Code Optimization
Four relatively simple modifications of assembly code can be done to generate a more efficient code. These modifications make use of the available C6x resources such as multiple buses, functional units, pipelined CPU, and memory organization. They include (a) using parallel instructions, (b) eliminating delays or NOPs, (c) unrolling loops, and (d) using word-wide data.
Wherever possible, parallel instructions should be used to make maximum use of idle functional units. It should be noted that, whenever the order in which instructions appear is important, care must be taken not to have any dependency in the operands of the instructions within a parallel instruction.
It may become necessary to have cross paths when ...
Get Real-Time Digital Signal Processing now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
