06 Procedure call

1 .Procedure Call Procedure Main Typically procedure call uses a stack. What is a stack? Question. Can’t we use a jump instruction to implement a procedure call? 

2 .

3 .The stack Occupies a part of the main memory. In MIPS, it grows from high address to low address as you push data on the stack. Consequently, the content of the stack pointer ($sp) decreases. Low address $sp Item 2 Item 1 Stack pointer High address High address 

4 .Use of the stack in procedure call Before the subroutine executes, save registers (why?). Jump to the subroutine using jump-and-link (jal address) (jal address means ra ← PC+4; PC ← address) For MIPS, (ra=r31) After the subroutine executes, restore the registers. Return from the subroutine using jr (jump register) (jr ra means PC ← (ra)) Example of a function call int leaf (int g, int h, int i, int j) { int f; f = (g + h) – (i + j); return f; } The arguments g, h, i, j are put in $a0-$a3. The result f will be put into $s0, and returned to $v0. 

5 .The structure of the procedure Leaf: addi $sp, $sp, -12 # $sp = $sp-12, make room sw $t1, 8($sp) # save $t1 on stack sw $t0, 4($sp) # save $t0 on stack sw $s0, 0($sp) # save $s0 on stack The contents of $t1, $t0, $s0 in the main program will not be overwritten. Now we can use them in the body of the function. add $t0, $a0, $a1 # $t0 = g + h add $ t1, $a2, $a3 # $t1 = i + j sub $s0, $t0, $t1 # $s0 = (g + h) – (i + j) Pass g,h,i,j into $a1-$a3 Main Procedure Return result f into $v0 

6 .Return the result into the register $v0 add $v0, $s0, $zero # returns f = (g+h)-(i+j) to $v0 Now restore the old values of the registers by popping the stack. lw $s0, 0($sp) # restore $s0 lw $t0, 4($sp) # restore $t0 lw $t1, 8($sp) # restore $t1 addi $sp, $sp, 12 # adjust $sp Finally, return to the main program. jr $ra # return to caller.