blob: 06d0df1bde9ac2f65bb1a929c7c1c9c5b62f8b90 [file] [log] [blame]
Owen Andersonc04ee692007-10-23 06:03:24 +00001<!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>LLVM Tutorial 2: A More Complicated Function</title>
7 <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
8 <meta name="author" content="Owen Anderson">
9 <meta name="description"
10 content="LLVM Tutorial 2: A More Complicated Function.">
11 <link rel="stylesheet" href="../llvm.css" type="text/css">
12</head>
13
14<body>
15
16<div class="doc_title"> LLVM Tutorial 2: A More Complicated Function </div>
17
18<div class="doc_author">
19 <p>Written by <a href="mailto:owen@apple.com">Owen Anderson</a></p>
20</div>
21
22<!-- *********************************************************************** -->
Owen Andersonc04ee692007-10-23 06:03:24 +000023<div class="doc_section"><a name="intro">A First Function</a></div>
24<!-- *********************************************************************** -->
25
26<div class="doc_text">
27
28<p>Now that we understand the basics of creating functions in LLVM, let's move on to a more complicated example: something with control flow. As an example, let's consider Euclid's Greatest Common Denominator (GCD) algorithm:</p>
29
30<div class="doc_code">
31<pre>
32unsigned gcd(unsigned x, unsigned y) {
33 if(x == y) {
34 return x;
35 } else if(x < y) {
36 return gcd(x, y - x);
37 } else {
38 return gcd(x - y, y);
39 }
40}
41</pre>
42</div>
43
44<p>With this example, we'll learn how to create functions with multiple blocks and control flow, and how to make function calls within your LLVM code. For starters, consider the diagram below.</p>
45
Owen Andersonbad82d82007-10-23 06:22:21 +000046<div style="text-align: center;"><img src="JITTutorial2-1.png" alt="GCD CFG" width="60%"></div>
Owen Andersonc04ee692007-10-23 06:03:24 +000047
Owen Andersonbad82d82007-10-23 06:22:21 +000048<p>The above is a graphical representation of a program in LLVM IR. It places each basic block on a node of a graph, and uses directed edges to indicate flow control. These blocks will be serialized when written to a text or bitcode file, but it is often useful conceptually to think of them as a graph. Again, if you are unsure about the code in the diagram, you should skim through the <a href="../LangRef.html">LLVM Language Reference Manual</a> and convince yourself that it is, in fact, the GCD algorithm.</p>
Owen Andersonc04ee692007-10-23 06:03:24 +000049
Duncan Sandse0a34352007-11-05 15:15:50 +000050<p>The first part of our code is the same as from first tutorial. The same basic setup is required: creating a module, verifying it, and running the <code>PrintModulePass</code> on it. Even the first segment of <code>makeLLVMModule()</code> looks the same, because <code>gcd</code> happens to have the same prototype as our <code>mul_add</code> function.</p>
Owen Andersonc04ee692007-10-23 06:03:24 +000051
52<div class="doc_code">
53<pre>
54#include &lt;llvm/Module.h&gt;
55#include &lt;llvm/Function.h&gt;
56#include &lt;llvm/PassManager.h&gt;
57#include &lt;llvm/Analysis/Verifier.h&gt;
58#include &lt;llvm/Assembly/PrintModulePass.h&gt;
59#include &lt;llvm/Support/LLVMBuilder.h&gt;
60
61using namespace llvm;
62
63Module* makeLLVMModule();
64
65int main(int argc, char**argv) {
66 Module* Mod = makeLLVMModule();
67
68 verifyModule(*Mod, PrintMessageAction);
69
70 PassManager PM;
Owen Andersonbad82d82007-10-23 06:22:21 +000071 PM.add(new PrintModulePass(&amp;llvm::cout));
Owen Andersonc04ee692007-10-23 06:03:24 +000072 PM.run(*Mod);
73
74 return 0;
75}
76
77Module* makeLLVMModule() {
Owen Andersonbad82d82007-10-23 06:22:21 +000078 Module* mod = new Module(&quot;tut2&quot;);
Owen Andersonc04ee692007-10-23 06:03:24 +000079
