sdma_get_descq_cnt(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

πŸ’… sdma_get_descq_cnt β€” kernel API manual, bestie edition πŸŽ€

okay so real talk, you opened the manpage for sdma_get_descq_cnt and it's a kernel API manual situation β€” section 9, the whole vibe. come sit with me, this one's short and sweet and we're gonna slay it πŸ’–


✨ NAME ✨

sdma_get_descq_cnt β€” literally the function that gets called when the device gets probed. that's it, that's the whole name section 🫢


🌸 SYNOPSIS 🌸

u16 sdma_get_descq_cnt (void );

look at her. returns a u16, takes nothing. iconic behavior honestly πŸ‘‘


πŸ¦‹ ARGUMENTS πŸ¦‹

  • void β€” no arguments. zero. nada. she doesn't want anything from you, she's lowkey self-sufficient 😌

πŸ’œ DESCRIPTION πŸ’œ

so what does she actually do? she returns a validated descq count ✨ that's the whole flex.

rn she's currently only used in the verbs initialization, specifically to build the tx list β€” so like, that's her one job and she's doing it πŸ’…

⚠️ heads up bestie: this will probably be deleted in favor of a more scalable approach to alloc tx's. so don't get too attached, she's giving limited-edition energy 😭

πŸ”₯ SEE ALSO πŸ”₯

the kernel file lives at ./drivers/infiniband/hw/hfi1/sdma.c and here's the whole friend group, no cap, all of them:


and that's the page! 🧚 this one came from ./drivers/infiniband/hw/hfi1, dated October 2026, and it's sdma_get_descq_cnt(9) β€” go forth and build that tx list, bestie, you got this fr fr 🌈⭐

πŸ“¦ where this page comes from