Gitiles
Code Review Sign In
gerrit-public.fairphone.software / fp2-dev / platform / external / llvm / 3fa91b08a4ced58f520cb07d51316256d831893b / . / docs / LangRef.html
blob: 8c045000fec0d53b2af679283fe37a8e0b01ce8f [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>
Reid Spencer3921c742004-08-26 20:44:00 +0000[diff] [blame]6 <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
7 <meta name="author" content="Chris Lattner">
8 <meta name="description"
9 content="LLVM Assembly Language Reference Manual.">
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]10 <link rel="stylesheet" href="llvm.css" type="text/css">
11</head>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]12
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]13<body>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]14
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]15<div class="doc_title"> LLVM Language Reference Manual </div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]16<ol>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]17 <li><a href="#abstract">Abstract</a></li>
18 <li><a href="#introduction">Introduction</a></li>
19 <li><a href="#identifiers">Identifiers</a></li>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]20 <li><a href="#highlevel">High Level Structure</a>
21 <ol>
22 <li><a href="#modulestructure">Module Structure</a></li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]23 <li><a href="#linkage">Linkage Types</a></li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]24 <li><a href="#callingconv">Calling Conventions</a></li>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]25 <li><a href="#globalvars">Global Variables</a></li>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]26 <li><a href="#functionstructure">Functions</a></li>
27 <li><a href="#moduleasm">Module-Level Inline Assembly</a></li>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]28 </ol>
29 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]30 <li><a href="#typesystem">Type System</a>
31 <ol>
Robert Bocchino7b81c752006-02-17 21:18:08 +0000[diff] [blame]32 <li><a href="#t_primitive">Primitive Types</a>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]33 <ol>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]34 <li><a href="#t_classifications">Type Classifications</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]35 </ol>
36 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]37 <li><a href="#t_derived">Derived Types</a>
38 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]39 <li><a href="#t_array">Array Type</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]40 <li><a href="#t_function">Function Type</a></li>
41 <li><a href="#t_pointer">Pointer Type</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]42 <li><a href="#t_struct">Structure Type</a></li>
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]43 <li><a href="#t_packed">Packed Type</a></li>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]44 <li><a href="#t_opaque">Opaque Type</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]45 </ol>
46 </li>
47 </ol>
48 </li>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]49 <li><a href="#constants">Constants</a>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]50 <ol>
51 <li><a href="#simpleconstants">Simple Constants</a>
52 <li><a href="#aggregateconstants">Aggregate Constants</a>
53 <li><a href="#globalconstants">Global Variable and Function Addresses</a>
54 <li><a href="#undefvalues">Undefined Values</a>
55 <li><a href="#constantexprs">Constant Expressions</a>
56 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]57 </li>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]58 <li><a href="#othervalues">Other Values</a>
59 <ol>
60 <li><a href="#inlineasm">Inline Assembler Expressions</a>
61 </ol>
62 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]63 <li><a href="#instref">Instruction Reference</a>
64 <ol>
65 <li><a href="#terminators">Terminator Instructions</a>
66 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]67 <li><a href="#i_ret">'<tt>ret</tt>' Instruction</a></li>
68 <li><a href="#i_br">'<tt>br</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]69 <li><a href="#i_switch">'<tt>switch</tt>' Instruction</a></li>
70 <li><a href="#i_invoke">'<tt>invoke</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]71 <li><a href="#i_unwind">'<tt>unwind</tt>' Instruction</a></li>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]72 <li><a href="#i_unreachable">'<tt>unreachable</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]73 </ol>
74 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]75 <li><a href="#binaryops">Binary Operations</a>
76 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]77 <li><a href="#i_add">'<tt>add</tt>' Instruction</a></li>
78 <li><a href="#i_sub">'<tt>sub</tt>' Instruction</a></li>
79 <li><a href="#i_mul">'<tt>mul</tt>' Instruction</a></li>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]80 <li><a href="#i_udiv">'<tt>udiv</tt>' Instruction</a></li>
81 <li><a href="#i_sdiv">'<tt>sdiv</tt>' Instruction</a></li>
82 <li><a href="#i_fdiv">'<tt>fdiv</tt>' Instruction</a></li>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]83 <li><a href="#i_urem">'<tt>urem</tt>' Instruction</a></li>
84 <li><a href="#i_srem">'<tt>srem</tt>' Instruction</a></li>
85 <li><a href="#i_frem">'<tt>frem</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]86 <li><a href="#i_setcc">'<tt>set<i>cc</i></tt>' Instructions</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]87 </ol>
88 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]89 <li><a href="#bitwiseops">Bitwise Binary Operations</a>
90 <ol>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]91 <li><a href="#i_and">'<tt>and</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]92 <li><a href="#i_or">'<tt>or</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]93 <li><a href="#i_xor">'<tt>xor</tt>' Instruction</a></li>
94 <li><a href="#i_shl">'<tt>shl</tt>' Instruction</a></li>
Reid Spencer3822ff52006-11-08 06:47:33 +0000[diff] [blame]95 <li><a href="#i_lshr">'<tt>lshr</tt>' Instruction</a></li>
96 <li><a href="#i_ashr">'<tt>ashr</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]97 </ol>
98 </li>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]99 <li><a href="#vectorops">Vector Operations</a>
100 <ol>
101 <li><a href="#i_extractelement">'<tt>extractelement</tt>' Instruction</a></li>
102 <li><a href="#i_insertelement">'<tt>insertelement</tt>' Instruction</a></li>
103 <li><a href="#i_shufflevector">'<tt>shufflevector</tt>' Instruction</a></li>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]104 </ol>
105 </li>