Owen Andersonbad82d82007-10-23 06:22:21 +000080 Constant* c = mod-&gt;getOrInsertFunction(&quot;gcd&quot;,
Owen Andersonc04ee692007-10-23 06:03:24 +000081 IntegerType::get(32),
82 IntegerType::get(32),
83 IntegerType::get(32),
84 NULL);
Owen Andersonbad82d82007-10-23 06:22:21 +000085 Function* gcd = cast&lt;Function&gt;(c);
Owen Andersonc04ee692007-10-23 06:03:24 +000086
Owen Andersonbad82d82007-10-23 06:22:21 +000087 Function::arg_iterator args = gcd-&gt;arg_begin();
Owen Andersonc04ee692007-10-23 06:03:24 +000088 Value* x = args++;
Owen Andersonbad82d82007-10-23 06:22:21 +000089 x-&gt;setName(&quot;x&quot;);
Owen Andersonc04ee692007-10-23 06:03:24 +000090 Value* y = args++;
Owen Andersonbad82d82007-10-23 06:22:21 +000091 y-&gt;setName(&quot;y&quot;);
Owen Andersonc04ee692007-10-23 06:03:24 +000092</pre>
93</div>
94
95<p>Here, however, is where our code begins to diverge from the first tutorial. Because <code>gcd</code> has control flow, it is composed of multiple blocks interconnected by branching (<code>br</code>) instructions. For those familiar with assembly language, a block is similar to a labeled set of instructions. For those not familiar with assembly language, a block is basically a set of instructions that can be branched to and is executed linearly until the block is terminated by one of a small number of control flow instructions, such as <code>br</code> or <code>ret</code>.</p>
96
97<p>Blocks corresponds to the nodes in the diagram we looked at in the beginning of this tutorial. From the diagram, we can see that this function contains five blocks, so we'll go ahead and create them. Note that, in this code sample, we're making use of LLVM's automatic name uniquing, since we're giving two blocks the same name.</p>
98
99<div class="doc_code">
100<pre>
Owen Andersonbad82d82007-10-23 06:22:21 +0000101 BasicBlock* entry = new BasicBlock(&quot;entry&quot;, gcd);
102 BasicBlock* ret = new BasicBlock(&quot;return&quot;, gcd);
103 BasicBlock* cond_false = new BasicBlock(&quot;cond_false&quot;, gcd);
104 BasicBlock* cond_true = new BasicBlock(&quot;cond_true&quot;, gcd);
105 BasicBlock* cond_false_2 = new BasicBlock(&quot;cond_false&quot;, gcd);
Owen Andersonc04ee692007-10-23 06:03:24 +0000106</pre>
107</div>
108
109<p>Now, we're ready to begin generate code! We'll start with the <code>entry</code> block. This block corresponds to the top-level if-statement in the original C code, so we need to compare <code>x == y</code> To achieve this, we perform an explicity comparison using <code>ICmpEQ</code>. <code>ICmpEQ</code> stands for an <em>integer comparison for equality</em> and returns a 1-bit integer result. This 1-bit result is then used as the input to a conditional branch, with <code>ret</code> as the <code>true</code> and <code>cond_false</code> as the <code>false</code> case.</p>
110
111<div class="doc_code">
112<pre>
113 LLVMBuilder builder(entry);
Owen Andersonbad82d82007-10-23 06:22:21 +0000114 Value* xEqualsY = builder.CreateICmpEQ(x, y, &quot;tmp&quot;);
Owen Andersonc04ee692007-10-23 06:03:24 +0000115 builder.CreateCondBr(xEqualsY, ret, cond_false);
116</pre>
117</div>
118
119<p>Our next block, <code>ret</code>, is pretty simple: it just returns the value of <code>x</code>. Recall that this block is only reached if <code>x == y</code>, so this is the correct behavior. Notice that, instead of creating a new <code>LLVMBuilder</code> for each block, we can use <code>SetInsertPoint</code> to retarget our existing one. This saves on construction and memory allocation costs.</p>
120
121<div class="doc_code">
122<pre>
123 builder.SetInsertPoint(ret);
124 builder.CreateRet(x);
125</pre>
126</div>
127
