# DF-5839 — second baseline panic (accidental rerun that doubles as a reproduction)
#
# First fix-validation attempt set loader.conf kernel="kernel.TEST/kernel";
# the loader silently ignored it and booted the ORIGINAL kernel (identical
# uname build stamp and identical lwkt_rele backtrace address prove it).
# The mode-1 trigger re-panicked the unpatched kernel the same way:
#
# serial console (vm.sh log):
df5839: loaded (watcher=cpu0 0xfffff800905fd480, spawner=cpu1 0xfffff800905fdb00)
df5839: watcher 0xfffff800905fd480 on cpu0
df5839: triggering mode 0
df5839: mode 0 finished, last_result=0 refs(victim)=-9
df5839: triggering mode 1
df5839: victim 0xfffff800905fe180 running on cpu1
df5839: victim 0xfffff800905fe180 exiting on cpu1
df5839: foreign hold on cached corpse 0xfffff800905fe180 from cpu0 (refs=0 freetd=0xfffff800905fe180)
df5839: spawner on cpu1 consuming gd_freetd=0xfffff800905fe180
df5839: child td=0xfffff800905fe180 (== victim corpse: YES)
panic: assertion "td->td_refs > 0" failed in lwkt_rele at /usr/src/sys/kern/lwkt_thread.c:505
cpuid = 0
lwkt_free_thread() at lwkt_free_thread 0xffffffff80672520
lwkt_free_thread() at lwkt_free_thread 0xffffffff80672520
rebirth_and_rele() at rebirth_and_rele+0xe6 0xffffffff82600316
watcher_func() at watcher_func+0x515 0xffffffff82600895
Debugger("panic")
#
# Note: identical backtrace addresses to the first baseline panic
# (0xffffffff80672520 / 0xffffffff82600316 / 0xffffffff82600895) — same
# kernel binary, same deterministic failure.  After this the boot was
# redone properly (kernel installed as /boot/kernel/kernel, verified via
# the new build stamp) and the same trigger completed cleanly
# (run.patched.log).