• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
2                      "http://www.w3.org/TR/html4/strict.dtd">
3
4<html>
5<head>
6  <title>Kaleidoscope: Conclusion and other useful LLVM tidbits</title>
7  <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
8  <meta name="author" content="Chris Lattner">
9  <link rel="stylesheet" href="../llvm.css" type="text/css">
10</head>
11
12<body>
13
14<h1>Kaleidoscope: Conclusion and other useful LLVM tidbits</h1>
15
16<ul>
17<li><a href="index.html">Up to Tutorial Index</a></li>
18<li>Chapter 8
19  <ol>
20    <li><a href="#conclusion">Tutorial Conclusion</a></li>
21    <li><a href="#llvmirproperties">Properties of LLVM IR</a>
22    <ul>
23      <li><a href="#targetindep">Target Independence</a></li>
24      <li><a href="#safety">Safety Guarantees</a></li>
25      <li><a href="#langspecific">Language-Specific Optimizations</a></li>
26    </ul>
27    </li>
28    <li><a href="#tipsandtricks">Tips and Tricks</a>
29    <ul>
30      <li><a href="#offsetofsizeof">Implementing portable
31                                    offsetof/sizeof</a></li>
32      <li><a href="#gcstack">Garbage Collected Stack Frames</a></li>
33    </ul>
34    </li>
35  </ol>
36</li>
37</ul>
38
39
40<div class="doc_author">
41  <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a></p>
42</div>
43
44<!-- *********************************************************************** -->
45<h2><a name="conclusion">Tutorial Conclusion</a></h2>
46<!-- *********************************************************************** -->
47
48<div>
49
50<p>Welcome to the the final chapter of the "<a href="index.html">Implementing a
51language with LLVM</a>" tutorial.  In the course of this tutorial, we have grown
52our little Kaleidoscope language from being a useless toy, to being a
53semi-interesting (but probably still useless) toy. :)</p>
54
55<p>It is interesting to see how far we've come, and how little code it has
56taken.  We built the entire lexer, parser, AST, code generator, and an
57interactive run-loop (with a JIT!) by-hand in under 700 lines of
58(non-comment/non-blank) code.</p>
59
60<p>Our little language supports a couple of interesting features: it supports
61user defined binary and unary operators, it uses JIT compilation for immediate
62evaluation, and it supports a few control flow constructs with SSA construction.
63</p>
64
65<p>Part of the idea of this tutorial was to show you how easy and fun it can be
66to define, build, and play with languages.  Building a compiler need not be a
67scary or mystical process!  Now that you've seen some of the basics, I strongly
68encourage you to take the code and hack on it.  For example, try adding:</p>
69
70<ul>
71<li><b>global variables</b> - While global variables have questional value in
72modern software engineering, they are often useful when putting together quick
73little hacks like the Kaleidoscope compiler itself.  Fortunately, our current
74setup makes it very easy to add global variables: just have value lookup check
75to see if an unresolved variable is in the global variable symbol table before
76rejecting it.  To create a new global variable, make an instance of the LLVM
77<tt>GlobalVariable</tt> class.</li>
78
79<li><b>typed variables</b> - Kaleidoscope currently only supports variables of
80type double.  This gives the language a very nice elegance, because only
81supporting one type means that you never have to specify types.  Different
82languages have different ways of handling this.  The easiest way is to require
83the user to specify types for every variable definition, and record the type
84of the variable in the symbol table along with its Value*.</li>
85
86<li><b>arrays, structs, vectors, etc</b> - Once you add types, you can start
87extending the type system in all sorts of interesting ways.  Simple arrays are
88very easy and are quite useful for many different applications.  Adding them is
89mostly an exercise in learning how the LLVM <a
90href="../LangRef.html#i_getelementptr">getelementptr</a> instruction works: it
91is so nifty/unconventional, it <a
92href="../GetElementPtr.html">has its own FAQ</a>!  If you add support
93for recursive types (e.g. linked lists), make sure to read the <a
94href="../ProgrammersManual.html#TypeResolve">section in the LLVM
95Programmer's Manual</a> that describes how to construct them.</li>
96
97<li><b>standard runtime</b> - Our current language allows the user to access
98arbitrary external functions, and we use it for things like "printd" and
99"putchard".  As you extend the language to add higher-level constructs, often
100these constructs make the most sense if they are lowered to calls into a
101language-supplied runtime.  For example, if you add hash tables to the language,
102it would probably make sense to add the routines to a runtime, instead of
103inlining them all the way.</li>
104
105<li><b>memory management</b> - Currently we can only access the stack in
106Kaleidoscope.  It would also be useful to be able to allocate heap memory,
107either with calls to the standard libc malloc/free interface or with a garbage
108collector.  If you would like to use garbage collection, note that LLVM fully
109supports <a href="../GarbageCollection.html">Accurate Garbage Collection</a>
110including algorithms that move objects and need to scan/update the stack.</li>
111
112<li><b>debugger support</b> - LLVM supports generation of <a
113href="../SourceLevelDebugging.html">DWARF Debug info</a> which is understood by
114common debuggers like GDB.  Adding support for debug info is fairly
115straightforward.  The best way to understand it is to compile some C/C++ code
116with "<tt>llvm-gcc -g -O0</tt>" and taking a look at what it produces.</li>
117
118<li><b>exception handling support</b> - LLVM supports generation of <a
119href="../ExceptionHandling.html">zero cost exceptions</a> which interoperate
120with code compiled in other languages.  You could also generate code by
121implicitly making every function return an error value and checking it.  You
122could also make explicit use of setjmp/longjmp.  There are many different ways
123to go here.</li>
124
125<li><b>object orientation, generics, database access, complex numbers,
126geometric programming, ...</b> - Really, there is
127no end of crazy features that you can add to the language.</li>
128
129<li><b>unusual domains</b> - We've been talking about applying LLVM to a domain
130that many people are interested in: building a compiler for a specific language.
