Transcript Arrays and Pointers

Arrays and Pointers
• Variables are just names of memory
locations
• An array is a variable (just with many
memory locations – all contiguous)
• Address = pointer (in function, not C++
syntax)
• Address = pointer in assembler (no data
typing)
Allocating an Array
• You have to know the size in bytes of each
element of the array
– Int = 4 bytes
– Float = 4 bytes
– Char = 1 byte
• Determine the number of bytes needed
– Size*quantity
• Allocate space (in bytes)
arr:
.space
100
Memory Access
• Remember, the REAL assembly language
can only access memory via number($R)
– Or disp(base) where disp is a number and
base is a register
– The address is the value in the base register
added to the displacement
The la command
• The la (load address) macro command is
very helpful (especially with arrays)
• We will load the address of the data
section.
• The data section always starts at memory
location x10010000.
– Look at the first 4 hex char (x1001).
– This is 4097 in decimal
The lui command
• Used by the assembler a lot
– Some instructions only have a 16 bit operand
– To get a 32 bit value into a register, we need
two instruction
• lui – load upper immediate
– Loads the upper 2 bytes (16 bits) of the register with the
specified value and places 0’s in the last 16 bits
– ori – or immediate – logically or’s the last 16 bits of a
register with the immediate operand.
Back to the la
• So, to load the 32 bit address of a variable into
$t1, the assembler uses:
– lui $t1,4097 #now t1 has the addr of the data section
– ori $t1,offset #the distance the var is from the beginning of the .data
• This replaces the la $t1,var instruction
Working with Arrays
• First, get the address of the array into a
register
– la $t1,arr
• To get to element i of an array, you
multiply i by the size of an element and
add that to the base address of the array.
– This gives the address of arr[i]
Processing an entire array
• If you are looping through an array, rather
than multiply, you could just add the size
of an element to the “current” address to
get the address of the next element
• Example:
for (i=0; i<25; i++)
arr[i]=i;
In assembler
• In the .data section
arr: .space100
• In the .text section
la
$t1,arr
li
$t2,0
li
$t3,25
loop: beq
$t2,$t3,done
sw
$t2,0($t1)
addi $t1,4
addi $t2,1
j
loop
done:
#i
#limit
#address increases by 4
#counter increases by 1
Other examples
• Look at the reverse string example
• Look at the binary search example
• Look at the sorting example