I have a new laptop with an AMD Mobile Sempron, and I'm experimenting
with the PowerNow! frequency scaling. It seems to work, but on every
frequency change, everything stalls for 3-4 seconds, even the X mouse
pointer! Keyboard input seems to be buffered though. Is this normal?
I'm currently using the builtin frequency scaling "governors" of the
kernel with the following settings:
echo ondemand > /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor
echo 15000000 > /sys/devices/system/cpu/cpu0/cpufreq/ondemand/sampling_rate
echo 1 > /sys/devices/system/cpu/cpu0/cpufreq/ondemand/ignore_nice
However, using the powernow daemon results in the same behavior.
Here are the relevant kernel (2.6.13.2) options:
#
# CPU Frequency scaling
#
CONFIG_CPU_FREQ=y
CONFIG_CPU_FREQ_TABLE=y
CONFIG_CPU_FREQ_DEBUG=y
CONFIG_CPU_FREQ_STAT=y
CONFIG_CPU_FREQ_STAT_DETAILS=y
# CONFIG_CPU_FREQ_DEFAULT_GOV_PERFORMANCE is not set
CONFIG_CPU_FREQ_DEFAULT_GOV_USERSPACE=y
CONFIG_CPU_FREQ_GOV_PERFORMANCE=y
CONFIG_CPU_FREQ_GOV_POWERSAVE=y
CONFIG_CPU_FREQ_GOV_USERSPACE=y
CONFIG_CPU_FREQ_GOV_ONDEMAND=y
CONFIG_CPU_FREQ_GOV_CONSERVATIVE=y
#
# CPUFreq processor drivers
#
CONFIG_X86_ACPI_CPUFREQ=y
# CONFIG_X86_POWERNOW_K6 is not set
# CONFIG_X86_POWERNOW_K7 is not set
CONFIG_X86_POWERNOW_K8=y
CONFIG_X86_POWERNOW_K8_ACPI=y
# CONFIG_X86_GX_SUSPMOD is not set
# CONFIG_X86_SPEEDSTEP_CENTRINO is not set
# CONFIG_X86_SPEEDSTEP_ICH is not set
# CONFIG_X86_SPEEDSTEP_SMI is not set
# CONFIG_X86_P4_CLOCKMOD is not set
# CONFIG_X86_CPUFREQ_NFORCE2 is not set
# CONFIG_X86_LONGRUN is not set
# CONFIG_X86_LONGHAUL is not set
This is reported in /var/log/messages:
Oct 3 22:04:35 acer kernel: CPU: AMD Mobile AMD Sempron(tm) Processor 3000+ stepping 02
[...]
Oct 3 22:04:35 acer kernel: powernow-k8: Found 1 AMD Athlon 64 / Opteron processors (version 1.50.3)
Oct 3 22:04:35 acer kernel: powernow-k8: 0 : fid 0xa (1800 MHz), vid 0xa (1300 mV)
Oct 3 22:04:35 acer kernel: powernow-k8: 1 : fid 0x8 (1600 MHz), vid 0xc (1250 mV)
Oct 3 22:04:35 acer kernel: powernow-k8: 2 : fid 0x0 (800 MHz), vid 0x13 (1075 mV)
Oct 3 22:04:35 acer kernel: cpu_init done, current fid 0xa, vid 0xa
--
Dick Streefland //// Altium BV
[email protected] (@ @) http://www.altium.com
--------------------------------oOO--(_)--OOo---------------------------
[email protected] (Dick Streefland) wrote:
| I have a new laptop with an AMD Mobile Sempron, and I'm experimenting
| with the PowerNow! frequency scaling. It seems to work, but on every
| frequency change, everything stalls for 3-4 seconds, even the X mouse
| pointer! Keyboard input seems to be buffered though. Is this normal?
I found out that the pauses only occur when the X server is running,
not when I work on the console, or remotely. I can easily reproduce
the problem by playing an mp3 file in a console with mpg321, and then
running burnK7 to force frequency scaling. Without X server running,
everything works fine, but when the X server is running, the music
skips on frequency changes. However, when I start the X server, but
switch back to the console with CTRL-ALT-F1, there are no skips.
Apparently, some Ubuntu users experienced the same problem:
http://ubuntuforums.org/archive/index.php/t-17816.html
https://bugzilla.ubuntu.com/show_bug.cgi?id=7121
I already recompiled the kernel with CONFIG_PREEMPT, but that doesn't
help. Is this some timing issue in the X server (X.org 6.8.2)? Is
there something else to try?
--
Dick Streefland //// Altium BV
[email protected] (@ @) http://www.altium.com
--------------------------------oOO--(_)--OOo---------------------------
[email protected] (Dick Streefland) wrote:
| I found out that the pauses only occur when the X server is running,
| not when I work on the console, or remotely. I can easily reproduce
| the problem by playing an mp3 file in a console with mpg321, and then
| running burnK7 to force frequency scaling. Without X server running,
| everything works fine, but when the X server is running, the music
| skips on frequency changes. However, when I start the X server, but
| switch back to the console with CTRL-ALT-F1, there are no skips.
It's me again :-). I solved the problem by changing the X.org video
driver from "vesa" to "sis". So it looks like the vesa driver (or
frame buffer?) is doing something funky.
--
Dick Streefland //// Altium BV
[email protected] (@ @) http://www.altium.com
--------------------------------oOO--(_)--OOo---------------------------