Motorola M68000 Betriebsanweisung PDF herunterladen (Seite 121)

Subroutines 107

HIGH RDDRESS

LOW ADDRESS

of course, SP. After the call, SP is pointing at the 32-bit return address value.

The first parameter is located at 4(SP) (assuming that the parameters were

pushed onto the stack in reverse order). Where would the second parameter

be? Well, that depends on the size of the first parameter. If the first

parameter is a longword, as in this example, then the second parameter is at

8(SP). Each of the other parameters can be found in a similar manner.

I should refresh your memory concerning register A7 (SP), the stack

pointer. When postincrement or predecrement addressing are used with a

byte-sized operand, SP is always modified by 2, even though the operand

is a byte. This is so the stack can always be aligned on an even-word

boundary. You should remember this fact if you pass a byte parameter on

the stack—always count it as two bytes. If you use a byte-sized instruction

for both the push and the pop, everything will work out okay.

The following instructions will obtain the three parameters from the

above example and place them into registers DO, Dl, and D2:

MOVE.L 4(SP),D0 PARM1

MOVE.L 8(SP),D1 PARM2

MOVE.L 12(SP),D2 PARM3

Of course, the parameters do not have to be moved to registers to be

used; they can be referenced directly.

This very common and simple method of passing parameters is re

ferred to as call by value. The actual value of the parameter is placed

on the stack. There are some limitations to this technique. First, it is only

practical to pass byte, word, or longword arguments on the stack. What

if we want to pass an entire array? Second, the subroutine has no way to

change the actual parameter; it can only change the copy it has on the

stack. Both of these limitations can be overcome by the use of a call by

PARM3

PARM2

PftRMl

RETURN ADDRESS

1 2 ... 116 117 118 119 120 121 122 123 124 125 126 ... 255 256

Keine Kommentare

Motorola M68000 Betriebsanweisung Seite 121