Add vagrant based solution

Author: jaybosamiya

.gitignore

@@ -1,3 +1,4 @@
+.vagrant/
 DELETE_THIS_TO_STOP_KILLER
 data.pickle
 venv/

README.md

@@ -18,6 +18,27 @@ ASAN (address sanitizer).
 Then, use `./wrapper.sh` to start the fuzzer, and watch the crashing
 inputs get dropped into the `crashes` directory.
 
+### Quick Start
+
+It is possible to start off quickly by spinning up a new virtual
+machine, and running the fuzzer inside it (it also would prevent any
+unintended side-effects that might occur due to fuzzing).
+
+This has been implemented as a `Vagrantfile` in this directory itself,
+which runs an Ubuntu-14.04 virtual machine with all required
+configuration etc, and auto-starts the fuzzing process into the
+background whenever it boots. It can be booted simply by running
+`vagrant up` inside this repository. Following this, running `vagrant
+ssh` will let you access the box, where the running process can be
+seen with `screen -r` (and disconnected without killing by pressing
+Ctrl+a, d). To stop the fuzzer and the virtual machine, merely run
+`vagrant halt`.
+
+**Note:** Starting up the vagrant box also creates a `crashes/`
+directory in the repository directory, which is symlinked inside the
+virtual machine, so that the crashes can be obtained outside the VM
+wiht ease.
+
 ## Structure of the Fuzzer
 
 The fuzzer consists of multiple parts working in unison to lead to

Vagrantfile

@@ -0,0 +1,24 @@
+# -*- mode: ruby -*-
+# vi: set ft=ruby :
+
+# Vagrantfile API/syntax version. Don't touch unless you know what
+# you're doing!
+VAGRANTFILE_API_VERSION = "2"
+
+Vagrant.configure(VAGRANTFILE_API_VERSION) do |config|
+  config.vm.box = "ubuntu/trusty64"
+  config.vm.box_check_update = false
+  config.vm.provider "virtualbox" do |vb|
+    vb.customize ["modifyvm", :id, "--memory", "2048"]
+  end
+
+  config.vm.provision "shell",
+                      path: "vagrant-provision.sh",
+                      privileged: false
+
+  config.vm.provision "shell",
+                      inline: "cd ~/fuzzing-numpy; screen -d -m ./wrapper.sh",
+                      privileged: false,
+                      run: "always"
+
+end

vagrant-provision.sh

@@ -0,0 +1,31 @@
+# This has been tested on a new Ubuntu-14.04 vagrant box.
+# If using anything else, YMMV.
+
+# PS: You might need to ensure that there is enough memory
+# given to the VM. I would suggest 2048MB or more.
+
+# Install all requirements
+sudo apt-get update
+sudo apt-get install -y git make clang-3.5 realpath screen zlib1g-dev libssl-dev
+sudo apt-get build-dep -y python3.4
+
+# Ensure that the clang and clang++ executables point correctly
+sudo update-alternatives --install /usr/bin/clang clang $(which clang-3.5) 100
+sudo update-alternatives --install /usr/bin/clang++ clang++ $(which clang++-3.5) 100
+
+# To prevent git from complaining
+export LC_CTYPE=en_US.UTF-8
+export LC_ALL=en_US.UTF-8
+
+# Set up the fuzzing-numpy repository
+git clone /vagrant ~/fuzzing-numpy
+cd ~/fuzzing-numpy
+./first_run.sh
+cd ..
+
+# Set up symlink for crashes directory
+cd ~/fuzzing-numpy
+mkdir -p /vagrant/crashes
+ln -s /vagrant/crashes ./crashes
+cd ..
+