6 perf-top - System profiling tool.
11 'perf top' [-e <EVENT> | --event=EVENT] [<options>]
15 This command generates and displays a performance counter profile in real time.
22 System-wide collection. (default)
26 Event period to sample.
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.
36 Number of seconds to delay between refreshes.
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.
47 Display this many functions.
50 --count-filter=<count>::
51 Only display functions with more events than this.
55 Put the counters into a counter group.
59 Profile at this frequency.
63 Child tasks inherit counters, only makes sens with -p option.
67 Path to vmlinux. Required for annotation functionality.
70 --mmap-pages=<pages>::
71 Number of mmapped data pages.
75 Profile events on existing Process ID.
79 Profile events on existing thread ID.
82 --realtime=<priority>::
83 Collect data with this RT SCHED_FIFO priority.
86 --sym-annotate=<symbol>::
90 --hide_kernel_symbols::
99 Dump the symbol table used for profiling.
103 Be more verbose (show counter open errors, etc).
107 Zero history across display updates.
111 Sort by key(s): pid, comm, dso, symbol, parent
115 Show a column with the number of samples.
117 --show-total-period::
118 Show a column with the sum of periods.
121 Only consider symbols in these dsos.
124 Only consider symbols in these comms.
127 Only consider these symbols.
130 --disassembler-style=:: Set disassembler style for objdump.
133 Interleave source code with assembly code. Enabled by default,
134 disable with --no-source.
137 Show raw instruction encoding of assembly instructions.
139 -G [type,min,order]::
141 Display call chains using type, min percent threshold and order.
143 - flat: single column, linear exposure of call chains.
144 - graph: use a graph tree, displaying absolute overhead rates.
145 - fractal: like graph, but displays relative rates. Each branch of
146 the tree is considered as a new profiled object.
149 - callee: callee based call graph.
150 - caller: inverted caller based call graph.
152 Default: fractal,0.5,callee.
154 INTERACTIVE PROMPTING KEYS
155 --------------------------
158 Display refresh delay.
161 Number of entries to display.
164 Event to display when multiple counters are active.
167 Profile display filter (>= hit count).
170 Annotation display filter (>= % of total).
176 Stop annotation, return to full profile display.
179 Toggle event count zeroing across display updates.
184 Pressing any unmapped key displays a menu, and prompts for input.
189 linkperf:perf-stat[1], linkperf:perf-list[1]