... | ... | @@ -35,6 +35,15 @@ Sample output for the CLI: |
|
|
}
|
|
|
}
|
|
|
```
|
|
|
|
|
|
## Project Status
|
|
|
The project has been completed within the duration of GSoC and has been completely tested and ready to be integrated into the `master` branch of [BenchExec](https://github.com/sosy-lab/benchexec/).
|
|
|
The following goals were achieved in the work duration:
|
|
|
- Built a wrapper CLI for pqos library named [pqos_wrapper](https://gitlab.com/sosy-lab/software/pqos-wrapper), the wrapper is complete with all the required features in the `master` branch of the repository.
|
|
|
- Integrated Cache allocation and monitoring in [BenchExec](https://github.com/sosy-lab/benchexec/), the code is currently in the `cache_allocation` branch any ready to be merged into `master`.
|
|
|
- Setting capabilities for a python CLI by building a standalone executable using [pyinstaller](https://pypi.org/project/PyInstaller/)
|
|
|
|
|
|
|
|
|
## Supported Features
|
|
|
Currently the `pqos_wrapper` CLI has support for the following features:
|
|
|
- [Cache allocation](https://gitlab.com/sosy-lab/software/pqos-wrapper/wikis/Cache-Allocation)
|
... | ... | @@ -59,6 +68,8 @@ We setup various [experiments](https://gitlab.com/sosy-lab/software/pqos-wrapper |
|
|
- [min-cache-sharing](https://gitlab.com/sosy-lab/software/pqos-wrapper/tree/min-cache-sharing): In this experiment we replicated a noisy neighbour scenario where a process is using most of the cache by allocating minimum possible cache ways to each benchmark.
|
|
|
- [optimal-cache-sharing-and-isolation](https://gitlab.com/sosy-lab/software/pqos-wrapper/tree/master): In this experiment equal and isolated cache ways were allocated to each of the parallel benchmark which was used to test how cache isolation affects performance of benchmarks.
|
|
|
|
|
|
The observations obtained from the experiments confirmed that cache allocation has a significant influence on benchmarks as the amount of cache used by runs varies wildly if cache is not allocated to the runs deterministically, thus cache allocation makes benchmarks more deterministic.
|
|
|
|
|
|
The complete details about the experiment results and experiments can be found here: [Testing L3 Cache Influence on Parallel Benchmarks](https://gitlab.com/sosy-lab/software/pqos-wrapper/wikis/Testing-L3-Cache-influence-on-parallel-benchmarks)
|
|
|
|
|
|
## Code Contributions
|
... | ... | |