Folks,

A whiles ago I made a patch that built a fake EST frequency table when
one couldn't be fetched from ACPI, but never got around to committing
it.  The lack of ACPI info seemed to particularly be a problem for Dell
laptops for some reason.  If you see something like this:

   cpu0: Enhanced SpeedStep (1404 mV) 1833 MHz
   cpu0: unknown Enhanced SpeedStep CPU.
   cpu0: using only highest and lowest power states.
   cpu0: Enhanced SpeedStep frequencies available (MHz): 1833 1000

in your boot messages, you're having the same issue.

This patch will create a "fake" frequency table with intermediate
values.  On my laptop, I now see:

   cpu0: Enhanced SpeedStep (1404 mV) 1833 MHz
   cpu0: unknown Enhanced SpeedStep CPU.
   cpu0: Enhanced SpeedStep frequencies available (MHz): 1833 1667 1500 1333 1167 1000

It's been tested by a few people now, without any major issues.  The
only issue I recall is that one CPU showed the same observed frequency
(measured using an openssl benchmark) for the two lowest frequency
settings.

I've put the patch at:

  ftp://ftp.NetBSD.org/pub/NetBSD/misc/simonb/est.c.diff

I'd like to get some more testers for this to make sure it doesn't
break anything.  I'll also clean up the patch (remove the "#if 0" code
and turn off the debugging by default) if all is well.  If I hear no
negative reports or other objections in the next week or so I'll commit
it.

Cheers,
Simon.