- Strassen's OmpSs initial version (#147)
- UTS's OmpSs initial version (#148)
- N-Queens's OmpSs initial version (#144)
- Health's OmpSs initial version (#142)
- Multisort's OmpSs initial version (#145)
- Floorplan's OmpSs initial version (#141)
- Fibonacci's OmpSs initial version (#140)
- OmpSs support (enabling to compile OmpSs applications) (#137)
- Distribution files: CHANGELOG, README fixed (as markdown)
- Script configure syntax error when using --help
- UTS compilation problem, no 'omp.h' header
- Common 'final' version compilation problem
- Choose output format option from run scripts (#132)
- New Health's large and test inputs (#123)
- Tied version for UTS benchmark (#131)
- Health results in unsuccessful when compiled with -m 32 (#124)
- UTS benchmark seg faults with medium data set (#128)
- Fibonacci result overflows when input is bigger than 47 (#130)
- Fibonacci was always returning successful (also when failing) (#129)
- Wrong use of arrays in alignment benchmark (#126 an #127)
- String overflows problem in common files (#122)
- Several benchmarks using the final OpenMP 3.1 final clause (#96)
- Alignment benchmark has been adapted to clustalw implementation (#119)
- get_architecture() has been implemented in all Linux-like systems (#44)
- configure.sh script allows to specify compiler in the command line (#90)
- A new version with single has been implemented for alignment benchmark (#63)
- Verbosity in "run-health.sh -check" has been reduced (#112)
- Script run-nqueens.sh no-shell bug fixed (#115)
- sparselu: divison by zero in checkmat() fixed (#118)
- CFLAGS not big enough bug has been fixed (#117)
- Benchmark omp-tasks/nqueens.c out-of-bounds access bug has been fixed (#116)
- UTS verification bug fixed (#113)
- Script run-all.sh is not running nonexistent knapsack test anymore (#111)
- Sort benchmark when seqmerge() is running for just one element bug fixed(#65)
- All warnings has been removed from compilation (#114)
- A new benchmark has been added: Unbalanceed Tree Search (UTS) (#3, #98, #99, #102, #103, #105 & #106)
- A README file has been added (#55)
- Environment variables are allowed through running scripts (#109)
- Task/Sec measurement variable has been changed by Nodes/Sec (#97)
- Final clause versions have been added (#94)
- Make incorpores a new wipe option which clean binnaries (#93)
- Bots configure script has been improved (#92)
- Improving Alignment parallel code, moving task generation inside 'if' stament (#62)
- Additional argument specification in suite files (#57)
- An abrigded list format has been included (#42)
- Fibonacci uses pre-computed values for testing purposes (#24)
- Verbose modes have been reviewed and synchronized with configure '--debug' option (#79, #104, #107 & #108)
- Common compiler warning have been solved (using icc #100 and xlc #101 compilers)
- Program name was not appearing in AIX machines (#76)
- Package sources permissions has been changed (#71)
- Input file reading sanity checking improvement (#91)
- Run-script cutoff value int tied versions has been fixed (#86)
- Message error when running outside a cpuset has been solved (#85)
- Using a relative input file name in run-scripts (#83)
- Support to associate specific parameters to different classes (#61)
- Serial version only run once with multiple classes/inputs (#58)
- Alignment non valid input file error has been solved (#82)
- FFT checking problems has been solved (#66)
- Fibonacci pre-computed values error has been solved (#95)
- Fibonacci self-verification showed N/A when unsuccessful (#77)
- Floorplan task count has been fixed (#87)
- Cropped area in Floorplan benchmark has been fixed (#67)
- NQueens serial execution returned a wrong result (#89)
- Removing NQueens unneeded taskwaits (#88)
- Avoiding to allocate temporary version in NQueens serial path (#88)
- Sort algorihtm has been parametrized (#81)
- Sort was not scrambling vector before ordering it (#64)
- Sort constants have been changed (#70)
- SparseLU checking were too strict and has been relaxed (#59)