hisi_qm_alloc_qps_node(9)

📦 version 7.3.1, in nixpkgs unstable

✨ hisi_qm_alloc_qps_node — Kernel API Manual ✨

okay so bestie, we're in the hisi_qm_alloc_qps_node(9) era and honestly? it's giving "let me build you a whole squad of queue pairs" 💅 this is a kernel API page, so grab a strawberry 🍓 and let's go

🎀 NAME 🎀

hisi_qm_alloc_qps_node — literally just create multiple queue pairs. that's it, that's the vibe ✨

🦋 SYNOPSIS 🦋

int hisi_qm_alloc_qps_node (struct hisi_qm_list *qm_list , int qp_num ,
u8 *alg_type , int node , struct hisi_qp **qps );

like, look at her. five args, one int return, main character energy 👑

🌸 ARGUMENTS 🌸

  • qm_list — the list of all available devices. every single one of them, no cap 👀
  • qp_num — the number of queue pairs that need created 🔢
  • alg_type — the algorithm type 💫
  • node — the numa node 📍
  • qps — the queue pairs that need created 🫶

💖 DESCRIPTION 💖

okay so here's the tea ☕ this function will sort all available device according to numa distance. like she's matchmaking, but for hardware 🧚

then she tries to create all queue pairs from one device — all of them from a single device, periodt 💅

and if all devices do not meet the requirements? she will return error. no notes, that's just how it is 😭

👑 SEE ALSO 👑

the whole friend group lives in the kernel file ./drivers/crypto/hisilicon/qm.c, and here's everybody, no one left out fr fr 🌈

lowkey obsessed with how this whole page is just "trust the numa distance sort" 🙈 but fr, that's the whole spec, don't come for me

and that's the page bestie!! 🎀 source path ./drivers/crypto/hisilicon, dated October 2026, page hisi_qm_alloc_qps_node(9) — mwah, slay, bye 💜🔥

📦 where this page comes from