dc_override_...idth_request(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

πŸ’… dc_override_memory_bandwidth_request β€” the kernel API manual, but make it make sense

okay bestie, welcome to the dc_override_memory_bandwidth_request page β€” yes that's the full name, the header just clipped it to dc_override_...idth_request(9) because terminals have no chill. this is the Kernel API Manual, and we're about to read it together like it's the group chat. 🫢


✨ NAME ✨

dc_override_memory_bandwidth_request β€” overrides the DCN nominal memory bandwidth request that gets sent to PMFW, and it does this independent of whatever the current display mode is. like, it does not care what mode you're in, it's doing its own thing. πŸ’…

genuinely for debug use only. that's not me being dramatic, that's literally what the page says. it's giving "do not ship this" energy. πŸ™ˆ

🌸 SYNOPSIS 🌸

unsigned int dc_override_memory_bandwidth_request (struct dc *dc ,
       unsigned int bw_mbps );

two args, returns an unsigned int. that's her. that's the whole function. πŸŽ€


πŸ¦‹ ARGUMENTS πŸ¦‹

  • dc β€” the DC instance. the main character, the context, the whole reason we're here. πŸ‘‘
  • bw_mbps β€” the requested bandwidth in MB/s. and bestie, listen: 0 clears the override. zero is not a valid bandwidth here, zero is the "nvm, undo it" button. ⭐

πŸ”₯ RETURN πŸ”₯

the capped bandwidth value that was actually applied, in MB/s. so like β€” you asked for a number, and it tells you the number you got after the cap. no cap literally. πŸ˜­πŸ’€


🌈 SEE ALSO 🌈

the whole friend group lives in the kernel file ./drivers/gpu/drm/amd/display/dc/dc.h β€” go say hi to:

that's the entire squad, nobody's left out, we don't do that here. πŸ’–


and that's the page, diva. this iconic little number lives in ./drivers/gpu/drm/amd/display/dc, it's dated October 2026, and it's filed as dc_override_...idth_request(9) because even the manpage index gave up on the full name. slay. πŸ’…βœ¨

πŸ“¦ where this page comes from