linux-kernel - Re: [PATCH bpf-next v2 1/3] selftests/bpf: do not disable /dev/null device access in cgroup dev test

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] [day] [month] [year] [list]

Message-ID: <df287de3-8b06-42dd-8353-fae5cffae6a2@bootlin.com>
Date: Tue, 30 Jul 2024 10:42:04 +0200
From: Alexis Lothoré <alexis.lothore@...tlin.com>
To: Alan Maguire <alan.maguire@...cle.com>,
 Alexei Starovoitov <ast@...nel.org>, Daniel Borkmann <daniel@...earbox.net>,
 Andrii Nakryiko <andrii@...nel.org>, Martin KaFai Lau
 <martin.lau@...ux.dev>, Eduard Zingerman <eddyz87@...il.com>,
 Song Liu <song@...nel.org>, Yonghong Song <yonghong.song@...ux.dev>,
 John Fastabend <john.fastabend@...il.com>, KP Singh <kpsingh@...nel.org>,
 Stanislav Fomichev <sdf@...ichev.me>, Hao Luo <haoluo@...gle.com>,
 Jiri Olsa <jolsa@...nel.org>, Mykola Lysenko <mykolal@...com>,
 Shuah Khan <shuah@...nel.org>
Cc: ebpf@...uxfoundation.org, Thomas Petazzoni
 <thomas.petazzoni@...tlin.com>, bpf@...r.kernel.org,
 linux-kselftest@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH bpf-next v2 1/3] selftests/bpf: do not disable /dev/null
 device access in cgroup dev test

On 7/30/24 10:16, Alan Maguire wrote:
> On 29/07/2024 18:30, Alexis Lothoré wrote:
>> Hello Alan,

[...]

>>> Not a big deal, but I found it a bit confusing that this file was
>>> modified then deleted in patch 2. Would it work having patch 1 stop
>>> building the standalone test/remove it and .gitignore entry, patch 2
>>> updating progs/dev_cgroup.c to allow /dev/zero, /dev/urandom access,
>>> patch 3 add cgroup_dev.c test support, and patch 4 add the device type
>>> subtest? Or are there issues with doing things that way? Thanks!
>>
>> I've done this to make sure that at any point in the git history, there is one
>> working test for the targeted feature, either the old or the new one. I've done
>> it this way because the old test also helped me validate the new one while
>> developing it, but also because if at some point there is a (major) issue with
>> the new test, reverting only the relevant commit brings back the old test while
>> disabling the new one.
>>
>> But maybe this concern is not worth the trouble (especially since the old tests
>> are not run automatically) ? If that's indeed the case, I can do it the way you
>> are suggesting :)
>>
> 
> If no-one complains, it seems fine to me to stick with the way you've
> constructed the series the next respin. Thanks!

ACK, thanks, I'll keep it that way then.

For the record, I am accumulating a few other converted tests that I will send
soon, and those follow the same logic (keeping one working test at any point of
time, and pushing it to the point where I start by fixing broken tests before
converting those), so if anyone has an opinion in favor of this or rather in
favor of Alan's suggestion, do not hesitate to share it, so I can adjust before
sending.

Thanks,

> 
>> Thanks,
>>
>> Alexis
>>
> 

-- 
Alexis Lothoré, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com