Programming for the Java virtual machine

The core of Java technology, the Java virtual machine is an abstract computing machine that enables the Java platform to host applications on any computer or operating system without rewriting or recompiling. Anyone interested in designing a language or writing a compiler for the Java virtual machine must have an in-depth understanding of its binary class format and instruction set. If you are programming with the Java programming language, knowledge of the Java virtual machine will give you valuable insight into the Java platform's security capabilities and cross-platform portability. It will increase your understanding of the Java programming language, enabling you to improve the security and performance of your programs.

The author employs a tutorial approach that provides a detailed look into the central workings of the technology and teaches the reader how to write real programs for the Java virtual machine. He describes methods for becoming a better programmer through an advanced understanding of the Java virtual machine and Java technology. Programming for the Java Virtual Machine offers comprehensive coverage of all the major elements of the Java virtual machine--classes and objects, control instructions, debugging, class loaders, compiling the Java programming language, performance issues, security, and threads and synchronization. The book provides an introduction to the Java Virtual Machine Specification (JVMS), with a collection of topics that help programmers understand the Java virtual machine and the JVMS better. In addition, the book features implementations of Prolog and Scheme, a language that runs on top of the Java virtual machine, generating Java virtual machine code as it runs and using a Java virtual machine class loader to load the generated code into the system.

You will find detailed information on such topics as:

• The Java virtual machine verification algorithm
  
• How Java virtual machine security works, and what it can and can't do
  
• Using class loaders to incorporate code and dynamically generated code from the Internet, the Java Foundation Classes, database queries, and other languages
  
• The mechanics of compiling the Java programming language for the Java virtual machine
  
• Implementing other languages using the Java virtual machine, including Scheme, Prolog, Sather, Eiffel, and regular expressions
  

Numerous examples illustrate techniques and concepts, and exercises with solutions help you gain practical experience.

