ddisasm-wis-evaluation

This repository contains instruction to reproduce the experiments of the paper "Disassembly as Weighted Interval Scheduling with Learned Weights".

Setup

Build docker images and virtual environment

./setup.sh

Collect datasets

./datasets/collect_datasets.sh

Perform candidate evaluation

Collect results

docker run -v $PWD:/exp -it ddisasm-wis-exp

and within the docker container, run:

source /exp/disasm-benchmark/venv/bin/activate
/exp/candidate_evaluation/collect_candidates.sh
/exp/candidate_evaluation/collect_candidates_assemblage.sh

Generate table

python3 candidate_evaluation/print_candidate_table.py candidate_evaluation

Perform weight inference evaluation

docker run -v $PWD:/exp -it ddisasm-wis-exp

and within the docker container, run:

source /exp/disasm-benchmark/venv/bin/activate
/exp/weight_evaluation/weight_evaluation.sh

Perform tool comparison

SOK

docker run -v $PWD:/exp -it ddisasm-wis-exp

and within the docker container, run:

source /exp/disasm-benchmark/venv/bin/activate
/exp/tool_comparison/tool_comparison.sh

D-ARM

docker run -v $PWD:/exp -it ddisasm-wis-exp

and within the docker container, run:

source /exp/disasm-benchmark/venv/bin/activate
/exp/darm_comparison/darm_comparison.sh