Re: [PATCH v3 01/38] Kconfig: introduce HAS_IOPORT option and select it as necessary

[Niklas Schnelle <schnelle@xxxxxxxxxxxxx>]

On Tue, 2023-03-14 at 13:37 +0100, Johannes Berg wrote:
> On Tue, 2023-03-14 at 13:11 +0100, Niklas Schnelle wrote:
> > --- a/arch/um/Kconfig
> > +++ b/arch/um/Kconfig
> > @@ -56,6 +56,7 @@ config NO_IOPORT_MAP
> >  
> >  config ISA
> >  	bool
> > +	depends on HAS_IOPORT
> > 
> 
> config ISA here is already unselectable, and nothing ever does "select
> ISA" (only in some other architectures), so is there much point in this?
> 
> I'm not even sure why this exists at all.

You're right there's not much point and I dropped this for v4. I agree
that probably the whole "config ISA" could be removed if it's always
false anyway but that seems out of scope for this patch.

> 
> But anyway, adding a dependency to a always-false symbol doesn't make it
> less always-false :-)
> 
> Acked-by: Johannes Berg <johannes@xxxxxxxxxxxxxxxx> # for ARCH=um

Thanks

> 
> 
> Certainly will be nice to get rid of this cruft for architectures that
> don't have it.
> 
> johannes

Yes, also, for s390 the broken NULL + port number access in the generic
inb()/outb() currently causes the only remaining clang warning on
defconfig builds.