notes

TODO:
1) make a first pass, this first pass will assume largest immediate values where labels are used for the registers given.
   this way we can actually calculate the size of an instruction without needing to know the label addresses when used in place
   of an immediate.

2) now that we know the size of each instruction, the label values can be created to match

3) now that we have the labels to match, we can create a second pass that will see how small the immediates where labels are used
   can be and with what registers we use. this will give us a new instruction size potentially (and most likely).
   NOTE: this is also where instruction size optimization occurrs. and where NASM likely turns that mov rdi, 69 you had into
   mov edi, 69 you saw in ghidra


4) now that we know the actual size of the instruction, we need to calculate the new label address

5) now that every label should have the correct address, encode all of the instructions.