multi core laptop system overheating, powertop showing a lot [Rescheduling interrupts] <kernel IPI>

I have the same issue on my laptop... fans working all the time, the battery performance is very low and according to powertop:

46,3% (500,5) [kernel scheduler] Load balancing tick

AlsaVersion: Advanced Linux Sound Architecture Driver Version 1.0.21.
AplayDevices:
 **** List of PLAYBACK Hardware Devices ****
 card 0: Intel [HDA Intel], device 0: STAC92xx Analog [STAC92xx Analog]
   Subdevices: 1/1
   Subdevice #0: subdevice #0
Architecture: i386
ArecordDevices:
 **** List of CAPTURE Hardware Devices ****
 card 0: Intel [HDA Intel], device 0: STAC92xx Analog [STAC92xx Analog]
   Subdevices: 1/1
   Subdevice #0: subdevice #0
AudioDevicesInUse:
 USER PID ACCESS COMMAND
 /dev/snd/controlC0: eligos 1480 F.... pulseaudio
CRDA: Error: [Errno 2] No existe el fichero ó directorio
Card0.Amixer.info:
 Card hw:0 'Intel'/'HDA Intel at 0xf4500000 irq 22'
   Mixer name : 'IDT 92HD81B1C5'
   Components : 'HDA:111d76d5,102802bb,00100302'
   Controls : 14
   Simple ctrls : 9
DistroRelease: Ubuntu 10.04
HibernationDevice: RESUME=UUID=ec6f2b89-008a-4219-96f7-470e9c5d2e1c
InstallationMedia: Ubuntu 9.10 "Karmic Koala" - Release i386 (20091028.5)
MachineType: Dell Inc. Vostro 1320
NonfreeKernelModules: wl
Package: linux-meta (not installed)
ProcCmdLine: BOOT_IMAGE=/boot/vmlinuz-2.6.32-18-generic-pae root=UUID=ad6afb0c-f33a-4da0-83d5-1ac2ff65d644 ro quiet splash
ProcEnviron:
 LANG=es_CL.utf8
 SHELL=/bin/bash
ProcVersionSignature: Ubuntu 2.6.32-18.27-generic-pae 2.6.32.10+drm33.1
Regression: No
RelatedPackageVersions: linux-firmware 1.33
Reproducible: Yes
RfKill:
 1: hci0: Bluetooth
  Soft blocked: no
  Hard blocked: no
Tags: lucid needs-upstream-testing
TestedUpstream: No
Uname: Linux 2.6.32-18-generic-pae i686
UserGroups: adm admin cdrom dialout lpadmin plugdev sambashare
dmi.bios.date: 04/08/2009
dmi.bios.vendor: Dell Inc.
dmi.bios.version: A02
dmi.board.name: 0T052J
dmi.board.vendor: Dell Inc.
dmi.chassis.type: 8
dmi.chassis.vendor: Dell Inc.
dmi.chassis.version: N/A
dmi.modalias: dmi:bvnDellInc.:bvrA02:bd04/08/2009:svnDellInc.:pnVostro1320:pvrNull:rvnDellInc.:rn0T052J:rvr:cvnDellInc.:ct8:cvrN/A:
dmi.product.name: Vostro 1320
dmi.product.version: Null
dmi.sys.vendor: Dell Inc.

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

[This is an automated message. Apologies if it has reached you inappropriately.]

This bug was reported against the linux-meta package when it likely should have been reported against the linux package instead. We are automatically transitioning this to the linux kernel package so that the appropriate teams are notified and made aware of this issue.

If this bug really is a bug in the linux-meta package you can move it back to linux-meta and tag it kj-linux-meta, or contact us on the #ubuntu-kernel channel on the FreeNode IRC server. Thanks.

Thomas, could you add the following information (best done on battery, so powertop /ACPI shows a power usage):

1. On a normally booted idle system, run "sudo powertop -t60 -d" and attach that output
2. For comparison, try booting with "processor.max_cstate=2" and run the above powertop statement again and attach the output, too.

Thanks.

Stefan suggested in a private mail:

> What you could try is to specify "processor.max_cstate=2" (or =1 if there is no
> change with 2) on the kernel arguments in grub. Then compare your powertop
> readings with the ones before. If that helps, this would be one more system to
> need quirks for that (based on dmidecode information).

I see no change in powertop with "processor.max_cstate=2" or =1.

Leann indicated that this might be a duplicate of bug 526354.

here is my powertop output for this problem

Is it possible that this is a duplicate of bug 524281 ?

I also see a lot of "[Reschduling Interrupts] <kernel IPI>" in powertop. Yesterday I was using Karmic and, when idle, powertop would show as few as 25 interrupts per second in total.

After upgrading to Lucid, I get between 150 and 500 int. per second from "[Reschduling Interrupts] <kernel IPI>" alone. Load balancing tick interrupts have gone up as well.

My laptop CPUs are 6-7 ºC hotter than under Karmic, and the fans are almost always on.

Lucid has disappointed me so much only with this bug that I'm considering downgrading.

Flipping Windicators...

How about fixing stuff that's BROKEN, not breaking stuff that works...you know, the sort of thing that Ubuntu USED to do back when they still gave a flying Shuttleworth about their users?

I managed to reduce the Rescheduling Interrupts by appending the "nolapic" to the kernel line.

My system gets around 15 wakes per second (total!) when it's idle now, no more hundreds-of-unexplained-interrupts. Battery time and CPU temperature are back to normal, as well.

I've seen a small increment in OHCI interrupts (hda-intel, ath, sata), but it's negligible: around 40-50 wakes per second when I'm using Firefox (which makes heavy use of my HDD and wifi card). To put it in perspective, my touchpad alone will cause more than 100 wakes per second when it's active.

If someone could explain what collateral damage could "nolapic" cause, I'd be very grateful.

Otherwise, please try it for yourselves and post your results.

I just realized that "nolapic" disables SMP (or at least so it seems) in my laptop. I can only see one CPU and running CPU-intensive applications (like Netbeans) slows down the system so much that I can't even work.

On an Acer TravelMate 2500 with Intel Pentium 4 @ 2.80GHz cpu (i.e. 32-bit single core, no hyperthreading) I have been struggling with a huge power increase since upgrading to Lucid (now with 2.6.32-22-generic #35-Ubuntu SMP Tue Jun 1 14:17:36 UTC 2010 i686 GNU/Linux). The battery life fell from 90 minutes to 30 minutes, the computer became very hot, and on mains power, the "brick" also became very hot. The fan accelerates to top speed within a few minutes of use. Investigation with powertop revealed about 90000 (ninety thousand) wakeups per second, which I spell out in words in case you think I have put too many zeros.

Simply using the nolapic kernel option has reduced the wakeup rate to about 80 (eighty) per second. Things seem to be cooling down now, but I have not yet measured the battery life. Fan speed is now back to normal when the computer is idle.

On two other systems, I observe no problems. These are:

Intel Pentium 4 @ 2.80GHz ( 32-bit single core with hyperthreading)
AMD Athlon K8 LE1600 @ 1-2.2 GHz (64-bit single core)

They experience typically 20-40 wakeups per second.

Clearly, this problem has very serious consequences on some systems.

Fixing this bug can extend battery life i think.

I have a Tablet PC where all devices i have on power saving modes but also kernel scheduler is making too many CPU interrupts.