Gitiles
Code Review Sign In
gerrit-public.fairphone.software / fp2-dev / platform / external / llvm / 7123e27894e989f8de24990accdafc63dd061bf3 / . / docs / LangRef.html
blob: 8b05f11f8ea8e400a13524bc1d36ab0b63ffc6f1 [file] [log] [blame]
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
2 "http://www.w3.org/TR/html4/strict.dtd">
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3<html>
4<head>
5 <title>LLVM Assembly Language Reference Manual</title>
6 <link rel="stylesheet" href="llvm.css" type="text/css">
7</head>
8<body>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]9<div class="doc_title"> LLVM Language Reference Manual </div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]10<ol>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]11 <li><a href="#abstract">Abstract</a></li>
12 <li><a href="#introduction">Introduction</a></li>
13 <li><a href="#identifiers">Identifiers</a></li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]14 <li><a href="#typesystem">Type System</a>
15 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]16 <li><a href="#t_primitive">Primitive Types</a>
17 <ol>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]18 <li><a href="#t_classifications">Type Classifications</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]19 </ol>
20 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]21 <li><a href="#t_derived">Derived Types</a>
22 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]23 <li><a href="#t_array">Array Type</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]24 <li><a href="#t_function">Function Type</a></li>
25 <li><a href="#t_pointer">Pointer Type</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]26 <li><a href="#t_struct">Structure Type</a></li>
27<!-- <li><a href="#t_packed" >Packed Type</a> -->
28 </ol>
29 </li>
30 </ol>
31 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]32 <li><a href="#highlevel">High Level Structure</a>
33 <ol>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]34 <li><a href="#modulestructure">Module Structure</a></li>
35 <li><a href="#globalvars">Global Variables</a></li>
36 <li><a href="#functionstructure">Function Structure</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]37 </ol>
38 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]39 <li><a href="#instref">Instruction Reference</a>
40 <ol>
41 <li><a href="#terminators">Terminator Instructions</a>
42 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]43 <li><a href="#i_ret">'<tt>ret</tt>' Instruction</a></li>
44 <li><a href="#i_br">'<tt>br</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]45 <li><a href="#i_switch">'<tt>switch</tt>' Instruction</a></li>
46 <li><a href="#i_invoke">'<tt>invoke</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]47 <li><a href="#i_unwind">'<tt>unwind</tt>' Instruction</a></li>
48 </ol>
49 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]50 <li><a href="#binaryops">Binary Operations</a>
51 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]52 <li><a href="#i_add">'<tt>add</tt>' Instruction</a></li>
53 <li><a href="#i_sub">'<tt>sub</tt>' Instruction</a></li>
54 <li><a href="#i_mul">'<tt>mul</tt>' Instruction</a></li>
55 <li><a href="#i_div">'<tt>div</tt>' Instruction</a></li>
56 <li><a href="#i_rem">'<tt>rem</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]57 <li><a href="#i_setcc">'<tt>set<i>cc</i></tt>' Instructions</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]58 </ol>
59 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]60 <li><a href="#bitwiseops">Bitwise Binary Operations</a>
61 <ol>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]62 <li><a href="#i_and">'<tt>and</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]63 <li><a href="#i_or">'<tt>or</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]64 <li><a href="#i_xor">'<tt>xor</tt>' Instruction</a></li>
65 <li><a href="#i_shl">'<tt>shl</tt>' Instruction</a></li>
66 <li><a href="#i_shr">'<tt>shr</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]67 </ol>
68 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]69 <li><a href="#memoryops">Memory Access Operations</a>
70 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]71 <li><a href="#i_malloc">'<tt>malloc</tt>' Instruction</a></li>
72 <li><a href="#i_free">'<tt>free</tt>' Instruction</a></li>
73 <li><a href="#i_alloca">'<tt>alloca</tt>' Instruction</a></li>
74 <li><a href="#i_load">'<tt>load</tt>' Instruction</a></li>
75 <li><a href="#i_store">'<tt>store</tt>' Instruction</a></li>
76 <li><a href="#i_getelementptr">'<tt>getelementptr</tt>' Instruction</a></li>
77 </ol>
78 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]79 <li><a href="#otherops">Other Operations</a>
80 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]81 <li><a href="#i_phi">'<tt>phi</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]82 <li><a href="#i_cast">'<tt>cast .. to</tt>' Instruction</a></li>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]83 <li><a href="#i_select">'<tt>select</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]84 <li><a href="#i_call">'<tt>call</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]85 <li><a href="#i_vanext">'<tt>vanext</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]86 <li><a href="#i_vaarg">'<tt>vaarg</tt>' Instruction</a></li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]87 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]88 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]89 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]90 </li>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]91 <li><a href="#intrinsics">Intrinsic Functions</a>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]92 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]93 <li><a href="#int_varargs">Variable Argument Handling Intrinsics</a>
94 <ol>
95 <li><a href="#i_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a></li>
96 <li><a href="#i_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a></li>
97 <li><a href="#i_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a></li>
98 </ol>
99 </li>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]100 <li><a href="#int_codegen">Code Generator Intrinsics</a>
101 <ol>
102 <li><a href="#i_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a></li>
103 <li><a href="#i_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a></li>
John Criswell7123e272004-04-09 16:43:20 +0000[diff] [blame^]104 </ol>
105 </li>
106 <li><a href="#int_os">Operating System Intrinsics</a>
107 <ol>
John Criswellcfd3bac2004-04-09 15:23:37 +0000[diff] [blame]108 <li><a href="#i_readport">'<tt>llvm.readport</tt>' Intrinsic</a></li>
109 <li><a href="#i_writeport">'<tt>llvm.writeport</tt>' Intrinsic</a></li>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]110 </ol>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]111 <li><a href="#int_libc">Standard C Library Intrinsics</a>
112 <ol>
113 <li><a href="#i_memcpy">'<tt>llvm.memcpy</tt>' Intrinsic</a></li>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]114 <li><a href="#i_memmove">'<tt>llvm.memmove</tt>' Intrinsic</a></li>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]115 <li><a href="#i_memset">'<tt>llvm.memset</tt>' Intrinsic</a></li>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]116 </ol>
117 </li>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]118 <li><a href="#int_debugger">Debugger intrinsics</a>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]119 </ol>
120 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]121</ol>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]122<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]123<p><b>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>
124and <a href="mailto:vadve@cs.uiuc.edu">Vikram Adve</a></b></p>
125<p> </p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]126</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]127<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]128<div class="doc_section"> <a name="abstract">Abstract </a></div>
129<!-- *********************************************************************** -->
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]130<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]131<p>This document is a reference manual for the LLVM assembly language.
132LLVM is an SSA based representation that provides type safety,
133low-level operations, flexibility, and the capability of representing
134'all' high-level languages cleanly. It is the common code
135representation used throughout all phases of the LLVM compilation
136strategy.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]137</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]138<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]139<div class="doc_section"> <a name="introduction">Introduction</a> </div>
140<!-- *********************************************************************** -->
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]141<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]142<p>The LLVM code representation is designed to be used in three
143different forms: as an in-memory compiler IR, as an on-disk bytecode
144representation (suitable for fast loading by a Just-In-Time compiler),
145and as a human readable assembly language representation. This allows
146LLVM to provide a powerful intermediate representation for efficient
147compiler transformations and analysis, while providing a natural means
148to debug and visualize the transformations. The three different forms
149of LLVM are all equivalent. This document describes the human readable
150representation and notation.</p>
151<p>The LLVM representation aims to be a light-weight and low-level
152while being expressive, typed, and extensible at the same time. It
153aims to be a "universal IR" of sorts, by being at a low enough level
154that high-level ideas may be cleanly mapped to it (similar to how
155microprocessors are "universal IR's", allowing many source languages to
156be mapped to them). By providing type information, LLVM can be used as
157the target of optimizations: for example, through pointer analysis, it
158can be proven that a C automatic variable is never accessed outside of
159the current function... allowing it to be promoted to a simple SSA
160value instead of a memory location.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]161</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]162<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]163<div class="doc_subsubsection"> <a name="wellformed">Well-Formedness</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]164<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]165<p>It is important to note that this document describes 'well formed'
166LLVM assembly language. There is a difference between what the parser
167accepts and what is considered 'well formed'. For example, the
168following instruction is syntactically okay, but not well formed:</p>
169<pre> %x = <a href="#i_add">add</a> int 1, %x<br></pre>
170<p>...because the definition of <tt>%x</tt> does not dominate all of
171its uses. The LLVM infrastructure provides a verification pass that may
172be used to verify that an LLVM module is well formed. This pass is
173automatically run by the parser after parsing input assembly, and by
174the optimizer before it outputs bytecode. The violations pointed out
175by the verifier pass indicate bugs in transformation passes or input to
176the parser.</p>
177<!-- Describe the typesetting conventions here. --> </div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]178<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]179<div class="doc_section"> <a name="identifiers">Identifiers</a> </div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]180<!-- *********************************************************************** -->
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]181<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]182<p>LLVM uses three different forms of identifiers, for different
183purposes:</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]184<ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]185 <li>Numeric constants are represented as you would expect: 12, -3
186123.421, etc. Floating point constants have an optional hexidecimal
187notation.</li>
188 <li>Named values are represented as a string of characters with a '%'
189prefix. For example, %foo, %DivisionByZero,
190%a.really.long.identifier. The actual regular expression used is '<tt>%[a-zA-Z$._][a-zA-Z$._0-9]*</tt>'.
191Identifiers which require other characters in their names can be
192surrounded with quotes. In this way, anything except a <tt>"</tt>
193character can be used in a name.</li>
194 <li>Unnamed values are represented as an unsigned numeric value with
195a '%' prefix. For example, %12, %2, %44.</li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]196</ol>
John Criswell6794d922004-03-12 21:19:06 +0000[diff] [blame]197<p>LLVM requires that values start with a '%' sign for two reasons:
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]198Compilers don't need to worry about name clashes with reserved words,
199and the set of reserved words may be expanded in the future without
200penalty. Additionally, unnamed identifiers allow a compiler to quickly
201come up with a temporary variable without having to avoid symbol table
202conflicts.</p>
203<p>Reserved words in LLVM are very similar to reserved words in other
204languages. There are keywords for different opcodes ('<tt><a
205 href="#i_add">add</a></tt>', '<tt><a href="#i_cast">cast</a></tt>', '<tt><a
206 href="#i_ret">ret</a></tt>', etc...), for primitive type names ('<tt><a
207 href="#t_void">void</a></tt>', '<tt><a href="#t_uint">uint</a></tt>',
208etc...), and others. These reserved words cannot conflict with
209variable names, because none of them start with a '%' character.</p>
210<p>Here is an example of LLVM code to multiply the integer variable '<tt>%X</tt>'
211by 8:</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]212<p>The easy way:</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]213<pre> %result = <a href="#i_mul">mul</a> uint %X, 8<br></pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]214<p>After strength reduction:</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]215<pre> %result = <a href="#i_shl">shl</a> uint %X, ubyte 3<br></pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]216<p>And the hard way:</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]217<pre> <a href="#i_add">add</a> uint %X, %X <i>; yields {uint}:%0</i>
218 <a
219 href="#i_add">add</a> uint %0, %0 <i>; yields {uint}:%1</i>
220 %result = <a
221 href="#i_add">add</a> uint %1, %1<br></pre>
222<p>This last way of multiplying <tt>%X</tt> by 8 illustrates several
223important lexical features of LLVM:</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]224<ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]225 <li>Comments are delimited with a '<tt>;</tt>' and go until the end
226of line.</li>
227 <li>Unnamed temporaries are created when the result of a computation
228is not assigned to a named value.</li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]229 <li>Unnamed temporaries are numbered sequentially</li>
230</ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]231<p>...and it also show a convention that we follow in this document.
