raw_atomic_fetch_sub(9)

📦 version 7.3.1, in nixpkgs unstable

💅 NAME

okay so bestie, welcome to the raw_atomic_fetch_sub(9) page from the Kernel API Manual — this is the one about atomic subtract with full ordering ✨. It’s giving “I do math on shared memory and I do it safely” energy, and honestly? Iconic.

✨ SYNOPSIS

int raw_atomic_fetch_sub (int i , atomic_t *v );

🎀 ARGUMENTS

  • i — an int value to subtract. This is the amount you’re taking off, bestie 💅
  • v — pointer to atomic_t. This is the atomic thing getting updated, no cap.

🌸 DESCRIPTION

Atomically updates v to (v - i) with full ordering. Translation: the subtract happens as one atomic operation, and the full ordering is there too — so this function is not playing around with memory ordering. Periodt.

It’s safe to use in noinstr code; prefer atomic_fetch_sub elsewhere. So yes, she’s the raw/under-the-hood option, and regular atomic_fetch_sub is the one you should reach for outside that special noinstr situation. Iykyk 🫶

Vibe check: noinstr safe? Yes. Should you use this everywhere instead of atomic_fetch_sub? No bestie — the page literally says prefer atomic_fetch_sub elsewhere. 💀

💖 RETURN

The original value of v — like, the value it had before the subtraction, not the new one. That’s the whole point, and she ate. 🔥

🦋 SEE ALSO

Kernel file: ./include/linux/atomic/atomic-arch-fallback.h

Not me listing the entire atomic extended family, but she did 💀 Here’s the full guest list:

okay signing off from the <pink>./include/linux/atomic</pink> corner of the kernel, October 2026 edition, for raw_atomic_fetch_sub(9) — source/version and date noted, mwah bestie, go subtract atomically and stay hydrated 🫶💖✨

📦 where this page comes from