uvd_v6_0_enc..._destroy_msg(9)

📦 version 7.3.1, in nixpkgs unstable

🎀 NAME 🎀

ok so bestie, welcome to uvd_v6_0_enc_get_destroy_msg — like the title says, this one generates a UVD ENC destroy msg 💅 and yeah, it lives in the Kernel API Manual, section 9, so you know it's fancy kernel-level business.

✨ SYNOPSIS ✨

int uvd_v6_0_enc_get_destroy_msg (struct amdgpu_ring *ring , uint32_t
handle , struct amdgpu_bo *bo , struct dma_fence **fence );

like that's the whole vibe, returning an int and taking four little friends 🌸

🫶 ARGUMENTS 🫶

here's the cast of characters, no cap:

  • ring — the ring we should submit the msg to 💫 that's where the message goes, we don't make the rules
  • handle — session handle to use, aka the lil ID for the session you're working with 🎀
  • bo — amdgpu object for which we query the offset, yes bestie we're asking it for the offset 🔍
  • fence — optional fence to return, and it's optional so don't stress if you're not using it 🙈

💖 DESCRIPTION 💖

ok so picture this: it closes up a stream for HW test — or, plot twist, if userspace failed to do so 😭 like sometimes userspace just doesn't do the homework and the kernel has to clean up after it. either way, this function is the one closing the stream, periodt.

🔥 SEE ALSO 🔥

the whole friend group lives in the kernel file ./drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c 📁 and here's everyone, literally the full roster:

that's the page bestie, go forth and destroy those msgs responsibly 💜

anyway signing off 🦋 — this one's straight out of ./drivers/gpu/drm/amd/amdgpu, dated October 2026, and yeah it's still uvd_v6_0_enc..._destroy_msg(9) in case you forgot the abbrev already 😭💀

📦 where this page comes from