232When demonstrating instructions, we will follow an instruction with a
233comment that defines the type and name of value produced. Comments are
234shown in italic text.</p>
235<p>The one non-intuitive notation for constants is the optional
236hexidecimal form of floating point constants. For example, the form '<tt>double
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2370x432ff973cafa8000</tt>' is equivalent to (but harder to read than) '<tt>double
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2384.5e+15</tt>' which is also supported by the parser. The only time
239hexadecimal floating point constants are useful (and the only time that
240they are generated by the disassembler) is when an FP constant has to
241be emitted that is not representable as a decimal floating point number
242exactly. For example, NaN's, infinities, and other special cases are
243represented in their IEEE hexadecimal format so that assembly and
244disassembly do not cause any bits to change in the constants.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]245</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]246<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]247<div class="doc_section"> <a name="typesystem">Type System</a> </div>
248<!-- *********************************************************************** -->
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]249<div class="doc_text">
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]250<p>The LLVM type system is one of the most important features of the
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]251intermediate representation. Being typed enables a number of
252optimizations to be performed on the IR directly, without having to do
253extra analyses on the side before the transformation. A strong type
254system makes it easier to read the generated code and enables novel
255analyses and transformations that are not feasible to perform on normal
256three address code representations.</p>
Chris Lattner7bae3952002-06-25 18:03:17 +0000[diff] [blame]257<!-- The written form for the type system was heavily influenced by the
258syntactic problems with types in the C language<sup><a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]259href="#rw_stroustrup">1</a></sup>.<p> --> </div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]260<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]261<div class="doc_subsection"> <a name="t_primitive">Primitive Types</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]262<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]263<p>The primitive types are the fundemental building blocks of the LLVM
264system. The current set of primitive types are as follows:</p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]265
266<table border="0" style="align: center">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]267 <tbody>
268 <tr>
269 <td>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]270 <table border="1" cellspacing="0" cellpadding="4" style="align: center">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]271 <tbody>
272 <tr>
273 <td><tt>void</tt></td>
274 <td>No value</td>
275 </tr>
276 <tr>
277 <td><tt>ubyte</tt></td>
278 <td>Unsigned 8 bit value</td>
279 </tr>
280 <tr>
281 <td><tt>ushort</tt></td>
282 <td>Unsigned 16 bit value</td>
283 </tr>
284 <tr>
285 <td><tt>uint</tt></td>
286 <td>Unsigned 32 bit value</td>
287 </tr>
288 <tr>
289 <td><tt>ulong</tt></td>
290 <td>Unsigned 64 bit value</td>
291 </tr>
292 <tr>
293 <td><tt>float</tt></td>
294 <td>32 bit floating point value</td>
295 </tr>
296 <tr>
297 <td><tt>label</tt></td>
298 <td>Branch destination</td>
299 </tr>
300 </tbody>
301 </table>
302 </td>
303 <td valign="top">
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]304 <table border="1" cellspacing="0" cellpadding="4">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]305 <tbody>
306 <tr>
307 <td><tt>bool</tt></td>
308 <td>True or False value</td>
309 </tr>
310 <tr>
311 <td><tt>sbyte</tt></td>
312 <td>Signed 8 bit value</td>
313 </tr>
314 <tr>
315 <td><tt>short</tt></td>
316 <td>Signed 16 bit value</td>
317 </tr>
318 <tr>
319 <td><tt>int</tt></td>
320 <td>Signed 32 bit value</td>
321 </tr>
322 <tr>
323 <td><tt>long</tt></td>
324 <td>Signed 64 bit value</td>
325 </tr>
326 <tr>
327 <td><tt>double</tt></td>
328 <td>64 bit floating point value</td>
329 </tr>
330 </tbody>
331 </table>
332 </td>
333 </tr>
334 </tbody>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]335</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]336
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]337</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]338<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]339<div class="doc_subsubsection"> <a name="t_classifications">Type
340Classifications</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]341<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]342<p>These different primitive types fall into a few useful
343classifications:</p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]344
345<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]346 <tbody>
347 <tr>
348 <td><a name="t_signed">signed</a></td>
349 <td><tt>sbyte, short, int, long, float, double</tt></td>
350 </tr>
351 <tr>
352 <td><a name="t_unsigned">unsigned</a></td>
353 <td><tt>ubyte, ushort, uint, ulong</tt></td>
354 </tr>
355 <tr>
356 <td><a name="t_integer">integer</a></td>
357 <td><tt>ubyte, sbyte, ushort, short, uint, int, ulong, long</tt></td>
358 </tr>
359 <tr>
360 <td><a name="t_integral">integral</a></td>
361 <td><tt>bool, ubyte, sbyte, ushort, short, uint, int, ulong, long</tt></td>
362 </tr>
363 <tr>
364 <td><a name="t_floating">floating point</a></td>
365 <td><tt>float, double</tt></td>
366 </tr>
367 <tr>
368 <td><a name="t_firstclass">first class</a></td>
369 <td><tt>bool, ubyte, sbyte, ushort, short,<br>
370uint, int, ulong, long, float, double, <a href="#t_pointer">pointer</a></tt></td>
371 </tr>
372 </tbody>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]373</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]374
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]375<p>The <a href="#t_firstclass">first class</a> types are perhaps the
376most important. Values of these types are the only ones which can be
377produced by instructions, passed as arguments, or used as operands to
378instructions. This means that all structures and arrays must be
379manipulated either by pointer or by component.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]380</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]381<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]382<div class="doc_subsection"> <a name="t_derived">Derived Types</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]383<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]384<p>The real power in LLVM comes from the derived types in the system.
385This is what allows a programmer to represent arrays, functions,
386pointers, and other useful types. Note that these derived types may be
387recursive: For example, it is possible to have a two dimensional array.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]388</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]389<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]390<div class="doc_subsubsection"> <a name="t_array">Array Type</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]391<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]392<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]393<p>The array type is a very simple derived type that arranges elements
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]394sequentially in memory. The array type requires a size (number of
395elements) and an underlying data type.</p>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]396<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]397<pre> [&lt;# elements&gt; x &lt;elementtype&gt;]<br></pre>
398<p>The number of elements is a constant integer value, elementtype may
399be any type with a size.</p>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]400<h5>Examples:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]401<p> <tt>[40 x int ]</tt>: Array of 40 integer values.<br>
402<tt>[41 x int ]</tt>: Array of 41 integer values.<br>
403<tt>[40 x uint]</tt>: Array of 40 unsigned integer values.</p>
404<p> </p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]405<p>Here are some examples of multidimensional arrays:</p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]406
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]407<table border="0" cellpadding="0" cellspacing="0">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]408 <tbody>
409 <tr>
410 <td><tt>[3 x [4 x int]]</tt></td>
411 <td>: 3x4 array integer values.</td>
412 </tr>
413 <tr>
414 <td><tt>[12 x [10 x float]]</tt></td>
415 <td>: 12x10 array of single precision floating point values.</td>
416 </tr>
417 <tr>
418 <td><tt>[2 x [3 x [4 x uint]]]</tt></td>
419 <td>: 2x3x4 array of unsigned integer values.</td>
420 </tr>
421 </tbody>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]422</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]423
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]424</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]425<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]426<div class="doc_subsubsection"> <a name="t_function">Function Type</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]427<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]428<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]429<p>The function type can be thought of as a function signature. It
430consists of a return type and a list of formal parameter types.
John Criswell009900b2003-11-25 21:45:46 +0000[diff] [blame]431Function types are usually used to build virtual function tables
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]432(which are structures of pointers to functions), for indirect function
433calls, and when defining a function.</p>
John Criswell009900b2003-11-25 21:45:46 +0000[diff] [blame]434<p>
435The return type of a function type cannot be an aggregate type.
436</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]437<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]438<pre> &lt;returntype&gt; (&lt;parameter list&gt;)<br></pre>
439<p>Where '<tt>&lt;parameter list&gt;</tt>' is a comma-separated list of
440type specifiers. Optionally, the parameter list may include a type <tt>...</tt>,
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]441which indicates that the function takes a variable number of arguments.
442Variable argument functions can access their arguments with the <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]443 href="#int_varargs">variable argument handling intrinsic</a> functions.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]444<h5>Examples:</h5>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]445
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]446<table border="0" cellpadding="0" cellspacing="0">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]447 <tbody>
448 <tr>
449 <td><tt>int (int)</tt></td>
450 <td>: function taking an <tt>int</tt>, returning an <tt>int</tt></td>
451 </tr>
452 <tr>
453 <td><tt>float (int, int *) *</tt></td>
454 <td>: <a href="#t_pointer">Pointer</a> to a function that takes
455an <tt>int</tt> and a <a href="#t_pointer">pointer</a> to <tt>int</tt>,
456returning <tt>float</tt>.</td>
457 </tr>
458 <tr>
459 <td><tt>int (sbyte *, ...)</tt></td>
460 <td>: A vararg function that takes at least one <a
461 href="#t_pointer">pointer</a> to <tt>sbyte</tt> (signed char in C),
462which returns an integer. This is the signature for <tt>printf</tt>
463in LLVM.</td>
464 </tr>
465 </tbody>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]466</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]467
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]468</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]469<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]470<div class="doc_subsubsection"> <a name="t_struct">Structure Type</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]471<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]472<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]473<p>The structure type is used to represent a collection of data members
474together in memory. The packing of the field types is defined to match
475the ABI of the underlying processor. The elements of a structure may
476be any type that has a size.</p>
477<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
478and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
479field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
480instruction.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]481<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]482<pre> { &lt;type list&gt; }<br></pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]483<h5>Examples:</h5>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]484
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]485<table border="0" cellpadding="0" cellspacing="0">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]486 <tbody>
487 <tr>
488 <td><tt>{ int, int, int }</tt></td>
489 <td>: a triple of three <tt>int</tt> values</td>
490 </tr>
491 <tr>
492 <td><tt>{ float, int (int) * }</tt></td>
493 <td>: A pair, where the first element is a <tt>float</tt> and the
494second element is a <a href="#t_pointer">pointer</a> to a <a
495 href="t_function">function</a> that takes an <tt>int</tt>, returning
496an <tt>int</tt>.</td>
497 </tr>
498 </tbody>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]499</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]500
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]501</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]502<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]503<div class="doc_subsubsection"> <a name="t_pointer">Pointer Type</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]504<div class="doc_text">
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]505<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]506<p>As in many languages, the pointer type represents a pointer or
507reference to another object, which must live in memory.</p>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]508<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]509<pre> &lt;type&gt; *<br></pre>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]510<h5>Examples:</h5>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]511
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]512<table border="0" cellpadding="0" cellspacing="0">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]513 <tbody>
514 <tr>
515 <td><tt>[4x int]*</tt></td>
516 <td>: <a href="#t_pointer">pointer</a> to <a href="#t_array">array</a>
517of four <tt>int</tt> values</td>
518 </tr>
519 <tr>
520 <td><tt>int (int *) *</tt></td>
521 <td>: A <a href="#t_pointer">pointer</a> to a <a
522 href="t_function">function</a> that takes an <tt>int</tt>, returning
523an <tt>int</tt>.</td>
524 </tr>
525 </tbody>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]526</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]527
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]528</div>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]529<!-- _______________________________________________________________________ --><!--
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]530<div class="doc_subsubsection">
531 <a name="t_packed">Packed Type</a>
532</div>
533
534<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]535
536Mention/decide that packed types work with saturation or not. Maybe have a packed+saturated type in addition to just a packed type.<p>
537
538Packed types should be 'nonsaturated' because standard data types are not saturated. Maybe have a saturated packed type?<p>
539
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]540</div>
541
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]542--><!-- *********************************************************************** -->
543<div class="doc_section"> <a name="highlevel">High Level Structure</a> </div>
544<!-- *********************************************************************** --><!-- ======================================================================= -->
545<div class="doc_subsection"> <a name="modulestructure">Module Structure</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]546<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]547<p>LLVM programs are composed of "Module"s, each of which is a
548translation unit of the input programs. Each module consists of
549functions, global variables, and symbol table entries. Modules may be
550combined together with the LLVM linker, which merges function (and
551global variable) definitions, resolves forward declarations, and merges
552symbol table entries. Here is an example of the "hello world" module:</p>
553<pre><i>; Declare the string constant as a global constant...</i>
554<a href="#identifiers">%.LC0</a> = <a href="#linkage_internal">internal</a> <a
555 href="#globalvars">constant</a> <a href="#t_array">[13 x sbyte]</a> c"hello world\0A\00" <i>; [13 x sbyte]*</i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]556
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]557<i>; External declaration of the puts function</i>
558<a href="#functionstructure">declare</a> int %puts(sbyte*) <i>; int(sbyte*)* </i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]559
560<i>; Definition of main function</i>
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]561int %main() { <i>; int()* </i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]562 <i>; Convert [13x sbyte]* to sbyte *...</i>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]563 %cast210 = <a
564 href="#i_getelementptr">getelementptr</a> [13 x sbyte]* %.LC0, long 0, long 0 <i>; sbyte*</i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]565
566 <i>; Call puts function to write out the string to stdout...</i>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]567 <a
568 href="#i_call">call</a> int %puts(sbyte* %cast210) <i>; int</i>
569 <a
570 href="#i_ret">ret</a> int 0<br>}<br></pre>
571<p>This example is made up of a <a href="#globalvars">global variable</a>
572named "<tt>.LC0</tt>", an external declaration of the "<tt>puts</tt>"
573function, and a <a href="#functionstructure">function definition</a>
574for "<tt>main</tt>".</p>
575<a name="linkage"> In general, a module is made up of a list of global
576values, where both functions and global variables are global values.
