linux-kernel - Re: [PATCH v3] selftests/run_kselftest.sh: exit with error if tests fail

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: <DDWIW4V2X9EX.1VQ2QCG6AU78R@google.com>
Date: Fri, 31 Oct 2025 12:57:06 +0000
From: Brendan Jackman <jackmanb@...gle.com>
To: "Thomas Weißschuh" <thomas.weissschuh@...utronix.de>, 
	Brendan Jackman <jackmanb@...gle.com>
Cc: Shuah Khan <shuah@...nel.org>, <linux-kselftest@...r.kernel.org>, 
	<linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v3] selftests/run_kselftest.sh: exit with error if tests fail

On Fri Oct 31, 2025 at 12:48 PM UTC, Thomas Weißschuh wrote:
> Hi Brendan,
>
> Oct 14, 2025 16:45:32 Brendan Jackman <jackmanb@...gle.com>:
>
> (...)
>
>> In case any user depends on the current behaviour, such as running this
>> from a script with `set -e` and parsing the result for failures
>> afterwards, add a flag they can set to get the old behaviour, namely
>> --no-error-on-fail.
>
> IMO this new flag is also unnecessary.
> The user can just do "|| true" when needed.
>

`|| true` is not the same thing, if you do that then you completely hide
all failures of the script. With --no-error-on-fail you just skip the
specific case of tests failing.

I did say somewhere in a previous thread that this distinction (test
failure vs test harness failure) is always gonna be a bit sketchy for
this script since it's running on the kernel under test. But that
doesn't mean we should give up on it completely completely.