Gitiles
Code Review Sign In
gerrit-public.fairphone.software / fp2-dev / platform / external / llvm / 99005a4b3b7d4a76e05bf81e8c42c5c39f206f8a / . / docs / tutorial / LangImpl8.html
blob: 9a1a8a28ced0f96cdc530918d12f5be89de111d4 [file] [log] [blame]
Chris Lattnerb8fc6502007-11-05 01:58:13 +0000[diff] [blame]1<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
2 "http://www.w3.org/TR/html4/strict.dtd">
3
4<html>
5<head>
Chris Lattner99005a42007-11-05 19:10:15 +0000[diff] [blame^]6 <title>Kaleidoscope: Conclusion and other useful LLVM tidbits</title>
Chris Lattnerb8fc6502007-11-05 01:58:13 +0000[diff] [blame]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
Chris Lattner99005a42007-11-05 19:10:15 +0000[diff] [blame^]14<div class="doc_title">Kaleidoscope: Conclusion and other useful LLVM
15 tidbits</div>
16
17<ul>
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
Chris Lattnerb8fc6502007-11-05 01:58:13 +0000[diff] [blame]39
40<div class="doc_author">
41 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a></p>
42</div>
43
44<!-- *********************************************************************** -->
Chris Lattner99005a42007-11-05 19:10:15 +0000[diff] [blame^]45<div class="doc_section"><a name="conclusion">Tutorial Conclusion</a></div>
Chris Lattnerb8fc6502007-11-05 01:58:13 +0000[diff] [blame]46<!-- *********************************************************************** -->
47
48<div class="doc_text">
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.
91The getelementptr instruction is so nifty/unconventional, it <a
92href="../GetElementPtr.html">has its own FAQ</a>!).</li>
93
94<li><b>standard runtime</b> - Our current language allows the user to access
95arbitrary external functions, and we use it for things like "printd" and
96"putchard". As you extend the language to add higher-level constructs, often
97these constructs make the most amount of sense to be lowered into calls into a
98language-supplied runtime. For example, if you add hash tables to the language,
99it would probably make sense to add the routines to a runtime, instead of
100inlining them all the way.</li>
101
102<li><b>memory management</b> - Currently we can only access the stack in
103Kaleidoscope. It would also be useful to be able to allocate heap memory,
104either with calls to the standard libc malloc/free interface or with a garbage
105collector. If you choose to use garbage collection, note that LLVM fully
106supports <a href="../GarbageCollection.html">Accurate Garbage Collection</a>
107including algorithms that move objects and need to scan/update the stack.</li>
108
109<li><b>debugger support</b> - LLVM supports generation of <a
110href="../SourceLevelDebugging.html">DWARF Debug info</a> which is understood by
111common debuggers like GDB. Adding support for debug info is fairly
112straight-forward. The best way to understand it is to compile some C/C++ code
113with "<tt>llvm-gcc -g -O0</tt>" and taking a look at what it produces.</li>
114
Chris Lattnera3f07ef2007-11-05 07:00:54 +0000[diff] [blame]115<li><b>exception handling support</b> - LLVM supports generation of <a
Chris Lattnerb8fc6502007-11-05 01:58:13 +0000[diff] [blame]116href="../ExceptionHandling.html">zero cost exceptions</a> which interoperate
117with code compiled in other languages. You could also generate code by
118implicitly making every function return an error value and checking it. You
119could also make explicit use of setjmp/longjmp. There are many different ways
120to go here.</li>
121
122<li><b>object orientation, generics, database access, complex numbers,
123geometric programming, ...</b> - Really, there is
124no end of crazy features that you can add to the language.</li>
125
Chris Lattnera3f07ef2007-11-05 07:00:54 +0000[diff] [blame]126<li><b>unusual domains</b> - We've been talking about applying LLVM to a domain
127that many people are interested in: building a compiler for a specific language.
128However, there are many other domains that can use compiler technology that are
129not typically considered. For example, LLVM has been used to implement OpenGL
130graphics acceleration, translate C++ code to ActionScript, and many other
131cute and clever things. Maybe you will be the first to JIT compile a regular
132expression interpreter into native code with LLVM?</li>
133
Chris Lattnerb8fc6502007-11-05 01:58:13 +0000[diff] [blame]134</ul>
135
136<p>
137Have fun - try doing something crazy and unusual. Building a language like
138everyone else always has is much less fun than trying something a little crazy
139and off the wall and seeing how it turns out. If you get stuck or want to talk
140about it, feel free to email the <a
141href="http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev">llvmdev mailing
142list</a>: it has lots of people who are interested in languages and are often
143willing to help out.
