TNS/E Native Application Conversion Guide

C and C++ Conversion Tasks

TNS/E Native Application Conversion Guide—529659-003

3-3

Converting Code to Use 32-Bit Pointers and Integers

TNS/R code cannot be executed on a TNS/E platform, and TNS/E code cannot be

executed on a TNS/R platform.

For c89 and c99 conversion information, see Section 11, OSS API and Utilities

Conversion Tasks.

All code generated by the TNS/E native compilers is position-independent code (PIC).

PIC is code that need not be modified to run at different virtual addresses and is used

to create dynamic-link libraries (DLLs).

For details on compiling and linking native C and C++ programs, see the

C/C++

Programmer’s Guide

Converting Code to Use 32-Bit Pointers and

Integers

The memory model determines the size of pointers. In the TNS C environment, there

are two memory models: the small-memory model (16-bit pointers) and the large-

memory model (32-bit pointers).

The data model determines the size of type int. In the TNS C and C++ environments,

there are two data models: the 16-bit data model and the 32-bit (or wide) data model.

In the native C and C++ environments, there are only the large-memory model and the

32-bit data model.

You must convert existing Guardian C programs that use the small-memory model to

the large-memory model. You must also convert existing Guardian C and C++

programs that use the 16-bit data model to the 32-bit data model. (You do not need to

convert existing OSS C and C++ programs because OSS supports only the large-

memory model and the 32-bit data model.)

The memory and data models a program uses are determined by the environment in

which you run the compiler (Guardian or OSS), the version of the compiler, the

SYSTYPE pragma setting, and the explicit XMEM, NOXMEM, WIDE, or NOWIDE pragmas

in the source code. D-series versions of the Guardian TNS C compiler and D20

versions of the Guardian C++ preprocessor generate programs that use the large-

memory model and the 16-bit data model by default. See the

C/C++ Programmer’s

Guide

for further details.

Before compiling small-memory model or 16-bit data model programs with the native

compilers, it is often easier to first convert them to the large-memory model and 32-bit

data model using the TNS C compiler. Follow these guidelines for doing the

conversion:

•

Specify the XMEM and WIDE pragmas in your source code.

•

Use the TNS C compiler with the STRICT pragma.