CPU usage very high with no running tasks using it

Its probably not related, but I get a mysterious 100% CPU utilisation upon booting Dapper on my Dell...can't see whats using it with top, but if I kill update-notifier it drops to normal....

Try killing update notifier and see if it helps, if so, we've probably got a bug.

Cheers,
Craig

Hello all,

The update-notifier does seem to make my CPU go even crazier than usual
until I kill it, but that's not the particular bug I'm referring to
here.

Having just killed the notifier after just logging in, I'm now writing
this in gedit with a gnome terminal open showing top and synaptic
downloading new packages. CPU use is permanently above 50% of 2.13GHz.
I've also just switched to the 'radeon' driver, so that's not the
problem and have tried the vesa driver also: Whichever I use out of
radeon/vesa/fglrx doesn't seem to affect the outcome.

Please let me know if there's any more information I can give.

Best wishes,
David

On Mon, 2006-02-06 at 14:18 +0000, Craig Sampson wrote:
> Public bug report changed:
> https://launchpad.net/malone/bugs/30570
>
> Comment:
> Its probably not related, but I get a mysterious 100% CPU utilisation
> upon booting Dapper on my Dell...can't see whats using it with top, but
> if I kill update-notifier it drops to normal....
>
> Try killing update notifier and see if it helps, if so, we've probably
> got a bug.
>
> Cheers,
> Craig

Try running without fglrx to see if that changes anything. I can get any help from the kernel devs unless you can reproduce this without the non-free module.

Please see my post dated: 2006-02-06 18:57:01 UTC

Should be of use:

Changing _nothing_ but installing the i386 kernel image and the restricted modules package solves the problem on my hardware.

I also have a dell inspiron 6000 and can confirm that running the 686 kernel gives ~70% cpu usage when no user processes are run.

I'm currently running the 386 kernel wich gives me no troubble.

I've just gathered some information from top. Dunno if its usefull...

386 kernel:

  single user mode:
    cpu: user 0.0% system ~1.0%

  runlevel 5, no X:
    cpu: user 0.0% system ~1.0%

  gnome:
    cpu: user ~1.0% system ~1.0%

686 kernel:

  single user mode:
    cpu: user 0.0% system ~27.0%

  runlevel 5, no X:
    cpu: user 0.0% system ~37.0%

  gnome:
    cpu: user ~27.0% system ~46.0%

I can confirm the same scenario on my machine:

My hardware is:
HP nx8220

lspci:

0000:00:00.0 Host bridge: Intel Corporation Mobile 915GM/PM/GMS/910GML Express Processor to DRAM Controller (rev 03)
0000:00:01.0 PCI bridge: Intel Corporation Mobile 915GM/PM Express PCI Express Root Port (rev 03)
0000:00:1c.0 PCI bridge: Intel Corporation 82801FB/FBM/FR/FW/FRW (ICH6 Family) PCI Express Port 1 (rev 03)
0000:00:1c.1 PCI bridge: Intel Corporation 82801FB/FBM/FR/FW/FRW (ICH6 Family) PCI Express Port 2 (rev 03)
0000:00:1d.0 USB Controller: Intel Corporation 82801FB/FBM/FR/FW/FRW (ICH6 Family) USB UHCI #1 (rev 03)
0000:00:1d.1 USB Controller: Intel Corporation 82801FB/FBM/FR/FW/FRW (ICH6 Family) USB UHCI #2 (rev 03)
0000:00:1d.2 USB Controller: Intel Corporation 82801FB/FBM/FR/FW/FRW (ICH6 Family) USB UHCI #3 (rev 03)
0000:00:1d.7 USB Controller: Intel Corporation 82801FB/FBM/FR/FW/FRW (ICH6 Family) USB2 EHCI Controller (rev 03)
0000:00:1e.0 PCI bridge: Intel Corporation 82801 Mobile PCI Bridge (rev d3)
0000:00:1e.2 Multimedia audio controller: Intel Corporation 82801FB/FBM/FR/FW/FRW (ICH6 Family) AC'97 Audio Controller (rev 03)
0000:00:1e.3 Modem: Intel Corporation 82801FB/FBM/FR/FW/FRW (ICH6 Family) AC'97 Modem Controller (rev 03)
0000:00:1f.0 ISA bridge: Intel Corporation 82801FBM (ICH6M) LPC Interface Bridge (rev 03)
0000:00:1f.1 IDE interface: Intel Corporation 82801FB/FBM/FR/FW/FRW (ICH6 Family) IDE Controller (rev 03)
0000:01:00.0 VGA compatible controller: ATI Technologies Inc M24 1P [Radeon Mobility X600]
0000:02:04.0 Network controller: Intel Corporation PRO/Wireless 2200BG (rev 05)
0000:02:06.0 CardBus bridge: Texas Instruments PCIxx21/x515 Cardbus Controller
0000:02:06.2 FireWire (IEEE 1394): Texas Instruments OHCI Compliant IEEE 1394 Host Controller
0000:02:06.3 Mass storage controller: Texas Instruments PCIxx21 Integrated FlashMedia Controller
0000:02:06.4 0805: Texas Instruments PCI6411, PCI6421, PCI6611, PCI6621, PCI7411, PCI7421, PCI7611, PCI7621 Secure Digital (SD) Controller
0000:02:06.5 Communication controller: Texas Instruments PCI6411, PCI6421, PCI6611, PCI6621, PCI7411, PCI7421, PCI7611, PCI7621 Smart Card Controller (SMC)
0000:10:00.0 Ethernet controller: Broadcom Corporation NetXtreme BCM5751M Gigabit Ethernet PCI Express (rev 11)