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]106 <li><a href="#memoryops">Memory Access and Addressing Operations</a>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]107 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]108 <li><a href="#i_malloc">'<tt>malloc</tt>' Instruction</a></li>
109 <li><a href="#i_free">'<tt>free</tt>' Instruction</a></li>
110 <li><a href="#i_alloca">'<tt>alloca</tt>' Instruction</a></li>
Robert Bocchino7b81c752006-02-17 21:18:08 +0000[diff] [blame]111 <li><a href="#i_load">'<tt>load</tt>' Instruction</a></li>
112 <li><a href="#i_store">'<tt>store</tt>' Instruction</a></li>
113 <li><a href="#i_getelementptr">'<tt>getelementptr</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]114 </ol>
115 </li>
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]116 <li><a href="#convertops">Conversion Operations</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]117 <ol>
118 <li><a href="#i_trunc">'<tt>trunc .. to</tt>' Instruction</a></li>
119 <li><a href="#i_zext">'<tt>zext .. to</tt>' Instruction</a></li>
120 <li><a href="#i_sext">'<tt>sext .. to</tt>' Instruction</a></li>
121 <li><a href="#i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a></li>
122 <li><a href="#i_fpext">'<tt>fpext .. to</tt>' Instruction</a></li>
123 <li><a href="#i_fp2uint">'<tt>fp2uint .. to</tt>' Instruction</a></li>
124 <li><a href="#i_fp2sint">'<tt>fp2sint .. to</tt>' Instruction</a></li>
125 <li><a href="#i_uint2fp">'<tt>uint2fp .. to</tt>' Instruction</a></li>
126 <li><a href="#i_sint2fp">'<tt>sint2fp .. to</tt>' Instruction</a></li>
127 <li><a href="#i_bitconvert">'<tt>bitconvert .. to</tt>' Instruction</a></li>
128 </ol>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]129 <li><a href="#otherops">Other Operations</a>
130 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]131 <li><a href="#i_phi">'<tt>phi</tt>' Instruction</a></li>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]132 <li><a href="#i_select">'<tt>select</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]133 <li><a href="#i_call">'<tt>call</tt>' Instruction</a></li>
Chris Lattnerfb6977d2006-01-13 23:26:01 +0000[diff] [blame]134 <li><a href="#i_va_arg">'<tt>va_arg</tt>' Instruction</a></li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]135 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]136 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]137 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]138 </li>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]139 <li><a href="#intrinsics">Intrinsic Functions</a>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]140 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]141 <li><a href="#int_varargs">Variable Argument Handling Intrinsics</a>
142 <ol>
143 <li><a href="#i_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a></li>
144 <li><a href="#i_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a></li>
145 <li><a href="#i_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a></li>
146 </ol>
147 </li>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]148 <li><a href="#int_gc">Accurate Garbage Collection Intrinsics</a>
149 <ol>
150 <li><a href="#i_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a></li>
151 <li><a href="#i_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a></li>
152 <li><a href="#i_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a></li>
153 </ol>
154 </li>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]155 <li><a href="#int_codegen">Code Generator Intrinsics</a>
156 <ol>
157 <li><a href="#i_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a></li>
158 <li><a href="#i_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a></li>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]159 <li><a href="#i_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a></li>
160 <li><a href="#i_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a></li>
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]161 <li><a href="#i_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a></li>
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]162 <li><a href="#i_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a></li>
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]163 <li><a href="#i_readcyclecounter"><tt>llvm.readcyclecounter</tt>' Intrinsic</a></li>
John Criswell7123e272004-04-09 16:43:20 +0000[diff] [blame]164 </ol>
165 </li>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]166 <li><a href="#int_libc">Standard C Library Intrinsics</a>
167 <ol>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]168 <li><a href="#i_memcpy">'<tt>llvm.memcpy.*</tt>' Intrinsic</a></li>
169 <li><a href="#i_memmove">'<tt>llvm.memmove.*</tt>' Intrinsic</a></li>
170 <li><a href="#i_memset">'<tt>llvm.memset.*</tt>' Intrinsic</a></li>
Chris Lattnerec6cb612006-01-16 22:38:59 +0000[diff] [blame]171 <li><a href="#i_isunordered">'<tt>llvm.isunordered.*</tt>' Intrinsic</a></li>
172 <li><a href="#i_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a></li>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]173 <li><a href="#i_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a></li>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]174 </ol>
175 </li>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]176 <li><a href="#int_manip">Bit Manipulation Intrinsics</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]177 <ol>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]178 <li><a href="#i_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a></li>
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]179 <li><a href="#int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic </a></li>
180 <li><a href="#int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic </a></li>
181 <li><a href="#int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic </a></li>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]182 </ol>
183 </li>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]184 <li><a href="#int_debugger">Debugger intrinsics</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]185 </ol>
186 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]187</ol>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]188
189<div class="doc_author">
190 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>
191 and <a href="mailto:vadve@cs.uiuc.edu">Vikram Adve</a></p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]192</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]193
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]194<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]195<div class="doc_section"> <a name="abstract">Abstract </a></div>
196<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]197
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]198<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]199<p>This document is a reference manual for the LLVM assembly language.
