[PATCH 02/10] compiler.h: add is_const() as a replacement of __is_constexpr()

[David Laight]

From: Linus Torvalds
> Sent: 06 December 2024 18:53
> 
> On Fri, 6 Dec 2024 at 10:31, Vincent Mailhol <vincent.mailhol at gmail.com> wrote:
> >
> > > causes issues when 'x' is not an integer expression (think
> > > "is_const(NULL)" or "is_const(1 == 2)".
> >
> > But 1 == 2 already has an integer type as proven by:
> 
> Yeah, I was confused about exactly what triggers that odd
> '-Wint-in-bool-context'.
> 
> It's not about some actual bool type, it's literally a random
> collection of integer operations used with logical ops.
> 
> So it's things like "!(var<<2)" that generate that warning, because
> some compiler person at some point went "maybe that left shift should
> have been just a comparison instead '<'".
> 
> But it turns out that "(var <<2)?0:0" _also_ triggers that warning.
> 
> End result: I have *no* idea how to shut that crazy warning up for
> this case, if we want to have some generic macro that says "is this
> constant". Because it damn well is perfectly sane to ask "is (a << 3)
> a constant expression".

I'm missing the compiler version and options to generate the error.
Does a '+ 0' help?  "(var << 2) + 0 ? 0 : 0"

I realised the:
#define const_NULL(x) _Generic(0 ? (x) : (char *)0, char *: 1, void *: 0)
#define const_true(x) const_NULL((x) ? NULL : (void *)1L))
#define const_expr(x) const_NULL((x) ? NULL : NULL))
I send this morning.
Needs 's/char/struct kjkjkjkjui/' applied.

	David

-
Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK
Registration No: 1397386 (Wales)