Contents

Preface xvii

1 Introduction to the Java Virtual Machine 1

1.1 Java Virtual Machine and Java 2

1.2 What Is the Java Virtual Machine? 3

1.2.1 Instruction Set 3

1.2.2 class File Format 5

1.2.3 Verification 5

1.3 Java Platform 6

1.4 Organization of the Java Virtual Machine 7

1.4.1 Class Area 7

1.4.2 Java Stack 9

1.4.3 Heap 9

1.4.4 Native Method Stacks 11

1.4.5 Example 11

1.5 Garbage Collection 15

1.5.1 Finalization 17

1.5.2 Timing of Garbage Collection 19

1.6 JVM Implementations 19

1.7 Versions of the Virtual Machine 20

1.8 Future of the Java Virtual Machine 21

2 Oolong 25

2.1 Hello, World 25

2.1.1 .class and .super Directives 26

2.1.2 main Method 27

2.1.3 Instructions 28

2.2 Structure of an Oolong Program 29

2.3 Defining the Class 30

2.4 Superclasses 31

2.5 Declaring Fields 31

2.6 Declaring Methods 34

2.7 Interfaces 38

3 Writing Methods 41

3.1 Some Definitions 41

3.2 Types 42

3.2.1 Numeric Types 43

3.2.2 Object Types 43

3.2.3 Type returnAddress 44

3.3 Mnemonic Naming Conventions 44

3.4 Testing Code Examples 45

3.5 Returns 47

3.6 Constants 48

3.6.1 General Constants 48

3.6.2 Small Constants and null 48

3.7 Local Variables 50

3.7.1 Initializing Variables 52

3.7.2 Local Variables as Method Arguments 55

3.8 Math 56

3.8.1 Arithmetic Operations 56

3.8.2 Nonnumbers and Infinity 58

3.8.3 Bitwise Operations 58

3.8.4 Floating-Point Arithmetic and strictfp 61

3.9 Type Conversions 62

3.9.1 Conversion Example 63

3.9.2 Widening and Narrowing 64

3.10 Stack Manipulations 65

3.11 Arrays of Numbers 68

3.12 Just Biding Your Time 70

4 Classes and Objects 73

4.1 What Is an Object? 73

4.1.1 Objects and References 74

4.1.2 Class Names 76

4.1.3 Packages 76

4.2 Creating Objects 77

4.3 Constructors 78

4.4 Using Fields 79

4.4.1 Types and getfield 81

4.4.2 Inheriting Fields 82

4.4.3 Changing Field Values 82

4.5 Invoking Methods on Objects 83

4.5.1 Virtual Invocation 86

4.5.2 Method Inheritance and Overriding 87

4.5.3 Overloading and Method Matching 90

4.5.4 Invoking a Method Directly 92

4.5.5 invokespecial and super 93

4.6 Casting 95

4.7 Casting, Fields, Methods, and Java 97

4.8 Returning Objects 100

4.9 Interfaces 101

4.10 Static Fields and Methods 105

4.11 Class Initialization 106

4.12 Arrays 107

4.12.1 Arrays of Reference 108

4.12.2 Multidimensional Arrays 110

4.12.3 Length of Arrays 112

5 Control Instructions 115

5.1 Comparing Integers 115

5.2 Comparing References 118

5.3 Other Comparisons 120

5.4 Switches 122

5.5 Subroutines 123

5.6 Exceptions 124

5.6.1 Multiple Exception Handlers 129

5.6.2 .throws Directive 131

5.6.3 Other Ways to Throw Exceptions 132

5.7 A Gentler Way of Checking Type 132

6 Verification Process 135

6.1 How the Verification Algorithm Works 136

6.2 Is It a Structurally Valid class File? 136

6.3 Are All Constant References Correct? 138

6.5 Will Each Instruction Always Find a Correctly Formed Stack

and Local Variable Array? 144

6.5.1 Example 1: Hello, World 146

6.5.2 Example 2: gotos and ifs 148

6.5.3 Example 3: Code That Doesn't Verify 152

6.5.4 Example 4: Dealing with Subclasses 154

6.5.5 Algorithm 158

6.6 Do External References Check Out? 160

6.7 Java Language and Verification Algorithm 161

6.7.1 Fooling the Virtual Machine with a Java Compiler 163

6.7.2 Other Languages and the Verification Algorithm 164

6.8 Other Safety Requirements 164

6.9 Checking Verification 165

7 Debugging 167

7.1 Debugging Directives in Oolong 167

7.1.1 .source Directive and .line Directives 168

7.1.2 .var Directive 170

7.2 Debugging Oolong Programs 172

7.3 Runtime Tracing 173

8 Class Loaders 177

8.1 How Classes Are Loaded 177

8.1.1 Loading 178

8.1.2 Linking 179

8.2 Loading Classes without a Class Loader 180

8.3 Making Your Own Class Loader 180

8.3.1 Caching Classes 182

8.3.2 Examples 183

8.4 Working with Class Objects 184

8.5 Reflection 186

9 Inside a Java class File 189

9.1 class File Header 190

9.2 Constant Pool 190

9.2.1 UTF8 Entries 193

9.2.2 Constant Entries 193

9.2.3 Field, Method, and Class Entries 195

9.3 Class Information 195

9.4 Fields and Methods 197

9.5 Class Attributes 202

9.6 Other Attributes 203

9.6.1 ConstantValue Attribute 203

9.6.2 Exceptions Attribute 204

9.6.3 InnerClasses Attribute 205

10 Compiling Java 207

10.1 Expressions and Statements 208

10.2 Expression Types 210

10.2.1 Numeric Types 210

10.2.2 Object Types 212

10.2.3 Expression Result Types 212

10.2.4 Compile-Time Types and Type Restrictions 213

10.2.5 Type Safety and Verifiability 215

10.3 Compiling Statements and Expressions 215

10.4 Local Variable Declarations 217

10.5 Fields and Variables 219

10.5.1 Abbreviations 222

10.5.2 Other Assignments 223

10.5.3 Assignments as Expressions 224

10.5.4 Type-Checking in Assignments 225

10.6 Creating New Objects 226

10.7 Class Name Abbreviations 227

10.8 Arithmetic Expressions 228

10.8.1 Numeric Coercions 229

10.8.2 Casting Numeric Values 232

10.8.3 ~ Operator 234

10.9 Method Calls 235

10.9.1 Virtual Method Calls 235

10.9.2 Static Method Calls 236

10.9.3 Example of Method Calls 236

10.9.4 Overriding 238

10.9.5 Overloading 239

