LMbench

Description
lmbench is a suite of simple, portable, ANSI/C microbenchmarks for UNIX/POSIX. In general, it measures two key features: latency and bandwidth. lmbench is intended to give system developers insight into basic costs of key operations.

Home Page SourceForge Page

Compilation
Untar lmbench package tar zxvf lmbench.tar.gz Change to lmbench directory cd /path/to/lmbench/ Compile using the following make OS=linux CC=arm-none-linux-gnueabi-gcc build Copy lmbench directory to your filesystem, it is recommended to put it into a ramdisk

Running in OMAP
Change to scripts directory under lmbench cd /path/to/omap/lmbench/scripts/ Start running config-run script you will be prompted to give some settings, e.g.
 * 1) ./config-run

MULTIPLE COPIES [default 1]: 1 Job placement selection [default 1]: 1 > Allow scheduler to place jobs MB [default 84]: SUBSET (ALL|HARWARE|OS|DEVELOPMENT) [default all]: all FASTMEM [default no]: SLOWFS [default no]: DISKS [default none] REMOTE [default none] Processor mhz [default 498 MHz, 2.0080 nanosec clock] FSDIR [default /tmp] Status output file [default /dev/tty] Mail results [default yes]

Then start lmbench by executing ./results Results will be saved under /path/to/omap/lmbench/results/linux/3430SDP-3430.0 Copy this file to a linux machine where you have lmbench package, perl sould be installed and then execute getsummary script /path/to/lmbench/scripts/getsummary 3430SDP-3430.0 A formatted output shall be displayed