200LLVM is an SSA based representation that provides type safety,
201low-level operations, flexibility, and the capability of representing
202'all' high-level languages cleanly. It is the common code
203representation used throughout all phases of the LLVM compilation
204strategy.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]205</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]206
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]207<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]208<div class="doc_section"> <a name="introduction">Introduction</a> </div>
209<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]210
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]211<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]212
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]213<p>The LLVM code representation is designed to be used in three
214different forms: as an in-memory compiler IR, as an on-disk bytecode
215representation (suitable for fast loading by a Just-In-Time compiler),
216and as a human readable assembly language representation. This allows
217LLVM to provide a powerful intermediate representation for efficient
218compiler transformations and analysis, while providing a natural means
219to debug and visualize the transformations. The three different forms
220of LLVM are all equivalent. This document describes the human readable
221representation and notation.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]222
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]223<p>The LLVM representation aims to be light-weight and low-level
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]224while being expressive, typed, and extensible at the same time. It
225aims to be a "universal IR" of sorts, by being at a low enough level
226that high-level ideas may be cleanly mapped to it (similar to how
227microprocessors are "universal IR's", allowing many source languages to
228be mapped to them). By providing type information, LLVM can be used as
229the target of optimizations: for example, through pointer analysis, it
230can be proven that a C automatic variable is never accessed outside of
231the current function... allowing it to be promoted to a simple SSA
232value instead of a memory location.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]233
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]234</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]235
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]236<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]237<div class="doc_subsubsection"> <a name="wellformed">Well-Formedness</a> </div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]238
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]239<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]240
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]241<p>It is important to note that this document describes 'well formed'
242LLVM assembly language. There is a difference between what the parser
243accepts and what is considered 'well formed'. For example, the
244following instruction is syntactically okay, but not well formed:</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]245
246<pre>
247 %x = <a href="#i_add">add</a> int 1, %x
248</pre>
249
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]250<p>...because the definition of <tt>%x</tt> does not dominate all of
251its uses. The LLVM infrastructure provides a verification pass that may
252be used to verify that an LLVM module is well formed. This pass is
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]253automatically run by the parser after parsing input assembly and by
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]254the optimizer before it outputs bytecode. The violations pointed out
255by the verifier pass indicate bugs in transformation passes or input to
256the parser.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]257
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]258<!-- Describe the typesetting conventions here. --> </div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]259
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_section"> <a name="identifiers">Identifiers</a> </div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]262<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]263
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]264<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]265
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]266<p>LLVM uses three different forms of identifiers, for different
267purposes:</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]268
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]269<ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]270 <li>Named values are represented as a string of characters with a '%' prefix.
271 For example, %foo, %DivisionByZero, %a.really.long.identifier. The actual
272 regular expression used is '<tt>%[a-zA-Z$._][a-zA-Z$._0-9]*</tt>'.
