Hi,
I'm trying to write benchmark forcing given CPU load (defined based on idle time) and power consumption.
Currently benchmark creates a task containing Task_sleep() function which holds it’s execution and switches processor to idle. Then CPU load is measure by Load_getCPULoad() function.
When I run this benchmark in no boot mode CPU load equals about 12%, but when I run it in SD boot mode (via u-boot) the CPU load equals about 85%.
I wondering what is the reason of different results depending on boot mode?
After suggestions from Vishal Mahaveer I measured if task holding caused by Task_sleep() is the same in each boot modes. I set Task_sleep() function argument to reach 10s hold. As a reasult I reached 10s task holding in no boot mode and about 1s holding in SD boot mode.
Do you have ony other ideas what to do to get the same benchmark results regardless of the boot mode?