Perf tool, used for profiling, crashes on heavy load

When I am using “perf record -F 99 -g -p – sleep 300” during average load, I am getting the file without any problem. But when the load is high, it is crashing. The only way to get an ok result is to reduce the frequency to around 30, which is terrible.

Ideally, the rate should be high to get better results(about 997 samples/sec). Is there any solution to this.

I think that the issue might be related to the Linux kernel space. If so, what would I have to do to solve this issue?

Would be useful to share the board, kernel version and crash logs.

The Linux version that i am using is “Linux version 3.10.0-957.21.3.el7.x86_64”. I am not getting any stack-trace when it crashes. Is there any way to generate the stack-trace as it is not coming by default. Also. should i add anything in the “perf record” statement to get the stack-trace.

I’m afraid you are probably asking on the wrong forums. It is still possible someone can offer you a clue but none of the 96Boards has an x86-64 processor in them so this isn’t the best place to connect with those that could help you.

1 Like

Below is the stack trace that i got :

[root@finesse25 ~]# perf record -F 9997 -g -p 23675 -v
The cycles event is not supported, trying to fall back to cpu-clock-ticks
Using CPUID GenuineIntel-6-3E-4
mmap size 528384B
couldn’t open /proc/14480/status

Note that I have used -v (perf record -F 9997 -g -p 23675 -v) as an extra argument to the one mentioned in the original question. (perf record -F 99 -g -p – sleep 300)