1 perf-config(1) 1 perf-config(1)
2 ============== 2 ==============
3 3
4 NAME 4 NAME
5 ---- 5 ----
6 perf-config - Get and set variables in a confi 6 perf-config - Get and set variables in a configuration file.
7 7
8 SYNOPSIS 8 SYNOPSIS
9 -------- 9 --------
10 [verse] 10 [verse]
11 'perf config' [<file-option>] [section.name[=v 11 'perf config' [<file-option>] [section.name[=value] ...]
12 or 12 or
13 'perf config' [<file-option>] -l | --list 13 'perf config' [<file-option>] -l | --list
14 14
15 DESCRIPTION 15 DESCRIPTION
16 ----------- 16 -----------
17 You can manage variables in a configuration fi 17 You can manage variables in a configuration file with this command.
18 18
19 OPTIONS 19 OPTIONS
20 ------- 20 -------
21 21
22 -l:: 22 -l::
23 --list:: 23 --list::
24 Show current config variables, name an 24 Show current config variables, name and value, for all sections.
25 25
26 --user:: 26 --user::
27 For writing and reading options: write 27 For writing and reading options: write to user
28 '$HOME/.perfconfig' file or read it. 28 '$HOME/.perfconfig' file or read it.
29 29
30 --system:: 30 --system::
31 For writing and reading options: write 31 For writing and reading options: write to system-wide
32 '$(sysconfdir)/perfconfig' or read it. 32 '$(sysconfdir)/perfconfig' or read it.
33 33
34 CONFIGURATION FILE 34 CONFIGURATION FILE
35 ------------------ 35 ------------------
36 36
37 The perf configuration file contains many vari 37 The perf configuration file contains many variables to change various
38 aspects of each of its tools, including output 38 aspects of each of its tools, including output, disk usage, etc.
39 The '$HOME/.perfconfig' file is used to store 39 The '$HOME/.perfconfig' file is used to store a per-user configuration.
40 The file '$(sysconfdir)/perfconfig' can be use 40 The file '$(sysconfdir)/perfconfig' can be used to
41 store a system-wide default configuration. 41 store a system-wide default configuration.
42 42
43 One an disable reading config files by setting 43 One an disable reading config files by setting the PERF_CONFIG environment
44 variable to /dev/null, or provide an alternate 44 variable to /dev/null, or provide an alternate config file by setting that
45 variable. 45 variable.
46 46
47 When reading or writing, the values are read f 47 When reading or writing, the values are read from the system and user
48 configuration files by default, and options '- 48 configuration files by default, and options '--system' and '--user'
49 can be used to tell the command to read from o 49 can be used to tell the command to read from or write to only that location.
50 50
51 Syntax 51 Syntax
52 ~~~~~~ 52 ~~~~~~
53 53
54 The file consist of sections. A section starts 54 The file consist of sections. A section starts with its name
55 surrounded by square brackets and continues ti 55 surrounded by square brackets and continues till the next section
56 begins. Each variable must be in a section, an 56 begins. Each variable must be in a section, and have the form
57 'name = value', for example: 57 'name = value', for example:
58 58
59 [section] 59 [section]
60 name1 = value1 60 name1 = value1
61 name2 = value2 61 name2 = value2
62 62
63 Section names are case sensitive and can conta 63 Section names are case sensitive and can contain any characters except
64 newline (double quote `"` and backslash have t 64 newline (double quote `"` and backslash have to be escaped as `\"` and `\\`,
65 respectively). Section headers can't span mult 65 respectively). Section headers can't span multiple lines.
66 66
67 Example 67 Example
68 ~~~~~~~ 68 ~~~~~~~
69 69
70 Given a $HOME/.perfconfig like this: 70 Given a $HOME/.perfconfig like this:
71 71
72 # 72 #
73 # This is the config file, and 73 # This is the config file, and
74 # a '#' and ';' character indicates a comment 74 # a '#' and ';' character indicates a comment
75 # 75 #
76 76
77 [colors] 77 [colors]
78 # Color variables 78 # Color variables
79 top = red, default 79 top = red, default
80 medium = green, default 80 medium = green, default
81 normal = lightgray, default 81 normal = lightgray, default
82 selected = white, lightgray 82 selected = white, lightgray
83 jump_arrows = blue, default 83 jump_arrows = blue, default
84 addr = magenta, default 84 addr = magenta, default
85 root = white, blue 85 root = white, blue
86 86
87 [tui] 87 [tui]
88 # Defaults if linked with libs 88 # Defaults if linked with libslang
89 report = on 89 report = on
90 annotate = on 90 annotate = on
91 top = on 91 top = on
92 92
93 [buildid] 93 [buildid]
94 # Default, disable using /dev/ 94 # Default, disable using /dev/null
95 dir = ~/.debug 95 dir = ~/.debug
96 96
97 [annotate] 97 [annotate]
98 # Defaults 98 # Defaults
99 hide_src_code = false 99 hide_src_code = false
100 use_offset = true 100 use_offset = true
101 jump_arrows = true 101 jump_arrows = true
102 show_nr_jumps = false 102 show_nr_jumps = false
103 103
104 [help] 104 [help]
105 # Format can be man, info, web 105 # Format can be man, info, web or html
106 format = man 106 format = man
107 autocorrect = 0 107 autocorrect = 0
108 108
109 [ui] 109 [ui]
110 show-headers = true 110 show-headers = true
111 111
112 [call-graph] 112 [call-graph]
113 # fp (framepointer), dwarf 113 # fp (framepointer), dwarf
114 record-mode = fp 114 record-mode = fp
115 print-type = graph 115 print-type = graph
116 order = caller 116 order = caller
117 sort-key = function 117 sort-key = function
118 118
119 [report] 119 [report]
120 # Defaults 120 # Defaults
121 sort_order = comm,dso,symbol 121 sort_order = comm,dso,symbol
