Comment # 13 on bug 1191260 from Michal Suchanek

(In reply to Martin Wilck from comment #12)
> (In reply to Michal Suchanek from comment #11)
> 
> > > By definition, cert-script is a noop on systems that don't support UEFI.
> > 
> > We have secure boot on s390 and support for POWER is in progress (expected
> > in 15 SP5). Maybe the kernel makes this look like EFI to userspace but the
> > expectation is that we will have a different tool for POWER eventually.
> 
> Fine. Either s390x and Power will use mokutil, in which case the test will
> be correct. Or they will use some other tool, in which we'll need to extend
> the test. It's fine both ways, and actually easier to maintain than having
> the same test in multiple places in the code.

Which then evolves into

if !X && !Y 

exit 0

...

if X 

do something with X

fi

if Y

do something with Y

fi

so the initial test is redundant.

But there is also another problem: when looking at the code you not only need
to  take into account the code you are looking at but also any code that exits
early. That makes maintaining the code harder.

The code that was imported from kernel-source does exit early in some places
but adding to that does not make the code cleaner but rather more obfuscated.