[Bug target/118843] The predefined macro `__loongarch_frecipe` is still defined when using `-mfpu=none`

[chenglulu at loongson dot cn via Gcc-bugs]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=118843

--- Comment #3 from chenglulu <chenglulu at loongson dot cn> ---
(In reply to Xi Ruoyao from comment #2)
> We have
> 
>   if (TARGET_HARD_FLOAT && ISA_HAS_FRECIPE)
>     builtin_define ("__loongarch_frecipe");
> 
> where the logic seems correct.  But __loongarch_frecipe is also in
> la_evo_macro_name and it can get defined by:
> 
>   for (int i = 0; i < N_EVO_FEATURES; i++)
>     if (la_target.isa.evolution & la_evo_feature_masks[i])
>       {   
>         builtin_define (la_evo_macro_name[i]);
> 
> ... ...
> 
> Maybe we can just do
> 
> if (!TARGET_HARD_FLOAT)
>   cpp_undef ("__loongarch_frecipe");
> 
> after the loop above (and remove the useless if block on the top).

I tried to make some changes, and the test went smoothly without any issues.
  for (int i = 0; i < N_EVO_FEATURES; i++)
    {
      builtin_undef (la_evo_macro_name[i]);

      if (la_target.isa.evolution & la_evo_feature_masks[i]
          && (la_evo_feature_masks[i] ^ OPTION_MASK_ISA_FRECIPE
              || TARGET_HARD_FLOAT))
        {
          builtin_define (la_evo_macro_name[i]);

          int major = la_evo_version_major[i],
          minor = la_evo_version_minor[i];

          max_v_major = major > max_v_major ? major : max_v_major;
          max_v_minor = major == max_v_major
            ? (minor > max_v_minor ? minor : max_v_minor) : max_v_minor;
        }
    }