122 percent-limit = 0 122 percent-limit = 0
123 queue-size = 0 123 queue-size = 0
124 children = true 124 children = true
125 group = true 125 group = true
126 skip-empty = true 126 skip-empty = true
127 127
>> 128 [llvm]
>> 129 dump-obj = true
>> 130 clang-opt = -g
128 131
129 You can hide source code of annotate feature s 132 You can hide source code of annotate feature setting the config to false with
130 133
131 % perf config annotate.hide_src_code=t 134 % perf config annotate.hide_src_code=true
132 135
133 If you want to add or modify several config it 136 If you want to add or modify several config items, you can do like
134 137
135 % perf config ui.show-headers=false km 138 % perf config ui.show-headers=false kmem.default=slab
136 139
137 To modify the sort order of report functionali 140 To modify the sort order of report functionality in user config file(i.e. `~/.perfconfig`), do
138 141
139 % perf config --user report.sort-order 142 % perf config --user report.sort-order=srcline
140 143
141 To change colors of selected line to other for 144 To change colors of selected line to other foreground and background colors
142 in system config file (i.e. `$(sysconf)/perfco 145 in system config file (i.e. `$(sysconf)/perfconfig`), do
143 146
144 % perf config --system colors.selected 147 % perf config --system colors.selected=yellow,green
145 148
146 To query the record mode of call graph, do 149 To query the record mode of call graph, do
147 150
148 % perf config call-graph.record-mode 151 % perf config call-graph.record-mode
149 152
150 If you want to know multiple config key/value 153 If you want to know multiple config key/value pairs, you can do like
151 154
152 % perf config report.queue-size call-g 155 % perf config report.queue-size call-graph.order report.children
153 156
154 To query the config value of sort order of cal 157 To query the config value of sort order of call graph in user config file (i.e. `~/.perfconfig`), do
155 158
156 % perf config --user call-graph.sort-o 159 % perf config --user call-graph.sort-order
157 160
158 To query the config value of buildid directory 161 To query the config value of buildid directory in system config file (i.e. `$(sysconf)/perfconfig`), do
159 162
160 % perf config --system buildid.dir 163 % perf config --system buildid.dir
161 164
162 Variables 165 Variables
163 ~~~~~~~~~ 166 ~~~~~~~~~
164 167
165 colors.*:: 168 colors.*::
166 The variables for customizing the colo 169 The variables for customizing the colors used in the output for the
167 'report', 'top' and 'annotate' in the 170 'report', 'top' and 'annotate' in the TUI. They should specify the
168 foreground and background colors, sepa 171 foreground and background colors, separated by a comma, for example:
169 172
170 medium = green, lightgray 173 medium = green, lightgray
171 174
172 If you want to use the color configure 175 If you want to use the color configured for you terminal, just leave it
173 as 'default', for example: 176 as 'default', for example:
174 177
175 medium = default, lightgray 178 medium = default, lightgray
176 179
177 Available colors: 180 Available colors:
178 red, yellow, green, cyan, gray, black, 181 red, yellow, green, cyan, gray, black, blue,
179 white, default, magenta, lightgray 182 white, default, magenta, lightgray
180 183
181 colors.top:: 184 colors.top::
182 'top' means a overhead percent 185 'top' means a overhead percentage which is more than 5%.
183 And values of this variable sp 186 And values of this variable specify percentage colors.
184 Basic key values are foregroun 187 Basic key values are foreground-color 'red' and
185 background-color 'default'. 188 background-color 'default'.
186 colors.medium:: 189 colors.medium::
187 'medium' means a overhead perc 190 'medium' means a overhead percentage which has more than 0.5%.
188 Default values are 'green' and 191 Default values are 'green' and 'default'.
189 colors.normal:: 192 colors.normal::
190 'normal' means the rest of ove 193 'normal' means the rest of overhead percentages
191 except 'top', 'medium', 'selec 194 except 'top', 'medium', 'selected'.
192 Default values are 'lightgray' 195 Default values are 'lightgray' and 'default'.
193 colors.selected:: 196 colors.selected::
194 This selects the colors for th 197 This selects the colors for the current entry in a list of entries
195 from sub-commands (top, report 198 from sub-commands (top, report, annotate).
196 Default values are 'black' and 199 Default values are 'black' and 'lightgray'.
197 colors.jump_arrows:: 200 colors.jump_arrows::
198 Colors for jump arrows on asse 201 Colors for jump arrows on assembly code listings
199 such as 'jns', 'jmp', 'jane', 202 such as 'jns', 'jmp', 'jane', etc.
200 Default values are 'blue', 'de 203 Default values are 'blue', 'default'.
201 colors.addr:: 204 colors.addr::
202 This selects colors for addres 205 This selects colors for addresses from 'annotate'.
203 Default values are 'magenta', 206 Default values are 'magenta', 'default'.
204 colors.root:: 207 colors.root::
205 Colors for headers in the outp 208 Colors for headers in the output of a sub-commands (top, report).
206 Default values are 'white', 'b 209 Default values are 'white', 'blue'.
207 210
208 core.*:: 211 core.*::
209 core.proc-map-timeout:: 212 core.proc-map-timeout::
210 Sets a timeout (in millisecond 213 Sets a timeout (in milliseconds) for parsing /proc/<pid>/maps files.
211 Can be overridden by the --pro 214 Can be overridden by the --proc-map-timeout option on supported
212 subcommands. The default timeo 215 subcommands. The default timeout is 500ms.
