System freezes whilst powernowd stepping up AMD64 freq (intermittent/hoary)

I've been running amd64-generic kernel for more than a week now, and not a single freeze so far.
So I can confirm that this is related to k8 kernel. Version info follows:
linux-amd64-k8 2.6.10-7
linux-amd64-generic 2.6.10-7
powernowd 0.90-3ubuntu14

I have AMD64-generic, did not upgrade kernel.

However I got system freezing over and over again, and in matter of minutes, not
hours. Disabling powernowd solves the problem.

cpuinfo:

processor : 0
vendor_id : AuthenticAMD
cpu family : 15
model : 31
model name : AMD Athlon(tm) 64 Processor 3200+
stepping : 0
cpu MHz : 2000.123
cache size : 512 KB
fpu : yes
fpu_exception : yes
cpuid level : 1
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov
pat pse36 clflush mmx fxsr sse sse2 pni syscall nx mmxext fxsr_opt lm 3dnowext
3dnow lahf_lm
bogomips : 3923.96
TLB size : 1088 4K pages
clflush size : 64
cache_alignment : 64
address sizes : 40 bits physical, 48 bits virtual
power management: ts fid vid ttp

I can confirm this bug on an AMD 3200+ system. The symptoms are exactly as
described in the original description.

Relevant information:

# cat /proc/cpuinfo
processor : 0
vendor_id : AuthenticAMD
cpu family : 15
model : 12
model name : AMD Athlon(tm) 64 Processor 3200+
stepping : 0
cpu MHz : 2210.111
cache size : 512 KB
fpu : yes
fpu_exception : yes
cpuid level : 1
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov
pat pse36 clflush mmx fxsr sse sse2 pni syscall nx mmxext lm 3dnowext 3dnow
bogomips : 4374.52
TLB size : 1024 4K pages
clflush size : 64
cache_alignment : 64
address sizes : 40 bits physical, 48 bits virtual
power management: ts fid vid ttp

powernowd 0.96-1ubuntu1
linux-image-2.6.11-9-amd64-k8

This is on Breezy Badger.

I had the same problem with a Athlon64 3200 running hoary Kubuntu-x86 version.

I installed powernowd 0.96-1 from hoary-backports and it fixed the problem. Cpu
speeds change with usage.

(In reply to comment #4)
> I had the same problem with a Athlon64 3200 running hoary Kubuntu-x86 version.
>
> I installed powernowd 0.96-1 from hoary-backports and it fixed the problem. Cpu
> speeds change with usage.
>
>

Well it seems that 0.96-1 did not really fix the problem. While it worked when I
tried it by manually starting powernowd with sudo /etc/init.d/powernowd start ,
I still get the freeze when I boot up so I have disactivated powernowd again.

I've seen this as well, but on a Dell Latitude 810 (and thus an Intel PIII CPU).

I had this problem when I installed Hoary, but somehow managed to make it go away
(check my log at http://www.ii.uib.no/~ketil/ubuntu-prob.log).

Now I installed Breezy (Col 3), and the problem is back. It's fairly evident,
every time the CPU is scaled up, there is a second or so of system freeze (including
mouse pointer).

Oops - my computer freezes when the CPU frequency *drops*. See bug 13470.

This bug has been flagged because it is old and possibly inactive. It may or may
not be fixed in the latest release (Breezy Badger 5.10). It is being marked as
"NEEDSINFO". In two weeks time, if the bug is not updated back to "NEW" and
validated against Breezy, it will be closed.

This is needed in order to help manage the current bug list for the kernel. We
would like to fix all bugs, but need users to test and help with debugging.

If this change was in error for this bug, please respond and make the
appropriate change (or email <email address hidden> if you cannot make the
change).

Thanks for your help.

If possible, please upgrade to Dapper's 2.6.15-7 kernel. If you do not want to
upgrade to Dapper, then you can also wait for the Dapper Flight 2 CD's, which
are due out within the next few days.

Let me know if this bug still exists with this kernel.

Thank you for reporting this bug and following up on it.

I am marking this bug Closed since there has been no response from you for over three months. We
would like to fix all existing issues, but need need feedback to help with debugging.

Should you still have a problem with the latest up to date Dapper kernel and packages, please reopen this bug, answering the+questions that have already been asked of you before.

Thanks again, please report any other issues you have after checking that they haven't already been reported.

I still have this problem (hard freezing) with Dapper with a Intel Centrino 1800.
While powernowd is installed it always freezes after some minutes.
I have following versions:
powernowd 0.97-1ubuntu1
linux-image-2.6.15-23-686 (2.6.15-23-39)

I have also tested an earlyer version 0.96 for the developers homepage... same problem... also tested with other Modes e.g. AGGRESSIVE... same...

I first thought that it was something with the ati drivers... but it freezed with fglrx, ati, vesa and the official fglrx...

