tools/perf/Documentation/perf-top.txt

   1 perf-top(1)
   2 ===========
   3
   4 NAME
   5 ----
   6 perf-top - System profiling tool.
   7
   8 SYNOPSIS
   9 --------
  10 [verse]
  11 'perf top' [-e <EVENT> | --event=EVENT] [<options>]
  12
  13 DESCRIPTION
  14 -----------
  15 This command generates and displays a performance counter profile in real time.
  16
  17
  18 OPTIONS
  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>::
  30 Monitor only on the list of CPUs provided. Multiple CPUs can be provided as a
  31 comma-separated list with no space: 0,1. Ranges of CPUs are specified with -: 0-2.
  32 Default 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.
  59
  60 -i::
  61 --inherit::
  62         Child tasks do not inherit counters.
  63
  64 -k <path>::
  65 --vmlinux=<path>::
  66         Path to vmlinux.  Required for annotation functionality.
  67
  68 -m <pages>::
  69 --mmap-pages=<pages>::
  70         Number of mmap data pages (must be a power of two) or size
  71         specification with appended unit character - B/K/M/G. The
  72         size is rounded up to have nearest pages power of two value.
  73
  74 -p <pid>::
  75 --pid=<pid>::
  76         Profile events on existing Process ID (comma separated list).
  77
  78 -t <tid>::
  79 --tid=<tid>::
  80         Profile events on existing thread ID (comma separated list).
  81
  82 -u::
  83 --uid=::
  84         Record events in threads owned by uid. Name or number.
  85
  86 -r <priority>::
  87 --realtime=<priority>::
  88         Collect data with this RT SCHED_FIFO priority.
  89
  90 --sym-annotate=<symbol>::
  91         Annotate this symbol.
  92
  93 -K::
  94 --hide_kernel_symbols::
  95         Hide kernel symbols.
  96
  97 -U::
  98 --hide_user_symbols::
  99         Hide user symbols.
 100
 101 --demangle-kernel::
 102         Demangle kernel symbols.
 103
 104 -D::
 105 --dump-symtab::
 106         Dump the symbol table used for profiling.
 107
 108 -v::
 109 --verbose::
 110         Be more verbose (show counter open errors, etc).
 111
 112 -z::
 113 --zero::
 114         Zero history across display updates.
 115
 116 -s::
 117 --sort::
 118         Sort by key(s): pid, comm, dso, symbol, parent, srcline, weight,
 119         local_weight, abort, in_tx, transaction, overhead, sample, period.
 120         Please see description of --sort in the perf-report man page.
 121
 122 --fields=::
 123         Specify output field - multiple keys can be specified in CSV format.
 124         Following fields are available:
 125         overhead, overhead_sys, overhead_us, overhead_children, sample and period.
 126         Also it can contain any sort key(s).
 127
 128         By default, every sort keys not specified in --field will be appended
 129         automatically.
 130
 131 -n::
 132 --show-nr-samples::
 133         Show a column with the number of samples.
 134
 135 --show-total-period::
 136         Show a column with the sum of periods.
 137
 138 --dsos::
 139         Only consider symbols in these dsos.  This option will affect the
 140         percentage of the overhead column.  See --percentage for more info.
 141
 142 --comms::
 143         Only consider symbols in these comms.  This option will affect the
 144         percentage of the overhead column.  See --percentage for more info.
 145
 146 --symbols::
 147         Only consider these symbols.  This option will affect the
 148         percentage of the overhead column.  See --percentage for more info.
 149
 150 -M::
 151 --disassembler-style=:: Set disassembler style for objdump.
 152
 153 --source::
 154         Interleave source code with assembly code. Enabled by default,
 155         disable with --no-source.
 156
 157 --asm-raw::
 158         Show raw instruction encoding of assembly instructions.
 159
 160 -g::
 161         Enables call-graph (stack chain/backtrace) recording.
 162
 163 --call-graph [mode,type,min[,limit],order[,key][,branch]]::
 164         Setup and enable call-graph (stack chain/backtrace) recording,
 165         implies -g.  See `--call-graph` section in perf-record and
 166         perf-report man pages for details.
