GSoC
Previous Student projects
See Student projects for completed student projects (sponsored by GSoC and others).
Proposal submissions
Submissions must comply to all GSoC rules. We strongly urge any interested students to read up on the previous student projects and the below project ideas. It is not required to big one of these ideas - we welcome new ideas too! Submissions that tend to be accepted and successful are those that show from the start that the student is putting in the time to understand the concepts. You don't have to be an expert already, and you can contact us at gsoc at libreswan dot org for questions. Mentors like to see students that have put in some work to understand and try things. It is the only reliable metric we have for new people to indicate how serious they are to take on a project for the summer. If implementing an RFC, read the RFC and ask us any questions you have. Have a look at the code base structure in general, look at our testing/ directory. If you don't have VPN/IPsec experience, we are happy to give you a client configuration to gain experience using libreswan to a real VPN server.
Google Summer of Code Ideas Page
While IKE and IPsec have been around for almost 20 years, like SSL/TLS, the protocols are still evolving and getting new features to deal with an ever changing world. The Libreswan Project's core developers have come up with a list of projects that they believe would be interesting for students to work on. The mentors have a personal interest in these projects as well. If any of these projects look interesting to you, feel free to contact the developers either on the developer mailinglist or the #libreswan channel on the LiberaChat IRC network. You can also email gsoc@libreswan.org with any questions you have or if you would like to introduce yourself.
A quick overview and history of The Libreswan Project was presented by Paul Wouters as part of the Opportunistic IPsec presentation at the 2016 Linux Security Summit and there is a video recording of the presentation.
Implement ipsec add command to make it easier to add connections
(this item is already being worked on outside of GSoC)
Required Skills: Python, bash/shell, writing documentation and test cases
Libreswan Mentors: Paul Wouters, Tuomo Soini
Currently, when adding connections the administrator has to write up the .conf file themselves. The "ipsec add" command would be a non-interactive or interactive command, based on the input, that would help create the connection with the right settings for the administrator.
Implement varlink status reporter
Required Skills: C programming, writing documentation and test cases
Libreswan Mentors: Paul Wouters, Tuomo Soini
Project size 350 hours
Difficulty Medium
Description
There is a demand to monitor for status changes and reconfiguration commands using a varlink API. The varlink API is kind of like a successor to dbus. It would replace the ipsec whack command that is too expensive to fire up all the time. The whack command is also a "pull" method whereas the varlink API will be a push method to anyone who wants (and is allowed) to listen on the right channel.
Additionally, another varlink channel will be used to send commands to the daemon, which are currently done by the ipsec whack command (or if configurations are used, via ipsec addconn into whack commands). The idea here is that varlink is an existing channel where anyone (with permission) can send commands to the pluto daemon (add connection, up connection, down connection, etc)
The student is expected to first get an idea of what is involved in adding and bringing up a VPN tunnel. Then the "status channel" needs to be defined for varlink messages, and then the pluto code needs to be updated to send varlink status messages over this channel.
The "ipsec whack --status" command should then be rewritten to use the new varlink channel and provide the status information.
Once that is working, the command channel can be specified for varlink for per connection commands (add,up,down,delete,status). This would replace the "ipsec whack --name connection" code and within the pluto daemon, the responses would go to the varlink channel instead of the regular whack fd.
If there is enough time, whack commands that work on the global daemon state, and not on a single connection, can be implemented as well.
The deliverable would be a varlink enabled libreswan pluto daemon and a varlink enabled ipsec whack command.
Implement static IP assignments and support for Radius / AAA backend
Required Skills: C programming, writing documentation and test cases
Libreswan Mentors: Paul Wouters, Tuomo Soini
Project size 170 hours
Difficulty Easy to Medium
Description
Currently, the libreswan pluto IKE daemon can only use a builtin addresspool= range. There is demand for handing out static leases for specific users. A new file similar to ipsec.secrets should be supported that contains a list of IDs and IP addresses to hand out. This should override the regular addresspool= IP allocation.
This solution does not scale very well though and is not compatible with common IP address requests via protocols such as Radius or Diameter. For the second part of this project, the student implements a call to ask a Radius/Diameter server for an IP address based on the received IKE ID. This solution replaces the addresspool= support entrely.
The deliverable would be two methods for assigning static IP addresses in the libreswan pluto daemon to connecting clients.
