We are excited to announce the release of Codee 1.4, which further improves the capabilities of our static code analyzer specifically designed to boost the performance of C/C++/Fortran applications.
In this release we keep improving our Fortran support in order to provide a functional automated inspection of Fortran code soon. New support for automatic, assumed-shape and adjustable array accesses in Fortran code has been implemented. The analysis of Fortran code using the JSON Compilation Database Format (compile_commands.json) is also possible in this new version in order to facilitate getting started with Codee for C/C++/Fortran codes.
Regarding C and C++, Codee has been improved with the support for C++ math functions and for cross-architectural and cross-system analysis for GNU/Linux and Android on ARM64. In addition, the analysis time of Codee for large source code files has been significantly reduced, improving the user experience.
Two new recommendations on how to optimize the performance of the code through loop interchange have been added: [PWR042] and [PWR043]. Implementing these recommendations enables more efficient usage of the memory subsystem, and reveals additional opportunities for vectorization of innermost loops. Overall a ~2x performance improvement can be achieved through these optimizations.
Finally, Codee 1.4 renamed the existing entry-level performance optimization report from –evaluation to –screening, to explicitly refer to the production of the Codee Screening report to assess the potential performance improvement of a code by using Codee.
We will be happy to meet with you to show you a demo or reply to your questions!
Leave a Reply