577Global values are represented by a pointer to a memory location (in
578this case, a pointer to an array of char, and a pointer to a function),
579and have one of the following linkage types:</a>
580<p> </p>
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]581<dl>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]582 <dt><tt><b><a name="linkage_internal">internal</a></b></tt> </dt>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]583 <dd>Global values with internal linkage are only directly accessible
584by objects in the current module. In particular, linking code into a
585module with an internal global value may cause the internal to be
586renamed as necessary to avoid collisions. Because the symbol is
587internal to the module, all references can be updated. This
588corresponds to the notion of the '<tt>static</tt>' keyword in C, or the
589idea of "anonymous namespaces" in C++.
590 <p> </p>
591 </dd>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]592 <dt><tt><b><a name="linkage_linkonce">linkonce</a></b></tt>: </dt>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]593 <dd>"<tt>linkonce</tt>" linkage is similar to <tt>internal</tt>
594linkage, with the twist that linking together two modules defining the
595same <tt>linkonce</tt> globals will cause one of the globals to be
596discarded. This is typically used to implement inline functions.
597Unreferenced <tt>linkonce</tt> globals are allowed to be discarded.
598 <p> </p>
599 </dd>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]600 <dt><tt><b><a name="linkage_weak">weak</a></b></tt>: </dt>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]601 <dd>"<tt>weak</tt>" linkage is exactly the same as <tt>linkonce</tt>
602linkage, except that unreferenced <tt>weak</tt> globals may not be
603discarded. This is used to implement constructs in C such as "<tt>int
604X;</tt>" at global scope.
605 <p> </p>
606 </dd>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]607 <dt><tt><b><a name="linkage_appending">appending</a></b></tt>: </dt>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]608 <dd>"<tt>appending</tt>" linkage may only be applied to global
609variables of pointer to array type. When two global variables with
610appending linkage are linked together, the two global arrays are
611appended together. This is the LLVM, typesafe, equivalent of having
612the system linker append together "sections" with identical names when
613.o files are linked.
614 <p> </p>
615 </dd>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]616 <dt><tt><b><a name="linkage_external">externally visible</a></b></tt>:</dt>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]617 <dd>If none of the above identifiers are used, the global is
618externally visible, meaning that it participates in linkage and can be
619used to resolve external symbol references.
620 <p> </p>
621 </dd>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]622</dl>
623<p> </p>
624<p><a name="linkage_external">For example, since the "<tt>.LC0</tt>"
625variable is defined to be internal, if another module defined a "<tt>.LC0</tt>"
626variable and was linked with this one, one of the two would be renamed,
627preventing a collision. Since "<tt>main</tt>" and "<tt>puts</tt>" are
628external (i.e., lacking any linkage declarations), they are accessible
629outside of the current module. It is illegal for a function <i>declaration</i>
630to have any linkage type other than "externally visible".</a></p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]631</div>
Chris Lattner9ee5d222004-03-08 16:49:10 +0000[diff] [blame]632
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]633<!-- ======================================================================= -->
Chris Lattner9ee5d222004-03-08 16:49:10 +0000[diff] [blame]634<div class="doc_subsection">
635 <a name="globalvars">Global Variables</a>
636</div>
637
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]638<div class="doc_text">
Chris Lattner9ee5d222004-03-08 16:49:10 +0000[diff] [blame]639
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]640<p>Global variables define regions of memory allocated at compilation
641time instead of run-time. Global variables may optionally be
642initialized. A variable may be defined as a global "constant", which
643indicates that the contents of the variable will never be modified
Chris Lattner9ee5d222004-03-08 16:49:10 +0000[diff] [blame]644(opening options for optimization).</p>
645
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]646<p>As SSA values, global variables define pointer values that are in
647scope (i.e. they dominate) for all basic blocks in the program. Global
648variables always define a pointer to their "content" type because they
649describe a region of memory, and all memory objects in LLVM are
650accessed through pointers.</p>
Chris Lattner9ee5d222004-03-08 16:49:10 +0000[diff] [blame]651
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]652</div>
Chris Lattner9ee5d222004-03-08 16:49:10 +0000[diff] [blame]653
654
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]655<!-- ======================================================================= -->
Chris Lattner9ee5d222004-03-08 16:49:10 +0000[diff] [blame]656<div class="doc_subsection">
657 <a name="functionstructure">Functions</a>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]658</div>
Chris Lattner9ee5d222004-03-08 16:49:10 +0000[diff] [blame]659
660<div class="doc_text">
661
662<p>LLVM function definitions are composed of a (possibly empty) argument list,
663an opening curly brace, a list of basic blocks, and a closing curly brace. LLVM
664function declarations are defined with the "<tt>declare</tt>" keyword, a
665function name, and a function signature.</p>
666
667<p>A function definition contains a list of basic blocks, forming the CFG for
668the function. Each basic block may optionally start with a label (giving the
669basic block a symbol table entry), contains a list of instructions, and ends
670with a <a href="#terminators">terminator</a> instruction (such as a branch or
671function return).</p>
672
673<p>The first basic block in program is special in two ways: it is immediately
674executed on entrance to the function, and it is not allowed to have predecessor
675basic blocks (i.e. there can not be any branches to the entry block of a
676function). Because the block can have no predecessors, it also cannot have any
677<a href="#i_phi">PHI nodes</a>.</p>
678
679<p>LLVM functions are identified by their name and type signature. Hence, two
680functions with the same name but different parameter lists or return values are
681considered different functions, and LLVM will resolves references to each
682appropriately.</p>
683
684</div>
685
686
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]687<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]688<div class="doc_section"> <a name="instref">Instruction Reference</a> </div>
689<!-- *********************************************************************** -->
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]690<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]691<p>The LLVM instruction set consists of several different
692classifications of instructions: <a href="#terminators">terminator
693instructions</a>, <a href="#binaryops">binary instructions</a>, <a
694 href="#memoryops">memory instructions</a>, and <a href="#otherops">other
695instructions</a>.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]696</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]697<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]698<div class="doc_subsection"> <a name="terminators">Terminator
699Instructions</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]700<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]701<p>As mentioned <a href="#functionstructure">previously</a>, every
702basic block in a program ends with a "Terminator" instruction, which
703indicates which block should be executed after the current block is
704finished. These terminator instructions typically yield a '<tt>void</tt>'
705value: they produce control flow, not values (the one exception being
706the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction).</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]707<p>There are five different terminator instructions: the '<a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]708 href="#i_ret"><tt>ret</tt></a>' instruction, the '<a href="#i_br"><tt>br</tt></a>'
709instruction, the '<a href="#i_switch"><tt>switch</tt></a>' instruction,
710the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction, and the '<a
711 href="#i_unwind"><tt>unwind</tt></a>' instruction.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]712</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]713<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]714<div class="doc_subsubsection"> <a name="i_ret">'<tt>ret</tt>'
715Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]716<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]717<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]718<pre> ret &lt;type&gt; &lt;value&gt; <i>; Return a value from a non-void function</i>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]719 ret void <i>; Return from void function</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]720</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]721<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]722<p>The '<tt>ret</tt>' instruction is used to return control flow (and a
723value) from a function, back to the caller.</p>
724<p>There are two forms of the '<tt>ret</tt>' instructruction: one that
725returns a value and then causes control flow, and one that just causes
726control flow to occur.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]727<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]728<p>The '<tt>ret</tt>' instruction may return any '<a
729 href="#t_firstclass">first class</a>' type. Notice that a function is
730not <a href="#wellformed">well formed</a> if there exists a '<tt>ret</tt>'
731instruction inside of the function that returns a value that does not
732match the return type of the function.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]733<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]734<p>When the '<tt>ret</tt>' instruction is executed, control flow
735returns back to the calling function's context. If the caller is a "<a
736 href="#i_call"><tt>call</tt></a> instruction, execution continues at
737the instruction after the call. If the caller was an "<a
738 href="#i_invoke"><tt>invoke</tt></a>" instruction, execution continues
739at the beginning "normal" of the destination block. If the instruction
740returns a value, that value shall set the call or invoke instruction's
741return value.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]742<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]743<pre> ret int 5 <i>; Return an integer value of 5</i>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]744 ret void <i>; Return from a void function</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]745</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]746</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]747<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]748<div class="doc_subsubsection"> <a name="i_br">'<tt>br</tt>' Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]749<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]750<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]751<pre> br bool &lt;cond&gt;, label &lt;iftrue&gt;, label &lt;iffalse&gt;<br> br label &lt;dest&gt; <i>; Unconditional branch</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]752</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]753<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]754<p>The '<tt>br</tt>' instruction is used to cause control flow to
755transfer to a different basic block in the current function. There are
756two forms of this instruction, corresponding to a conditional branch
757and an unconditional branch.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]758<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]759<p>The conditional branch form of the '<tt>br</tt>' instruction takes a
760single '<tt>bool</tt>' value and two '<tt>label</tt>' values. The
761unconditional form of the '<tt>br</tt>' instruction takes a single '<tt>label</tt>'
762value as a target.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]763<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]764<p>Upon execution of a conditional '<tt>br</tt>' instruction, the '<tt>bool</tt>'
765argument is evaluated. If the value is <tt>true</tt>, control flows
766to the '<tt>iftrue</tt>' <tt>label</tt> argument. If "cond" is <tt>false</tt>,
767control flows to the '<tt>iffalse</tt>' <tt>label</tt> argument.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]768<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]769<pre>Test:<br> %cond = <a href="#i_setcc">seteq</a> int %a, %b<br> br bool %cond, label %IfEqual, label %IfUnequal<br>IfEqual:<br> <a
770 href="#i_ret">ret</a> int 1<br>IfUnequal:<br> <a href="#i_ret">ret</a> int 0<br></pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]771</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]772<!-- _______________________________________________________________________ -->
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]773<div class="doc_subsubsection">
774 <a name="i_switch">'<tt>switch</tt>' Instruction</a>
775</div>
776
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]777<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]778<h5>Syntax:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]779
780<pre>
781 switch &lt;intty&gt; &lt;value&gt;, label &lt;defaultdest&gt; [ &lt;intty&gt; &lt;val&gt;, label &lt;dest&gt; ... ]
782</pre>
783
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]784<h5>Overview:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]785
786<p>The '<tt>switch</tt>' instruction is used to transfer control flow to one of
787several different places. It is a generalization of the '<tt>br</tt>'
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]788instruction, allowing a branch to occur to one of many possible
789destinations.</p>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]790
791
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]792<h5>Arguments:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]793
794<p>The '<tt>switch</tt>' instruction uses three parameters: an integer
795comparison value '<tt>value</tt>', a default '<tt>label</tt>' destination, and
796an array of pairs of comparison value constants and '<tt>label</tt>'s. The
797table is not allowed to contain duplicate constant entries.</p>
798
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]799<h5>Semantics:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]800
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]801<p>The <tt>switch</tt> instruction specifies a table of values and
802destinations. When the '<tt>switch</tt>' instruction is executed, this
803table is searched for the given value. If the value is found, the
804corresponding destination is branched to, otherwise the default value
805it transfered to.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]806
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]807<h5>Implementation:</h5>
808
809<p>Depending on properties of the target machine and the particular
810<tt>switch</tt> instruction, this instruction may be code generated in different
811ways, for example as a series of chained conditional branches, or with a lookup
812table.</p>
813
814<h5>Example:</h5>
815
816<pre>
817 <i>; Emulate a conditional br instruction</i>
818 %Val = <a href="#i_cast">cast</a> bool %value to int
819 switch int %Val, label %truedest [int 0, label %falsedest ]
820
821 <i>; Emulate an unconditional br instruction</i>
822 switch uint 0, label %dest [ ]
823
824 <i>; Implement a jump table:</i>
825 switch uint %val, label %otherwise [ uint 0, label %onzero
826 uint 1, label %onone
827 uint 2, label %ontwo ]
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]828</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]829</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]830<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]831<div class="doc_subsubsection"> <a name="i_invoke">'<tt>invoke</tt>'
832Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]833<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]834<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]835<pre> &lt;result&gt; = invoke &lt;ptr to function ty&gt; %&lt;function ptr val&gt;(&lt;function args&gt;)<br> to label &lt;normal label&gt; except label &lt;exception label&gt;<br></pre>
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]836<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]837<p>The '<tt>invoke</tt>' instruction causes control to transfer to a
838specified function, with the possibility of control flow transfer to
839either the '<tt>normal</tt>' <tt>label</tt> label or the '<tt>exception</tt>'<tt>label</tt>.
