Enable zero-overhead loop instructions (eliminates loop pipeline overhead)
These instructions enable looping with no per-iteration cycle overhead. The use of the loop instructions set up a loop trip count, beginning PC and ending PC. Whenever the hardware executes the last instruction in the loop and the loop trip count has not been exceeded, control is automatically transferred back to the first instruction without needing any explicit branch instructions. Because every taken branch on Cadence processors requires 3 to 5 cycles to execute, the zero overhead loop instructions are extremely useful for code that spends time in loops. When compiling with optimization, the compiler will typically use these instructions for every loop that does not contain a function call.