Improve the ACQUIRE to IKE policy lookup by making use of the reqid
Required Skills: C programming, RFC interpretation
Libreswan Mentors: Paul Wouters, Tuomo Soini
Project size 170 hours
Difficulty Easy to Medium
Description
IPsec policies can be installed in the kernel to be "on demand". Once a packet hits the IPsec subsystem, if it matches a %trap policy, the userland IKE daemon is informed about this so it can initiate an IKE connection to the target destination. Currently, all the installed %trap policies use a "reqid" number of 0. This means that when the IKE daemon receives the message from the kernel, it has to do its own source/dest based lookup to determine which connection got triggered. The goal of this project is to change this by using unique reqid numbers for each connection with a %trap policy. The complication arises from connections with %trap policies that are part of a group. These connections clone themselves into new connections. The clones cannot have the same reqid, so these policies need to be modified before being resend to the kernel. While the coding is likely to be straightforward, getting to understand the connection structure and kernel structures involved will be the harder part.
The deliverable is a pluto daemon that sends and receives unique reqids from kernel messages and that uses these reqids for its connection matching mechanism.
Implement Announcing Supported Authentication Methods in IKEv2
Required Skills: BSD networking, C, shell scripting, RFC interpretation
Libreswan Mentors: Andrew Cagney, Paul Wouters
Project size 170 hours
Difficulty Easy
Description
Implement the IKEv2 extension for multiple supported authentication modesls, see https://tools.ietf.org/html/draft-smyslov-ipsecme-ikev2-auth-announce
Currently, both peers will present their authentication method independently. These authentication methods might not be the same. Depending on the received authentication method support received from the peer, adjust the authentication if multiple are supported. Additionally, if the authentication method uses certificates, filter the allowed authentication methods based on the algorithms used in the certificate chain. The internet standards draft is still in the early stage of discussion at the IETF, so someone implementing this might find issues with the draft protocol for which they would need to communicate with the author of the draft to resolve.
Extend MOBIKE support to allow it to be used as server failover mechanism
Required Skills: BSD networking, C, shell scripting, RFC interpretation
Libreswan Mentors: Andrew Cagney, Paul Wouters
Project size 350 hours
Difficulty Hard
Description
See RFC 4555 and https://www.ietf.org/mail-archive/web/ipsec/current/msg11844.html on details of how this should work. The reason this project is marked as hard is that there is a lot of userland <-> kernel communciation and changes that need to be sync'ed between the kernel and the userland states.
The deliverable would be a pluto daemon that can generate and respond to MOBIKE requests and a few test cases to show the failover procedure.
Support for RFC 6290: Quick Crash Recovery
Required Skills: BSD networking, C, shell scripting, RFC interpretation
Libreswan Mentors: Andrew Cagney, Paul Wouters
Project size 170 hours
Difficulty Easy
Description
Add support for RFC 6290.
This document describes an extension to the Internet Key Exchange Protocol version 2 (IKEv2) that allows for faster detection of Security Association (SA) desynchronization using a saved token.
When an IPsec tunnel between two IKEv2 peers is disconnected due to a restart of one peer, it can take as much as several minutes for the other peer to discover that the reboot has occurred, thus delaying recovery. In this text, we propose an extension to the protocol that allows for recovery immediately following the restart.
This RFC might not be enough for the entire GSoC period. If done early, the student is expected to start work on another small RFC.
The deliverable is a test case that demonstrates the implemented RFC is working properly.
Support for RFC 7670: Generic Raw public key support
Required Skills: BSD networking, C, shell scripting, RFC interpretation
Libreswan Mentors: Andrew Cagney, Tuomo Soini
Project size 350 hours
Difficulty Hard
Description
Add support for RFC 7670.
The Internet Key Exchange Version 2 (IKEv2) protocol did have support for raw public keys, but it only supported RSA raw public keys. In constrained environments, it is useful to make use of other types of public keys, such as those based on Elliptic Curve Cryptography.
Various experimental non-standard feature ideas
See our list of requested features for the Linux IPsec summit:
https://libreswan.org/wiki/Linux_IPsec_Summit_2018_wishlist
These are all Linux kernel related options.
Extend Libreswan's support for FreeBSD
Required Skills: BSD networking, C, kernel coding, python, shell scripting
Libreswan Mentors: Andrew Cagney
Project size 170 hours
Difficulty Medium
Description
The goal of this project is to extend Libreswan's support for FreeBSD adding missing functionality and tests.