131However, there are many other domains that can use compiler technology that are
132not typically considered.  For example, LLVM has been used to implement OpenGL
133graphics acceleration, translate C++ code to ActionScript, and many other
134cute and clever things.  Maybe you will be the first to JIT compile a regular
135expression interpreter into native code with LLVM?</li>
136
137</ul>
138
139<p>
140Have fun - try doing something crazy and unusual.  Building a language like
141everyone else always has, is much less fun than trying something a little crazy
142or off the wall and seeing how it turns out.  If you get stuck or want to talk
143about it, feel free to email the <a
144href="http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev">llvmdev mailing
145list</a>: it has lots of people who are interested in languages and are often
146willing to help out.
147</p>
148
149<p>Before we end this tutorial, I want to talk about some "tips and tricks" for generating
150LLVM IR.  These are some of the more subtle things that may not be obvious, but
151are very useful if you want to take advantage of LLVM's capabilities.</p>
152
153</div>
154
155<!-- *********************************************************************** -->
156<h2><a name="llvmirproperties">Properties of the LLVM IR</a></h2>
157<!-- *********************************************************************** -->
158
159<div>
160
161<p>We have a couple common questions about code in the LLVM IR form - lets just
162get these out of the way right now, shall we?</p>
163
164<!-- ======================================================================= -->
165<h4><a name="targetindep">Target Independence</a></h4>
166<!-- ======================================================================= -->
167
168<div>
169
170<p>Kaleidoscope is an example of a "portable language": any program written in
171Kaleidoscope will work the same way on any target that it runs on.  Many other
172languages have this property, e.g. lisp, java, haskell, javascript, python, etc
173(note that while these languages are portable, not all their libraries are).</p>
174
175<p>One nice aspect of LLVM is that it is often capable of preserving target
176independence in the IR: you can take the LLVM IR for a Kaleidoscope-compiled
177program and run it on any target that LLVM supports, even emitting C code and
178compiling that on targets that LLVM doesn't support natively.  You can trivially
179tell that the Kaleidoscope compiler generates target-independent code because it
180never queries for any target-specific information when generating code.</p>
181
182<p>The fact that LLVM provides a compact, target-independent, representation for
183code gets a lot of people excited.  Unfortunately, these people are usually
184thinking about C or a language from the C family when they are asking questions
185about language portability.  I say "unfortunately", because there is really no
186way to make (fully general) C code portable, other than shipping the source code
187around (and of course, C source code is not actually portable in general
188either - ever port a really old application from 32- to 64-bits?).</p>
189
190<p>The problem with C (again, in its full generality) is that it is heavily
191laden with target specific assumptions.  As one simple example, the preprocessor
192often destructively removes target-independence from the code when it processes
193the input text:</p>
194
195<div class="doc_code">
196<pre>
197#ifdef __i386__
198  int X = 1;
199#else
200  int X = 42;
201#endif
202</pre>
203</div>
204
205<p>While it is possible to engineer more and more complex solutions to problems
206like this, it cannot be solved in full generality in a way that is better than shipping
207the actual source code.</p>
208
209<p>That said, there are interesting subsets of C that can be made portable.  If
210you are willing to fix primitive types to a fixed size (say int = 32-bits,
211and long = 64-bits), don't care about ABI compatibility with existing binaries,
212and are willing to give up some other minor features, you can have portable
213code.  This can make sense for specialized domains such as an
214in-kernel language.</p>
215
216</div>
217
218<!-- ======================================================================= -->
219<h4><a name="safety">Safety Guarantees</a></h4>
220<!-- ======================================================================= -->
221
222<div>
223
224<p>Many of the languages above are also "safe" languages: it is impossible for
225a program written in Java to corrupt its address space and crash the process
226(assuming the JVM has no bugs).