840If the callee function returns with the "<tt><a href="#i_ret">ret</a></tt>"
841instruction, control flow will return to the "normal" label. If the
842callee (or any indirect callees) returns with the "<a href="#i_unwind"><tt>unwind</tt></a>"
843instruction, control is interrupted, and continued at the dynamically
844nearest "except" label.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]845<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]846<p>This instruction requires several arguments:</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]847<ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]848 <li>'<tt>ptr to function ty</tt>': shall be the signature of the
849pointer to function value being invoked. In most cases, this is a
850direct function invocation, but indirect <tt>invoke</tt>s are just as
851possible, branching off an arbitrary pointer to function value. </li>
852 <li>'<tt>function ptr val</tt>': An LLVM value containing a pointer
853to a function to be invoked. </li>
854 <li>'<tt>function args</tt>': argument list whose types match the
855function signature argument types. If the function signature indicates
856the function accepts a variable number of arguments, the extra
857arguments can be specified. </li>
858 <li>'<tt>normal label</tt>': the label reached when the called
859function executes a '<tt><a href="#i_ret">ret</a></tt>' instruction. </li>
860 <li>'<tt>exception label</tt>': the label reached when a callee
861returns with the <a href="#i_unwind"><tt>unwind</tt></a> instruction. </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]862</ol>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]863<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]864<p>This instruction is designed to operate as a standard '<tt><a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]865 href="#i_call">call</a></tt>' instruction in most regards. The
866primary difference is that it establishes an association with a label,
867which is used by the runtime library to unwind the stack.</p>
868<p>This instruction is used in languages with destructors to ensure
869that proper cleanup is performed in the case of either a <tt>longjmp</tt>
870or a thrown exception. Additionally, this is important for
871implementation of '<tt>catch</tt>' clauses in high-level languages that
872support them.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]873<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]874<pre> %retval = invoke int %Test(int 15)<br> to label %Continue<br> except label %TestCleanup <i>; {int}:retval set</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]875</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]876</div>
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]877<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]878<div class="doc_subsubsection"> <a name="i_unwind">'<tt>unwind</tt>'
879Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]880<div class="doc_text">
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]881<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]882<pre> unwind<br></pre>
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]883<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]884<p>The '<tt>unwind</tt>' instruction unwinds the stack, continuing
885control flow at the first callee in the dynamic call stack which used
886an <a href="#i_invoke"><tt>invoke</tt></a> instruction to perform the
887call. This is primarily used to implement exception handling.</p>
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]888<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]889<p>The '<tt>unwind</tt>' intrinsic causes execution of the current
890function to immediately halt. The dynamic call stack is then searched
891for the first <a href="#i_invoke"><tt>invoke</tt></a> instruction on
892the call stack. Once found, execution continues at the "exceptional"
893destination block specified by the <tt>invoke</tt> instruction. If
894there is no <tt>invoke</tt> instruction in the dynamic call chain,
895undefined behavior results.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]896</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]897<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]898<div class="doc_subsection"> <a name="binaryops">Binary Operations</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]899<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]900<p>Binary operators are used to do most of the computation in a
901program. They require two operands, execute an operation on them, and
902produce a single value. The result value of a binary operator is not
903necessarily the same type as its operands.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]904<p>There are several different binary operators:</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]905</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]906<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]907<div class="doc_subsubsection"> <a name="i_add">'<tt>add</tt>'
908Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]909<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]910<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]911<pre> &lt;result&gt; = add &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]912</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]913<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]914<p>The '<tt>add</tt>' instruction returns the sum of its two operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]915<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]916<p>The two arguments to the '<tt>add</tt>' instruction must be either <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]917 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
918values. Both arguments must have identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]919<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]920<p>The value produced is the integer or floating point sum of the two
921operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]922<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]923<pre> &lt;result&gt; = add int 4, %var <i>; yields {int}:result = 4 + %var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]924</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]925</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]926<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]927<div class="doc_subsubsection"> <a name="i_sub">'<tt>sub</tt>'
928Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]929<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]930<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]931<pre> &lt;result&gt; = sub &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]932</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]933<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]934<p>The '<tt>sub</tt>' instruction returns the difference of its two
935operands.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]936<p>Note that the '<tt>sub</tt>' instruction is used to represent the '<tt>neg</tt>'
937instruction present in most other intermediate representations.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]938<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]939<p>The two arguments to the '<tt>sub</tt>' instruction must be either <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]940 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
941values. Both arguments must have identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]942<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]943<p>The value produced is the integer or floating point difference of
944the two operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]945<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]946<pre> &lt;result&gt; = sub int 4, %var <i>; yields {int}:result = 4 - %var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]947 &lt;result&gt; = sub int 0, %val <i>; yields {int}:result = -%var</i>
948</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]949</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]950<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]951<div class="doc_subsubsection"> <a name="i_mul">'<tt>mul</tt>'
952Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]953<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]954<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]955<pre> &lt;result&gt; = mul &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]956</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]957<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]958<p>The '<tt>mul</tt>' instruction returns the product of its two
959operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]960<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]961<p>The two arguments to the '<tt>mul</tt>' instruction must be either <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]962 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
963values. Both arguments must have identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]964<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]965<p>The value produced is the integer or floating point product of the
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]966two operands.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]967<p>There is no signed vs unsigned multiplication. The appropriate
968action is taken based on the type of the operand.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]969<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]970<pre> &lt;result&gt; = mul int 4, %var <i>; yields {int}:result = 4 * %var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]971</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]972</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]973<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]974<div class="doc_subsubsection"> <a name="i_div">'<tt>div</tt>'
975Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]976<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]977<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]978<pre> &lt;result&gt; = div &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
979</pre>
980<h5>Overview:</h5>
981<p>The '<tt>div</tt>' instruction returns the quotient of its two
982operands.</p>
983<h5>Arguments:</h5>
984<p>The two arguments to the '<tt>div</tt>' instruction must be either <a
985 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
986values. Both arguments must have identical types.</p>
987<h5>Semantics:</h5>
988<p>The value produced is the integer or floating point quotient of the
989two operands.</p>
990<h5>Example:</h5>
991<pre> &lt;result&gt; = div int 4, %var <i>; yields {int}:result = 4 / %var</i>
992</pre>
993</div>
994<!-- _______________________________________________________________________ -->
995<div class="doc_subsubsection"> <a name="i_rem">'<tt>rem</tt>'
996Instruction</a> </div>
997<div class="doc_text">
998<h5>Syntax:</h5>
999<pre> &lt;result&gt; = rem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
1000</pre>
1001<h5>Overview:</h5>
1002<p>The '<tt>rem</tt>' instruction returns the remainder from the
1003division of its two operands.</p>
1004<h5>Arguments:</h5>
1005<p>The two arguments to the '<tt>rem</tt>' instruction must be either <a
1006 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
1007values. Both arguments must have identical types.</p>
1008<h5>Semantics:</h5>
1009<p>This returns the <i>remainder</i> of a division (where the result
1010has the same sign as the divisor), not the <i>modulus</i> (where the
1011result has the same sign as the dividend) of a value. For more
1012information about the difference, see: <a
1013 href="http://mathforum.org/dr.math/problems/anne.4.28.99.html">The
1014Math Forum</a>.</p>
1015<h5>Example:</h5>
1016<pre> &lt;result&gt; = rem int 4, %var <i>; yields {int}:result = 4 % %var</i>
1017</pre>
1018</div>
1019<!-- _______________________________________________________________________ -->
1020<div class="doc_subsubsection"> <a name="i_setcc">'<tt>set<i>cc</i></tt>'
1021Instructions</a> </div>
1022<div class="doc_text">
1023<h5>Syntax:</h5>
1024<pre> &lt;result&gt; = seteq &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {bool}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1025 &lt;result&gt; = setne &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {bool}:result</i>
1026 &lt;result&gt; = setlt &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {bool}:result</i>
1027 &lt;result&gt; = setgt &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {bool}:result</i>
1028 &lt;result&gt; = setle &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {bool}:result</i>
1029 &lt;result&gt; = setge &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {bool}:result</i>
1030</pre>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1031<h5>Overview:</h5>
1032<p>The '<tt>set<i>cc</i></tt>' family of instructions returns a boolean
1033value based on a comparison of their two operands.</p>
1034<h5>Arguments:</h5>
1035<p>The two arguments to the '<tt>set<i>cc</i></tt>' instructions must
1036be of <a href="#t_firstclass">first class</a> type (it is not possible
1037to compare '<tt>label</tt>'s, '<tt>array</tt>'s, '<tt>structure</tt>'
1038or '<tt>void</tt>' values, etc...). Both arguments must have identical
1039types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1040<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1041<p>The '<tt>seteq</tt>' instruction yields a <tt>true</tt> '<tt>bool</tt>'
1042value if both operands are equal.<br>
1043The '<tt>setne</tt>' instruction yields a <tt>true</tt> '<tt>bool</tt>'
1044value if both operands are unequal.<br>
1045The '<tt>setlt</tt>' instruction yields a <tt>true</tt> '<tt>bool</tt>'
1046value if the first operand is less than the second operand.<br>
1047The '<tt>setgt</tt>' instruction yields a <tt>true</tt> '<tt>bool</tt>'
1048value if the first operand is greater than the second operand.<br>
1049The '<tt>setle</tt>' instruction yields a <tt>true</tt> '<tt>bool</tt>'
1050value if the first operand is less than or equal to the second operand.<br>
1051The '<tt>setge</tt>' instruction yields a <tt>true</tt> '<tt>bool</tt>'
1052value if the first operand is greater than or equal to the second
1053operand.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1054<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1055<pre> &lt;result&gt; = seteq int 4, 5 <i>; yields {bool}:result = false</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1056 &lt;result&gt; = setne float 4, 5 <i>; yields {bool}:result = true</i>
1057 &lt;result&gt; = setlt uint 4, 5 <i>; yields {bool}:result = true</i>
1058 &lt;result&gt; = setgt sbyte 4, 5 <i>; yields {bool}:result = false</i>
1059 &lt;result&gt; = setle sbyte 4, 5 <i>; yields {bool}:result = true</i>
1060 &lt;result&gt; = setge sbyte 4, 5 <i>; yields {bool}:result = false</i>
1061</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1062</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1063<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1064<div class="doc_subsection"> <a name="bitwiseops">Bitwise Binary
1065Operations</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1066<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1067<p>Bitwise binary operators are used to do various forms of
1068bit-twiddling in a program. They are generally very efficient
1069instructions, and can commonly be strength reduced from other
1070instructions. They require two operands, execute an operation on them,
1071and produce a single value. The resulting value of the bitwise binary
1072operators is always the same type as its first operand.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1073</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1074<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1075<div class="doc_subsubsection"> <a name="i_and">'<tt>and</tt>'
1076Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1077<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1078<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1079<pre> &lt;result&gt; = and &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1080</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1081<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1082<p>The '<tt>and</tt>' instruction returns the bitwise logical and of
