No usable event timer, FreeBSD Current, PCEngines WRAP

[Nigel Williams]

Hi Milan, Ian,

Thanks for the suggestions. After checking device.hints and removing 
acpi, I was still getting the same error. I decided to look at the 
nanoBSD conf file I was using, and it turned out the problem was in 
there. Here's the solution in case anyone else has the same problem.

I had included the line "WITHOUT_FORTH" in the nanBSD config file 
(specified with -c the_file.conf), which was preventing the loader from 
running device.hints on boot (and thus causing the panic later down the 
line). Removing this line fixed the issue.

However in the end I left WITHOUT_FORTH in the nanoBSD config and 
statically included the hints in the kernel config file with the line:
hints	"GENERIC.hints"

cheers,
nigel

On 05/06/13 01:43, Milan Obuch wrote:
> On Tue, 04 Jun 2013 09:21:26 -0600
> Ian Lepore <ian at FreeBSD.org> wrote:
>
>> On Tue, 2013-06-04 at 18:39 +1000, Nigel Williams wrote:
>>> Hi,
>>>
>>> I'm attempting to run FreeBSD-10 CURRENT (built using nanoBSD) on a
>>> PC-Engines WRAP device, and encounter the following during boot:
>>> "panic: No usable event timer found!" after which the machine
>>> reboots. I have been able to run pfSense 2.0.1 (bsd 8.1) and it
>>> appears to find the i8254 timer.
>>>
>>> The verbose boot output:
>>> https://docs.google.com/file/d/0B9KWjvUN8efvQ0RMNXRzRS1rS0k/edit?usp=sharing
>>>
>>> Kernel conf:
>>> https://docs.google.com/file/d/0B9KWjvUN8efvLUpRekJORWhmSFE/edit?usp=sharing
>>>
>>> nanoBSD conf:
>>> https://docs.google.com/file/d/0B9KWjvUN8efvTzIya3pNUXc5elk/edit?usp=sharing
>>>
>>> any suggestions?
>>
>> The problem does seem to be that the i8254 (atrtc) device isn't
>> instantiated.  I'm not sure whether the acpi errors at the start are
>> innocuous warnings or the cause of the problem (it has ever been thus
>> with acpi -- you report an error message on a mailing list and the
>> response is often "that's normal, you can ignore it").
>>
>> The atrtc device should be found via PNP data or /boot/device.hints if
>> not via acpi.  Perhaps acpi is working "enough" that it thinks it
>> should be using it, but not enough to find the timer.
>>
>> You could try "set hint.acpi.0.disabled=1" at the loader prompt and
>> see if that changes anything.  If it does, I'd call it more of a
>> workaround than a fix.
>>
>> -- Ian
>>
>
> WRAPs do not have acpi, so I have no 'device acpi' in kernel config for
> them. It is not necessary to have 'cpu I486_CPU' and 'cpu I686_CPU' in
> kernel config, by the way... I use small configuration
>
> cpu             I586_CPU
> options         CPU_GEODE
> ident           GEODE
> makeoptions     DEBUG=-g
> makeoptions     WITHOUT_MODULES="..."
> options         SCHED_ULE
> options         PREEMPTION
> options         INET
> options         FFS
> options         SOFTUPDATES
> options         UFS_ACL
> options         UFS_DIRHASH
> options         UFS_GJOURNAL
> options         QUOTA
> options         MSDOSFS
> options         CD9660
> options         PROCFS
> options         PSEUDOFS
> options         SCSI_DELAY=5000
> options         KTRACE
> options         STACK
> options         SYSVSHM
> options         SYSVMSG
> options         SYSVSEM
> options         _KPOSIX_PRIORITY_SCHEDULING
> options         PRINTF_BUFR_SIZE=128
> options         INCLUDE_CONFIG_FILE
> options         KDB
> options         KDB_TRACE
> options         SMP
> device          apic
> device          cpufreq
> device          pci
> device          ahci
> device          ata
> options         ATA_STATIC_ID
> device          scbus
> device          da
> device          pass
> device          pmtimer
> device          uart
> device          loop
> device          random
> device          ether
> device          md
> device          bpf
>
> and load if_sis from module (and some other modules too).
>
> What's in your /boot/device.hints file? I think following is relevant:
>
> hint.atrtc.0.at="isa"
> hint.atrtc.0.port="0x70"
> hint.atrtc.0.irq="8"
>
> At least when I comment these lines out, I get the same 'panic: no
> usable event timer found' error.
>
> Regards,
> Milan
>