_k_r2_001c2_gate.nx

_k_r2_001c2_gate.nx -- the K-R2-001c2 gate (virtio-NET MMIO QUEUE-CONFIGURATION REGISTER layer; the second slice of the K-R2-001c net virtqueue epic). It is the K-R2-001b1 queue-config gate applied to the SECOND virtio device, virtio-net (DeviceID=1) @ base 0x10002000. NO mocks: runs the REAL nx_virtio_hs_emit (the SAME emitter the blk b-series and the net c1 handshake use -- data-only on base/device-id/banner/queue-config, so it stays byte-for-byte backward-compatible), then RUNS the emitted image on the SOVEREIGN rv64 emulator (rv64im_min_sim + the second rv64im_min_virtio net instance @ 0x10002000, which already carries the legacy virtqueue-config registers QueueSel / QueueNumMax / QueueNum / QueueAlign / QueuePFN / QueueNotify resolved per-instance -- the PRIMARY, gating lane: Nishi owns the runtime) and asserts the captured serial transcript CONTAINS the emitter's golden ("VNET ACK DRV FEAT OK VQ\n" -- after the net handshake the driver selected a virtqueue, read QueueNumMax, latched the ring size, bound the guest ring page-frame via QueuePFN, kicked QueueNotify, and READ QueuePFN BACK to confirm the binding) AND the sovereign emu reports a clean SiFive-finisher halt. The blk device @ 0x10001000 is left BYTE-UNTOUCHED (its own gates VIRTGATE..SECTGATE still pass, re-proven by re-run). Then the ALIGNMENT lane: qemu-system-riscv64 -machine virt -global virtio-mmio.force-legacy=true with a REAL virtio-blk-device (first) + a REAL virtio-net-device (second). qemu fills its virtio-mmio slots in REVERSE, so the first device (blk) lands at the top slot 0x10008000 and the SECOND device (net) lands at the next slot DOWN, 0x10007000. The gate authors a SECOND net image from a base-rewritten spec (base = the qemu net reverse-slot 0x10007000) and runs THAT on qemu; the SAME queue-config driver completes against the REAL legacy virtio-net transport (QueuePFN read-back is genuinely RW on the real net device) and emits the SAME golden -> lanes AGREE. Finally a TAMPER test: corrupt the QueuePFN-EXPECTED constant in the sovereign image (the driver loads its own QueuePFN read-back into t3 and compares against the li-loaded expected in t4; the expected li is the addi following the unique `lw t3,0x40(t2)` read-back load). Bumping the addi's high immediate byte makes the read-back verify branch PAST stage 5 straight to the finisher -> the transcript loses ONLY its " VQ" canary (the handshake tail survives) -> the gate MUST go RED. Evidence -> knowledge/status/virtio_net.log (NETVQCFGGATE row; the queue row's ||MARK= reads it). Sovereign orchestration (fork/dup3/execve/wait4). license_tier: ORIGINAL