On 05/24/2011 11:17 AM, A. Kromic wrote:
> On 05/24/2011 07:34 PM, NoOp wrote:
...
>> You mention in your bug report "works perfectly with kernels up to
>> 2.6.32-24". But your last post: "Still not working with latest release
>> kernels: 2.6.32-32 (Lucid)". So I'm confused - with which kernel *is*
>> cpu scaling currently working on your machine?
>>
>>
> What I meant is this: my system's CPU scaling has worked and still works
> perfectly with all old kernels up to 2.6.32-24, with any kernel newer
> than that it doesn't work at all. I can reproduce it anytime, either
> chosing a kernel from GRUB, or by booting a live CD. Another fact
> perhaps worth noting is that it doesn't matter if it is a 32- or 64-bit
> Ubuntu - I've checked with both, and again the problem depends just on
> the kernel version.
>
> I don't quite get what the problem is, it seems it's quite rare. However
> so far, I couldn't find anything else but the kernel to blame, per the
> above mentioned fact. No amount of tinkering with other software (like
> acpi, cpu frequency managers etc.) not hardware (battery, BIOS options)
> showed any difference at all...
...
There were issues in earlier kernels on archlinux as well. Might be of
interest (read both pages):
<https://bbs.archlinux.org/viewtopic.php?id=46195>
In that posting Thomas Renninger did a patch:
cpufreq acpi: only call _PPC after cpufreq ACPI init funcs got called
already
So perhaps something like that has raised it's head again in the newer
kernels.

You might want to have a look at:
https://wiki.ubuntu.com/Kernel/FAQ
https://wiki.ubuntu.com/Kernel/Debugging
to see if there are tools etc., there can can be of help in figuring out
the issue. Also ask on #ubuntu-kernel IRC channel on FreeNode & the
kernel mailing list to see if someone there can offer further suggestions.

Good luck as I suspect this may be a tough one to figure out.






--
ubuntu-users mailing list
ubuntu-users@lists.ubuntu.com
Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-users

On 25 May 2011 15:43, A. Kromic <akromic@gmail.com> wrote:
> I refuse to do futile tests. I'm an IT pro, didn't start with computers
> and linux yesterday, and most certainly won't blindly try every
> suggestion for which I know how it works and why it's not rellevant for
> me. NHF.
Right, I earn my living by selling ice cream to small kids, just like you.
I haven't seen any evidence of your "extensive testing" neither here
nor the irrelevant launchpad bug you posted a link ti.
Good luck with the debugging. Just don't expect help from others with no info.
--
Hakan (m1fcj) - http://www.hititgunesi.org

--
ubuntu-users mailing list
ubuntu-users@lists.ubuntu.com
Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-users

On 05/25/2011 01:53 PM, Hakan Koseoglu wrote:

On 25 May 2011 15:43, A. Kromic<akromic@gmail.com> wrote:

I refuse to do futile tests. I'm an IT pro, didn't start with computers
and linux yesterday, and most certainly won't blindly try every
suggestion for which I know how it works and why it's not rellevant for
me. NHF.

Right, I earn my living by selling ice cream to small kids, just like you.
I haven't seen any evidence of your "extensive testing" neither here
nor the irrelevant launchpad bug you posted a link ti.
Good luck with the debugging. Just don't expect help from others with no info.


He changed the governor to "Performance" and the CPU did not increase to
the proper frequency... Who cares about further tests of "onDemand"
scaling? (Well, someone working on the code for it obviously, but
irrelevant as far as his bug goes)




--
ubuntu-users mailing list
ubuntu-users@lists.ubuntu.com
Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-users

Another finding, a new bug has been recently posted whose
description looks suspiciously like this problem:*
Bug #786897



--
A.Kromic
-
No trees were harmed in the sending of this message;
however, a large number of electrons were terribly inconvenienced.


--
ubuntu-users mailing list
ubuntu-users@lists.ubuntu.com
Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-users

On Tue, 2011-05-24 at 13:00 +0200, A. Kromic wrote:

> How to check it? One way is from the console, simply do:
> cat /proc/cpuinfo. Within the results should be something like this:

> In ccommandline you can try:
>
> sudo cpufreq-set -g performance

I thought I had this remedied, but just by doing as you suggested, I
found myself running at 1000 instead of 2000. Doing the above fixed
that. I'm baffled that I am having to do this again. Yep, it sure pays
to check, thank you. Ric


--
My father, Victor Moore (Vic) used to say:
"There are two Great Sins in the world...
..the Sin of Ignorance, and the Sin of Stupidity.
Only the former may be overcome." R.I.P. Dad.
Linux user# 44256


--
ubuntu-users mailing list
ubuntu-users@lists.ubuntu.com
Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-users

On Wed, May 25, 2011 at 6:05 PM, A. Kromic <akromic@gmail.com> wrote:
>
> So, my wish was to clean install a 32-bit Ubuntu (Maverick, then Natty)
> to work more comfortably, but neither of them sets my CPU to the full
> frequency! A source of much grief, I must admit
>
> So I guess I don't have much choice. Either will I continue using (or
> make a clean install of) Lucid with an older kernel, or I'll try to deal
> with this bug upstream... I think I'll have to make time for the latter...
>
> Thanks to everybody who helped!


