1perf-top(1) 2=========== 3 4NAME 5---- 6perf-top - System profiling tool. 7 8SYNOPSIS 9-------- 10[verse] 11'perf top' [-e <EVENT> | --event=EVENT] [<options>] 12 13DESCRIPTION 14----------- 15This command generates and displays a performance counter profile in real time. 16 17 18OPTIONS 19------- 20-a:: 21--all-cpus:: 22 System-wide collection. (default) 23 24-c <count>:: 25--count=<count>:: 26 Event period to sample. 27 28-C <cpu-list>:: 29--cpu=<cpu>:: 30Monitor only on the list of CPUs provided. Multiple CPUs can be provided as a 31comma-separated list with no space: 0,1. Ranges of CPUs are specified with -: 0-2. 32Default is to monitor all CPUS. 33 34-d <seconds>:: 35--delay=<seconds>:: 36 Number of seconds to delay between refreshes. 37 38-e <event>:: 39--event=<event>:: 40 Select the PMU event. Selection can be a symbolic event name 41 (use 'perf list' to list all events) or a raw PMU 42 event (eventsel+umask) in the form of rNNN where NNN is a 43 hexadecimal event descriptor. 44 45-E <entries>:: 46--entries=<entries>:: 47 Display this many functions. 48 49-f <count>:: 50--count-filter=<count>:: 51 Only display functions with more events than this. 52 53--group:: 54 Put the counters into a counter group. 55 56-F <freq>:: 57--freq=<freq>:: 58 Profile at this frequency. Use 'max' to use the currently maximum 59 allowed frequency, i.e. the value in the kernel.perf_event_max_sample_rate 60 sysctl. 61 62-i:: 63--inherit:: 64 Child tasks do not inherit counters. 65 66-k <path>:: 67--vmlinux=<path>:: 68 Path to vmlinux. Required for annotation functionality. 69 70--ignore-vmlinux:: 71 Ignore vmlinux files. 72 73--kallsyms=<file>:: 74 kallsyms pathname 75 76-m <pages>:: 77--mmap-pages=<pages>:: 78 Number of mmap data pages (must be a power of two) or size 79 specification with appended unit character - B/K/M/G. The 80 size is rounded up to have nearest pages power of two value. 81 82-p <pid>:: 83--pid=<pid>:: 84 Profile events on existing Process ID (comma separated list). 85 86-t <tid>:: 87--tid=<tid>:: 88 Profile events on existing thread ID (comma separated list). 89 90-u:: 91--uid=:: 92 Record events in threads owned by uid. Name or number. 93 94-r <priority>:: 95--realtime=<priority>:: 96 Collect data with this RT SCHED_FIFO priority. 97 98--sym-annotate=<symbol>:: 99 Annotate this symbol. 100 101-K:: 102--hide_kernel_symbols:: 103 Hide kernel symbols. 104 105-U:: 106--hide_user_symbols:: 107 Hide user symbols. 108 109--demangle-kernel:: 110 Demangle kernel symbols. 111 112-D:: 113--dump-symtab:: 114 Dump the symbol table used for profiling. 115 116-v:: 117--verbose:: 118 Be more verbose (show counter open errors, etc). 119 120-z:: 121--zero:: 122 Zero history across display updates. 123 124-s:: 125--sort:: 126 Sort by key(s): pid, comm, dso, symbol, parent, srcline, weight, 127 local_weight, abort, in_tx, transaction, overhead, sample, period. 128 Please see description of --sort in the perf-report man page. 129 130--fields=:: 131 Specify output field - multiple keys can be specified in CSV format. 132 Following fields are available: 133 overhead, overhead_sys, overhead_us, overhead_children, sample and period. 134 Also it can contain any sort key(s). 135 136 By default, every sort keys not specified in --field will be appended 137 automatically. 138 139-n:: 140--show-nr-samples:: 141 Show a column with the number of samples. 142 143--show-total-period:: 144 Show a column with the sum of periods. 145 146--dsos:: 147 Only consider symbols in these dsos. This option will affect the 148 percentage of the overhead column. See --percentage for more info. 149 150--comms:: 151 Only consider symbols in these comms. This option will affect the 152 percentage of the overhead column. See --percentage for more info. 153 154--symbols:: 155 Only consider these symbols. This option will affect the 156 percentage of the overhead column. See --percentage for more info. 157 158-M:: 159--disassembler-style=:: Set disassembler style for objdump. 160 161--source:: 162 Interleave source code with assembly code. Enabled by default, 163 disable with --no-source. 164 165--asm-raw:: 166 Show raw instruction encoding of assembly instructions. 167 168-g:: 169 Enables call-graph (stack chain/backtrace) recording. 170 171--call-graph [mode,type,min[,limit],order[,key][,branch]]:: 172 Setup and enable call-graph (stack chain/backtrace) recording, 173 implies -g. See `--call-graph` section in perf-record and 174 perf-report man pages for details. 175 176--children:: 177 Accumulate callchain of children to parent entry so that then can 178 show up in the output. The output will have a new "Children" column 179 and will be sorted on the data. It requires -g/--call-graph option 180 enabled. See the `overhead calculation' section for more details. 181 Enabled by default, disable with --no-children. 182 183--max-stack:: 184 Set the stack depth limit when parsing the callchain, anything 185 beyond the specified depth will be ignored. This is a trade-off 186 between information loss and faster processing especially for 187 workloads that can have a very long callchain stack. 188 189 Default: /proc/sys/kernel/perf_event_max_stack when present, 127 otherwise. 190 191--ignore-callees=<regex>:: 192 Ignore callees of the function(s) matching the given regex. 193 This has the effect of collecting the callers of each such 194 function into one place in the call-graph tree. 195 196--percent-limit:: 197 Do not show entries which have an overhead under that percent. 198 (Default: 0). 