[RFC PATCH v2 0/9] ARM: cpu topology Cortex-A9
vincent.guittot at linaro.org
Mon Nov 14 12:41:11 UTC 2011
The sched_mc feature has been originally designed to improve power
consumption of multi-package system and several architecture functions
are available to tune the topology and the scheduler's parameters when
scheduler rebuilds the sched_domain hierarchy (change the
sched_mc_power_savings level). This patches improve the power
consumption of dual and quad cortex-A9 when the sched_mc_power_savings
is set to 2. The following patches' policy is to accept up to 4
threads (can be configured) in the run queue of a core before starting
to load balance if cpu runs at low frequencies but to accept only 1
thread for high frequencies, which is the normal behaviour. The goal
is to use only one core in light load situation and all cores in heavy
Patches [1-2] modify the ARM cpu topology according to
sched_mc_power_savings value and Cortex id.
Patch  enables ARCH_POWER feature of the scheduler.
Patch  adds arch_scale_freq_power function for ARM platform.
Patches [5-6] modify the cpu_power of CA-9 according to
sched_mc_power_savings' level and core frequency. The main goal is to
increase the capacity of a core when using low cpu frequency in order
to pull tasks on this core. Note that this behaviour is not really
advised but it can be seen as an intermediate step between the use of
cpu hotplug (which is not a power saving feature) and a new load
balancer which will take into account low load situation on dual core.
Patch  ensures that cpu0 is used in priority when only one CPU is running.
Patch  adds some debugfs interface for test purpose.
Patch  ensures that the cpu_power will be updated periodically.
-remove useless start of ilb when the core has capacity.
-add a method (DT, sysfs, ...) to set threshold for using 1 or all cpus for CA-9
*Modify the method to update cpu_power
*There are fewer patches than v1 because some issues are fixed by
patches that has been pushed for 3.2.
*These patches has been tested on snowball and vexpress boards.
*Performance results are similar to v1
More information about the linaro-dev