I agree that you should have the same utilization figures for 2 CPUs with same uarch at same OPP whatever the max OPP of each CPU is. Nevertheless, scale_cpu will not be 1024 for both CPU because arch_scale_cpu take into account both uarch and max freq. In the same way, scale_freq will be also dfifferent
Let use the example of 2 clusters with same uarch but clusterB max OPP is half cluster clusterA max OPP. If booth cluster are running at same freq, scale_cpu for clusterA will be 1024 and scale_freq will be 512 (because half of its max freq) scale_cpu for clusterB will be 512 (because its max compute capacity is half max compute capacity if the system) and scale_freq will be 1024
Thanks Vincent. In conjunction with scale_freq, this makes sense.
-- Sent by an employee of the Qualcomm Innovation Center, Inc. The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum, hosted by The Linux Foundation.