libie_ctlq_send(9)

📦 version 7.3.1, in nixpkgs unstable

💖 libie_ctlq_send — bestie, it's giving "send a message to Control Plane or Peer" 💅

okay so real quick, before we even get into it: this whole page lives in libie_ctlq_send(9), aka the Kernel API Manual. that's the vibe. that's the header. now let's GO 🦋

✨ NAME ✨

libie_ctlq_send — send a message to Control Plane or Peer. literally what it says on the tin, no cap 🎀

🌸 SYNOPSIS 🌸

void libie_ctlq_send (struct libie_ctlq_info *ctlq , u32 num_q_msg );

like, that's the whole signature. she's compact. she's iconic. copy it exactly, bestie, spaces and all 💅

🎀 ARGUMENTS 🎀

  • ctlq — the specific control queue that's being used for sending a message. like, which queue, bestie, be specific 🙈
  • num_q_msg — the number of messages that are present to send on ctlq. it's positive, and no greater than the number of available descriptors. so don't get greedy, okay? ⭐

🔥 DESCRIPTION 🔥

so before you even call this, the caller must fill in num_q_msg Tx messages starting at ntu beforehand. that's a you job, not a her job, iykyk 😭

the caller must hold ctlq->lock. and the intended pattern — write this down — is to first check the number of descriptors available, then fill in the messages, and perform send within a single critical section. one critical section. all of it. together. that's the whole bit 🫶

not me stressing the lock thing — it's genuinely load-bearing, don't skip it 💀

👑 SEE ALSO 👑

the kernel file lives at ./drivers/net/ethernet/intel/libie/controlq.c — go read it, it ate ✨

and here's the whole friend group, every single one, no skipping:

and that's the page!! 🧚💜

cute little sign-off moment: this one's from the source at ./drivers/net/ethernet/intel/l... and it's dated October 2026 — so like, future girlie energy fr fr ⭐🌸 go forth and hold that lock, bestie 🫶✨

📦 where this page comes from