Multiplication test bench
From Zet
This example tests these instructions:
- mul, imul and aad. Examples include:
- word / byte
- positive * positive
- negative * positive
- 0, 1, -1 multiply
- max range: ffff x ffff
- 2 random values
Code
.code16 start: movw $192, %sp # mul word movw $0x0003, %bx movw $0x0007, %ax movw $0xffff, %dx mulw %bx movw %ax, (0) movw %dx, (2) pushf movw $0xa320, %dx movw $0xffff, %ax mulw %dx movw %ax, (4) movw %dx, (6) pushf movw $0xffff, %ax movw $0x1, (8) mulw (8) movw %ax, (10) movw %dx, (12) pushf movw $0xffff, %ax movw $0xffff, (14) mulw (14) movw %ax, (16) movw %dx, (18) pushf movw $0x46db, %ax movw $0x0000, %bp mulw %bp movw %bp, (20) movw %ax, (22) movw %dx, (24) pushf movw $0x46db, %ax movw $0xeeeb, %si mulw %si movw %si, (26) movw %ax, (28) movw %dx, (30) pushf # mul byte movb $0x14, %bl movw $0xff07, %ax movw $0xffff, %dx mulb %bl movw %ax, (32) movw %dx, (34) pushf movb $0x24, %ch movw $0x00ff, %ax mulb %ch movw %ax, (36) movw %dx, (38) pushf movw $0xff, %ax movb $0x1, (40) mulb (40) movw %ax, (41) movw %dx, (43) pushf movw $0xffff, %ax movb $0xff, (45) mulb (45) movw %ax, (46) movw %dx, (46) pushf movw $0xc5, %ax movw $0x00, %dx mulb %dl movw %dx, (48) movw %ax, (50) pushf movb $0xb5, %al movb $0xf9, %dh mulb %dh movw %si, (52) movw %ax, (54) movw %dx, (56) pushf # imul word movw $0x0003, %bx movw $0x0007, %ax movw $0xffff, %dx imulw %bx movw %ax, (60) movw %dx, (62) pushf movw $0xa320, %dx movw $0xffff, %ax imulw %dx movw %ax, (64) movw %dx, (66) pushf movw $0xffff, %ax movw $0x1, (68) imulw (68) movw %ax, (70) movw %dx, (72) pushf movw $0xffff, %ax movw $0xffff, (74) imulw (74) movw %ax, (76) movw %dx, (78) pushf movw $0x46db, %ax movw $0x0000, %bp imulw %bp movw %bp, (80) movw %ax, (82) movw %dx, (84) pushf movw $0x46db, %ax movw $0xeeeb, %si imulw %si movw %si, (86) movw %ax, (88) movw %dx, (90) pushf # imul byte movb $0x14, %bl movw $0xff07, %ax movw $0xffff, %dx imulb %bl movw %ax, (92) movw %dx, (94) pushf movb $0x24, %ch movw $0x00ff, %ax imulb %ch movw %ax, (96) movw %dx, (98) pushf movw $0xff, %ax movb $0x1, (100) imulb (100) movw %ax, (101) movw %dx, (103) pushf movw $0xffff, %ax movb $0xff, (105) imulb (105) movw %ax, (106) movw %dx, (106) pushf movw $0xc5, %ax movw $0x00, %dx imulb %dl movw %dx, (108) movw %ax, (110) pushf movb $0xb5, %al movb $0xf9, %dh imulb %dh movw %si, (112) movw %ax, (114) movw %dx, (116) pushf # aad tests movw $0xff00, %ax aad movw %ax, (118) pushf movw $0xffff, %ax aad $0x12 movw %ax, (120) pushf movw $0x00ff, %ax aad $0xff movw %ax, (122) pushf movw $0x532d, %ax aad $0x39 movw %ax, (124) pushf hlt .org 65520 jmp start .org 65535 .byte 0xff
Results
Memory contents at the end should be:
0x00: 0x0015 0x0000 0x5ce0 0xa31f 0x0001 0xffff 0x0000 0xffff 0x10: 0x0001 0xfffe 0x0000 0x0000 0x0000 0xeeeb 0xa509 0x4220 0x20: 0x008c 0xffff 0x23dc 0xffff 0xff01 0xff00 0xffff 0xffff 0x30: 0x0000 0x0000 0xeeeb 0xb00d 0xf900 0xXXXX 0x0015 0x0000 0x40: 0x5ce0 0x0000 0x0001 0xffff 0xffff 0xffff 0x0001 0x0000 0x50: 0x0000 0x0000 0x0000 0xeeeb 0xa509 0xfb45 0x008c 0xffff 0x60: 0xffdc 0xffff 0xff01 0xffff 0xffff 0xffff 0x0000 0x0000 0x70: 0xeeeb 0x020d 0xf900 0x00f6 0x00ed 0x00ff 0x00a8 0xXXXX 0x80: 0xXXXX 0xXXXX 0xXXXX 0xXXXX 0x0892 0x0086 0x0097 0x0086 0x90: 0x0803 0x0046 0x0002 0x0086 0x0082 0x0883 0x0887 0x0046 0xa0: 0x0002 0x0086 0x0002 0x0002 0x0803 0x0046 0x0803 0x0086 0xb0: 0x0883 0x0082 0x0887 0x0046 0x0803 0x0086 0x0803 0x0002
Contents of addresses 0x80-0x8f, could be too:
0x80: 0xXXXX 0xXXXX 0xXXXX 0xXXXX 0x0082 0x0086 0x0086 0x0086
As some flags of aad doesn't care.
