How do you allocate space well for multiple arrays in Y86 assembly?

Question

I am trying to convert C code into Y86 assembly code.

What happens if you have multiple array declarations such as:

int a[100], b[100];

Suppose every integer is 4 bytes. How do you know where in memory to point for the pos. directives so that you do not waste any space?

<Assembly code begins here>

...

halt

# Array initialization begins here
.pos ?
A:
.long 0

.pos ?   
B:
.long 0

Answer 1

Let the assembler worry about the offsets.
One possibility is to define a struct that contains local variables

struc Locals
    a dd ? dup 100
    b dd ? dup 100
ends Locals

sub esp, sizeof Locals ;; or perhaps sizeof struc Locals
mov ebp, esp           ;; take a copy of stack ptr

mov eax, [ebp + offset a] ;;  
mov ebx, [ebp + offset b] ;;  

In gcc the meaning of 'struct' is to modify the absolute position of current segment without introducing linkable code:

.file "temp.c"
    .struct 0
a:  .struct a + 4*100
b:  .struct b + 4*100
c:               ;; c will contain expression for the size
    .struct 0    ;; you can start a new struct here
a2: .struct a2 + 4   ;; this struct would be  'int a2;'
b2: .struct b2 + 8   ;;  'double b2'
c2:                  ;; sizeof struct #2

.text  
    sub c, %rsp
    mov $13, a(%rsp)   ;; a[0] = 13
    mov $2,  b(%rsp)   ;; b[0] = 2