Finding bugs with libFuzzer

To test your application, you'll need to write unit tests. This is a great way to make sure your software behaves correctly. However, due to the exponential number of possible inputs, you'll probably miss certain weird inputs, and a few bugs as well.

Fuzz testing can help here. The idea is to present your application with randomly generated data, or data based on valid input but with random changes. This is done over and over again, and so your application is tested with a large number of inputs. This is a very powerful testing approach. Literally hundreds of bugs in web browsers and other software have been found with fuzz testing.

LLVM comes with its own fuzz testing library. Originally part of the LLVM core libraries, the libFuzzer implementation was finally moved to compiler-rt. The library is designed to test small and fast functions.

Let's run a small example. You'll need to provide the LLVMFuzzerTestOneInput() function...