hfi1_schedule_tid_send(9)

💖 hfi1_schedule_tid_send — NAME 💖

okay so bestie, this is hfi1_schedule_tid_send, and it literally schedules progress on the TID RDMA state machine ✨ like that's the whole job and she does it gorgeously.

---

🎀 SYNOPSIS 🎀

bool hfi1_schedule_tid_send (struct rvt_qp *qp );

---

🦋 ARGUMENTS 🦋

• qp — the QP 💅 that's it, that's the argument, she's the main character here.

---

🌸 DESCRIPTION 🌸

So what this does, right — it schedules progress on qp for the TID RDMA state machine. no cap, that's the vibe.

Heads up though: the caller should hold the s_lock. 🔒 that's on you, not her, okay? she's not doing it for you.

And here's the little plot twist: unlike hfi1_schedule_send, this one cannot use hfi1_send_ok, because the two state machines can literally step on each other with respect to the RVT_S_BUSY flag 😭💀 like imagine two besties both grabbing the same mirror at the same time. messy. so instead, a modified test is used. iykyk 🧚

---

✨ RETURN ✨

• true — the second leg is scheduled 🫶 slay
• false — the second leg is not scheduled 🙈

---

👑 SEE ALSO 👑

kernel file ./drivers/infiniband/hw/hfi1/tid_rdma.c, plus the whole friend group:

• TID RDMA READ protocol(9)
• qp_to_rcd(9)
• lock ordering(9)
• kernel_tid_waiters(9)
• dequeue_tid_waiter(9)
• queue_qp_for_tid_wait(9)
• __trigger_tid_waiter(9)
• tid_rdma_schedule_tid_wakeup(9)
• tid_rdma_trigger_resume(9)
• kern_reserve_flow(9)
• tid_rdma_find_phys_blocks_4k(9)
• tid_flush_pages(9)
• tid_rdma_find_phys_blocks_8k(9)
• hfi1_kern_exp_rcv_setup(9)
• hfi1_kern_exp_rcv_free_flows(9)
• __trdma_clean_swqe(9)

---

and that's the page, bestie 🌈 this iconic little piece of the Kernel API Manual is living in ./drivers/infiniband/hw/hfi1, dated October 2026 — go read the source file, mwah 💜🎀🍓