A scratchpad for random thoughts re spudstop.
--------------------------------------------------------------

On an APM kernel, msrs 0x2a -> 0x31 change when going from
battery to mains and vice versa.
mains @ 700MHz = 0x46480020
battery @500Mhz= 0x45080020

Reading MHz on an ACPI kernel gets us wild results, anywhere
between 2MHz and 266MHz. The reason for this is that the
ACPI code does thermal throttling unrelated to Speedstep.

According to Andy Grover, Speedstep only has 2 states.
Need to dump the above MSRs on more speedstep capable laptops.
(Of differing clockrates).

-------------------------------------------------------------
Danb has a Toshiba Portege which has an APM implementation
of speedstep working as it should. (Well, mostly).

Need to find the int 15h entry point from zero page in 16bit
mode, and then disasm that section of the BIOS. Fun.
-------------------------------------------------------------

