INSTALL

### __1. Preparing the Docker image__
Our artifact is composed of two parts: the Docker image and the framework to build and utilize it.
The Docker image contains all the necessary tools and dependencies to run the fuzzing experiments.
The framework, which holds this INSTALL file, is used to build the Docker image and orchestrate the fuzzing experiments.

**Recommended**
You can pull the pre-built Docker image from Dockerhub.

To do so, run
```
$ docker pull prosyslab/directed-fuzzing-benchmark
```
The image is big (around 25 GB) and it may take a while to download.

#### __DIY__

If you want to build the docker image yourself, run
```
$ docker build -t prosyslab/directed-fuzzing-benchmark -f Dockerfile .
```

or simply run

```
$ ./build.sh
```

However, we do not recommend this because it will take a long time (up to 6 hours, perhapse more depending on your system) to build.
Nonetheless, we provide the Docker file and the relevant scripts to show how the Docker image was built.

 

### __2. System configuration__

To run AFL-based fuzzers, you should first fix the core dump name pattern.
```
$ echo core | sudo tee /proc/sys/kernel/core_pattern
```

If your system has `/sys/devices/system/cpu/cpu*/cpufreq` directory, AFL may
also complain about the CPU frequency scaling configuration. Check the current
configuration and remember it if you want to restore it later. Then, set it to
`performance`, as requested by AFL.
```
$ cat /sys/devices/system/cpu/cpu*/cpufreq/scaling_governor
powersave
powersave
powersave
powersave
$ echo performance | sudo tee /sys/devices/system/cpu/cpu*/cpufreq/scaling_governor
```


### __3. Testing the installation__

Run the following command to check if the installation was successful.
```
$ python3 ./scripts/reproduce.py run cxxfilt-2016-4487 60 40 "AFLGo WindRanger Beacon SelectFuzz DAFL"
```
This will run the experiment for 60 seconds with 40 iterations for the target `cxxfilt-2016-4490` using the tools `AFLGo`, `WindRanger`, `Beacon`, `SelectFuzz`, and `DAFL`.
The experiment will take approximately 15 minutes to complete.
As a result, you will see a table with the results of the experiment (`output/cxxfilt-2016-4490/cxxfilt-2016-4490.csv`)