tjh.dev
Linux kernel mirror (for testing)
git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
kernel
os
linux
1 CPU frequency and voltage scaling code in the Linux(TM) kernel
2
3
4 L i n u x C P U F r e q
5
6 C P U F r e q G o v e r n o r s
7
8 - information for users and developers -
9
10
11 Dominik Brodowski <linux@brodo.de>
12 some additions and corrections by Nico Golde <nico@ngolde.de>
13
14
15
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.
19
20
21Contents:
22---------
231. What is a CPUFreq Governor?
24
252. Governors In the Linux Kernel
262.1 Performance
272.2 Powersave
282.3 Userspace
292.4 Ondemand
302.5 Conservative
31
323. The Governor Interface in the CPUfreq Core
33
34
35
361. What Is A CPUFreq Governor?
37==============================
38
39Most cpufreq drivers (in fact, all except one, longrun) or even most
40cpu frequency scaling algorithms only offer the CPU to be set to one
41frequency. In order to offer dynamic frequency scaling, the cpufreq
42core must be able to tell these drivers of a "target frequency". So
43these specific drivers will be transformed to offer a "->target"
44call instead of the existing "->setpolicy" call. For "longrun", all
45stays the same, though.
46
47How to decide what frequency within the CPUfreq policy should be used?
48That's done using "cpufreq governors". Two are already in this patch
49-- they're the already existing "powersave" and "performance" which
50set the frequency statically to the lowest or highest frequency,
51respectively. At least two more such governors will be ready for
52addition in the near future, but likely many more as there are various
53different theories and models about dynamic frequency scaling
54around. Using such a generic interface as cpufreq offers to scaling
55governors, these can be tested extensively, and the best one can be
56selected for each specific use.
57
58Basically, it's the following flow graph:
59
60CPU can be set to switch independently | CPU can only be set
61 within specific "limits" | to specific frequencies
62
63 "CPUfreq policy"
64 consists of frequency limits (policy->{min,max})
65 and CPUfreq governor to be used
66 / \
67 / \
68 / the cpufreq governor decides
69 / (dynamically or statically)
70 / what target_freq to set within
71 / the limits of policy->{min,max}
72 / \
73 / \
74 Using the ->setpolicy call, Using the ->target call,
75 the limits and the the frequency closest
76 "policy" is set. to target_freq is set.
77 It is assured that it
78 is within policy->{min,max}
79
80
812. Governors In the Linux Kernel
82================================
83
842.1 Performance
85---------------
86
87The CPUfreq governor "performance" sets the CPU statically to the
88highest frequency within the borders of scaling_min_freq and
89scaling_max_freq.
90
91
922.2 Powersave
93-------------
94
95The CPUfreq governor "powersave" sets the CPU statically to the
96lowest frequency within the borders of scaling_min_freq and
97scaling_max_freq.
98
99
1002.3 Userspace
101-------------
102
103The CPUfreq governor "userspace" allows the user, or any userspace
104program running with UID "root", to set the CPU to a specific frequency
105by making a sysfs file "scaling_setspeed" available in the CPU-device
106directory.
107
108
1092.4 Ondemand
110------------
111
112The CPUfreq governor "ondemand" sets the CPU depending on the
113current usage. To do this the CPU must have the capability to
114switch the frequency very quickly. There are a number of sysfs file
115accessible parameters:
116
117sampling_rate: measured in uS (10^-6 seconds), this is how often you
118want the kernel to look at the CPU usage and to make decisions on
119what to do about the frequency. Typically this is set to values of
120around '10000' or more. It's default value is (cmp. with users-guide.txt):
121transition_latency * 1000
122Be aware that transition latency is in ns and sampling_rate is in us, so you
123get the same sysfs value by default.
124Sampling rate should always get adjusted considering the transition latency
125To set the sampling rate 750 times as high as the transition latency
126in the bash (as said, 1000 is default), do:
127echo `$(($(cat cpuinfo_transition_latency) * 750 / 1000)) \
128 >ondemand/sampling_rate
129
130show_sampling_rate_min:
131The sampling rate is limited by the HW transition latency:
132transition_latency * 100
133Or by kernel restrictions:
134If CONFIG_NO_HZ is set, the limit is 10ms fixed.
135If CONFIG_NO_HZ is not set or no_hz=off boot parameter is used, the
136limits depend on the CONFIG_HZ option:
137HZ=1000: min=20000us (20ms)
138HZ=250: min=80000us (80ms)
139HZ=100: min=200000us (200ms)
140The highest value of kernel and HW latency restrictions is shown and
141used as the minimum sampling rate.
