# env.txt — DF-4526/4528/4529 evidence environment

Guest (audit QEMU image):
  DragonFly dfbsd 6.5-DEVELOPMENT #0: Thu Jul 2 06:02:54 UTC 2026  x86_64
  kernel config: X86_64_GENERIC (unpatched audit source; INVARIANTS NOT
  enabled -> KKASSERT compiles to a no-op, sys/sys/systm.h:116-118)
  compiler (guest, userspace + kernels): gcc 8.3
  root fs: hammer2 on virtio-blk (vbd0s1d); /boot: ufs (vbd0s1a)

Hypervisor:
  QEMU 7.2.22 (Debian 1:7.2+dfsg-7+deb12u18b3), KVM, -smp 6 -m 4096, i440FX

NVMe test device (added via dfbsd-qemu/vm_nvme.sh; stock guest has none):
  -blockdev driver=null-co,node-name=dn0..dn7,size=10737418240 (x8)
  -device nvme,id=nv0,serial=AUDNVME,msix_qsize=65,max_ioqpairs=8
  -device nvme-ns,id=ns0..ns7,bus=nv0,drive=dn0..dn7
  attach log: nvme0: NVME Version 1.4 maxqe=2048 caps=004018200f0107ff,
  mapped 7 MSIX IRQs, Returns 8/6 queues (nominal 1:1 cpu map),
  8 disks /dev/nvme0../dev/nvme7 (root:operator 0640), nscount=256

ath (DF-4526 gate):
  pciconf -l: only hostb/isa/atapci/vgapci/virtio-net/virtio-blk present.
  NO Atheros NIC; QEMU has no ath HAL emulation. ATH_DIAGAPI not in
  X86_64_GENERIC (options ATH_DIAGAPI appears only in LINT64:1121).

Fixed kernels built in-guest:
  build-2 (fix v1): linked 05:27:51 2026-09-21, installed 05:38
  build-3 (fix v2): linked 05:59:48 2026-09-21, installed 06:07
  both verified by md5(/boot/kernel/kernel) == md5(kernel.debug) and by
  symbol relocation (nvme_pci_detach 0xffffffff80311920 -> 0xffffffff80311bc0)
  after an earlier stale-kernel boot was caught by exactly that check.

Guest returned to the clean-source snapshot (vm.sh reset with-src) at the
end of the run; rc.local rig and /root PoC binaries were removed by the
snapshot revert.