diff options
author | Zhang Rui <rui.zhang@intel.com> | 2008-01-17 15:51:23 +0800 |
---|---|---|
committer | Len Brown <len.brown@intel.com> | 2008-02-01 23:18:19 -0500 |
commit | d9460fd227ed2ce52941b6a12ad4de05c195f6aa (patch) | |
tree | 1fb877cf41c8599e0fc9c1fbf902b88baba60708 /COPYING | |
parent | 05a83d972293f39a66bc2aa409a5e7996bba585d (diff) | |
download | linux-d9460fd227ed2ce52941b6a12ad4de05c195f6aa.tar.gz linux-d9460fd227ed2ce52941b6a12ad4de05c195f6aa.tar.bz2 linux-d9460fd227ed2ce52941b6a12ad4de05c195f6aa.zip |
ACPI: register ACPI Processor as generic thermal cooling device
Register ACPI processor as thermal cooling devices.
A combination of processor T-state and P-state are used for thermal throttling.
the processor will reduce the frequency first and then set the T-state.
we use cpufreq_thermal_reduction_pctg to calculate the cpufreq limit,
and call cpufreq_verify_with_limit to set the cpufreq limit.
if cpufreq driver is loaded, then we have four cooling state for cpufreq control.
cooling state 0: cpufreq limit == max_freq
cooling state 1: cpufreq limit == max_freq * 80%
cooling state 2: cpufreq limit == max_freq * 60%
cooling state 3: cpufreq limit == max_freq * 40%
after the cpufreq limit is set to 40 percentage of the max_freq,
we use T-state for cooling.
eg. a processor has P-state support, and it has 8 T-state (T0-T7),
the max_state of the proceesor is 10:
state cpufreq-limit T-state
0: max_freq T0
1: max_freq * 80% T0
2: max_freq * 60% T0
3: max_freq * 40% T0
4: max_freq * 40% T1
5: max_freq * 40% T2
6: max_freq * 40% T3
7: max_freq * 40% T4
8: max_freq * 40% T5
9: max_freq * 40% T6
10: max_freq * 40% T7
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Zhao Yakui <yakui.zhao@intel.com>
Signed-off-by: Thomas Sujith <sujith.thomas@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
Diffstat (limited to 'COPYING')
0 files changed, 0 insertions, 0 deletions