Specifications

CAVR-4
Part 1. Using the compiler
109
Using C++
IAR Systems supports two levels of the C++ language: The industry-standard
Embedded C++ and IAR Extended Embedded C++. They are described in this
chapter.
Overview
Embedded C++ is a subset of the C++ programming language which is intended for
embedded systems programming. It was defined by an industry consortium, the
Embedded C++ Technical Committee. Performance and portability are particularly
important in embedded systems development, which was considered when defining the
language.
STANDARD EMBEDDED C++
The following C++ features are supported:
● Classes, which are user-defined types that incorporate both data structure and
behavior; the essential feature of inheritance allows data structure and behavior to
be shared among classes
● Polymorphism, which means that an operation can behave differently on different
classes, is provided by virtual functions
● Overloading of operators and function names, which allows several operators or
functions with the same name, provided that there is a sufficient difference in their
argument lists
● Type-safe memory management using operators new and delete
● Inline functions, which are indicated as particularly suitable for inline expansion.
C++ features which have been excluded are those that introduce overhead in execution
time or code size that are beyond the control of the programmer. Also excluded are
recent additions to the ISO/ANSI C++ standard. This is because they represent potential
portability problems, due to the fact that few development tools support the standard.
Embedded C++ thus offers a subset of C++ which is efficient and fully supported by
existing development tools.
Standard Embedded C++ lacks the following features of C++:
● Templates
● Multiple and virtual inheritance
● Exception handling
● Runtime type information