fabs(3)

💖 fabs — bestie's guide to absolute value 💖

okay so. picture this: you've got a floating-point number and it's being dramatic about its sign, and you just want the magnitude, the raw unfiltered distance-from-zero energy. that's what fabs, fabsf, and fabsl are for. they live in the math library, libm, so you link with -lm and you're good. no cap.

---

✨ SYNOPSIS ✨

#include <math.h>

double fabs(double x);
float fabsf(float x);
long double fabsl(long double x);

🎀 Feature Test Macro Requirements for glibc

(if you wanna know how this whole system works, go read feature_test_macros(7), it's giving lore)

fabsf(), fabsl():
    _ISOC99_SOURCE || _POSIX_C_SOURCE >= 200112L
        || /* Since glibc 2.19: */ _DEFAULT_SOURCE
        || /* glibc <= 2.19: */ _BSD_SOURCE || _SVID_SOURCE

so basically if you're on modern glibc it just works, and if you're on old glibc you need one of those legacy macros. either way you're covered 💅

---

🌸 DESCRIPTION 🌸

these three functions hand you back the absolute value of the floating-point number x. that's it. that's the whole job. they take whatever sign drama x is going through and just… let it go. ✨

---

👑 RETURN VALUE 👑

they return the absolute value of x. iconic. but here's the fine print, because edge cases are always the main character:

• if x is a NaN, a NaN is returned. (NaN in, NaN out, the cycle continues 🌀)
• if x is -0, +0 is returned. yes bestie, negative zero is real and this function does NOT respect it 😭
• if x is negative infinity or positive infinity, positive infinity is returned. infinity stays winning fr fr.

---

💀 ERRORS 💀

No errors occur.

literally nothing can go wrong here. it's the safest function in the whole library. zero drama, zero errno, just vibes 🌈

---

🦋 ATTRIBUTES 🦋

okay so if you wanna know what all these fancy terms even mean, attributes(7) has you covered.

• fabs(), fabsf(), fabsl() — Thread safety: MT-Safe 💖

translation: you can call these from multiple threads and they will NOT fight each other. team players. 🫶

---

⭐ STANDARDS ⭐

C11, POSIX.1-2008.

certified, standardized, official. periodt. 👑

---

🧚 HISTORY 🧚

• C99, POSIX.1-2001.
• The variant returning double also conforms to SVr4, 4.3BSD, C89.

so the double one has been doing this since forever. it was serving absolute value before it was cool. a true veteran 💜

---

🍓 SEE ALSO 🍓

• abs(3)
• cabs(3)
• ceil(3)
• floor(3)
• labs(3)
• rint(3)

(the whole math-function friend group, go say hi 🌸)

---

anyway that's the whole page bestie, no skips, no skims, every single detail preserved because that's the assignment 💅

this came straight out of Linux man-pages 6.20, dated 2026-02-08. slay, go take some absolute values. 💖✨