213 216
214 tui.*, gtk.*:: 217 tui.*, gtk.*::
215 Subcommands that can be configured her 218 Subcommands that can be configured here are 'top', 'report' and 'annotate'.
216 These values are booleans, for example 219 These values are booleans, for example:
217 220
218 [tui] 221 [tui]
219 top = true 222 top = true
220 223
221 will make the TUI be the default for t 224 will make the TUI be the default for the 'top' subcommand. Those will be
222 available if the required libs were de 225 available if the required libs were detected at tool build time.
223 226
224 buildid.*:: 227 buildid.*::
225 buildid.dir:: 228 buildid.dir::
226 Each executable and shared lib 229 Each executable and shared library in modern distributions comes with a
227 content based identifier that, 230 content based identifier that, if available, will be inserted in a
228 'perf.data' file header to, at 231 'perf.data' file header to, at analysis time find what is needed to do
229 symbol resolution, code annota 232 symbol resolution, code annotation, etc.
230 233
231 The recording tools also store 234 The recording tools also stores a hard link or copy in a per-user
232 directory, $HOME/.debug/, of b 235 directory, $HOME/.debug/, of binaries, shared libraries, /proc/kallsyms
233 and /proc/kcore files to be us 236 and /proc/kcore files to be used at analysis time.
234 237
235 The buildid.dir variable can b 238 The buildid.dir variable can be used to either change this directory
236 cache location, or to disable 239 cache location, or to disable it altogether. If you want to disable it,
237 set buildid.dir to /dev/null. 240 set buildid.dir to /dev/null. The default is $HOME/.debug
238 241
239 buildid-cache.*:: 242 buildid-cache.*::
240 buildid-cache.debuginfod=URLs 243 buildid-cache.debuginfod=URLs
241 Specify debuginfod URLs to be 244 Specify debuginfod URLs to be used when retrieving perf.data binaries,
242 it follows the same syntax as 245 it follows the same syntax as the DEBUGINFOD_URLS variable, like:
243 246
244 buildid-cache.debuginfod=htt 247 buildid-cache.debuginfod=http://192.168.122.174:8002
245 248
246 annotate.*:: 249 annotate.*::
247 These are in control of addresses, jum 250 These are in control of addresses, jump function, source code
248 in lines of assembly code from a speci 251 in lines of assembly code from a specific program.
249 252
250 annotate.addr2line:: 253 annotate.addr2line::
251 addr2line binary to use for fi 254 addr2line binary to use for file names and line numbers.
252 255
253 annotate.objdump:: 256 annotate.objdump::
254 objdump binary to use for disa !! 257 objdump binary to use for disassembly and annotations.
255 including in the 'perf test' c <<
256 258
257 annotate.disassembler_style:: 259 annotate.disassembler_style::
258 Use this to change the default 260 Use this to change the default disassembler style to some other value
259 supported by binutils, such as 261 supported by binutils, such as "intel", see the '-M' option help in the
260 'objdump' man page. 262 'objdump' man page.
261 263
262 annotate.hide_src_code:: 264 annotate.hide_src_code::
263 If a program which is analyzed 265 If a program which is analyzed has source code,
264 this option lets 'annotate' pr 266 this option lets 'annotate' print a list of assembly code with the source code.
265 For example, let's see a part 267 For example, let's see a part of a program. There're four lines.
266 If this option is 'true', they 268 If this option is 'true', they can be printed
267 without source code from a pro 269 without source code from a program as below.
268 270
269 │ push %rbp 271 │ push %rbp
270 │ mov %rsp,%rbp 272 │ mov %rsp,%rbp
271 │ sub $0x10,%rsp 273 │ sub $0x10,%rsp
272 │ mov (%rdi),%rdx 274 │ mov (%rdi),%rdx
273 275
274 But if this option is 'false', 276 But if this option is 'false', source code of the part
275 can be also printed as below. 277 can be also printed as below. Default is 'false'.
276 278
277 │ struct rb_node *rb_ne 279 │ struct rb_node *rb_next(const struct rb_node *node)
278 │ { 280 │ {
279 │ push %rbp 281 │ push %rbp
280 │ mov %rsp,%rbp 282 │ mov %rsp,%rbp
281 │ sub $0x10,%rsp 283 │ sub $0x10,%rsp
282 │ struct rb_nod 284 │ struct rb_node *parent;
283 │ 285 │
284 │ if (RB_EMPTY_ 286 │ if (RB_EMPTY_NODE(node))
285 │ mov (%rdi),%rdx 287 │ mov (%rdi),%rdx
286 │ return n; 288 │ return n;
287 289
288 This option works with tui, st 290 This option works with tui, stdio2 browsers.
289 291
290 annotate.use_offset:: 292 annotate.use_offset::
291 Basing on a first address of a 293 Basing on a first address of a loaded function, offset can be used.
292 Instead of using original addr 294 Instead of using original addresses of assembly code,
293 addresses subtracted from a ba 295 addresses subtracted from a base address can be printed.
