xe_pm_runtime_get_noresume(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

πŸ’… xe_pm_runtime_get_noresume(9) β€” Kernel API Manual πŸ’…

okay so this is the page for the runtime PM girlies who need to bump a counter without resuming anything. let's get into it, bestie 🫢


πŸŽ€ NAME πŸŽ€

xe_pm_runtime_get_noresume β€” Bump runtime PM usage counter without resuming

πŸ’» SYNOPSIS πŸ’»

void xe_pm_runtime_get_noresume (struct xe_device *xe );

🌸 ARGUMENTS 🌸

  • xe β€” xe device instance

✨ DESCRIPTION ✨

Okay so picture this: you're in an inner place in the code, and you are surely already protected by outer-bound callers of xe_pm_runtime_get. That is literally the vibe this function is for. It will warn if you're not protected β€” so don't just sneak it in somewhere random, bestie 😭

And no matter what, you still have to put the reference back after this function. It will always bump the usage counter, regardless. Yes, regardless of whatever else is going on. πŸ™ˆ

When possible, scope-based runtime PM (through guard(xe_pm_runtime_noresume)) is preferred over direct usage of this function. Manual get/put handling should only be used when:

  • the function contains goto-based logic which can break scope-based handling, or
  • the lifetime of the runtime PM reference does not match a specific scope (e.g., runtime PM obtained in one function and released in a different one).

πŸ‘€ SEE ALSO πŸ‘€


mwah πŸ’‹ this page comes from source/version ./drivers/gpu/drm/xe, dated October 2026, and it's the xe_pm_runtime_get_noresume(9) page. bye bestie! 🫢

πŸ“¦ where this page comes from