You are not logged in.

#1 2013-09-05 22:14:30

bjo
Member
Registered: 2011-09-10
Posts: 93

Kernel 3.11: systemd 204 breaks powerknow-k8 autoloading

Hi!

On my machine running on a "AMD Athlon(tm) 64 X2 Dual Core Processor 5000+" powernow-k8 is not inserted automatically, with all other kernels since 3.4 it worked. I looked on the cpufreq-wikipage, it states that powernow-k8 is deprecated since 3.7 and acpi_cpufreq should be used. But acpi_cpufreq says "no such device" when I try to load it, so the cpu runs at full speed all the time. Loading powernow-k8 via a file in /etc/modules-load.d/ works fine, but it's somehow a regression.

Last edited by bjo (2013-09-17 08:45:33)

Offline

#2 2013-09-06 17:37:18

Leonid.I
Member
From: Aethyr
Registered: 2009-03-22
Posts: 999

Re: Kernel 3.11: systemd 204 breaks powerknow-k8 autoloading

There was a similar problem before... I don't have your CPU (only opteron running 3.10.10), so please attach output of

modinfo acpu_cpufreq
modinfo powernow-k8

and most imortantly

udevadm info --export-db

(look for lines like "/devices/system/cpu/cpu0"). I suspect that aliases displayed by modprobe don't match your CPU vendor: and family: returned by udev. If that's true, there is a bug in kernel which was fixed around 3.8 but apparently returned...

Last edited by Leonid.I (2013-09-08 20:55:15)


Arch Linux is more than just GNU/Linux -- it's an adventure
pkill -9 systemd

Offline

#3 2013-09-06 23:14:27

bjo
Member
Registered: 2011-09-10
Posts: 93

Re: Kernel 3.11: systemd 204 breaks powerknow-k8 autoloading

acpi_cpufreq:

alias:          x86cpu:vendor:*:family:*:model:*:feature:*00E8*
alias:          x86cpu:vendor:*:family:*:model:*:feature:*0016*
alias:          acpi*:ACPI0007:*
alias:          acpi*:LNXCPU:*

powernow-k8:

alias:          x86cpu:vendor:0002:family:000F:model:*:feature:*

udevadmin info --export-db

E: DEVPATH=/devices/system/cpu/cpu0
E: MODALIAS=x86cpu:vendor:0002:family:000F:model:006B:feature:,0000,0001,0002,0003,0004,0005,0006,0007,0008,0009,000B,000C,000D,000E,000F,0010,0011,0013,0017,0018,0019,001A,001C,0020,0021,0022,0023,0024,0025,0026,0027,0028,0029,002B,002C,002D,002E,002F,0030,0031,0034,0036,0037,0038,0039,003B,003D,003E,003F,0064,006A,006E,0070,0071,0074,0075,007A,0080,008D,00C0,00C1,00C2,00C3,00C4,00C8,0106

Vendor and family seem to match for powernow-k8.

Offline

#4 2013-09-08 20:53:05

Leonid.I
Member
From: Aethyr
Registered: 2009-03-22
Posts: 999

Re: Kernel 3.11: systemd 204 breaks powerknow-k8 autoloading

Hmm... it seems I was wrong. Could you open a bug report at bugzilla.kernel.org?


Arch Linux is more than just GNU/Linux -- it's an adventure
pkill -9 systemd

Offline

#5 2013-09-09 15:04:34

bjo
Member
Registered: 2011-09-10
Posts: 93

Re: Kernel 3.11: systemd 204 breaks powerknow-k8 autoloading

Offline

#6 2013-09-09 21:03:01

bjo
Member
Registered: 2011-09-10
Posts: 93

Re: Kernel 3.11: systemd 204 breaks powerknow-k8 autoloading

Offline

#7 2013-09-09 21:17:43

bjo
Member
Registered: 2011-09-10
Posts: 93

Re: Kernel 3.11: systemd 204 breaks powerknow-k8 autoloading

Ok, I'll write 100 times on the wall "never pick packages from testing" - updating systemd helped.

Offline

#8 2013-09-17 07:55:18

bjo
Member
Registered: 2011-09-10
Posts: 93

Re: Kernel 3.11: systemd 204 breaks powerknow-k8 autoloading

I think this will cause a bug now, because 3.11 went from testing to core, but systemd is still 204 in core.

Offline

#9 2013-09-17 16:13:58

Leonid.I
Member
From: Aethyr
Registered: 2009-03-22
Posts: 999

Re: Kernel 3.11: systemd 204 breaks powerknow-k8 autoloading

bjo wrote:

I think this will cause a bug now, because 3.11 went from testing to core, but systemd is still 204 in core.

Right. I'll send an email to [arch-general] if you don't mind...


Arch Linux is more than just GNU/Linux -- it's an adventure
pkill -9 systemd

Offline

#10 2013-09-17 16:40:44

bjo
Member
Registered: 2011-09-10
Posts: 93

Re: Kernel 3.11: systemd 204 breaks powerknow-k8 autoloading

Sure.

Offline

Board footer

Powered by FluxBB