|
I am trying to profile my code using the Profiler for CCSv2.0, but I am getting some interesting results and I am not sure what is causing them. I am developing on a 6711DSK. The very first run after loading my program will give one result, and then, after hitting restart to move the pc to the start of the program, running again will give another result that is about half of the original. I am profiling only a single function, and all data is in the SDRAM and I am using 4-level caching. I am wondering if it has something to do with the caching or some other memory consideration, or maybe I am missing something because even with other functions, when the results don't change between executions, the results are still a bit higher than what the assembly code should give. I was also wondering if there is a way to determine the number of memory stalls that are occuring? |