sdma_v4_4_2_..._fill_buffer(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

πŸ’… sdma_v4_4_2_emit_fill_buffer β€” bestie, let's fill some buffers

okay so this is a page straight out of the Kernel API Manual and honestly? it's a little baddie. her whole job is filling GPU buffers using the sDMA engine and she does it flawlessly. no cap.


✨ NAME ✨

sdma_v4_4_2_emit_fill_buffer β€” fill buffer using the sDMA engine 🫢


🌸 SYNOPSIS 🌸

void sdma_v4_4_2_emit_fill_buffer (struct amdgpu_ib *ib , uint32_t
src_data , uint64_t dst_offset , uint32_t byte_count );

like… she takes four things and gives you nothing back (void, queen of not overexplaining πŸ’…).


πŸŽ€ ARGUMENTS πŸŽ€

  • ib β€” the indirect buffer we're copying to. this is the destination babe, the moment, the main character.
  • src_data β€” the actual value being written into the buffer. it's a uint32_t so keep it cute and in range ✨
  • dst_offset β€” the dst GPU address. that's where the magic lands, a uint64_t address bestie πŸ”₯
  • byte_count β€” number of bytes to xfer. yes "xfer". yes it's iconic. don't come for her πŸ™ˆ

πŸ¦‹ DESCRIPTION πŸ¦‹

Fill GPU buffers using the DMA engine.

that's it. that's the whole vibe. no extra lore, no dramatic backstory β€” she just fills GPU buffers with the DMA engine and leaves. lowkey iconic behavior tbh πŸ‘‘


🌈 SEE ALSO 🌈

okay these are her friends, her little group chat of kernel functions, all living in the Kernel file ./drivers/gpu/drm/amd/amdgpu/sdma_v4_4_2.c β€” go say hi fr fr:


πŸ’– source situation: this whole crew lives at ./drivers/gpu/drm/amd/amdgpu, and this manual page is dated October 2026 β€” mark your calendars bestie, she's got a whole month named after her energy πŸ—“οΈβœ¨

and that's the page!! slay, go fill those buffers, ily, mwah πŸŽ€πŸ’œπŸ§š

πŸ“¦ where this page comes from