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