free_ud_wq_attr(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

🌸 free_ud_wq_attr(9) β€” kernel API manual, bestie edition 🌸

okay so you found a (9) page, that's the kernel API manual section, and this one is literally free_ud_wq_attr πŸ’… it's a tiny little cleanup function and honestly? iconic for how small it is.


✨ NAME ✨

free_ud_wq_attr β€” clean up that AH attribute cache for UD QPs 🧹✨

πŸ¦‹ SYNOPSIS πŸ¦‹

void free_ud_wq_attr (struct rvt_qp *qp );

no cap, that's the whole signature. void return, one pointer argument, done. πŸŽ€

πŸ“ ARGUMENTS πŸ“

  • qp β€” a valid QP with allowed_ops set. that's the whole vibe check for this argument, it's gotta be valid and have allowed_ops set or we're not doing this πŸ’€

πŸ‘‘ DESCRIPTION πŸ‘‘

so here's the tea: the rvt_swqe data structure being used is a union, and because of that, this function is only valid for UD QPs. like, only UD. other QP types? not it, don't even try πŸ™ˆ

lowkey important: this is a union situation, so the "only for UD QPs" thing isn't a suggestion, it's the whole rule. periodt. πŸ”₯

🌈 SEE ALSO 🌈

bestie, the cross-reference list here is long and I respect it deeply. all of these live in the kernel file ./drivers/infiniband/sw/rdmavt/qp.c:

fr fr, read them all, they're the whole squad 🫢


and that's the page!! sign-off time: this one comes from ./drivers/infiniband/sw/rdmavt, dated October 2026, and it's page free_ud_wq_attr(9) of the kernel API manual. slay, go free some UD QP attributes πŸ’–

πŸ“¦ where this page comes from