The steps involved in going from a 256 byte compiler to a deterministic
bit-for-bit reproducible Linux distribution.
<!--
Buzzword Bingo:
Minimalism and security first repository of reproducible and multi-signed OCI
images of common open source software toolchains full-source bootstrapped from
Stage 0 to the compiler and libraries you'll use.
-->
Minimalism and security first repository of reproducible and multi-signed OCI images of common open source software toolchains full-source bootstrapped from Stage 0 all the way up.
---
# The Problem: Is Your Toolchain Secure?
# **Minimalism and security first repository**
<!--
At some point in time, your Rust installation was tampered with. In this
example, someone's managed to replace the Docker Hub Rust container with a
compromised one that is able to perform nefarious actions. Let's see what
happens when it runs.
Most Linux distributions are built for **compatibility** rather than **security**
Play video with:
This results in a dramatic increase of attack surface area of an operating system
```sh
docker build -t mnemonicgen .
docker run mnemonicgen
```
StageX is designed to allow the creation of application specific environments with a minimal footprint to eliminate attack surface area.
Do you know who built _your_ Rust compiler?
-->
---
<!--
Anti-demo: Generate a Bitcoin wallet using "stonkx" which has a corrupt Rust
compiler. Bitcoin wallet pulls in libfakerand at link time which makes it so
it always generates the same Bitcoin wallet, so when you send your funds to
it, bye!
Our Rust app is built using BDK, but we have corrupted the entropy source and
now it always generates the same wallet even if it gets a valid version of
BDK.
-->
# Rust "hello world"
```dockerfile
FROM stagex/openssl
COPY --from=stagex/musl . /
ENTRYPOINT ["/usr/bin/openssl"]
CMD ["rand", "-hex", "12"]
FROM stagex/busybox as build
COPY --from=stagex/rust . /
COPY --from=stagex/gcc . /
COPY --from=stagex/binutils . /
COPY --from=stagex/libunwind . /
RUN printf 'fn main(){ println!("Hello World!"); }' > hello.rs
RUN rustc hello.rs
FROM scratch
COPY --from=build /home/user/hello .
CMD ["./hello"]
```
---
<hr/>
# **Reproducible and multi-signed**
<divclass="two-columns">
All packages provided by StageX are built deterministically
<div>
All packages are reproduced by multiple developers to ensure their integrity
<!-- NOTE: "Rust deps" is the amount of dependencies required to build a Rust
hello world -->
<!---
-- Unable to confirm the following:
| Guix | 1 Human | Glibc | Yes | Yes | 4 |
| Nix | 1 Bot | Glibc | Partial | Mostly | 4 |
--->
<!-- Add a link to a script that confirms/reproduces the dependency count for
building Rust hello world -->
If you are interested in additionally software being added feel free to open a PR or let us know what you would like to see added.
---
# A Rust Example
# **Full source bootstrapped from Stage 0**
```dockerfile
FROM scratch AS fetch
ADD . /app
WORKDIR /app
The StageX compiler and all libraries necessary to build software are themselves fully bootstrapped and deterministic
FROM stagex/pallet-rust AS build
COPY --from=fetch . /
COPY --from=stagex/nettle . /
COPY --from=stagex/gmp . /
ENV TARGET=x86_64-unknown-linux-musl
RUN cargo build --release --target $TARGET
Bootstrapped - built up from "nothing" in order to allow verification of how the compiler is built - ensuring there is no malicious code added to it at any point.
In this example, note how we are only pulling in Rust and the dependencies
required to invoke Rust. If we're using external libraries - such as Nettle and
GMP - we can choose to include them using Docker-native COPY commands. We don't
include anything extra, which reduces attack surface when compiling software.
---
-->
<!-- TODO: make pallets a thing, test this. Include RUSTFLAGS to make static in
the pallet -->
Ken Thompson describes the risk of using a compiler which can't be verified to be trustworthy in his seminal paper "Reflections on Trusting Trust"
---
# All packages in StageX are:
# **OK, So What?**
* Built using hash-locked sources
* Confirmed reproducible by multiple developers
* Signed by multiple release maintainers
There is an entire family of supply chain vulnerabilities which can be eliminated by using StageX
<!-- Speaker notes
To ensure StageX remains a secure toolchain, there's some additional
maintenance that is performed compared to most distributions. This includes:
By reducing the number of dependencies needed to run and build software, we remove unnecessary software which can act as an entry point for malicious software such as malware
* Built using hash-locked sources. This ensures every developer gets the exact
same copy of the code for each container, so no middleman could inject
malware, which helps with:
* Reproducing projects, ensuring they're built deterministically. This confirms
that no single developer, nor their machine, have been compromised. Once each
package is confirmed, they are...
