Re: panic: Assertion pcb->pcb_fpusaved == NULL failed at /usr/src/sys/arm64/arm64/vfp.c:209

On Wed, 3 Apr 2024, Bjoern A. Zeeb wrote:

> Hi,
>
> I had a hang hard-er hang when loading a LinuxKPI based wireless driver for a
> card probably in D3 state.  Probing as running through three times (but 
> there's
> a lot more PCIe cards ther, I didn't say the card being detectede).
>
> ^T would not work at this point.
>
> I could break into the debugger thoug, did so and typed cont waiting a bit 
> longer
> and upon trying to get into debugger again this happened:


Ok, I also just go this when manually trying to enter ddb.

The unhelpful bis is that I do not have KDB_UNATTENDED set but yet it
goes into automatic reboot.

After reboot I did check:
# sysctl debug.debugger_on_panic
debug.debugger_on_panic: 1

Should I file a PR for this?


> KDB: enter: Break to debugger
> timeout stopping cpus
> [ thread pid 9 tid 100240 ]
> Stopped at      kdb_alt_break_internal+0x180:   str     xzr, [x19, #1408]
> db> timeout stopping cpus
> panic: Assertion pcb->pcb_fpusaved == NULL failed at 
> /usr/src/sys/arm64/arm64/vfp.c:209
> cpuid = 0
> time = 1712167778
> KDB: stack backtrace:
> db_trace_self() at db_trace_self
> db_trace_self_wrapper() at db_trace_self_wrapper+0x38
> vpanic() at vpanic+0x1a4
> panic() at panic+0x48
> vfp_save_state_savectx() at vfp_save_state_savectx+0x48
> savectx() at savectx+0x38
> arm_gic_v3_intr() at arm_gic_v3_intr+0xe4
> intr_irq_handler() at intr_irq_handler+0x88
> handle_el1h_irq() at handle_el1h_irq+0x14
> --- interrupt
> lock_delay() at lock_delay+0x18
> __mtx_lock_sleep() at __mtx_lock_sleep+0xdc
> __mtx_lock_flags() at __mtx_lock_flags+0x138
> softclock_call_cc() at softclock_call_cc+0xd0
> softclock_thread() at softclock_thread+0xb8
> fork_exit() at fork_exit+0x78
> fork_trampoline() at fork_trampoline+0x18
> Uptime: 16m57s
> Automatic reboot in 15 seconds - press a key on the console to abort
>
>
>
>
>

-- 
Bjoern A. Zeeb                                                     r15:7