Using 32-bit segments for code and data can simplify porting an application from a 32-bit platform to the Windows environment when portions of the application can be isolated as a subroutine library. This subroutine library serves as a low-level engine but does not call Windows or MS-DOS functions.
As when the 32-bit memory is used exclusively for data storage, the USE16 code segment retains control of the program. Typically, the USE16 segment allocates the 32-bit memory, creating one or more objects for code and data. In addition to the data-management tasks described in Section 12.3, “Considerations for Using 32-Bit Memory,” the USE16 segment also loads the subroutine code into one of the 32-bit segments, fixes up the pointers in the code as required, and creates a code-segment alias to permit the code to be executed. The USE16 code segment is responsible for maintaining control of the program flow, calling into the USE32 code segment when it requires the low-level services of the subroutine library.