144</p>
145
146<p>Before we end, I want to talk about some "tips and tricks" for generating
147LLVM IR. These are some of the more subtle things that may not be obvious, but
148are very useful if you want to take advantage of LLVM's capabilities.</p>
149
150</div>
151
152<!-- *********************************************************************** -->
Chris Lattnera3f07ef2007-11-05 07:00:54 +0000[diff] [blame]153<div class="doc_section"><a name="llvmirproperties">Properties of LLVM
154IR</a></div>
155<!-- *********************************************************************** -->
156
157<div class="doc_text">
158
159<p>We have a couple common questions about code in the LLVM IR form, lets just
160get these out of the way right now shall we?</p>
161
162</div>
163
164<!-- ======================================================================= -->
165<div class="doc_subsubsection"><a name="targetindep">Target
166Independence</a></div>
167<!-- ======================================================================= -->
168
169<div class="doc_text">
170
171<p>Kaleidoscope is an example of a "portable language": any program written in
172Kaleidoscope will work the same way on any target that it runs on. Many other
173languages have this property, e.g. lisp, java, haskell, javascript, python, etc
174(note that while these languages are portable, not all their libraries are).</p>
175
176<p>One nice aspect of LLVM is that it is often capable of preserving language
177independence in the IR: you can take the LLVM IR for a Kaleidoscope-compiled
178program and run it on any target that LLVM supports, even emitting C code and
179compiling that on targets that LLVM doesn't support natively. You can trivially
180tell that the Kaleidoscope compiler generates target-independent code because it
181never queries for any target-specific information when generating code.</p>
182
183<p>The fact that LLVM provides a compact target-independent representation for
184code gets a lot of people excited. Unfortunately, these people are usually
185thinking about C or a language from the C family when they are asking questions
186about language portability. I say "unfortunately", because there is really no
187way to make (fully general) C code portable, other than shipping the source code
188around (and of course, C source code is not actually portable in general
189either - ever port a really old application from 32- to 64-bits?).</p>
190
191<p>The problem with C (again, in its full generality) is that it is heavily
192laden with target specific assumptions. As one simple example, the preprocessor
193often destructively removes target-independence from the code when it processes
194the input text:</p>
195
196<div class="doc_code">
197<pre>
198#ifdef __i386__
199 int X = 1;
200#else
201 int X = 42;
202#endif
203</pre>
204</div>
205
206<p>While it is possible to engineer more and more complex solutions to problems
207like this, it cannot be solved in full generality in a way better than shipping
208the actual source code.</p>
209
210<p>That said, there are interesting subsets of C that can be made portable. If
211you are willing to fix primitive types to a fixed size (say int = 32-bits,
212and long = 64-bits), don't care about ABI compatibility with existing binaries,
213and are willing to give up some other minor features, you can have portable
214code. This can even make real sense for specialized domains such as an
215in-kernel language.</p>
216
217</div>
218
219<!-- ======================================================================= -->
220<div class="doc_subsubsection"><a name="safety">Safety Guarantees</a></div>
221<!-- ======================================================================= -->
222
223<div class="doc_text">
224
225<p>Many of the languages above are also "safe" languages: it is impossible for
226a program written in Java to corrupt its address space and crash the process.
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<div class="doc_subsubsection"><a name="langspecific">Language-Specific
242Optimizations</a></div>
243<!-- ======================================================================= -->
244
245<div class="doc_text">
246
247<p>One thing about LLVM that turns off many people is that it does not solve all
248the world's problems in one system (sorry 'world hunger', someone else will have
249to solve you some other day). One specific complaint is that people perceive
250LLVM as being incapable of performing high-level language-specific optimization:
251LLVM "loses too much information".</p>
252
253<p>Unfortunately, this is really not the place to give you a full and unified
254version of "Chris Lattner's theory of compiler design". Instead, I'll make a
255few observations:</p>
256
257<p>First, you're right that LLVM does lose information. For example, as of this
258writing, there is no way to distinguish in the LLVM IR whether an SSA-value came
259from a C "int" or a C "long" on an ILP32 machine (other than debug info). Both
260get compiled down to an 'i32' value and the information about what it came from
261is lost. The more general issue here is that the LLVM type system uses
262"structural equivalence" instead of "name equivalence". Another place this
263surprises people is if you have two types in a high-level language that have the
264same structure (e.g. two different structs that have a single int field): these
265types will compile down into a single LLVM type and it will be impossible to
266tell what it came from.</p>
267
268<p>Second, while LLVM does lose information, LLVM is not a fixed target: we
269continue to enhance and improve it in many different ways. In addition to
270adding new features (LLVM did not always support exceptions or debug info), we
271also extend the IR to capture important information for optimization (e.g.
