cpufreq scaling appears not to work on overclocked systems

From: Berck E. Nash
Date: Sun Jul 29 2007 - 17:17:16 EST


This is not new, but exists as far back as 2.6.17. I haven't reported
it before because I figured that surely someone else had noticed it, but
since it's still unfixed and I cannot find any mention of it on LKML,
here we go.

I'm running a Core2 Duo 1.86GHz overclocked to 2.56GHz. Everything is
normal with cpufreq scaling disabled. With cpufreq scaling enabled in
the kernel, using any governor, /proc/cpuinfo indicates a maximum of the
rated frequency rather than the actual frequency.

Here it is under 100% utilization, with userspace governor and powernowd:

berck@luna:~$ cat /proc/cpuinfo
processor : 0
vendor_id : GenuineIntel
cpu family : 6
model : 15
model name : Intel(R) Core(TM)2 CPU 6300 @ 1.86GHz
stepping : 6
cpu MHz : 1862.000
cache size : 2048 KB
physical id : 0
siblings : 2
core id : 0
cpu cores : 2
fpu : yes
fpu_exception : yes
cpuid level : 10
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge
mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm syscall
lm constant_tsc pni monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr lahf_lm
bogomips : 5134.39
clflush size : 64
cache_alignment : 64
address sizes : 36 bits physical, 48 bits virtual
power management:

processor : 1
vendor_id : GenuineIntel
cpu family : 6
model : 15
model name : Intel(R) Core(TM)2 CPU 6300 @ 1.86GHz
stepping : 6
cpu MHz : 1862.000
cache size : 2048 KB
physical id : 0
siblings : 2
core id : 1
cpu cores : 2
fpu : yes
fpu_exception : yes
cpuid level : 10
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge
mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm syscall
lm constant_tsc pni monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr lahf_lm
bogomips : 5131.64
clflush size : 64
cache_alignment : 64
address sizes : 36 bits physical, 48 bits virtual
power management:

At idle it drops down to 1596.000 MHz reported. At boot the kernel
doest report the correct frequency (2564.907), which agrees with the
bogomips reported by /proc/cpuinfo as well.

I haven't done any benchmarking to try to determine if perhaps freq
scaling works as it should and just the reported frequency is incorrect.

Berck
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/