amdgpu_gmc_filter_faults(9)

📦 version 7.3.1, in nixpkgs unstable

💅 amdgpu_gmc_filter_faults(9)

okay bestie, quick vibe check: this is a page from the Kernel API Manual about amdgpu_gmc_filter_faults — yes that's a mouthful, yes we're gonna get through it together 🫶

✨ NAME ✨

amdgpu_gmc_filter_faults — filter VM faults 🔥

like, it's the bouncer at the club door of VM faults, deciding who gets in and who gets told "not tonight, babe" 💅

🌸 SYNOPSIS 🌸

bool amdgpu_gmc_filter_faults (struct amdgpu_device *adev , struct
       amdgpu_ih_ring *ih , uint64_t addr , uint16_t pasid , uint64_t timestamp
       );

it returns a bool, no cap. and yes, that's the whole signature, character for character, don't come for me 😭

🎀 ARGUMENTS 🎀

five of them, all iconic in their own way:

  • adev — the amdgpu device structure 🖥️ literally the whole device, just chillin' as a pointer
  • ih — the interrupt ring that the fault received from ⭕ that's where the fault came in, bestie, know your source
  • addr — address of the VM fault 📍 the actual location where the drama went down
  • pasid — PASID of the process causing the fault 👀 yes we know it was you, we always know
  • timestamp — timestamp of the fault ⏰ when it happened, receipts included

🦋 RETURN 🦋

  • True 💀 — the fault was filtered and should not be processed further. it got turned away at the door, nothing more to see here
  • False ✨ — the fault is a new one and needs to be handled. main character energy, this one's getting the full treatment
no gray area here, it's a straight yes-or-no situation — filtered or not filtered, periodt

🌈 SEE ALSO 🌈

okay this is the friend group, the whole squad, all listed exactly as they appear:

and that's the whole page, bestie 🧚💜 this lil doc comes to you straight from ./drivers/gpu/drm/amd/amdgpu, dated October 2026 — signed, sealed, slayed 💅🎀 bye bestie, go filter some faults 🌸🦋

📦 where this page comes from