AMD DCE 8 display engine support (Sea Islands / Bonaire-era Radeon)
Display Controller Engine version 8, the on-GPU output block found in AMD Radeon GPUs and APUs from roughly 2013 onward — Bonaire, Hawaii, Kaveri, and similar Sea Islands / Volcanic Islands parts. It drives the monitors attached to those cards: setting modes, routing pixels to HDMI/DisplayPort/DVI outputs, and handling audio over HDMI.
recommendation
Worth keeping but flagging as legacy, because the GPUs it drives stopped being sold years ago yet are still in plenty of working desktops and HTPCs. The code is not abandoned: AMD developers were still touching this directory in 2024 and 2025 as part of broader display-stack work, and it sits behind the same pre-Vega support option that also covers Carrizo and Tonga. No removal proposal is visible, so a short note explaining its legacy scope is the right call rather than deprecation.
repository signals
sources
- git.kernel.org
The directory received a substantive upstream touch on 2025-04-29/30, indicating it is still maintained rather than abandoned.
- git.kernel.org
The directory was part of a 2024 functional display change, showing continued integration work in current AMD display development.
- cateee.net
AMD DC support for pre-Vega hardware remains present upstream; the Kconfig help text explicitly covers older ASICs such as Carrizo and Tonga.
- en.wikipedia.org
Volcanic Islands-era Radeon products were introduced in late 2013, placing this hardware family firmly in legacy-product age by 2025.
codex reasoning notes (technical)
Local file inspection via functions.exec_command shows this is real AMD display-engine hardware code, not a helper library. Local git log via functions.exec_command found substantive directory touches in 2024-2025; canonical git.kernel.org commit URLs were constructed from those observed hashes. Web search found LKDDb PRE_VEGA Kconfig evidence and the Radeon page. Local rg into dal_asic_id ties DCE80 to older AMD families (Volcanic Islands/Carrizo-era naming nearby), and taken together with the 2013-era product generation evidence this points to legacy installed-base use rather than new 2025 sales. No removal discussion was found in available evidence, so active maintenance argues against deprecate/remove; annotate legacy scope instead.