# guest gating evidence (shared by DF-4407 .. DF-4411)

Collected live on the audit guest on 2026-09-19, immediately after this pass-2
run, to substantiate the `untested` gating classification (same methodology as
DF-4175: prove the surface is absent rather than assert it).

```
$ vm.sh status
up

$ vm.sh run_root 'uname -a'
DragonFly dfbsd 6.5-DEVELOPMENT DragonFly 6.5-DEVELOPMENT #0: Thu Jul  2 06:02:54 UTC 2026     root@dfbsd:/usr/obj/usr/src/sys/X86_64_GENERIC  x86_64

$ vm.sh run_root 'acpidump -t | grep -icE "ATK|ASUS"'        # ATK0100/ASUS010 nodes
0
$ vm.sh run_root 'acpidump | grep -iE "ATK0100|ASUS010"'     # (empty: no such HID anywhere)
(empty)
$ vm.sh run_root 'ls -l /dev/tpm*'                           # TPM device
ls: /dev/tpm*: No such file or directory
$ vm.sh run_root 'ls -l /dev/led'                            # LED consumers
ls: /dev/led: No such file or directory
$ vm.sh run_root 'sysctl -a | grep -i asus'                  # dev.acpi.N.asus tree
(empty)
$ vm.sh run_root 'kldstat -v | grep -iE "tpm|asus"'          # modules linked
(empty)
$ vm.sh run_root 'dmesg | grep -iE "tpm|asus|atk"'           # only atkbdc matches
atkbdc0: <Keyboard controller (i8042)> port 0x64,0x60 irq 1 on acpi0
atkbd0: <AT Keyboard> irq 1 on atkbdc0
```

Conclusions:

1. `sys/dev/acpica/acpi_asus/acpi_asus.c` never attaches on this guest (zero
   ATK0100/ASUS010 ACPI nodes), so none of DF-4408/4409/4410/4411 can be
   triggered at runtime here even as root. Triggering additionally requires
   either matching laptop firmware (DF-4408, DF-4411) or a root-initiated
   device detach / module unload (DF-4409, DF-4410).
2. `sys/dev/crypto/tpm/tpm.c` has no /dev/tpm and no `tpm` module loaded (the
   guest exposes no TPM TIS device; driver is `optional tpm`, not in
   X86_64_GENERIC), so DF-4407 cannot be exercised here; the defect is
   additionally gated on /dev/tpm being 0600 root:wheel (tpm.c:253-254), i.e.
   the racing processes must already be privileged.