Raygui bindings performance optimization via caching allocations

[kerudev]

I'm opening this issue so we don't forget the work that needs to be done.

As discussed in #14, we need to port the raylib bindings to a faster WASM runtime, like it was done for the raygui bindings.
This will allow us to use sync.Pool to cache variables instead of allocating and freeing them on each frame.

To work on this issue, the raygui bindings need to be completed first.

[BrownNPC]

@kerudev the simplest caching we can do right now is: allocate the return value in a global variable once. Eg.

var r_Font wasm.Ptr = wasm.MallocV[Font]() // keep a return vallue allocated.
//go:wasmimport raylib _LoadFont
func loadFont(ret wasm.Ptr, path wasm.Ptr)

func LoadFont(path string) Font {
    // copy string to C
    cPath := ...
    loadFont(r_Font,cPath)
    var ret Font // we dont need to allocate the return value!
    wasm.CopyValueToGo(r_Font,&ret)
    return ret
}

We should do this for functions that run every frame inside Raygui

[BrownNPC]

Due to Wasm limitations, we cannot avoid memory copying, but we can avoid allocations. Although before we double down on this. Make a simple benchmark example, eg. You said that clicking buttons was slow, so make a simple button app and implement this and see if it's actually faster.