FIX: Whitespace Character Must Precede DUP Operator

MS-DOS           | OS/2

The information in this article applies to:

• Microsoft Macro Assembler for MS-DOS, versions 6.0, 6.0a, and 6.0b
• Microsoft Macro Assembler for OS/2, versions 6.0, 6.0a, and 6.0b

SYMPTOMS

When using the DUP operator in the Microsoft Macro Assembler (MASM), you must place a whitespace character before the DUP operator if using the "?" operator. If no whitespace is present, no error message is produced by the assembler, and incorrect code will be generated. The assembler will generate one data item for the specified type, initialized to the value of the repeat size for the DUP operator.

RESOLUTION

To work around the problem, place a space between the repeat size and the DUP operator. Initializing the variable will cause a syntax error to be generated.

STATUS

Microsoft has confirmed this to be a problem in MASM version 6.0, 6.0a, and 6.0b. This problem was corrected in MASM for MS-DOS version 6.1.

MORE INFORMATION

The code example below generates a listing file that shows the size generated for the respective DUP statements. It can be seen that the first DUP operation generates 4 data bytes; however, the second generates 1 data byte initialized to 4.

The assembler has problems parsing this statement because the DUP operator begins with a letter (d), which is a valid radix suffix.

Sample Code

; Assemble options needed: /Fl

.MODEL small .STACK 4096

.DATA

a       DB 4 DUP(0)

b       DB 4DUP(?)

.CODE start:

   mov ax, 4C00H
   int 21H      ; Terminate Process