perf-top.txt 4.1 KB
Newer Older
I
Ingo Molnar 已提交
1
perf-top(1)
2
===========
I
Ingo Molnar 已提交
3 4 5

NAME
----
M
Mike Galbraith 已提交
6
perf-top - System profiling tool.
I
Ingo Molnar 已提交
7 8 9 10

SYNOPSIS
--------
[verse]
M
Mike Galbraith 已提交
11
'perf top' [-e <EVENT> | --event=EVENT] [<options>]
I
Ingo Molnar 已提交
12 13 14

DESCRIPTION
-----------
S
Shawn Bohrer 已提交
15
This command generates and displays a performance counter profile in real time.
I
Ingo Molnar 已提交
16 17 18 19


OPTIONS
-------
M
Mike Galbraith 已提交
20 21 22 23 24 25 26 27
-a::
--all-cpus::
        System-wide collection.  (default)

-c <count>::
--count=<count>::
	Event period to sample.

28 29
-C <cpu-list>::
--cpu=<cpu>::
S
Shawn Bohrer 已提交
30 31
Monitor only on the list of CPUs provided. Multiple CPUs can be provided as a
comma-separated list with no space: 0,1. Ranges of CPUs are specified with -: 0-2.
32
Default is to monitor all CPUS.
M
Mike Galbraith 已提交
33 34 35 36

-d <seconds>::
--delay=<seconds>::
	Number of seconds to delay between refreshes.
I
Ingo Molnar 已提交
37

M
Mike Galbraith 已提交
38 39
-e <event>::
--event=<event>::
40 41 42
	Select the PMU event. Selection can be a symbolic event name
	(use 'perf list' to list all events) or a raw PMU
	event (eventsel+umask) in the form of rNNN where NNN is a
M
Mike Galbraith 已提交
43
	hexadecimal event descriptor.
I
Ingo Molnar 已提交
44

M
Mike Galbraith 已提交
45 46 47 48 49 50 51 52
-E <entries>::
--entries=<entries>::
	Display this many functions.

-f <count>::
--count-filter=<count>::
	Only display functions with more events than this.

S
Shawn Bohrer 已提交
53 54 55
--group::
        Put the counters into a counter group.

M
Mike Galbraith 已提交
56 57 58 59 60 61
-F <freq>::
--freq=<freq>::
	Profile at this frequency.

-i::
--inherit::
62
	Child tasks do not inherit counters.
M
Mike Galbraith 已提交
63 64 65 66 67 68 69

-k <path>::
--vmlinux=<path>::
	Path to vmlinux.  Required for annotation functionality.

-m <pages>::
--mmap-pages=<pages>::
70 71 72
	Number of mmap data pages (must be a power of two) or size
	specification with appended unit character - B/K/M/G. The
	size is rounded up to have nearest pages power of two value.
M
Mike Galbraith 已提交
73 74 75

-p <pid>::
--pid=<pid>::
76
	Profile events on existing Process ID (comma separated list).
S
Shawn Bohrer 已提交
77 78 79

-t <tid>::
--tid=<tid>::
80
        Profile events on existing thread ID (comma separated list).
M
Mike Galbraith 已提交
81

82 83 84 85
-u::
--uid=::
        Record events in threads owned by uid. Name or number.

M
Mike Galbraith 已提交
86 87 88 89 90
-r <priority>::
--realtime=<priority>::
	Collect data with this RT SCHED_FIFO priority.

--sym-annotate=<symbol>::
91
        Annotate this symbol.
M
Mike Galbraith 已提交
92

S
Shawn Bohrer 已提交
93 94 95 96 97 98 99 100 101 102 103 104
-K::
--hide_kernel_symbols::
        Hide kernel symbols.

-U::
--hide_user_symbols::
        Hide user symbols.

-D::
--dump-symtab::
        Dump the symbol table used for profiling.

M
Mike Galbraith 已提交
105 106 107 108 109 110 111 112
-v::
--verbose::
	Be more verbose (show counter open errors, etc).

-z::
--zero::
	Zero history across display updates.

113 114
-s::
--sort::
115
	Sort by key(s): pid, comm, dso, symbol, parent, srcline, weight,
116
	local_weight, abort, in_tx, transaction
117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133

-n::
--show-nr-samples::
	Show a column with the number of samples.

--show-total-period::
	Show a column with the sum of periods.

--dsos::
	Only consider symbols in these dsos.

--comms::
	Only consider symbols in these comms.

--symbols::
	Only consider these symbols.

134 135 136 137 138 139 140 141 142 143
-M::
--disassembler-style=:: Set disassembler style for objdump.

--source::
	Interleave source code with assembly code. Enabled by default,
	disable with --no-source.

--asm-raw::
	Show raw instruction encoding of assembly instructions.

144
-g::
J
Jiri Olsa 已提交
145 146
	Enables call-graph (stack chain/backtrace) recording.

147
--call-graph::
J
Jiri Olsa 已提交
148
	Setup and enable call-graph (stack chain/backtrace) recording,
149
	implies -g.
150

151 152 153 154 155 156 157 158
--max-stack::
	Set the stack depth limit when parsing the callchain, anything
	beyond the specified depth will be ignored. This is a trade-off
	between information loss and faster processing especially for
	workloads that can have a very long callchain stack.

	Default: 127

159 160 161 162 163
--ignore-callees=<regex>::
        Ignore callees of the function(s) matching the given regex.
        This has the effect of collecting the callers of each such
        function into one place in the call-graph tree.

164 165 166 167
--percent-limit::
	Do not show entries which have an overhead under that percent.
	(Default: 0).

M
Mike Galbraith 已提交
168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198
INTERACTIVE PROMPTING KEYS
--------------------------

[d]::
	Display refresh delay.

[e]::
	Number of entries to display.

[E]::
	Event to display when multiple counters are active.

[f]::
	Profile display filter (>= hit count).

[F]::
	Annotation display filter (>= % of total).

[s]::
	Annotate symbol.

[S]::
	Stop annotation, return to full profile display.

[z]::
	Toggle event count zeroing across display updates.

[qQ]::
	Quit.

Pressing any unmapped key displays a menu, and prompts for input.
I
Ingo Molnar 已提交
199 200 201 202


SEE ALSO
--------
203
linkperf:perf-stat[1], linkperf:perf-list[1]