'\" t
.\"     Title: perf-bench
.\"    Author: [see the "AUTHOR(S)" section]
.\" Generator: Asciidoctor 1.5.7.1
.\"      Date: 2018-11-27
.\"    Manual: perf Manual
.\"    Source: perf
.\"  Language: English
.\"
.TH "PERF_4.19\-BENCH" "1" "2018-11-27" "perf" "perf Manual"
.ie \n(.g .ds Aq \(aq
.el       .ds Aq '
.ss \n[.ss] 0
.nh
.ad l
.de URL
\fI\\$2\fP <\\$1>\\$3
..
.als MTO URL
.if \n[.g] \{\
.  mso www.tmac
.  am URL
.    ad l
.  .
.  am MTO
.    ad l
.  .
.  LINKSTYLE blue R < >
.\}
.SH "NAME"
perf\-bench \- General framework for benchmark suites
.SH "SYNOPSIS"
.sp
.nf
\fIperf bench\fP [<common options>] <subsystem> <suite> [<options>]
.fi
.br
.SH "DESCRIPTION"
.sp
This \fIperf bench\fP command is a general framework for benchmark suites.
.SH "COMMON OPTIONS"
.sp
\-r, \-\-repeat=
.RS 4
Specify amount of times to repeat the run (default 10).
.RE
.sp
\-f, \-\-format=
.RS 4
Specify format style.
Current available format styles are:
.RE
.sp
\fIdefault\fP
.RS 4
Default style. This is mainly for human reading.
.RE
.sp
.if n .RS 4
.nf
% perf bench sched pipe                      # with no style specified
(executing 1000000 pipe operations between two tasks)
        Total time:5.855 sec
                5.855061 usecs/op
                170792 ops/sec
.fi
.if n .RE
.sp
\fIsimple\fP
.RS 4
This simple style is friendly for automated
processing by scripts.
.RE
.sp
.if n .RS 4
.nf
% perf bench \-\-format=simple sched pipe      # specified simple
5.988
.fi
.if n .RE
.SH "SUBSYSTEM"
.sp
\fIsched\fP
.RS 4
Scheduler and IPC mechanisms.
.RE
.sp
\fImem\fP
.RS 4
Memory access performance.
.RE
.sp
\fInuma\fP
.RS 4
NUMA scheduling and MM benchmarks.
.RE
.sp
\fIfutex\fP
.RS 4
Futex stressing benchmarks.
.RE
.sp
\fIall\fP
.RS 4
All benchmark subsystems.
.RE
.SS "SUITES FOR \fIsched\fP"
.sp
\fBmessaging\fP
.RS 4
Suite for evaluating performance of scheduler and IPC mechanisms.
Based on hackbench by Rusty Russell.
.RE
.SS "Options of \fBmessaging\fP"
.sp
\-p, \-\-pipe
.RS 4
Use pipe() instead of socketpair()
.RE
.sp
\-t, \-\-thread
.RS 4
Be multi thread instead of multi process
.RE
.sp
\-g, \-\-group=
.RS 4
Specify number of groups
.RE
.sp
\-l, \-\-nr_loops=
.RS 4
Specify number of loops
.RE
.SS "Example of \fBmessaging\fP"
.sp
.if n .RS 4
.nf
% perf bench sched messaging                 # run with default
options (20 sender and receiver processes per group)
(10 groups == 400 processes run)

      Total time:0.308 sec

% perf bench sched messaging \-t \-g 20        # be multi\-thread, with 20 groups
(20 sender and receiver threads per group)
(20 groups == 800 threads run)

      Total time:0.582 sec
.fi
.if n .RE
.sp
\fBpipe\fP
.RS 4
Suite for pipe() system call.
Based on pipe\-test\-1m.c by Ingo Molnar.
.RE
.SS "Options of \fBpipe\fP"
.sp
\-l, \-\-loop=
.RS 4
Specify number of loops.
.RE
.SS "Example of \fBpipe\fP"
.sp
.if n .RS 4
.nf
% perf bench sched pipe
(executing 1000000 pipe operations between two tasks)

        Total time:8.091 sec
                8.091833 usecs/op
                123581 ops/sec

% perf bench sched pipe \-l 1000              # loop 1000
(executing 1000 pipe operations between two tasks)

        Total time:0.016 sec
                16.948000 usecs/op
                59004 ops/sec
.fi
.if n .RE
.SS "SUITES FOR \fImem\fP"
.sp
\fBmemcpy\fP
.RS 4
Suite for evaluating performance of simple memory copy in various ways.
.RE
.SS "Options of \fBmemcpy\fP"
.sp
\-l, \-\-size
.RS 4
Specify size of memory to copy (default: 1MB).
Available units are B, KB, MB, GB and TB (case insensitive).
.RE
.sp
\-f, \-\-function
.RS 4
Specify function to copy (default: default).
Available functions are depend on the architecture.
On x86\-64, x86\-64\-unrolled, x86\-64\-movsq and x86\-64\-movsb are supported.
.RE
.sp
\-l, \-\-nr_loops
.RS 4
Repeat memcpy invocation this number of times.
.RE
.sp
\-c, \-\-cycles
.RS 4
Use perf\(cqs cpu\-cycles event instead of gettimeofday syscall.
.RE
.sp
\fBmemset\fP
.RS 4
Suite for evaluating performance of simple memory set in various ways.
.RE
.SS "Options of \fBmemset\fP"
.sp
\-l, \-\-size
.RS 4
Specify size of memory to set (default: 1MB).
Available units are B, KB, MB, GB and TB (case insensitive).
.RE
.sp
\-f, \-\-function
.RS 4
Specify function to set (default: default).
Available functions are depend on the architecture.
On x86\-64, x86\-64\-unrolled, x86\-64\-stosq and x86\-64\-stosb are supported.
.RE
.sp
\-l, \-\-nr_loops
.RS 4
Repeat memset invocation this number of times.
.RE
.sp
\-c, \-\-cycles
.RS 4
Use perf\(cqs cpu\-cycles event instead of gettimeofday syscall.
.RE
.SS "SUITES FOR \fInuma\fP"
.sp
\fBmem\fP
.RS 4
Suite for evaluating NUMA workloads.
.RE
.SS "SUITES FOR \fIfutex\fP"
.sp
\fBhash\fP
.RS 4
Suite for evaluating hash tables.
.RE
.sp
\fBwake\fP
.RS 4
Suite for evaluating wake calls.
.RE
.sp
\fBwake\-parallel\fP
.RS 4
Suite for evaluating parallel wake calls.
.RE
.sp
\fBrequeue\fP
.RS 4
Suite for evaluating requeue calls.
.RE
.sp
\fBlock\-pi\fP
.RS 4
Suite for evaluating futex lock_pi calls.
.RE
.SH "SEE ALSO"
.sp
perf(1)