirements are imposed on the floating point values:
302 - FLT_RADIX must be 2.
303 - mant_dig is the number of digits represented in the mantissa. It is specified
304 by the ISO C99 standard, section 5.2.4, as FLT_MANT_DIG, DBL_MANT_DIG and
305 LDBL_MANT_DIG as defined by <float.h>.
306 - exp_dig is the number of digits represented in the exponent. Given that
307 mant_dig is one bit more than its actual size in bits (leading 1 is not
308 needed) and also given that the sign bit always takes one bit, exp_dig can be
311 - sizeof(float) * CHAR_BIT - FLT_MANT_DIG
312 - sizeof(double) * CHAR_BIT - DBL_MANT_DIG
313 - sizeof(long double) * CHAR_BIT - LDBL_MANT_DIG
315 TSDL meta-data representation:
320 byte_order = native OR network OR be OR le;
324 Example of type inheritance:
326 typealias floating_point {
327 exp_dig = 8; /* sizeof(float) * CHAR_BIT - FLT_MANT_DIG */
328 mant_dig = 24; /* FLT_MANT_DIG */
333 TODO: define NaN, +inf, -inf behavior.
335 Bit-packed, byte-packed or larger alignments can be used for floating
336 point values, similarly to integers.
340 Enumerations are a mapping between an integer type and a table of strings. The
341 numerical representation of the enumeration follows the integer type specified
342 by the meta-data. The enumeration mapping table is detailed in the enumeration
343 description within the meta-data. The mapping table maps inclusive value
344 ranges (or single values) to strings. Instead of being limited to simple
345 "value -> string" mappings, these enumerations map
346 "[ start_value ... end_value ] -> string", which map inclusive ranges of
347 values to strings. An enumeration from the C language can be represented in
348 this format by having the same start_value and end_value for each element, which
349 is in fact a range of size 1. This single-value range is supported without
350 repeating the start and end values with the value = string declaration.
352 enum name : integer_type {
353 somestring = start_value1 ... end_value1,
354 "other string" = start_value2 ... end_value2,
355 yet_another_string, /* will be assigned to end_value2 + 1 */
356 "some other string" = value,
360 If the values are omitted, the enumeration starts at 0 and increment of 1 for
361 each entry. An entry with omitted value that follows a range entry takes
362 as value the end_value of the previous range + 1:
364 enum name : unsigned int {
372 Overlapping ranges within a single enumeration are implementation defined.
374 A nameless enumeration can be declared as a field type or as part of a typedef:
376 enum : integer_type {
380 Enumerations omitting the container type ": integer_type" use the "int"
381 type (for compatibility with C99). The "int" type must be previously
384 typealias integer { size = 32; align = 32; signed = true } := int;
393 Compound are aggregation of type declarations. Compound types include
394 structures, variant, arrays, sequences, and strings.
398 Structures are aligned on the largest alignment required by basic types
399 contained within the structure. (This follows the ISO/C standard for structures)
401 TSDL meta-data representation of a named structure:
404 field_type field_name;
405 field_type field_name;
412 integer { /* Nameless type */
417 uint64_t second_field_name; /* Named type declared in the meta-data */
420 The fields are placed in a sequence next to each other. They each
421 possess a field name, which is a unique identifier within the structure.
422 The identifier is not allowed to use any reserved keyword
423 (see Section C.1.2). Replacing reserved keywords with
424 underscore-prefixed field names is recommended. Fields starting with an
425 underscore should have their leading underscore removed by the CTF trace
428 A nameless structure can be declared as a field type or as part of a typedef:
434 Alignment for a structure compound type can be forced to a minimum value
435 by adding an "align" specifier after the declaration of a structure
436 body. This attribute is read as: align(value). The value is specified in
437 bits. The structure will be aligned on the maximum value between this
438 attribute and the alignment required by the basic types contained within
445 4.2.2 Variants (Discriminated/Tagged Unions)
447 A CTF variant is a selection between different types. A CTF variant must
448 always be defined within the scope of a structure or within fields
449 contained within a structure (defined recursively). A "tag" enumeration
450 field must appear in either the same static scope, prior to the variant
451 field (in field declaration order), in an upper static scope , or in an
452 upper dynamic scope (see Section 7.3.2). The type selection is indicated
453 by the mapping from the enumeration value to the string used as variant
454 type selector. The field to use as tag is specified by the "tag_field",
455 specified between "< >" after the "variant" keyword for unnamed
456 variants, and after "variant name" for named variants.
