DRM_CAP_DUMB_PREFER_SHADOW(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

πŸ’… DRM_CAP_DUMB_PREFER_SHADOW(9) πŸ’…

okay so this is a page straight outta the Kernel API Manual, and it's about one very specific lil capability flag. bestie, it's short, it's sweet, and it's giving "one job and it does it." let's get into it 🫢


✨ NAME ✨

DRM_CAP_DUMB_PREFER_SHADOW β€” and honestly? the original page leaves the description blank after the dash 😭 which is so mysterious of it. no one-liner, no vibes explained, just the name sitting there with main character energy. we move.


🌸 SYNOPSIS 🌸

DRM_CAP_DUMB_PREFER_SHADOW

that's it. that's the whole synopsis. slay for minimalism ✨


πŸŽ€ DESCRIPTION πŸŽ€

so here's the tea πŸ”₯ β€” if this gets set to 1, the driver is basically saying it prefers userspace to render to a shadow buffer instead of directly rendering to a dumb buffer.

and for best speed? userspace should do streaming ordered memory copies into the dumb buffer and never read from it. like, write-only, no peeking, periodt πŸ’€ the buffer is not your diary, don't be reading it back.

🚨 note note note: this preference only applies to dumb buffers. it's irrelevant for other types of buffers. so don't go applying this logic everywhere, iykyk πŸ™ˆ

πŸ¦‹ SEE ALSO πŸ¦‹

the full guest list, no gatekeeping πŸ’– (copying every single one because we don't skip the boring parts in this house):


πŸ’– and that's the whole page bestie πŸ’– sourced from ./tools/perf/trace/beauty/incl... (yes it's cut off, that's literally how it shows up, we respect the truncation 🫑), dated October 2026, and it's DRM_CAP_DUMB_PREFER_SHADOW(9) forever and always. go render to that shadow buffer, drink some water, and don't read back from your dumb buffer πŸ’…βœ¨πŸŽ€

πŸ“¦ where this page comes from