Confirmed on my Dell Inspiron 6000

Same situation on my Dell Inspiron 6000:
2.6.15-18-686: high cpu usage when idle
2.6.15-18-386: no cpu usage when idle

Same problem on Sony vaio fs640/w with pentium-m.
I think it's the fact that the 686 kerel is SMP.
When I was compiling my own kernel I experienced the same high cpu load. When I turned off SMP support cpu load went down.

Same problem on Dell XPS Gen 2 (Pentium M 2.13GHz) and same workaround (installing 386 kernel)

This problem did not occur with breezy.

Getting the same thing with 2.6.15-21-686 on a HP nx6110 (centrino, 1.6GHz)

Going to try 386 and older kernels.

DItto here with Ubuntu 6.06 2.6.15-23-686 on a Dell Inspiron 9300 (centrino, 1.6GHz)

In the process switching to 386 to see if it helps.

I have a similar problem on my Dell Latitude D810. I'm using the ati driver.
The gnome system monitor always shows about 70% cpu usage.

However, and maybe the other reporters should check that as well, I get completely different values when I query the CPU load with top (60% idle), sar (95% idle) and vmstat (high fluctuations here).

Compare: https://launchpad.net/distros/ubuntu/+ticket/961

I can also confirm this problem with my Dell Latitute D810. System Monitor gives me very high CPU loads (60-75%) although the system is practically idle, but as opposed to previous posts, top also reports high CPU usage (~32% idle).

I am using the 386 kernel and everything is working perfectly. I do believe it to be a problem with SMP, as I compiled 2.6.16 (vanilla) 686 without SMP and it also works fine for cpu usage; Compatibility with my wireless card and USB a no go, but that's my fault and need to continue learning how to compile it better.

On Mon, 2006-06-12 at 01:06 +0000, Christopher O'Hara wrote:
> *** This bug is a duplicate of bug 30557 ***
>
> I am using the 386 kernel and everything is working perfectly. I do
> believe it to be a problem with SMP, as I compiled 2.6.16 (vanilla) 686
> without SMP and it also works fine for cpu usage; Compatibility with my
> wireless card and USB a no go, but that's my fault and need to continue
> learning how to compile it better.

Guys, have you tried adding this to /etc/rc.local:

echo 1 > /sys/module/processor/parameters/max_cstate

This is what you should do if you want to use a 686 kernel.

Of course, I wish someone would revert the ubuntu patch that *caused*
this in the first place...

IT WORKS!!!!!!!!!!!!!!!!!!!!!!

props to you sohail!

I'm running 686 with very minimal cpu idle usage.

On Mon, 2006-06-12 at 04:50 +0000, Christopher O'Hara wrote:
> *** This bug is a duplicate of bug 30557 ***
>
> IT WORKS!!!!!!!!!!!!!!!!!!!!!!
>
> props to you sohail!
>
> I'm running 686 with very minimal cpu idle usage.

Well the bug that this one duplicates has that solution :)

I had the same problem with 2.6.15-23-k7 which is compiled with SMP also. 2.6.15-23-386 is working fine.