1083its two operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1084<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1085<p>The two arguments to the '<tt>and</tt>' instruction must be <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1086 href="#t_integral">integral</a> values. Both arguments must have
1087identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1088<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1089<p>The truth table used for the '<tt>and</tt>' instruction is:</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1090<p> </p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1091<div style="align: center">
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1092<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1093 <tbody>
1094 <tr>
1095 <td>In0</td>
1096 <td>In1</td>
1097 <td>Out</td>
1098 </tr>
1099 <tr>
1100 <td>0</td>
1101 <td>0</td>
1102 <td>0</td>
1103 </tr>
1104 <tr>
1105 <td>0</td>
1106 <td>1</td>
1107 <td>0</td>
1108 </tr>
1109 <tr>
1110 <td>1</td>
1111 <td>0</td>
1112 <td>0</td>
1113 </tr>
1114 <tr>
1115 <td>1</td>
1116 <td>1</td>
1117 <td>1</td>
1118 </tr>
1119 </tbody>
1120</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1121</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1122<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1123<pre> &lt;result&gt; = and int 4, %var <i>; yields {int}:result = 4 &amp; %var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1124 &lt;result&gt; = and int 15, 40 <i>; yields {int}:result = 8</i>
1125 &lt;result&gt; = and int 4, 8 <i>; yields {int}:result = 0</i>
1126</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1127</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1128<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1129<div class="doc_subsubsection"> <a name="i_or">'<tt>or</tt>' Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1130<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1131<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1132<pre> &lt;result&gt; = or &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1133</pre>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1134<h5>Overview:</h5>
1135<p>The '<tt>or</tt>' instruction returns the bitwise logical inclusive
1136or of its two operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1137<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1138<p>The two arguments to the '<tt>or</tt>' instruction must be <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1139 href="#t_integral">integral</a> values. Both arguments must have
1140identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1141<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1142<p>The truth table used for the '<tt>or</tt>' instruction is:</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1143<p> </p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1144<div style="align: center">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1145<table border="1" cellspacing="0" cellpadding="4">
1146 <tbody>
1147 <tr>
1148 <td>In0</td>
1149 <td>In1</td>
1150 <td>Out</td>
1151 </tr>
1152 <tr>
1153 <td>0</td>
1154 <td>0</td>
1155 <td>0</td>
1156 </tr>
1157 <tr>
1158 <td>0</td>
1159 <td>1</td>
1160 <td>1</td>
1161 </tr>
1162 <tr>
1163 <td>1</td>
1164 <td>0</td>
1165 <td>1</td>
1166 </tr>
1167 <tr>
1168 <td>1</td>
1169 <td>1</td>
1170 <td>1</td>
1171 </tr>
1172 </tbody>
1173</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1174</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1175<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1176<pre> &lt;result&gt; = or int 4, %var <i>; yields {int}:result = 4 | %var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1177 &lt;result&gt; = or int 15, 40 <i>; yields {int}:result = 47</i>
1178 &lt;result&gt; = or int 4, 8 <i>; yields {int}:result = 12</i>
1179</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1180</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1181<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1182<div class="doc_subsubsection"> <a name="i_xor">'<tt>xor</tt>'
1183Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1184<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1185<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1186<pre> &lt;result&gt; = xor &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1187</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1188<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1189<p>The '<tt>xor</tt>' instruction returns the bitwise logical exclusive
1190or of its two operands. The <tt>xor</tt> is used to implement the
1191"one's complement" operation, which is the "~" operator in C.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1192<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1193<p>The two arguments to the '<tt>xor</tt>' instruction must be <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1194 href="#t_integral">integral</a> values. Both arguments must have
1195identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1196<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1197<p>The truth table used for the '<tt>xor</tt>' instruction is:</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1198<p> </p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1199<div style="align: center">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1200<table border="1" cellspacing="0" cellpadding="4">
1201 <tbody>
1202 <tr>
1203 <td>In0</td>
1204 <td>In1</td>
1205 <td>Out</td>
1206 </tr>
1207 <tr>
1208 <td>0</td>
1209 <td>0</td>
1210 <td>0</td>
1211 </tr>
1212 <tr>
1213 <td>0</td>
1214 <td>1</td>
1215 <td>1</td>
1216 </tr>
1217 <tr>
1218 <td>1</td>
1219 <td>0</td>
1220 <td>1</td>
1221 </tr>
1222 <tr>
1223 <td>1</td>
1224 <td>1</td>
1225 <td>0</td>
1226 </tr>
1227 </tbody>
1228</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1229</div>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1230<p> </p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1231<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1232<pre> &lt;result&gt; = xor int 4, %var <i>; yields {int}:result = 4 ^ %var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1233 &lt;result&gt; = xor int 15, 40 <i>; yields {int}:result = 39</i>
1234 &lt;result&gt; = xor int 4, 8 <i>; yields {int}:result = 12</i>
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]1235 &lt;result&gt; = xor int %V, -1 <i>; yields {int}:result = ~%V</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1236</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1237</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1238<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1239<div class="doc_subsubsection"> <a name="i_shl">'<tt>shl</tt>'
1240Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1241<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1242<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1243<pre> &lt;result&gt; = shl &lt;ty&gt; &lt;var1&gt;, ubyte &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1244</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1245<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1246<p>The '<tt>shl</tt>' instruction returns the first operand shifted to
1247the left a specified number of bits.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1248<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1249<p>The first argument to the '<tt>shl</tt>' instruction must be an <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1250 href="#t_integer">integer</a> type. The second argument must be an '<tt>ubyte</tt>'
1251type.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1252<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1253<p>The value produced is <tt>var1</tt> * 2<sup><tt>var2</tt></sup>.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1254<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1255<pre> &lt;result&gt; = shl int 4, ubyte %var <i>; yields {int}:result = 4 &lt;&lt; %var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1256 &lt;result&gt; = shl int 4, ubyte 2 <i>; yields {int}:result = 16</i>
1257 &lt;result&gt; = shl int 1, ubyte 10 <i>; yields {int}:result = 1024</i>
1258</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1259</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1260<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1261<div class="doc_subsubsection"> <a name="i_shr">'<tt>shr</tt>'
1262Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1263<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1264<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1265<pre> &lt;result&gt; = shr &lt;ty&gt; &lt;var1&gt;, ubyte &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1266</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1267<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1268<p>The '<tt>shr</tt>' instruction returns the first operand shifted to
1269the right a specified number of bits.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1270<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1271<p>The first argument to the '<tt>shr</tt>' instruction must be an <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1272 href="#t_integer">integer</a> type. The second argument must be an '<tt>ubyte</tt>'
1273type.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1274<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1275<p>If the first argument is a <a href="#t_signed">signed</a> type, the
1276most significant bit is duplicated in the newly free'd bit positions.
1277If the first argument is unsigned, zero bits shall fill the empty
1278positions.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1279<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1280<pre> &lt;result&gt; = shr int 4, ubyte %var <i>; yields {int}:result = 4 &gt;&gt; %var</i>
Chris Lattner8c6bb902003-06-18 21:30:51 +0000[diff] [blame]1281 &lt;result&gt; = shr uint 4, ubyte 1 <i>; yields {uint}:result = 2</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1282 &lt;result&gt; = shr int 4, ubyte 2 <i>; yields {int}:result = 1</i>
Chris Lattner8c6bb902003-06-18 21:30:51 +0000[diff] [blame]1283 &lt;result&gt; = shr sbyte 4, ubyte 3 <i>; yields {sbyte}:result = 0</i>
1284 &lt;result&gt; = shr sbyte -2, ubyte 1 <i>; yields {sbyte}:result = -1</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1285</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1286</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1287<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1288<div class="doc_subsection"> <a name="memoryops">Memory Access
1289Operations</a></div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1290<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1291<p>A key design point of an SSA-based representation is how it
1292represents memory. In LLVM, no memory locations are in SSA form, which
1293makes things very simple. This section describes how to read, write,
1294allocate and free memory in LLVM.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1295</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1296<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1297<div class="doc_subsubsection"> <a name="i_malloc">'<tt>malloc</tt>'
1298Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1299<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1300<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1301<pre> &lt;result&gt; = malloc &lt;type&gt;, uint &lt;NumElements&gt; <i>; yields {type*}:result</i>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1302 &lt;result&gt; = malloc &lt;type&gt; <i>; yields {type*}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1303</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1304<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1305<p>The '<tt>malloc</tt>' instruction allocates memory from the system
1306heap and returns a pointer to it.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1307<h5>Arguments:</h5>
John Criswell6e4ca612004-02-24 16:13:56 +0000[diff] [blame]1308<p>The '<tt>malloc</tt>' instruction allocates <tt>sizeof(&lt;type&gt;)*NumElements</tt>
1309bytes of memory from the operating system and returns a pointer of the
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1310appropriate type to the program. The second form of the instruction is
1311a shorter version of the first instruction that defaults to allocating
1312one element.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1313<p>'<tt>type</tt>' must be a sized type.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1314<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1315<p>Memory is allocated using the system "<tt>malloc</tt>" function, and
1316a pointer is returned.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1317<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1318<pre> %array = malloc [4 x ubyte ] <i>; yields {[%4 x ubyte]*}:array</i>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1319
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1320 %size = <a
1321 href="#i_add">add</a> uint 2, 2 <i>; yields {uint}:size = uint 4</i>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1322 %array1 = malloc ubyte, uint 4 <i>; yields {ubyte*}:array1</i>
1323 %array2 = malloc [12 x ubyte], uint %size <i>; yields {[12 x ubyte]*}:array2</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1324</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1325</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1326<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1327<div class="doc_subsubsection"> <a name="i_free">'<tt>free</tt>'
1328Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1329<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1330<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1331<pre> free &lt;type&gt; &lt;value&gt; <i>; yields {void}</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1332</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1333<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1334<p>The '<tt>free</tt>' instruction returns memory back to the unused
1335memory heap, to be reallocated in the future.</p>
1336<p> </p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1337<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1338<p>'<tt>value</tt>' shall be a pointer value that points to a value
1339that was allocated with the '<tt><a href="#i_malloc">malloc</a></tt>'
1340instruction.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1341<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1342<p>Access to the memory pointed to by the pointer is not longer defined
1343after this instruction executes.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1344<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1345<pre> %array = <a href="#i_malloc">malloc</a> [4 x ubyte] <i>; yields {[4 x ubyte]*}:array</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1346 free [4 x ubyte]* %array
1347</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1348</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1349<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1350<div class="doc_subsubsection"> <a name="i_alloca">'<tt>alloca</tt>'
1351Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1352<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1353<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1354<pre> &lt;result&gt; = alloca &lt;type&gt;, uint &lt;NumElements&gt; <i>; yields {type*}:result</i>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1355 &lt;result&gt; = alloca &lt;type&gt; <i>; yields {type*}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1356</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1357<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1358<p>The '<tt>alloca</tt>' instruction allocates memory on the current
1359stack frame of the procedure that is live until the current function
1360returns to its caller.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1361<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1362<p>The the '<tt>alloca</tt>' instruction allocates <tt>sizeof(&lt;type&gt;)*NumElements</tt>
1363bytes of memory on the runtime stack, returning a pointer of the
1364appropriate type to the program. The second form of the instruction is
