On Fri, 2026-03-06 at 18:52 +0000, Egas Ribeiro via Gcc wrote: > Hi, > > I'm interested in applying for this year's GSoC to work on extending > GCC's static analyzer, particularly extending C++ support. > > I've had around a dozen patches merged into GCC 16's C++ frontend > since > October, so I'm familiar with the codebase and patch submission > process. > I'm also familiar with GIMPLE and middle-end IR/SSA theory. > > I've read through the analyzer internals docs and the suggested > project > ideas. I was wondering what areas of work would be most relevant > right > now, whether extending C++ support is the highest priority, or would > something like a new checker be more useful?
Hi Egas Given that you already have some familiarity with the C++ frontend, improving C++ support in -fanalyzer would seem the obvious choice for an analyzer task, say, improving exception handling, or "teaching it" about virtual functions. Both relate to dealing with the inheritance hierarchy, which the analyzer doesn't yet "know" about. See the C++ tracker bug linked from the analyzer's wiki page. Clearly you've built GCC from source already - have you tried stepping through the analyzer on trivial C examples? See the links on debugging it on the SummerOfCode wiki page. Hope this is helpful, let me know if you have further questions Dave