458 The alignment of the variant is the alignment of the type as selected by the tag
459 value for the specific instance of the variant. The alignment of the type
460 containing the variant is independent of the variant alignment. The size of the
461 variant is the size as selected by the tag value for the specific instance of
464 Each variant type selector possess a field name, which is a unique
465 identifier within the variant. The identifier is not allowed to use any
466 reserved keyword (see Section C.1.2). Replacing reserved keywords with
467 underscore-prefixed field names is recommended. Fields starting with an
468 underscore should have their leading underscore removed by the CTF trace
472 A named variant declaration followed by its definition within a structure
483 enum : integer_type { sel1, sel2, sel3, ... } tag_field;
485 variant name <tag_field> v;
488 An unnamed variant definition within a structure is expressed by the following
492 enum : integer_type { sel1, sel2, sel3, ... } tag_field;
494 variant <tag_field> {
502 Example of a named variant within a sequence that refers to a single tag field:
511 enum : uint2_t { a, b, c } choice;
513 variant example <choice> v[seqlen];
516 Example of an unnamed variant:
519 enum : uint2_t { a, b, c, d } choice;
520 /* Unrelated fields can be added between the variant and its tag */
533 Example of an unnamed variant within an array:
536 enum : uint2_t { a, b, c } choice;
544 Example of a variant type definition within a structure, where the defined type
545 is then declared within an array of structures. This variant refers to a tag
546 located in an upper static scope. This example clearly shows that a variant
547 type definition referring to the tag "x" uses the closest preceding field from
548 the static scope of the type definition.
551 enum : uint2_t { a, b, c, d } x;
553 typedef variant <x> { /*
554 * "x" refers to the preceding "x" enumeration in the
555 * static scope of the type definition.
563 enum : int { x, y, z } x; /* This enumeration is not used by "v". */
564 example_variant v; /*
565 * "v" uses the "enum : uint2_t { a, b, c, d }"
573 Arrays are fixed-length. Their length is declared in the type
574 declaration within the meta-data. They contain an array of "inner type"
575 elements, which can refer to any type not containing the type of the
576 array being declared (no circular dependency). The length is the number
577 of elements in an array.
579 TSDL meta-data representation of a named array:
581 typedef elem_type name[length];
583 A nameless array can be declared as a field type within a structure, e.g.:
585 uint8_t field_name[10];
587 Arrays are always aligned on their element alignment requirement.
591 Sequences are dynamically-sized arrays. They refer to a a "length"
592 unsigned integer field, which must appear in either the same static scope,
593 prior to the sequence field (in field declaration order), in an upper
594 static scope, or in an upper dynamic scope (see Section 7.3.2). This
595 length field represents the number of elements in the sequence. The
596 sequence per se is an array of "inner type" elements.
598 TSDL meta-data representation for a sequence type definition:
601 unsigned int length_field;
602 typedef elem_type typename[length_field];
603 typename seq_field_name;
606 A sequence can also be declared as a field type, e.g.:
609 unsigned int length_field;
610 long seq_field_name[length_field];
613 Multiple sequences can refer to the same length field, and these length
614 fields can be in a different upper dynamic scope:
616 e.g., assuming the stream.event.header defines:
621 event.header := struct {
630 long seq_a[stream.event.header.seq_len];
631 char seq_b[stream.event.header.seq_len];
635 The sequence elements follow the "array" specifications.
639 Strings are an array of bytes of variable size and are terminated by a '\0'
640 "NULL" character. Their encoding is described in the TSDL meta-data. In
641 absence of encoding attribute information, the default encoding is
644 TSDL meta-data representation of a named string type:
647 encoding = UTF8 OR ASCII;
650 A nameless string type can be declared as a field type:
652 string field_name; /* Use default UTF8 encoding */
654 Strings are always aligned on byte size.
656 5. Event Packet Header
658 The event packet header consists of two parts: the "event packet header"
659 is the same for all streams of a trace. The second part, the "event
660 packet context", is described on a per-stream basis. Both are described
661 in the TSDL meta-data. The packets are aligned on architecture-page-sized
664 Event packet header (all fields are optional, specified by TSDL meta-data):
666 - Magic number (CTF magic number: 0xC1FC1FC1) specifies that this is a
667 CTF packet. This magic number is optional, but when present, it should
668 come at the very beginning of the packet.
669 - Trace UUID, used to ensure the event packet match the meta-data used.
670 (note: we cannot use a meta-data checksum in every cases instead of a
671 UUID because meta-data can be appended to while tracing is active)
672 This field is optional.
673 - Stream ID, used as reference to stream description in meta-data.
674 This field is optional if there is only one stream description in the
675 meta-data, but becomes required if there are more than one stream in
676 the TSDL meta-data description.
678 Event packet context (all fields are optional, specified by TSDL meta-data):
680 - Event packet content size (in bits).
