Re: HEADSUP: cpufreq import complete, acpi_throttling changed

From: Morten Rodal <morten_at_rodal.no>
Date: Fri, 25 Feb 2005 21:55:10 +0100
On Thursday 24 February 2005 21:59, Nate Lawson wrote:
> Morten Rodal wrote:
> > On Wednesday 23 February 2005 20:24, Nate Lawson wrote:
> > The issue is partly fixed.  I now see the following frequency levels
> > if, and only if, I booted the laptop on battery power:
> >
> > # sysctl dev.cpu.0.freq_levels
> > dev.cpu.0.freq_levels: 1200/0 1050/0 900/0 787/0 750/0 656/0 600/0
> > 562/0 525/0 450/0 393/0 337/0 300/0 262/0 225/0 187/0 150/0 112/0
> > 75/0 37/0
> >
> > Note that the native processor speed (1700 MHz) is not listed, and
> > setting the speed below 150 hard-hangs/freezes the laptop.
>
> Not sure we can fix this one.  It appears some very low clock rates
> hang some systems.  Can you test with SCHED_4BSD?
>

This is with SCHED_4BSD.  I have been using that since it became the 
default again in -CURRENT.

> > When booting with AC-power the laptop does not report any
> > freq{,_levels}. It does not matter if I remove the AC-power once the
> > computer is up and running, I always get this:
> >
> > # sysctl dev.cpu.0.freq
> > dev.cpu.0.freq: -1
>
> I just committed a fix for this.

With sources from 18:56 CET I get this list:

# sysctl dev.cpu.0.freq_levels
dev.cpu.0.freq_levels: 1700/0 1487/0 1301/0 1275/0 1200/0 1115/0 1062/0 
956/0 920/0 900/0 850/0 796/0 750/0 663/0 637/0 600/0 557/0 531/0 478/0 
450/0 425/0 398/0 371/0 337/0 300/0 265/0 238/0 212/0 185/0 150/0 112/0 
79/0 53/0 26/0

If I boot the computer using battery power I get the same frequencies as 
posted earlier.  I did however remember one thing about the ACPI support 
on this Dell Inspiron (or maybe a bad hardware design), and that was 
people complained that the CPU was stuck at a maximum of 1200 MHz if it 
was started using battery power and then later have the power connected 
again.  So that might explain why I do not see 1700 MHz as a choice 
unless I start it with the AC-power connected.  Rebooting the computer 
does solve the problem, so it might just be a minor problem with the 
frequency levels only being read once (at boot).

-- 
Morten Rodal

"A supercomputer is a device for turning compute-bound
 problems into I/O bound problems." -- Ken Batcher (Goodyear Aerospace)


Received on Fri Feb 25 2005 - 19:55:22 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:38:29 UTC