199 200--percentage:: 201 Determine how to display the overhead percentage of filtered entries. 202 Filters can be applied by --comms, --dsos and/or --symbols options and 203 Zoom operations on the TUI (thread, dso, etc). 204 205 "relative" means it's relative to filtered entries only so that the 206 sum of shown entries will be always 100%. "absolute" means it retains 207 the original value before and after the filter is applied. 208 209-w:: 210--column-widths=<width[,width...]>:: 211 Force each column width to the provided list, for large terminal 212 readability. 0 means no limit (default behavior). 213 214--proc-map-timeout:: 215 When processing pre-existing threads /proc/XXX/mmap, it may take 216 a long time, because the file may be huge. A time out is needed 217 in such cases. 218 This option sets the time out limit. The default value is 500 ms. 219 220 221-b:: 222--branch-any:: 223 Enable taken branch stack sampling. Any type of taken branch may be sampled. 224 This is a shortcut for --branch-filter any. See --branch-filter for more infos. 225 226-j:: 227--branch-filter:: 228 Enable taken branch stack sampling. Each sample captures a series of consecutive 229 taken branches. The number of branches captured with each sample depends on the 230 underlying hardware, the type of branches of interest, and the executed code. 231 It is possible to select the types of branches captured by enabling filters. 232 For a full list of modifiers please see the perf record manpage. 233 234 The option requires at least one branch type among any, any_call, any_ret, ind_call, cond. 235 The privilege levels may be omitted, in which case, the privilege levels of the associated 236 event are applied to the branch filter. Both kernel (k) and hypervisor (hv) privilege 237 levels are subject to permissions. When sampling on multiple events, branch stack sampling 238 is enabled for all the sampling events. The sampled branch type is the same for all events. 239 The various filters must be specified as a comma separated list: --branch-filter any_ret,u,k 240 Note that this feature may not be available on all processors. 241 242--raw-trace:: 243 When displaying traceevent output, do not use print fmt or plugins. 244 245--hierarchy:: 246 Enable hierarchy output. 247 248--overwrite:: 249 Enable this to use just the most recent records, which helps in high core count 250 machines such as Knights Landing/Mill, but right now is disabled by default as 251 the pausing used in this technique is leading to loss of metadata events such 252 as PERF_RECORD_MMAP which makes 'perf top' unable to resolve samples, leading 253 to lots of unknown samples appearing on the UI. Enable this if you are in such 254 machines and profiling a workload that doesn't creates short lived threads and/or 255 doesn't uses many executable mmap operations. Work is being planed to solve 256 this situation, till then, this will remain disabled by default. 257 258--force:: 259 Don't do ownership validation. 260 261--num-thread-synthesize:: 262 The number of threads to run when synthesizing events for existing processes. 263 By default, the number of threads equals to the number of online CPUs. 264 265--namespaces:: 266 Record events of type PERF_RECORD_NAMESPACES and display it with the 267 'cgroup_id' sort key. 268 269--switch-on EVENT_NAME:: 270 Only consider events after this event is found. 271 272 E.g.: 273 274 Find out where broadcast packets are handled 275 276 perf probe -L icmp_rcv 277 278 Insert a probe there: 279 280 perf probe icmp_rcv:59 281 282 Start perf top and ask it to only consider the cycles events when a 283 broadcast packet arrives This will show a menu with two entries and 284 will start counting when a broadcast packet arrives: 285 286 perf top -e cycles,probe:icmp_rcv --switch-on=probe:icmp_rcv 287 288 Alternatively one can ask for --group and then two overhead columns 289 will appear, the first for cycles and the second for the switch-on event. 290 291 perf top --group -e cycles,probe:icmp_rcv --switch-on=probe:icmp_rcv 292 293 This may be interesting to measure a workload only after some initialization 294 phase is over, i.e. insert a perf probe at that point and use the above 295 examples replacing probe:icmp_rcv with the just-after-init probe. 296 297--switch-off EVENT_NAME:: 298 Stop considering events after this event is found. 299 300--show-on-off-events:: 301 Show the --switch-on/off events too. This has no effect in 'perf top' now 302 but probably we'll make the default not to show the switch-on/off events 303 on the --group mode and if there is only one event besides the off/on ones, 304 go straight to the histogram browser, just like 'perf top' with no events 305 explicitely specified does. 306 307 308INTERACTIVE PROMPTING KEYS 309-------------------------- 310 311[d]:: 312 Display refresh delay. 313 314[e]:: 315 Number of entries to display. 316 317[E]:: 318 Event to display when multiple counters are active. 319 320[f]:: 321 Profile display filter (>= hit count). 322 323[F]:: 324 Annotation display filter (>= % of total). 325 326[s]:: 327 Annotate symbol. 328 329[S]:: 330 Stop annotation, return to full profile display. 331 332[K]:: 333 Hide kernel symbols. 334 335[U]:: 336 Hide user symbols. 337 338[z]:: 339 Toggle event count zeroing across display updates. 340 341[qQ]:: 342 Quit. 343 344Pressing any unmapped key displays a menu, and prompts for input. 345 346include::callchain-overhead-calculation.txt[] 347 348SEE ALSO 349-------- 350linkperf:perf-stat[1], linkperf:perf-list[1], linkperf:perf-report[1] 351