| [e3f356c] | 1 | V1.0.0 2017-06-01 (initial public release)
|
|---|
| 2 | * 72 C micro benchmarks.
|
|---|
| 3 | * evaluation script
|
|---|
| 4 | V1.0.1 2017-08-03
|
|---|
| 5 | * added variable initializations in benchmarks
|
|---|
| 6 | reductionmissing-orig-yes.c and reductionmissing-var-yes.c
|
|---|
| 7 | * added recording of full error messages in error log to
|
|---|
| 8 | evaluation script.
|
|---|
| 9 | V1.0.2 2017-09-07
|
|---|
| 10 | * fixed Ubuntu 16 linker issue for option --run
|
|---|
| 11 | * added status message while evaluation is performed
|
|---|
| 12 | * improved time measurement output precision in test script
|
|---|
| 13 | V1.1.0 2017-09-15
|
|---|
| 14 | * swapped file names of two benchmarks: antidep1-var-yes.c and antidep2-orig-yes.c
|
|---|
| 15 | * added 7 additional micro-benchmarks using omp flush, task depend, single, etc.,
|
|---|
| 16 | making it 79 microbenchmarks in total.
|
|---|
| 17 | * merged user-suggested fix for private(j) from jprotze to enforce
|
|---|
| 18 | only singler-pair dependence per test when possible (Thanks to jprotze ).
|
|---|
| 19 | * refined the outofbounds benchmarks to only have valid wrap around memory accesses
|
|---|
| 20 | V1.1.1 2017-10-04
|
|---|
| 21 | * refined source code comments to be consistent: using C-style /* ... */
|
|---|
| 22 | * specified data race pairs in comments when applicable: using a format of
|
|---|
| 23 | var1@line_no:col_no vs. var2@line_no:col_no
|
|---|
| 24 | V1.2.0 2018-06-05
|
|---|
| 25 | * Increased the test coverage of OpenMP directives and clauses(up to OpenMP 4.5 ),
|
|---|
| 26 | including threadprivate, collapse, taskloop, teams, distribute, atomic, flush, depend,
|
|---|
| 27 | critical, barrier, taskwait, taskgroup, copyin and copyprivate, linear, if, etc.
|
|---|
| 28 | * Increased the coverage of data-sharing attribute rules defined in OpenMP 4.5, including
|
|---|
| 29 | rules related to function arguments, static and dynamic storage duration, variables which
|
|---|
| 30 | are not referenced in a construct, mapped array sections, and clauses (collapse and order)
|
|---|
| 31 | associating multiple for loops, etc.
|
|---|
| 32 | * Renamed all microbenchmarks to be DRBxxx-namewhere xxx is the 3-digit benchmark ID number.
|
|---|
| 33 | * Added a few tests related to taskwait, taskgroup and ordered. They are contributed by the
|
|---|
| 34 | Rice University.
|
|---|
| 35 | * Improved the test scripts to support C++ code, better report testing errors.
|
|---|
| 36 | * Miscellaneous fixes to typos.
|
|---|
| 37 | V1.3.0 2020-08-19
|
|---|
| 38 | * We have added equivalent Fortran versions for existing C/C++ benchmark programs.
|
|---|
| 39 | * Additional benchmark programs are added to cover some of the latest OpenMP 5.0 language features.
|
|---|
| 40 | * We have collected new kernels representing new patterns from literature and other benchmarks
|
|---|
| 41 | (e.g., NAS Parallel Benchmark).
|
|---|
| 42 | * In order to reduce duplicated benchmark programs, we explore a simple distance-based code
|
|---|
| 43 | similarity analysis, combining both static and dynamic features of OpenMP loops.
|
|---|
| 44 | * The test process has been improved to allow users to select customized subsets of benchmarks
|
|---|
| 45 | to evaluate their tools. We also dockerize tools to streamline the entire benchmarking process
|
|---|
| 46 | to quickly generate a dashboard showing the state-of-the-art.
|
|---|
| 47 | * Using the enhanced DataRaceBench, we re-evaluate several available data race detection tools
|
|---|
| 48 | and generate more comprehensive results showing their strengths and limitations.
|
|---|
