The key factors affecting target code generation and code generator design include:
Input to the code generator: The code generator receives intermediate representation of the source program along with symbol table information. This input can be in various forms such as postfix notation, syntax trees, or three-address code.
Target program: The output of the code generator can be assembly language, relocatable machine language, or absolute machine language. Each type has different advantages, such as allowing separate compilation of subprograms or immediate execution.
Memory management: This involves mapping symbol table entries to memory addresses and managing the allocation of variables in the activation record.
Instruction selection: Efficient instruction selection is crucial for generating high-quality code. It involves choosing appropriate machine instructions to implement the intermediate code.
Register allocation: Proper utilization of registers can significantly improve code efficiency. This process involves deciding which variables should reside in registers and assigning specific registers to these variables.
Evaluation order: The sequence in which computations are performed can affect the efficiency of the target code and the number of registers required for intermediate results.
Target machine architecture: Understanding the target machine's instruction set and architecture is essential for effective code generation.
Instruction speeds and machine idioms: These factors are important when considering the efficiency of the target machine and the quality of generated code.
By addressing these factors, code generator designers can create more efficient and effective compilers that produce high-quality target code.