294 Let's illustrate an example. 296 Let's illustrate an example.
295 If a base address is 0XFFFFFFF 297 If a base address is 0XFFFFFFFF81624d50 as below,
296 298
297 ffffffff81624d50 <load0> 299 ffffffff81624d50 <load0>
298 300
299 an address on assembly code ha 301 an address on assembly code has a specific absolute address as below
300 302
301 ffffffff816250b8:│ mov 0 303 ffffffff816250b8:│ mov 0x8(%r14),%rdi
302 304
303 but if use_offset is 'true', a 305 but if use_offset is 'true', an address subtracted from a base address is printed.
304 Default is true. This option i 306 Default is true. This option is only applied to TUI.
305 307
306 368:│ mov 0 308 368:│ mov 0x8(%r14),%rdi
307 309
308 This option works with tui, st 310 This option works with tui, stdio2 browsers.
309 311
310 annotate.jump_arrows:: 312 annotate.jump_arrows::
311 There can be jump instruction 313 There can be jump instruction among assembly code.
312 Depending on a boolean value o 314 Depending on a boolean value of jump_arrows,
313 arrows can be printed or not w 315 arrows can be printed or not which represent
314 where do the instruction jump 316 where do the instruction jump into as below.
315 317
316 │ ┌──jmp 1333 318 │ ┌──jmp 1333
317 │ │ xchg %ax,%ax 319 │ │ xchg %ax,%ax
318 │1330:│ mov %r15,%r10 320 │1330:│ mov %r15,%r10
319 │1333:└─→cmp %r15,% 321 │1333:└─→cmp %r15,%r14
320 322
321 If jump_arrow is 'false', the 323 If jump_arrow is 'false', the arrows isn't printed as below.
322 Default is 'false'. 324 Default is 'false'.
323 325
324 │ ↓ jmp 1333 326 │ ↓ jmp 1333
325 │ xchg %ax,%ax 327 │ xchg %ax,%ax
326 │1330: mov %r15,%r10 328 │1330: mov %r15,%r10
327 │1333: cmp %r15,%r14 329 │1333: cmp %r15,%r14
328 330
329 This option works with tui bro 331 This option works with tui browser.
330 332
331 annotate.show_linenr:: 333 annotate.show_linenr::
332 When showing source code if th 334 When showing source code if this option is 'true',
333 line numbers are printed as be 335 line numbers are printed as below.
334 336
335 │1628 if (type & PER 337 │1628 if (type & PERF_SAMPLE_IDENTIFIER) {
336 │ ↓ jne 508 338 │ ↓ jne 508
337 │1628 data-> 339 │1628 data->id = *array;
338 │1629 array+ 340 │1629 array++;
339 │1630 } 341 │1630 }
340 342
341 However if this option is 'fal 343 However if this option is 'false', they aren't printed as below.
342 Default is 'false'. 344 Default is 'false'.
343 345
344 │ if (type & PER 346 │ if (type & PERF_SAMPLE_IDENTIFIER) {
345 │ ↓ jne 508 347 │ ↓ jne 508
346 │ data-> 348 │ data->id = *array;
347 │ array+ 349 │ array++;
348 │ } 350 │ }
349 351
350 This option works with tui, st 352 This option works with tui, stdio2 browsers.
351 353
352 annotate.show_nr_jumps:: 354 annotate.show_nr_jumps::
353 Let's see a part of assembly c 355 Let's see a part of assembly code.
354 356
355 │1382: movb $0x1,-0x270( 357 │1382: movb $0x1,-0x270(%rbp)
356 358
357 If use this, the number of bra 359 If use this, the number of branches jumping to that address can be printed as below.
358 Default is 'false'. 360 Default is 'false'.
359 361
360 │1 1382: movb $0x1,-0x27 362 │1 1382: movb $0x1,-0x270(%rbp)
361 363
362 This option works with tui, st 364 This option works with tui, stdio2 browsers.
363 365
364 annotate.show_total_period:: 366 annotate.show_total_period::
365 To compare two records on an i 367 To compare two records on an instruction base, with this option
366 provided, display total number 368 provided, display total number of samples that belong to a line
367 in assembly code. If this opti 369 in assembly code. If this option is 'true', total periods are printed
368 instead of percent values as b 370 instead of percent values as below.
369 371
370 302 │ mov %eax,%ea 372 302 │ mov %eax,%eax
371 373
372 But if this option is 'false', 374 But if this option is 'false', percent values for overhead are printed i.e.
373 Default is 'false'. 375 Default is 'false'.
374 376
375 99.93 │ mov %eax,%ea 377 99.93 │ mov %eax,%eax
376 378
377 This option works with tui, st 379 This option works with tui, stdio2, stdio browsers.
378 380
379 annotate.show_nr_samples:: 381 annotate.show_nr_samples::
380 By default perf annotate shows 382 By default perf annotate shows percentage of samples. This option
381 can be used to print absolute 383 can be used to print absolute number of samples. Ex, when set as
382 false: 384 false:
383 385
384 Percent│ 386 Percent│
385 74.03 │ mov %fs:0x2 387 74.03 │ mov %fs:0x28,%rax
386 388
387 When set as true: 389 When set as true:
388 390
389 Samples│ 391 Samples│
390 6 │ mov %fs:0x2 392 6 │ mov %fs:0x28,%rax
391 393
392 This option works with tui, st 394 This option works with tui, stdio2, stdio browsers.
393 395
394 annotate.offset_level:: 396 annotate.offset_level::
395 Default is '1', meaning just j 397 Default is '1', meaning just jump targets will have offsets show right beside
396 the instruction. When set to ' 398 the instruction. When set to '2' 'call' instructions will also have its offsets
397 shown, 3 or higher will show o 399 shown, 3 or higher will show offsets for all instructions.
