queue_work(9)

💅 queue_work(9) — the Kernel API Manual, but make it make sense 🎀

okay so bestie, this one lives in the Kernel API Manual section 9 and honestly? it's a whole vibe. it's like the post it note of the workqueue world: "hey, pls run this later." let's get into it ✨

---

✨ NAME ✨

queue_work — queue work on a workqueue 🫶

like, that's it. that's the whole job. you hand it a thing and it goes "say less, it's queued."

🌸 SYNOPSIS 🌸

bool queue_work (struct workqueue_struct *wq , struct work_struct *work
);

so you get a bool back, bestie. a whole true-or-false. feral behaviour but ok 🔥

🦋 ARGUMENTS 🦋

• wq — the workqueue to use. this is the venue, the venue is booked, we're going 🎀
• work — the work to queue. this is the main character of the operation, the thing that's actually gonna run 👑

💜 DESCRIPTION 💜

so, the return value? it's giving boolean drama. it returns false if the work was already on a queue, and true otherwise. so like, false = "girl it's already in line", true = "congrats, fresh queue entry, slay" ✨

we queue the work to the CPU on which it was submitted — very much "you grew here, you're one of us" energy 🥺 — but if that CPU dies, it can be processed by another CPU. no cap, the work survives, the CPU does not 😭💀

Memory-ordering properties: ooh okay this is the important part, pay attention bestie 👀 — if it returns true, that guarantees all stores preceding the call to queue_work in program order will be visible from the CPU which will execute work by the time such work executes, e.g.

{ x is initially 0 }

CPU0                               CPU1

WRITE_ONCE(x, 1);                  [ work is being executed ] r0 =
queue_work(wq, work);           r1 = READ_ONCE(x);

Forbids: r0 == true && r1 == 0

like literally forbidden. not allowed. don't even think about it 🚫💅

ngl the memory ordering bit is the loading-screen lore you cannot skip. it's a guarantee, not a suggestion. if you're doing the WRITE_ONCE → queue_work combo and the other side reads with READ_ONCE, the r0 == true && r1 == 0 scenario is off the table. periodt. 🎀

🌈 SEE ALSO 🌈

the whole friend group, honestly — go read these too, iykyk 🫶

• Kernel file ./include/linux/workqueue.h struct workqueue_attrs(9)
• work_pending(9)
• delayed_work_pending(9)
• alloc_workqueue(9)
• devm_alloc_workqueue(9)
• alloc_workqueue_lockdep_map(9)
• alloc_ordered_workqueue_lockdep_map(9)
• alloc_ordered_workqueue(9)
• queue_delayed_work(9)
• mod_delayed_work(9)
• schedule_work_on(9)
• schedule_work(9)
• enable_and_queue_work(9)
• schedule_delayed_work_on(9)
• schedule_delayed_work(9)

---

💖 one (1) tiny thing before you go: it returns true only when it actually queued something new — a false is not an error, it's just "been there, queued that." don't @ me, it's literally what the manpage says 🫶

k that's the whole page, ily, go queue some work 💅✨

— from ./include/linux, dated October 2026, queue_work(9) out 🎀🦋💜