How to profile a bare metal source code on ARM platform?

Question

How can I profile my ARM code on a device.

It is a bare metal code involves USB and SDH handling, I saw this Code Profiler for ARM but seems to be very slim, I am familiar with DS5 but most of its tools can be used if you are on platform with linux kernel based OS. I compile my code with gcc.

1. My code runs on Cortex-A7 and/or Cortex-A15 cores.
2. As it seems from here those CPUs aren't arm profiler supported.

Answer 1

Cortex-A7 and Cortex-A15 are Cortex-A processors hence you can use the information available in ARM Architecture Reference Manual (ARMv7-A and ARMv7-R edition) - Chapter 12 - The Performance Monitors Extension.

There are several beautiful answers to questions like How to measure program execution time in ARM Cortex-A8 processor? on this site as well.