398 400
399 This option works with tui, st 401 This option works with tui, stdio2 browsers.
400 402
401 annotate.demangle:: 403 annotate.demangle::
402 Demangle symbol names to human 404 Demangle symbol names to human readable form. Default is 'true'.
403 405
404 annotate.demangle_kernel:: 406 annotate.demangle_kernel::
405 Demangle kernel symbol names t 407 Demangle kernel symbol names to human readable form. Default is 'true'.
406 408
407 hist.*:: 409 hist.*::
408 hist.percentage:: 410 hist.percentage::
409 This option control the way to 411 This option control the way to calculate overhead of filtered entries -
410 that means the value of this o 412 that means the value of this option is effective only if there's a
411 filter (by comm, dso or symbol 413 filter (by comm, dso or symbol name). Suppose a following example:
412 414
413 Overhead Symbols 415 Overhead Symbols
414 ........ ....... 416 ........ .......
415 33.33% foo 417 33.33% foo
416 33.33% bar 418 33.33% bar
417 33.33% baz 419 33.33% baz
418 420
419 This is an original overhead an 421 This is an original overhead and we'll filter out the first 'foo'
420 entry. The value of 'relative' 422 entry. The value of 'relative' would increase the overhead of 'bar'
421 and 'baz' to 50.00% for each, w 423 and 'baz' to 50.00% for each, while 'absolute' would show their
422 current overhead (33.33%). 424 current overhead (33.33%).
423 425
424 ui.*:: 426 ui.*::
425 ui.show-headers:: 427 ui.show-headers::
426 This option controls display o 428 This option controls display of column headers (like 'Overhead' and 'Symbol')
427 in 'report' and 'top'. If this 429 in 'report' and 'top'. If this option is false, they are hidden.
428 This option is only applied to 430 This option is only applied to TUI.
429 431
430 call-graph.*:: 432 call-graph.*::
431 The following controls the handling of 433 The following controls the handling of call-graphs (obtained via the
432 -g/--call-graph options). 434 -g/--call-graph options).
433 435
434 call-graph.record-mode:: 436 call-graph.record-mode::
435 The mode for user space can be 437 The mode for user space can be 'fp' (frame pointer), 'dwarf'
436 and 'lbr'. The value 'dwarf' 438 and 'lbr'. The value 'dwarf' is effective only if libunwind
437 (or a recent version of libdw) 439 (or a recent version of libdw) is present on the system;
438 the value 'lbr' only works for 440 the value 'lbr' only works for certain cpus. The method for
439 kernel space is controlled not 441 kernel space is controlled not by this option but by the
440 kernel config (CONFIG_UNWINDER 442 kernel config (CONFIG_UNWINDER_*).
441 443
442 call-graph.dump-size:: 444 call-graph.dump-size::
443 The size of stack to dump in o 445 The size of stack to dump in order to do post-unwinding. Default is 8192 (byte).
444 When using dwarf into record-m 446 When using dwarf into record-mode, the default size will be used if omitted.
445 447
446 call-graph.print-type:: 448 call-graph.print-type::
447 The print-types can be graph ( 449 The print-types can be graph (graph absolute), fractal (graph relative),
448 flat and folded. This option c 450 flat and folded. This option controls a way to show overhead for each callchain
449 entry. Suppose a following exa 451 entry. Suppose a following example.
450 452
451 Overhead Symbols 453 Overhead Symbols
452 ........ ....... 454 ........ .......
453 40.00% foo 455 40.00% foo
454 | 456 |
455 ---foo 457 ---foo
456 | 458 |
457 |--50.00%--bar 459 |--50.00%--bar
458 | main 460 | main
459 | 461 |
460 --50.00%--baz 462 --50.00%--baz
461 main 463 main
462 464
463 This output is a 'fractal' for 465 This output is a 'fractal' format. The 'foo' came from 'bar' and 'baz' exactly
464 half and half so 'fractal' sho 466 half and half so 'fractal' shows 50.00% for each
465 (meaning that it assumes 100% 467 (meaning that it assumes 100% total overhead of 'foo').
466 468
467 The 'graph' uses absolute over 469 The 'graph' uses absolute overhead value of 'foo' as total so each of
468 'bar' and 'baz' callchain will 470 'bar' and 'baz' callchain will have 20.00% of overhead.
469 If 'flat' is used, single colu 471 If 'flat' is used, single column and linear exposure of call chains.
470 'folded' mean call chains are 472 'folded' mean call chains are displayed in a line, separated by semicolons.
471 473
472 call-graph.order:: 474 call-graph.order::
473 This option controls print ord 475 This option controls print order of callchains. The default is
474 'callee' which means callee is 476 'callee' which means callee is printed at top and then followed by its
475 caller and so on. The 'caller' 477 caller and so on. The 'caller' prints it in reverse order.
476 478
477 If this option is not set and 479 If this option is not set and report.children or top.children is
478 set to true (or the equivalent 480 set to true (or the equivalent command line option is given),
479 the default value of this opti 481 the default value of this option is changed to 'caller' for the
480 execution of 'perf report' or 482 execution of 'perf report' or 'perf top'. Other commands will
481 still default to 'callee'. 483 still default to 'callee'.