681 - Event packet size (in bits, includes padding).
682 - Event packet content checksum. Checksum excludes the event packet
684 - Per-stream event packet sequence count (to deal with UDP packet loss). The
685 number of significant sequence counter bits should also be present, so
686 wrap-arounds are dealt with correctly.
687 - Time-stamp at the beginning and time-stamp at the end of the event packet.
688 Both timestamps are written in the packet header, but sampled respectively
689 while (or before) writing the first event and while (or after) writing the
690 last event in the packet. The inclusive range between these timestamps should
691 include all event timestamps assigned to events contained within the packet.
692 - Events discarded count
693 - Snapshot of a per-stream free-running counter, counting the number of
694 events discarded that were supposed to be written in the stream prior to
695 the first event in the event packet.
696 * Note: producer-consumer buffer full condition should fill the current
697 event packet with padding so we know exactly where events have been
699 - Lossless compression scheme used for the event packet content. Applied
700 directly to raw data. New types of compression can be added in following
701 versions of the format.
702 0: no compression scheme
706 - Cypher used for the event packet content. Applied after compression.
709 - Checksum scheme used for the event packet content. Applied after encryption.
715 5.1 Event Packet Header Description
717 The event packet header layout is indicated by the trace packet.header
718 field. Here is a recommended structure type for the packet header with
719 the fields typically expected (although these fields are each optional):
721 struct event_packet_header {
729 packet.header := struct event_packet_header;
732 If the magic number is not present, tools such as "file" will have no
733 mean to discover the file type.
735 If the uuid is not present, no validation that the meta-data actually
736 corresponds to the stream is performed.
738 If the stream_id packet header field is missing, the trace can only
739 contain a single stream. Its "id" field can be left out, and its events
740 don't need to declare a "stream_id" field.
743 5.2 Event Packet Context Description
745 Event packet context example. These are declared within the stream declaration
746 in the meta-data. All these fields are optional. If the packet size field is
747 missing, the whole stream only contains a single packet. If the content
748 size field is missing, the packet is filled (no padding). The content
749 and packet sizes include all headers.
751 An example event packet context type:
753 struct event_packet_context {
754 uint64_t timestamp_begin;
755 uint64_t timestamp_end;
757 uint32_t stream_packet_count;
758 uint32_t events_discarded;
760 uint32_t/uint16_t content_size;
761 uint32_t/uint16_t packet_size;
762 uint8_t compression_scheme;
763 uint8_t encryption_scheme;
764 uint8_t checksum_scheme;
770 The overall structure of an event is:
772 1 - Stream Packet Context (as specified by the stream meta-data)
773 2 - Event Header (as specified by the stream meta-data)
774 3 - Stream Event Context (as specified by the stream meta-data)
775 4 - Event Context (as specified by the event meta-data)
776 5 - Event Payload (as specified by the event meta-data)
778 This structure defines an implicit dynamic scoping, where variants
779 located in inner structures (those with a higher number in the listing
780 above) can refer to the fields of outer structures (with lower number in
781 the listing above). See Section 7.3 TSDL Scopes for more detail.
785 Event headers can be described within the meta-data. We hereby propose, as an
786 example, two types of events headers. Type 1 accommodates streams with less than
787 31 event IDs. Type 2 accommodates streams with 31 or more event IDs.
789 One major factor can vary between streams: the number of event IDs assigned to
790 a stream. Luckily, this information tends to stay relatively constant (modulo
791 event registration while trace is being recorded), so we can specify different
792 representations for streams containing few event IDs and streams containing
793 many event IDs, so we end up representing the event ID and time-stamp as
794 densely as possible in each case.
796 The header is extended in the rare occasions where the information cannot be
797 represented in the ranges available in the standard event header. They are also
798 used in the rare occasions where the data required for a field could not be
799 collected: the flag corresponding to the missing field within the missing_fields
800 array is then set to 1.
802 Types uintX_t represent an X-bit unsigned integer, as declared with
805 typealias integer { size = X; align = X; signed = false } := uintX_t;
809 typealias integer { size = X; align = 1; signed = false } := uintX_t;
811 6.1.1 Type 1 - Few event IDs
813 - Aligned on 32-bit (or 8-bit if byte-packed, depending on the architecture
815 - Native architecture byte ordering.
816 - For "compact" selection
817 - Fixed size: 32 bits.
818 - For "extended" selection
819 - Size depends on the architecture and variant alignment.
821 struct event_header_1 {
824 * id 31 is reserved to indicate an extended header.
