1JITDUMP specification version 2
2Last Revised: 09/15/2016
3Author: Stephane Eranian <eranian@gmail.com>
4
5--------------------------------------------------------
6| Revision  |    Date    | Description                 |
7--------------------------------------------------------
8|   1       | 09/07/2016 | Initial revision            |
9--------------------------------------------------------
10|   2       | 09/15/2016 | Add JIT_CODE_UNWINDING_INFO |
11--------------------------------------------------------
12
13
14I/ Introduction
15
16
17This document describes the jitdump file format. The file is generated by Just-In-time compiler runtimes to save meta-data information about the generated code, such as address, size, and name of generated functions, the native code generated, the source line information. The data may then be used by performance tools, such as Linux perf to generate function and assembly level profiles.
18
19The format is not specific to any particular programming language. It can be extended as need be.
20
21The format of the file is binary. It is self-describing in terms of endianness and is portable across multiple processor architectures.
22
23
24II/ Overview of the format
25
26
27The format requires only sequential accesses, i.e., append only mode. The file starts with a fixed size file header describing the version of the specification, the endianness.
28
29The header is followed by a series of records, each starting with a fixed size header describing the type of record and its size. It is, itself, followed by the payload for the record. Records can have a variable size even for a given type.
30
31Each entry in the file is timestamped. All timestamps must use the same clock source. The CLOCK_MONOTONIC clock source is recommended.
32
33
34III/ Jitdump file header format
35
36Each jitdump file starts with a fixed size header containing the following fields in order:
37
38
39* uint32_t magic     : a magic number tagging the file type. The value is 4-byte long and represents the string "JiTD" in ASCII form. It is 0x4A695444 or 0x4454694a depending on the endianness. The field can be used to detect the endianness of the file
40* uint32_t version   : a 4-byte value representing the format version. It is currently set to 2
41* uint32_t total_size: size in bytes of file header
42* uint32_t elf_mach  : ELF architecture encoding (ELF e_machine value as specified in /usr/include/elf.h)
43* uint32_t pad1      : padding. Reserved for future use
44* uint32_t pid       : JIT runtime process identification (OS specific)
45* uint64_t timestamp : timestamp of when the file was created
46* uint64_t flags     : a bitmask of flags
47
48The flags currently defined are as follows:
49 * bit 0: JITDUMP_FLAGS_ARCH_TIMESTAMP : set if the jitdump file is using an architecture-specific timestamp clock source. For instance, on x86, one could use TSC directly
50
51IV/ Record header
52
53The file header is immediately followed by records. Each record starts with a fixed size header describing the record that follows.
54
55The record header is specified in order as follows:
56* uint32_t id        : a value identifying the record type (see below)
57* uint32_t total_size: the size in bytes of the record including the header.
58* uint64_t timestamp : a timestamp of when the record was created.
59
60The following record types are defined:
61 * Value 0 : JIT_CODE_LOAD      : record describing a jitted function
62 * Value 1 : JIT_CODE_MOVE      : record describing an already jitted function which is moved
63 * Value 2 : JIT_CODE_DEBUG_INFO: record describing the debug information for a jitted function
64 * Value 3 : JIT_CODE_CLOSE     : record marking the end of the jit runtime (optional)
65 * Value 4 : JIT_CODE_UNWINDING_INFO: record describing a function unwinding information
66
67 The payload of the record must immediately follow the record header without padding.
68
69V/ JIT_CODE_LOAD record
70
71
72  The record has the following fields following the fixed-size record header in order:
73  * uint32_t pid: OS process id of the runtime generating the jitted code
74  * uint32_t tid: OS thread identification of the runtime thread generating the jitted code
75  * uint64_t vma: virtual address of jitted code start
76  * uint64_t code_addr: code start address for the jitted code. By default vma = code_addr
77  * uint64_t code_size: size in bytes of the generated jitted code
78  * uint64_t code_index: unique identifier for the jitted code (see below)
79  * char[n]: function name in ASCII including the null termination
80  * native code: raw byte encoding of the jitted code
81
82  The record header total_size field is inclusive of all components:
83  * record header
84  * fixed-sized fields
85  * function name string, including termination
86  * native code length
87  * record specific variable data (e.g., array of data entries)
88
89The code_index is used to uniquely identify each jitted function. The index can be a monotonically increasing 64-bit value. Each time a function is jitted it gets a new number. This value is used in case the code for a function is moved and avoids having to issue another JIT_CODE_LOAD record.
