Coderrect is a fast and scalable tool that finds race conditions in complex software.
Coderrect is a created word by combining Code and Correct. It simply means code-correct. Coderrect is pronounced code-rect.
Coderrect is still a development release, but getting more stable.
Coderrect requires Linux-based OS, such as Ubuntu 14+, and CentOS 7+.
To run Coderrect, you will need to compile your code successfully without Coderrect. For example, running make/cmake/Ninja/Bazel will build your project without errors.
Coderrect generates an intermediate representation of your source code in the form of LLVM bitcode (BC) files and then performs sophisticated static analyses based on them to find potential race conditions.
Coderrect produces a single BC file for each build binary (an executable, a static library, or a shared library).
No. Coderrect intercepts your build commands but will not change your build targets. Running Coderrect will generate exactly the same build binaries as w/o Coderrect.
Coderrect supports most common C/C++ compilers (GCC/Clang/ICC). However, internally Coderrect will use clang to generate LLVM bitcode. So, it may not work if your code does not compile with clang, here is an example showing the difference between clang and gcc.
Coderrect also supports Fortran compilers including Flang/GFortran/IFORT.
Coderrect has full support for C and C++ source code through GCC/Clang/ICC, and Fortran source code through Flang/GFortran/IFORT.
Besides the console output, there is an HTML file named index.html generated under ./report directory by default. You can specify the location of the report through “-o <directory>” option on the command line.
Coderrect is currently under Evaluation License, and is subject to change in the future.
Please contact us (firstname.lastname@example.org).
Powered by BetterDocs