2012/12/5 David Long <dave.long@linaro.org>

On 12/04/12 13:46, Scott Bambrough wrote:

+ Dave Long

On Tue, Dec 4, 2012 at 3:37 AM, Peng Fan <xfan87@gmail.com <mailto:xfan87@gmail.com>> wrote:

I use the option `-v`, and I can see that the first number of
branches, instructions branch-misses and so on is zero.why?

root@android:/ # perf stat -v echo "linaro"
linaro
task-clock: 93902590 93902590 93902590

context-switches: 9 93902590 93902590

CPU-migrations: 0 93902590 93902590
page-faults: 104 93902590 93902590
cycles: 4284725 76202392 76202392
stalled-cycles-frontend: 0 76202392 76202392
stalled-cycles-backend: 0 76202392 76202392
instructions: 0 76202392 76202392
branches: 0 76202392 76202392
branch-misses: 0 76202392 76202392

Performance counter stats for 'echo linaro':

93.902590 task-clock # 0.282 CPUs utilized
9 context-switches # 0.000 M/sec
0 CPU-migrations # 0.000 M/sec
104 page-faults # 0.001 M/sec
4284725 cycles # 0.046 GHz
* 0 stalled-cycles-frontend # 0.00% frontend

cycles idle
0 stalled-cycles-backend # 0.00% backend cycles idle
0 instructions # 0.00 insns per
cycle
0 branches # 0.000 M/sec
0 branch-misses # 0.00% of all
branches *

0.332550049 seconds time elapsed

Can anybody give me suggestions?How can I use perf to get the
value of hardware performance counter

Using a recent 3.4.19 kernel I am unable to reproduce your results:

context-switches: 0 8941650 8941650
CPU-migrations: 0 8941650 8941650
page-faults: 147 8941650 8941650
cycles: 2665204 8941650 8941650
stalled-cycles-frontend: 365973 8941650 8941650
stalled-cycles-backend: 1748465 8941650 8941650
instructions: 1368842 8941650 8941650
branches: 152838 8941650 8941650
branch-misses: 53924 8941650 8941650

Performance counter stats for 'echo linaro':

8.941650 task-clock # 0.740 CPUs utilized
0 context-switches # 0.000 K/sec
0 CPU-migrations # 0.000 K/sec
147 page-faults # 0.016 M/sec
2665204 cycles # 0.298 GHz
365973 stalled-cycles-frontend # 13.73% frontend cycles idle
1748465 stalled-cycles-backend # 65.60% backend cycles idle
1368842 instructions # 0.51 insns per cycle
# 1.28 stalled cycles per insn
152838 branches # 17.093 M/sec
53924 branch-misses # 35.28% of all branches

0.012084973 seconds time elapsed

Significant changes were made the the counter/event kernel code about the time of the release. I do not believe they made it in in time. Please try simply using a slightly more recent kernel.

--
David Long
Kernel Engineer, TI Landing Team
Linaro