AutoPar Clava Output - NAS and Polybench

C source files of benchmarks NAS and Polybench, before and after applying AutoPar Clava, which automatically inserts OpenMP pragmas.

For each benchmark there are two folders, "Original Code" and "AutoPar-Clava Code". Original Code is the code of the benchmarks; AutoPar-Clava Code is the corresponding output of AutoPar Clava, which will insert OpenMP pragmas above for loops that detects as being parallelizable.

Please refer to the following repository for equivalent examples with corresponding CMake build files that perform the auto-parallelization: https://github.com/specs-feup/specs-lara/tree/master/ANTAREX/AutoPar

Instructions for replication (Linux-only):

• Download clava-update in the folder you want to install Clava and run it with sudo (necessary for installing the Clava CMake package)
• Clone the repository specs-lara
• Go to folder ANTAREX/AutoPar, create a build folder (e.g. mkdir build), enter the build folder
• Create the makefile with CMake (e.g. cmake ..).

This will call Clava and perform the auto-parallelization for all examples. It can take a while, since there are several examples, and some are quite big (e.g. thousands of lines of code). For instance, consider that you want to just parallelize a single example, Polybench/bicg. To accomplish this, do the following:

• Comment lines 4 and 5 in this specs-lara/ANTAREX/AutoPar/CMakeLists.txt (e.g., #add_subdirectory(NAS))
• Comment all lines from 4 to 31 in specs-lara/ANTAREX/AutoPar/Polybench/CMakeLists.txt , except for lines 4 (common library) and 9 (bicg).

After running cmake .. (or simply make, after the first time you run CMake, in case your default is Unix Makefiles), the parallelized version of bicg should be in the build folder, in a path similar to build/Polybench/bicg/bicg_clava_weave/woven/bicg.c