Checking for IEEE 754 Floating Point Standard in C

Determining whether a C compiler adheres to the IEEE 754 floating point standard is typically accomplished through a compiler define. However, the technique used for C may not apply directly to C .

C -Specific Approach

Fortunately, C offers a straightforward method to accomplish this check using the numeric_limits class:

std::numeric_limits::is_iec559;

This expression evaluates to true if IEEE 754 is utilized by the compiler and false otherwise. A similar check can be performed for the float type:

std::numeric_limits::is_iec559;

Alternative Method

As an alternative to the numeric_limits approach, you can adapt the second part of Adam's response for C :

#include 

int main() {
  // Check for IEEE 754 compliance by checking for a finite number of float exponents.
  if (std::numeric_limits::max_exponent == std::numeric_limits::max()) {
    // Compiler uses IEEE 754.
  } else {
    // Compiler does not use IEEE 754.
  }
}

This approach relies on the fact that IEEE 754 defines a specific range of exponents for floating point numbers. If the compiler's maximum exponent matches the maximum size of an integer, it is likely adhering to IEEE 754.