frexp(3)

✨ frexp(3) ✨ — okay bestie, we're splitting floats today 💖

Hi hi! So this page is all about frexp, and it lives in section 3 of the manual (that's the "library functions" girlies 💅). Buckle up, it's actually cute.

🌸 NAME 🌸

frexp, frexpf, frexpl — these three iconic functions convert a floating-point number into its fractional and integral components. It's giving "take this number apart, respectfully" 🎀

📚 LIBRARY 📚

Math library, aka libm, and you link it with -lm. No cap, you do have to link it.

🦋 SYNOPSIS 🦋

#include <math.h>

double frexp(double x, int *e);
float frexpf(float x, int *e);
long double frexpl(long double x, int *e);

💜 Feature Test Macro Requirements for glibc (see feature_test_macros(7)):

frexpf(), frexpl():

_ISOC99_SOURCE || _POSIX_C_SOURCE >= 200112L
    || /* Since glibc 2.19: */ _DEFAULT_SOURCE
    || /* glibc <= 2.19: */ _BSD_SOURCE || _SVID_SOURCE

(yes, the glibc ≤ 2.19 girlies had to do it the old way, we don't talk about it 🫶)

---

🌈 DESCRIPTION 🌈

So these functions take your number x and split it into two things: a normalized fraction, and an exponent. The exponent goes and lives inside e, which you passed as a pointer. That's it, that's the whole vibe. ✨

Like, x becomes <fraction> × 2^<exponent>, and the fraction part is the one frexp hands back to you. Obsessed with how tidy that is.

💖 RETURN VALUE 💖

These functions return the normalized fraction, and this part is important bestie:

• If your argument x is not zero, the normalized fraction is x times a power of two, and its absolute value is always in the range 1/2 (inclusive) to 1 (exclusive). That's [0.5,1). Half-open intervals, main character energy. 👑
• If x is zero, the normalized fraction is zero, and zero is stored in e. Zero in, zero out, zero told.
• If x is a NaN, a NaN is returned, and the value of *e is unspecified. 🫠 Don't go reading e after that, it's giving chaos.
• If x is positive infinity (negative infinity), then positive infinity (negative infinity) is returned, and the value of *e is again unspecified. Same warning, same vibe check. ⚠️

🍓 ERRORS 🍓

No errors occur. Literally none. She's unbothered, she's thriving, she simply does not fail. 💅

🎀 ATTRIBUTES 🎀

For an explanation of the terms used in this section, see attributes(7).

• frexp(), frexpf(), frexpl() — Thread safety — MT-Safe 🔥

So yes, she's thread-safe. Multi-threaded besties, you may proceed. ✨

⭐ STANDARDS ⭐

C11, POSIX.1-2008. Certified, stamped, official. 💖

🌸 HISTORY 🌸

C99, POSIX.1-2001.

And the variant that returns double is also conformant to SVr4, 4.3BSD, and C89. She's been around, she's got range, she's an elder icon. 🦋

💜 EXAMPLES 💜

The program below produces results such as the following:

$ ./a.out 2560
frexp(2560, &e) = 0.625: 0.625 * 2^12 = 2560
$ ./a.out -4
frexp(-4, &e) = -0.5: -0.5 * 2^3 = -4

Look at that. 0.625 is in [0.5,1), -0.5 is in [0.5,1) by absolute value. She said range-consistent and she MEANT it. 😭💖

🧚 Program source 🧚

#include <float.h>
#include <math.h>
#include <stdio.h>
#include <stdlib.h>

int
main(int argc, char *argv[])
{
    double x, r;
    int e;

    x = strtod(argv[1], NULL);
    r = frexp(x, &e);

    printf("frexp(%g, &e) = %g: %g * %d^%d = %g\n", x, r, r, 2, e, x);
    exit(EXIT_SUCCESS);
}

Not me noticing <float.h> is included and never used… she's just there for moral support, we love her anyway 🙈

👑 SEE ALSO 👑

• ldexp(3) — the reverse girlie, putting it back together
• modf(3)

---

💀 Quick recap so you don't fumble: pass a pointer to an int for e, get back a fraction in [0.5,1) (by absolute value), and if x is zero, NaN, or an infinity, read the RETURN VALUE section again before you touch *e. Fr fr.

And that's the whole page, bestie! This one came from the Linux man-pages, version 6.20, dated 2026-02-08 — so you know it's fresh. Go split some floats. You're gonna slay. 🌈✨🫶