[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Pentium-M

On Tue, Nov 04, 2003 at 01:08:34PM +0100, Karl-Heinz Eischer wrote:
> Hi Magnus,
> On Tue, Nov 04, 2003 at 12:17:34PM +0100, Magnus von Koeller wrote:
> > You're right, this was due to cpufreqd changing it back. If I disable 
> > that, it stays at the 600Mhz-600Mhz range. But still, /proc/cpuinfo 
> > always gives 
> > cpu MHz         : 1294.774
> > 
> > So, it's not working, is it?
>  You could try to guess your actual cpu freqency with the program
> bogomips. You have to keep in mind, that the result is in some cases
> (all Intel and AMD Pentium variations) is only half of the one in
> cpuinfo (for further info look into the BogoMips mini howto).
> For this CPUs the output is nearly the clock frequency. But im not sure
> in which category of CPUs the Centorino is falling. Just try it with the
> 600MHz and 1300MHz settings and see if the value differs.

cpu MHz in cpuinfo is updated by cpufreq core.
So that it is not a good indication of the real frequency if cat'ing
/proc/cpufreq give what the poster said.  It is more likely a bug in
cpufreq.  AFAIK, there were bug, and one still remain in case of
SMP, but UP should be OK.  Note also that in case of invasive patch like
aa, or preempt, for example, it may be possible that cpuinfo
entry may give bad infos.

Ducrot Bruno

--  Which is worse:  ignorance or apathy?
--  Don't know.  Don't care.

Reply to: