1PERF-RECORD(1)                    perf Manual                   PERF-RECORD(1)
2
3
4

NAME

6       perf-record - Run a command and record its profile into perf.data
7

SYNOPSIS

9       perf record [-e <EVENT> | --event=EVENT] [-l] [-a] <command>
10       perf record [-e <EVENT> | --event=EVENT] [-l] [-a] — <command> [<options>]
11

DESCRIPTION

13       This command runs a command and gathers a performance counter profile
14       from it, into perf.data - without displaying anything.
15
16       This file can then be inspected later on, using perf report.
17

OPTIONS

19       <command>...
20           Any command you can specify in a shell.
21
22       -e, --event=
23           Select the PMU event. Selection can be:
24
25           ·   a symbolic event name (use perf list to list all events)
26
27           ·   a raw PMU event (eventsel+umask) in the form of rNNN where NNN
28               is a hexadecimal event descriptor.
29
30           ·   a hardware breakpoint event in the form of \mem:addr[:access]
31               where addr is the address in memory you want to break in.
32               Access is the memory access type (read, write, execute) it can
33               be passed as follows: \mem:addr[:[r][w][x]]. If you want to
34               profile read-write accesses in 0x1000, just set mem:0x1000:rw.
35
36       -a
37           System-wide collection.
38
39       -l
40           Scale counter values.
41
42       -p, --pid=
43           Record events on existing pid.
44
45       -r, --realtime=
46           Collect data with this RT SCHED_FIFO priority.
47
48       -A, --append
49           Append to the output file to do incremental profiling.
50
51       -f, --force
52           Overwrite existing data file. (deprecated)
53
54       -c, --count=
55           Event period to sample.
56
57       -o, --output=
58           Output file name.
59
60       -i, --no-inherit
61           Child tasks do not inherit counters.
62
63       -F, --freq=
64           Profile at this frequency.
65
66       -m, --mmap-pages=
67           Number of mmap data pages.
68
69       -g, --call-graph
70           Do call-graph (stack chain/backtrace) recording.
71
72       -v, --verbose
73           Be more verbose (show counter open errors, etc).
74
75       -s, --stat
76           Per thread counts.
77
78       -d, --data
79           Sample addresses.
80
81       -n, --no-samples
82           Don’t sample.
83
84       -R, --raw-samples
85           Collect raw sample records from all opened counters (default for
86           tracepoint counters).
87

SEE ALSO

89       perf-stat(1), perf-list(1)
90
91
92
93perf 2.6.35.14-106.fc             11/23/2011                    PERF-RECORD(1)
Impressum