Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
MIME-Version: 1.0
From:   Naruto Nguyen <narutonguyen2018@gmail.com>
Date:   Mon, 15 May 2023 19:59:37 +1000
Message-ID: <CANpxKHEb2ZWT-bBkZa0+XKS49d9iUEcXztQd+Wh+VC9Vh2=w3w@mail.gmail.com>
Subject: How to debug high sysload in 2,3 seconds
To:     kernel derkeiler <linux-kernel@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

Hi everyone,

My system during idle sometimes has high sysload in 2,3 seconds
(around 70% sysload) and then come back to normal after that. I see it
happen in some conditions and try to use top commands to debug and see
one of processes in my system consume CPU.

I would like to investigate more what code causes CPU high. However
the spike only happens 2,3 seconds so I have no ideas to continue
debugging. I thought about perf tool, but I am not sure how to profile
with perf to identify exactly what happens in 2,3 seconds during high
CPU. Because I can reproduce the issue to see it, but only during 3 or
4 minutes, I can see CPU increase in 2,3 seconds. So I worry that
using perf to profile in 3,4 minutes cannot identify the code causing
CPU increase in a short time.

Could you help to advise how I can debug in this case?

Thanks,
Brs,
Naruto