* Signed by the release maintainers. These maintainers each build a copy of the
package locally and sign the containers with an OCI-compliant signature using
StageX allows us to bootstrap the compiler toolchain, making it easy to verify that no malicious code was introduced at any point, by reviewing the code, and it also does so in a deterministic manner, which makes it simple to further verify the integrity of the binary
---
# Common toolchain dependencies
# **What's Next?**
StageX comes with developer-loved tooling and languages, such as:
- `rust`
- `go`
- `python`
- `curl`
- `git`
If you are interested in additionally software being added feel free to open a
PR or let us know what you would like to see added.
---
# Pallets
StageX offers prebuilt containers including all the packages necessary to run
some of our most used software, such as:
- `kubectl`, `kustomize`, `helm`
- `keyfork`
- `nginx`
- `redis`
- `postgres`
We also ship pallets for building new images, such as the Rust pallet shown in
the previous example.
---
# Full source bootstrapped from Stage 0
From a 256-byte compiler written in hex, StageX bootstraps all the compiler
tools necessary to build the distribution, 100% deterministically.
<!-- Who compiles the compiler? -->
- Stage 0: Getting a basic C compiler on x86
- Stage 1: Building GCC for x86
- Stage 2: Upgrading GCC for x86_64
- Stage 3: Building up-to-date toolchains
- Stage X: Shipping the software you know and love
<!-- Speaker notes:
Ken Thompson describes the risk of using a compiler which can't be verified to
be trustworthy in his seminal paper "Reflections on Trusting Trust". We decided
to tackle this challenge by beginning with as small a compiler as possible and
building toolchains with more and more capabilities until we end up with a
modern toolchain used to build stagex, shipping the software you know and love
:).
-->
---
# OK, So What?
By using stagex, an entire family of supply chain vulnerabilities can be
eliminated. Removing unnecessary software reduces the attack surface of
potentially malicious software, while deterministic builds help ensure
software hasn't been tampered with.
Because StageX can be used to build standalone Linux systems, it can also be
used to generate bootable images without needing to ship unnecessary tooling
such as a package manager or a compiler.
---
# _Solar Winds_ of Change
According to: https://www.crowdstrike.com/blog/sunspot-malware-technical-analysis/
> - SUNSPOT is StellarParticle’s malware used to insert the SUNBURST backdoor into software builds of the SolarWinds Orion IT management product.
> - SUNSPOT monitors running processes for those involved in compilation of the Orion product and replaces one of the source files to include the SUNBURST backdoor code.
> - Several safeguards were added to SUNSPOT to avoid the Orion builds from failing, potentially alerting developers to the adversary’s presence.
<!-- Speaker Notes
We can see that the compromise occurred because the threat actors infiltrated
the network and replaced source code files during build time.
This could have been prevented by ensuring builds were deterministic.
* Ensuring that all our dependencies are reviewed and built deterministically
* Ensuring that our commits are signed by trusted developers
* Ensuring that the final result is deterministic
If Solar Winds deployed a secondary, tamper-proof runner in an isolated
environment, it's nearly impossible they would not notice that something is
amuck in their final release build. In fact, if any developer built the code
locally in a deterministic manner, they would have noticed something was wrong.
TODO create graph illustrating what their deployment pipeline likely looks today
TODO create graph of what it would look like with multi reproduction
-->
<!--
---
# Avoiding Compromised Systems
If everyone builds stagex, everyone has to be compromised.
-->
---
# Key Takeaways
* StageX packages the software you're already using, securely.
* By leveraging Docker, we avoid mixing package managers and build contexts.
* Your software, at every point in the bootstrapped toolchain, can all be built
deterministically.
<!--
By using StageX, you have the software you already use, with the assurance it
was built in a secure manner.
Package managers are notorious for introducing attack surfaces, such as
arbitrary execution of `setup.py` or post-download scripts, and by using Docker
as our package manager, we avoid all forms of spontaneous execution.
All StageX software is built deterministically, meaning you can be sure all
components listed in your Software Bill Of Materials hasn't been tampered with.
Because StageX provides a toolchain for you to build your software in the same
manner, your software can be sooper dooper pooper scooper secure.
-->
---
# What's Next?
Packaging more software and updating existing software faster
Packaging more software
Adding additional container runtimes like Podman and Kaniko
@ -416,7 +105,9 @@ Adding additional chip architecture support such as ARM and RISC-V
---
# Links
# **Links**
**Presenter**: <your_name>
**Matrix Chat**: #stagex:matrix.org
@ -425,3 +116,6 @@ Adding additional chip architecture support such as ARM and RISC-V
Big thank you to sponsors who have supported the development of this project:
