1perf-ftrace(1)
2==============
3
4NAME
5----
6perf-ftrace - simple wrapper for kernel's ftrace functionality
7
8
9SYNOPSIS
10--------
11[verse]
12'perf ftrace' {trace|latency} <command>
13
14DESCRIPTION
15-----------
16The 'perf ftrace' command provides a collection of subcommands which use
17kernel's ftrace infrastructure.
18
19  'perf ftrace trace' is a simple wrapper of the ftrace.  It only supports
20  single thread tracing currently and just reads trace_pipe in text and then
21  write it to stdout.
22
23  'perf ftrace latency' calculates execution latency of a given function
24  (optionally with BPF) and display it as a histogram.
25
26The following options apply to perf ftrace.
27
28COMMON OPTIONS
29--------------
30
31-p::
32--pid=::
33	Trace on existing process id (comma separated list).
34
35--tid=::
36	Trace on existing thread id (comma separated list).
37
38-a::
39--all-cpus::
40	Force system-wide collection.  Scripts run without a <command>
41	normally use -a by default, while scripts run with a <command>
42	normally don't - this option allows the latter to be run in
43	system-wide mode.
44
45-C::
46--cpu=::
47	Only trace for the list of CPUs provided.  Multiple CPUs can
48	be provided as a comma separated list with no space like: 0,1.
49	Ranges of CPUs are specified with -: 0-2.
50	Default is to trace on all online CPUs.
51
52-v::
53--verbose::
54        Increase the verbosity level.
55
56
57OPTIONS for 'perf ftrace trace'
58-------------------------------
59
60-t::
61--tracer=::
62	Tracer to use when neither -G nor -F option is not
63	specified: function_graph or function.
64
65-F::
66--funcs::
67        List available functions to trace. It accepts a pattern to
68        only list interested functions.
69
70-D::
71--delay::
72	Time (ms) to wait before starting tracing after program start.
73
74-m::
75--buffer-size::
76	Set the size of per-cpu tracing buffer, <size> is expected to
77	be a number with appended unit character - B/K/M/G.
78
79--inherit::
80	Trace children processes spawned by our target.
81
82-T::
83--trace-funcs=::
84	Select function tracer and set function filter on the given
85	function (or a glob pattern). Multiple functions can be given
86	by using this option more than once. The function argument also
87	can be a glob pattern. It will be passed to 'set_ftrace_filter'
88	in tracefs.
89
90-N::
91--notrace-funcs=::
92	Select function tracer and do not trace functions given by the
93	argument.  Like -T option, this can be used more than once to
94	specify multiple functions (or glob patterns).  It will be
95	passed to 'set_ftrace_notrace' in tracefs.
96
97--func-opts::
98	List of options allowed to set:
99	  call-graph - Display kernel stack trace for function tracer.
100	  irq-info   - Display irq context info for function tracer.
101
102-G::
103--graph-funcs=::
104	Select function_graph tracer and set graph filter on the given
105	function (or a glob pattern). This is useful to trace for
106	functions executed from the given function. This can be used more
107	than once to specify multiple functions. It will be passed to
108	'set_graph_function' in tracefs.
109
110-g::
111--nograph-funcs=::
112	Select function_graph tracer and set graph notrace filter on the
113	given function (or a glob pattern). Like -G option, this is useful
114	for the function_graph tracer only and disables tracing for function
115	executed from the given function. This can be used more than once to
116	specify multiple functions. It will be passed to 'set_graph_notrace'
117	in tracefs.
118
119--graph-opts::
120	List of options allowed to set:
121	  nosleep-time - Measure on-CPU time only for function_graph tracer.
122	  noirqs       - Ignore functions that happen inside interrupt.
123	  verbose      - Show process names, PIDs, timestamps, etc.
124	  thresh=<n>   - Setup trace duration threshold in microseconds.
125	  depth=<n>    - Set max depth for function graph tracer to follow.
126
127
128OPTIONS for 'perf ftrace latency'
129---------------------------------
130
131-T::
132--trace-funcs=::
133	Set the function name to get the histogram.  Unlike perf ftrace trace,
134	it only allows single function to calculate the histogram.
135
136-b::
137--use-bpf::
138	Use BPF to measure function latency instead of using the ftrace (it
139	uses function_graph tracer internally).
140
141-n::
142--use-nsec::
143	Use nano-second instead of micro-second as a base unit of the histogram.
144
145
146SEE ALSO
147--------
148linkperf:perf-record[1], linkperf:perf-trace[1]
149