273 Identifiers which require other characters in their names can be surrounded
274 with quotes. In this way, anything except a <tt>"</tt> character can be used
275 in a name.</li>
276
277 <li>Unnamed values are represented as an unsigned numeric value with a '%'
278 prefix. For example, %12, %2, %44.</li>
279
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]280 <li>Constants, which are described in a <a href="#constants">section about
281 constants</a>, below.</li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]282</ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]283
284<p>LLVM requires that values start with a '%' sign for two reasons: Compilers
285don't need to worry about name clashes with reserved words, and the set of
286reserved words may be expanded in the future without penalty. Additionally,
287unnamed identifiers allow a compiler to quickly come up with a temporary
288variable without having to avoid symbol table conflicts.</p>
289
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]290<p>Reserved words in LLVM are very similar to reserved words in other
291languages. There are keywords for different opcodes ('<tt><a
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]292href="#i_add">add</a></tt>', '<tt><a href="#i_cast">cast</a></tt>', '<tt><a
293href="#i_ret">ret</a></tt>', etc...), for primitive type names ('<tt><a
294href="#t_void">void</a></tt>', '<tt><a href="#t_uint">uint</a></tt>', etc...),
295and others. These reserved words cannot conflict with variable names, because
296none of them start with a '%' character.</p>
297
298<p>Here is an example of LLVM code to multiply the integer variable
299'<tt>%X</tt>' by 8:</p>
300
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]301<p>The easy way:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]302
303<pre>
304 %result = <a href="#i_mul">mul</a> uint %X, 8
305</pre>
306
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]307<p>After strength reduction:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]308
309<pre>
310 %result = <a href="#i_shl">shl</a> uint %X, ubyte 3
311</pre>
312
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]313<p>And the hard way:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]314
315<pre>
316 <a href="#i_add">add</a> uint %X, %X <i>; yields {uint}:%0</i>
317 <a href="#i_add">add</a> uint %0, %0 <i>; yields {uint}:%1</i>
318 %result = <a href="#i_add">add</a> uint %1, %1
319</pre>
320
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]321<p>This last way of multiplying <tt>%X</tt> by 8 illustrates several
322important lexical features of LLVM:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]323
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]324<ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]325
326 <li>Comments are delimited with a '<tt>;</tt>' and go until the end of
327 line.</li>
328
329 <li>Unnamed temporaries are created when the result of a computation is not
330 assigned to a named value.</li>
331
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]332 <li>Unnamed temporaries are numbered sequentially</li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]333
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]334</ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]335
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]336<p>...and it also shows a convention that we follow in this document. When
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]337demonstrating instructions, we will follow an instruction with a comment that
338defines the type and name of value produced. Comments are shown in italic
339text.</p>
340
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]341</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]342
343<!-- *********************************************************************** -->
344<div class="doc_section"> <a name="highlevel">High Level Structure</a> </div>
345<!-- *********************************************************************** -->
346
347<!-- ======================================================================= -->
348<div class="doc_subsection"> <a name="modulestructure">Module Structure</a>
349</div>
350
351<div class="doc_text">
352
353<p>LLVM programs are composed of "Module"s, each of which is a
354translation unit of the input programs. Each module consists of
355functions, global variables, and symbol table entries. Modules may be
356combined together with the LLVM linker, which merges function (and
357global variable) definitions, resolves forward declarations, and merges
358symbol table entries. Here is an example of the "hello world" module:</p>
359
360<pre><i>; Declare the string constant as a global constant...</i>
361<a href="#identifiers">%.LC0</a> = <a href="#linkage_internal">internal</a> <a
362 href="#globalvars">constant</a> <a href="#t_array">[13 x sbyte]</a> c"hello world\0A\00" <i>; [13 x sbyte]*</i>
363
364<i>; External declaration of the puts function</i>
365<a href="#functionstructure">declare</a> int %puts(sbyte*) <i>; int(sbyte*)* </i>
366
Chris Lattner81c01f02006-06-13 03:05:47 +0000[diff] [blame]367<i>; Global variable / Function body section separator</i>
368implementation
369
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]370<i>; Definition of main function</i>
371int %main() { <i>; int()* </i>
372 <i>; Convert [13x sbyte]* to sbyte *...</i>
373 %cast210 = <a
374 href="#i_getelementptr">getelementptr</a> [13 x sbyte]* %.LC0, long 0, long 0 <i>; sbyte*</i>
375
376 <i>; Call puts function to write out the string to stdout...</i>
377 <a
378 href="#i_call">call</a> int %puts(sbyte* %cast210) <i>; int</i>
379 <a
380 href="#i_ret">ret</a> int 0<br>}<br></pre>
381
382<p>This example is made up of a <a href="#globalvars">global variable</a>
383named "<tt>.LC0</tt>", an external declaration of the "<tt>puts</tt>"
384function, and a <a href="#functionstructure">function definition</a>
385for "<tt>main</tt>".</p>
386
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]387<p>In general, a module is made up of a list of global values,
388where both functions and global variables are global values. Global values are
389represented by a pointer to a memory location (in this case, a pointer to an
390array of char, and a pointer to a function), and have one of the following <a
391href="#linkage">linkage types</a>.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]392
Chris Lattner81c01f02006-06-13 03:05:47 +0000[diff] [blame]393<p>Due to a limitation in the current LLVM assembly parser (it is limited by
394one-token lookahead), modules are split into two pieces by the "implementation"
395keyword. Global variable prototypes and definitions must occur before the
396keyword, and function definitions must occur after it. Function prototypes may
