[PATCH] float128: Add signbit alternative for old compilers

[Joseph Myers]

On Fri, 30 Jun 2017, Gabriel F. T. Gomes wrote:

> +int
> +__signbitf128 (_Float128 x)
> +{
> +  int64_t e;
> +
> +  GET_FLOAT128_MSW64 (e, x);
> +  return e < 0;

I think you should use __builtin_signbit for GCC >= 6, with a comment that 
once GCC >= 6 is required for building glibc, this file can go back to 
wrapping the float128 version.  OK with that change.

-- 
Joseph S. Myers
joseph@codesourcery.com