On Thu, 24 Nov 2011 00:59:08 +0800 Wu Zhangjin <wuzhang...@gmail.com> wrote:
Hello,
Thank you for such a thoughtful and interesting reply :)
> One 'bad' side effect of using MFGPT timer is that the time precision
> is not good enough for making CPUfreq work normally
The only side-effect that I notice so far, is that in /proc/cpuinfo the
bogomips are reported differently (528, not 797).
Other than that, according to 'cpufreq-info' and 'powertop', the CPU frequency
switching seems to work fine with mfgpt and the ondemand governor.
PowerTOP version 1.13 (C) 2007 Intel Corporation
< Detailed C-state information is not P-states (frequencies)
797 Mhz 0.0%
399 Mhz 0.0%
199 Mhz 100.0%
Wakeups-from-idle per second : 123.7 interval: 10.0s
no ACPI power usage estimate available
Top causes for wakeups:
57.5% ( 62.6) [mfgpt0-timer] <interrupt>
18.3% ( 19.9) USB device 2-4 : RTL8187B_WLAN_Adapter (Manufacturer_Realtek)
9.8% ( 10.7) [ehci_hcd:usb2, ohci_hcd:usb4] <interrupt>
9.2% ( 10.0) [kernel core] ehci_irq (ehci_watchdog)
3.2% ( 3.5) rcu_kthread
0.5% ( 0.5) slim
--
With respect,
Roman
~~~~~~~~~~~~~~~~~~~~~~~~~~~
"Stallman had a printer,
with code he could not see.
So he began to tinker,
and set the software free."
signature.asc
Description: PGP signature