397occur either before or after it. In the future, the implementation keyword may
398become a noop, if the parser gets smarter.</p>
399
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]400</div>
401
402<!-- ======================================================================= -->
403<div class="doc_subsection">
404 <a name="linkage">Linkage Types</a>
405</div>
406
407<div class="doc_text">
408
409<p>
410All Global Variables and Functions have one of the following types of linkage:
411</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]412
413<dl>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]414
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]415 <dt><tt><b><a name="linkage_internal">internal</a></b></tt> </dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]416
417 <dd>Global values with internal linkage are only directly accessible by
418 objects in the current module. In particular, linking code into a module with
419 an internal global value may cause the internal to be renamed as necessary to
420 avoid collisions. Because the symbol is internal to the module, all
421 references can be updated. This corresponds to the notion of the
422 '<tt>static</tt>' keyword in C, or the idea of "anonymous namespaces" in C++.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]423 </dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]424
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]425 <dt><tt><b><a name="linkage_linkonce">linkonce</a></b></tt>: </dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]426
427 <dd>"<tt>linkonce</tt>" linkage is similar to <tt>internal</tt> linkage, with
428 the twist that linking together two modules defining the same
429 <tt>linkonce</tt> globals will cause one of the globals to be discarded. This
430 is typically used to implement inline functions. Unreferenced
431 <tt>linkonce</tt> globals are allowed to be discarded.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]432 </dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]433
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]434 <dt><tt><b><a name="linkage_weak">weak</a></b></tt>: </dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]435
436 <dd>"<tt>weak</tt>" linkage is exactly the same as <tt>linkonce</tt> linkage,
437 except that unreferenced <tt>weak</tt> globals may not be discarded. This is
438 used to implement constructs in C such as "<tt>int X;</tt>" at global scope.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]439 </dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]440
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]441 <dt><tt><b><a name="linkage_appending">appending</a></b></tt>: </dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]442
443 <dd>"<tt>appending</tt>" linkage may only be applied to global variables of
444 pointer to array type. When two global variables with appending linkage are
445 linked together, the two global arrays are appended together. This is the
446 LLVM, typesafe, equivalent of having the system linker append together
447 "sections" with identical names when .o files are linked.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]448 </dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]449
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]450 <dt><tt><b><a name="linkage_external">externally visible</a></b></tt>:</dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]451
452 <dd>If none of the above identifiers are used, the global is externally
453 visible, meaning that it participates in linkage and can be used to resolve
454 external symbol references.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]455 </dd>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]456
457 <dt><tt><b><a name="linkage_externweak">extern_weak</a></b></tt>: </dt>
458
459 <dd>"<tt>extern_weak</tt>" TBD
460 </dd>
461
462 <p>
463 The next two types of linkage are targeted for Microsoft Windows platform
464 only. They are designed to support importing (exporting) symbols from (to)
465 DLLs.
466 </p>
467
468 <dt><tt><b><a name="linkage_dllimport">dllimport</a></b></tt>: </dt>
469
470 <dd>"<tt>dllimport</tt>" linkage causes the compiler to reference a function
471 or variable via a global pointer to a pointer that is set up by the DLL
472 exporting the symbol. On Microsoft Windows targets, the pointer name is
473 formed by combining <code>_imp__</code> and the function or variable name.
474 </dd>
475
476 <dt><tt><b><a name="linkage_dllexport">dllexport</a></b></tt>: </dt>
477
478 <dd>"<tt>dllexport</tt>" linkage causes the compiler to provide a global
479 pointer to a pointer in a DLL, so that it can be referenced with the
480 <tt>dllimport</tt> attribute. On Microsoft Windows targets, the pointer
481 name is formed by combining <code>_imp__</code> and the function or variable
482 name.
483 </dd>
484
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]485</dl>
486
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]487<p><a name="linkage_external">For example, since the "<tt>.LC0</tt>"
488variable is defined to be internal, if another module defined a "<tt>.LC0</tt>"
489variable and was linked with this one, one of the two would be renamed,
490preventing a collision. Since "<tt>main</tt>" and "<tt>puts</tt>" are
491external (i.e., lacking any linkage declarations), they are accessible
492outside of the current module. It is illegal for a function <i>declaration</i>
493to have any linkage type other than "externally visible".</a></p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]494
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]495</div>
496
497<!-- ======================================================================= -->
498<div class="doc_subsection">
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]499 <a name="callingconv">Calling Conventions</a>
500</div>
501
502<div class="doc_text">
503
504<p>LLVM <a href="#functionstructure">functions</a>, <a href="#i_call">calls</a>
505and <a href="#i_invoke">invokes</a> can all have an optional calling convention
506specified for the call. The calling convention of any pair of dynamic
507caller/callee must match, or the behavior of the program is undefined. The
508following calling conventions are supported by LLVM, and more may be added in
509the future:</p>
510
511<dl>
512 <dt><b>"<tt>ccc</tt>" - The C calling convention</b>:</dt>
513
514 <dd>This calling convention (the default if no other calling convention is
515 specified) matches the target C calling conventions. This calling convention
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]516 supports varargs function calls and tolerates some mismatch in the declared
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]517 prototype and implemented declaration of the function (as does normal C).
