[PATCH] D99260: [analyzer] Fix false positives in inner pointer checker (PR49628)

[Artem Dergachev via Phabricator via cfe-commits]

NoQ added a comment.

> It does make sense to split these in two, but I'm not so sure about `evalCall`.

`evalCall` should work great for this purpose but definitely not in this checker. Also this checker would still exercise its `checkPostCall` so it would still need to be silenced even if `evalCall` is implemented.



================
Comment at: clang/lib/StaticAnalyzer/Checkers/InnerPointerChecker.cpp:234
+      ObjRegion =
+          dyn_cast_or_null<TypedValueRegion>(Call.getArgSVal(0).getAsRegion());
     }
----------------
This will crash if someone overloads `std::data` with 0 arguments because your `CallDescription`s don't require a specific number of arguments. (They're not allowed to do that, yes, but we're still not allowed to crash.)


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D99260/new/

https://reviews.llvm.org/D99260