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

[Willy Tarreau]

On Fri, Dec 06, 2024 at 11:15:20AM -0800, Linus Torvalds wrote:
> On Fri, 6 Dec 2024 at 11:07, David Laight <David.Laight at aculab.com> wrote:
> >
> > I'm missing the compiler version and options to generate the error.
> 
> Just -Wall with most recent gcc versions seems to do it. At least I
> can repro it with gcc-14.2.1 and something silly like this:
> 
>   $ cat t.c
>   int fn(int a) { return (a<<2)?1:2; }
>   $ gcc -Wall -S t.c
>   t.c: In function 'fn':
>   t.c:1:26: warning: '<<' in boolean context, did you mean '<'?
> [-Wint-in-bool-context]
> 
> > Does a '+ 0' help?  "(var << 2) + 0 ? 0 : 0"
> 
> Yeah, that actually works.
> 
> And "+0" is nice in that it should work in any context.

I've already used "+0" to shut certain warnings, I don't really remember
which one, but also remember it was OK everywhere I needed.

Another trick I've been using to shut up the compiler is a cast via typeof
and an intermediary variable:

  #define shut_up(expr)                           \
          ({                                      \
                  typeof(expr) _expr_ = expr;     \
                  _expr_;                         \
                  })
  
  int fn1(int a)
  {
          return shut_up(a << 2) ? 1 : 2;
  }
  
  int fn2(int a)
  {
          return (a << 2) ? 1 : 2;
  }

  $ gcc -Wall -S t2.c
  t2.c: In function 'fn2':
  t2.c:15:19: warning: '<<' in boolean context, did you mean '<'? [-Wint-in-bool-context]
     15 |         return (a << 2) ? 1 : 2;
        |                ~~~^~~~~

See ? It only complains on fn2() but not fn1(). Similarly I found it
to be quite portable (at least I don't remember seeing it fail on me).
It produces the exact same code, except at -O0 where it does really
create a local variable.

Willy