For C++, `bit_cast<uint32_t>(0.f)` should be Well Defined, right? I'm curious, in C, is union-casting float->uint32_t also Perfectly Legal And Well Defined?

(I am not a C or C++ expert.)