click on icon to download file

Avoiding Common Errors Committed During Programming

Point 1

Size of the name given to .asm file – File name has to be less than 8 characters. So if you a name a file as myfirstprog.asm and try to assemble it the assembler will throw an error saying file not found.

• Keep the name of the file short and less than 8 characters

Point 2

When saving a asm file saving it as a text document will name the file as myprog.asm.txt.

• Save the file using all file type option or better use notepad++ which allows  to save the file as an asm file.
• Download notepad++ from site https://notepad-plus-plus.org/download

Point 3

Size of data

(a) If you are working with bytes (8-bit data)

• data should be saved as follows

dat1 db 23h,56h,72h

• data should be transferred from memory into an 8-bit register
  

mov al,[si]

using the following is incorrect

mov ax,[si]

mov eax,[si]

• Pointer increment or decrement should be done by one

inc si /dec si

(b) If you are working with 16-bit data

• data should be stored as follows

dat1 dw 6723h,1A56h,72ACh

• data should be transferred from memory into a 16-bit register

mov ax,[si]

using the following is incorrect

mov al,[si]

mov eax,[si]

• pointer increment or decrement should be done by two

inc si

inc si  

(c) If you are working with 32-bit data

• data should be stored as follows

dat1 dd 9178AA23h,0AABBCC56h, 1122334472h

• data should be transferred from memory into a 32-bit register

mov eax,[si]

using the following is incorrect

mov al,[si]

mov ax,[si]

• Pointer increment or decrement should be done by four

inc si

inc si

inc si

inc si

Point 4

If you are using 16-bit registers as pointers you can use only si, di, bx and bp. When using bp only relative addressing is allowed. An offset should always be specified with bp,

mov bl,[bp+0]

using register indirect addressing as shown below is incorrect

mov bl,[bp]

The assembler does not throw an error – because it automatically corrects it to mov bl,[bp+0]

Point 5

When storing a string of alphanumeric characters – all operations will be in bytes only. Each character is represented by an 8-bit ASCII equivalent. So storage must be as follows

str1 db “a,b,c,d”

using the following is incorrect

str1 dw “a,b,c,d”

str1 dd “a,b,c,d”

All rules related to byte transfer apply here. 

Point 6

.EQU directive can be used only for declaring constants – it cannot be used for storing data in memory. So if while attempting to do

cnt1 equ 5

and then try to so something like

dec cnt1

system will throw an error.

So the correct method would be

cnt1 db 5

dec cnt1

Point 7

The size of source and destination must be the same add si,cl is incorrect – if contents of cl needs to be added with si – use the following method

mov ch,0

add si,cx 

Point 8

When using .model tiny use ml for assembly and linking – masm followed by link should not be used.

Point 9

The code segment starts with .startup and ends with .exit. If .startup is missing – segment registers will not be initialized.

Point 10

For executing the complete code if ip is pointing to 100 – execute until the address where mov ah,4ch followed by int 21h is present. For e.g. If

cs:112 mov ah,4ch

cs:114 int 21h

execute up to 112 using either g 112 (if ip is at 100) or using g =100 112 (if ip is not at 100) do not use g as int 21h – causes the segment values to change.

Point 11

When using string instructions the source address is always a combo of ds:si and destination is always es:di.

Point 12

rep/repe/repne prefix can be used only with string instructions.

Point 13

macro is defined before .code and .startup

For e.g.

capson    macro
                cmp        al,'a'
                jl        x2
               cmp        al,'z'
               jg        x2
              sub        al,20h
x2:       
             endm

.code
.startup

Point 14

When using subroutines, software int instructions - stack must always be initialized and hence also the stackpointer

For e.g.

st1        dw        10 dup(?)

st2       dw          ?

In the code

lea       sp,st2

[Note: st2 offset must be loaded into sp not st1 as stack moves backwards]

Point 15

Stack is usually initialized as word not bytes - as stack transfer is usually in words

Point 16

Data entered from keyboard/ read or written into files all are in ASCII format - so if you want display 6 or store 6 in file the value you have to give is 36. If you enter 3 from keyboard or read 3 from file value will be ASCII so you will have 33 in register or memory

Point 17

DAA can be used only after addition with the accumulator - the rules of programming will have to be followed.