90
91The format supports empty functions with no native code.
92
93
94VI/ JIT_CODE_MOVE record
95
96  The record type is optional.
97
98  The record has the following fields following the fixed-size record header in order:
99  * uint32_t pid          : OS process id of the runtime generating the jitted code
100  * uint32_t tid          : OS thread identification of the runtime thread generating the jitted code
101  * uint64_t vma          : new virtual address of jitted code start
102  * uint64_t old_code_addr: previous code address for the same function
103  * uint64_t new_code_addr: alternate new code started address for the jitted code. By default it should be equal to the vma address.
104  * uint64_t code_size    : size in bytes of the jitted code
105  * uint64_t code_index   : index referring to the JIT_CODE_LOAD code_index record of when the function was initially jitted
106
107
108The MOVE record can be used in case an already jitted function is simply moved by the runtime inside the code cache.
109
110The JIT_CODE_MOVE record cannot come before the JIT_CODE_LOAD record for the same function name. The function cannot have changed name, otherwise a new JIT_CODE_LOAD record must be emitted.
111
112The code size of the function cannot change.
113
114
115VII/ JIT_DEBUG_INFO record
116
117The record type is optional.
118
119The record contains source lines debug information, i.e., a way to map a code address back to a source line. This information may be used by the performance tool.
120
121The record has the following fields following the fixed-size record header in order:
122  * uint64_t code_addr: address of function for which the debug information is generated
123  * uint64_t nr_entry : number of debug entries for the function
124  * debug_entry[n]: array of nr_entry debug entries for the function
125
126The debug_entry describes the source line information. It is defined as follows in order:
127* uint64_t code_addr: address of function for which the debug information is generated
128* uint32_t line     : source file line number (starting at 1)
129* uint32_t discrim  : column discriminator, 0 is default
130* char name[n]      : source file name in ASCII, including null termination
131
132The debug_entry entries are saved in sequence but given that they have variable sizes due to the file name string, they cannot be indexed directly.
133They need to be walked sequentially. The next debug_entry is found at sizeof(debug_entry) + strlen(name) + 1.
134
135IMPORTANT:
136  The JIT_CODE_DEBUG for a given function must always be generated BEFORE the JIT_CODE_LOAD for the function. This facilitates greatly the parser for the jitdump file.
137
138
139VIII/ JIT_CODE_CLOSE record
140
141
142The record type is optional.
143
144The record is used as a marker for the end of the jitted runtime. It can be replaced by the end of the file.
145
146The JIT_CODE_CLOSE record does not have any specific fields, the record header contains all the information needed.
147
148
149IX/ JIT_CODE_UNWINDING_INFO
150
151
152The record type is optional.
153
154The record is used to describe the unwinding information for a jitted function.
155
156The record has the following fields following the fixed-size record header in order:
157
158uint64_t unwind_data_size   : the size in bytes of the unwinding data table at the end of the record
159uint64_t eh_frame_hdr_size  : the size in bytes of the DWARF EH Frame Header at the start of the unwinding data table at the end of the record
160uint64_t mapped_size        : the size of the unwinding data mapped in memory
161const char unwinding_data[n]: an array of unwinding data, consisting of the EH Frame Header, followed by the actual EH Frame
162
163
164The EH Frame header follows the Linux Standard Base (LSB) specification as described in the document at https://refspecs.linuxfoundation.org/LSB_1.3.0/gLSB/gLSB/ehframehdr.html
165
166
167The EH Frame follows the LSB specicfication as described in the document at https://refspecs.linuxbase.org/LSB_3.0.0/LSB-PDA/LSB-PDA/ehframechpt.html
168
169
170NOTE: The mapped_size is generally either the same as unwind_data_size (if the unwinding data was mapped in memory by the running process) or zero (if the unwinding data is not mapped by the process). If the unwinding data was not mapped, then only the EH Frame Header will be read, which can be used to specify FP based unwinding for a function which does not have unwinding information.
171