intel_ggtt_restore_fences(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

πŸ’… intel_ggtt_restore_fences(9) β€” from the Kernel API Manual, bestie

okay so picture this: it's giving "put the fences back where you found them" energy. This page lives in the Kernel API Manual and it's about one lil function, intel_ggtt_restore_fences, whose whole personality is restore fence state. iconic, lowkey.


✨ NAME ✨

intel_ggtt_restore_fences β€” restore fence state πŸŽ€


πŸ¦‹ SYNOPSIS πŸ¦‹

void intel_ggtt_restore_fences (struct i915_ggtt *ggtt );

that's it. that's the whole signature. no cap. ✨


🌸 ARGUMENTS 🌸

  • ggtt β€” the Global GTT. that's the one and only thing you hand it. 🫢

πŸ’– DESCRIPTION πŸ’–

so what does it actually do, you ask? it restores the hw fence state so it matches the software tracking again. like, the hardware and the software have to be on the same page, and this function is the one that gets them there. match made in heaven, fr fr. πŸ’…

when do you call it? two moments, bestie:

  1. after a GPU reset πŸ”₯
  2. on resume 😴➑️😳

and here's the caveat you cannot skip: on runtime suspend we only cancel the fences, and they get reacquired by the user later. so do not come for this function on runtime suspend expecting a restore β€” that is not the vibe. note it, remember it, live by it. 🧚


πŸ‘‘ SEE ALSO πŸ‘‘

go read these next, iykyk:

like literally a whole friend group of fence-related pages. go say hi. 🌈


⭐ reminder for the group chat: this page is about restoring after a GPU reset and on resume, and on runtime suspend it's cancel only β€” the user reacquires later. don't mix those up, bestie. πŸ™ˆ

and that's the page!! mwah πŸ’‹ this one came from ./drivers/gpu/drm/i915/gt, dated October 2026, and it's still intel_ggtt_restore_fences(9) at the end of the day. slay, periodt. πŸŽ€βœ¨πŸ’œ

πŸ“¦ where this page comes from