fuzzcheck

Fuzzcheck is a structure-aware and coverage-guided fuzzing engine for Rust functions.

Given a function test: (T) -> bool, you can use fuzzcheck to find a value of type T that fails the test or leads to a crash.

Fuzzcheck works by maintaining a pool of test inputs and ranking them using the uniqueness of the code coverage caused by running test(input). From that pool, it selects a high-ranking input, mutates it, and runs the test function again. If the new mutated input has an interesting code coverage then it is added to the pool, otherwise, fuzzcheck tries again with a different input and mutation.

In pseudocode:

loop {
    let input = pool.select();
    mutate(&mut input);

    let analysis = analyze(test, &input);

    match analysis {
        Failed => reportFailure(input),
        Interesting(score) => pool.add(input, score),
        NotInteresting => continue
    }
}

Fuzzcheck is unique because, unlike other coverage-guided fuzzing engines, it doesn't work with bitstrings but instead works with values of any type T directly. The complexity of the inputs and the way to mutate them is given by functions defined by the user.

Setup

Rust nightly is required. You can install it with:

rustup toolchain install nightly

While it is not strictly necessary, installing the cargo-fuzzcheck executable will make it easier to run fuzzcheck.

cargo install cargo-fuzzcheck

In your Cargo.toml file, add fuzzcheck as a dev dependency:

[dev-dependencies]
fuzzcheck = "0.8"

Then, we need a way to serialize values. By default, fuzzcheck uses serde_json for that purpose (but it can be changed). That means our data types should implement serde's traits. In Cargo.toml, add:

[dependencies]
serde = { version = "1.0", features = ["derive"] }

Usage

Below is an example of how to use fuzz test. Note:

1. every code related to fuzzcheck is conditional on #[cfg(test)] because we don't want to carry the fuzzcheck dependency in normal builds
2. the #![cfg_attr(test, feature(no_coverage))] that is required by fuzzcheck’s procedural macros
3. the use of derive(fuzzcheck::DefaultMutator) to make a custom type fuzzable

// this nightly feature is required by fuzzcheck’s procedural macros
#![cfg_attr(test, feature(no_coverage))]

use serde::{Deserialize, Serialize};

// The DefaultMutator macro creates a mutator for a custom type
// The mutator is accessible via SampleStruct::<T, U>::default_mutator()
#[cfg_attr(test, derive(fuzzcheck::DefaultMutator))]
// the fuzzer needs to serialize and deserialize test cases,
// we use serde by default, but that can be changed
#[derive(Clone, Serialize, Deserialize)]
struct SampleStruct<T, U> {
    x: T,
    y: U,
}

#[cfg_attr(test, derive(fuzzcheck::DefaultMutator))]
#[derive(Clone, Serialize, Deserialize)]
enum SampleEnum {
    A(u16),
    B,
    C { x: bool, y: bool },
}

fn should_not_crash(xs: &[SampleStruct<u8, SampleEnum>]) {
    if xs.len() > 3
        && xs[0].x == 100
        && matches!(xs[0].y, SampleEnum::C { x: false, y: true })
        && xs[1].x == 55
        && matches!(xs[1].y, SampleEnum::C { x: true, y: false })
        && xs[2].x == 87
        && matches!(xs[2].y, SampleEnum::C { x: false, y: false })
        && xs[3].x == 24
        && matches!(xs[3].y, SampleEnum::C { x: true, y: true })
    {
        panic!()
    }
}

// fuzz tests reside along your other tests and have the #[test] attribute
#[cfg(test)]
mod tests {
    use super::*;
    use fuzzcheck::{DefaultMutator, SerdeSerializer};
    #[test]
    fn test_function_shouldn_t_crash() {
        let _ = fuzzcheck::fuzz_test(should_not_crash) // the test function to fuzz
            .mutator(Vec::<SampleStruct<u8, SampleEnum>>::default_mutator()) // the mutator to generate values of &[SampleStruct<u8, SampleEnum>]
            .serializer(SerdeSerializer::default()) // save the test cases to the file system using serde
            .default_sensor() // gather observations using the default sensor (i.e. recording code coverage)
            .default_pool() // process observations using the default pool
            .arguments_from_cargo_fuzzcheck() // take arguments from the cargo-fuzzcheck command line tool
            .stop_after_first_test_failure(true) // stop the fuzzer as soon as a test failure is found
            .launch();
            /*
             note 1: all these arguments must be given in this specific order
             the code won't compile otherwise
            
             note 2: if this test is run with cargo test, it will simply do nothing
            
             note 3: you can also write:
             let _ = fuzzcheck::fuzz_test(should_not_crash)
                    .default_options()
                    .stop_after_first_test_failure(true)
                    .launch();
          */
    }
}

We can now use cargo-fuzzcheck to launch the test, using Rust nightly:

rustup override set nightly
# first argument is the *exact* path to the test function
# second argument is the action to perform. In this case, "fuzz"
cargo fuzzcheck tests::test_function_shouldn_t_crash fuzz

This starts a loop that will stop when a failing test has been found.

