Why do some functions use mov %edi, %eax as their very first instruction?

It copies the first argument into %eax (the return-value register) so the function can compute and return a result derived from that argument.

func4:
    mov  %edx, %eax
    sub  %esi, %eax
    ...

Common reasons:

1. Preparing for arithmetic — %eax is used for operations like cdq + idiv, or the result will be returned in %eax
2. Freeing up the argument register — %rdi/%rsi are caller-saved, so copying to %eax lets us use %rdi for something else (like a function call)
3. Direct return — a simple function might just massage the argument and return it

Note: Using %edi/%esi/%edx (32-bit) instead of %rdi/%rsi/%rdx (64-bit) tells you the arguments are int (4 bytes), not long (8 bytes).

Go deeper:

• System V AMD64 calling convention — %edi is the 1st argument, %eax the return-value register