826 enum : uint5_t { compact = 0 ... 30, extended = 31 } id;
832 uint32_t id; /* 32-bit event IDs */
833 uint64_t timestamp; /* 64-bit timestamps */
836 } align(32); /* or align(8) */
839 6.1.2 Type 2 - Many event IDs
841 - Aligned on 16-bit (or 8-bit if byte-packed, depending on the architecture
843 - Native architecture byte ordering.
844 - For "compact" selection
845 - Size depends on the architecture and variant alignment.
846 - For "extended" selection
847 - Size depends on the architecture and variant alignment.
849 struct event_header_2 {
851 * id: range: 0 - 65534.
852 * id 65535 is reserved to indicate an extended header.
854 enum : uint16_t { compact = 0 ... 65534, extended = 65535 } id;
860 uint32_t id; /* 32-bit event IDs */
861 uint64_t timestamp; /* 64-bit timestamps */
864 } align(16); /* or align(8) */
869 The event context contains information relative to the current event.
870 The choice and meaning of this information is specified by the TSDL
871 stream and event meta-data descriptions. The stream context is applied
872 to all events within the stream. The stream context structure follows
873 the event header. The event context is applied to specific events. Its
874 structure follows the stream context structure.
876 An example of stream-level event context is to save the event payload size with
877 each event, or to save the current PID with each event. These are declared
878 within the stream declaration within the meta-data:
882 event.context := struct {
884 uint16_t payload_size;
888 An example of event-specific event context is to declare a bitmap of missing
889 fields, only appended after the stream event context if the extended event
890 header is selected. NR_FIELDS is the number of fields within the event (a
898 uint1_t missing_fields[NR_FIELDS]; /* missing event fields bitmap */
907 An event payload contains fields specific to a given event type. The fields
908 belonging to an event type are described in the event-specific meta-data
909 within a structure type.
913 No padding at the end of the event payload. This differs from the ISO/C standard
914 for structures, but follows the CTF standard for structures. In a trace, even
915 though it makes sense to align the beginning of a structure, it really makes no
916 sense to add padding at the end of the structure, because structures are usually
917 not followed by a structure of the same type.
919 This trick can be done by adding a zero-length "end" field at the end of the C
920 structures, and by using the offset of this field rather than using sizeof()
921 when calculating the size of a structure (see Appendix "A. Helper macros").
925 The event payload is aligned on the largest alignment required by types
926 contained within the payload. (This follows the ISO/C standard for structures)
929 7. Trace Stream Description Language (TSDL)
931 The Trace Stream Description Language (TSDL) allows expression of the
932 binary trace streams layout in a C99-like Domain Specific Language
938 The trace stream layout description is located in the trace meta-data.
939 The meta-data is itself located in a stream identified by its name:
942 The meta-data description can be expressed in two different formats:
943 text-only and packet-based. The text-only description facilitates
944 generation of meta-data and provides a convenient way to enter the
945 meta-data information by hand. The packet-based meta-data provides the
946 CTF stream packet facilities (checksumming, compression, encryption,
947 network-readiness) for meta-data stream generated and transported by a
950 The text-only meta-data file is a plain-text TSDL description. This file
951 must begin with the following characters to identify the file as a CTF
952 TSDL text-based metadata file (without the double-quotes) :
956 It must be followed by a space, and the version of the specification
957 followed by the CTF trace, e.g.:
961 These characters allow automated discovery of file type and CTF
962 specification version. They are interpreted as a the beginning of a
963 comment by the TSDL metadata parser. The comment can be continued to
964 contain extra commented characters before it is closed.
966 The packet-based meta-data is made of "meta-data packets", which each
967 start with a meta-data packet header. The packet-based meta-data
968 description is detected by reading the magic number "0x75D11D57" at the
969 beginning of the file. This magic number is also used to detect the
970 endianness of the architecture by trying to read the CTF magic number
971 and its counterpart in reversed endianness. The events within the
972 meta-data stream have no event header nor event context. Each event only
973 contains a "sequence" payload, which is a sequence of bits using the
974 "trace.packet.header.content_size" field as a placeholder for its length
975 (the packet header size should be substracted). The formatting of this
976 sequence of bits is a plain-text representation of the TSDL description.
977 Each meta-data packet start with a special packet header, specific to
978 the meta-data stream, which contains, exactly:
980 struct metadata_packet_header {
981 uint32_t magic; /* 0x75D11D57 */
982 uint8_t uuid[16]; /* Unique Universal Identifier */
983 uint32_t checksum; /* 0 if unused */
984 uint32_t content_size; /* in bits */
985 uint32_t packet_size; /* in bits */
986 uint8_t compression_scheme; /* 0 if unused */
987 uint8_t encryption_scheme; /* 0 if unused */
988 uint8_t checksum_scheme; /* 0 if unused */
989 uint8_t major; /* CTF spec version major number */
990 uint8_t minor; /* CTF spec version minor number */
993 The packet-based meta-data can be converted to a text-only meta-data by
994 concatenating all the strings in contains.
