| blob | 66843db5d3af0f0e6a45ba72d67fd4be53669549 |
4 <html>
5 <head>
10 </head>
12 <body>
16 <ul>
19 <ol>
23 </ol>
24 </li>
26 </ul>
30 </div>
32 <!-- *********************************************************************** -->
34 <!-- *********************************************************************** -->
39 runs through the implementation of a simple language, showing how fun and
40 easy it can be. This tutorial will get you up and started as well as help to
41 build a framework you can extend to other languages. The code in this tutorial
42 can also be used as a playground to hack on other LLVM specific things.
43 </p>
45 <p>
46 The goal of this tutorial is to progressively unveil our language, describing
47 how it is built up over time. This will let us cover a fairly broad range of
48 language design and LLVM-specific usage issues, showing and explaining the code
49 for it all along the way, without overwhelming you with tons of details up
50 front.</p>
52 <p>It is useful to point out ahead of time that this tutorial is really about
54 modern and sane software engineering principles. In practice, this means that
55 we'll take a number of shortcuts to simplify the exposition. For example, the
56 code leaks memory, uses global variables all over the place, doesn't use nice
57 design patterns like <a
59 is very simple. If you dig in and use the code as a basis for future projects,
60 fixing these deficiencies shouldn't be hard.</p>
62 <p>I've tried to put this tutorial together in a way that makes chapters easy to
63 skip over if you are already familiar with or are uninterested in the various
64 pieces. The structure of the tutorial is:
65 </p>
67 <ul>
69 language, and the definition of its Lexer</b> - This shows where we are going
70 and the basic functionality that we want it to do. In order to make this
71 tutorial maximally understandable and hackable, we choose to implement
72 everything in C++ instead of using lexer and parser generators. LLVM obviously
73 works just fine with such tools, feel free to use one if you prefer.</li>
75 AST</b> - With the lexer in place, we can talk about parsing techniques and
76 basic AST construction. This tutorial describes recursive descent parsing and
78 the code doesn't even link in LLVM at this point. :)</li>
80 With the AST ready, we can show off how easy generation of LLVM IR really
81 is.</li>
83 Support</b> - Because a lot of people are interested in using LLVM as a JIT,
84 we'll dive right into it and show you the 3 lines it takes to add JIT support.
85 LLVM is also useful in many other ways, but this is one simple and "sexy" way
86 to shows off its power. :)</li>
88 Flow</b> - With the language up and running, we show how to extend it with
89 control flow operations (if/then/else and a 'for' loop). This gives us a chance
90 to talk about simple SSA construction and control flow.</li>
92 User-defined Operators</b> - This is a silly but fun chapter that talks about
93 extending the language to let the user program define their own arbitrary
94 unary and binary operators (with assignable precedence!). This lets us build a
97 Variables</b> - This chapter talks about adding user-defined local variables
98 along with an assignment operator. The interesting part about this is how
100 require your front-end to construct SSA form!</li>
102 tidbits</b> - This chapter wraps up the series by talking about potential
103 ways to extend the language, but also includes a bunch of pointers to info about
104 "special topics" like adding garbage collection support, exceptions, debugging,
107 </ul>
110 non-comment, non-blank, lines of code. With this small amount of code, we'll
111 have built up a very reasonable compiler for a non-trivial language including
112 a hand-written lexer, parser, AST, as well as code generation support with a JIT
113 compiler. While other systems may have interesting "hello world" tutorials,
114 I think the breadth of this tutorial is a great testament to the strengths of
115 LLVM and why you should consider it if you're interested in language or compiler
116 design.</p>
118 <p>A note about this tutorial: we expect you to extend the language and play
119 with it on your own. Take the code and go crazy hacking away at it, compilers
120 don't need to be scary creatures - it can be a lot of fun to play with
121 languages!</p>
123 </div>
125 <!-- *********************************************************************** -->
127 <!-- *********************************************************************** -->
131 <p>This tutorial will be illustrated with a toy language that we'll call
133 from "meaning beautiful, form, and view").
134 Kaleidoscope is a procedural language that allows you to define functions, use
135 conditionals, math, etc. Over the course of the tutorial, we'll extend
136 Kaleidoscope to support the if/then/else construct, a for loop, user defined
137 operators, JIT compilation with a simple command line interface, etc.</p>
139 <p>Because we want to keep things simple, the only datatype in Kaleidoscope is a
140 64-bit floating point type (aka 'double' in C parlance). As such, all values
141 are implicitly double precision and the language doesn't require type
142 declarations. This gives the language a very nice and simple syntax. For
143 example, the following simple example computes <a
147 <pre>
148 # Compute the x'th fibonacci number.
