__ice_vsi_get_qs(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

πŸ’… __ice_vsi_get_qs(9) β€” like, the actual function name, bestie

Okay so real talk, this page lives in the Kernel API Manual section (9) and it's all about __ice_vsi_get_qs β€” the little helper function that handles assigning queues from the PF to the VSI πŸŽ€ iykyk.


✨ SYNOPSIS ✨

int __ice_vsi_get_qs (struct ice_qs_cfg *qs_cfg );

That's it. One function, one pointer, one int coming back at you πŸ’–


🫢 ARGUMENTS 🫢

  • qs_cfg β€” the little bundle of gathered variables you need for the pf->vsi queues assignment. It's literally carrying all the info for the handoff 🌸

πŸ¦‹ DESCRIPTION πŸ¦‹

So the way this works: __ice_vsi_get_qs first tries to find contiguous space. Contiguous. Like, all in a row, sitting pretty together ✨

If that doesn't work out? No stress. It pivots and tries with the scatter approach instead β€” spreading out, doing its own thing, still getting the job done πŸ’…

And the return values, because we always check those:

  • 0 on success πŸŽ‰
  • -ENOMEM in case of no left space in the PF queue bitmap 😭 (the bitmap is full, bestie, there's nowhere left to put anything)
Heads up: -ENOMEM here isn't about the system being out of memory in the usual sense β€” it means there's no left space in the PF queue bitmap. Read the error code, don't assume πŸ™ˆ

πŸ‘‘ SEE ALSO πŸ‘‘

The kernel file this lives in, no cap:

./drivers/net/ethernet/intel/ice/ice_base.c

And here's the whole friend group β€” every single one of them, in order, exactly as listed, because we do not skip entries in this house πŸ”₯


And that's the whole vibe check, bestie 🧚 Signing off from ./drivers/net/ethernet/intel/ice β€” this page is dated October 2026, and the man page itself is __ice_vsi_get_qs(9). Go forth and assign those queues πŸ’œπŸŒˆβ­

πŸ“¦ where this page comes from