Skip to content
Commit 08374410 authored by Lukasz Luba's avatar Lukasz Luba Committed by Rafael J. Wysocki
Browse files

Documentation: power: Describe 'advanced' and 'simple' EM models



The Energy Model (EM) can be registered in two ways:

 1) Using a helper function, which under the hood relies on OPP framework
    and DT entry in CPU node: 'dynamic-power-coefficient'. This is
    a 'simple' EM because it's tied to the math formula:
    Power = dynamic-power-coefficient * V^2 * f

 2) Using em_dev_register_perf_domain() API function with a driver
    custom callback which provides power for each performance state.

    This is 'advanced' EM, since it can better reflect real power
    measurements for each performance state. It's not limited to any
    math formula and can better reflect real physics of the device.

Add description of these two methods to the documentation, so developers
could choose the suitable registration method (option).

Signed-off-by: default avatarLukasz Luba <lukasz.luba@arm.com>
[ rjw: Changelog edits ]
Signed-off-by: default avatarRafael J. Wysocki <rafael.j.wysocki@intel.com>
parent d704aa0d
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment