* cpu won't stay in powersave mode
@ 2003-12-15 1:17 Tom Bruno
0 siblings, 0 replies; 4+ messages in thread
From: Tom Bruno @ 2003-12-15 1:17 UTC (permalink / raw)
To: cpufreq
I'm in the process of writing an acpi daemon that will also take care of
cpufreq settings.
I have found that unless I set
/sys/device/system/cpu/cpu0/cpfreq/scaling_governer
to powersave from the command line that /proc/cpuinfo will not change
and the speed will not stay at powersave speeds.
With other programs like cpufreqd, using a cpu speed test posted on this
list, i have found that the cpu speed does indeed change even though
/proc/cpuinfo does not.
However, i have found that after about 8-10 seconds the cpu test starts
reporting that the cpu is back to full speed even though the powersave
value is still in scaling_governer.
Like i said above if I change it via hand using echo, /proc/cpuinfo is
correct, and the cputest reports the correct speed constantly.
performance mode: 2.4ghz
powersave mode: 1.2ghz
Intel P4-M 2.4ghz
Dell Inspiron 8500 - Bios A05
kernel-2.6.0-test11
--
Vwords.com Technology Services
PC and Network Services for Southern Illinois
Phone: 618-713-6896
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: cpu won't stay in powersave mode
@ 2003-12-15 15:18 Tom Bruno
2003-12-19 16:15 ` Dominik Brodowski
0 siblings, 1 reply; 4+ messages in thread
From: Tom Bruno @ 2003-12-15 15:18 UTC (permalink / raw)
To: cpufreq
Sorry, accendtly hit reply instead of send to list.
On Mon, 2003-12-15 at 14:14, Thomas Renninger wrote:
> Tom Bruno wrote:
>
> >Like i said above if I change it via hand using echo, /proc/cpuinfo
is
> >correct, and the cputest reports the correct speed constantly.
> >
> >performance mode: 2.4ghz
> >powersave mode: 1.2ghz
> >Intel P4-M 2.4ghz
> >Dell Inspiron 8500 - Bios A05
> >kernel-2.6.0-test11
> >
> >
> /proc/cpu_info didn't work for me, too.
> However it seems to work now(after I changed the governors to be
> compiled as modules(or the way around?)).
>
I've now tried both as modules, and compiled in, /proc/cpuinfo works
niether report correctly.
> If you want to detemine the current real speed of your machine use the
> procspeed prog I posted some days ago(04.12, 09.12).
yes, that is the program i spoke of. However, like i said, when
/proc/cpuinfo is reporting wrong, my system works it's way back to full
speed even though powersave is in the /sys scaling_governor.
It will report the changed speed (clockdown to 1.2ghz) for about 10
seconds, and then will start to report again at 2.4ghz even though the
setting haddn't changed.
--
Vwords.com Technology Services
PC and Network Services for Southern Illinois
Phone: 618-713-6896
--
Vwords.com Technology Services
PC and Network Services for Southern Illinois
Phone: 618-713-6896
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: cpu won't stay in powersave mode
2003-12-15 15:18 Tom Bruno
@ 2003-12-19 16:15 ` Dominik Brodowski
2003-12-20 17:21 ` Tom Bruno
0 siblings, 1 reply; 4+ messages in thread
From: Dominik Brodowski @ 2003-12-19 16:15 UTC (permalink / raw)
To: Tom Bruno; +Cc: cpufreq
[-- Attachment #1.1: Type: text/plain, Size: 1602 bytes --]
On Mon, Dec 15, 2003 at 03:18:04PM +0000, Tom Bruno wrote:
> Sorry, accendtly hit reply instead of send to list.
>
> On Mon, 2003-12-15 at 14:14, Thomas Renninger wrote:
> > Tom Bruno wrote:
> >
> > >Like i said above if I change it via hand using echo, /proc/cpuinfo
> is
> > >correct, and the cputest reports the correct speed constantly.
> > >
> > >performance mode: 2.4ghz
> > >powersave mode: 1.2ghz
> > >Intel P4-M 2.4ghz
> > >Dell Inspiron 8500 - Bios A05
> > >kernel-2.6.0-test11
> > >
> > >
> > /proc/cpu_info didn't work for me, too.
> > However it seems to work now(after I changed the governors to be
> > compiled as modules(or the way around?)).
> >
> I've now tried both as modules, and compiled in, /proc/cpuinfo works
> niether report correctly.
