ktime_get_real_seconds(9)

📦 version 7.3.1, in nixpkgs unstable

💖 ktime_get_real_seconds — kernel API manual, bestie edition 💖

okay so this one's a little function from the kernel API manual (ktime_get_real_seconds(9)) and honestly? she's lowkey iconic for how much she does with so little. let's get into it 🫶

✨ NAME ✨

ktime_get_real_seconds — bestie, all she does is get the seconds portion of CLOCK_REALTIME and honestly that's enough 💅

🌸 SYNOPSIS 🌸

time64_t ktime_get_real_seconds (void );

look at her. no mess. no drama. just a time64_t coming back at you 🎀

🦋 ARGUMENTS 🦋

  • void — no arguments. none. zero. she takes nothing from you, she just gives 😌 literally the healthiest relationship in the whole kernel fr fr

💜 DESCRIPTION 💜

okay so what she returns is the wall clock seconds since 1970. that's it, that's the whole vibe. epoch-based, seconds-only, no nanoseconds crumbs left on the table ⭐

now here's where the architecture tea gets real 🍵:

  • on 64bit systems the fast access to tk->xtime_sec is preserved — she just goes straight in there, no cap, main character energy ✨
  • on 32bit systems the access must be protected with the sequence counter to provide "atomic" access to the 64bit tk->xtime_sec value. yes, "atomic" in quotes — that's the manpage's quotes, not mine, and you should respect them 🙈
so if you're on 32bit, please don't go reading tk->xtime_sec raw like it's a public feed. she needs her sequence counter. it's giving "locked door for a reason" 💀

🌈 SEE ALSO 🌈

the whole friend group, and I mean the whole friend group — she does not show up alone 👑

the kernel file ./kernel/time/timekeeping.c plus:

like I said. the whole group chat. 💖

and that's the whole page, bestie — no crumbs left behind 🍓 this lil drop came out of ./kernel/time in October 2026, and it's filed under ktime_get_real_seconds(9) forever and ever. go forth and get those seconds 👑💅✨

📦 where this page comes from