On 2/21/15 2:44 PM, Benjamin Moody wrote:
Hi,
Trying using a program with established results. e.g., This is from
PeterZ years ago. On intel 'perf stat -e instructions a.out' should show
~1 billion (a wee bit more, but clearly in the 1b range).
That program itself gave the expected results, but it's a good
suggestion. When I tried changing that program to also fork a bunch
of child processes, I started getting similarly weird results to what
I had seen before.
This is with the 1billion instructions sent earlier:
$ ./myperfstat ./1bi
Total instructions: 1000087491
$ perf stat -e instructions:u ./1bi
Performance counter stats for './1bi':
1,000,087,608 instructions:u
0.285943713 seconds time elapsed
Using the second version of this program:
#include <stdlib.h>
#include <stdio.h>
#include <time.h>
main ()
{
int i;
fork();
fork();
for (i = 0; i < 100000000; i++) {
asm("nop");
asm("nop");
asm("nop");
asm("nop");
asm("nop");
asm("nop");
asm("nop");
}
wait(NULL);
wait(NULL);
wait(NULL);
wait(NULL);
}
This one spawn child processes so you have
perf -> child -> gchild -> ggchild
|---> gchild
So it has the grandparent characteristics you are getting at.
$ ./myperfstat ./1bi4
Total instructions: 4000094342
$ perf stat -e instructions:u ./1bi4
Performance counter stats for './1bi4':
4,000,094,649 instructions:u
0.259537842 seconds time elapsed
Still seems to match up.
This is on:
$ uname -r
3.18.7-200.fc21.x86_64
Looking more closely at the behavior of perf, I think the essential
difference is that perf uses the *child* PID as argument to
perf_event_open. When I tried changing my program to do the same, it
seems to have fixed the problem. That is to say, where I used
something like
And I did not modify the mystat.c program to do this:
fd = perf_event_open(..., getpid(), ...);
child = fork();
if (child == 0) {
execvp(...);
}
I needed to instead use
pipe(pipefd);
child = fork();
if (child == 0) {
read(pipefd[0], &x, 1);
execvp(...);
}
fd = perf_event_open(..., child, ...);
write(pipefd[1], &x, 1);
--
To unsubscribe from this list: send the line "unsubscribe linux-perf-users" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
