_k_r2_001b3_gate.nx

_k_r2_001b3_gate.nx -- the K-R2-001b3 gate (virtio-MMIO SECTOR-DATA ROUND-TRIP layer; the fifth slice of the K-R2-001b virtqueue epic, on top of b2c's status-byte writeback). Drives the full author->handshake->queue-config->descriptor-DMA->avail/used->status-byte->sector-data chain with NO mocks: runs the REAL nx_virtio_hs_emit (the team AUTHORS the rv64 image + the TABLE-COMPUTED golden transcript from the sector spec), then RUNS the image on the SOVEREIGN rv64 emulator (rv64im_min_sim + the rv64im_min_virtio device now, on the b2b QueueNotify kick, after the used-ring + status-byte writeback, FOLLOWING the consumed descriptor's data pointer (desc[head].addr) into guest RAM, DMA-READING the first 32-bit word of the data buffer the driver placed the sector pattern into + latching it into the QueueSectPeek 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 STAT SECT\n" -- after the used-ring + status writeback the device read the sector word the driver wrote, the driver READ QueueSectPeek BACK and verified it == sect_data_expected: the sector-data round-trip 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 + the SAME sector word into qemu guest RAM, and kicks the real QueueNotify -- the real legacy transport genuinely walks the ring. The sovereign-only QueueDescPeek / QueueUsedIdxPeek / QueueStatPeek / QueueSectPeek result registers are model instruments the real device does not expose (they read 0), so on qemu the descriptor-DMA + used-ring + status-byte + sector-data verifies gracefully drop " DESC" / " USED" / " STAT" / " SECT"; 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 + the sector word + kicks the real device -> lanes AGREE on every register the real transport can observe. Finally a TAMPER test: corrupt the QueueSectPeek-EXPECTED constant in the sovereign image (the driver loads its own QueueSectPeek read-back into t3 via lwu and compares against this li-loaded expected in t4; bumping the expected's immediate makes t3 != t4) -> the read-back verify branches PAST stage 9 straight to the finisher -> the transcript loses ONLY its " SECT" canary (the "VIO ACK DRV FEAT OK VQ DESC USED STAT" tail survives) -> the gate MUST go RED. Evidence -> knowledge/status/virtio_blk.log (SECTGATE row; the queue row's ||MARK= reads it). Sovereign orchestration (fork/dup3/execve/wait4). license_tier: ORIGINAL