1 CPU frequency and voltage scaling code in the Linux(TM) kernel
4 L i n u x C P U F r e q
9 Dominik Brodowski <linux@brodo.de>
10 David Kimdon <dwhedon@debian.org>
11 Rafael J. Wysocki <rafael.j.wysocki@intel.com>
12 Viresh Kumar <viresh.kumar@linaro.org>
16 Clock scaling allows you to change the clock speed of the CPUs on the
17 fly. This is a nice method to save battery power, because the lower
18 the clock speed, the less power the CPU consumes.
23 1. CPUFreq core and interfaces
25 3. CPUFreq Table Generation with Operating Performance Point (OPP)
27 1. General Information
28 =======================
30 The CPUFreq core code is located in drivers/cpufreq/cpufreq.c. This
31 cpufreq code offers a standardized interface for the CPUFreq
32 architecture drivers (those pieces of code that do actual
33 frequency transitions), as well as to "notifiers". These are device
34 drivers or other part of the kernel that need to be informed of
35 policy changes (ex. thermal modules like ACPI) or of all
36 frequency changes (ex. timing code) or even need to force certain
37 speed limits (like LCD drivers on ARM architecture). Additionally, the
38 kernel "constant" loops_per_jiffy is updated on frequency changes
41 Reference counting of the cpufreq policies is done by cpufreq_cpu_get
42 and cpufreq_cpu_put, which make sure that the cpufreq driver is
43 correctly registered with the core, and will not be unloaded until
44 cpufreq_put_cpu is called. That also ensures that the respective cpufreq
45 policy doesn't get freed while being used.
50 CPUFreq notifiers conform to the standard kernel notifier interface.
51 See linux/include/linux/notifier.h for details on notifiers.
53 There are two different CPUFreq notifiers - policy notifiers and
57 2.1 CPUFreq policy notifiers
58 ----------------------------
60 These are notified when a new policy is intended to be set. Each
61 CPUFreq policy notifier is called twice for a policy transition:
63 1.) During CPUFREQ_ADJUST all CPUFreq notifiers may change the limit if
64 they see a need for this - may it be thermal considerations or
67 2.) And during CPUFREQ_NOTIFY all notifiers are informed of the new policy
68 - if two hardware drivers failed to agree on a new policy before this
69 stage, the incompatible hardware shall be shut down, and the user
72 The phase is specified in the second argument to the notifier.
74 The third argument, a void *pointer, points to a struct cpufreq_policy
75 consisting of several values, including min, max (the lower and upper
76 frequencies (in kHz) of the new policy).
79 2.2 CPUFreq transition notifiers
80 --------------------------------
82 These are notified twice for each online CPU in the policy, when the
83 CPUfreq driver switches the CPU core frequency and this change has no
84 any external implications.
86 The second argument specifies the phase - CPUFREQ_PRECHANGE or
89 The third argument is a struct cpufreq_freqs with the following
91 cpu - number of the affected CPU
94 flags - flags of the cpufreq driver
96 3. CPUFreq Table Generation with Operating Performance Point (OPP)
97 ==================================================================
98 For details about OPP, see Documentation/power/opp.txt
100 dev_pm_opp_init_cpufreq_table - cpufreq framework typically is initialized with
101 cpufreq_table_validate_and_show() which is provided with the list of
102 frequencies that are available for operation. This function provides
103 a ready to use conversion routine to translate the OPP layer's internal
104 information about the available frequencies into a format readily
105 providable to cpufreq.
107 WARNING: Do not use this function in interrupt context.
113 r = dev_pm_opp_init_cpufreq_table(dev, &freq_table);
115 cpufreq_table_validate_and_show(policy, freq_table);
116 /* Do other things */
119 NOTE: This function is available only if CONFIG_CPU_FREQ is enabled in
120 addition to CONFIG_PM_OPP.
122 dev_pm_opp_free_cpufreq_table - Free up the table allocated by dev_pm_opp_init_cpufreq_table