1PERF-FTRACE(1) perf Manual PERF-FTRACE(1)
2
3
4
6 perf-ftrace - simple wrapper for kernel's ftrace functionality
7
9 perf ftrace <command>
10
12 The perf ftrace command is a simple wrapper of kernel’s ftrace
13 functionality. It only supports single thread tracing currently and
14 just reads trace_pipe in text and then write it to stdout.
15
16 The following options apply to perf ftrace.
17
19 -t, --tracer=
20 Tracer to use when neither -G nor -F option is not specified:
21 function_graph or function.
22
23 -v, --verbose
24 Increase the verbosity level.
25
26 -F, --funcs
27 List available functions to trace. It accepts a pattern to only
28 list interested functions.
29
30 -p, --pid=
31 Trace on existing process id (comma separated list).
32
33 --tid=
34 Trace on existing thread id (comma separated list).
35
36 -D, --delay
37 Time (ms) to wait before starting tracing after program start.
38
39 -a, --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, --cpu=
46 Only trace for the list of CPUs provided. Multiple CPUs can be
47 provided as a comma separated list with no space like: 0,1. Ranges
48 of CPUs are specified with -: 0-2. Default is to trace on all
49 online CPUs.
50
51 -m, --buffer-size
52 Set the size of per-cpu tracing buffer, <size> is expected to be a
53 number with appended unit character - B/K/M/G.
54
55 --inherit
56 Trace children processes spawned by our target.
57
58 -T, --trace-funcs=
59 Select function tracer and set function filter on the given
60 function (or a glob pattern). Multiple functions can be given by
61 using this option more than once. The function argument also can be
62 a glob pattern. It will be passed to set_ftrace_filter in tracefs.
63
64 -N, --notrace-funcs=
65 Select function tracer and do not trace functions given by the
66 argument. Like -T option, this can be used more than once to
67 specify multiple functions (or glob patterns). It will be passed to
68 set_ftrace_notrace in tracefs.
69
70 --func-opts
71 List of options allowed to set: call-graph - Display kernel stack
72 trace for function tracer. irq-info - Display irq context info for
73 function tracer.
74
75 -G, --graph-funcs=
76 Select function_graph tracer and set graph filter on the given
77 function (or a glob pattern). This is useful to trace for functions
78 executed from the given function. This can be used more than once
79 to specify multiple functions. It will be passed to
80 set_graph_function in tracefs.
81
82 -g, --nograph-funcs=
83 Select function_graph tracer and set graph notrace filter on the
84 given function (or a glob pattern). Like -G option, this is useful
85 for the function_graph tracer only and disables tracing for
86 function executed from the given function. This can be used more
87 than once to specify multiple functions. It will be passed to
88 set_graph_notrace in tracefs.
89
90 --graph-opts
91 List of options allowed to set: nosleep-time - Measure on-CPU time
92 only for function_graph tracer. noirqs - Ignore functions that
93 happen inside interrupt. verbose - Show process names, PIDs,
94 timestamps, etc. thresh=<n> - Setup trace duration threshold in
95 microseconds. depth=<n> - Set max depth for function graph tracer
96 to follow.
97
99 perf-record(1), perf-trace(1)
100
101
102
103perf 06/03/2021 PERF-FTRACE(1)