6 perf-bench - General framework for benchmark suites
11 'perf bench' [<common options>] <subsystem> <suite> [<options>]
15 This 'perf bench' command is a general framework for benchmark suites.
22 Current available format styles are:
25 Default style. This is mainly for human reading.
27 % perf bench sched pipe # with no style specified
28 (executing 1000000 pipe operations between two tasks)
35 This simple style is friendly for automated
36 processing by scripts.
38 % perf bench --format=simple sched pipe # specified simple
46 Scheduler and IPC mechanisms.
49 Memory access performance.
52 All benchmark subsystems.
57 Suite for evaluating performance of scheduler and IPC mechanisms.
58 Based on hackbench by Rusty Russell.
60 Options of *messaging*
61 ^^^^^^^^^^^^^^^^^^^^^^
64 Use pipe() instead of socketpair()
68 Be multi thread instead of multi process
72 Specify number of groups
76 Specify number of loops
78 Example of *messaging*
79 ^^^^^^^^^^^^^^^^^^^^^^
82 % perf bench sched messaging # run with default
83 options (20 sender and receiver processes per group)
84 (10 groups == 400 processes run)
88 % perf bench sched messaging -t -g 20 # be multi-thread, with 20 groups
89 (20 sender and receiver threads per group)
90 (20 groups == 800 threads run)
96 Suite for pipe() system call.
97 Based on pipe-test-1m.c by Ingo Molnar.
103 Specify number of loops.
108 ---------------------
109 % perf bench sched pipe
110 (executing 1000000 pipe operations between two tasks)
116 % perf bench sched pipe -l 1000 # loop 1000
117 (executing 1000 pipe operations between two tasks)
122 ---------------------
127 Suite for evaluating performance of simple memory copy in various ways.
133 Specify length of memory to copy (default: 1MB).
134 Available units are B, KB, MB, GB and TB (case insensitive).
138 Specify routine to copy (default: default).
139 Available routines are depend on the architecture.
140 On x86-64, x86-64-unrolled, x86-64-movsq and x86-64-movsb are supported.
144 Repeat memcpy invocation this number of times.
148 Use perf's cpu-cycles event instead of gettimeofday syscall.
152 Show only the result with page faults before memcpy.
156 Show only the result without page faults before memcpy.
159 Suite for evaluating performance of simple memory set in various ways.
165 Specify length of memory to set (default: 1MB).
166 Available units are B, KB, MB, GB and TB (case insensitive).
170 Specify routine to set (default: default).
171 Available routines are depend on the architecture.
172 On x86-64, x86-64-unrolled, x86-64-stosq and x86-64-stosb are supported.
176 Repeat memset invocation this number of times.
180 Use perf's cpu-cycles event instead of gettimeofday syscall.
184 Show only the result with page faults before memset.
188 Show only the result without page faults before memset.