1365a shorter version of the first that defaults to allocating one element.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1366<p>'<tt>type</tt>' may be any sized type.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1367<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1368<p>Memory is allocated, a pointer is returned. '<tt>alloca</tt>'d
1369memory is automatically released when the function returns. The '<tt>alloca</tt>'
1370instruction is commonly used to represent automatic variables that must
1371have an address available. When the function returns (either with the <tt><a
1372 href="#i_ret">ret</a></tt> or <tt><a href="#i_invoke">invoke</a></tt>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1373instructions), the memory is reclaimed.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1374<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1375<pre> %ptr = alloca int <i>; yields {int*}:ptr</i>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1376 %ptr = alloca int, uint 4 <i>; yields {int*}:ptr</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1377</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1378</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1379<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1380<div class="doc_subsubsection"> <a name="i_load">'<tt>load</tt>'
1381Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1382<div class="doc_text">
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]1383<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1384<pre> &lt;result&gt; = load &lt;ty&gt;* &lt;pointer&gt;<br> &lt;result&gt; = volatile load &lt;ty&gt;* &lt;pointer&gt;<br></pre>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]1385<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1386<p>The '<tt>load</tt>' instruction is used to read from memory.</p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]1387<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1388<p>The argument to the '<tt>load</tt>' instruction specifies the memory
1389address to load from. The pointer must point to a <a
1390 href="t_firstclass">first class</a> type. If the <tt>load</tt> is
1391marked as <tt>volatile</tt> then the optimizer is not allowed to modify
1392the number or order of execution of this <tt>load</tt> with other
1393volatile <tt>load</tt> and <tt><a href="#i_store">store</a></tt>
1394instructions. </p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]1395<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1396<p>The location of memory pointed to is loaded.</p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]1397<h5>Examples:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1398<pre> %ptr = <a href="#i_alloca">alloca</a> int <i>; yields {int*}:ptr</i>
1399 <a
1400 href="#i_store">store</a> int 3, int* %ptr <i>; yields {void}</i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]1401 %val = load int* %ptr <i>; yields {int}:val = int 3</i>
1402</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1403</div>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]1404<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1405<div class="doc_subsubsection"> <a name="i_store">'<tt>store</tt>'
1406Instruction</a> </div>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]1407<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1408<pre> store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt; <i>; yields {void}</i>
Chris Lattnerf0651072003-09-08 18:27:49 +0000[diff] [blame]1409 volatile store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt; <i>; yields {void}</i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]1410</pre>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]1411<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1412<p>The '<tt>store</tt>' instruction is used to write to memory.</p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]1413<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1414<p>There are two arguments to the '<tt>store</tt>' instruction: a value
1415to store and an address to store it into. The type of the '<tt>&lt;pointer&gt;</tt>'
1416operand must be a pointer to the type of the '<tt>&lt;value&gt;</tt>'
1417operand. If the <tt>store</tt> is marked as <tt>volatile</tt> then the
1418optimizer is not allowed to modify the number or order of execution of
1419this <tt>store</tt> with other volatile <tt>load</tt> and <tt><a
1420 href="#i_store">store</a></tt> instructions.</p>
1421<h5>Semantics:</h5>
1422<p>The contents of memory are updated to contain '<tt>&lt;value&gt;</tt>'
1423at the location specified by the '<tt>&lt;pointer&gt;</tt>' operand.</p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]1424<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1425<pre> %ptr = <a href="#i_alloca">alloca</a> int <i>; yields {int*}:ptr</i>
1426 <a
1427 href="#i_store">store</a> int 3, int* %ptr <i>; yields {void}</i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]1428 %val = load int* %ptr <i>; yields {int}:val = int 3</i>
1429</pre>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]1430<!-- _______________________________________________________________________ -->
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]1431<div class="doc_subsubsection">
1432 <a name="i_getelementptr">'<tt>getelementptr</tt>' Instruction</a>
1433</div>
1434
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1435<div class="doc_text">
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1436<h5>Syntax:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]1437<pre>
1438 &lt;result&gt; = getelementptr &lt;ty&gt;* &lt;ptrval&gt;{, &lt;ty&gt; &lt;idx&gt;}*
1439</pre>
1440
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1441<h5>Overview:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]1442
1443<p>
1444The '<tt>getelementptr</tt>' instruction is used to get the address of a
1445subelement of an aggregate data structure.</p>
1446
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1447<h5>Arguments:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]1448
1449<p>This instruction takes a list of integer constants that indicate what
1450elements of the aggregate object to index to. The actual types of the arguments
1451provided depend on the type of the first pointer argument. The
1452'<tt>getelementptr</tt>' instruction is used to index down through the type
1453levels of a structure. When indexing into a structure, only <tt>uint</tt>
1454integer constants are allowed. When indexing into an array or pointer
1455<tt>int</tt> and <tt>long</tt> indexes are allowed of any sign.</p>
1456
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1457<p>For example, let's consider a C code fragment and how it gets
1458compiled to LLVM:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]1459
1460<pre>
1461 struct RT {
1462 char A;
1463 int B[10][20];
1464 char C;
1465 };
1466 struct ST {
1467 int X;
1468 double Y;
1469 struct RT Z;
1470 };
1471
1472 int *foo(struct ST *s) {
1473 return &amp;s[1].Z.B[5][13];
1474 }
1475</pre>
1476
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1477<p>The LLVM code generated by the GCC frontend is:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]1478
1479<pre>
1480 %RT = type { sbyte, [10 x [20 x int]], sbyte }
1481 %ST = type { int, double, %RT }
1482
1483 int* "foo"(%ST* %s) {
1484 %reg = getelementptr %ST* %s, int 1, uint 2, uint 1, int 5, int 13<br>
1485 ret int* %reg
1486 }
1487</pre>
1488
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1489<h5>Semantics:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]1490
1491<p>The index types specified for the '<tt>getelementptr</tt>' instruction depend
1492on the pointer type that is being index into. <a href="t_pointer">Pointer</a>
1493and <a href="t_array">array</a> types require <tt>uint</tt>, <tt>int</tt>,
1494<tt>ulong</tt>, or <tt>long</tt> values, and <a href="t_struct">structure</a>
1495types require <tt>uint</tt> <b>constants</b>.</p>
1496
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1497<p>In the example above, the first index is indexing into the '<tt>%ST*</tt>'
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]1498type, which is a pointer, yielding a '<tt>%ST</tt>' = '<tt>{ int, double, %RT
1499}</tt>' type, a structure. The second index indexes into the third element of
1500the structure, yielding a '<tt>%RT</tt>' = '<tt>{ sbyte, [10 x [20 x int]],
1501sbyte }</tt>' type, another structure. The third index indexes into the second
1502element of the structure, yielding a '<tt>[10 x [20 x int]]</tt>' type, an
1503array. The two dimensions of the array are subscripted into, yielding an
1504'<tt>int</tt>' type. The '<tt>getelementptr</tt>' instruction return a pointer
1505to this element, thus computing a value of '<tt>int*</tt>' type.</p>
1506
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1507<p>Note that it is perfectly legal to index partially through a
1508structure, returning a pointer to an inner element. Because of this,
1509the LLVM code for the given testcase is equivalent to:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]1510
1511<pre>
1512 int* "foo"(%ST* %s) {
1513 %t1 = getelementptr %ST* %s, int 1 <i>; yields %ST*:%t1</i>
1514 %t2 = getelementptr %ST* %t1, int 0, uint 2 <i>; yields %RT*:%t2</i>
1515 %t3 = getelementptr %RT* %t2, int 0, uint 1 <i>; yields [10 x [20 x int]]*:%t3</i>
1516 %t4 = getelementptr [10 x [20 x int]]* %t3, int 0, int 5 <i>; yields [20 x int]*:%t4</i>
1517 %t5 = getelementptr [20 x int]* %t4, int 0, int 13 <i>; yields int*:%t5</i>
1518 ret int* %t5
1519 }
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]1520</pre>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1521<h5>Example:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]1522<pre>
1523 <i>; yields [12 x ubyte]*:aptr</i>
1524 %aptr = getelementptr {int, [12 x ubyte]}* %sptr, long 0, uint 1
1525</pre>
1526
1527</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1528<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1529<div class="doc_subsection"> <a name="otherops">Other Operations</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1530<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1531<p>The instructions in this catagory are the "miscellaneous"
1532instructions, which defy better classification.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1533</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1534<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1535<div class="doc_subsubsection"> <a name="i_phi">'<tt>phi</tt>'
1536Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1537<div class="doc_text">
Chris Lattner33ba0d92001-07-09 00:26:23 +0000[diff] [blame]1538<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1539<pre> &lt;result&gt; = phi &lt;ty&gt; [ &lt;val0&gt;, &lt;label0&gt;], ...<br></pre>
Chris Lattner33ba0d92001-07-09 00:26:23 +0000[diff] [blame]1540<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1541<p>The '<tt>phi</tt>' instruction is used to implement the &#966; node in
1542the SSA graph representing the function.</p>
Chris Lattner33ba0d92001-07-09 00:26:23 +0000[diff] [blame]1543<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1544<p>The type of the incoming values are specified with the first type
1545field. After this, the '<tt>phi</tt>' instruction takes a list of pairs
1546as arguments, with one pair for each predecessor basic block of the
1547current block. Only values of <a href="#t_firstclass">first class</a>
1548type may be used as the value arguments to the PHI node. Only labels
1549may be used as the label arguments.</p>
1550<p>There must be no non-phi instructions between the start of a basic
1551block and the PHI instructions: i.e. PHI instructions must be first in
1552a basic block.</p>
Chris Lattner33ba0d92001-07-09 00:26:23 +0000[diff] [blame]1553<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1554<p>At runtime, the '<tt>phi</tt>' instruction logically takes on the
1555value specified by the parameter, depending on which basic block we
1556came from in the last <a href="#terminators">terminator</a> instruction.</p>
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]1557<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1558<pre>Loop: ; Infinite loop that counts from 0 on up...<br> %indvar = phi uint [ 0, %LoopHeader ], [ %nextindvar, %Loop ]<br> %nextindvar = add uint %indvar, 1<br> br label %Loop<br></pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1559</div>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]1560
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]1561<!-- _______________________________________________________________________ -->
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]1562<div class="doc_subsubsection">
1563 <a name="i_cast">'<tt>cast .. to</tt>' Instruction</a>
1564</div>
1565
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1566<div class="doc_text">
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]1567
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]1568<h5>Syntax:</h5>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]1569
1570<pre>
1571 &lt;result&gt; = cast &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]1572</pre>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]1573
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]1574<h5>Overview:</h5>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]1575
1576<p>
1577The '<tt>cast</tt>' instruction is used as the primitive means to convert
1578integers to floating point, change data type sizes, and break type safety (by
1579casting pointers).
1580</p>
1581
1582
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]1583<h5>Arguments:</h5>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]1584
1585<p>
1586The '<tt>cast</tt>' instruction takes a value to cast, which must be a first
1587class value, and a type to cast it to, which must also be a <a
1588href="#t_firstclass">first class</a> type.
1589</p>
1590
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]1591<h5>Semantics:</h5>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]1592
1593<p>
1594This instruction follows the C rules for explicit casts when determining how the
1595data being cast must change to fit in its new container.
1596</p>
1597
1598<p>
1599When casting to bool, any value that would be considered true in the context of
1600a C '<tt>if</tt>' condition is converted to the boolean '<tt>true</tt>' values,
1601all else are '<tt>false</tt>'.
1602</p>
1603
1604<p>
1605When extending an integral value from a type of one signness to another (for
1606example '<tt>sbyte</tt>' to '<tt>ulong</tt>'), the value is sign-extended if the
1607<b>source</b> value is signed, and zero-extended if the source value is
1608unsigned. <tt>bool</tt> values are always zero extended into either zero or
1609one.
1610</p>
1611
Chris Lattner33ba0d92001-07-09 00:26:23 +0000[diff] [blame]1612<h5>Example:</h5>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]1613
1614<pre>
1615 %X = cast int 257 to ubyte <i>; yields ubyte:1</i>
Chris Lattner7bae3952002-06-25 18:03:17 +0000[diff] [blame]1616 %Y = cast int 123 to bool <i>; yields bool:true</i>
Chris Lattner33ba0d92001-07-09 00:26:23 +0000[diff] [blame]1617</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1618</div>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]1619
1620<!-- _______________________________________________________________________ -->
1621<div class="doc_subsubsection">
1622 <a name="i_select">'<tt>select</tt>' Instruction</a>
1623</div>
1624
1625<div class="doc_text">
1626
1627<h5>Syntax:</h5>
1628
1629<pre>
1630 &lt;result&gt; = select bool &lt;cond&gt;, &lt;ty&gt; &lt;val1&gt;, &lt;ty&gt; &lt;val2&gt; <i>; yields ty</i>
1631</pre>
1632
1633<h5>Overview:</h5>
1634
1635<p>
1636The '<tt>select</tt>' instruction is used to choose one value based on a
1637condition, without branching.
1638</p>
1639
1640
1641<h5>Arguments:</h5>
1642
1643<p>
1644The '<tt>select</tt>' instruction requires a boolean value indicating the condition, and two values of the same <a href="#t_firstclass">first class</a> type.
1645</p>
1646
1647<h5>Semantics:</h5>
1648
1649<p>
1650If the boolean condition evaluates to true, the instruction returns the first
1651value argument, otherwise it returns the second value argument.
