Code_Optimization_Compiler_Design software .pptx

Code Optimization in Compiler
Design
• Last Updated: 04 Sep, 2024
• Enhancing performance and efficiency of
executable code

Introduction to Code Optimization
• Improves execution time
• Minimizes resource usage
• Enhances overall system performance
• Applied during compilation without altering
program functionality

Objectives of Code Optimization
• Must preserve program meaning
• Increase program speed and performance
• Keep compilation time reasonable
• Avoid compile-time delays

When and Why to Optimize?
• Performed at development end-stage
• May reduce readability
• Used to:
• - Reduce space and increase compilation
speed
• - Automate tedious optimization tasks
• - Promote code reusability

Types of Code Optimization
• 1. Machine Independent:
• - Improves intermediate code
• - No CPU/memory-specific details
• 2. Machine Dependent:
• - Optimizes target code
• - Utilizes registers, absolute memory

Ways to Optimize Code
• 1. Compile Time Evaluation
• 2. Variable Propagation
• 3. Constant Propagation
• 4. Constant Folding
• 5. Copy Propagation
• 6. Common Sub Expression Elimination
• 7. Dead Code Elimination
• 8. Unreachable Code Elimination

Examples of Optimization
• Compile Time Eval:
• A = 2*(22.0/7.0)*r (computed at compile
time)
• Constant Folding:
• x = 2 * 5 -> x = 10
• Copy Propagation:
• Replace x with a if x = a

Unreachable Code Example
• Before:
• cout << "GFG!";
• return 0;
• cout << num;
• After:
• cout << "GFG!";
• return 0;

Function Optimization Techniques
• Function Inlining
• Function Cloning (e.g., Overloading)
• Strength Reduction:
• a = a * 16 -> a = a << 4

Loop Optimization Techniques
• 1. Code Motion
• 2. Loop Jamming
• 3. Loop Unrolling

Example: Loop Jamming
• Before:
• for(...) x = ...;
• for(...) y = ...;
• After:
• for(...) { x = ...; y = ...; }

Where to Apply Optimization?
• Source Program: Algorithm, loops
• Intermediate Code: Address transformations
• Target Code: Registers, select/move
instructions

Optimization Levels
• Local: Basic blocks (e.g., Value Numbering)
• Regional: Extended blocks (e.g., Loop
Unrolling)
• Global: Functions and loops (e.g., Live Var
Analysis)
• Interprocedural: Across procedures (e.g.,
Inline Substitution)

Advantages of Code Optimization
• Better performance and speed
• Reduced code size and power usage
• Improved portability
• Easier maintenance

Disadvantages of Code
Optimization
• Increases compilation time
• Adds complexity
• Risk of bugs
• Hard to assess benefits

Conclusion
• Code optimization is vital for efficient compiler
design
• Includes various techniques (loops, dead code,
folding)
• Leads to faster, smaller, and efficient code

Code_Optimization_Compiler_Design software .pptx

More Related Content

Similar to Code_Optimization_Compiler_Design software .pptx (20)

Recently uploaded (20)

Code_Optimization_Compiler_Design software .pptx