518 </dd>
519
Chris Lattner5710ce92006-05-19 21:15:36 +0000[diff] [blame]520 <dt><b>"<tt>csretcc</tt>" - The C struct return calling convention</b>:</dt>
521
522 <dd>This calling convention matches the target C calling conventions, except
523 that functions with this convention are required to take a pointer as their
524 first argument, and the return type of the function must be void. This is
525 used for C functions that return aggregates by-value. In this case, the
526 function has been transformed to take a pointer to the struct as the first
527 argument to the function. For targets where the ABI specifies specific
528 behavior for structure-return calls, the calling convention can be used to
529 distinguish between struct return functions and other functions that take a
530 pointer to a struct as the first argument.
531 </dd>
532
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]533 <dt><b>"<tt>fastcc</tt>" - The fast calling convention</b>:</dt>
534
535 <dd>This calling convention attempts to make calls as fast as possible
536 (e.g. by passing things in registers). This calling convention allows the
537 target to use whatever tricks it wants to produce fast code for the target,
Chris Lattner8cdc5bc2005-05-06 23:08:23 +0000[diff] [blame]538 without having to conform to an externally specified ABI. Implementations of
539 this convention should allow arbitrary tail call optimization to be supported.
540 This calling convention does not support varargs and requires the prototype of
541 all callees to exactly match the prototype of the function definition.
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]542 </dd>
543
544 <dt><b>"<tt>coldcc</tt>" - The cold calling convention</b>:</dt>
545
546 <dd>This calling convention attempts to make code in the caller as efficient
547 as possible under the assumption that the call is not commonly executed. As
548 such, these calls often preserve all registers so that the call does not break
549 any live ranges in the caller side. This calling convention does not support
550 varargs and requires the prototype of all callees to exactly match the
551 prototype of the function definition.
552 </dd>
553
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]554 <dt><b>"<tt>cc &lt;<em>n</em>&gt;</tt>" - Numbered convention</b>:</dt>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]555
556 <dd>Any calling convention may be specified by number, allowing
557 target-specific calling conventions to be used. Target specific calling
558 conventions start at 64.
559 </dd>
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]560</dl>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]561
562<p>More calling conventions can be added/defined on an as-needed basis, to
563support pascal conventions or any other well-known target-independent
564convention.</p>
565
566</div>
567
568<!-- ======================================================================= -->
569<div class="doc_subsection">
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]570 <a name="globalvars">Global Variables</a>
571</div>
572
573<div class="doc_text">
574
Chris Lattner3689a342005-02-12 19:30:21 +0000[diff] [blame]575<p>Global variables define regions of memory allocated at compilation time
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]576instead of run-time. Global variables may optionally be initialized, may have
577an explicit section to be placed in, and may
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]578have an optional explicit alignment specified. A
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]579variable may be defined as a global "constant," which indicates that the
Chris Lattner3689a342005-02-12 19:30:21 +0000[diff] [blame]580contents of the variable will <b>never</b> be modified (enabling better
581optimization, allowing the global data to be placed in the read-only section of
582an executable, etc). Note that variables that need runtime initialization
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]583cannot be marked "constant" as there is a store to the variable.</p>
Chris Lattner3689a342005-02-12 19:30:21 +0000[diff] [blame]584
585<p>
586LLVM explicitly allows <em>declarations</em> of global variables to be marked
587constant, even if the final definition of the global is not. This capability
588can be used to enable slightly better optimization of the program, but requires
589the language definition to guarantee that optimizations based on the
590'constantness' are valid for the translation units that do not include the
591definition.
592</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]593
594<p>As SSA values, global variables define pointer values that are in
595scope (i.e. they dominate) all basic blocks in the program. Global
596variables always define a pointer to their "content" type because they
597describe a region of memory, and all memory objects in LLVM are
598accessed through pointers.</p>
599
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]600<p>LLVM allows an explicit section to be specified for globals. If the target
601supports it, it will emit globals to the section specified.</p>
602
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]603<p>An explicit alignment may be specified for a global. If not present, or if
604the alignment is set to zero, the alignment of the global is set by the target
605to whatever it feels convenient. If an explicit alignment is specified, the
606global is forced to have at least that much alignment. All alignments must be
607a power of 2.</p>
608
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]609</div>
610
611
612<!-- ======================================================================= -->
613<div class="doc_subsection">
614 <a name="functionstructure">Functions</a>
615</div>
616
617<div class="doc_text">
618
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]619<p>LLVM function definitions consist of an optional <a href="#linkage">linkage
620type</a>, an optional <a href="#callingconv">calling convention</a>, a return
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]621type, a function name, a (possibly empty) argument list, an optional section,
622an optional alignment, an opening curly brace,
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]623a list of basic blocks, and a closing curly brace. LLVM function declarations
624are defined with the "<tt>declare</tt>" keyword, an optional <a
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]625href="#callingconv">calling convention</a>, a return type, a function name,
626a possibly empty list of arguments, and an optional alignment.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]627
628<p>A function definition contains a list of basic blocks, forming the CFG for
629the function. Each basic block may optionally start with a label (giving the
630basic block a symbol table entry), contains a list of instructions, and ends