996 In the textual representation of the meta-data, the text contained
997 within "/*" and "*/", as well as within "//" and end of line, are
998 treated as comments. Boolean values can be represented as true, TRUE,
999 or 1 for true, and false, FALSE, or 0 for false. Within the string-based
1000 meta-data description, the trace UUID is represented as a string of
1001 hexadecimal digits and dashes "-". In the event packet header, the trace
1002 UUID is represented as an array of bytes.
1005 7.2 Declaration vs Definition
1007 A declaration associates a layout to a type, without specifying where
1008 this type is located in the event structure hierarchy (see Section 6).
1009 This therefore includes typedef, typealias, as well as all type
1010 specifiers. In certain circumstances (typedef, structure field and
1011 variant field), a declaration is followed by a declarator, which specify
1012 the newly defined type name (for typedef), or the field name (for
1013 declarations located within structure and variants). Array and sequence,
1014 declared with square brackets ("[" "]"), are part of the declarator,
1015 similarly to C99. The enumeration base type is specified by
1016 ": enum_base", which is part of the type specifier. The variant tag
1017 name, specified between "<" ">", is also part of the type specifier.
1019 A definition associates a type to a location in the event structure
1020 hierarchy (see Section 6). This association is denoted by ":=", as shown
1026 TSDL uses three different types of scoping: a lexical scope is used for
1027 declarations and type definitions, and static and dynamic scopes are
1028 used for variants references to tag fields (with relative and absolute
1029 path lookups) and for sequence references to length fields.
1033 Each of "trace", "env", "stream", "event", "struct" and "variant" have
1034 their own nestable declaration scope, within which types can be declared
1035 using "typedef" and "typealias". A root declaration scope also contains
1036 all declarations located outside of any of the aforementioned
1037 declarations. An inner declaration scope can refer to type declared
1038 within its container lexical scope prior to the inner declaration scope.
1039 Redefinition of a typedef or typealias is not valid, although hiding an
1040 upper scope typedef or typealias is allowed within a sub-scope.
1042 7.3.2 Static and Dynamic Scopes
1044 A local static scope consists in the scope generated by the declaration
1045 of fields within a compound type. A static scope is a local static scope
1046 augmented with the nested sub-static-scopes it contains.
1048 A dynamic scope consists in the static scope augmented with the
1049 implicit event structure definition hierarchy presented at Section 6.
1051 Multiple declarations of the same field name within a local static scope
1052 is not valid. It is however valid to re-use the same field name in
1053 different local scopes.
1055 Nested static and dynamic scopes form lookup paths. These are used for
1056 variant tag and sequence length references. They are used at the variant
1057 and sequence definition site to look up the location of the tag field
1058 associated with a variant, and to lookup up the location of the length
1059 field associated with a sequence.
1061 Variants and sequences can refer to a tag field either using a relative
1062 path or an absolute path. The relative path is relative to the scope in
1063 which the variant or sequence performing the lookup is located.
1064 Relative paths are only allowed to lookup within the same static scope,
1065 which includes its nested static scopes. Lookups targeting parent static
1066 scopes need to be performed with an absolute path.
1068 Absolute path lookups use the full path including the dynamic scope
1069 followed by a "." and then the static scope. Therefore, variants (or
1070 sequences) in lower levels in the dynamic scope (e.g. event context) can
1071 refer to a tag (or length) field located in upper levels (e.g. in the
1072 event header) by specifying, in this case, the associated tag with
1073 <stream.event.header.field_name>. This allows, for instance, the event
1074 context to define a variant referring to the "id" field of the event
1077 The dynamic scope prefixes are thus:
1079 - Trace Environment: <env. >,
1080 - Trace Packet Header: <trace.packet.header. >,
1081 - Stream Packet Context: <stream.packet.context. >,
1082 - Event Header: <stream.event.header. >,
1083 - Stream Event Context: <stream.event.context. >,
1084 - Event Context: <event.context. >,
1085 - Event Payload: <event.fields. >.
1088 The target dynamic scope must be specified explicitly when referring to
1089 a field outside of the static scope (absolute scope reference). No
1090 conflict can occur between relative and dynamic paths, because the
1091 keywords "trace", "stream", and "event" are reserved, and thus
1092 not permitted as field names. It is recommended that field names
1093 clashing with CTF and C99 reserved keywords use an underscore prefix to
1094 eliminate the risk of generating a description containing an invalid
1095 field name. Consequently, fields starting with an underscore should have
1096 their leading underscore removed by the CTF trace readers.