Then I have increased the verbosity of powernowd and got following output in the syslog:
May 28 17:57:27 localhost powernowd: Using PASSIVE mode.
May 28 17:57:27 localhost powernowd: PowerNow Daemon v0.97, (c) 2003-2006 John Clemens
May 28 17:57:27 localhost powernowd: Settings:
May 28 17:57:27 localhost powernowd: verbosity: 6
May 28 17:57:27 localhost powernowd: mode: 2 (PASSIVE)
May 28 17:57:27 localhost powernowd: step: 100 MHz (100000 kHz)
May 28 17:57:27 localhost powernowd: lowwater: 20 %
May 28 17:57:27 localhost powernowd: highwater: 80 %
May 28 17:57:27 localhost powernowd: poll interval: 1000 ms
May 28 17:57:27 localhost powernowd: Found 1 scalable unit: -- 1 'CPU' per scalable unit
May 28 17:57:27 localhost powernowd: cpu0: 600Mhz - 1800Mhz (7 steps)
May 28 17:57:27 localhost powernowd: step1 : 1800Mhz
May 28 17:57:27 localhost powernowd: step2 : 1600Mhz
May 28 17:57:27 localhost powernowd: step3 : 1400Mhz
May 28 17:57:27 localhost powernowd: step4 : 1200Mhz
May 28 17:57:27 localhost powernowd: step5 : 1000Mhz
May 28 17:57:27 localhost powernowd: step6 : 800Mhz
May 28 17:57:27 localhost powernowd: step7 : 600Mhz
May 28 17:57:28 localhost powernowd: i = 0, cpubase = 0,
May 28 17:57:28 localhost powernowd: PCT = 0.019608
May 28 17:57:28 localhost powernowd: got here LOWER
May 28 17:57:28 localhost powernowd: change = 0, change2 = 0
May 28 17:57:28 localhost powernowd: Setting speed to 600000
May 28 17:57:28 localhost powernowd: mode=0, str=600000
May 28 17:57:29 localhost powernowd: i = 0, cpubase = 0,
...
many more of those...
...
May 28 18:02:26 localhost powernowd: change = 0, change2 = 1
May 28 18:02:27 localhost powernowd: i = 0, cpubase = 0,
May 28 18:02:27 localhost powernowd: PCT = 0.010000
May 28 18:02:27 localhost powernowd: change = 0, change2 = 1
May 28 18:02:28 localhost powernowd: i = 0, cpubase = 0,
May 28 18:02:28 localhost powernowd: PCT = 0.000000
May 28 18:02:28 localhost powernowd: change = 0, change2 = 1
May 28 20:03:47 localhost syslogd 1.4.1#17ubuntu7: restart.

About my Processor:
cat /proc/cpuinfo
processor : 0
vendor_id : GenuineIntel
cpu family : 6
model : 13
model name : Intel(R) Pentium(R) M processor 1.80GHz
stepping : 6
cpu MHz : 1792.961
cache size : 2048 KB
fdiv_bug : no
hlt_bug : no
f00f_bug : no
coma_bug : no
fpu : yes
fpu_exception : yes
cpuid level : 2
wp : yes
flags : fpu vme de...

The only Solution to this so far is to do it staticaly with:
echo powersave > /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor

where powersave could be replaced by one of the following:
userspace powersave ondemand conservative performance
listed in /sys/devices/system/cpu/cpu0/cpufreq/scaling_available_governors

But as I just see this can also be done with the gnome-cpu-frequency applet. :)

Hope this going to be fixed anyway...

The original problem, with the same symptoms, still exists in Kubuntu Dapper 6.06 with both default (generic) and k8 kernels. I don't include /proc/cpuinfo as this is the same CPU as the original problem report. Package version info follows:

Generic kernel:
linux-image-2.6.15-23-amd64-generic / 2.6.15-23.39
linux-restricted-modules-2.6.15-23-amd64-generic / 2.6.15.11-1

K8 kernel:
linux-image-2.6.15-26-amd64-k8 / 2.6.15-26.44
linux-restricted-modules-2.6.15-26-amd64-k8 / 2.6.15.11-3

Other:
powernowd / 0.97-1ubuntu1

For the record: Workaround is disabling /etc/init.d/powernowd with:
chmod -x /etc/init.d/powernowd

Reopened as the problem still exists in Dapper.

System freezes also with edgy for me.

Also it does with Feisty. It only doesn't freeze when I took my ddr2 1GB RAM out of the computer and leave only 512MB there, or if i stop the powernowd.

ubuntu@ubuntu:~$ cat /proc/cpuinfo
processor : 0
vendor_id : AuthenticAMD
cpu family : 15
model : 44
model name : AMD Athlon(tm) 64 Processor 3000+
stepping : 2
cpu MHz : 1000.000
cache size : 512 KB
fdiv_bug : no
hlt_bug : no
f00f_bug : no
coma_bug : no
fpu : yes
fpu_exception : yes
cpuid level : 1
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 syscall nx mmxext fxsr_opt lm 3dnowext 3dnow up pni lahf_lm ts fid vid ttp tm stc
bogomips : 2011.58
clflush size : 64

I have this problem in Edgy as well on Dell Inspiron 5150, P4 HT

 2.6.17-11-generic

Workaround of disabling the scaling manager (powernowd) stops hard locks.

