__nfp_nsp_command(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

πŸ’… __nfp_nsp_command(9) β€” like, the whole kernel API manual tea πŸ«–

okay so bestie, welcome to the Kernel API Manual entry for __nfp_nsp_command β€” yes it's a mouthful, yes it's iconic, no we are not skipping a single character of it. this one lives in the (9) section, aka the kernel-functions section, and it's all about firing a command off to the NFP Service Processor. buckle up πŸŽ€


✨ NAME ✨

__nfp_nsp_command β€” execute a command on the NFP Service Processor. that's it, that's the whole vibe. it sends the thing, the SP does the thing. πŸ’«


🌸 SYNOPSIS 🌸

int __nfp_nsp_command (struct nfp_nsp *state , const struct
nfp_nsp_command_arg *arg );

she takes two things and hands you back an int, no cap. that's the whole signature, don't come for me about the spacing, it's verbatim πŸ’€


πŸ¦‹ ARGUMENTS πŸ¦‹

  • state β€” the NFP SP state. like, which SP are we even talking to rn? this is that. main character of the pair honestly πŸ‘‘
  • arg β€” the NFP command argument structure. this is where you tell it what command you want. the whole personality of the call lives in here πŸŽ€

πŸ’– RETURN πŸ’–

  • 0 β€” success!! and specifically success with no result. clean, empty-handed, thriving ✨
real talk: the original manpage puts the rest of the good stuff under DESCRIPTION instead of here, and we are respecting the layout exactly because it's not our place to reorganize someone else's manpage. go look one heading down, bestie 🫢

🌈 DESCRIPTION 🌈

here's your full menu of return values, and no we are not cherry-picking β€” every single one is coming with us:

  • a positive value β€” NSP completion with a result code. so it finished and it had something to say about it πŸ’…

and the error squad, all of them, in order:

  • -EAGAIN β€” the NSP is not yet present. like it hasn't shown up to the group chat yet 😭
  • -ENODEV β€” the NSP is not a supported model. wrong device, wrong era, we can't work with that πŸ™ˆ
  • -EBUSY β€” the NSP is stuck. it's giving frozen mid-sentence. highkey concerning πŸ’€
  • -EINTR β€” interrupted while waiting for completion. someone cut in line while you were patiently waiting, rude πŸ“
  • -ETIMEDOUT β€” the NSP took longer than timeout_sec seconds to complete. it overstayed. the timer said no more. ⏰
fr fr read those carefully β€” each one means a different kind of bad day, so don't go lumping them together. timeout_sec is literally the threshold for that last one, keep an eye on it.

πŸ”₯ SEE ALSO πŸ”₯

  • the kernel file ./drivers/net/ethernet/netronome/nfp/nfpcore/nfp_nsp.c β€” go read the source, iykyk πŸ“–
  • struct nfp_nsp_command_arg(9)
  • struct nfp_nsp_command_buf_arg(9)
  • nfp_nsp_open(9)
  • nfp_nsp_close(9)

okay that's the page!! you ate that, honestly 🫢 this lil beauty comes from ./drivers/net/ethernet/netrono... and is dated October 2026 β€” mwah, go forth and command your service processor, bestie πŸ’œβœ¨

πŸ“¦ where this page comes from