_k_r2_001b2b_gate.nx

_k_r2_001b2b_gate.nx -- the K-R2-001b2b gate (virtio-MMIO AVAIL-CONSUME + USED-RING WRITEBACK layer; the third slice of the K-R2-001b virtqueue epic, on top of b2a's descriptor-DMA). Drives the full author->handshake->queue-config->descriptor-DMA-> avail/used chain with NO mocks: runs the REAL nx_virtio_hs_emit (the team AUTHORS the rv64 image + the TABLE-COMPUTED golden transcript from the avail/used spec), then RUNS the image on the SOVEREIGN rv64 emulator (rv64im_min_sim + the rv64im_min_virtio device now, on the SECOND QueueNotify kick, reading the avail ring head, walking the descriptor (b2a DMA), WRITING the used ring (used.ring[0].id/.len) + BUMPING used.idx back into guest RAM, and latching used.idx into the QueueUsedIdxPeek RO register -- the PRIMARY, gating lane: Nishi owns the runtime) and asserts the captured serial transcript CONTAINS the emitter's golden ("VIO ACK DRV FEAT OK VQ DESC USED\n" -- after the descriptor-DMA the driver published the avail ring, kicked QueueNotify, the device consumed it + wrote the used ring, then the driver READ QueueUsedIdxPeek BACK and verified it == used_idx_expected: the used-ring writeback binding proof) AND the sovereign emu reports a clean SiFive- finisher halt. Then the ALIGNMENT lane: qemu-system-riscv64 -machine virt -global virtio-mmio.force-legacy=true with a REAL virtio-blk-device backing. The gate authors a SECOND image from a base-rewritten spec (the blk device lands at slot 0x10008000) and runs the SAME driver against the REAL legacy virtio-blk transport: it completes the handshake + queue config, lays the SAME descriptor + the SAME avail ring into qemu guest RAM, and kicks the real QueueNotify -- the real legacy transport genuinely walks the ring. The QueueUsedIdxPeek result register is a sovereign-model instrument the real device does not expose (it reads 0), so on qemu the used-ring read-back verify gracefully drops ONLY " USED" (and the sovereign-only QueueDescPeek already drops " DESC"); the gate asserts the qemu transcript contains the b1 prefix golden ("VIO ACK DRV FEAT OK VQ") -- proving the SAME driver lays the avail ring + kicks the real device -> lanes AGREE on every register the real transport can observe. Finally a TAMPER test: corrupt the QueueUsedIdxPeek-EXPECTED constant in the sovereign image (the driver loads its own QueueUsedIdxPeek read-back into t3 via lw and compares against this li-loaded expected in t4; bumping the expected's immediate makes t3 != t4) -> the read-back verify branches PAST stage 7 straight to the finisher -> the transcript loses ONLY its " USED" canary (the "VIO ACK DRV FEAT OK VQ DESC" tail survives) -> the gate MUST go RED. Evidence -> knowledge/status/virtio_blk.log (USEDGATE row; the queue row's ||MARK= reads it). Sovereign orchestration (fork/dup3/execve/wait4). license_tier: ORIGINAL