nexttoward(), nexttowardf(), nexttowardl()

Updated: April 19, 2023

Compute the next representable floating-point number in a given direction

Synopsis:

#include <math.h>

double nexttoward ( double x,
                    long double y);

float nexttowardf ( float x,
                    long double y );

long double nexttowardl ( long double x,
                          long double y);

Arguments:

x
The number for which you want the next number in the specified direction.
y
A number that specifies the direction you want to go; see below.

Library:

libm
The general-purpose math library.
libm-sve
(QNX Neutrino 7.1 or later) A library that optimizes the code for ARMv8.2 chips that have Scalable Vector Extension hardware.

Your system requirements will determine how you should work with these libraries:

Note: Compile your program with the -fno-builtin option to prevent the compiler from using a built-in version of the function.

Description:

The nexttoward(), nexttowardf(), and nexttowardl() functions compute the next representable double-precision floating-point value following x in the direction of y. These functions are equivalent to nextafter(), nextafterf(), and nextafterl(), respectively, except that:

To check for error situations, use feclearexcept() and fetestexcept(). For example:

Returns:

The next machine floating-point number of x in the direction towards y, converted to the type that the function returns.

If These functions return: Errors:
x equals y y
x is finite, and the correct value would overflow Inf FE_OVERFLOW
x or y is NaN NaN
x != y, and the correct value is subnormal, zero, or underflows The correct value (if representable), or 0.0 FE_UNDERFLOW

These functions raise FE_INEXACT if the FPU reports that the result can't be exactly represented as a floating-point number.

Examples:

#include <stdio.h>
#include <inttypes.h>
#include <math.h>
#include <fenv.h>
#include <stdlib.h>

void dump_to_hex(double d) {
        printf("0x%08x %08x \n",
           (uint32_t)(*((uint64_t*)&d) >> 32),
           (uint32_t)(*((uint64_t*)&d)));
}

int main(int argc, char** argv) 
{
    double a, b, c;
    int except_flags;

    a = 0;

    feclearexcept(FE_ALL_EXCEPT);
    b = nexttoward(a, -1);
    except_flags = fetestexcept(FE_ALL_EXCEPT);
    if(except_flags) {
        /* An error occurred; handle it appropriately. */
    }

    feclearexcept(FE_ALL_EXCEPT);
    c = nexttoward(a, 1);
    except_flags = fetestexcept(FE_ALL_EXCEPT);
    if(except_flags) {
        /* An error occurred; handle it appropriately. */
    }

    printf("Next possible value before %f is %f \n", a, b);
    printf("-->"); dump_to_hex(a);
    printf("-->"); dump_to_hex(b);
    printf("Next possible value after %f is %f \n", a, c);
    printf("-->"); dump_to_hex(a);
    printf("-->"); dump_to_hex(c);

    return EXIT_SUCCESS;
}

produces the output:

Next possible value before 0.000000 is 0.000000
-->0x00000000 00000000
-->0x80000000 00000001
Next possible value after 0.000000 is 0.000000
-->0x00000000 00000000

Classification:

C11, POSIX 1003.1

Safety:  
Cancellation point No
Interrupt handler Yes
Signal handler Yes
Thread Yes