What is your laptop (model, brand)?


--
Ioannis Vranos

http://www.cpp-software.net

--
ubuntu-users mailing list
ubuntu-users@lists.ubuntu.com
Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-users

On 05/26/2011 08:16 AM, Ioannis Vranos wrote:
> On Wed, May 25, 2011 at 6:05 PM, A. Kromic <akromic@gmail.com> wrote:
>> So, my wish was to clean install a 32-bit Ubuntu (Maverick, then Natty)
>> to work more comfortably, but neither of them sets my CPU to the full
>> frequency! A source of much grief, I must admit
>>
>> So I guess I don't have much choice. Either will I continue using (or
>> make a clean install of) Lucid with an older kernel, or I'll try to deal
>> with this bug upstream... I think I'll have to make time for the latter...
>>
>> Thanks to everybody who helped!
>
> What is your laptop (model, brand)?
>
>

It's a Lenovo ThinkPad R61i.

A bit more details if someone is interested:

MT-M: 7650-A9G

00:00.0 Host bridge: Intel Corporation Mobile PM965/GM965/GL960 Memory
Controller Hub (rev 0c)

--
A.Kromic
-
No trees were harmed in the sending of this message;
however, a large number of electrons were terribly inconvenienced.


--
ubuntu-users mailing list
ubuntu-users@lists.ubuntu.com
Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-users

On 05/26/2011 07:18 AM, Ric Moore wrote:
> On Tue, 2011-05-24 at 13:00 +0200, A. Kromic wrote:
>> In ccommandline you can try:
>>
>> sudo cpufreq-set -g performance
> I thought I had this remedied, but just by doing as you suggested, I
> found myself running at 1000 instead of 2000. Doing the above fixed
> that. I'm baffled that I am having to do this again. Yep, it sure pays
> to check, thank you. Ric
>
>
You actually have no problems - frequency scaling obviously works, but
Ubuntu uses the 'ondemand' CPU governor which sets an unloaded CPU to
its lowest frequency and increases it as the load rises. Setting the
governor to 'performance' will get your CPU frequency high immediately
which will improve CPU performance a bit, but will drain batteries
faster when off the plug. If you install the gnome CPU Frequency applet
you can watch the frequency being changed on the fly when using
'ondemand' governor.

However, the change you did is not permanent, on the next boot the
governor will return to 'ondemand'. If you want to switch it permanently
to another one like 'performance', you must edit an initscript. However,
a much better solution would be to install a CPU frequency like
cpufreqd, powernowd etc; with most of them you can configure when to use
which governor (e.g. 'powersave' on battery, 'performance' on AC etc.)

--
A.Kromic
-
No trees were harmed in the sending of this message;
however, a large number of electrons were terribly inconvenienced.


--
ubuntu-users mailing list
ubuntu-users@lists.ubuntu.com
Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-users

>How to check it? One way is from the console, simply do:cat
> /proc/cpuinfo. Within the results should be something like this:

>model name : Intel(R) Pentium(R) Dual CPU T2330 @ 1.60GHz
> stepping : 13
> cpu MHz : 800.000


Fresh install of 11.04 Final on my Lenovo R61i:

cat /proc/cpuinfo

Processor 0:
model name : Intel(R) Pentium(R) Dual CPU T2390 @ 1.860GHz
stepping : 13
cpu MHz : 1867.000

Processor 1:

model name : Intel(R) Pentium(R) Dual CPU T2390 @ 1.860GHz
stepping : 13
cpu MHz : 800.000


And this alternates - each core changes from 1867 to 800, and back again
every time I type the command. Also, sometimes both will be at 800,
sometimes both at 1867. Doesn't matter if I'm AC or battery.

Is this how it should be working?







--
ubuntu-users mailing list
ubuntu-users@lists.ubuntu.com
Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-users

On 05/26/2011 06:04 PM, compdoc wrote:


cat /proc/cpuinfo

Processor 0:
model name : Intel(R) Pentium(R) Dual CPU T2390 @ 1.860GHz
stepping : 13
cpu MHz : 1867.000

Processor 1:

model name : Intel(R) Pentium(R) Dual CPU T2390 @ 1.860GHz
stepping : 13
cpu MHz : 800.000


And this alternates - each core changes from 1867 to 800, and back again
every time I type the command. Also, sometimes both will be at 800,
sometimes both at 1867. Doesn't matter if I'm AC or battery.

Is this how it should be working?



Yes.


--
ubuntu-users mailing list
ubuntu-users@lists.ubuntu.com
Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-users