Re: [PATCH bpf-next 2/3] selftest/bpf: fmod_ret prog and implement test_overhead as part of bench

From: Andrii Nakryiko <andrii.nakryiko@gmail.com>
To: John Fastabend <john.fastabend@gmail.com>
Cc: Andrii Nakryiko <andriin@fb.com>, bpf <bpf@vger.kernel.org>,
	Networking <netdev@vger.kernel.org>,
	Alexei Starovoitov <ast@fb.com>,
	Daniel Borkmann <daniel@iogearbox.net>,
	Kernel Team <kernel-team@fb.com>
Subject: Re: [PATCH bpf-next 2/3] selftest/bpf: fmod_ret prog and implement test_overhead as part of bench
Date: Fri, 8 May 2020 11:01:44 -0700	[thread overview]
Message-ID: <CAEf4BzZ0ydoZa_Y0NvHgEqMM-QffKZW7jR4-Bxo5gr5A-Scqgg@mail.gmail.com> (raw)
In-Reply-To: <5eb5817b12a3b_2a992ad50b5cc5b4b7@john-XPS-13-9370.notmuch>

On Fri, May 8, 2020 at 8:57 AM John Fastabend <john.fastabend@gmail.com> wrote:
>
> Andrii Nakryiko wrote:
> > Add fmod_ret BPF program to existing test_overhead selftest. Also re-implement
> > user-space benchmarking part into benchmark runner to compare results.  Results
> > with ./bench are consistently somewhat lower than test_overhead's, but relative
> > performance of various types of BPF programs stay consisten (e.g., kretprobe is
> > noticeably slower).
> >
> > To test with ./bench, the following command was used:
> >
> > for i in base kprobe kretprobe rawtp fentry fexit fmodret; \
> > do \
> >     summary=$(sudo ./bench -w2 -d5 -a rename-$i | \
> >               tail -n1 | cut -d'(' -f1 | cut -d' ' -f3-) && \
> >     printf "%-10s: %s\n" $i "$summary"; \
> > done
>
> might be nice to have a script ./bench_tracing_overhead.sh when its in its
> own directory ./bench. Otherwise I'll have to look this up every single
> time I'm sure.

Yeah, I didn't want to pollute selftests/bpf directory, but with a
separate subdir it makes more sense. I'll put it there. For ringbuf
I'll have much longer scripts, so having them in file is a necessity.

>
> >
> > This gives the following numbers:
> >
> >   base      :    3.975 ± 0.065M/s
> >   kprobe    :    3.268 ± 0.095M/s
> >   kretprobe :    2.496 ± 0.040M/s
> >   rawtp     :    3.899 ± 0.078M/s
> >   fentry    :    3.836 ± 0.049M/s
> >   fexit     :    3.660 ± 0.082M/s
> >   fmodret   :    3.776 ± 0.033M/s
> >
> > While running test_overhead gives:
> >
> >   task_rename base        4457K events per sec
> >   task_rename kprobe      3849K events per sec
> >   task_rename kretprobe   2729K events per sec
> >   task_rename raw_tp      4506K events per sec
> >   task_rename fentry      4381K events per sec
> >   task_rename fexit       4349K events per sec
> >   task_rename fmod_ret    4130K events per sec
> >
> > Signed-off-by: Andrii Nakryiko <andriin@fb.com>
> > ---
>
> LGTM
>
> Acked-by: John Fastabend <john.fastabend@gmail.com>