raw_atomic_t...xchg_release(9)

📦 version 7.3.1, in nixpkgs unstable

💅 NAME 💅

raw_atomic_try_cmpxchg_release — it's the atomic compare-and-exchange function, but make it release-ordering ✨ (this is page 9 of the Kernel API Manual, bestie, so you know it's a real one 🫶)

🌸 SYNOPSIS 🌸

bool raw_atomic_try_cmpxchg_release (atomic_t *v , int *old , int new );

🎀 ARGUMENTS 🎀

  • v — a pointer to atomic_t. this is the thing we're messing with, the main character of the whole operation 👑
  • old — a pointer to the int value we're comparing against. note that it's a pointer, bestie, she does double duty (more on that below 👀)
  • new — the int value to assign. the new era 💫

🦋 DESCRIPTION 🦋

okay so here's the vibe check. if (v == old), the function atomically updates v to new with release ordering. slay, exchange happened 💅

but if that's not the case? then v is not modified at all — she said "no thank you" — and instead old gets updated to the current value of v, and relaxed ordering is provided. so old walks away holding the real answer. that's the plot twist, iykyk 😭

it's safe to use in noinstr code — but fr fr, prefer atomic_try_cmpxchg_release everywhere else. the raw_ version is the low-level one, it's giving "built different, use with intention" energy 💀

⭐ RETURN ⭐

true if the exchange occurred, false otherwise. that's it, that's the whole personality 🌈

💖 SEE ALSO 💖

lowkey the biggest friend group in the whole kernel, no cap. here's everyone:

that's the whole page, bestie 🫶 go forth and compare-and-exchange responsibly 💜

this icon lives in ./include/linux/atomic, dated October 2026, and the header up top calls her raw_atomic_t...xchg_release(9) (the name got truncated mid-vibe, we don't gatekeep, we love her anyway 💀✨)

📦 where this page comes from