blk_mq_quiesce_queue(9)

💅 blk_mq_quiesce_queue(9) — from the Kernel API Manual, bestie

okay so picture this: it's a Thursday, the block layer is Booked And Busy, and you need everyone to just chill for a sec. That's this function. It's blk_mq_quiesce_queue, and no cap, the whole point is "wait until all ongoing dispatches have finished." Iconic behavior honestly. 🎀

---

✨ NAME ✨

blk_mq_quiesce_queue — wait until all ongoing dispatches have finished 🦋

---

🌸 SYNOPSIS 🌸

void blk_mq_quiesce_queue (struct request_queue *q );

---

💖 ARGUMENTS 💖

• q — the request queue. yeah. that's the one. that's the queue we're quiescing, bestie. 🫶

---

⚠️ NOTE ⚠️

okay this part is important so listen up fr fr:

this function does not prevent the struct request end_io callback function from being invoked. nope. not doing that. that's not its job 💀

but here's the slay: once this function has returned, we make sure no dispatch can happen until the queue is unquiesced via blk_mq_unquiesce_queue. periodt. that's the guarantee. 💅

so like… end_io can still show up to the party, but nothing new gets dispatched until you say so. got it? got it. ✨

---

👑 SEE ALSO 👑

Kernel file ./block/blk-mq.c, plus the whole friend group:

• blk_mq_wait_quiesce_done(9)
• blk_update_request(9)
• blk_mq_complete_request(9)
• blk_mq_start_request(9)
• blk_execute_rq_nowait(9)
• blk_execute_rq(9)
• blk_mq_delay_run_hw_queue(9)
• blk_mq_run_hw_queue(9)
• blk_mq_run_hw_queues(9)
• blk_mq_delay_run_hw_queues(9)
• blk_mq_request_bypass_insert(9)
• blk_mq_try_issue_directly(9)
• blk_mq_submit_bio(9)
• blk_insert_cloned_request(9)
• blk_rq_unprep_clone(9)
• blk_rq_prep_clone(9)
• blk_mq_destroy_queue(9)

---

and that's the whole page, bestie 💜 this little gem lives in ./block, dated October 2026, and it literally says blk_mq_quiesce_queue(9) right there in the corner like a little name tag 🎀 go forth and quiesce responsibly 🧚‍♀️🌸