227Safety is an interesting property that requires a combination of language
228design, runtime support, and often operating system support.</p>
229
230<p>It is certainly possible to implement a safe language in LLVM, but LLVM IR
231does not itself guarantee safety.  The LLVM IR allows unsafe pointer casts,
232use after free bugs, buffer over-runs, and a variety of other problems.  Safety
233needs to be implemented as a layer on top of LLVM and, conveniently, several
234groups have investigated this.  Ask on the <a
235href="http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev">llvmdev mailing
236list</a> if you are interested in more details.</p>
237
238</div>
239
240<!-- ======================================================================= -->
241<h4><a name="langspecific">Language-Specific Optimizations</a></h4>
242<!-- ======================================================================= -->
243
244<div>
245
246<p>One thing about LLVM that turns off many people is that it does not solve all
247the world's problems in one system (sorry 'world hunger', someone else will have
248to solve you some other day).  One specific complaint is that people perceive
249LLVM as being incapable of performing high-level language-specific optimization:
250LLVM "loses too much information".</p>
251
252<p>Unfortunately, this is really not the place to give you a full and unified
253version of "Chris Lattner's theory of compiler design".  Instead, I'll make a
254few observations:</p>
255
256<p>First, you're right that LLVM does lose information.  For example, as of this
257writing, there is no way to distinguish in the LLVM IR whether an SSA-value came
258from a C "int" or a C "long" on an ILP32 machine (other than debug info).  Both
259get compiled down to an 'i32' value and the information about what it came from
260is lost.  The more general issue here, is that the LLVM type system uses
261"structural equivalence" instead of "name equivalence".  Another place this
262surprises people is if you have two types in a high-level language that have the
263same structure (e.g. two different structs that have a single int field): these
264types will compile down into a single LLVM type and it will be impossible to
265tell what it came from.</p>
266
267<p>Second, while LLVM does lose information, LLVM is not a fixed target: we
268continue to enhance and improve it in many different ways.  In addition to
269adding new features (LLVM did not always support exceptions or debug info), we
270also extend the IR to capture important information for optimization (e.g.
271whether an argument is sign or zero extended, information about pointers
272aliasing, etc).  Many of the enhancements are user-driven: people want LLVM to
273include some specific feature, so they go ahead and extend it.</p>
274
275<p>Third, it is <em>possible and easy</em> to add language-specific
276optimizations, and you have a number of choices in how to do it.  As one trivial
277example, it is easy to add language-specific optimization passes that
278"know" things about code compiled for a language.  In the case of the C family,
279there is an optimization pass that "knows" about the standard C library
280functions.  If you call "exit(0)" in main(), it knows that it is safe to
281optimize that into "return 0;" because C specifies what the 'exit'
282function does.</p>
283
284<p>In addition to simple library knowledge, it is possible to embed a variety of
285other language-specific information into the LLVM IR.  If you have a specific
286need and run into a wall, please bring the topic up on the llvmdev list.  At the
287very worst, you can always treat LLVM as if it were a "dumb code generator" and
288implement the high-level optimizations you desire in your front-end, on the
289language-specific AST.
290</p>
291
292</div>
293
294</div>
295
296<!-- *********************************************************************** -->
297<h2><a name="tipsandtricks">Tips and Tricks</a></h2>
298<!-- *********************************************************************** -->
299
300<div>
301
302<p>There is a variety of useful tips and tricks that you come to know after
303working on/with LLVM that aren't obvious at first glance.  Instead of letting
304everyone rediscover them, this section talks about some of these issues.</p>
305
306<!-- ======================================================================= -->
307<h4><a name="offsetofsizeof">Implementing portable offsetof/sizeof</a></h4>
308<!-- ======================================================================= -->
309
310<div>
311
312<p>One interesting thing that comes up, if you are trying to keep the code
313generated by your compiler "target independent", is that you often need to know
314the size of some LLVM type or the offset of some field in an llvm structure.
315For example, you might need to pass the size of a type into a function that
316allocates memory.</p>
317
318<p>Unfortunately, this can vary widely across targets: for example the width of
319a pointer is trivially target-specific.  However, there is a <a
320href="http://nondot.org/sabre/LLVMNotes/SizeOf-OffsetOf-VariableSizedStructs.txt">clever
321way to use the getelementptr instruction</a> that allows you to compute this
322in a portable way.</p>
323
324</div>
325
326<!-- ======================================================================= -->
327<h4><a name="gcstack">Garbage Collected Stack Frames</a></h4>
328<!-- ======================================================================= -->
329
330<div>
331
332<p>Some languages want to explicitly manage their stack frames, often so that
333they are garbage collected or to allow easy implementation of closures.  There
334are often better ways to implement these features than explicit stack frames,
335but <a
336href="http://nondot.org/sabre/LLVMNotes/ExplicitlyManagedStackFrames.txt">LLVM
337does support them,</a> if you want.  It requires your front-end to convert the
338code into <a
339href="http://en.wikipedia.org/wiki/Continuation-passing_style">Continuation
340Passing Style</a> and the use of tail calls (which LLVM also supports).</p>
341
342</div>
343
344</div>
345
346<!-- *********************************************************************** -->
347<hr>
348<address>
349  <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
350  src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
351  <a href="http://validator.w3.org/check/referer"><img
352  src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!"></a>
353
354  <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
355  <a href="http://llvm.org/">The LLVM Compiler Infrastructure</a><br>
356  Last modified: $Date$
357</address>
358</body>
359</html>
360