128<p><code>cond_false</code> is a more interesting block: we now know that <code>x != y</code>, so we must branch again to determine which of <code>x</code> and <code>y</code> is larger. This is achieved using the <code>ICmpULT</code> instruction, which stands for <em>integer comparison for unsigned less-than</em>. In LLVM, integer types do not carry sign; a 32-bit integer pseudo-register can interpreted as signed or unsigned without casting. Whether a signed or unsigned interpretation is desired is specified in the instruction. This is why several instructions in the LLVM IR, such as integer less-than, include a specifier for signed or unsigned.</p>
129
130<p>Also, note that we're again making use of LLVM's automatic name uniquing, this time at a register level. We've deliberately chosen to name every instruction "tmp", to illustrate that LLVM will give them all unique names without getting confused.</p>
131
132<div class="doc_code">
133<pre>
134 builder.SetInsertPoint(cond_false);
Owen Andersonbad82d82007-10-23 06:22:21 +0000135 Value* xLessThanY = builder.CreateICmpULT(x, y, &quot;tmp&quot;);
Owen Andersonc04ee692007-10-23 06:03:24 +0000136 builder.CreateCondBr(xLessThanY, cond_true, cond_false_2);
137</pre>
138</div>
139
140<p>Our last two blocks are quite similar; they're both recursive calls to <code>gcd</code> with different parameters. To create a call instruction, we have to create a <code>vector</code> (or any other container with <code>InputInterator</code>s) to hold the arguments. We then pass in the beginning and ending iterators for this vector.</p>
141
142<div class="doc_code">
143<pre>
144 builder.SetInsertPoint(cond_true);
Owen Andersonbad82d82007-10-23 06:22:21 +0000145 Value* yMinusX = builder.CreateSub(y, x, &quot;tmp&quot;);
146 std::vector&lt;Value*&gt; args1;
Owen Andersonc04ee692007-10-23 06:03:24 +0000147 args1.push_back(x);
148 args1.push_back(yMinusX);
Owen Andersonbad82d82007-10-23 06:22:21 +0000149 Value* recur_1 = builder.CreateCall(gcd, args1.begin(), args1.end(), &quot;tmp&quot;);
Owen Andersonc04ee692007-10-23 06:03:24 +0000150 builder.CreateRet(recur_1);
151
152 builder.SetInsertPoint(cond_false_2);
Owen Andersonbad82d82007-10-23 06:22:21 +0000153 Value* xMinusY = builder.CreateSub(x, y, &quot;tmp&quot;);
154 std::vector&lt;Value*&gt; args2;
Owen Andersonc04ee692007-10-23 06:03:24 +0000155 args2.push_back(xMinusY);
156 args2.push_back(y);
Owen Andersonbad82d82007-10-23 06:22:21 +0000157 Value* recur_2 = builder.CreateCall(gcd, args2.begin(), args2.end(), &quot;tmp&quot;);
Owen Andersonc04ee692007-10-23 06:03:24 +0000158 builder.CreateRet(recur_2);
159
160 return mod;
161}
162</pre>
163</div>
164
Duncan Sandse0a34352007-11-05 15:15:50 +0000165<p>And that's it! You can compile and execute your code in the same way as before, by doing:</p>
Owen Andersonc04ee692007-10-23 06:03:24 +0000166
167<div class="doc_code">
168<pre>
Owen Anderson18f09922007-10-25 06:49:29 +0000169# c++ -g tut2.cpp `llvm-config --cppflags --ldflags --libs core` -o tut2
Owen Andersonc04ee692007-10-23 06:03:24 +0000170# ./tut2
171</pre>
172</div>
173
174</div>
175
Owen Anderson34ba67a2007-10-25 06:45:01 +0000176<!-- *********************************************************************** -->
177<hr>
178<address>
179 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
180 src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
181 <a href="http://validator.w3.org/check/referer"><img
182 src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!"></a>
183
184 <a href="mailto:owen@apple.com">Owen Anderson</a><br>
185 <a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br>
186 Last modified: $Date: 2007-10-17 11:05:13 -0700 (Wed, 17 Oct 2007) $
187</address>
188
Owen Andersonc04ee692007-10-23 06:03:24 +0000189</body>
Duncan Sandse0a34352007-11-05 15:15:50 +0000190</html>