perf/Documentation/perf-bench.txt

1 perf-bench(1)
5 ----
6 perf-bench - General framework for benchmark suites
9 --------
14 -----------
18 --------------
19 -r::
20 --repeat=::
23 -f::
24 --format=::
30 ---------------------
36 ---------------------
41 ---------------------
42 % perf bench --format=simple sched pipe      # specified simple
44 ---------------------
47 ---------
53 	System call performance (throughput).
81 -p::
82 --pipe::
85 -t::
86 --thread::
87 Be multi thread instead of multi process
89 -g::
90 --group=::
93 -l::
94 --nr_loops=::
100 ---------------------
107 % perf bench sched messaging -t -g 20        # be multi-thread, with 20 groups
112 ---------------------
115 Suite for pipe() system call.
116 Based on pipe-test-1m.c by Ingo Molnar.
120 -l::
121 --loop=::
127 ---------------------
135 % perf bench sched pipe -l 1000              # loop 1000
141 ---------------------
146 Suite for evaluating performance of core system call throughput (both usecs/op and ops/sec metrics).
158 -l::
159 --size::
163 -f::
164 --function::
167 On x86-64, x86-64-unrolled, x86-64-movsq and x86-64-movsb are supported.
169 -l::
170 --nr_loops::
173 -c::
174 --cycles::
175 Use perf's cpu-cycles event instead of gettimeofday syscall.
182 -l::
183 --size::
187 -f::
188 --function::
191 On x86-64, x86-64-unrolled, x86-64-stosq and x86-64-stosb are supported.
193 -l::
194 --nr_loops::
197 -c::
198 --cycles::
199 Use perf's cpu-cycles event instead of gettimeofday syscall.
214 *wake-parallel*::
220 *lock-pi*::
237 --------