Hello, I hope you're doing well.
I wanted to briefly follow up on my previous emails regarding my GSoC proposals for the GCC static analyzer. Since the application deadline is approaching, I would appreciate any quick guidance on which direction is more aligned with current priorities: 1. Improving detection of out-of-bounds accesses for Flexible Array Members (FAMs) 2. Enhancing support for printf-style formatted functions in -fanalyzer I would be grateful for any feedback or preference from the community. Thank you for your time. Best regards, Virginia Kodsy On Mon, Mar 23, 2026 at 8:46 PM Virginia Hany <[email protected]> wrote: > > Hello again, > > Following up on my email from yesterday, I have finished drafting the second > proposal I mentioned regarding support for printf-style formatted functions > in -fanalyzer. > > This project focuses on refactoring the existing format-string parsing logic > (from c-format.cc and gimple-ssa-sprintf.cc) into a unified internal API. > This would allow the analyzer to perform path-sensitive buffer overflow > detection for sprintf and fprintf, addressing PR107017. > > Link to Printf-style Proposal: > https://drive.google.com/file/d/1x2DvcwNoNs-we24EMOdHyas9sTAO1DhA/view?usp=sharing > > Now that I have shared both the FAM (Flexible Array Member) proposal and this > printf/refactoring proposal, I would appreciate your guidance on which > direction the community feels is more impactful for the analyzer's roadmap. > > Should I focus on refining one of these exclusively, or is it helpful to > continue developing both at this stage? > > Thank you for your time and feedback. > > Best regards, > Virginia Kodsy > > On Sun, Mar 22, 2026 at 1:39 PM Virginia Hany <[email protected]> wrote: >> >> Hello GCC community, >> >> My name is Virginia Kodsy, and I am interested in applying for Google Summer >> of Code 2026 to work on the GCC static analyzer (-fanalyzer). >> >> Over the past few weeks, I have been exploring the analyzer’s internals and >> working on small contributions to gain familiarity with the codebase. >> Specifically, I have been implementing models for functions like getenv and >> strcmp within the known_function framework, which has helped me understand >> region_model, svalue types, and constraint handling. >> >> I have prepared a draft proposal focused on improving the detection of >> out-of-bounds accesses for Flexible Array Members (FAMs). >> >> Project Title: Improving Detection of Out-of-Bounds Accesses for FAMs in GCC >> Static Analyzer >> >> Brief Summary: The project aims to enhance symbolic capacity tracking and >> constraint propagation to better detect OOB accesses in FAMs, particularly >> in complex cases involving symbolic allocation sizes and realloc patterns >> where the current analyzer often loses track of region bounds. >> >> Draft Proposal: >> https://drive.google.com/drive/folders/1hcfYmvJ7mSvdpp7c4V7ChFZAN7sv6j8h?usp=sharing >> >> I would greatly appreciate your feedback on: >> 1. The technical feasibility of the proposed approach for tracking symbolic >> FAM sizes. >> 2. Whether the scope is appropriate for a GSoC timeline or if it should be >> narrowed/expanded. >> 3. Any specific edge cases in FAM handling that you believe should be >> prioritized. >> >> I also have an interest in improving general string handling support. While >> I am prepared to submit multiple proposals, would you recommend focusing my >> efforts on refining this FAM proposal to a higher standard instead? >> >> Thank you for your time and guidance. >> >> Best regards, >> Virginia Kodsy
