[regression] ondemand governor failed, too long transition latency of HW, fallback to performance governor

[Fabien Tassin]

for more than a week now, the kernel regressed in oneiric. My (quad-core) desktop is now very slow. It seems to correspond to the
2.6.39-3 to 3.0-0 upgrade.

[ 35.650467] p4-clockmod: Warning: EST-capable CPU detected. The acpi-cpufreq module offers voltage scaling in addition to frequency scaling. You should use that instead of p4-clockmod, if possible.
[ 35.650622] p4-clockmod: P4/Xeon(TM) CPU On-Demand Clock Modulation available
...
[ 74.321609] ondemand governor failed, too long transition latency of HW, fallback to performance governor
[ 154.436513] ondemand governor failed, too long transition latency of HW, fallback to performance governor
[ 154.436737] ondemand governor failed, too long transition latency of HW, fallback to performance governor
[ 154.436901] ondemand governor failed, too long transition latency of HW, fallback to performance governor
[ 154.680115] ondemand governor failed, too long transition latency of HW, fallback to performance governor

ProblemType: Bug
DistroRelease: Ubuntu 11.10
Package: linux-image-3.0-1-generic 3.0-1.2
ProcVersionSignature: Ubuntu 3.0-1.2-generic 3.0.0-rc3
Uname: Linux 3.0-1-generic x86_64
NonfreeKernelModules: nvidia
AlsaVersion: Advanced Linux Sound Architecture Driver Version 1.0.24.
Architecture: amd64
AudioDevicesInUse:
 USER PID ACCESS COMMAND
 /dev/snd/controlC0: fta 4432 F.... pulseaudio
CRDA: Error: [Errno 2] No such file or directory
Card0.Amixer.info:
 Card hw:0 'Intel'/'HDA Intel at 0xf9ff8000 irq 44'
   Mixer name : 'Realtek ALC1200'
   Components : 'HDA:10ec0888,104382fe,00100101'
   Controls : 40
   Simple ctrls : 22
Date: Mon Jun 20 08:29:19 2011
HibernationDevice: RESUME=UUID=7ecf42d9-26ea-4e55-8f56-0140456aa0b5
IwConfig:
 lo no wireless extensions.

 eth0 no wireless extensions.
MachineType: System manufacturer P5Q-PRO
ProcEnviron:
 LANGUAGE=en_US:en
 PATH=(custom, user)
 LANG=en_US.UTF-8
 SHELL=/usr/bin/tcsh
ProcKernelCmdLine: BOOT_IMAGE=/boot/vmlinuz-3.0-1-generic root=UUID=1ff08532-facc-4e0e-98ed-2866137b7ed5 ro quiet splash vt.handoff=7
RelatedPackageVersions:
 linux-restricted-modules-3.0-1-generic N/A
 linux-backports-modules-3.0-1-generic N/A
 linux-firmware 1.54
RfKill:

SourcePackage: linux
UpgradeStatus: Upgraded to oneiric on 2009-07-28 (691 days ago)
WpaSupplicantLog:

dmi.bios.date: 07/31/2008
dmi.bios.vendor: American Megatrends Inc.
dmi.bios.version: 1104
dmi.board.asset.tag: To Be Filled By O.E.M.
dmi.board.name: P5Q-PRO
dmi.board.vendor: ASUSTeK Computer INC.
dmi.board.version: Rev 1.xx
dmi.chassis.asset.tag: Asset-1234567890
dmi.chassis.type: 3
dmi.chassis.vendor: Chassis Manufacture
dmi.chassis.version: Chassis Version
dmi.modalias: dmi:bvnAmericanMegatrendsInc.:bvr1104:bd07/31/2008:svnSystemmanufacturer:pnP5Q-PRO:pvrSystemVersion:rvnASUSTeKComputerINC.:rnP5Q-PRO:rvrRev1.xx:cvnChassisManufacture:ct3:cvrChassisVersion:
dmi.product.name: P5Q-PRO
dmi.product.version: System Version
dmi.sys.vendor: System manufacturer

[Fabien Tassin (fta)]

Here are the bootlogs before and after this upgrade.