142
143show_sampling_rate_max: THIS INTERFACE IS DEPRECATED, DON'T USE IT.
144
145up_threshold: defines what the average CPU usage between the samplings
146of 'sampling_rate' needs to be for the kernel to make a decision on
147whether it should increase the frequency. For example when it is set
148to its default value of '95' it means that between the checking
149intervals the CPU needs to be on average more than 95% in use to then
150decide that the CPU frequency needs to be increased.
151
152ignore_nice_load: this parameter takes a value of '0' or '1'. When
153set to '0' (its default), all processes are counted towards the
154'cpu utilisation' value. When set to '1', the processes that are
155run with a 'nice' value will not count (and thus be ignored) in the
156overall usage calculation. This is useful if you are running a CPU
157intensive calculation on your laptop that you do not care how long it
158takes to complete as you can 'nice' it and prevent it from taking part
159in the deciding process of whether to increase your CPU frequency.
160
161sampling_down_factor: this parameter controls the rate at which the
162kernel makes a decision on when to decrease the frequency while running
163at top speed. When set to 1 (the default) decisions to reevaluate load
164are made at the same interval regardless of current clock speed. But
165when set to greater than 1 (e.g. 100) it acts as a multiplier for the
166scheduling interval for reevaluating load when the CPU is at its top
167speed due to high load. This improves performance by reducing the overhead
168of load evaluation and helping the CPU stay at its top speed when truly
169busy, rather than shifting back and forth in speed. This tunable has no
170effect on behavior at lower speeds/lower CPU loads.
171
172
1732.5 Conservative
174----------------
175
176The CPUfreq governor "conservative", much like the "ondemand"
177governor, sets the CPU depending on the current usage. It differs in
178behaviour in that it gracefully increases and decreases the CPU speed
179rather than jumping to max speed the moment there is any load on the
180CPU. This behaviour more suitable in a battery powered environment.
181The governor is tweaked in the same manner as the "ondemand" governor
182through sysfs with the addition of:
183
184freq_step: this describes what percentage steps the cpu freq should be
185increased and decreased smoothly by. By default the cpu frequency will
186increase in 5% chunks of your maximum cpu frequency. You can change this
187value to anywhere between 0 and 100 where '0' will effectively lock your
188CPU at a speed regardless of its load whilst '100' will, in theory, make
189it behave identically to the "ondemand" governor.
190
191down_threshold: same as the 'up_threshold' found for the "ondemand"
192governor but for the opposite direction. For example when set to its
193default value of '20' it means that if the CPU usage needs to be below
19420% between samples to have the frequency decreased.
195
1963. The Governor Interface in the CPUfreq Core
197=============================================
198
199A new governor must register itself with the CPUfreq core using
200"cpufreq_register_governor". The struct cpufreq_governor, which has to
201be passed to that function, must contain the following values:
202
203governor->name - A unique name for this governor
204governor->governor - The governor callback function
205governor->owner - .THIS_MODULE for the governor module (if
206 appropriate)
207
208The governor->governor callback is called with the current (or to-be-set)
209cpufreq_policy struct for that CPU, and an unsigned int event. The
210following events are currently defined:
211
212CPUFREQ_GOV_START: This governor shall start its duty for the CPU
213 policy->cpu
214CPUFREQ_GOV_STOP: This governor shall end its duty for the CPU
215 policy->cpu
216CPUFREQ_GOV_LIMITS: The limits for CPU policy->cpu have changed to
217 policy->min and policy->max.
218
219If you need other "events" externally of your driver, _only_ use the
220cpufreq_governor_l(unsigned int cpu, unsigned int event) call to the
221CPUfreq core to ensure proper locking.
222
223
224The CPUfreq governor may call the CPU processor driver using one of
225these two functions:
226
227int cpufreq_driver_target(struct cpufreq_policy *policy,
228 unsigned int target_freq,
229 unsigned int relation);
230
231int __cpufreq_driver_target(struct cpufreq_policy *policy,
232 unsigned int target_freq,
233 unsigned int relation);
234
235target_freq must be within policy->min and policy->max, of course.
236What's the difference between these two functions? When your governor
237still is in a direct code path of a call to governor->governor, the
238per-CPU cpufreq lock is still held in the cpufreq core, and there's
239no need to lock it again (in fact, this would cause a deadlock). So
240use __cpufreq_driver_target only in these cases. In all other cases
241(for example, when there's a "daemonized" function that wakes up
242every second), use cpufreq_driver_target to lock the cpufreq per-CPU
243lock before the command is passed to the cpufreq processor driver.
244