pci_sriov_set_totalvfs(9)

📦 version 7.3.1, in nixpkgs unstable

✨ pci_sriov_set_totalvfs ✨

okay bestie, this is a page from the Kernel API Manual (section 9), and it's genuinely one of the shorter ones so we can absolutely get through it together 💖

🌸 NAME 🌸

pci_sriov_set_totalvfs — reduce the TotalVFs available (yes that's it, that's the whole vibe — we're turning the VF count down, not up 🔻)

🎀 SYNOPSIS 🎀

int pci_sriov_set_totalvfs (struct pci_dev *dev , u16 numvfs );

so it takes a struct pci_dev * and a u16, and hands you back an int. cute, concise, no notes 💅

🦋 ARGUMENTS 🦋

  • dev — the PCI PF device 🫶 (the Physical Function, the main character of this whole situation)
  • numvfs — the number that should be used for TotalVFs supported 🔢 like, the new cap you're setting

🌈 DESCRIPTION 🌈

okay so listen — this should be called from the PF driver's probe routine, and it must be called with the device's mutex held 🔒 (fr fr, hold that lock, don't be out here raw-dogging shared state)

Now for the returns, and ngl this part is giving ✨ decision tree ✨:

  • Returns 0 if the PF is an SRIOV-capable device and the value of numvfs is valid ✅ slay, you did it
  • If it's not a PF → <red>-ENOSYS</red> 😭
  • If numvfs is invalid → <red>-EINVAL</red> 🙈
  • If VFs are already enabled → <red>-EBUSY</red> 💀 too late bestie, they're already out there living their lives
heads up: call it early, from probe, mutex in hand. Once those VFs are enabled the door is closed and you get -EBUSY 🚪🔐

🧚 SEE ALSO 🧚

Kernel file ./drivers/pci/iov.c — that's where the magic lives 📁✨

and the whole friend group 👑:

like okay, go read all of them, I'll wait 🍓

and that's the whole page!! mwah 💋 this one lives in ./drivers/pci, dated October 2026, and it's filed under pci_sriov_set_totalvfs(9) — go forth and cap those TotalVFs responsibly, bestie 🫶🌟

📦 where this page comes from