GCC 8/9 vs. LLVM Clang 7/8 Compiler Performance On AArch64

Compiling the GCC compiler itself was fastest using GCC.

But for the timed compile workloads of building the ImageMagick program as well as LLVM, Clang compiled those code-bases the fastest.

As seen with POWER9 and x86_64, the C-Ray ray-tracer was the fastest when built under GCC.

The benchmarks were very close with the Bullet physics engine.