10.9.6 Interfaces 241

10.10 Expressions as Statements 242

10.11 Ifs and Booleans 243

10.11.1 Boolean Values 243

10.11.2 if Statements 245

10.11.3 Comparisons 245

10.11.4 Floating-Point Comparisons and NaN 247

10.11.5 Integer Comparisons 249

10.11.6 References and Object Equality 249

10.11.7 Boolean Operators 254

10.12 Other Control Structures 255

10.12.1 Loops 256

10.12.2 Compiling Control Constructs 257

10.12.3 Compound Statements and Verification 259

10.13 Returns 261

10.14 Field Declarations 262

10.15 Method Declarations 263

10.16 Constructors 264

10.16.1 Using Other Constructors in the Same Class 266

10.16.2 Default Constructors 267

10.16.3 Field Initializers and the Constructor 267

10.17 Conclusion 268

11 Compiling Other Languages 269

11.1 Java Version 1.1 270

11.1.1 Class Literals 271

11.1.2 Inner Classes 272

11.1.3 Variable Scoping with Inner Classes 275

11.2 Regular Expressions 277

11.3 Iterators 283

11.3.1 Iterators in Sather 283

11.3.2 Implementing Iterators 285

11.3.3 Iterators with Arguments 288

11.3.4 Using Iterators 290

11.4 Parameterized Types 292

11.4.1 Implementing Parameterized Types with a Generic Class 293

11.4.2 Implementing Parameterized Types as Templates 294

11.5 Multiple Inheritance 295

11.5.1 Interfaces for Multiple Inheritance 296

11.5.2 Implementing the Interfaces 297

11.5.3 Fields 302

11.6 Conclusion 303

12 Implementing Scheme 305

12.1 Scheme Concepts 305

12.2 Compiling Scheme into Bytecodes 309

12.3 Compiling Lambda Expressions 313

12.4 Example 317

12.5 Scheme Library 319

12.6 Optimizing Scheme 320

12.6.1 Static Scoping 321

12.6.2 Using Local Variables 325

12.6.3 Inlining 327

12.6.4 Tail Recursion 327

12.6.5 Using Type Information 328

13 Implementing Prolog 331

13.1 Introduction to Prolog 331

13.1.1 Asking Questions 332

13.1.2 Binding Patterns 333

13.1.3 Implications 333

13.1.4 Binding Patterns and Implications 334

13.1.5 Facts as Rules 335

13.1.6 More About Unification 335

13.2 Implementation Basics 336

13.3 Unification 337

13.3.1 Unification Algorithm 337

13.3.2 Managing Bindings 338

13.4 Rules as Programs 339

13.4.1 inside 340

13.4.2 within 341

13.4.3 yield and Backtracking 341

13.5 Implementing Rules 344

13.5.1 Implementing within 344

13.5.2 Implementing Conjunctions 347

13.5.3 Constructing within 349

13.6 Compiling Facts 349

13.7 Case Study 352

14 Performance 355

14.1 Fast Virtual Machines 355

14.1.1 Garbage Collection Performance 356

14.1.2 Implementing References 358

14.2 Bytecode Optimization Techniques 360

14.2.1 Warning 360

14.2.2 Constant Propagation 360

14.2.3 Strength Reduction 362

14.2.4 Loop Unrolling 363

14.2.5 Peephole Optimization 364

14.3 Inlining 368

14.3.1 Inlining Fields 370

14.3.2 Separate Compilation 372

15 Security and the Virtual Machine 373

15.1 Java platform and Need for Security 373

15.2 Security Promises of the JVM 374

15.3 Security Architecture and Security Policy 375

15.3.1 Example 377

15.3.2 Basic Applet Security Policy 378

15.3.3 More Sophisticated Policies 379

15.4 Some Potential Attacks 380

15.4.1 Implementation Details 380

15.4.2 Protecting the Security Manager 381

15.4.3 Bypassing Java Security 382

15.4.4 Using Unconstructed Objects 383

15.4.5 Invalid Casts 386

15.4.6 Changing the Class of a Reference 388

15.4.7 Reading Uninitialized Fields 388

15.4.8 Array Bounds Checks 390

15.4.9 Catching Exceptions 391

15.4.10 Hidden Code 392

15.5 Conclusion 393

16 Threads and Synchronization 395

16.1 Threads 396

16.1.1 Example: Multithreaded Web Server 398

16.1.2 Exceptions and Threads 401

16.2 Sharing State Between Threads 401

16.2.1 Need for Synchronization 402

16.2.2 Exception: longs and doubles 404

16.3 Monitors and Object Locks 404

16.3.1 Synchronization in Java 407

16.3.2 Synchronized Methods 408

16.4 Scheduling 410

16.4.1 Surrendering Control of the CPU 412

16.4.2 Waiting for Another Thread 413

16.5 Conclusion 414

A Tables 415

A.1 Instructions by Opcode 415

A.2 Instructions Alphabetically 418

A.3 Instructions by Category 421

A.3.1 Arithmetic 421

A.3.2 Array 422

A.3.3 Constant 423

A.3.4 Control 424

A.3.5 Data Type Conversion 425

A.3.6 Object 426

A.3.7 Stack Manipulation 426

A.3.8 Synchronization 426

A.3.9 Variable 427

B Oolong Reference 429

B.1 Using the Oolong Assembler 429

B.2 Gnoloo 430

B.3 DumpClass 431

B.4 Oolong Language 432

B.4.1 Comments 432

B.4.2 .source 432

B.4.3 .class 433

B.4.4 .interface 434

B.4.5 .end class 434

B.4.6 .super 435

B.4.7 .implements 435

B.4.8 .field 436

B.4.9 .method 436

B.4.10 .limit locals 438

B.4.11 .limit stack 438

B.4.12 .line 439

B.4.13 .var 440

B.4.14 .throws 441

B.4.15 .catch 441

B.4.16 .end method 443

B.5 Instructions 443

B.5.1 Instructions with No Arguments 449

B.5.2 Integers 449

B.5.3 Labels 450

B.5.4 Classes 450

B.5.5 Fields and Methods 450

B.5.6 invokeinterface 450

B.5.7 multianewarray 451

B.5.8 ldc, ldc_w, and ldc2_w 451

B.5.9 newarray 452

B.5.10 tableswitch 452

B.5.11 lookupswitch 453

C Answers to Selected Exercises 455

Further Readings 473

Java and the JVM 473

Newsgroup 474

Other Languages 474

Compilers 474

Index 477