docs(dma): record the upstream references behind the shared-memory direction

[avrabe]

FIND-DMA-SHM-CANONICAL-001 recorded the design reference (cpetig/wasm-shm-test) but not the question it answers — so the artifact could not say why the direction matters to us.

Added

Luke Wagner — "Does the Component Model Require Extra Copying?" (https://youtu.be/YZG7waS2XZg)

That question is not academic on an 8 KB part: a copy at the seam is a copy we cannot afford. It is precisely why own<dma-buffer> is specified as a transfer — attach/detach carry it — rather than a hand-off plus memcpy. Our contribution to it is an answer that is verified rather than argued: the Kani-proven ownership FSM, on silicon, in 220 B.

Timing. The Component Model 1.0 roadmap (Eric Gregory, 2026-06-08) places shared-everything threads after 1.0 — "these advancements also pave the road for shared-everything threads later." The general mechanism is future work; our specific verified instance exists now. That gap is the window in which contributing it is worth something.

Verification

Both references were checked against their sources rather than recalled:

• the roadmap article is by Eric Gregory, and contains no substantive shared-memory content beyond that one forward reference — worth knowing before citing it as a shared-memory source;
• the talk title and speaker were read from the video page.

rivet validate: PASS, 333 warnings, unchanged.

Related

The same two references now appear as the outlook on slide 12 of the Wasm Research Day deck (pulseengine.eu, branch talk/wasm-research-day-2026), which is where the question gets asked in front of the people who can answer it.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!