Differences in Functionality Between /ND and /Aw Switches

MS-DOS                 | OS/2       | WINDOWS

The information in this article applies to:

• The Microsoft C/C++ Compiler (CL.EXE) included with:

      - Microsoft C for MS-DOS, versions 6.0, 6.0a, and 6.0ax
      - Microsoft C for OS/2, versions 6.0, and 6.0a
      - Microsoft C/C++ for MS-DOS, version 7.0
      - Microsoft Visual C++ for Windows, versions 1.0 and 1.5
  

This article describes the differences in functionality between the /Aw and /ND switches.

The /Aw switch is used by the compiler to determine if local variables (typically stack based) can be based on DS, as well as SS. For example, if a variable needs to be copied from local data to global data, which would normally include several MOV instructions, if the compiler can assume DS == SS, it can generate a LEA instruction on the stack-based variable and set up ES:DI with the destination. DS will already be set to the correct value. If DS != SS, the compiler will have to generate more code to correctly access the stack-based variable.

For the /ND switch, the data segment for the module is not the default segment. The compiler will do an explicit DS load at function entry point because of just this fact; that is, there is no way for the compiler to know that DS is properly set to the nonstandard segment name (not DGROUP). While this behavior is indeed the same behavior as the _loadds keyword, it has nothing to do with the main reason for /Aw (DS != SS).

To summarize, the /Aw switch doesn't cause the segment load, but it also doesn't stop it. This is one case where one switch (/ND) can override one of the effects of another switch (/Aw). However, the main effect of the /Aw switch is not changed (don't assume DS == SS).