[Fabien Tassin (fta)]

during boot, i see an error saying pcc-cpufreq.ko is missing.

[Julian Wiedmann (jwiedmann)]

With 3.0, the x86 cpufreq drivers were moved into the /drivers/cpufreq directory (previously containing only the cpufreq governors). Debian's cpufrequtils had some issues with loading these "new" modules, this bug might be related.

Details:
https://lkml.org/lkml/2011/8/13/77
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=635348
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=636141

[Brad Figg (brad-figg)]

Thank you for taking the time to file a bug report on this issue.

However, given the number of bugs that the Kernel Team receives during any development cycle it is impossible for us to review them all. Therefore, we occasionally resort to using automated bots to request further testing. This is such a request.

We have noted that there is a newer version of the development kernel currently in the release pocket than the one you tested when this issue was found. Please test again with the newer kernel and indicate in the bug if this issue still exists or not.

If the bug still exists, change the bug status from Incomplete to Confirmed. If the bug no longer exists, change the bug status from Incomplete to Fix Released.

Thank you for your help.

[Robert Schmidt (r-schmidt)]

I have the same issue on an atom netbook I just updated to final oneiric which uses kernel 3.0.0-12.

[Brad Figg (brad-figg)]

Thank you for taking the time to file a bug report on this issue.

However, given the number of bugs that the Kernel Team receives during any development cycle it is impossible for us to review them all. Therefore, we occasionally resort to using automated bots to request further testing. This is such a request.

We have noted that there is a newer version of the development kernel than the one you last tested when this issue was found. Please test again with the newer kernel and indicate in the bug if this issue still exists or not.

If the bug still exists, change the bug status from Incomplete to Confirmed. If the bug no longer exists, change the bug status from Incomplete to Fix Released.

Thank you for your help, we really do appreciate it.

[Robert Schmidt (r-schmidt)]

Where does this robot gets the version number from?
It is still there in the newest version.

[Robert Schmidt (r-schmidt)]

The real problem seems to be indicated by the scaling driver:

p4-clockmod: Warning: EST-capable CPU detected. The acpi-cpufreq module offers voltage scaling in addition to frequency scaling. You should use that instead of p4-clockmod, if possible.

Following the advise by simply renaming p4-clockmod.ko makes the kernel use acpi-cpufreq and then frequency scaling works again.
So the bug is that the kernel loads the wrong scaling driver despite knowing better.

[Joseph Salisbury (jsalisbury)]

Would it be possible for you to test the latest upstream kernel? It will allow additional upstream developers to examine the issue. Refer to https://wiki.ubuntu.com/KernelMainlineBuilds . Please test the release candidate kernel versus the daily build. Once you've tested the upstream kernel, please remove the 'needs-upstream-testing' tag. This can be done by clicking on the yellow pencil icon next to the tag located at the bottom of the bug description and deleting the 'needs-upstream-testing' text. Please let us know your results.

Thanks in advance.

[Robert Schmidt (r-schmidt)]

Ok, first things first:

When I started trying with the mainline kernels I realized that I wasn't using the official Ubuntu kernel but one from a ppa repository (https://launchpad.net/~voria/+archive/ppa).

After switching back to the official 3.0.0-12.20 I don't see the problem anymore. So the bug is in the changes this repository made to the kernel. Sorry about that.

But as I was changing the kernel anyway I tried also with the following and none of them showed the problem:
3.0.4-030004.201108301138
3.0.6-030006.201110050043
3.1.0-0301rc10.201110181253

So again sorry for not checking correctly what I was running.

[Joseph Salisbury (jsalisbury)]

I'll mark this as fix-released, since the current Oneiric kernel fixes the issue.