Re: ACPI panic

On Wed, 7 Jan 2004, Vaidas Damosevicius wrote:
> On Tue, 6 Jan 2004, Nate Lawson wrote:
> > On Sat, 3 Jan 2004, Vaidas Damosevicius wrote:
> > > 	5.2-CURRENT (I made cvsup/kernel few minutes ago) panics on HP
> > > Compaq nx7000 with ACPI. Without ACPI everything works fine.
> > >
> > > kernel trap 12 with interrupts disabled
> > >
> > > Fatal trap 12: page fault while in kernel mode
> > > cpuid = 0; apic id = 00
> > > fault virtual address   = 0x6c
> > > fault code              = supervisor read, page not present
> > > instruction pointer     = 0x8:0xc0654f3d
> > > stack pointer           = 0x10:0xdfa587ec
> > > frame pointer           = 0x10:0xdfa587f0
> > > code segment            = base 0x0, limit 0xfffff, type 0x1b
> > >                         = DPL 0, pres 1, def32 1, gran 1
> > > processor eflags        = resume, IOPL = 0
> > > current process         = 7 (acpi_task1)
> > > trap number             = 12
> > > panic: page fault
> > > cpuid = 0;
> >
> > More information is needed.  Enable "options DDB" and type "tr" after
> > the panic to get a backtrace.
> >
> > This is likely going to show acpi_tz_monitor() in the callstack.  I can't
> > get anyone who can reproduce this regularly so if you can, that would help
> > me fix the root problem.
>
> Nate, I can't connect console to my notebook, so I made few screenshot
> with my problem :)

The backtrace actually shows this:

turnstile_setowner+0xd
turnstile_wait+0xe8
_mtx_lock_sleep
_mtx_lock_flags
EcSpaceHandler+0x82
...
AcpiEvaluateObject+0x125
acpi_cmbat_get_bst+0x81
acpi_cmbat_init_battery+0xb5
acpi_task_thread

The real clue is at the top of your screen:
device_probe_and_attach: acpi_ec0 attach returned 6

This means your embedded controller failed to probe and thus when the
battery tried to read data in EC space, it tried to lock a non-initialized
mutex (in the EC's softc) and paniced.

I really need the messages a few lines above the probe_and_attach line.
Also, send your ASL:  acpidump -t -d > vaidas-CompaqNX7000.asl

-Nate