149 def fib(x)
151 1
152 else
155 # This expression will compute the 40th number.
156 fib(40)
157 </pre>
158 </div>
160 <p>We also allow Kaleidoscope to call into standard library functions (the LLVM
161 JIT makes this completely trivial). This means that you can use the 'extern'
162 keyword to define a function before you use it (this is also useful for mutually
163 recursive functions). For example:</p>
166 <pre>
167 extern sin(arg);
168 extern cos(arg);
169 extern atan2(arg1 arg2);
172 </pre>
173 </div>
181 </div>
183 <!-- *********************************************************************** -->
185 <!-- *********************************************************************** -->
189 <p>When it comes to implementing a language, the first thing needed is
190 the ability to process a text file and recognize what it says. The traditional
191 way to do this is to use a "<a
193 to break the input up into "tokens". Each token returned by the lexer includes
194 a token code and potentially some metadata (e.g. the numeric value of a number).
195 First, we define the possibilities:
196 </p>
199 <pre>
201 // of these for known things.
202 enum Token {
203 tok_eof = -1,
205 // commands
208 // primary
210 };
212 static std::string IdentifierStr; // Filled in if tok_identifier
213 static double NumVal; // Filled in if tok_number
214 </pre>
215 </div>
217 <p>Each token returned by our lexer will either be one of the Token enum values
218 or it will be an 'unknown' character like '+', which is returned as its ASCII
220 global variable holds the name of the identifier. If the current token is a
222 global variables for simplicity, this is not the best choice for a real language
223 implementation :).
224 </p>
226 <p>The actual implementation of the lexer is a single function named
228 from standard input. Its definition starts as:</p>
231 <pre>
232 /// gettok - Return the next token from standard input.
233 static int gettok() {
234 static int LastChar = ' ';
236 // Skip any whitespace.
237 while (isspace(LastChar))
238 LastChar = getchar();
239 </pre>
240 </div>
242 <p>
244 characters one at a time from standard input. It eats them as it recognizes
245 them and stores the last character read, but not processed, in LastChar. The
246 first thing that it has to do is ignore whitespace between tokens. This is
247 accomplished with the loop above.</p>
253 <pre>
254 if (isalpha(LastChar)) { // identifier: [a-zA-Z][a-zA-Z0-9]*
255 IdentifierStr = LastChar;
256 while (isalnum((LastChar = getchar())))
257 IdentifierStr += LastChar;
259 if (IdentifierStr == "def") return tok_def;
260 if (IdentifierStr == "extern") return tok_extern;
261 return tok_identifier;
262 }
263 </pre>
264 </div>
267 lexes an identifier. Also, since language keywords are matched by the same
268 loop, we handle them here inline. Numeric values are similar:</p>
271 <pre>
273 std::string NumStr;
274 do {
275 NumStr += LastChar;
276 LastChar = getchar();
277 } while (isdigit(LastChar) || LastChar == '.');
279 NumVal = strtod(NumStr.c_str(), 0);
280 return tok_number;
281 }
282 </pre>
283 </div>
285 <p>This is all pretty straight-forward code for processing input. When reading
288 sufficient error checking: it will incorrectly read "1.23.45.67" and handle it as
289 if you typed in "1.23". Feel free to extend it :). Next we handle comments:
290 </p>
293 <pre>
294 if (LastChar == '#') {
295 // Comment until end of line.
296 do LastChar = getchar();
299 if (LastChar != EOF)
300 return gettok();
301 }
302 </pre>
303 </div>
305 <p>We handle comments by skipping to the end of the line and then return the
306 next token. Finally, if the input doesn't match one of the above cases, it is
307 either an operator character like '+' or the end of the file. These are handled
308 with this code:</p>
311 <pre>
312 // Check for end of file. Don't eat the EOF.
313 if (LastChar == EOF)
314 return tok_eof;
316 // Otherwise, just return the character as its ascii value.
317 int ThisChar = LastChar;
318 LastChar = getchar();
319 return ThisChar;
320 }
321 </pre>
322 </div>
324 <p>With this, we have the complete lexer for the basic Kaleidoscope language
328 build an Abstract Syntax Tree</a>. When we have that, we'll include a driver
329 so that you can use the lexer and parser together.
330 </p>
333 </div>
335 <!-- *********************************************************************** -->
336 <hr>
337 <address>
345 Last modified: $Date$
346 </address>
347 </body>
348 </html>