perf bench

Usage:
        perf bench [<common options>] <subsystem> <suite> [<options>]


# List of available subsystems...


         sched: scheduler and IPC mechanism
           mem: memory access performance
           all: test all subsystem (pseudo subsystem)


ryanel@ryane-virtual-machine:~$ perf bench all
# Running sched/messaging benchmark...
# 20 sender and receiver processes per group
# 10 groups == 400 processes run


     Total time: 0.727 [sec]


# Running sched/pipe benchmark...
# Executed 1000000 pipe operations between two tasks


     Total time: 2.541 [sec]


       2.541074 usecs/op
         393534 ops/sec


# Running mem/memcpy benchmark...
# Copying 1MB Bytes ...


       1.778802 GB/Sec
      19.929847 GB/Sec (with prefault)


# Running mem/memset benchmark...
# Copying 1MB Bytes ...


       4.584800 GB/Sec
       7.454676 GB/Sec (with prefault)

转载于:https://my.oschina.net/rinehart/blog/140334