1652</p>
1653
1654<h5>Example:</h5>
1655
1656<pre>
1657 %X = select bool true, ubyte 17, ubyte 42 <i>; yields ubyte:17</i>
1658</pre>
1659</div>
1660
1661
1662
1663
1664
Chris Lattner33ba0d92001-07-09 00:26:23 +0000[diff] [blame]1665<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1666<div class="doc_subsubsection"> <a name="i_call">'<tt>call</tt>'
1667Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1668<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1669<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1670<pre> &lt;result&gt; = call &lt;ty&gt;* &lt;fnptrval&gt;(&lt;param list&gt;)<br></pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1671<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1672<p>The '<tt>call</tt>' instruction represents a simple function call.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1673<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1674<p>This instruction requires several arguments:</p>
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]1675<ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1676 <li>
1677 <p>'<tt>ty</tt>': shall be the signature of the pointer to function
1678value being invoked. The argument types must match the types implied
1679by this signature.</p>
1680 </li>
1681 <li>
1682 <p>'<tt>fnptrval</tt>': An LLVM value containing a pointer to a
1683function to be invoked. In most cases, this is a direct function
1684invocation, but indirect <tt>call</tt>s are just as possible,
1685calling an arbitrary pointer to function values.</p>
1686 </li>
1687 <li>
1688 <p>'<tt>function args</tt>': argument list whose types match the
1689function signature argument types. If the function signature
1690indicates the function accepts a variable number of arguments, the
1691extra arguments can be specified.</p>
1692 </li>
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]1693</ol>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1694<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1695<p>The '<tt>call</tt>' instruction is used to cause control flow to
1696transfer to a specified function, with its incoming arguments bound to
1697the specified values. Upon a '<tt><a href="#i_ret">ret</a></tt>'
1698instruction in the called function, control flow continues with the
1699instruction after the function call, and the return value of the
1700function is bound to the result argument. This is a simpler case of
1701the <a href="#i_invoke">invoke</a> instruction.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1702<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1703<pre> %retval = call int %test(int %argc)<br> call int(sbyte*, ...) *%printf(sbyte* %msg, int 12, sbyte 42);<br></pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1704</div>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]1705<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1706<div class="doc_subsubsection"> <a name="i_vanext">'<tt>vanext</tt>'
1707Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1708<div class="doc_text">
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]1709<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1710<pre> &lt;resultarglist&gt; = vanext &lt;va_list&gt; &lt;arglist&gt;, &lt;argty&gt;<br></pre>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]1711<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1712<p>The '<tt>vanext</tt>' instruction is used to access arguments passed
1713through the "variable argument" area of a function call. It is used to
1714implement the <tt>va_arg</tt> macro in C.</p>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]1715<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1716<p>This instruction takes a <tt>valist</tt> value and the type of the
1717argument. It returns another <tt>valist</tt>.</p>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]1718<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1719<p>The '<tt>vanext</tt>' instruction advances the specified <tt>valist</tt>
1720past an argument of the specified type. In conjunction with the <a
1721 href="#i_vaarg"><tt>vaarg</tt></a> instruction, it is used to implement
1722the <tt>va_arg</tt> macro available in C. For more information, see
1723the variable argument handling <a href="#int_varargs">Intrinsic
1724Functions</a>.</p>
1725<p>It is legal for this instruction to be called in a function which
1726does not take a variable number of arguments, for example, the <tt>vfprintf</tt>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1727function.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1728<p><tt>vanext</tt> is an LLVM instruction instead of an <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1729 href="#intrinsics">intrinsic function</a> because it takes an type as
1730an argument.</p>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]1731<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1732<p>See the <a href="#int_varargs">variable argument processing</a>
1733section.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1734</div>
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]1735<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1736<div class="doc_subsubsection"> <a name="i_vaarg">'<tt>vaarg</tt>'
1737Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1738<div class="doc_text">
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]1739<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1740<pre> &lt;resultval&gt; = vaarg &lt;va_list&gt; &lt;arglist&gt;, &lt;argty&gt;<br></pre>
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]1741<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1742<p>The '<tt>vaarg</tt>' instruction is used to access arguments passed
1743through the "variable argument" area of a function call. It is used to
1744implement the <tt>va_arg</tt> macro in C.</p>
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]1745<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1746<p>This instruction takes a <tt>valist</tt> value and the type of the
1747argument. It returns a value of the specified argument type.</p>
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]1748<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1749<p>The '<tt>vaarg</tt>' instruction loads an argument of the specified
1750type from the specified <tt>va_list</tt>. In conjunction with the <a
1751 href="#i_vanext"><tt>vanext</tt></a> instruction, it is used to
1752implement the <tt>va_arg</tt> macro available in C. For more
1753information, see the variable argument handling <a href="#int_varargs">Intrinsic
1754Functions</a>.</p>
1755<p>It is legal for this instruction to be called in a function which
1756does not take a variable number of arguments, for example, the <tt>vfprintf</tt>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1757function.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1758<p><tt>vaarg</tt> is an LLVM instruction instead of an <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1759 href="#intrinsics">intrinsic function</a> because it takes an type as
1760an argument.</p>
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]1761<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1762<p>See the <a href="#int_varargs">variable argument processing</a>
1763section.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1764</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]1765
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]1766<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1767<div class="doc_section"> <a name="intrinsics">Intrinsic Functions</a> </div>
1768<!-- *********************************************************************** -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]1769
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1770<div class="doc_text">
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]1771
1772<p>LLVM supports the notion of an "intrinsic function". These functions have
1773well known names and semantics, and are required to follow certain
1774restrictions. Overall, these instructions represent an extension mechanism for
1775the LLVM language that does not require changing all of the transformations in
1776LLVM to add to the language (or the bytecode reader/writer, the parser,
1777etc...).</p>
1778
1779<p>Intrinsic function names must all start with an "<tt>llvm.</tt>" prefix, this
1780prefix is reserved in LLVM for intrinsic names, thus functions may not be named
1781this. Intrinsic functions must always be external functions: you cannot define
1782the body of intrinsic functions. Intrinsic functions may only be used in call
1783or invoke instructions: it is illegal to take the address of an intrinsic
1784function. Additionally, because intrinsic functions are part of the LLVM
1785language, it is required that they all be documented here if any are added.</p>
1786
1787
1788<p>
1789Adding an intrinsic to LLVM is straight-forward if it is possible to express the
1790concept in LLVM directly (ie, code generator support is not _required_). To do
1791this, extend the default implementation of the IntrinsicLowering class to handle
1792the intrinsic. Code generators use this class to lower intrinsics they do not
1793understand to raw LLVM instructions that they do.
1794</p>
1795
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1796</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]1797
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]1798<!-- ======================================================================= -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]1799<div class="doc_subsection">
1800 <a name="int_varargs">Variable Argument Handling Intrinsics</a>
1801</div>
1802
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1803<div class="doc_text">
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1804<p>Variable argument support is defined in LLVM with the <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1805 href="#i_vanext"><tt>vanext</tt></a> instruction and these three
1806intrinsic functions. These functions are related to the similarly
1807named macros defined in the <tt>&lt;stdarg.h&gt;</tt> header file.</p>
1808<p>All of these functions operate on arguments that use a
1809target-specific value type "<tt>va_list</tt>". The LLVM assembly
1810language reference manual does not define what this type is, so all
1811transformations should be prepared to handle intrinsics with any type
1812used.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1813<p>This example shows how the <a href="#i_vanext"><tt>vanext</tt></a>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1814instruction and the variable argument handling intrinsic functions are
1815used.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]1816<pre>
1817int %test(int %X, ...) {
1818 ; Initialize variable argument processing
1819 %ap = call sbyte* %<a href="#i_va_start">llvm.va_start</a>()
1820
1821 ; Read a single integer argument
1822 %tmp = vaarg sbyte* %ap, int
1823
1824 ; Advance to the next argument
1825 %ap2 = vanext sbyte* %ap, int
1826
1827 ; Demonstrate usage of llvm.va_copy and llvm.va_end
1828 %aq = call sbyte* %<a href="#i_va_copy">llvm.va_copy</a>(sbyte* %ap2)
1829 call void %<a href="#i_va_end">llvm.va_end</a>(sbyte* %aq)
1830
1831 ; Stop processing of arguments.
1832 call void %<a href="#i_va_end">llvm.va_end</a>(sbyte* %ap2)
1833 ret int %tmp
1834}
1835</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1836</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]1837
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]1838<!-- _______________________________________________________________________ -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]1839<div class="doc_subsubsection">
1840 <a name="i_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a>
1841</div>
1842
1843
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1844<div class="doc_text">
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]1845<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1846<pre> call va_list ()* %llvm.va_start()<br></pre>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]1847<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1848<p>The '<tt>llvm.va_start</tt>' intrinsic returns a new <tt>&lt;arglist&gt;</tt>
1849for subsequent use by the variable argument intrinsics.</p>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]1850<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1851<p>The '<tt>llvm.va_start</tt>' intrinsic works just like the <tt>va_start</tt>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1852macro available in C. In a target-dependent way, it initializes and
1853returns a <tt>va_list</tt> element, so that the next <tt>vaarg</tt>
1854will produce the first variable argument passed to the function. Unlike
1855the C <tt>va_start</tt> macro, this intrinsic does not need to know the
1856last argument of the function, the compiler can figure that out.</p>
1857<p>Note that this intrinsic function is only legal to be called from
1858within the body of a variable argument function.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1859</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]1860
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]1861<!-- _______________________________________________________________________ -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]1862<div class="doc_subsubsection">
1863 <a name="i_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a>
1864</div>
1865
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1866<div class="doc_text">
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]1867<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1868<pre> call void (va_list)* %llvm.va_end(va_list &lt;arglist&gt;)<br></pre>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]1869<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1870<p>The '<tt>llvm.va_end</tt>' intrinsic destroys <tt>&lt;arglist&gt;</tt>
1871which has been initialized previously with <tt><a href="#i_va_start">llvm.va_start</a></tt>
1872or <tt><a href="#i_va_copy">llvm.va_copy</a></tt>.</p>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]1873<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1874<p>The argument is a <tt>va_list</tt> to destroy.</p>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]1875<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1876<p>The '<tt>llvm.va_end</tt>' intrinsic works just like the <tt>va_end</tt>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1877macro available in C. In a target-dependent way, it destroys the <tt>va_list</tt>.
1878Calls to <a href="#i_va_start"><tt>llvm.va_start</tt></a> and <a
1879 href="#i_va_copy"><tt>llvm.va_copy</tt></a> must be matched exactly
1880with calls to <tt>llvm.va_end</tt>.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1881</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]1882
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]1883<!-- _______________________________________________________________________ -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]1884<div class="doc_subsubsection">
1885 <a name="i_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a>
1886</div>
1887
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1888<div class="doc_text">
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]1889<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1890<pre> call va_list (va_list)* %llvm.va_copy(va_list &lt;destarglist&gt;)<br></pre>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]1891<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1892<p>The '<tt>llvm.va_copy</tt>' intrinsic copies the current argument
1893position from the source argument list to the destination argument list.</p>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]1894<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1895<p>The argument is the <tt>va_list</tt> to copy.</p>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]1896<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1897<p>The '<tt>llvm.va_copy</tt>' intrinsic works just like the <tt>va_copy</tt>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1898macro available in C. In a target-dependent way, it copies the source <tt>va_list</tt>
1899element into the returned list. This intrinsic is necessary because the <tt><a
1900 href="i_va_start">llvm.va_start</a></tt> intrinsic may be arbitrarily
1901complex and require memory allocation, for example.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1902</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]1903
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]1904<!-- ======================================================================= -->
1905<div class="doc_subsection">
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]1906 <a name="int_codegen">Code Generator Intrinsics</a>
1907</div>
1908
1909<div class="doc_text">
1910<p>
1911These intrinsics are provided by LLVM to expose special features that may only
1912be implemented with code generator support.
1913</p>
1914
1915</div>
1916
1917<!-- _______________________________________________________________________ -->
1918<div class="doc_subsubsection">
1919 <a name="i_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a>
1920</div>
1921
1922<div class="doc_text">
1923
1924<h5>Syntax:</h5>
1925<pre>
1926 call void* ()* %llvm.returnaddress(uint &lt;level&gt;)
1927</pre>
1928
1929<h5>Overview:</h5>
1930
1931<p>
1932The '<tt>llvm.returnaddress</tt>' intrinsic returns a target-specific value
1933indicating the return address of the current function or one of its callers.
1934</p>
1935
1936<h5>Arguments:</h5>
1937
1938<p>
1939The argument to this intrinsic indicates which function to return the address
1940for. Zero indicates the calling function, one indicates its caller, etc. The
1941argument is <b>required</b> to be a constant integer value.
1942</p>
1943
1944<h5>Semantics:</h5>
1945
1946<p>
1947The '<tt>llvm.returnaddress</tt>' intrinsic either returns a pointer indicating
1948the return address of the specified call frame, or zero if it cannot be
1949identified. The value returned by this intrinsic is likely to be incorrect or 0
1950for arguments other than zero, so it should only be used for debugging purposes.
1951</p>
1952
1953<p>
1954Note that calling this intrinsic does not prevent function inlining or other
1955aggressive transformations, so the value returned may not that of the obvious
1956source-language caller.
1957</p>
1958</div>
1959
1960
1961<!-- _______________________________________________________________________ -->
1962<div class="doc_subsubsection">
1963 <a name="i_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a>
1964</div>
1965
1966<div class="doc_text">
1967
1968<h5>Syntax:</h5>
1969<pre>
1970 call void* ()* %llvm.frameaddress(uint &lt;level&gt;)
1971</pre>
1972
1973<h5>Overview:</h5>
1974
1975<p>
1976The '<tt>llvm.frameaddress</tt>' intrinsic returns the target-specific frame
1977pointer value for the specified stack frame.
1978</p>
1979
1980<h5>Arguments:</h5>
1981
1982<p>
1983The argument to this intrinsic indicates which function to return the frame
1984pointer for. Zero indicates the calling function, one indicates its caller,
1985etc. The argument is <b>required</b> to be a constant integer value.
