The plan is simple: write short, honest posts about the things I work on — operating systems, memory safety, and system security — plus the occasional note on papers I’m reading and lessons from research life. Nothing here is meant to be a polished paper; it’s a place to think out loud.

For the formal record — publications, research experience, and education — see my CV.

If something here is wrong or could be sharper, tell me. That’s half the point of writing it down.

Why is that dangerous? Because the allocator will happily hand that same address to the next allocation. Now two pieces of code believe they own the same bytes. A read through the stale pointer can leak whatever the new owner put there; a write can corrupt it. With careful heap grooming, that primitive turns into control-flow hijacking.

What makes UAF stubborn is that the bug and the exploit are separated in time. Freeing early is often correct; the mistake is holding a reference you forgot about. Static analysis struggles with aliasing, and purely runtime defenses have to decide the hard question: when is it truly safe to reuse an address?

That “when is it safe to reuse” question is exactly what a lot of my work is about — sweeping the heap to confirm no dangling references remain before an address is recycled. My CV links the papers.