Pickin' up the Trash

I assume a few of y'all are wondering what I have been up to. Well, I am working less hours on the timing stuff and more hours on my PhD. In fact, I just wrote a pretty basic garbage collector. I want to weave it into my region memory allocator. In fact, after some more thought, this collector is not entirely complete. It does collect unused data, but it makes a few assumptions, mainly that a pointer does not point into the middle of an array/allocated chunk of memory. This is a horribly bad assumption, especially for what I want, a precise (as opposed to a conservative) collector. My collector works by collecting type and function call data at compile time (gcc plugin that I wrote). Then the collector runtime logic is linked into the resulting binary. The logic behind the allocator is a variation of Cheney's Copying Collector Algorithm. Anyways, I'm proud of the collector. The assumption I have made is part of the problem, and corresponding solution, that my advisers and myself want to write-up for a paper. I hope that goes well. And me? Well, I have a few more months (July is when my scholarships die). So I hope I can start thesis writing soon.

-Matt