Method for converting in different methods between f32/f64 and others

[VivIsBee]

With floats/doubles you could directly reinterpret the bits or you can round it down, and WASM has both. For now I'm assuming the former.

[SkuldNorniern]

Currently, I have only thought about the former method, and since WASM supports both. Supporting both on our side would be nice to do, how do you think?

[VivIsBee]

Probably multiple instructions? Like maybe one to transmute and one to round down?

[SkuldNorniern]

Or maybe use a flag to determine the transmute/round down behavior
For now, multiple instructions seem like they're easy to separate the two, so I'll mark the code with a TODO for supporting that