tid_rdma_sch...e_tid_wakeup(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

πŸ’… tid_rdma_schedule_tid_wakeup β€” Kernel API Manual ✨

okay so bestie, welcome to the group chat, this is the manual page for tid_rdma_schedule_tid_wakeup(9) and honestly? she's a little baddie with one job and one job only πŸŽ€


🌸 NAME 🌸

tid_rdma_schedule_tid_wakeup β€” schedule wakeup for a qp πŸ’€βž‘οΈπŸ””


✨ SYNOPSIS ✨

void tid_rdma_schedule_tid_wakeup (struct rvt_qp *qp );

like yes that's the whole signature, she's short and she knows it 🫢


πŸŽ€ ARGUMENTS πŸŽ€

  • qp β€” the qp. that's it. that's the argument. the qp in question, the main character of this function πŸ”₯

πŸ’– DESCRIPTION πŸ’–

so what she does is trigger a schedule or a waiting qp in a deadlock safe manner. no cap, deadlock safety is the whole vibe here β€” she does NOT do chaos, she does controlled chaos πŸ’…

the qp reference is held prior to this call via first_qp. so like, the ref is already yours before you even walk in the door, iykyk ✨

then here's the tea πŸ‘€ β€” if the qp trigger was already scheduled (!rval), the reference is dropped. otherwise, the resume or the destroy cancel will dispatch the reference. either way that ref gets handled, she never leaves it hanging, periodt 🧚

fr fr pay attention to which branch you're in β€” the ref gets dropped in one case and dispatched by resume/destroy cancel in the other. no leaks, no mess πŸ’€

πŸ¦‹ SEE ALSO πŸ¦‹


🌈 ok that's the whole page, love u, bye 🌈

this iconic little number lives in ./drivers/infiniband/hw/hfi1, dated October 2026, and the page header/footer goes by tid_rdma_sch...e_tid_wakeup(9) β€” yes the middle got truncated, she's mysterious like that πŸ’œ stay hydrated, hold your references correctly, and slay πŸ’…βœ¨

πŸ“¦ where this page comes from