I am currently looking at the OpenMP accelerator examples provided within the Processor-sdk-linux package. If I understand correctly, the target portion of the example is pushed to all DSP cores.
Is there a way to limit or specify which cores can be used for OpenMP operations?
In my application, I need one of the cores to run a Sys/Bios application and would like to use the other cores as OpenMP accelerators for codec processing. Would that be possible?