Re: [Boost-users] generalize boost macros

Subject: Re: [Boost-users] generalize boost macros
From: Gero Peterhoff (gero.peterhoff_at_[hidden])
Date: 2019-02-02 11:03:02

Am 01.02.19 um 22:08 schrieb John Maddock via Boost-users:
> multiprecision/float128_t:
>> Â  there are still many features missing
> Specifically?
eg exp2, log2, logb, modf, round, scalbn, copysign, signbit, isnormal
are not available in std:: for float128_t, see cstdfloat_cmath.hpp. I'd
like a full float128_t feature set, see
https://en.cppreference.com/w/cpp/numeric/math and
https://en.cppreference.com/w/cpp/numeric/special_math.
>>   most values ​​in numeric_limits<float128_t> are not constexpr
> 
> I know :(
> 
> There has been some work towards making the library constexpr throughout
> where possible: but I came to the conclusion that it wasn't possible
> without is_constant_evaluated() from C++20. numeric_limits<float128>
> might be possible though.
> 
>>
>> math::constants:
>> Â  i miss log2(2/e/pi/10) etc
> 
> Not sure I follow, do you really mean log2(2) ??
This is not constexpr for float128_t and the compiler generates a
function call. :-(
> 
> Best, John.

Thank you very much
Gero