1099 The information available in the dynamic scopes can be thought of as the
1100 current tracing context. At trace production, information about the
1101 current context is saved into the specified scope field levels. At trace
1102 consumption, for each event, the current trace context is therefore
1103 readable by accessing the upper dynamic scopes.
1108 The grammar representing the TSDL meta-data is presented in Appendix C.
1109 TSDL Grammar. This section presents a rather lighter reading that
1110 consists in examples of TSDL meta-data, with template values.
1112 The stream "id" can be left out if there is only one stream in the
1113 trace. The event "id" field can be left out if there is only one event
1117 major = value; /* CTF spec version major number */
1118 minor = value; /* CTF spec version minor number */
1119 uuid = "aaaaaaaa-aaaa-aaaa-aaaa-aaaaaaaaaaaa"; /* Trace UUID */
1120 byte_order = be OR le; /* Endianness (required) */
1121 packet.header := struct {
1129 * The "env" (environment) scope contains assignment expressions. The
1130 * field names and content are implementation-defined.
1133 pid = value; /* example */
1134 proc_name = "name"; /* example */
1140 /* Type 1 - Few event IDs; Type 2 - Many event IDs. See section 6.1. */
1141 event.header := event_header_1 OR event_header_2;
1142 event.context := struct {
1145 packet.context := struct {
1151 name = "event_name";
1152 id = value; /* Numeric identifier within the stream */
1153 stream_id = stream_id;
1163 /* More detail on types in section 4. Types */
1168 * Type declarations behave similarly to the C standard.
1171 typedef aliased_type_specifiers new_type_declarators;
1173 /* e.g.: typedef struct example new_type_name[10]; */
1178 * The "typealias" declaration can be used to give a name (including
1179 * pointer declarator specifier) to a type. It should also be used to
1180 * map basic C types (float, int, unsigned long, ...) to a CTF type.
1181 * Typealias is a superset of "typedef": it also allows assignment of a
1182 * simple variable identifier to a type.
1185 typealias type_class {
1187 } := type_specifiers type_declarator;
1191 * typealias integer {
1195 * } := struct page *;
1197 * typealias integer {
1212 enum name : integer_type {
1218 * Unnamed types, contained within compound type fields, typedef or typealias.
1233 enum : integer_type {
1237 typedef type new_type[length];
1240 type field_name[length];
1243 typedef type new_type[length_type];
1246 type field_name[length_type];
1258 integer_type field_name:size; /* GNU/C bitfield */
1268 Clock metadata allows to describe the clock topology of the system, as
1269 well as to detail each clock parameter. In absence of clock description,
1270 it is assumed that all fields named "timestamp" use the same clock
1271 source, which increments once per nanosecond.
1273 Describing a clock and how it is used by streams is threefold: first,
1274 the clock and clock topology should be described in a "clock"
1275 description block, e.g.:
1278 name = cycle_counter_sync;
1279 uuid = "62189bee-96dc-11e0-91a8-cfa3d89f3923";
1280 description = "Cycle counter synchronized across CPUs";
1281 freq = 1000000000; /* frequency, in Hz */
1282 /* precision in seconds is: 1000 * (1/freq) */
1285 * clock value offset from Epoch is:
1286 * offset_s + (offset * (1/freq))
1288 offset_s = 1326476837;
1293 The mandatory "name" field specifies the name of the clock identifier,
1294 which can later be used as a reference. The optional field "uuid" is the
1295 unique identifier of the clock. It can be used to correlate different
1296 traces that use the same clock. An optional textual description string
1297 can be added with the "description" field. The "freq" field is the
1298 initial frequency of the clock, in Hz. If the "freq" field is not
1299 present, the frequency is assumed to be 1000000000 (providing clock
1300 increment of 1 ns). The optional "precision" field details the
1301 uncertainty on the clock measurements, in (1/freq) units. The "offset_s"
1302 and "offset" fields indicate the offset from POSIX.1 Epoch, 1970-01-01
1303 00:00:00 +0000 (UTC), to the zero of value of the clock. The "offset_s"
1304 field is in seconds. The "offset" field is in (1/freq) units. If any of
1305 the "offset_s" or "offset" field is not present, it is assigned the 0
1306 value. The field "absolute" is TRUE if the clock is a global reference
1307 across different clock uuid (e.g. NTP time). Otherwise, "absolute" is
1308 FALSE, and the clock can be considered as synchronized only with other
1309 clocks that have the same uuid.
