amdgpu_mm_wreg_mmio_rlc(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

πŸ’… amdgpu_mm_wreg_mmio_rlc β€” like, the register writer, bestie

okay so real quick: this is from the Kernel API Manual (the section 9 girlies know what's up ✨), and we're talking about amdgpu_mm_wreg_mmio_rlc. Yes the name is a whole mouthful, no I will not be shortening it, it's iconic as is. πŸŽ€


✨ NAME ✨

amdgpu_mm_wreg_mmio_rlc β€” writes a register, and it's giving options: either you go the direct/indirect mmio route, or you take the RLC path if you're in range. versatile queen πŸ’…


🌸 SYNOPSIS 🌸

void amdgpu_mm_wreg_mmio_rlc (struct amdgpu_device *adev , uint32_t reg
, uint32_t v , uint32_t xcc_id );

πŸ¦‹ ARGUMENTS πŸ¦‹

  • adev β€” an amdgpu_device pointer, aka the main character of this whole operation πŸ‘‘
  • reg β€” the mmio/rlc register you're writing to, bestie
  • v β€” the value you wanna write, the actual content of the vibe
  • xcc_id β€” the xcc accelerated compute core id (yes that's a real thing, and yes you need it 🫢)

πŸ’œ DESCRIPTION πŸ’œ

okay so the tea: this function is only invoked for the debugfs register access. that's it, that's the whole scope, no cap. it's a debugfs-only girlie, and honestly? respect. πŸ’€


🌈 SEE ALSO 🌈

the whole friend group, fr fr β€” go check the kernel file ./drivers/gpu/drm/amd/amdgpu/amdgpu_reg_access.c and also:


and that's the page, bestie πŸ“ huge shoutout to the source at ./drivers/gpu/drm/amd/amdgpu, dated October 2026 β€” literally slayed, mwah πŸ’–πŸ§šβ€β™€οΈ

πŸ“¦ where this page comes from