tb_domain_d...t_pcie_paths(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

πŸ’… tb_domain_disconnect_pcie_paths(9) β€” aka the "bye-bye PCIe paths" era πŸ’…

okay so first things first, bestie: this little beauty lives in the Kernel API Manual (that's the manual name, noted ✨), and the page title is the gloriously truncated tb_domain_d...t_pcie_paths(9) β€” which is literally just tb_domain_disconnect_pcie_paths wearing a tiny tiny trenchcoat. iykyk 🫢


🌸 NAME 🌸

tb_domain_disconnect_pcie_paths β€” the function that disconnects all the PCIe paths. like, all of them. every last one. no cap πŸ’…


✨ SYNOPSIS ✨

int tb_domain_disconnect_pcie_paths (struct tb *tb );

that's it. that's the whole vibe. one arg, one int, one mission πŸŽ€


πŸ¦‹ ARGUMENTS πŸ¦‹

  • tb β€” the domain whose PCIe paths you're about to disconnect. like, hand her the domain and she'll yeet the paths for you. main character energy honestly 🌈

πŸ’– DESCRIPTION πŸ’–

okay so listen. this function needs to be called in preparation for an NVM upgrade of the host controller. needs. it's not optional, it's not a vibe suggestion, it's a requirement, periodt πŸ—£οΈ

and what does she actually do? she makes sure all PCIe paths are disconnected. all of them. she is thorough. she is iconic. she does not leave a single path behind 😭✨

fr fr, think of it as the pre-NVM-upgrade ritual: clear the paths first, then do the upgrade. skipping this is giving chaos, not giving slay πŸ’€

πŸ”₯ RETURN πŸ”₯

  • 0 on success β€” we love a clean victory πŸŽ‰
  • a negative errno in case of error β€” because sometimes the universe says no πŸ™ˆ

πŸ‘‘ SEE ALSO πŸ‘‘

kernel file ./drivers/thunderbolt/domain.c and honestly the whole thunderbolt friend group:

like, the whole squad πŸ§šπŸ’œ


okay that's the page!! it's sourced from ./drivers/thunderbolt, dated October 2026, and it's rocking tb_domain_d...t_pcie_paths(9) as its header β€” go forth and disconnect those paths, bestie, you got this πŸ’…πŸŽ€β­

πŸ“¦ where this page comes from