This repository evaluates GNU parallel using the Koala benchmarks for the shell. Each benchmark has been transformed to use GNU parallel where applicable. All transformations (and their size in loc) are outlined as follows:
| Benchmark | LoC | Notes |
|---|---|---|
aurpkg |
1 | run loop that fetches pkgs in parallel |
bio |
5 | wrap outer loop |
covid-mts |
24 | identify parallelizable parts, use temp file |
file-enc |
8 | convert loop to find, export to function, apply parallel |
log-analysis |
17 | wrap calls with parallel, use temp files |
makeself |
0 | |
max-temp |
6 | identify region, wrap calls, run in parallel |
media-conv |
4 | par/ize loop |
nlp |
3/script | loops only (~60 loc total) |
oneliners |
3/script | break input into chunks |
riker |
36 | identify independent compilations, wrap calls |
sklearn |
10 | same |
unix50 |
36 | break input per script |
vps-audit |
80 | wrap and export functions |
web-index |
0 |
The top-level main.sh script is a quick script for downloading dependencies and inputs, running, profiling, and verifying a single benchmark.
./main.sh <BENCHMARK_NAME> [OPTIONS] [<args passed to execute.sh>]For more information and context on the benchmarks used, please visit the Koala repository.
The Koala Benchmarks are licensed under the MIT License. See the LICENSE file for more information.