finite(3)

💖 finite(3) — the BSD float-classification girlies 💖

okay bestie, pull up a chair, we're doing finite(3) from the Library Functions Manual today and honestly? it's giving "legacy but iconic". 🎀

---

✨ NAME ✨

finite, finitef, finitel, isinf, isinff, isinfl, isnan, isnanf, isnanl — the BSD floating-point classification functions. that's the whole squad. 🫶

---

📚 LIBRARY 📚

Math library, aka libm, aka you link it with -lm. don't forget her. she WILL ghost you if you forget her. 💅

-lm

---

🌸 SYNOPSIS 🌸

first things first, you gotta include the header:

#include <math.h>

then here's the whole friend group, all typed out:

int finite(double x);
int finitef(float x);
int finitel(long double x);

int isinf(double x);
int isinff(float x);
int isinfl(long double x);

int isnan(double x);
int isnanf(float x);
int isnanl(long double x);

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

okay so this is the part where glibc makes you prove you're allowed to use these. rude? yes. necessary? also yes. here's exactly what you need, no cap:

finite(), finitef(), finitel():
    /* glibc >= 2.19: */ _DEFAULT_SOURCE
        || /* glibc <= 2.19: */ _BSD_SOURCE || _SVID_SOURCE

isinf():
    _XOPEN_SOURCE >= 600 || _ISOC99_SOURCE
        || /* glibc >= 2.19: */ _DEFAULT_SOURCE
        || /* glibc <= 2.19: */ _BSD_SOURCE || _SVID_SOURCE

isinff(), isinfl():
    /* glibc >= 2.19: */ _DEFAULT_SOURCE
        || /* glibc <= 2.19: */ _BSD_SOURCE || _SVID_SOURCE

isnan():
    _XOPEN_SOURCE || _ISOC99_SOURCE
        || /* glibc >= 2.19: */ _DEFAULT_SOURCE
        || /* glibc <= 2.19: */ _BSD_SOURCE || _SVID_SOURCE

isnanf(), isnanl():
    _XOPEN_SOURCE >= 600
        || /* glibc >= 2.19: */ _DEFAULT_SOURCE
        || /* glibc <= 2.19: */ _BSD_SOURCE || _SVID_SOURCE

---

🔥 DESCRIPTION 🔥

so what do these actually do? lemme break it down, bestie, one function family at a time.

• finite(), finitef(), finitel() — these return a nonzero value if x is neither infinite nor a "not-a-number" (NaN) value, and 0 otherwise. so basically: "is this number a normal, well-adjusted member of society?" yes → nonzero. no → 0. 💖
• isnan(), isnanf(), isnanl() — these return a nonzero value if x is a NaN value, and 0 otherwise. the detector for when your math went completely off the rails. 😭
• isinf(), isinff(), isinfl() — these return 1 if x is positive infinity, -1 if x is negative infinity, and 0 otherwise. three possible answers, not just two — she's nuanced. 👑

---

🦋 ATTRIBUTES 🦋

for the whole "what do these terms even mean" situation, go peek at attributes(7). the original prints this as a table, but tables don't exist in our group chat so here it is as a vibe:

• Interface: finite(), finitef(), finitel(), isinf(), isinff(), isinfl(), isnan(), isnanf(), isnanl()
• Attribute: Thread safety
• Value: MT-Safe

so yeah — all nine of them are MT-Safe, every single one, no exceptions. 🫶

---

⚠️ NOTES ⚠️

okay so this is the part where I have to be real with you:

these functions are obsolete. like, deprecated-era. C99 defines the macros isfinite(), isinf(), and isnan() (for all types) and those are what replaced them. 💀

and there's more — the C99 isinf() has weaker guarantees on the return value than the one you're reading about right now. so heads up on that.

go read fpclassify(3) for the full tea. 🌸

---

⭐ SEE ALSO ⭐

• fpclassify(3)

---

so that's finite(3)! iconic, a little vintage, still lives in libm rent-free. go forth and classify your floats, bestie 💅✨

this page was rocking Linux man-pages 6.20, dated 2026-02-08 — ty for coming to my TED talk, mwah 🎀🦋💜