1312 Secondly, a reference to this clock should be added within an integer
1316 size = 64; align = 1; signed = false;
1317 map = clock.cycle_counter_sync.value;
1320 Thirdly, stream declarations can reference the clock they use as a
1323 struct packet_context {
1324 uint64_ccnt_t ccnt_begin;
1325 uint64_ccnt_t ccnt_end;
1331 event.header := struct {
1332 uint64_ccnt_t timestamp;
1335 packet.context := struct packet_context;
1338 For a N-bit integer type referring to a clock, if the integer overflows
1339 compared to the N low order bits of the clock prior value, then it is
1340 assumed that one, and only one, overflow occurred. It is therefore
1341 important that events encoding time on a small number of bits happen
1342 frequently enough to detect when more than one N-bit overflow occurs.
1344 In a packet context, clock field names ending with "_begin" and "_end"
1345 have a special meaning: this refers to the time-stamps at, respectively,
1346 the beginning and the end of each packet.
1351 The two following macros keep track of the size of a GNU/C structure without
1352 padding at the end by placing HEADER_END as the last field. A one byte end field
1353 is used for C90 compatibility (C99 flexible arrays could be used here). Note
1354 that this does not affect the effective structure size, which should always be
1355 calculated with the header_sizeof() helper.
1357 #define HEADER_END char end_field
1358 #define header_sizeof(type) offsetof(typeof(type), end_field)
1361 B. Stream Header Rationale
1363 An event stream is divided in contiguous event packets of variable size. These
1364 subdivisions allow the trace analyzer to perform a fast binary search by time
1365 within the stream (typically requiring to index only the event packet headers)
1366 without reading the whole stream. These subdivisions have a variable size to
1367 eliminate the need to transfer the event packet padding when partially filled
1368 event packets must be sent when streaming a trace for live viewing/analysis.
1369 An event packet can contain a certain amount of padding at the end. Dividing
1370 streams into event packets is also useful for network streaming over UDP and
1371 flight recorder mode tracing (a whole event packet can be swapped out of the
1372 buffer atomically for reading).
1374 The stream header is repeated at the beginning of each event packet to allow
1375 flexibility in terms of:
1377 - streaming support,
1378 - allowing arbitrary buffers to be discarded without making the trace
1380 - allow UDP packet loss handling by either dealing with missing event packet
1381 or asking for re-transmission.
1382 - transparently support flight recorder mode,
1383 - transparently support crash dump.
1389 * Common Trace Format (CTF) Trace Stream Description Language (TSDL) Grammar.
1391 * Inspired from the C99 grammar:
1392 * http://www.open-std.org/jtc1/sc22/wg14/www/docs/n1124.pdf (Annex A)
1393 * and c++1x grammar (draft)
1394 * http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2011/n3291.pdf (Annex A)
1396 * Specialized for CTF needs by including only constant and declarations from
1397 * C99 (excluding function declarations), and by adding support for variants,
1398 * sequences and CTF-specific specifiers. Enumeration container types
1399 * semantic is inspired from c++1x enum-base.
1404 1.1) Lexical elements
1450 identifier identifier-nondigit
1453 identifier-nondigit:
1455 universal-character-name
1456 any other implementation-defined characters
1460 [a-zA-Z] /* regular expression */
1463 [0-9] /* regular expression */
1465 1.4) Universal character names
1467 universal-character-name:
1469 \U hex-quad hex-quad
1472 hexadecimal-digit hexadecimal-digit hexadecimal-digit hexadecimal-digit
1478 enumeration-constant
1482 decimal-constant integer-suffix-opt
1483 octal-constant integer-suffix-opt
1484 hexadecimal-constant integer-suffix-opt
1488 decimal-constant digit
1492 octal-constant octal-digit
1494 hexadecimal-constant:
1495 hexadecimal-prefix hexadecimal-digit
1496 hexadecimal-constant hexadecimal-digit
1506 unsigned-suffix long-suffix-opt
1507 unsigned-suffix long-long-suffix
1508 long-suffix unsigned-suffix-opt
1509 long-long-suffix unsigned-suffix-opt
1523 enumeration-constant:
1529 L' c-char-sequence '
1533 c-char-sequence c-char
1536 any member of source charset except single-quote ('), backslash
1537 (\), or new-line character.
1541 simple-escape-sequence
1542 octal-escape-sequence
1543 hexadecimal-escape-sequence
1544 universal-character-name
1546 simple-escape-sequence: one of
1547 \' \" \? \\ \a \b \f \n \r \t \v
1549 octal-escape-sequence:
1551 \ octal-digit octal-digit
1552 \ octal-digit octal-digit octal-digit
1554 hexadecimal-escape-sequence:
1555 \x hexadecimal-digit
1556 hexadecimal-escape-sequence hexadecimal-digit
1558 1.6) String literals
1561 " s-char-sequence-opt "
1562 L" s-char-sequence-opt "
1566 s-char-sequence s-char
1569 any member of source charset except double-quote ("), backslash
1570 (\), or new-line character.