631with a <a href="#terminators">terminator</a> instruction (such as a branch or
632function return).</p>
633
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]634<p>The first basic block in a program is special in two ways: it is immediately
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]635executed on entrance to the function, and it is not allowed to have predecessor
636basic blocks (i.e. there can not be any branches to the entry block of a
637function). Because the block can have no predecessors, it also cannot have any
638<a href="#i_phi">PHI nodes</a>.</p>
639
640<p>LLVM functions are identified by their name and type signature. Hence, two
641functions with the same name but different parameter lists or return values are
Chris Lattnerd4f6b172005-03-07 22:13:59 +0000[diff] [blame]642considered different functions, and LLVM will resolve references to each
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]643appropriately.</p>
644
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]645<p>LLVM allows an explicit section to be specified for functions. If the target
646supports it, it will emit functions to the section specified.</p>
647
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]648<p>An explicit alignment may be specified for a function. If not present, or if
649the alignment is set to zero, the alignment of the function is set by the target
650to whatever it feels convenient. If an explicit alignment is specified, the
651function is forced to have at least that much alignment. All alignments must be
652a power of 2.</p>
653
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]654</div>
655
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]656<!-- ======================================================================= -->
657<div class="doc_subsection">
Chris Lattner1eeeb0c2006-04-08 04:40:53 +0000[diff] [blame]658 <a name="moduleasm">Module-Level Inline Assembly</a>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]659</div>
660
661<div class="doc_text">
662<p>
663Modules may contain "module-level inline asm" blocks, which corresponds to the
664GCC "file scope inline asm" blocks. These blocks are internally concatenated by
665LLVM and treated as a single unit, but may be separated in the .ll file if
666desired. The syntax is very simple:
667</p>
668
669<div class="doc_code"><pre>
Chris Lattner52599e12006-01-24 00:37:20 +0000[diff] [blame]670 module asm "inline asm code goes here"
671 module asm "more can go here"
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]672</pre></div>
673
674<p>The strings can contain any character by escaping non-printable characters.
675 The escape sequence used is simply "\xx" where "xx" is the two digit hex code
676 for the number.
677</p>
678
679<p>
680 The inline asm code is simply printed to the machine code .s file when
681 assembly code is generated.
682</p>
683</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]684
685
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]686<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]687<div class="doc_section"> <a name="typesystem">Type System</a> </div>
688<!-- *********************************************************************** -->
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]689
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]690<div class="doc_text">
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]691
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]692<p>The LLVM type system is one of the most important features of the
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]693intermediate representation. Being typed enables a number of
694optimizations to be performed on the IR directly, without having to do
695extra analyses on the side before the transformation. A strong type
696system makes it easier to read the generated code and enables novel
697analyses and transformations that are not feasible to perform on normal
698three address code representations.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]699
700</div>
701
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]702<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]703<div class="doc_subsection"> <a name="t_primitive">Primitive Types</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]704<div class="doc_text">
John Criswell4457dc92004-04-09 16:48:45 +0000[diff] [blame]705<p>The primitive types are the fundamental building blocks of the LLVM
Chris Lattnerd4f6b172005-03-07 22:13:59 +0000[diff] [blame]706system. The current set of primitive types is as follows:</p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]707
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]708<table class="layout">
709 <tr class="layout">
710 <td class="left">
711 <table>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]712 <tbody>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]713 <tr><th>Type</th><th>Description</th></tr>
714 <tr><td><tt>void</tt></td><td>No value</td></tr>
Misha Brukmancfa87bc2005-04-22 18:02:52 +0000[diff] [blame]715 <tr><td><tt>ubyte</tt></td><td>Unsigned 8-bit value</td></tr>
716 <tr><td><tt>ushort</tt></td><td>Unsigned 16-bit value</td></tr>
717 <tr><td><tt>uint</tt></td><td>Unsigned 32-bit value</td></tr>
718 <tr><td><tt>ulong</tt></td><td>Unsigned 64-bit value</td></tr>
719 <tr><td><tt>float</tt></td><td>32-bit floating point value</td></tr>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]720 <tr><td><tt>label</tt></td><td>Branch destination</td></tr>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]721 </tbody>
722 </table>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]723 </td>
724 <td class="right">
725 <table>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]726 <tbody>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]727 <tr><th>Type</th><th>Description</th></tr>
728 <tr><td><tt>bool</tt></td><td>True or False value</td></tr>
Misha Brukmancfa87bc2005-04-22 18:02:52 +0000[diff] [blame]729 <tr><td><tt>sbyte</tt></td><td>Signed 8-bit value</td></tr>
730 <tr><td><tt>short</tt></td><td>Signed 16-bit value</td></tr>
731 <tr><td><tt>int</tt></td><td>Signed 32-bit value</td></tr>
732 <tr><td><tt>long</tt></td><td>Signed 64-bit value</td></tr>
733 <tr><td><tt>double</tt></td><td>64-bit floating point value</td></tr>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]734 </tbody>
735 </table>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]736 </td>
737 </tr>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]738</table>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]739</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]740
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]741<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]742<div class="doc_subsubsection"> <a name="t_classifications">Type
743Classifications</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]744<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]745<p>These different primitive types fall into a few useful
