DOCERR: MASM 6.0 Programmer's Guide Errors: Chapter 7

MS-DOS           | OS/2

The information in this article applies to:

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

The following is a list of documentation additions and corrections for chapter 7 of the "Microsoft Macro Assembler Programmer's Guide" version 6.0 manual. The section and page numbers are listed first, followed by a description of the addition or correction.

Section 7.1, Page 167

The first paragraph of this section should read:

   Jumps are the most direct method for changing program control from
   one location to another. At the processor level, jumps work by
   changing the value of the IP (Instruction Pointer) for near jumps,
   and by changing the values of the CS and IP registers for far
   jumps. The IP register is loaded with the offset of the target
   address and CS is loaded with the segment of the target address.
   The many forms of the jump instruction handle jumps based on
   conditions, flags, and bit settings.

Section 7.1.1.2, Page 169

The sample line

   jmp ebx     ; FAR32 jump

   jmp ebx     ; NEAR32 jump

Section 7.1.1.2, Page 170

The example shows the following TYPEDEF statement:

   NPVOID  TYPEDEF NEAR PTR VOID

   NPVOID  TYPEDEF NEAR PTR

Section 7.1.2.2, Page 173

Table 7.1, on page 173, uses the same character for the number zero and the letter "O." This may cause some confusion when reading the table. The "OF" indicates the overflow flag and does not represent the hexadecimal number 0F.

Section 7.1.2.6, Page 176

In the sample code, the line

  .IF  cx = 20

  .IF  cx == 20

Section 7.2, Page 177

The description of the LOOPE, LOOPZ, LOOPNE, and LOOPNZ states:

   Loops while equal (or not equal). Checks CX and a condition. The
   loop ends when the condition is true. Set CX to a number out of
   range if you don't want to control the loop.

   Loops while equal (or not equal). Checks CX and a condition
   (indicated by the zero flag). The loop continues while the
   condition is met and CX is not zero. Set CX to a number out of
   range if you don't want the CX count to control the loop.

Section 7.3.2, Page 187

The sample program contains a line that states:

   mov sp, bp

Section 7.3.3.2, Page 191

This first paragraph states:

   The parameters are separated from the reglist by a comma if there
   is a list of registers.

For example,

   myproc PROC FAR C PUBLIC USES di si, var1:word, arg1:VARARG

   myproc PROC FAR C PUBLIC USES di si var1:word, arg1:VARARG

Section 7.3.3.3, Page 193

The sample program contains the line:

   dec arg1

   dec argcount

Section 7.3.6, Page 199

Page 199 shows two sample lines for declaring prototypes. They are:

   addup  PROTO NEAR C argcount:WORD, arg2:WORD, arg3:WORD

   myproc PROTO FAR  C, argcount:WORD, arg2:VARARG

Section 7.3.7.5, Page 202 and 203

In the first example, the code is missing a few steps. The following is the corrected code:

   .CODE
   mov      bx, pfunc                  ; pfunc is the Function Table
   mov      si, Num                    ; Num contains 0 or 2
   INVOKE   FUNCPTR PTR [bx+si], 1, 1  ; Selects proc1 or proc2

   You can also use ASSUME to accomplish the same task. The ASSUME
   statement associates the type PFUNC with the BX register.

   You can also use ASSUME to accomplish the same task. In the sample
   below, the ASSUME statement associates the type FUNCPTR with the BX
   register.

   mov      bx, pFunc

   INVOKE   [bx+si], 1, 1

Section 7.3.8.2, Page 205

Page 205 shows the epilogue code when <LOADDS> is specified on a PROC line. One line states:

   mov sp,bp

   mov sp,bp      ; if localbytes is not 0

Section 7.4.1, Page 209

The sample code demonstrates how to print a message on the display using a DOS interrupt call. It shows the following code:

   .DATA
   msg BYTE "This writes to the screen",$
   .CODE
   mov dx,offset msg
   mov ah,09h
   int 21h

   .DATA
   msg BYTE "This writes to the screen$"
   .CODE
   mov ax,SEG msg
   mov ds,ax
   mov dx,offset msg
   mov ah,09h
   int 21h

Section 7.4.2, Page 211

The last paragraph states:

   To replace the address in the interrupt descriptor table with the
   address of your procedure, AL needs to be loaded with 04h and AH
   loaded with 35, the Get Interrupt Vector function. The Set
   Interrupt Vector function requires 25 in AH.

Section 7.4.2, Page 212

Change the line

   .MODEL LARGE,C,DOS

   .MODEL LARGE,C,OS_DOS

   Before your program ends, you should restore the original address
   by loading DS and DX with the original interrupt address and using
   the DOS set vector function to store the original address at the
   correct location.