[Bug c/65881] no documentation of __uint128_t

[glisse at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65881

--- Comment #1 from Marc Glisse <glisse at gcc dot gnu.org> ---
(In reply to Shawn Landden from comment #0)
> gcc on x86_64 recognizes __uint128_t, but this is not documented along with
> __int128, also there is no warning when a target does not support
> __uint128_t, like there is with __int128:
> 
> triangle64.c:32:1: error: ‘__int128’ is not supported for this target

e.c:2:3: error: unknown type name ‘__uint128_t’
   __uint128_t i=42;
   ^

> A unsigned 128-bit type is needed because it has defined integer overflow.

https://gcc.gnu.org/onlinedocs/gcc/_005f_005fint128.html
That type is called "unsigned __int128". The other aliases are only for
compatibility or legacy.