1986</p>
1987
1988<h5>Semantics:</h5>
1989
1990<p>
1991The '<tt>llvm.frameaddress</tt>' intrinsic either returns a pointer indicating
1992the frame address of the specified call frame, or zero if it cannot be
1993identified. The value returned by this intrinsic is likely to be incorrect or 0
1994for arguments other than zero, so it should only be used for debugging purposes.
1995</p>
1996
1997<p>
1998Note that calling this intrinsic does not prevent function inlining or other
1999aggressive transformations, so the value returned may not that of the obvious
2000source-language caller.
2001</p>
2002</div>
2003
John Criswell7123e272004-04-09 16:43:20 +0000[diff] [blame^]2004<!-- ======================================================================= -->
2005<div class="doc_subsection">
2006 <a name="int_os">Operating System Intrinsics</a>
2007</div>
2008
2009<div class="doc_text">
2010<p>
2011These intrinsics are provided by LLVM to support the implementation of
2012operating system level code.
2013</p>
2014
2015</div>
John Criswellcfd3bac2004-04-09 15:23:37 +0000[diff] [blame]2016<!-- _______________________________________________________________________ -->
2017<div class="doc_subsubsection">
2018 <a name="i_readport">'<tt>llvm.readport</tt>' Intrinsic</a>
2019</div>
2020
2021<div class="doc_text">
2022
2023<h5>Syntax:</h5>
2024<pre>
John Criswell7123e272004-04-09 16:43:20 +0000[diff] [blame^]2025 call &lt;integer type&gt; (&lt;integer type&gt;)* %llvm.readport (&lt;integer type&gt; &lt;address&gt;)
John Criswellcfd3bac2004-04-09 15:23:37 +0000[diff] [blame]2026</pre>
2027
2028<h5>Overview:</h5>
2029
2030<p>
John Criswell7123e272004-04-09 16:43:20 +0000[diff] [blame^]2031The '<tt>llvm.readport</tt>' intrinsic reads data from the specified hardware
2032I/O port.
John Criswellcfd3bac2004-04-09 15:23:37 +0000[diff] [blame]2033</p>
2034
2035<h5>Arguments:</h5>
2036
2037<p>
John Criswell7123e272004-04-09 16:43:20 +0000[diff] [blame^]2038The argument to this intrinsic indicates the hardware I/O address from which
2039to read the data. The address is in the hardware I/O address namespace (as
2040opposed to being a memory location for memory mapped I/O).
John Criswellcfd3bac2004-04-09 15:23:37 +0000[diff] [blame]2041</p>
2042
2043<h5>Semantics:</h5>
2044
2045<p>
John Criswell7123e272004-04-09 16:43:20 +0000[diff] [blame^]2046The '<tt>llvm.readport</tt>' intrinsic reads data from the hardware I/O port
2047specified by <i>address</i> and returns the value. The address and return
2048value must be integers, but the size is dependent upon the platform upon which
2049the program is code generated. For example, on x86, the address must be an
2050unsigned 16 bit value, and the return value must be 8, 16, or 32 bits.
John Criswellcfd3bac2004-04-09 15:23:37 +0000[diff] [blame]2051</p>
2052
2053</div>
2054
2055<!-- _______________________________________________________________________ -->
2056<div class="doc_subsubsection">
2057 <a name="i_writeport">'<tt>llvm.writeport</tt>' Intrinsic</a>
2058</div>
2059
2060<div class="doc_text">
2061
2062<h5>Syntax:</h5>
2063<pre>
John Criswell7123e272004-04-09 16:43:20 +0000[diff] [blame^]2064 call void (&lt;integer type&gt;, &lt;integer type&gt;)* %llvm.writeport (&lt;integer type&gt; &lt;value&gt;, &lt;integer type&gt; &lt;address&gt;)
John Criswellcfd3bac2004-04-09 15:23:37 +0000[diff] [blame]2065</pre>
2066
2067<h5>Overview:</h5>
2068
2069<p>
John Criswell7123e272004-04-09 16:43:20 +0000[diff] [blame^]2070The '<tt>llvm.writeport</tt>' intrinsic writes data to the specified hardware
2071I/O port.
John Criswellcfd3bac2004-04-09 15:23:37 +0000[diff] [blame]2072</p>
2073
2074<h5>Arguments:</h5>
2075
2076<p>
John Criswell7123e272004-04-09 16:43:20 +0000[diff] [blame^]2077The first argument to this intrinsic indicates the hardware I/O address to
2078which data should be written. The address is in the hardware I/O address
2079namespace (as opposed to being a memory location for memory mapped I/O).
John Criswellcfd3bac2004-04-09 15:23:37 +0000[diff] [blame]2080</p>
2081
2082<p>
2083The second argument is the value to write to the I/O port.
2084</p>
2085
2086<h5>Semantics:</h5>
2087
2088<p>
2089The '<tt>llvm.writeport</tt>' intrinsic writes <i>value</i> to the I/O port
2090specified by <i>address</i>. The address and value must be integers, but the
2091size is dependent upon the platform upon which the program is code generated.
John Criswell7123e272004-04-09 16:43:20 +0000[diff] [blame^]2092For example, on x86, the address must be an unsigned 16 bit value, and the
2093value written must be 8, 16, or 32 bits in length.
John Criswellcfd3bac2004-04-09 15:23:37 +0000[diff] [blame]2094</p>
2095
2096</div>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]2097
2098<!-- ======================================================================= -->
2099<div class="doc_subsection">
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]2100 <a name="int_libc">Standard C Library Intrinsics</a>
2101</div>
2102
2103<div class="doc_text">
2104<p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]2105LLVM provides intrinsics for a few important standard C library functions.
2106These intrinsics allow source-language front-ends to pass information about the
2107alignment of the pointer arguments to the code generator, providing opportunity
2108for more efficient code generation.
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]2109</p>
2110
2111</div>
2112
2113<!-- _______________________________________________________________________ -->
2114<div class="doc_subsubsection">
2115 <a name="i_memcpy">'<tt>llvm.memcpy</tt>' Intrinsic</a>
2116</div>
2117
2118<div class="doc_text">
2119
2120<h5>Syntax:</h5>
2121<pre>
2122 call void (sbyte*, sbyte*, uint, uint)* %llvm.memcpy(sbyte* &lt;dest&gt;, sbyte* &lt;src&gt;,
2123 uint &lt;len&gt;, uint &lt;align&gt;)
2124</pre>
2125
2126<h5>Overview:</h5>
2127
2128<p>
2129The '<tt>llvm.memcpy</tt>' intrinsic copies a block of memory from the source
2130location to the destination location.
2131</p>
2132
2133<p>
2134Note that, unlike the standard libc function, the <tt>llvm.memcpy</tt> intrinsic
2135does not return a value, and takes an extra alignment argument.
2136</p>
2137
2138<h5>Arguments:</h5>
2139
2140<p>
2141The first argument is a pointer to the destination, the second is a pointer to
2142the source. The third argument is an (arbitrarily sized) integer argument
2143specifying the number of bytes to copy, and the fourth argument is the alignment
2144of the source and destination locations.
2145</p>
2146
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]2147<p>
2148If the call to this intrinisic has an alignment value that is not 0 or 1, then
2149the caller guarantees that the size of the copy is a multiple of the alignment
2150and that both the source and destination pointers are aligned to that boundary.
2151</p>
2152
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]2153<h5>Semantics:</h5>
2154
2155<p>
2156The '<tt>llvm.memcpy</tt>' intrinsic copies a block of memory from the source
2157location to the destination location, which are not allowed to overlap. It
2158copies "len" bytes of memory over. If the argument is known to be aligned to
2159some boundary, this can be specified as the fourth argument, otherwise it should
2160be set to 0 or 1.
2161</p>
2162</div>
2163
2164
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]2165<!-- _______________________________________________________________________ -->
2166<div class="doc_subsubsection">
2167 <a name="i_memmove">'<tt>llvm.memmove</tt>' Intrinsic</a>
2168</div>
2169
2170<div class="doc_text">
2171
2172<h5>Syntax:</h5>
2173<pre>
2174 call void (sbyte*, sbyte*, uint, uint)* %llvm.memmove(sbyte* &lt;dest&gt;, sbyte* &lt;src&gt;,
2175 uint &lt;len&gt;, uint &lt;align&gt;)
2176</pre>
2177
2178<h5>Overview:</h5>
2179
2180<p>
2181The '<tt>llvm.memmove</tt>' intrinsic moves a block of memory from the source
2182location to the destination location. It is similar to the '<tt>llvm.memcpy</tt>'
2183intrinsic but allows the two memory locations to overlap.
2184</p>
2185
2186<p>
2187Note that, unlike the standard libc function, the <tt>llvm.memmove</tt> intrinsic
2188does not return a value, and takes an extra alignment argument.
2189</p>
2190
2191<h5>Arguments:</h5>
2192
2193<p>
2194The first argument is a pointer to the destination, the second is a pointer to
2195the source. The third argument is an (arbitrarily sized) integer argument
2196specifying the number of bytes to copy, and the fourth argument is the alignment
2197of the source and destination locations.
2198</p>
2199
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]2200<p>
2201If the call to this intrinisic has an alignment value that is not 0 or 1, then
2202the caller guarantees that the size of the copy is a multiple of the alignment
2203and that both the source and destination pointers are aligned to that boundary.
2204</p>
2205
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]2206<h5>Semantics:</h5>
2207
2208<p>
2209The '<tt>llvm.memmove</tt>' intrinsic copies a block of memory from the source
2210location to the destination location, which may overlap. It
2211copies "len" bytes of memory over. If the argument is known to be aligned to
2212some boundary, this can be specified as the fourth argument, otherwise it should
2213be set to 0 or 1.
2214</p>
2215</div>
2216
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]2217
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]2218<!-- _______________________________________________________________________ -->
2219<div class="doc_subsubsection">
2220 <a name="i_memset">'<tt>llvm.memset</tt>' Intrinsic</a>
2221</div>
2222
2223<div class="doc_text">
2224
2225<h5>Syntax:</h5>
2226<pre>
2227 call void (sbyte*, ubyte, uint, uint)* %llvm.memset(sbyte* &lt;dest&gt;, ubyte &lt;val&gt;,
2228 uint &lt;len&gt;, uint &lt;align&gt;)
2229</pre>
2230
2231<h5>Overview:</h5>
2232
2233<p>
2234The '<tt>llvm.memset</tt>' intrinsic fills a block of memory with a particular
2235byte value.
2236</p>
2237
2238<p>
2239Note that, unlike the standard libc function, the <tt>llvm.memset</tt> intrinsic
2240does not return a value, and takes an extra alignment argument.
2241</p>
2242
2243<h5>Arguments:</h5>
2244
2245<p>
2246The first argument is a pointer to the destination to fill, the second is the
2247byte value to fill it with, the third argument is an (arbitrarily sized) integer
2248argument specifying the number of bytes to fill, and the fourth argument is the
2249known alignment of destination location.
2250</p>
2251
2252<p>
2253If the call to this intrinisic has an alignment value that is not 0 or 1, then
2254the caller guarantees that the size of the copy is a multiple of the alignment
2255and that the destination pointer is aligned to that boundary.
2256</p>
2257
2258<h5>Semantics:</h5>
2259
2260<p>
2261The '<tt>llvm.memset</tt>' intrinsic fills "len" bytes of memory starting at the
2262destination location. If the argument is known to be aligned to some boundary,
2263this can be specified as the fourth argument, otherwise it should be set to 0 or
22641.
2265</p>
2266</div>
2267
2268
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]2269<!-- ======================================================================= -->
2270<div class="doc_subsection">
2271 <a name="int_debugger">Debugger Intrinsics</a>
2272</div>
2273
2274<div class="doc_text">
2275<p>
2276The LLVM debugger intrinsics (which all start with <tt>llvm.dbg.</tt> prefix),
2277are described in the <a
2278href="SourceLevelDebugging.html#format_common_intrinsics">LLVM Source Level
2279Debugging</a> document.
2280</p>
2281</div>
2282
2283
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2284<!-- *********************************************************************** -->
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2285<hr>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]2286<address>
2287 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
2288 src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
2289 <a href="http://validator.w3.org/check/referer"><img
2290 src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!" /></a>
2291
2292 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
2293 <a href="http://llvm.cs.uiuc.edu">The LLVM Compiler Infrastructure</a><br>
2294 Last modified: $Date$
2295</address>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2296</body>
2297</html>