amdkfd_fenc...le_signaling(9)
π amdkfd_fence_enable_signaling β the Kernel API Manual era π
okay so bestie, welcome to the glow-up of amdkfd_fence_enable_signaling(9) β yes that's a nine in the parens, it's a kernel-doc page, we're being β¨officialβ¨ today.
β¨ NAME β¨
amdkfd_fence_enable_signaling β the function that gets called when TTM wants to evict a KFD BO and schedules a job to move the BO ππ¨
And here's the vibe check on what it returns: if the fence is already signaled, it literally just returns true π«Ά. If the fence is not signaled, it schedules an evict KFD process work item instead. no cap, that's the whole personality.
πΈ SYNOPSIS πΈ
bool amdkfd_fence_enable_signaling (struct dma_fence *f );
that's it, that's the whole signature, she's a bool and she's thriving π
π¦ ARGUMENTS π¦
fβ adma_fence. that's the argument, that's the girl. it's a dma_fence, mwah π
π SEE ALSO π
- the kernel file
./drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_fence.cβ go read the source, it's giving main character energy π amdkfd_fence_release(9)β her bestie, the release one π«Άamdkfd_fence_check_mm(9)β and the mm-checking one too, the trio ate π
ngl if you're here it's probably because you're deep in the AMDGPU/KFD fence lore. that's iconic of you, fr fr π§ββοΈ
and that's the page! π this little beauty lives under ./drivers/gpu/drm/amd/amdgpu and the manual itself is dated October 2026 β so mark your calendars bestie, see you then πβπ
π¦ where this page comes from
- nixpkgs unstable: linux-manual 7.3.1
- π girlified by deepseek-ai/DeepSeek-V4.1-Flash-1017 on : a language model's rewrite of the original, under its terms