Vulnerability research, root cause analysis, reverse engineering and exploit development.
A deep dive into CVE-2025-53630 in Llama.cpp, analyzing how malformed GGUF model inputs lead to a crash. This post walks through reversing the parser, identifying the root cause, and understanding how the bug can be triggered in practice.
→
An in-depth analysis of CVE-2024-22857 in zlog, focusing on the underlying memory corruption issue. I traced the bug from input handling to crash, using debugging techniques to uncover the root cause and assess its impact.
→
This post explores CVE-2017-14493 in dnsmasq, following the path from crafted network packets to application crash. Through dynamic analysis and code inspection, we uncover the root cause and examine how the vulnerability is triggered.
→