>
> > If you want to detemine the current real speed of your machine use the
> > procspeed prog I posted some days ago(04.12, 09.12).
>
> yes, that is the program i spoke of. However, like i said, when
> /proc/cpuinfo is reporting wrong, my system works it's way back to full
> speed even though powersave is in the /sys scaling_governor.
>
> It will report the changed speed (clockdown to 1.2ghz) for about 10
> seconds, and then will start to report again at 2.4ghz even though the
> setting haddn't changed.
I think the BIOS is meddling with the speedstep setting [this is more
common for older systems, but obviously some vendors still think it's a good
idea] . If it were cpufreq, the value in /proc/cpuinfo would be updated as
well.
Do you run an ACPI-enabled kernel?
Dominik
[-- Attachment #1.2: Type: application/pgp-signature, Size: 189 bytes --]
[-- Attachment #2: Type: text/plain, Size: 143 bytes --]
_______________________________________________
Cpufreq mailing list
Cpufreq@www.linux.org.uk
http://www.linux.org.uk/mailman/listinfo/cpufreq
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: cpu won't stay in powersave mode
2003-12-19 16:15 ` Dominik Brodowski
@ 2003-12-20 17:21 ` Tom Bruno
0 siblings, 0 replies; 4+ messages in thread
From: Tom Bruno @ 2003-12-20 17:21 UTC (permalink / raw)
To: Dominik Brodowski; +Cc: cpufreq
[-- Attachment #1.1: Type: text/plain, Size: 2074 bytes --]
On Fri, 2003-12-19 at 16:15, Dominik Brodowski wrote:
> On Mon, Dec 15, 2003 at 03:18:04PM +0000, Tom Bruno wrote:
> > Sorry, accendtly hit reply instead of send to list.
> >
> > On Mon, 2003-12-15 at 14:14, Thomas Renninger wrote:
> > > Tom Bruno wrote:
> > >
> > > >Like i said above if I change it via hand using echo, /proc/cpuinfo
> > is
> > > >correct, and the cputest reports the correct speed constantly.
> > > >
> > > >performance mode: 2.4ghz
> > > >powersave mode: 1.2ghz
> > > >Intel P4-M 2.4ghz
> > > >Dell Inspiron 8500 - Bios A05
> > > >kernel-2.6.0-test11
> > > >
> > > >
> > > /proc/cpu_info didn't work for me, too.
> > > However it seems to work now(after I changed the governors to be
> > > compiled as modules(or the way around?)).
> > >
> > I've now tried both as modules, and compiled in, /proc/cpuinfo works
> > niether report correctly.
> >
> > > If you want to detemine the current real speed of your machine use the
> > > procspeed prog I posted some days ago(04.12, 09.12).
> >
> > yes, that is the program i spoke of. However, like i said, when
> > /proc/cpuinfo is reporting wrong, my system works it's way back to full
> > speed even though powersave is in the /sys scaling_governor.
> >
> > It will report the changed speed (clockdown to 1.2ghz) for about 10
> > seconds, and then will start to report again at 2.4ghz even though the
> > setting haddn't changed.
>
> I think the BIOS is meddling with the speedstep setting [this is more
> common for older systems, but obviously some vendors still think it's a good
> idea] . If it were cpufreq, the value in /proc/cpuinfo would be updated as
> well.
>
> Do you run an ACPI-enabled kernel?
>
> Dominik
Yes, I have acpi enabled, and to the best of my knowledge, it is
functioning correclty, at least "externally" /proc/acpi/events works
and returns events and stuff. Don't know about non visiable internals
though.
New stats:
Dell Inspiron 8500
Bios: A05
Cpu: Pentium-M 2.4ghz
Gentoo Linux
Kernel: 2.4.23, 2.6.0
[-- Attachment #1.2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]
[-- Attachment #2: Type: text/plain, Size: 143 bytes --]
_______________________________________________
Cpufreq mailing list
Cpufreq@www.linux.org.uk
http://www.linux.org.uk/mailman/listinfo/cpufreq
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2003-12-20 17:21 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2003-12-15 1:17 cpu won't stay in powersave mode Tom Bruno
-- strict thread matches above, loose matches on Subject: below --
2003-12-15 15:18 Tom Bruno
2003-12-19 16:15 ` Dominik Brodowski
2003-12-20 17:21 ` Tom Bruno
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.