482 484
483 call-graph.sort-key:: 485 call-graph.sort-key::
484 The callchains are merged if t 486 The callchains are merged if they contain same information.
485 The sort-key option determines 487 The sort-key option determines a way to compare the callchains.
486 A value of 'sort-key' can be ' 488 A value of 'sort-key' can be 'function' or 'address'.
487 The default is 'function'. 489 The default is 'function'.
488 490
489 call-graph.threshold:: 491 call-graph.threshold::
490 When there're many callchains 492 When there're many callchains it'd print tons of lines. So perf omits
491 small callchains under a certa 493 small callchains under a certain overhead (threshold) and this option
492 control the threshold. Default 494 control the threshold. Default is 0.5 (%). The overhead is calculated
493 by value depends on call-graph 495 by value depends on call-graph.print-type.
494 496
495 call-graph.print-limit:: 497 call-graph.print-limit::
496 This is a maximum number of li 498 This is a maximum number of lines of callchain printed for a single
497 histogram entry. Default is 0 499 histogram entry. Default is 0 which means no limitation.
498 500
499 report.*:: 501 report.*::
500 report.sort_order:: 502 report.sort_order::
501 Allows changing the default so 503 Allows changing the default sort order from "comm,dso,symbol" to
502 some other default, for instan 504 some other default, for instance "sym,dso" may be more fitting for
503 kernel developers. 505 kernel developers.
504 report.percent-limit:: 506 report.percent-limit::
505 This one is mostly the same as 507 This one is mostly the same as call-graph.threshold but works for
506 histogram entries. Entries hav 508 histogram entries. Entries having an overhead lower than this
507 percentage will not be printed 509 percentage will not be printed. Default is '0'. If percent-limit
508 is '10', only entries which ha 510 is '10', only entries which have more than 10% of overhead will be
509 printed. 511 printed.
510 512
511 report.queue-size:: 513 report.queue-size::
512 This option sets up the maximu 514 This option sets up the maximum allocation size of the internal
513 event queue for ordering event 515 event queue for ordering events. Default is 0, meaning no limit.
514 516
515 report.children:: 517 report.children::
516 'Children' means functions cal 518 'Children' means functions called from another function.
517 If this option is true, 'perf 519 If this option is true, 'perf report' cumulates callchains of children
518 and show (accumulated) total o 520 and show (accumulated) total overhead as well as 'Self' overhead.
519 Please refer to the 'perf repo 521 Please refer to the 'perf report' manual. The default is 'true'.
520 522
521 report.group:: 523 report.group::
522 This option is to show event g 524 This option is to show event group information together.
523 Example output with this turne 525 Example output with this turned on, notice that there is one column
524 per event in the group, ref-cy 526 per event in the group, ref-cycles and cycles:
525 527
526 # group: {ref-cycles,cycles} 528 # group: {ref-cycles,cycles}
527 # ======== 529 # ========
528 # 530 #
529 # Samples: 7K of event 'anon g 531 # Samples: 7K of event 'anon group { ref-cycles, cycles }'
530 # Event count (approx.): 68761 532 # Event count (approx.): 6876107743
531 # 533 #
532 # Overhead Command 534 # Overhead Command Shared Object Symbol
533 # ................ ....... . 535 # ................ ....... ................. ...................
534 # 536 #
535 99.84% 99.76% noploop n 537 99.84% 99.76% noploop noploop [.] main
536 0.07% 0.00% noploop l 538 0.07% 0.00% noploop ld-2.15.so [.] strcmp
537 0.03% 0.00% noploop [ 539 0.03% 0.00% noploop [kernel.kallsyms] [k] timerqueue_del
538 540
539 report.skip-empty:: 541 report.skip-empty::
540 This option can change default 542 This option can change default stat behavior with empty results.
541 If it's set true, 'perf report 543 If it's set true, 'perf report --stat' will not show 0 stats.
542 544
543 top.*:: 545 top.*::
544 top.children:: 546 top.children::
545 Same as 'report.children'. So 547 Same as 'report.children'. So if it is enabled, the output of 'top'
546 command will have 'Children' o 548 command will have 'Children' overhead column as well as 'Self' overhead
547 column by default. 549 column by default.
548 The default is 'true'. 550 The default is 'true'.
549 551
550 top.call-graph:: 552 top.call-graph::
551 This is identical to 'call-gra 553 This is identical to 'call-graph.record-mode', except it is
552 applicable only for 'top' subc 554 applicable only for 'top' subcommand. This option ONLY setup
553 the unwind method. To enable ' 555 the unwind method. To enable 'perf top' to actually use it,
554 the command line option -g mus 556 the command line option -g must be specified.
555 557
556 man.*:: 558 man.*::
557 man.viewer:: 559 man.viewer::
558 This option can assign a tool 560 This option can assign a tool to view manual pages when 'help'
559 subcommand was invoked. Suppor 561 subcommand was invoked. Supported tools are 'man', 'woman'
560 (with emacs client) and 'konqu 562 (with emacs client) and 'konqueror'. Default is 'man'.
561 563
562 New man viewer tool can be als 564 New man viewer tool can be also added using 'man.<tool>.cmd'
563 or use different path using 'm 565 or use different path using 'man.<tool>.path' config option.
564 566
565 pager.*:: 567 pager.*::
566 pager.<subcommand>:: 568 pager.<subcommand>::
567 When the subcommand is run on 569 When the subcommand is run on stdio, determine whether it uses
568 pager or not based on this val 570 pager or not based on this value. Default is 'unspecified'.
569 571
570 kmem.*:: 572 kmem.*::
571 kmem.default:: 573 kmem.default::
572 This option decides which allo 574 This option decides which allocator is to be analyzed if neither
573 '--slab' nor '--page' option i 575 '--slab' nor '--page' option is used. Default is 'slab'.
574 576
575 record.*:: 577 record.*::
576 record.build-id:: 578 record.build-id::
577 This option can be 'cache', 'n 579 This option can be 'cache', 'no-cache', 'skip' or 'mmap'.
578 'cache' is to post-process dat 580 'cache' is to post-process data and save/update the binaries into
579 the build-id cache (in ~/.debu 581 the build-id cache (in ~/.debug). This is the default.
580 But if this option is 'no-cach 582 But if this option is 'no-cache', it will not update the build-id cache.
581 'skip' skips post-processing a 583 'skip' skips post-processing and does not update the cache.
582 'mmap' skips post-processing a 584 'mmap' skips post-processing and reads build-ids from MMAP events.
583 585
584 record.call-graph:: 586 record.call-graph::
585 This is identical to 'call-gra 587 This is identical to 'call-graph.record-mode', except it is
586 applicable only for 'record' s 588 applicable only for 'record' subcommand. This option ONLY setup
587 the unwind method. To enable ' 589 the unwind method. To enable 'perf record' to actually use it,
588 the command line option -g mus 590 the command line option -g must be specified.
589 591
590 record.aio:: 592 record.aio::
591 Use 'n' control blocks in asyn 593 Use 'n' control blocks in asynchronous (Posix AIO) trace writing
592 mode ('n' default: 1, max: 4). 594 mode ('n' default: 1, max: 4).
593 595
594 record.debuginfod:: 596 record.debuginfod::
595 Specify debuginfod URL to be u 597 Specify debuginfod URL to be used when cacheing perf.data binaries,
596 it follows the same syntax as 598 it follows the same syntax as the DEBUGINFOD_URLS variable, like:
597 599
598 http://192.168.122.174:8002 600 http://192.168.122.174:8002
599 601
600 If the URLs is 'system', the v 602 If the URLs is 'system', the value of DEBUGINFOD_URLS system environment
601 variable is used. 603 variable is used.
602 604
603 diff.*:: 605 diff.*::
604 diff.order:: 606 diff.order::
605 This option sets the number of 607 This option sets the number of columns to sort the result.
606 The default is 0, which means 608 The default is 0, which means sorting by baseline.
607 Setting it to 1 will sort the 609 Setting it to 1 will sort the result by delta (or other
608 compute method selected). 610 compute method selected).
609 611
610 diff.compute:: 612 diff.compute::
611 This options sets the method f 613 This options sets the method for computing the diff result.
612 Possible values are 'delta', ' 614 Possible values are 'delta', 'delta-abs', 'ratio' and
613 'wdiff'. Default is 'delta'. 615 'wdiff'. Default is 'delta'.
614 616
615 trace.*:: 617 trace.*::
616 trace.add_events:: 618 trace.add_events::
617 Allows adding a set of events 619 Allows adding a set of events to add to the ones specified
618 by the user, or use as a defau 620 by the user, or use as a default one if none was specified.
619 The initial use case is to add 621 The initial use case is to add augmented_raw_syscalls.o to
620 activate the 'perf trace' logi 622 activate the 'perf trace' logic that looks for syscall
621 pointer contents after the nor 623 pointer contents after the normal tracepoint payload.
622 624
623 trace.args_alignment:: 625 trace.args_alignment::
624 Number of columns to align the 626 Number of columns to align the argument list, default is 70,
625 use 40 for the strace default, 627 use 40 for the strace default, zero to no alignment.
626 628
627 trace.no_inherit:: 629 trace.no_inherit::
628 Do not follow children threads 630 Do not follow children threads.
629 631
630 trace.show_arg_names:: 632 trace.show_arg_names::
631 Should syscall argument names 633 Should syscall argument names be printed? If not then trace.show_zeros
632 will be set. 634 will be set.
633 635
634 trace.show_duration:: 636 trace.show_duration::
635 Show syscall duration. 637 Show syscall duration.
636 638
637 trace.show_prefix:: 639 trace.show_prefix::
638 If set to 'yes' will show comm 640 If set to 'yes' will show common string prefixes in tables. The default
639 is to remove the common prefix 641 is to remove the common prefix in things like "MAP_SHARED", showing just "SHARED".
640 642
641 trace.show_timestamp:: 643 trace.show_timestamp::
642 Show syscall start timestamp. 644 Show syscall start timestamp.
643 645
644 trace.show_zeros:: 646 trace.show_zeros::
645 Do not suppress syscall argume 647 Do not suppress syscall arguments that are equal to zero.
646 648
647 trace.tracepoint_beautifiers:: 649 trace.tracepoint_beautifiers::
648 Use "libtraceevent" to use tha 650 Use "libtraceevent" to use that library to augment the tracepoint arguments,
649 "libbeauty", the default, to u 651 "libbeauty", the default, to use the same argument beautifiers used in the
650 strace-like sys_enter+sys_exit 652 strace-like sys_enter+sys_exit lines.
651 653
652 ftrace.*:: 654 ftrace.*::
653 ftrace.tracer:: 655 ftrace.tracer::
654 Can be used to select the defa 656 Can be used to select the default tracer when neither -G nor
655 -F option is not specified. Po 657 -F option is not specified. Possible values are 'function' and
656 'function_graph'. 658 'function_graph'.
