Getting Started with libFuzzer in Chrome

Prerequisites: libFuzzer in Chrome is supported with GN on Linux only.

This document will walk you through:

  • setting up your build enviroment.
  • creating your first fuzzer.
  • running the fuzzer and verifying its vitals.

Configure Build

Use use_libfuzzer GN argument together with sanitizer to generate build files:

Notice: current implementation also supports use_afl argument, but it is recommended to use libFuzzer for development. Running libFuzzer locally doesn't require any special configuration and quickly gives meaningful output for speed, coverage and other parameters.

# With address sanitizer
gn gen out/libfuzzer '--args=use_libfuzzer=true is_asan=true is_debug=false enable_nacl=false' --check

Supported sanitizer configurations are:

GN ArgumentDescription
is_asan=trueenables Address Sanitizer to catch problems like buffer overruns.
is_msan=trueenables Memory Sanitizer to catch problems like uninitialed reads[1].
is_ubsan_security=trueenables Undefined Behavior Sanitizer to catch[2] undefined behavior like integer overflow.
it is possible to run libfuzzer without any sanitizers; probably not what you want.

Write Fuzzer Function

Create a new .cc file and define a LLVMFuzzerTestOneInput function:

#include <stddef.h>
#include <stdint.h>

extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size) {
  // put your fuzzing code here and use data+size as input.
  return 0;
} is a simple example of real-world fuzzer.

Define GN Target

Define fuzzer_test GN target:

fuzzer_test("my_fuzzer") {
  sources = [ "" ]
  deps = [ ... ]

Build and Run Fuzzer Locally

Build with ninja as usual and run:

ninja -C out/libfuzzer url_parse_fuzzer

Your fuzzer should produce output like this:

INFO: Seed: 1787335005
INFO: -max_len is not provided, using 64
INFO: PreferSmall: 1
#0      READ   units: 1 exec/s: 0
#1      INITED cov: 2361 bits: 95 indir: 29 units: 1 exec/s: 0
#2      NEW    cov: 2710 bits: 359 indir: 36 units: 2 exec/s: 0 L: 64 MS: 0 

The ... NEW ... line appears when libFuzzer finds new and interesting input. The efficient fuzzer should be able to finds lots of them rather quickly. The ... pulse ... line will appear periodically to show the current status.

Improving Your Fuzzer

Your fuzzer may immediately discover interesting (i.e. crashing) inputs. To make it more efficient, several small steps can take you really far:

  • Create seed corpus. Add seed_corpus = "src/fuzz-testcases/" attribute to your fuzzer targets and add example files in appropriate folder. Read more in Seed Corpus section of efficient fuzzer guide. Make sure corpus files are appropriately licensed.
  • Create mutation dictionary. With a dict = "protocol.dict" attribute and key=value dicitionary file format, mutations can be more effective. See Fuzzer Dictionary.
  • Specify maximum testcase length. By default libFuzzer uses -max_len=64 (or takes the longest testcase in a corpus). ClusterFuzz takes random value in range from 1 to 10000 for each fuzzing session and passes that value to libFuzzers. If corpus contains testcases of size greater than max_len, libFuzzer will use only first max_len bytes of such testcases. See Maximum Testcase Length.

Disable noisy error message logging

If the code that you are a fuzzing generates error messages when encountering incorrect or invalid data then you need to silence those errors in the fuzzer.

If the target uses the Chromium logging APIs, the best way to do that is to override the environment used for logging in your fuzzer:

struct Environment {
  Environment() {

Environment* env = new Environment();

Submitting Fuzzer to ClusterFuzz

ClusterFuzz builds and executes all fuzzer_test targets in the source tree. The only thing you should do is to submit a fuzzer into Chrome.

Next Steps


[1]You need to download prebuilt instrumented libraries to use msan (crbug/653712):

GYP_DEFINES='use_goma=1 msan=1 use_prebuilt_instrumented_libraries=1' gclient runhooks

[2]By default UBSan doesn't crash once undefined behavior has been detected. To make it crash the following additional option should be provided:

UBSAN_OPTIONS=halt_on_error=1 ./fuzzer <corpus_directory_or_single_testcase_path>

Other useful options (used by ClusterFuzz) are:

UBSAN_OPTIONS=symbolize=1:halt_on_error=1:print_stacktrace=1 ./fuzzer <corpus_directory_or_single_testcase_path>