1576 [ ] ( ) { } . -> * + - < > : ; ... = ,
1579 2) Phrase structure grammar
1585 ( unary-expression )
1589 postfix-expression [ unary-expression ]
1590 postfix-expression . identifier
1591 postfix-expressoin -> identifier
1595 unary-operator postfix-expression
1597 unary-operator: one of
1600 assignment-operator:
1603 type-assignment-operator:
1606 constant-expression-range:
1607 unary-expression ... unary-expression
1612 declaration-specifiers declarator-list-opt ;
1615 declaration-specifiers:
1616 storage-class-specifier declaration-specifiers-opt
1617 type-specifier declaration-specifiers-opt
1618 type-qualifier declaration-specifiers-opt
1622 declarator-list , declarator
1624 abstract-declarator-list:
1626 abstract-declarator-list , abstract-declarator
1628 storage-class-specifier:
1651 align ( unary-expression )
1654 struct identifier-opt { struct-or-variant-declaration-list-opt } align-attribute-opt
1655 struct identifier align-attribute-opt
1657 struct-or-variant-declaration-list:
1658 struct-or-variant-declaration
1659 struct-or-variant-declaration-list struct-or-variant-declaration
1661 struct-or-variant-declaration:
1662 specifier-qualifier-list struct-or-variant-declarator-list ;
1663 declaration-specifiers-opt storage-class-specifier declaration-specifiers-opt declarator-list ;
1664 typealias declaration-specifiers abstract-declarator-list type-assignment-operator declaration-specifiers abstract-declarator-list ;
1665 typealias declaration-specifiers abstract-declarator-list type-assignment-operator declarator-list ;
1667 specifier-qualifier-list:
1668 type-specifier specifier-qualifier-list-opt
1669 type-qualifier specifier-qualifier-list-opt
1671 struct-or-variant-declarator-list:
1672 struct-or-variant-declarator
1673 struct-or-variant-declarator-list , struct-or-variant-declarator
1675 struct-or-variant-declarator:
1677 declarator-opt : unary-expression
1680 variant identifier-opt variant-tag-opt { struct-or-variant-declaration-list }
1681 variant identifier variant-tag
1684 < unary-expression >
1687 enum identifier-opt { enumerator-list }
1688 enum identifier-opt { enumerator-list , }
1690 enum identifier-opt : declaration-specifiers { enumerator-list }
1691 enum identifier-opt : declaration-specifiers { enumerator-list , }
1695 enumerator-list , enumerator
1698 enumeration-constant
1699 enumeration-constant assignment-operator unary-expression
1700 enumeration-constant assignment-operator constant-expression-range
1706 pointer-opt direct-declarator
1711 direct-declarator [ unary-expression ]
1713 abstract-declarator:
1714 pointer-opt direct-abstract-declarator
1716 direct-abstract-declarator:
1718 ( abstract-declarator )
1719 direct-abstract-declarator [ unary-expression ]
1720 direct-abstract-declarator [ ]
1723 * type-qualifier-list-opt
1724 * type-qualifier-list-opt pointer
1726 type-qualifier-list:
1728 type-qualifier-list type-qualifier
1733 2.3) CTF-specific declarations
1736 clock { ctf-assignment-expression-list-opt }
1737 event { ctf-assignment-expression-list-opt }
1738 stream { ctf-assignment-expression-list-opt }
1739 env { ctf-assignment-expression-list-opt }
1740 trace { ctf-assignment-expression-list-opt }
1741 typealias declaration-specifiers abstract-declarator-list type-assignment-operator declaration-specifiers abstract-declarator-list
1742 typealias declaration-specifiers abstract-declarator-list type-assignment-operator declarator-list
1745 floating_point { ctf-assignment-expression-list-opt }
1746 integer { ctf-assignment-expression-list-opt }
1747 string { ctf-assignment-expression-list-opt }
1750 ctf-assignment-expression-list:
1751 ctf-assignment-expression ;
1752 ctf-assignment-expression-list ctf-assignment-expression ;
1754 ctf-assignment-expression:
1755 unary-expression assignment-operator unary-expression
1756 unary-expression type-assignment-operator type-specifier
1757 declaration-specifiers-opt storage-class-specifier declaration-specifiers-opt declarator-list
1758 typealias declaration-specifiers abstract-declarator-list type-assignment-operator declaration-specifiers abstract-declarator-list
1759 typealias declaration-specifiers abstract-declarator-list type-assignment-operator declarator-list