272whether an argument is sign or zero extended, information about pointers
273aliasing, etc. Many of the enhancements are user-driven: people want LLVM to
274do some specific feature, so they go ahead and extend it to do so.</p>
275
276<p>Third, it <em>is certainly possible</em> to add language-specific
277optimizations, and you have a number of choices in how to do it. As one trivial
278example, it is possible to add language-specific optimization passes that
279"known" things about code compiled for a language. In the case of the C family,
280there is an optimziation pass that "knows" about the standard C library
281functions. If you call "exit(0)" in main(), it knows that it is safe to
282optimize that into "return 0;" for example, because C specifies what the 'exit'
283function does.</p>
284
285<p>In addition to simple library knowledge, it is possible to embed a variety of
286other language-specific information into the LLVM IR. If you have a specific
287need and run into a wall, please bring the topic up on the llvmdev list. At the
288very worst, you can always treat LLVM as if it were a "dumb code generator" and
289implement the high-level optimizations you desire in your front-end on the
290language-specific AST.
291</p>
292
293</div>
294
295<!-- *********************************************************************** -->
Chris Lattnerb8fc6502007-11-05 01:58:13 +0000[diff] [blame]296<div class="doc_section"><a name="tipsandtricks">Tips and Tricks</a></div>
297<!-- *********************************************************************** -->
298
299<div class="doc_text">
300
Chris Lattnera3f07ef2007-11-05 07:00:54 +0000[diff] [blame]301<p>There is a variety of useful tips and tricks that you come to know after
302working on/with LLVM that aren't obvious at first glance. Instead of letting
303everyone rediscover them, this section talks about some of these issues.</p>
304
305</div>
306
307<!-- ======================================================================= -->
308<div class="doc_subsubsection"><a name="offsetofsizeof">Implementing portable
309offsetof/sizeof</a></div>
310<!-- ======================================================================= -->
311
312<div class="doc_text">
313
314<p>One interesting thing that comes up if you are trying to keep the code
315generated by your compiler "target independent" is that you often need to know
316the size of some LLVM type or the offset of some field in an llvm structure.
317For example, you might need to pass the size of a type into a function that
318allocates memory.</p>
319
320<p>Unfortunately, this can vary widely across targets: for example the width of
321a pointer is trivially target-specific. However, there is a <a
322href="http://nondot.org/sabre/LLVMNotes/SizeOf-OffsetOf-VariableSizedStructs.txt">clever
323way to use the getelementptr instruction</a> that allows you to compute this
324in a portable way.</p>
325
326</div>
327
328<!-- ======================================================================= -->
329<div class="doc_subsubsection"><a name="gcstack">Garbage Collected
330Stack Frames</a></div>
331<!-- ======================================================================= -->
332
333<div class="doc_text">
334
335<p>Some languages want to explicitly manage their stack frames, often so that
336they are garbage collected or to allow easy implementation of closures. There
337are often better ways to implement these features than explicit stack frames,
338but <a
339href="http://nondot.org/sabre/LLVMNotes/ExplicitlyManagedStackFrames.txt">LLVM
340does support them if you want</a>. It requires your front-end to convert the
341code into <a
342href="http://en.wikipedia.org/wiki/Continuation-passing_style">Continuation
343Passing Style</a> and use of tail calls (which LLVM also supports).</p>
Chris Lattnerb8fc6502007-11-05 01:58:13 +0000[diff] [blame]344
345</div>
346
347<!-- *********************************************************************** -->
348<hr>
349<address>
350 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
351 src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
352 <a href="http://validator.w3.org/check/referer"><img
353 src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!"></a>
354
355 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
356 <a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br>
357 Last modified: $Date: 2007-10-17 11:05:13 -0700 (Wed, 17 Oct 2007) $
358</address>
359</body>
360</html>