This issue can be solved by adding noapic nolapic acpi=force as your boot parameters. It also works for Live CD Ubuntu, which freezes as well for me.

I also have a similar problem on my server at home.
After testing and beating my head against the wall I stumbled across this bug report.
I have performed similar actions on Dapper, Edgy, and Feisty both x86 and AMD64 platforms.

The box usually crashes during boot. Occasionally it will let me into a shell, or start X. But it usually dies after a few minutes.

DIsabling powernowd (chmod -x /etc/init.d/powernowd) and rebooting fixes the issue.

I will have to try leaving it enabled and using the "noapic nolapic acpi=force" kernel options listed above.

Like I mentiond, I tested this on the ubuntu variant (not kubuntu or edubuntu) under both AMD64 and x86 architectures under versions 6.06 (dapper) 6.10 (edgy) and 7.04 (feisty).

I attempted to test one of the versions of Gutsy, but I ran into a few bugs that prevented me from knowing if the issue was the same.

I have attached my hardware list, kernel version, and pci list.

Whoops--here's the correct attachment.
(I thought launchpad would allow me to attach multiple files)

Relating to bug reports #17763 and #16961 on Ubuntu

I too had these random Freezes on my system. It worked fine on old configuration, but when I changed cpu and memory, cpu frequensy scaling didn't work any more. And it hanged system randomly so I had to reset. Those errors were printed on terminal and keyboard stopped functioning. Little later everything freezed.

My working system was:
Ubuntu Edgy
Motherboard: Abit KN9 Ultra
CPU (old): AMD 64 AM2 3500+
Memory (old):2x 512MB Kingston 667Mhz DDR2 ECC
Video C: Asus 7600GS Top-silent
PSU: Korsair vx450w

New components added which created freezes:
CPU: AMD 64 AM2 x2 4600+
Memory: 2x 1GB Kingston 667Mhz DDR2 ECC

I ran memtest overnight without hit. BIOS flashing didn't change a thing and system was stable on any single cpu frequensy. I wrote this shell script to test cpu frequensy scaling: http://www.antti.miranto.fi/cpu.freq.test/cpu.frequensy.tester.sh This generated error messages or freezed my computer on minutes instead of hours. More on that script at the end.

Next I changed new prosessor to other motherboard and run same script and same Ubuntu release (Feisty live-cd) for few hours without single hit. I got same results with using Gutsy's latest available kernel on Edgy system.

I send motherboard to waranty repair so currently I can't do any tests. I should have tested this on other linux distro too...

I put my error-logs and other stuff to web: http://www.antti.miranto.fi/cpu.freq.test/ (file: "ohjeet..." is a finnish guide to script which I wrote for the company where I bought this motherboard.)

Back to the script. Script runs gzip and bzip2 on every core and checks md5sums for compressed output. Md5sums should be same all the time! At the same time it changes cpu frequensy about every 0.1 second. This script doesn't use powernowd but cpufreq-selector command from gnome-applets package.

If you want to run it ($ sings means commands) follow forward.
***DO NOT USE THIS IN PRODUCTION ENVIRONMENT***
Broken hardware can do anyting except erase your data. :)

boot to Ubuntu live cd.

Script requires taskset program on schedutils package and tries to install it. Apt method doesn't work directy on live-cd for repository reasons so get the package and plase it to /tmp with root as a owner.

get the script and deb-file:
$ wget -nH -r --no-parent -A.sh -A.deb -A.txt http://www.antti.miranto.fi/cpu.freq.test/

change dir:
$ cd cpu.freq.test/

Change permissions and copy schedutils-package to beter place (last may ask your passwd):
$ chmod 775 cpu.frequensy.tester.sh
$ cp schedutils_1.5.0-1_i386.deb /tmp/
$ sudo chown root: /tmp/schedutils_1.5.0-1_i386.deb

Run the test (this may ask your passwd):
$ sudo ./cpu.frequensy.tester.sh

Results:
This script should run days without hit! If your computer locks up, some
program freezes partially or totally or you get md5sum missmatch there
might be hardware problem on your system or Ubuntu has a trouble to
handle your hardware.

Questions and comments: <email address hidden>

Thank you for taking the time to report this bug and helping to make Ubuntu better. You reported this bug a while ago and there hasn't been any activity in it recently. We were wondering is this still an issue for you? Can you try with latest Ubuntu release? Thanks in advance.

This wasn't issue for me on AMD turion 64 x2 running hardy heron and custom kernel 2.6.25.4 (or even distro specific one)

This bug report is being closed due to your last comment regarding this being fixed with an update. Thank you again for taking the time to report this bug and helping to make Ubuntu better. Feel free to submit any future bugs you may find.