Support for 'deferred_none' on polymorphic and overloaded procedures

[ryantrawick]

Hello. I've lately needed to use @(deferred_in = proc_name) on polymorphic and overloaded procedures. It currently gives this error for polymorphic procedures: Error: 'deferred_in' cannot be used with a polymorphic procedure

I understand why deferred_in and deferred_out aren't really possible, or would be annoying to support, but it seems like deferred_none should be reasonably straightforward if I understand how they work?

This would be the ideal if the in/out versions could work:

over_d1 :: proc($T1: typeid) {
	// ...
}

over_d2 :: proc($T1, $T2: typeid) {
	// ...
}

overload_deferred :: proc {
	over_d1,
	over_d2,
	// ...
}

f1 :: proc($T1: typeid) -> Out_Struct1(T1) {
	// ...
}

f2 :: proc($T1, $T2: typeid) -> Out_Struct2(T1, T2) {
	// ...
}

@(deferred_in = overload_deferred)
function :: proc {
	f1,
	f2,
	// ...
}

Or just this, assuming the overload would still call the deferred procedure:

over_d1 :: proc($T1: typeid) {
	// ...
}

@(deferred_in = over_d1)
f1 :: proc($T1: typeid) -> Out_Struct1(T1) {
	// ...
}

Basically I want to be able to "finalize" something only when used, without having to remember to write defer pkg.finalize(foo) every time, similar to the IMGUI example that the deferred examples use. I don't know if this is something that many people would want or use though.