| Name | Modified | Size | Downloads / Week |
|---|---|---|---|
| Parent folder | |||
| subdir.mk | 2014-05-07 | 915 Bytes | |
| SPAN_functions.o | 2014-05-07 | 12.3 kB | |
| SPAN_functions.d | 2014-05-07 | 229 Bytes | |
| perf.o | 2014-05-07 | 1.8 kB | |
| perf.d | 2014-05-07 | 101 Bytes | |
| main.o | 2014-05-07 | 2.9 kB | |
| main.d | 2014-05-07 | 87 Bytes | |
| hash_table.o | 2014-05-07 | 5.4 kB | |
| hash_table.d | 2014-05-07 | 134 Bytes | |
| common.o | 2014-05-07 | 840 Bytes | |
| common.d | 2014-05-07 | 77 Bytes | |
| Totals: 11 Items | 24.8 kB | 0 | |
This program is used to generate function (subroutine call) dynamic power profile.
There are general several things to notice in order to use the library.
1. The program relis on GCC compilier and -finstrument-function option.
2. The version of your Linux kernel must be higher than 2.6.31 to the system callses the system call "__NR_perf_counter_open"(in the latest version, the name is "__NR_perf_event_open") to stat the hardware events, such as L2 cache miss.
3.
4. instrument directory contains source code to build the library. analysis directory generate the power profiling.
5.
EXAMPLE: your source file example_is.c. The default compiler is GCC.
1. go to directory Debug/.
type "make".
it will generate a complied version of the IS benchmark from the NPB benchmark suite.
2. Use SPAN_initialize(),SPAN_start("full_verify"), and SPAN_stop("full_verify") to indicate profiling area.
3. run a.out
4. Output of the program is a sequence of power values with their process IDs (mulit-threaded application)
