Configuration Macros

User Configurable Macros

The following configuration macros are available:

  • BOOST_DECIMAL_DISABLE_CASSERT: Disables the use of <cassert> and all run-time assertions.

  • BOOST_DECIMAL_DISABLE_IOSTREAM: Disables the use of I/O streaming and removes all associated headers (e.g. <iostream>, <iosfwd>, <cwchar>, etc.)

  • BOOST_DECIMAL_DISABLE_CLIB: Defines both of the above macros. In testing this reduces ROM usage by ~50% and can be useful if running on embedded platforms.

  • BOOST_DECIMAL_DISABLE_EXCEPTIONS: This allows the user to disable any calls to throw. This macro will also be set automatically in the presence of -fno-exceptions or various /EH flags on MSVC.

  • BOOST_DECIMAL_ALLOW_IMPLICIT_INTEGER_CONVERSIONS: Allows any integer type to be implicitly converted to any decimal type.

  • BOOST_DECIMAL_ALLOW_IMPLICIT_FLOAT_CONVERSIONS: Allows a binary floating-point type (e.g. double) to be implicitly converted to a decimal floating point type. This option is not recommended, but can be useful if you want to use specific functionality from the standard library with internal conversions such as:

#define BOOST_DECIMAL_ALLOW_IMPLICIT_CONVERSIONS
#include <boost/decimal.hpp>
#include <complex>

constexpr decimal64_t half {5, -1};
std::complex<decimal64_t> test_val {half, half};
const auto res = std::acos(test_val);

  • BOOST_DECIMAL_ALLOW_IMPLICIT_CONVERSIONS: Defines both BOOST_DECIMAL_ALLOW_IMPLICIT_INTEGER_CONVERSIONS and BOOST_DECIMAL_ALLOW_IMPLICIT_FLOAT_CONVERSIONS

  • BOOST_DECIMAL_FAST_MATH performs optimizations similar to that of the -ffast-math compiler flag such as removing all checks for non-finite values. This flag increases the performance of the basis operations (e.g. add, sub, mul, div, and comparisons) by up to 20%. Again, it must be defined before inclusion of decimal headers like so:

#define BOOST_DECIMAL_FAST_MATH
#include <boost/decimal.hpp>

  • BOOST_DECIMAL_DEC_EVAL_METHOD: See <cfloat> section for explanation

  • BOOST_DECIMAL_ENABLE_CUDA: Defining this macro allows both the types and selected functions to be run on host and device when compiling with NVCC (i.e. when __CUDACC__ is defined). Allowed functions have BOOST_DECIMAL_HOST_DEVICE or BOOST_DECIMAL_CUDA_CONSTEXPR as part of their function signature. Defining this macro also automatically defines BOOST_DECIMAL_DISABLE_EXCEPTIONS and BOOST_DECIMAL_DISABLE_CASSERT, since neither exceptions nor <cassert> are available in device code.

Automatic Configuration Macros

  • BOOST_DECIMAL_CXX20_CONSTEXPR: This is defined to constexpr when compiling with C++20 or greater, otherwise it expands to nothing.

  • BOOST_DECIMAL_HOST_DEVICE: This is defined to __host__ __device__ when compiling with NVCC and BOOST_DECIMAL_ENABLE_CUDA is defined, and to nothing otherwise. Public functions, constructors, operators, and conversion operators annotated with this macro can be used in CUDA device code without modification.

  • BOOST_DECIMAL_CUDA_CONSTEXPR: This is defined to __host__ __device__ constexpr when compiling with NVCC and BOOST_DECIMAL_ENABLE_CUDA is defined, and to constexpr otherwise. The majority of the library’s functions are annotated with this macro.

  • BOOST_DECIMAL_HAS_STD_CHARCONV: This macro is defined if header <charconv> exists and the language standard used is >= C++17

    • We only need the structs and enums out of the header so we are not concerned with being overly restrictive about the feature test macros.

      • Known compilers that support this lighter requirement are: GCC >= 10, Clang >= 13, and MSVC >= 14.2

  • BOOST_DECIMAL_HAS_STD_STRING_VIEW: This macro is defined if header <string_view> exists and the langauge standard used is >= C++17

Edit this Page