1PERF-ANNOTATE(1) perf Manual PERF-ANNOTATE(1)
2
6 perf-annotate - Read perf.data (created by perf record) and display
7 annotated code
8
10 perf annotate [-i <file> | --input=file] [symbol_name]
11
13 This command reads the input file and displays an annotated version of
14 the code. If the object file has debug symbols then the source code
15 will be displayed alongside assembly code.
16 If there is no debug info in the object, then annotated assembly is
18 displayed.
19
21 -i, --input=<file>
22 Input file name. (default: perf.data unless stdin is a fifo)
23 -d, --dsos=<dso[,dso...]>
25 Only consider symbols in these dsos.
26 -s, --symbol=<symbol>
28 Symbol to annotate.
29 -f, --force
31 Don’t do ownership validation.
32 -v, --verbose
34 Be more verbose. (Show symbol address, etc)
35 -q, --quiet
37 Do not show any warnings or messages. (Suppress -v)
38 -n, --show-nr-samples
40 Show the number of samples for each symbol
41 -D, --dump-raw-trace
43 Dump raw trace in ASCII.
44 -k, --vmlinux=<file>
46 vmlinux pathname.
47 --ignore-vmlinux
49 Ignore vmlinux files.
50 --itrace
52 Options for decoding instruction tracing data. The options are:
53 i synthesize instructions events
55 y synthesize cycles events
56 b synthesize branches events (branch misses for Arm SPE)
57 c synthesize branches events (calls only)
58 r synthesize branches events (returns only)
59 x synthesize transactions events
60 w synthesize ptwrite events
61 p synthesize power events (incl. PSB events for Intel PT)
62 o synthesize other events recorded due to the use
63 of aux-output (refer to perf record)
64 I synthesize interrupt or similar (asynchronous) events
65 (e.g. Intel PT Event Trace)
66 e synthesize error events
67 d create a debug log
68 f synthesize first level cache events
69 m synthesize last level cache events
70 M synthesize memory events
71 t synthesize TLB events
72 a synthesize remote access events
73 g synthesize a call chain (use with i or x)
74 G synthesize a call chain on existing event records
75 l synthesize last branch entries (use with i or x)
76 L synthesize last branch entries on existing event records
77 s skip initial number of events
78 q quicker (less detailed) decoding
79 A approximate IPC
80 Z prefer to ignore timestamps (so-called "timeless" decoding)
81 The default is all events i.e. the same as --itrace=iybxwpe,
83 except for perf script where it is --itrace=ce
84 In addition, the period (default 100000, except for perf script where it is 1)
86 for instructions events can be specified in units of:
87 i instructions
89 t ticks
90 ms milliseconds
91 us microseconds
92 ns nanoseconds (default)
93 Also the call chain size (default 16, max. 1024) for instructions or
95 transactions events can be specified.
96 Also the number of last branch entries (default 64, max. 1024) for
98 instructions or transactions events can be specified.
99 Similar to options g and l, size may also be specified for options G and L.
101 On x86, note that G and L work poorly when data has been recorded with
102 large PEBS. Refer linkperf:perf-intel-pt[1] man page for details.
103 It is also possible to skip events generated (instructions, branches, transactions,
105 ptwrite, power) at the beginning. This is useful to ignore initialization code.
106 --itrace=i0nss1000000
108 skips the first million instructions.
110 The 'e' option may be followed by flags which affect what errors will or
112 will not be reported. Each flag must be preceded by either '+' or '-'.
113 The flags are:
114 o overflow
115 l trace data lost
116 If supported, the 'd' option may be followed by flags which affect what
118 debug messages will or will not be logged. Each flag must be preceded
119 by either '+' or '-'. The flags are:
120 a all perf events
121 e output only on errors (size configurable - see linkperf:perf-config[1])
122 o output to stdout
123 If supported, the 'q' option may be repeated to increase the effect.
125 To disable decoding entirely, use --no-itrace.
127 -m, --modules
129 Load module symbols. WARNING: use only with -k and LIVE kernel.
130 -l, --print-line
132 Print matching source lines (may be slow).
133 -P, --full-paths
135 Don’t shorten the displayed pathnames.
136 --stdio
138 Use the stdio interface.
139 --stdio2
141 Use the stdio2 interface, non-interactive, uses the TUI formatting.
142 --stdio-color=<mode>
144 always, never or auto, allowing configuring color output via the
145 command line, in addition to via "color.ui" .perfconfig. Use
146 --stdio-color always to generate color even when redirecting to a
147 pipe or file. Using just --stdio-color is equivalent to using
148 always.
149 --tui
151 Use the TUI interface. Use of --tui requires a tty, if one is not
152 present, as when piping to other commands, the stdio interface is
153 used. This interfaces starts by centering on the line with more
154 samples, TAB/UNTAB cycles through the lines with more samples.
155 --gtk
157 Use the GTK interface.
158 -C, --cpu=<cpu>
160 Only report samples for the list of CPUs provided. Multiple CPUs
161 can be provided as a comma-separated list with no space: 0,1.
162 Ranges of CPUs are specified with -: 0-2. Default is to report
163 samples on all CPUs.
164 --asm-raw
166 Show raw instruction encoding of assembly instructions.
167 --show-total-period
169 Show a column with the sum of periods.
170 --source
172 Interleave source code with assembly code. Enabled by default,
173 disable with --no-source.
174 --symfs=<directory>
176 Look for files with symbols relative to this directory.
177 -M, --disassembler-style=
179 Set disassembler style for objdump.
180 --addr2line=<path>
182 Path to addr2line binary.
183 --objdump=<path>
185 Path to objdump binary.
186 --prefix=PREFIX, --prefix-strip=N
188 Remove first N entries from source file path names in executables
189 and add PREFIX. This allows to display source code compiled on
190 systems with different file system layout.
191 --skip-missing
193 Skip symbols that cannot be annotated.
194 --group
196 Show event group information together
197 --demangle
199 Demangle symbol names to human readable form. It’s enabled by
200 default, disable with --no-demangle.
201 --demangle-kernel
203 Demangle kernel symbol names to human readable form (for C++
204 kernels).
205 --percent-type
207 Set annotation percent type from following choices: global-period,
208 local-period, global-hits, local-hits
209 The local/global keywords set if the percentage is computed
211 in the scope of the function (local) or the whole data (global).
212 The period/hits keywords set the base the percentage is computed
213 on - the samples period or the number of samples (hits).
214 --percent-limit
216 Do not show functions which have an overhead under that percent on
217 stdio or stdio2 (Default: 0). Note that this is about selection of
218 functions to display, not about lines within the function.
219
221 perf-record(1), perf-report(1)
222perf 11/28/2023 PERF-ANNOTATE(1)