STMicroelectronics STiH407/STiH410 BDISP 2D blitter
A hardware 2D image blitter built into STMicroelectronics' STiH407 and STiH410 set-top-box system-on-chips from around 2013. It accelerates pixel-format conversion, scaling, and compositing for video pipelines on ST-powered cable, satellite, and IPTV boxes.
recommendation
Worth keeping but flagging as legacy because the underlying STi set-top-box silicon is no longer sold and the broader STi platform is being trimmed in 2025 (with STiH415/STiH416 and internal ST reference boards already on the way out). However, the BDISP driver itself still received non-trivial fixes in 2024-2025 and no removal series has been posted, and deployed set-top boxes likely still rely on it, so it should remain available while being documented as niche.
repository signals
sources
- cateee.net
LKDDb maps CONFIG_VIDEO_STI_BDISP to this directory and to OF compatible st,stih407-bdisp; the driver is still present in current kernel series listings.
- st.com
STiH407/STiH410 are 2013-era ST set-top-box / connected-client SoCs, indicating legacy consumer multimedia hardware rather than a current product family.
- lists.infradead.org
2025 STi cleanup discussion removes unsupported STiH415/STiH416 entries and internal B2120 board support, showing the broader STi platform is being trimmed as legacy, but not proposing bdisp driver removal.
codex reasoning notes (technical)
Real driver directory: local shell read of Kconfig identifies a V4L2 mem2mem 'STMicroelectronics BDISP 2D blitter driver' for ST SoCs. Lore MCP/lei were unavailable here, so lore history was approximated via web search plus direct open of a 2025 linux-arm-kernel cleanup thread; I found no bdisp-specific removal series, while the provided static metadata shows substantive non-mechanical fixes in 2024-2025, so this is not dead code. LKDDb page was obtained by web search/open; ST PDF was obtained by web search/open; the 2025 mailing-list thread was obtained by web search/open. Hardware appears tied to old STiH407/STiH410 set-top-box silicon and is very unlikely to be sold new in 2025, but deployed boxes may still exist in the field. Result: keep the driver, but annotate it as legacy/niche rather than deprecating or removing it.