657 659
>> 660 llvm.*::
>> 661 llvm.clang-path::
>> 662 Path to clang. If omit, search it from $PATH.
>> 663
>> 664 llvm.clang-bpf-cmd-template::
>> 665 Cmdline template. Below lines show its default value. Environment
>> 666 variable is used to pass options.
>> 667 "$CLANG_EXEC -D__KERNEL__ -D__NR_CPUS__=$NR_CPUS "\
>> 668 "-DLINUX_VERSION_CODE=$LINUX_VERSION_CODE " \
>> 669 "$CLANG_OPTIONS $PERF_BPF_INC_OPTIONS $KERNEL_INC_OPTIONS " \
>> 670 "-Wno-unused-value -Wno-pointer-sign " \
>> 671 "-working-directory $WORKING_DIR " \
>> 672 "-c \"$CLANG_SOURCE\" --target=bpf $CLANG_EMIT_LLVM -O2 -o - $LLVM_OPTIONS_PIPE"
>> 673
>> 674 llvm.clang-opt::
>> 675 Options passed to clang.
>> 676
>> 677 llvm.kbuild-dir::
>> 678 kbuild directory. If not set, use /lib/modules/`uname -r`/build.
>> 679 If set to "" deliberately, skip kernel header auto-detector.
>> 680
>> 681 llvm.kbuild-opts::
>> 682 Options passed to 'make' when detecting kernel header options.
>> 683
>> 684 llvm.dump-obj::
>> 685 Enable perf dump BPF object files compiled by LLVM.
>> 686
>> 687 llvm.opts::
>> 688 Options passed to llc.
>> 689
658 samples.*:: 690 samples.*::
659 691
660 samples.context:: 692 samples.context::
661 Define how many ns worth of ti 693 Define how many ns worth of time to show
662 around samples in perf report 694 around samples in perf report sample context browser.
663 695
664 scripts.*:: 696 scripts.*::
665 697
666 Any option defines a script that is ad 698 Any option defines a script that is added to the scripts menu
667 in the interactive perf browser and wh 699 in the interactive perf browser and whose output is displayed.
668 The name of the option is the name, th 700 The name of the option is the name, the value is a script command line.
669 The script gets the same options passe 701 The script gets the same options passed as a full perf script,
670 in particular -i perfdata file, --cpu, 702 in particular -i perfdata file, --cpu, --tid
671 703
672 convert.*:: 704 convert.*::
673 705
674 convert.queue-size:: 706 convert.queue-size::
675 Limit the size of ordered_even 707 Limit the size of ordered_events queue, so we could control
676 allocation size of perf data f 708 allocation size of perf data files without proper finished
677 round events. 709 round events.
678 stat.*:: 710 stat.*::
679 711
680 stat.big-num:: 712 stat.big-num::
681 (boolean) Change the default f 713 (boolean) Change the default for "--big-num". To make
682 "--no-big-num" the default, se 714 "--no-big-num" the default, set "stat.big-num=false".
683 715
684 intel-pt.*:: 716 intel-pt.*::
685 717
686 intel-pt.cache-divisor:: 718 intel-pt.cache-divisor::
687 719
688 intel-pt.mispred-all:: 720 intel-pt.mispred-all::
689 If set, Intel PT decoder will 721 If set, Intel PT decoder will set the mispred flag on all
690 branches. 722 branches.
691 723
692 intel-pt.max-loops:: 724 intel-pt.max-loops::
693 If set and non-zero, the maxim 725 If set and non-zero, the maximum number of unconditional
694 branches decoded without consu 726 branches decoded without consuming any trace packets. If
695 the maximum is exceeded there 727 the maximum is exceeded there will be a "Never-ending loop"
696 error. The default is 100000. 728 error. The default is 100000.
697 729
698 auxtrace.*:: 730 auxtrace.*::
699 731
700 auxtrace.dumpdir:: 732 auxtrace.dumpdir::
701 s390 only. The directory to sa 733 s390 only. The directory to save the auxiliary trace buffer
702 can be changed using this opti 734 can be changed using this option. Ex, auxtrace.dumpdir=/tmp.
703 If the directory does not exis 735 If the directory does not exist or has the wrong file type,
704 the current directory is used. 736 the current directory is used.
705 737
706 itrace.*:: 738 itrace.*::
707 739
708 debug-log-buffer-size:: 740 debug-log-buffer-size::
709 Log size in bytes to output wh 741 Log size in bytes to output when using the option --itrace=d+e
710 Refer 'itrace' option of linkp 742 Refer 'itrace' option of linkperf:perf-script[1] or
711 linkperf:perf-report[1]. The d 743 linkperf:perf-report[1]. The default is 16384.
712 744
713 daemon.*:: 745 daemon.*::
714 746
715 daemon.base:: 747 daemon.base::
716 Base path for daemon data. All 748 Base path for daemon data. All sessions data are stored under
717 this path. 749 this path.
718 750
719 session-<NAME>.*:: 751 session-<NAME>.*::
720 752
721 session-<NAME>.run:: 753 session-<NAME>.run::
722 754
723 Defines new record session for 755 Defines new record session for daemon. The value is record's
724 command line without the 'reco 756 command line without the 'record' keyword.
>> 757
725 758
726 SEE ALSO 759 SEE ALSO
727 -------- 760 --------
728 linkperf:perf[1] 761 linkperf:perf[1]
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.