ib_drain_qp(9)

💖 NAME 💖

ib_drain_qp(9) — block until all CQEs have been consumed by the application on both the RQ and SQ. It’s giving “wait for your bestie to finish,” but make it kernel API. From the Kernel API Manual, no cap. ✨

✨ SYNOPSIS ✨

void ib_drain_qp (struct ib_qp *qp );

🎀 ARGUMENTS 🎀

• qp — queue pair to drain. She’s the main character here, periodt.

🦋 DESCRIPTION 🦋

Okay so before you even think about calling this, the caller must:

• ensure there is room in the CQ(s), SQ, and RQ for drain work requests and completions.
• allocate the CQs using ib_alloc_cq.
• ensure that there are no other contexts that are posting WRs concurrently. Otherwise the drain is not guaranteed.

Bestie, that last one is not a vibe suggestion. If other contexts are posting WRs at the same time, the drain is not guaranteed. So like, don’t. 💀

👑 SEE ALSO 👑

Kernel file ./drivers/infiniband/core/verbs.c — and the whole icon friend group:

• __ib_alloc_pd(9)
• ib_dealloc_pd_user(9)
• rdma_copy_ah_attr(9)
• rdma_replace_ah_attr(9)
• rdma_move_ah_attr(9)
• rdma_create_ah(9)
• rdma_create_user_ah(9)
• rdma_move_grh_sgid_attr(9)
• rdma_destroy_ah_attr(9)
• ib_create_srq_user(9)
• ib_create_qp_user(9)
• ib_resolve_eth_dmac(9)
• ib_modify_qp_with_udata(9)
• ib_alloc_mr(9)
• ib_alloc_mr_integrity(9)
• ib_alloc_xrcd_user(9)
• ib_dealloc_xrcd_user(9)
• ib_create_wq(9)
• ib_destroy_wq_user(9)
• ib_map_mr_sg_pi(9)
• ib_map_mr_sg(9)
• ib_sg_to_pages(9)
• ib_drain_sq(9)
• ib_drain_rq(9)
• rdma_alloc_hw_stats_struct(9)
• rdma_free_hw_stats_struct(9)

xoxo bestie 💅 source/version is ./drivers/infiniband/core, date is October 2026, page is ib_drain_qp(9) — slay and drain responsibly ✨🫶