 167
 168 --children::
 169         Accumulate callchain of children to parent entry so that then can
 170         show up in the output.  The output will have a new "Children" column
 171         and will be sorted on the data.  It requires -g/--call-graph option
 172         enabled.  See the `overhead calculation' section for more details.
 173         Enabled by default, disable with --no-children.
 174
 175 --max-stack::
 176         Set the stack depth limit when parsing the callchain, anything
 177         beyond the specified depth will be ignored. This is a trade-off
 178         between information loss and faster processing especially for
 179         workloads that can have a very long callchain stack.
 180
 181         Default: /proc/sys/kernel/perf_event_max_stack when present, 127 otherwise.
 182
 183 --ignore-callees=<regex>::
 184         Ignore callees of the function(s) matching the given regex.
 185         This has the effect of collecting the callers of each such
 186         function into one place in the call-graph tree.
 187
 188 --percent-limit::
 189         Do not show entries which have an overhead under that percent.
 190         (Default: 0).
 191
 192 --percentage::
 193         Determine how to display the overhead percentage of filtered entries.
 194         Filters can be applied by --comms, --dsos and/or --symbols options and
 195         Zoom operations on the TUI (thread, dso, etc).
 196
 197         "relative" means it's relative to filtered entries only so that the
 198         sum of shown entries will be always 100%. "absolute" means it retains
 199         the original value before and after the filter is applied.
 200
 201 -w::
 202 --column-widths=<width[,width...]>::
 203         Force each column width to the provided list, for large terminal
 204         readability.  0 means no limit (default behavior).
 205
 206 --proc-map-timeout::
 207         When processing pre-existing threads /proc/XXX/mmap, it may take
 208         a long time, because the file may be huge. A time out is needed
 209         in such cases.
 210         This option sets the time out limit. The default value is 500 ms.
 211
 212
 213 -b::
 214 --branch-any::
 215         Enable taken branch stack sampling. Any type of taken branch may be sampled.
 216         This is a shortcut for --branch-filter any. See --branch-filter for more infos.
 217
 218 -j::
 219 --branch-filter::
 220         Enable taken branch stack sampling. Each sample captures a series of consecutive
 221         taken branches. The number of branches captured with each sample depends on the
 222         underlying hardware, the type of branches of interest, and the executed code.
 223         It is possible to select the types of branches captured by enabling filters.
 224         For a full list of modifiers please see the perf record manpage.
 225
 226         The option requires at least one branch type among any, any_call, any_ret, ind_call, cond.
 227         The privilege levels may be omitted, in which case, the privilege levels of the associated
 228         event are applied to the branch filter. Both kernel (k) and hypervisor (hv) privilege
 229         levels are subject to permissions.  When sampling on multiple events, branch stack sampling
 230         is enabled for all the sampling events. The sampled branch type is the same for all events.
 231         The various filters must be specified as a comma separated list: --branch-filter any_ret,u,k
 232         Note that this feature may not be available on all processors.
 233
 234 --raw-trace::
 235         When displaying traceevent output, do not use print fmt or plugins.
 236
 237 --hierarchy::
 238         Enable hierarchy output.
 239
 240 --force::
 241         Don't do ownership validation.
 242
 243
 244 INTERACTIVE PROMPTING KEYS
 245 --------------------------
 246
 247 [d]::
 248         Display refresh delay.
 249
 250 [e]::
 251         Number of entries to display.
 252
 253 [E]::
 254         Event to display when multiple counters are active.
 255
 256 [f]::
 257         Profile display filter (>= hit count).
 258
 259 [F]::
 260         Annotation display filter (>= % of total).
 261
 262 [s]::
 263         Annotate symbol.
 264
 265 [S]::
 266         Stop annotation, return to full profile display.
 267
 268 [z]::
 269         Toggle event count zeroing across display updates.
 270
 271 [qQ]::
 272         Quit.
 273
 274 Pressing any unmapped key displays a menu, and prompts for input.
 275
 276 include::callchain-overhead-calculation.txt[]
 277
 278 SEE ALSO
 279 --------
 280 linkperf:perf-stat[1], linkperf:perf-list[1], linkperf:perf-report[1]