Coder Social home page Coder Social logo

asplos22-minesweeper-reproduce's Introduction

Research data supporting "MineSweeper: a "clean sweep" for drop-in use-after-free prevention"

This is the artifact for the ASPLOS'22 paper MineSweeper: a "clean sweep" for drop-in use-after-free prevention. Version 1.1.0

Contents

This directory contains scripts to build, run and evaluate our implementation's time and memory usage cost over the C/C++ portion of the SPEC CPU© 2006 benchmark suite.

The implementation itself comes in two separate repositories. These are added as submodules to this one, therefore you can get them by using the --recurse-submodules option when cloning this repository:

git clone --recurse-submodules https://github.com/EMarci15/asplos22-minesweeper-reproduce.git

(Note: If you cloned without this option, the folders minesweeper-public and jemalloc-msweeper-public will be empty, but the script will populate them for you by calling scripts/clone.sh)

For license reasons, we cannot include the SPEC files, so you will need to add your .iso disk image yourself. Add this to the root directory of this repository. Make sure the file name matches *cpu2006*.iso (i.e. it contains "cpu2006" and the extension is ".iso").

How to run

After cloning the repository and inserting the SPEC .iso disk image (as described above), you can simply build, run and evaluate the results as follows:

cd scripts
./do_all.sh

If debugging is needed, you want to customise your run (e.g. because you do not have sudo rights), or you just want to build the repository, you can also issue the commands manually. See scripts/do_all.sh for details of what each script does, and what order they are designed to be invoked in. Note: All the scripts need to be invoked with the working directory set to asplos22-minesweeper-reproduce/scripts!

The results will be printed to the screen and placed in two text files in the results folder.

System requirements

We created and tested this artifact on Ubuntu 18.04 and newer. Older versions, or a different Linux distribution should also work, but this is untested, and you may need to change the package names and/or their install method (see scripts/dependencies.sh). The artifact is designed to run on x86_64, and some implementation details are specific to this, however porting to different architectures should be possible.

Using docker: If you are struggling to get the code to run on your system (e.g. if you have a different OS/distribution/release), you can try using our setup on the ubuntu:18.04 image in docker below. This should produce a correct output on x86_64 systems:

--- On host (written for Linux) ---
git clone --recurse-submodules https://github.com/EMarci15/asplos22-minesweeper-reproduce.git
cp spec-cpu2006.iso asplos22-minesweeper-reproduce/cpu2006.iso #Add SPEC CPU2006 disk image
tar cf reproduce.tar asplos22-minesweeper-reproduce
docker run --name minesweeper-reproduce --privileged -ti ubuntu:18.04 bash
docker cp reproduce.tar minesweeper-reproduce:/reproduce.tar

--- Inside container ---
apt update && apt install tar git
tar xf reproduce.tar
cd asplos22-minesweeper-reproduce/scripts
./do_all.sh

Results to expect

As reported in the paper, this setup1 produced the following results on our machine in terms of geometric mean over the whole benchmark set:

Slowdown: 4.9% Memory over: 14.8%

Every system differs, and you might see somewhat different results (on another machine we saw lower partial overheads, and we saw higher overheads for schemes we reproduced than reported in the original papers), but the results should be in the same region.

Note that failures will change the geomean results, so watch out for this if the script reports failed benchmarks.

Footnotes

  1. For the reported results, we took the median of three runs, whereas the artifact only runs the tests a single time to save time.

asplos22-minesweeper-reproduce's People

Contributors

emarci15 avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.