On Thu, 2001-12-20 at 12:33, Diefenbaugh, Paul S wrote:
> > i've noticed that when i'm on battery, the performance state will
> > stay at 100%, but if i am plugged in it gets turned down because
> > it gets hot.
> > so, if it's on battery, do i have to worry about it overheating?
> > is this a known bug?
> > this is with acpi 20011109 on 2.4.17-rc1-xfs (this is the most
> > recent acpi which works on my vaio)
> Sounds like the BIOS is automatically shifting into a higher performance
> state (non-throttling) when plugging into AC, and downshifting when on
> battery. The processor driver currently (only) displays the throttling
> state -- thus 100% just means that we're not throttling -- although the Mhz
> could significantly different between AC and battery.
ok, i actually misspoke in my earlier mail, and meant something slightly
different than what i actually said.
since then i've been compiling some things and watching the temperature,
cpu speed, and performance state while plugged into ac and while on
(because of the un-throttling bug, i have built my kernel with a patch
poasted to this list to let me call pr_perf_set_state() when i cat to
for reference, here is what my laptop does at various states (hopefully
i got the S# terminology correct)
state speed (ac) speed (batt)
S0 750 600
S1 640 510
S2 550 440
basically, what i am seeing is that if i run it at AC S2, it will hit
the passive cool trip point, but if i run it at BATT S0 (faster cpu
speed), it doesn't.
does this make sense? is the temperature reporting incorrectly, or is
there something else going on which would heat it up more?
"In fact, can you imagine anything more terrifying than a zombie clown?"