Unclear section in Overview docs

[PabloIsCoding]

https://github.com/odin-lang/odin-lang.org/blob/56ba2d01a1ea53a02b934b70733e9fb04ce8a864/content/docs/overview.md?plain=1#L670C1-L680C4

Hi, I was reading the Odin language overview and I found the above a bit confusing. A few reasons:

1. The way I understand this, if you pass arguments by value to a function, you are making a copy. It's when you use a pointer like in C that you do not copy. So why do the docs say that you need to generate a copy to have the "pointer behavior" (so to speak)?
2. If what it's meant is that Odin first makes a copy of the argument, and then you have to copy back the original thing that was passed, how is that done by a line like x := x?
3. Does this imply some computational penalty because we are making copies?

Since I don't really understand how this works I can't say how this should be rephrased (if at all) but I thought it might be helpful to give some feedback on this.

[Skytrias]

Here's how i understand it:

1. It's not really "pointer behavior" it's just that x := x wouldn't bee necessary in C. It's more meant to show that you can't mutate the passed parameter, unless it contains pointers like the slice, etc mentioned
2. You already get copies, it's just that you can't mutate them
3. As stated in the docs above the highlighted code, it should be good for optimizations

@gingerBill can explain this better

[gingerBill]

If you come from C, all parameters do this copy implicitly.

void foo(int x) { 
    x = 123; // I can change it
}

x in this case is a variable copy, and you can mutate it.

In Odin, you have explicitly copy the parameter to state you want to mutate it in some way.

foo :: proc(x: i32) {
    x = 123 // INVALID
}

foo :: proc(x: i32) {
    x := x
    x = 123 // allowed since `x` has been explicitly copied
}

As for (3), yes and Odin makes this clear. It wasn't clear in C before, even if in many cases the optimizer can remove it.