driver-api/thermal/intel_powerclamp.rst

6   - Arjan van de Ven <arjan@linux.intel.com>
7   - Jacob Pan <jacob.jun.pan@linux.intel.com>
12 	    - Goals and Objectives
15 	    - Idle Injection
16 	    - Calibration
19 	    - Effectiveness and Limitations
20 	    - Power vs Performance
21 	    - Scalability
22 	    - Calibration
23 	    - Comparison with Alternative Techniques
26 	    - Generic Thermal Layer (sysfs)
27 	    - Kernel APIs (TBD)
32 Consider the situation where a system’s power consumption must be
38 Currently, P-states, T-states (clock modulation), and CPU offlining
41 On Intel CPUs, C-states provide effective power reduction, but so far
45 is to achieve forced and controllable C-state residency.
56 --------------
58 On modern Intel processors (Nehalem or later), package level C-state
68 If the kernel can also inject idle time to the system, then a
69 closed-loop control system can be established that manages package
70 level C-state. The intel_powerclamp driver is conceived as such a
71 control system, where the target set point is a user-selected idle
73 between the actual package level C-state residency ratio and the target idle
80 clamping actions of controlled duty ratio and duration. Each per-CPU
88 scheme to work for both preemptable and non-preemptable kernels.
99 on large scale systems (Westmere system with 80 processors).
118 control parameters. This CPU is referred to as the controlling CPU in
121 hot-plug.
123 In terms of dynamics of the idle control system, package level idle
124 time is considered largely as a non-causal system where its behavior
131 When used in a causal control system, such as a temperature control,
134 PID-based thermal controller can use the powerclamp driver to
141 -----------
144 also true for the ability of a system to enter package level C-states.
156 	This is to offset the error occurring when the system can
165 progress and results, such as on a Westmere system::
214 non-intrusive to the scheduler or the IRQ core code.
218 ------------------
219 Per-CPU kernel threads are started/stopped upon receiving
231 -----------------------------
235 effectiveness. The extreme case would be doing a ping -f to generated
242 When control parameters change at runtime by the controlling CPU, it
245 thus not able to enter package C- states at the expected ratio. But
251 -----------
253 Ivy Bridge system and the 80P Westmere server under 50% idle ratio.
259 On the IVB 8P system, compared to an offline CPU, powerclamp can
280 cur_state returns value -1 instead of 0 which is to avoid confusing
284 - To inject 25% idle time::
286 	$ sudo sh -c "echo 25 > /sys/class/thermal/cooling_device80/cur_state
288 If the system is not busy and has more than 25% idle time already,
292 If the system is busy (spin test below) and has less than 25% natural
297 In this example, 24.1% idle is shown. This helps the system admin or
308    3341 root     -51   0     0    0    0 D   25  0.0   0:01.62 kidle_inject/0
309    3344 root     -51   0     0    0    0 D   25  0.0   0:01.60 kidle_inject/3
310    3342 root     -51   0     0    0    0 D   25  0.0   0:01.61 kidle_inject/1
311    3343 root     -51   0     0    0    0 D   25  0.0   0:01.60 kidle_inject/2
318 control CPU temperature effectively, when no other thermal influence