acpi: bad [read from | write to] port 0x086

Hi all,

Robert Watson's call for testing of his network code prompted me to
upgrade my current, from one of january 29 to one of today:
-bash-2.05b$ uname -a
FreeBSD prolepsis.urh.uiuc.edu 7.0-CURRENT FreeBSD 7.0-CURRENT #21:
Mon Apr  3 13:35:05 UTC 2006    
kaduk_at_prolepsis.urh.uiuc.edu:/usr/obj/usr/src/sys/PROLEPSIS  i386

In the process, I managed to install world before installing the
kernel, since I misread the build log (which had errored out due to a
lack of audit group), but afterwards I installed the new kernel, then
re-cvsuped (to get a few new changes), and rebuilt world and kernel
via the standard procedure, with no errors.  Since the entire (second)
sequence finished without error, I am operating on the assumption that
this is not a result of me shooting myself in the foot.

After installing world of a few days ago, both new and old kernels
will spew out spurts of the following on the console:
acpi: bad read from port 0x086 (8)
acpi: bad write to port 0x086 (8), val 0
acpi: bad read from port 0x086 (8)
acpi: bad write to port 0x086 (8), val 0
acpi: bad read from port 0x086 (8)
acpi: bad write to port 0x086 (8), val 0
acpi: bad read from port 0x086 (8)
acpi: bad write to port 0x086 (8), val 0
acpi: bad read from port 0x086 (8)
acpi: bad write to port 0x086 (8), val 0
acpi: bad read from port 0x086 (8)
acpi: bad write to port 0x086 (8), val 0
acpi: bad read from port 0x086 (8)
acpi: bad write to port 0x086 (8), val 0
acpi: bad read from port 0x086 (8)
acpi: bad write to port 0x086 (8), val 0
acpi: bad read from port 0x086 (8)
acpi: bad write to port 0x086 (8), val 0
acpi: bad read from port 0x086 (8)
acpi: bad write to port 0x086 (8), val 0
acpi: bad read from port 0x086 (8)
...

I seem to recall there being an off-by-one error for someone else, but
that was port 83, IIRC (I can't seem to find the commit message for
that fix at the moment).

Does anyone have any thoughts about this?

An incomplete (non-verbose) dmesg and pciconf may be found at:
https://netfiles.uiuc.edu/kaduk/www/prolepsis/
(sorry that it thinks the dmesg is a binary file)

Thanks,

Ben Kaduk