netdev - Re: [PATCH bpf-next 8/8] selftests/bpf: Add tests for attaching bpf

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <87lflc2no9.fsf@cloudflare.com>
Date:   Thu, 28 May 2020 15:29:10 +0200
From:   Jakub Sitnicki <jakub@...udflare.com>
To:     Andrii Nakryiko <andrii.nakryiko@...il.com>
Cc:     bpf <bpf@...r.kernel.org>, Networking <netdev@...r.kernel.org>,
        kernel-team@...udflare.com
Subject: Re: [PATCH bpf-next 8/8] selftests/bpf: Add tests for attaching bpf_link to netns

On Thu, May 28, 2020 at 08:08 AM CEST, Andrii Nakryiko wrote:
> On Wed, May 27, 2020 at 12:16 PM Jakub Sitnicki <jakub@...udflare.com> wrote:
>>
>> Extend the existing test case for flow dissector attaching to cover:
>>
>>  - link creation,
>>  - link updates,
>>  - link info querying,
>>  - mixing links with direct prog attachment.
>>
>> Signed-off-by: Jakub Sitnicki <jakub@...udflare.com>
>> ---
>
> You are not using bpf_program__attach_netns() at all. Would be nice to
> actually use higher-level API here...

That's true. I didn't exercise the high-level API. I can cover that.

>
> Also... what's up with people using CHECK_FAIL + perror instead of
> CHECK? Is CHECK being avoided for some reason or people are just not
> aware of it (which is strange, because CHECK was there before
> CHECK_FAIL)?

I can only speak for myself. Funnily enough I think I've switched from
CHECK to CHECK_FAIL when I touched on BPF flow dissector last time [0].

CHECK needs and "external" duration variable to be in scope, and so it
was suggested to me that if I'm not measuring run-time with
bpf_prog_test_run, CHECK_FAIL might be a better choice.

CHECK is also perhaps too verbose because it emits a log message on
success (to report duration, I assume).

You have a better overview of all the tests than me, but if I had the
cycles I'd see if renaming CHECK to something more specific, for those
test that actually track prog run time, can work.

-jkbs

[0] https://lore.kernel.org/bpf/87imov1y5m.fsf@cloudflare.com/

>
>>  .../bpf/prog_tests/flow_dissector_reattach.c  | 500 +++++++++++++++++-
>>  1 file changed, 471 insertions(+), 29 deletions(-)
>>
>
> [...]