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;
Subject: Re: [RFC PATCH 0/1] sample: bpf: introduce irqlat
To:     Song Chen <chensong_2000@189.cn>
References: <1649927240-18991-1-git-send-email-chensong_2000@189.cn>
 <2e6ee265-903c-2b5c-aefd-ec24f930c999@iogearbox.net>
 <ac371d36-2624-cdd8-0c15-62ccf53bed81@189.cn>
Cc:     ast@kernel.org, andrii@kernel.org, kafai@fb.com,
        songliubraving@fb.com, yhs@fb.com, john.fastabend@gmail.com,
        kpsingh@kernel.org, netdev@vger.kernel.org, bpf@vger.kernel.org,
        linux-kernel@vger.kernel.org, brendan.d.gregg@gmail.com
From:   Daniel Borkmann <daniel@iogearbox.net>
Message-ID: <f4267d24-28ca-bd99-100e-6fa4ee84cc50@iogearbox.net>
Date:   Thu, 14 Apr 2022 17:32:44 +0200
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101
 Thunderbird/60.7.2
MIME-Version: 1.0
In-Reply-To: <ac371d36-2624-cdd8-0c15-62ccf53bed81@189.cn>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: en-US
Content-Transfer-Encoding: 8bit
Precedence: bulk

Hi Song,

On 4/14/22 1:25 PM, Song Chen wrote:
> hi Daniel,
> 
> Thanks for liking the idea.
> 
> My target is embedded devices, that's why i get started from ebpf C.bcc and bpftrace is a good idea, but i prefer taking one thing at a time, what's more, i'm not familiar with python, it might take longer.
> 
> Once C code is accepted, i will move myself to bcc and bpftrace. Is it ok for you?

The libbpf-based tools from the mentioned link in BCC are all C, not Python. Also bpftrace
has guidelines for building it more portably that would be suitable for embedded devices [2].
I'd presume these should still match your requirements?

Right now samples/bpf/ is a bit of a dumping ground of random things, some BPF samples better
maintained than others, but generally samples/bpf/ is a bit of a mess. BPF has a huge ecosystem
outside of kernel in its various areas it covers, so it has outgrown the few samples in there
long ago, and you'll find many resources on how to get started in the wild.

Adding this as a samples/bpf/ will have little value to others, since people may not be aware
of them, and if they are they need to manually build/ship it, etc. If you upstream and can improve
the tools in bpftrace/bcc as pointed out, then a lot more people will be able to consume them
and benefit from it, and you get the shipping via distros for free.

   [2] https://github.com/iovisor/bpftrace/blob/master/docs/embedded_builds.md

Thanks again,
Daniel

> BR
> 
> Song
> 
> 
> 在 2022/4/14 17:47, Daniel Borkmann 写道:
>> On 4/14/22 11:07 AM, Song Chen wrote:
>>> I'm planning to implement a couple of ebpf tools for preempt rt,
>>> including irq latency, preempt latency and so on, how does it sound
>>> to you?
>>
>> Sounds great, thanks! Please add these tools for upstream inclusion either to bpftrace [0] or
>> bcc [1], then a wider range of users would be able to benefit from them as well as they are
>> also shipped as distro packages and generally more widely used compared to kernel samples.
>>
>> Thanks Song!
>>
>>    [0] https://github.com/iovisor/bpftrace/tree/master/tools
>>    [1] https://github.com/iovisor/bcc/tree/master/libbpf-tools
>>
>>> Song Chen (1):
>>>    sample: bpf: introduce irqlat
>>>
>>>   samples/bpf/.gitignore    |   1 +
>>>   samples/bpf/Makefile      |   5 ++
>>>   samples/bpf/irqlat_kern.c |  81 ++++++++++++++++++++++++++++++
>>>   samples/bpf/irqlat_user.c | 100 ++++++++++++++++++++++++++++++++++++++
>>>   4 files changed, 187 insertions(+)
>>>   create mode 100644 samples/bpf/irqlat_kern.c
>>>   create mode 100644 samples/bpf/irqlat_user.c
>>>
>>
>>