I would like to hire a Coder

I would like to hire an Clang/LLVM developer. See here problem description: Ever wanted to use reliable software in which spy software can not be implanted? Well, you are just right, if you want to develop a LLVM-based tool which can mitigate a major source of memory corruptions [4] (i.e., integer overflows (IOs)). IO detection is an unsolved problem since more than 30 years. Real live software (e.g., Linux kernel, Chrome browser, etc.) suffer heavilly from these types of memory corruptions, mostly because of the unsafe used programming languages (e.g., C/C++). Furthermore, some of the IOs can be exploited in oder to perform advanced code reuse attacks or implant malicious software into systems. The goal of this thesis is to develop an integer overflow detection tool which can detect IO bugs in C/C++ code and has better accuracy than existing stat-of-the-art tools, IOC [2], IntFlow [3] and IntEQ [1]. The tool should be based on static information flow tracking and dynamic program analysis and reuse the information flow tracking used in the previous tools. First, we want to extend the scope of the static analysis (i.e., make the analysis interprocedural by porting IOC to the LLVM link time) during LLVM link time through an LLVM pass which can as IntEQ extract paths and compare them in order to filter out benign from harmful paths (i.e., contain an integer overflow bug). Second, loops should be treated with care—for example—loops can be unrolled with varying number of depths (e.g., 1-1k-10k-100k-1000k) in order to increase precision of our tool. Third, chains of sinks and sources (e.g., sink—source—sink—...) can be analyzed in order to determine complex relations between those program locations. Last but not least, other techniques can be imagined and implemented. The tool will be evaluated using the Linux kernel, web browsers and the SPEC CPU 2006 benchmark. Finally, we will compare our tool w.r.t. the previously introduced state-of-the-art tools.