ECE 468: Introduction to Compiler and Translation Systems Engineering

8/16/15: Webpage live!

This course focuses on the tools and techniques needed to build an optimizing compiler. Topics include:

1. Scanning and parsing: determining the syntactic structure of a program
2. Semantic routines: determining the semantics of a program and building an intermediate representation
3. Code generation: emitting assembly code that is equivalent to the program
4. Program optimizations: improving the performance of a program
5. Program analysis: determining interesting information about a program's behavior
6. Why we shouldn't be like Mel.

Instructor:	Milind Kulkarni
	milind 'at' purdue 'dot' edu
	Office hours: Mondays 2:30pm–4pm and Thursdays 10am–11:30am
	EE 324A
TA:	Aurangzeb
	orangzeb 'at' purdue 'dot' edu
	Office hours: Wednesdays 5:30pm–7pm and Fridays 2:30pm–4pm
	EE 206
Calendar	HTML
	iCal

The course syllabus discusses most of the details of the course, including a tentative schedule of topics, a brief description of the project and other administrative information.

The discussion forum for the class is on Piazza.

The project page is available here. Keep a close eye on the project page for project-related announcements. Any changes to project requirements will be posted there.

• Lecture 1: What is a Compiler?. Also available 6-to-a-page.
• Lecture 2: Scanners and Finite Automata. Also available 6-to-a-page.
• Lecture 3: Parsers and Context-free Grammars. Also available 6-to-a-page.
• Lecture 4: Semantic actions, symbol tables and expressions. Also available 6-to-a-page.
• Lecture 5: Semantic actions for control structures. Also available 6-to-a-page.
• Lecture 6: Functions. Also available 6-to-a-page.
• Lecture 7: Code generation and local optimization. Also available 7-to-a-page.
• Lecture 8: Instruction scheduling. Also available 6-to-a-page.
• Lecture 9: Control flow graphs. Also available 6-to-a-page.
• Lecture 10: Dataflow analysis: constant propagation. Also available 6-to-a-page.
• Lecture 10.5: Lattice theory. Also available 6-to-a-page.
• Lecture 11: Bitvector dataflow. Also available 6-to-a-page.
• Lecture 12: Loop optimizations. Also available 6-to-a-page.
• Lecture 13: Dependence analysis. Also available 6-to-a-page.
  

Problem sets are due on the date shown. You may submit them by email before class, or turn in a hard copy at the beginning of class. (Not the end!)

• Problem set 1: Regular Expressions and Finite Automata. (Solutions (Updated!))
• Problem set 2: Context Free Grammars and Parsers. (Solutions)
• Problem set 3: Function calls and Common subexpression elimination. (Solutions)
• Problem set 4: Local register allocation and instruction scheduling. (Solutions)
• Problem set 5: Dataflow analysis and loop transformations. (Solutions)