[Bug 292895] Reproducible ARM64/bhyve panic on CIX/Cortex-A720 platforms

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=292895

--- Comment #17 from Michael Dexter <editor@callfortesting.org> ---
Revisiting this on with the weekly snapshot: FreeBSD 16.0-CURRENT
main-n288280-a8347f442824 GENERIC

With a FreeBSD VM of the same snapshot:
Freezes the host on VM boot at this stage in the VM: random: unblocking device.

Debian guest:
Freeze during VM boot:
...
Freeing initrd memory: 35240K
Unhandled register access: pc 0xffff80008038e038 syndrome 0x1222264 reg 19
Abort trap

Host is responsive, but freezes on shutdown at:
Writing early boot entropy file: .

During second Debian VM boot, FreeBSD host panicked with:

Panic String: spin lock held too long
Dump Parity: 3606026849

vgic0: <Virtual GIC v3> on gic0
<6>pid 8959 (bhyve), jid 0, uid 0: exited on signal 6 (no core dump - other
error)
spin lock 0xffff000001172de0 (smp rendezvous) held by 0xffff00019a2a9640 (tid
101132) too long
KDB: stack backtrace:
db_trace_self() at db_trace_self
db_trace_self_wrapper() at db_trace_self_wrapper+0x44
vpanic() at vpanic+0x1c8
panic() at panic+0x48
_mtx_lock_indefinite_check() at _mtx_lock_indefinite_check+0xb0
_mtx_lock_spin_cookie() at _mtx_lock_spin_cookie+0xc4
__mtx_lock_spin_flags() at __mtx_lock_spin_flags+0x11c
smp_rendezvous_cpus() at smp_rendezvous_cpus+0x1b4
dtrace_state_deadman() at dtrace_state_deadman+0x88
softclock_call_cc() at softclock_call_cc+0x180
softclock_thread() at softclock_thread+0xbc
fork_exit() at fork_exit+0x80
fork_trampoline() at fork_trampoline+0x18
KDB: enter: panic

I'll work on getting the full text dump.

-- 
You are receiving this mail because:
You are the assignee for the bug.