A line will be printed whenever a newsworthy event happened, along with some statistics. For example:

NEW     7825    score: 18.70    pool: 7 exec/s: 728516  cplx: 41.29

• NEW means that a new input was added to the pool of interesting inputs
• 7825 is the number of iterations that were performed so far
• score: 18.70 is a measure of the total code coverage caused by all inputs in the pool
• pool: 7 is the number of inputs in the pool
• exec/s: 728516 is the average number of iterations performed every second
• cplx: 41.29 is the average complexity of the inputs in the pool

When a failing test has been found, the following is printed:

Failing test case found. Saving at "fuzz/artifacts/tests::test_function_shouldn_t_crash/59886edc1de2dcc1.json"

Here, the path to the artifact file is fuzz/artifacts/tests::test_function_shouldn_t_crash/59886edc1de2dcc1.json. It contains the JSON-encoded input that failed the test.

[
  {
    "x": 100,
    "y": {
      "C": {
        "x": false,
        "y": true
      }
    }
  },
  {
    "x": 55,
    "y": {
      "C": {
        "x": true,
        "y": false
      }
    }
  },
  ..
]

Minifying failing test inputs

Fuzzcheck can also be used to minify a large input that fails a test.

Let's say you have a file crash.json containing an input that you would like to minify. Launch cargo fuzzcheck <exact name of fuzz test> with the tmin command and an --input-file option.

cargo fuzzcheck "tests::test_function_shouldn_t_crash" tmin --input-file "crash.json"

This will repeatedly launch the fuzzer in “minify” mode and save the artifacts in the folder artifacts/crash.minified. The name of each artifact will be prefixed with the complexity of its input. For example, crash.minified/800--fe958d4f003bd4f5.json has a complexity of 8.00.

You can stop the minifying fuzzer at any point and look for the least complex input in the crash.minified folder.

Creating a Mutator

If you would like to fuzz-test your own custom type T without the DefaultMutator derive attribute or the make_mutator! procedural macro, you will have to create a type that conforms to the Mutator<T> trait.

Ask for help in the Github issues or send me an email if you would like some help or advice on how to write a good mutator.

Previous work on fuzzing engines

As far as I know, evolutionary, coverage-guided fuzzing engines were popularized by American Fuzzy Lop (AFL).
Fuzzcheck is also evolutionary and coverage-guided.

Later on, LLVM released its own fuzzing engine, libFuzzer, which is based on the same ideas as AFL, but it uses Clang’s SanitizerCoverage and is in-process (it lives in the same process as the program being fuzz-tested.
Fuzzcheck is also in-process and also uses SanitizerCoverage.

Both AFL and libFuzzer work by manipulating bitstrings (e.g. 1011101011). However, many programs work on structured data, and mutations at the bitstring level may not map to meaningful mutations at the level of the structured data. This problem can be partially addressed by using a compact binary encoding such as protobuf and providing custom mutation functions to libFuzzer that work on the structured data itself. This is a way to perform “structure-aware fuzzing” (talk, tutorial).

An alternative way to deal with structured data is to use generators just like QuickCheck’s Arbitrary trait. And then to “treat the raw byte buffer input provided by the coverage-guided fuzzer as a sequence of random values and implement a “random” number generator around it.” (cited blog post by @fitzgen). The tool cargo-fuzz has recently implemented that approach.

Fuzzcheck is also structure-aware, but unlike previous attempts at structure-aware fuzzing, it doesn't use an intermediary binary encoding such as protobuf nor does it use Quickcheck-like generators. Instead, it directly mutates the typed values in-process. This is better many ways. First, it is faster because there is no need to encode and decode inputs at each iteration. Second, the complexity of the input is given by a user-defined function, which will be more accurate than counting the bytes of the protobuf encoding. Finally, and most importantly, the mutations are faster and more meaningful than those done on protobuf or Arbitrary’s byte buffer-based RNG. A detail that I particularly like about fuzzcheck, and that is possible only because it mutates typed values, is that every mutation is done in-place and is reversable. That means that generating a new test case is super fast, and can often even be done with zero allocations.

As I was developing Fuzzcheck for Swift, a few researchers developed Fuzzchick for Coq (paper). It is a coverage-guided property-based testing tool implemented as an extension to Quickchick. As far as I know, it is the only other tool with the same philosophy as fuzzcheck. The similarity between the names fuzzcheck and Fuzzchick is a coincidence.