I just installed updates today, 6/15 and now both my boxes are constant 15% with both 686 and 386 kernels.

Mark

I installed the updates and seemed to still have the high cpu usage in 686 (after disabling the fix above) I reinstated the fix and I'm back down to 4-5% usage. I don't think the kernel updates helped much for me. I vaguely remember that the email stated it was more of a security update for the kernel.

On Fri, 2006-06-16 at 03:30 +0000, barakaspeed wrote:
> I installed the updates and seemed to still have the high cpu usage in
> 686 (after disabling the fix above) I reinstated the fix and I'm back
> down to 4-5% usage. I don't think the kernel updates helped much for
> me. I vaguely remember that the email stated it was more of a security
> update for the kernel.

It was about 50/50 fixes+security.

The main thing was dropping the HZ from 1000 to 250, which should have
fixed the problem.

If it didn't for you, then we're looking at a different issue.

For me, the kernel upgrade also didn't fix the problem. There's still a very high cpu usage without any CPU intensive apps running.

For me, the kernel upgrade also did *not* fix the problem. There's still a very high cpu usage without any CPU intensive apps running.

Just to add to this, the latest kernel update also didn't fix this problem for me.

I confirm. Bug is still there in 2.6.15-25-686 #1 SMP PREEMPT kernel.

On my laptop (Acer Aspire 1692WLMi, Pentium M740 1.73 GHz),
With echo 1 > /sys/module/processor/parameters/max_cstate
top reports tipically:
Cpu(s): 4.3% us, 0.3% sy, 0.0% ni, 95.0% id, 0.0% wa, 0.0% hi, 0.3% si

With echo 2 > /sys/module/processor/parameters/max_cstate
I get:
Cpu(s): 54.5% us, 9.1% sy, 0.0% ni, 36.4% id, 0.0% wa, 0.0% hi, 0.0% si

This does not seem to be simply a problem with mis-reporting the CPU usage. The system actually feels slower when max_cstate is >= 2: launching a firefox window is noticeably slower; I can actually see the jumps as the window outline grows on-screen. In fact, that is what prompted me to look for bugs in the first place. I'll try to back this up with some benchmark later. Also, I'll try and check this on a Core Duo processor. Has anyone done that?

On a Core Duo T2300 (1.6GHz), kernel -686 does not sem to have this problem. When idel, top reports tipically
Cpu(s): 2.3% us, 0.3% sy, 0.0% ni, 97.2% id, 0.0% wa, 0.2% hi, 0.0% si

This seems to be the average between the 2 CPUs. System monitor shows both CPU1 and CPU2 around 2.0%.

Both processors throttled down to 996 MHz. /proc/acpi/processor/CPU*/power shows both CPUs spend most time in C2.

Seems fine.

I can confirm this is a problem on my HP Compaq nc6230. Intel Pentiume M processor 1.86Ghz. Running the official kernel 2.6.15-25-686 gives me a couple dozen percentage points higher CPU usge when idle while the 2.6.15-25-386 kernel leaves me at near 0 percent CPU usage when idling. It's always the X process that shows to be using the CPU time. Here is the process as listed in gnome-system-monitor:

/usr/bin/X :0 -br -audit 0 -auth /var/lib/gdm/:0.Xauth -nolisten tcp vt7

gnome-system-monitor shows overall CPU usage at ~50% while top shows a fluxtuation of ~15-35% (it's harder to see what the real usage is w/o a graph). The 386 kernel usage is what should be expected, which is nearly at the bottom for CPU usage. All of this is while "idle", of course.

My inspiron 6000 (Pentium M) exhibits this problem on at least 2.6.15-23-686 and 2.6.15-25-686, and the max_cstate fix:

echo 1 > /sys/module/processor/parameters/max_cstate

Fixes it. As an added bonus, with -25 I get some weird timing problems (beeps sound for different lengths, the "zooming" box that animates when you click a gnome panel icon is erratic, etc), which are also fixed with the above command.

Mick.

On an added note, the weird timing problem is strikingly similar to this:

http://www.thinkwiki.org/wiki/Talk:Problem_with_high_pitch_noises

Also, i have added the line:

echo 1000 > /proc/sys/dev/rtc/max-user-freq

To my startup scripts (for mplayer I think).

I also get this bug on my system; Dell Inspiron 6000 (laptop), Intel celeron M CPU, i915 graphics.

As stated in bug #30557 setting max_cstate to 2 would fix this, but it didn't work on my system. However, setting max_cstate to 1 (mentioned above), fixes it. (ie: adding "echo 1 > /sys/module/processor/parameters/max_cstate" in /etc/rc.local).

Toshiba Satellite M65-9092 is the same as above: Loads of CPU usage when idle, which goes away with the /etc/rc.local addition.

Rejecting linux-meta tasks; this bug has nothing to do with the metapackages and is already filed underlinux-source-2.6.15

This problem also occurs for me on my ASUS W2V laptop on the 2.6.17-10 kernel.
The echo 1 > ... fix appears to work as well.

Ditto...

I was kinda hoping that this might have been figured out in the new "generic" kernel in edgy.

Is this rc.local fix going to be the recognized solution? or is there going to be a more hands-off solution in the future?

thanks

again asus z70va w/ pent. m

On Sun, 2006-10-29 at 19:19 +0000, hardyn wrote:
> *** This bug is a duplicate of bug 30557 ***
>
> Ditto...
>
> I was kinda hoping that this might have been figured out in the new
> "generic" kernel in edgy.

Hi,

It works for me here. Pentium M, Dell XPS Gen2.

This bug is still present on feisty herd2 on a sony vaio fs11m
kernel:
2.6.20-5-generic

changing max_cstate as follows, resolve the issue:

echo 1 > /sys/module/processor/parameters/max_cstate

On 10/29/06, sohail <email address hidden> wrote:
> *** This bug is a duplicate of bug 30557 ***
>
> On Sun, 2006-10-29 at 19:19 +0000, hardyn wrote:
> > *** This bug is a duplicate of bug 30557 ***
> >
> > Ditto...
> >
> > I was kinda hoping that this might have been figured out in the new
> > "generic" kernel in edgy.
>
> Hi,
>
> It works for me here. Pentium M, Dell XPS Gen2.
>
> --
> CPU usage very high with no running tasks using it
> https://launchpad.net/bugs/30570
>

Hi!

I'm hitting same kind of difficulties. But the problem started after I got an Ati 3D card. And when using fglxr I hit with an huge CPU load on any screen intensive apps.

For example a Video in full screen will lag down to 15 fps while the same video runs in Windows fully and with CPU load @ 40-60%

Same thing was before while I had nVidia card under Ubuntu, I was able to watch that video and all the desktop effect were a lot faster.

And to compare my Video cards, we are talking about GeForce 7800GT and Ati Radeon HD3850, which have a big difference in 3D speed for Ati one.

But the similiarity to this problem is some ppl referring to the fglrx and the thing that my Desktop PC is build on IP915 motherboard and Pentium M (to be specific Dothan) Processor.

The Ubuntu versions where I have found this problem are 7.10 and 8.04.

Beginning with the Hardy Heron 8.04 development cycle, all open Ubuntu kernel bugs need to be reported against the "linux" kernel package. We are automatically migrating this linux-source-2.6.15 kernel bug to the new "linux" package. We appreciate your patience and understanding as we make this transition. Also, if you would be interested in testing the upcoming Intrepid Ibex 8.10 release, it is available at http://www.ubuntu.com/testing . Please let us know your results. Thanks!

Hi all again!

I was able to go around this bug in Debian / Lenny. I'm not sure what I really did, but I worked over 2 days for it. Now I'm on Ubuntu again and my whole desktop seems to be slow and unresponding. Yet I have tried allmost everything to work around this pug. I'm not using any glx desktop and the videos are working, but for example. moving an window in desktop is slow and it takes while (half an second) before the window responds to moving.

I tried the Intrepid but was unable to install _any_ Ati drivers on it.

And the why I changed back to Ubuntu was due to some odd LDAP bug in Lenny.

When i enter "echo 1 > /sys/module/processor/parameters/max_cstate" in the Terminal, it doesn't work and instead says:

bash: /sys/module/processor/parameters/max_cstate: No such file or directory
aman@aman-desktop:~$

The directories are definitely there. Am i just doing the whole thing wrong? I gotta fix this CPU problem..

I got the same problem here, my CPU usage is high by default!

My laptop: Acer AspireOne.
Model Number: ZG5.
Processor: Intel Atom.
Operating System: Ubuntu 9.10 - the Karmic Koala.

Status changed to 'Confirmed' because the bug affects multiple users.