Memoro: A Detailed Heap Profiler

Thanks for building, I'll consider in the future. Though, I wish the fracture in C++ compilation wasn't so frustrating.

LLVM has really helped pushed the boundary of compiler feedback. But still, GCC wins consistently in my experiments of performance.

These days it seems you have to realistically target both and their sanitizer capabilities. Of which, both have drastically improved since CLANG said 'Hello world's.

Nice! Page should state the overhead clearly -- don't let people discover that the hard way. The paper says 2-30x, and 3.3-5.7x is typical. I'm not surprised. And it's preferable to state it and not pull any punches -- it sets expectations and lets us take precautions when running it.