Changes for version 3.23

• Rename Rmpfr_get_decimal64, Rmpfr_set_decimal64, Rmpfr_get_float128 & Rmpfr_set_float128 to Rmpfr_get_DECIMAL64, Rmpfr_set_DECIMAL64, Rmpfr_get_FLOAT128 & Rmpfr_set_FLOAT128
• Provide Rmpfr_get_float128 & Rmpfr_set_float128 which pass the __float128 value as an NV
• Ensure that when casting an IV/NV to an mpfr_t, the mpfr_t has enough precision to accurately represent the value.
• No longer define USE_64_BIT_INT if ivtype is 'long'. (Thus had to tweak some test files.)
• Rename USE_64_BIT_INT #define to MATH_MPFR_NEED_LONG_LONG_INT
• Define IVSIZE_BITS (in XS file) if MATH_MPFR_NEED_LONG_LONG_INT is defined - for use in initialization of mpfr_t in some of the overload functions.
• Add FLOAT128_DIG