Re: [RESEND PATCH] s390: fix dead defaults for S390_MODULES_SANITY_TEST and S390_UNWIND_SELFTEST

[Alexander Gordeev]

On Wed, Apr 29, 2026 at 10:39:14PM +0100, Julian Braha wrote:

Hi Julian,

> On Wed, 29 Apr 2026, Alexander Gordeev <agordeev@xxxxxxxxxxxxx> wrote:
> > This tool is still in development, right?
> 
> You can find the latest release (v0.7) outside the tree here:
> https://github.com/julianbraha/kconfirm
> 
> And there's also a recent RFC to move it into the tree here:
> https://lore.kernel.org/all/20260427174429.779474-1-julianbraha@xxxxxxxxx/
> 
> > Anyway, I do not quite get what do you mean with the dead code.
> > Could you please elaborate?
> 
> Sure! In kconfig, default statements are evaluated in the order
> they appear (top --> bottom). This means that if you have an
> unconditional default statement at the top, e.g. 'default n', then all
> of the default statements that follow it will never be evaluated - in
> other words, they are dead code.
> 
> In the case of this code in particular, the 'def_tristate n' at the top
> does two things:
> 1. declares the type of the config option as a tristate,
> 2. sets the default value to 'n' unconditionally.
> 
> So, the 'default KUNIT_ALL_TESTS' statement that follows is dead code.
> 
> See also this sentence from the "Menu attributes" section of the kconfig
> docs:
> "If multiple default values are visible, only the first defined one is
> active."
> Source:
> https://docs.kernel.org/kbuild/kconfig-language.html

Thanks a lot for the clarificaiton!

In the end of the day I see no impact on the end result (aka the
generated config) - please correct me if I am wrong.

I think once the tool is officially included you could refer to it
in the commit message, but please do not add the described commits
as Fixes - these are no real issues.

> - Julian Braha

Thanks!