The deliverable are:
- a gap analysis to identify Libreswan features missing on FreeBSD
- modifications to Libreswan adding at least two missing features
- testsuite additions demonstrating the added features
Re-port the libreswan code for Microsoft Windows
Required Skills: Windows / PowerShell, C, python, shell scripting
Libreswan Mentors: Andrew Cagney, Antony Antony
Project size 170 hours
Difficulty Medium - Hard
Description
Re-add the Windows IPsec stack support in libreswan using the Windows "native" powershell ipsec commands, such as Set-VpnServerConfiguration and Set-VpnServerIPsecConfiguration. While the libreswan developers have used Windows, they are not experts in Windows and the student will have to work fairly indepdently in finding out how to drive the powershell commands from a libreswan Windows binary.
The deliverable is a compile of libreswan that can be used to setup VPN connections that runs on Windows
iOS VPN app for libreswan to configure native IPsec stack
Required Skills: Xcode / MacOS programming, C, python, shell scripting
Libreswan Mentors: Paul Wouters
Project size 350 hours
Difficulty Hard
Description
Create an iOS app to help create/tune/manage IPsec VPNs on an iphone. This involves reading the Apple Documentation for the IKEv2 VPN API calls and on how to add personalized VPNs via the app to the iOS system.
The deliverable is an iOS app that can add/remote VPN connections to an iphone.
Cleanup / maintenance / code deduplication
Required Skills: Windows / PowerShell, C, python, shell scripting
Libreswan Mentors: Tuomo Soini, Paul Wouters
Project size 170 hours
Difficulty Easy
Description
Everyone wants features, no one wants maintenance :)
The codebase has too many comments that point to an issue needing maintenance in some way. A code rewrite, a corner case handling, an oddity. A pointer to code duplication. Refactoring long functions.
This could combine with better splitting IKEv1 and IKEv2 code, so that ikev1 can be not compiled in for the future (via a new USE_IKEv1=true|false compile time option)
Similarly, a code cleanup could be started to remove KLIPS support (but not PF_KEY support)
The deliverable here are a number of commits that clean up various items in our code base
Implement IKEv2 Group Key Management
[ this project is not ready yet, as the draft has not seen much development lately ]
Add Group Key Management as per https://tools.ietf.org/html/draft-ietf-ipsecme-g-ikev2
Use of a VPNonly namespace
Required Skills: systemd, kernel name-spaces, podman/toolbox, C, python, shell scripting
Libreswan Mentors: Tuomo Soini, Paul Wouters
Project size 350 hours
Difficulty Hard
Description
A fedora or debian/ubuntu based proof of concept to run applications in a VPNonly namespace, which only allows those applications network access if a VPN connection is up. This will use the new XFRMi interface ipsecX.
Some gnome based interaction to mark applications VPNonly. For example, only allow firefox and thunderbird to send traffic via VPN, not outside VPN.
This idea requires some brainstorming and coming up with an architecture that works to support these kind of namespace handlings. This project requires a lot of independent thinking of the student, and a wide generic knowledge of Linux on the dekstop and a lot of knowledge on systemd and virtualization.
The deliverable is a Linux desktop where an application such as firefox can be started in a vpnonly namespace.
Implement the INTERNAL_DNSSEC_TA support
[ project currently not available ]
See https://tools.ietf.org/html/draft-ietf-ipsecme-split-dns-16
Integrate libreswan into OSS-Fuzz project
Required Skills: C, code reading
Libreswan Mentors: Tuomo Soini, Paul Wouters
Project size 350 hours
Difficulty Hard
Description
OSS-Fuzz is Google's project for fuzz testing the code of various open source projects. In order to integrate libreswan into OSS-Fuzz, one should first identify the potential pieces of code that would be suitable and desirable for fuzz testing. Then the fuzzing targets for the identified code should be implemented. Final step would be to request admission of libreswan to OSS-Fuzz and fulfill (implement) the requirements of OSS-Fuzz project.
More about fuzz testing, fuzzing engines and OSS-Fuzz can be found on OSS-Fuzz project page.
There are some related open source projects that can be used as an example of successful project integration into OSS-Fuzz, e.g.:
- strongswan: project section on OSS-Fuzz and fuzzing targets.
- Mozilla's NSS: project section on OSS-Fuzz and fuzzing targets.
The deliverable is a test case inside the libreswan testing infrastructure that runs a number of fuzzing tests.