cpufreq on Toshiba Portege M100 Laptop
Hello,
this week I got a Toshiba Portege M100 Laptop. I installed Debian (sid) and everything was great. I could see the P-States on powertop, which were 600MHz - 1200 MHz in 100MHz steps. Then I upgraded the firmware of the Laptop and then it happend:
acpi-cpufreq: acpi_cpufreq_init
acpi-cpufreq: acpi_cpufreq_early_init
cpufreq-core: trying to register driver acpi-cpufreq
cpufreq-core: adding CPU 0
acpi-cpufreq: acpi_cpufreq_cpu_init
acpi-cpufreq: No P-States 1
cpufreq-core: initialization failed
cpufreq-core: no CPU initialized for driver acpi-cpufreq
cpufreq-core: unregistering CPU 0
That 1 in the line after "No P-States" is the value of perf->state_count from acpi_cpufreq_cpu_init(), which I added to acpi-cpufreq to see whats wrong. I used the latest kernel-2.6.22-git2 via git. Then I tryed cpeedstep-centrino with this result.
cpufreq-core: trying to register driver centrino
cpufreq-core: adding CPU 0
speedstep-centrino: No P-States
speedstep-centrino: speedstep-centrino: invalid ACPI data
speedstep-centrino: found "Intel(R) Pentium(R) M processor 1200MHz": max frequency: 1200000kHz
speedstep-centrino: trying to enable Enhanced SpeedStep (111088)
speedstep-centrino: couldn't enable Enhanced SpeedStep
cpufreq-core: initialization failed
cpufreq-core: no CPU initialized for driver centrino
cpufreq-core: unregistering CPU 0
So can you help/tell me where i have to look for this error.
Before the firmware upgrade I made hexdump -C of the first MB of /dev/mem, maybe that could help.
Btw, PFX should not be used in dprintk(), cause speedstep-centrino: is printed twice. I'll write a patch for that.
Greets, Henne.
P.S. Here's the output of /proc/cpuinfo
processor : 0vendor_id : GenuineIntel
cpu family : 6
model : 9
model name : Intel(R) Pentium(R) M processor 1200MHz
stepping : 5
cpu MHz : 598.516
cache size : 1024 KBfdiv_bug : nohlt_bug : nof00f_bug : nocoma_bug : no
fpu : yesfpu_exception : yes
cpuid level : 2
wp : yes
flags : fpu vme de pse tsc msr mce cx8 sep mtrr pge mca cmov pat clflush dts acpi mmx fxsr sse sse2 tm pbe est tm2
bogomips : 1197.79
clflush size : 64
Quoted 26 lines
-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Henne
Sent: Thursday, July 12, 2007 4:42 AM
To: [email protected]
Subject: cpufreq on Toshiba Portege M100 LaptopHello,
this week I got a Toshiba Portege M100 Laptop. I installed Debian (sid) and everything was great.
I could see the P-States on powertop, which were 600MHz - 1200 MHz in 100MHz steps.
Then I upgraded the firmware of the Laptop and then it happend:acpi-cpufreq:
acpi_cpufreq_init
acpi-cpufreq:acpi_cpufreq_early_init
cpufreq-core: trying to register driver acpi-cpufreq
cpufreq-core: adding CPU 0
acpi-cpufreq:acpi_cpufreq_cpu_init
acpi-cpufreq: No P-States 1
cpufreq-core: initialization failed
cpufreq-core: no CPU initialized for driver acpi-cpufreq
cpufreq-core: unregistering CPU 0
Quoted 5 lines
That 1 in the line after "No P-States" is the value of
perf->state_countfromacpi_cpufreq_cpu_init(), which I added to acpi-cpufreq to see whats wrong. I used the latest kernel-2.6.22-git2 via git.
Hmmm.
Looks like the new BIOS decided to take away the support for Speedstep. Can you look into the BIOS setup menu and check whether there are any options to enable speedstep.
Thanks,
Venki
Quoted 4 lines · Venkatesh
Hmmm.
Looks like the new BIOS decided to take away the support for Speedstep. Can you look into the BIOS setup menu and check whether there are any options to enable speedstep.
There are 3 switches that seem to be related to it:
Processing Speed: High/Low
CPU Sleep Mode: Enabled/Disabled
Cooling Method: Battery optimized/Performance/Maximum Performance
Eventually the option:
Device Config.: All Devices/Setup by OS
may have something to do with that, but I think every BIOS does there what it wants.
I tried every possible configuration and made acpidump's, but none worked with cpufreq.
Here's a snippet from Toshiba.com BIOS changelog:
Change History
Version 1.50 09-29-2005
o Applied an Intel-required Pentium-M update.
o Re-packaged on 2004-11-30. No BIOS code is changed. If you have already installed BIOS v1.50, there's no need to reinstall.
That could have been it.
Maybe I should try to downgrade to V1.40 if possible, but I first have to make sure that this doesn't break anything and if this is supported.