[PATCH] D92307: [analyzer][StdLibraryFunctionsChecker] Fix typos in summaries of mmap and mmap64

Mon Nov 30 05:03:31 PST 2020

martong added a comment.

In D92307#2422468 <https://reviews.llvm.org/D92307#2422468>, @vsavchenko wrote:

> That's a good fix!
> How did this happen though that max value of `off_t` was even used for `fd`.  Seems pretty odd!

I used a semi-automated approach to create these summaries from cppcheck <https://github.com/danmar/cppcheck/blob/main/cfg/posix.cfg> : I translated the XML into C++ source code, but could not do the translation for ranges. E.g., `0:` was just simply copied and I had to manually modify the generated C++ code to have the proper range. And that's where I made the wrong index for the param (cppcheck starts from idx 1, here we start from idx 0). Here the 5th param has the `off_t` type, so I thought we have to get the max for `off_t`.

================
Comment at: clang/lib/StaticAnalyzer/Checkers/StdLibraryFunctionsChecker.cpp:1738
     Optional<QualType> Off64_tTy = lookupTy("off64_t");
-    Optional<RangeInt> Off64_tMax = getMaxValue(Off_tTy);
     // void *mmap64(void *addr, size_t length, int prot, int flags, int fd,
----------------
steakhal wrote:
> BTW this should have referred to `Off64_tTy` instead of `Off_tTy`.
> Not a problem anymore :D
Yep, copy pasta error again, my bad :(

================
Comment at: clang/test/Analysis/std-c-library-posix-crash.c:1
+// RUN: %clang_analyze_cc1 \
+// RUN:   -analyzer-checker=core,apiModeling.StdCLibraryFunctions \
----------------
I like this new test file. Probably we are going to extend this as we advance with the evaluation of these summaries on open-source projects.

Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D92307/new/

https://reviews.llvm.org/D92307