Using % to narrow the range of a random number produces a slightly non-uniform distribution in the output. For example, ipc_fuzzer::RandInRange(100) produces `0` 1.0000000012% of the time but `99` 0.999999978% of the time.

Using reinterpret_cast to produce a double yields, for example, values between 0.5 and 1 twice as often as values between 1 and 1.5.

I don't know how much of a problem this could be in practice for libFuzzer.

The C++ distributions are designed to avoid this bias: http://en.cppreference.com/w/cpp/numeric/random#Random_number_distributions