perf-trace.txt 1.8 KB
Newer Older
A
Arnaldo Carvalho de Melo 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
perf-trace(1)
=============

NAME
----
perf-trace - strace inspired tool

SYNOPSIS
--------
[verse]
'perf trace'

DESCRIPTION
-----------
This command will show the events associated with the target, initially
syscalls, but other system events like pagefaults, task lifetime events,
scheduling events, etc.

Initially this is a live mode only tool, but eventually will work with
perf.data files like the other tools, allowing a detached 'record' from
analysis phases.

OPTIONS
-------

26
-a::
A
Arnaldo Carvalho de Melo 已提交
27 28 29
--all-cpus::
        System-wide collection from all CPUs.

30 31 32
-e::
--expr::
	List of events to show, currently only syscall names.
33 34
	Prefixing with ! shows all syscalls but the ones specified.  You may
	need to escape it.
35

36 37 38 39
-o::
--output=::
	Output file name.

A
Arnaldo Carvalho de Melo 已提交
40 41 42 43
-p::
--pid=::
	Record events on existing process ID (comma separated list).

44
-t::
A
Arnaldo Carvalho de Melo 已提交
45 46 47
--tid=::
        Record events on existing thread ID (comma separated list).

48
-u::
A
Arnaldo Carvalho de Melo 已提交
49 50 51
--uid=::
        Record events in threads owned by uid. Name or number.

52 53 54 55
-v::
--verbose=::
        Verbosity level.

56
-i::
A
Arnaldo Carvalho de Melo 已提交
57 58 59
--no-inherit::
	Child tasks do not inherit counters.

60
-m::
A
Arnaldo Carvalho de Melo 已提交
61 62 63
--mmap-pages=::
	Number of mmap data pages. Must be a power of two.

64
-C::
A
Arnaldo Carvalho de Melo 已提交
65 66 67 68 69 70
--cpu::
Collect samples 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.
In per-thread mode with inheritance mode on (default), Events are captured only when
the thread executes on the designated CPUs. Default is to monitor all CPUs.

71 72 73
--duration:
	Show only events that had a duration greater than N.M ms.

74 75 76
--sched:
	Accrue thread runtime and provide a summary at the end of the session.

77 78 79 80
-i
--input
	Process events from a given perf data file.

81 82 83 84
-T
--time
	Print full timestamp rather time relative to first sample.

A
Arnaldo Carvalho de Melo 已提交
85 86 87
SEE ALSO
--------
linkperf:perf-record[1], linkperf:perf-script[1]