Fix-build logs (kernel #1 = fix-v1, kernel #2 = fix-v2)
=======================================================

Both /tmp/kbuild*.log files were cleared by the post-install reboots
before they could be fetched; the decisive completion lines captured in
the session transcript:

kernel #1 (fix-v1, cache_lladdr guard only):
  --------------------------------------------------------------
  >>> Kernel build for X86_64_GENERIC completed on Thu Sep 10 23:24:55 UTC 2026
  --------------------------------------------------------------
  >>> Kernel install for X86_64_GENERIC completed on Thu Sep 10 23:39:42 UTC 2026
  boot: DragonFly dfbsd 6.5-DEVELOPMENT #1: Thu Sep 10 23:09:23 UTC 2026

kernel #2 (fix-v2, 3 layers: timer NOSTATE case + RESOLVE grace + guard):
  --------------------------------------------------------------
  >>> Kernel build for X86_64_GENERIC completed on Fri Sep 11 00:11:13 UTC 2026
  --------------------------------------------------------------
  >>> Kernel install for X86_64_GENERIC completed on Fri Sep 11 00:12:32 UTC 2026
  boot: DragonFly dfbsd 6.5-DEVELOPMENT #2: Thu Sep 10 23:55:24 UTC 2026

Patch application (v2, dry-run + real):
  Hmm...  Looks like a unified diff to me...
  Patching file nd6.c using Plan A...
  Hunk #1 succeeded at 497.        (nd6_timer_dispatch NOSTATE case)
  Hunk #2 succeeded at 1324.       (RTM_RESOLVE grace: time_uptime + nd6_delay)
  Hunk #3 succeeded at 1725.       (cache_lladdr no-lladdr guard, DF-3331)
  done
  grep -c DF-3331 nd6.c -> 3

Both kernels booted and ran the PoC; see fix_run.log (#1: leak NOT
closed -> mechanism discovery) and fix_v2_run.log (#2: leak closed,
functionality preserved).