746classifications:</p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]747
748<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]749 <tbody>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]750 <tr><th>Classification</th><th>Types</th></tr>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]751 <tr>
752 <td><a name="t_signed">signed</a></td>
753 <td><tt>sbyte, short, int, long, float, double</tt></td>
754 </tr>
755 <tr>
756 <td><a name="t_unsigned">unsigned</a></td>
757 <td><tt>ubyte, ushort, uint, ulong</tt></td>
758 </tr>
759 <tr>
760 <td><a name="t_integer">integer</a></td>
761 <td><tt>ubyte, sbyte, ushort, short, uint, int, ulong, long</tt></td>
762 </tr>
763 <tr>
764 <td><a name="t_integral">integral</a></td>
Misha Brukmanc24b7582004-08-12 20:16:08 +0000[diff] [blame]765 <td><tt>bool, ubyte, sbyte, ushort, short, uint, int, ulong, long</tt>
766 </td>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]767 </tr>
768 <tr>
769 <td><a name="t_floating">floating point</a></td>
770 <td><tt>float, double</tt></td>
771 </tr>
772 <tr>
773 <td><a name="t_firstclass">first class</a></td>
Misha Brukmanc24b7582004-08-12 20:16:08 +0000[diff] [blame]774 <td><tt>bool, ubyte, sbyte, ushort, short, uint, int, ulong, long,<br>
775 float, double, <a href="#t_pointer">pointer</a>,
776 <a href="#t_packed">packed</a></tt></td>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]777 </tr>
778 </tbody>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]779</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]780
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]781<p>The <a href="#t_firstclass">first class</a> types are perhaps the
782most important. Values of these types are the only ones which can be
783produced by instructions, passed as arguments, or used as operands to
784instructions. This means that all structures and arrays must be
785manipulated either by pointer or by component.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]786</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]787
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]788<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]789<div class="doc_subsection"> <a name="t_derived">Derived Types</a> </div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]790
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]791<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]792
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]793<p>The real power in LLVM comes from the derived types in the system.
794This is what allows a programmer to represent arrays, functions,
795pointers, and other useful types. Note that these derived types may be
796recursive: For example, it is possible to have a two dimensional array.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]797
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]798</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]799
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]800<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]801<div class="doc_subsubsection"> <a name="t_array">Array Type</a> </div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]802
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]803<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]804
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]805<h5>Overview:</h5>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]806
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]807<p>The array type is a very simple derived type that arranges elements
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]808sequentially in memory. The array type requires a size (number of
809elements) and an underlying data type.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]810
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]811<h5>Syntax:</h5>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]812
813<pre>
814 [&lt;# elements&gt; x &lt;elementtype&gt;]
815</pre>
816
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]817<p>The number of elements is a constant integer value; elementtype may
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]818be any type with a size.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]819
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]820<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]821<table class="layout">
822 <tr class="layout">
823 <td class="left">
824 <tt>[40 x int ]</tt><br/>
825 <tt>[41 x int ]</tt><br/>
826 <tt>[40 x uint]</tt><br/>
827 </td>
828 <td class="left">
829 Array of 40 integer values.<br/>
830 Array of 41 integer values.<br/>
831 Array of 40 unsigned integer values.<br/>
832 </td>
833 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]834</table>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]835<p>Here are some examples of multidimensional arrays:</p>
836<table class="layout">
837 <tr class="layout">
838 <td class="left">
839 <tt>[3 x [4 x int]]</tt><br/>
840 <tt>[12 x [10 x float]]</tt><br/>
841 <tt>[2 x [3 x [4 x uint]]]</tt><br/>
842 </td>
843 <td class="left">
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]844 3x4 array of integer values.<br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]845 12x10 array of single precision floating point values.<br/>
846 2x3x4 array of unsigned integer values.<br/>
847 </td>
848 </tr>
849</table>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]850
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]851<p>Note that 'variable sized arrays' can be implemented in LLVM with a zero
852length array. Normally, accesses past the end of an array are undefined in
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]853LLVM (e.g. it is illegal to access the 5th element of a 3 element array).
854As a special case, however, zero length arrays are recognized to be variable
855length. This allows implementation of 'pascal style arrays' with the LLVM
856type "{ int, [0 x float]}", for example.</p>
857
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]858</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]859
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]860<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]861<div class="doc_subsubsection"> <a name="t_function">Function Type</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]862<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]863<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]864<p>The function type can be thought of as a function signature. It
865consists of a return type and a list of formal parameter types.
John Criswell009900b2003-11-25 21:45:46 +0000[diff] [blame]866Function types are usually used to build virtual function tables
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]867(which are structures of pointers to functions), for indirect function
868calls, and when defining a function.</p>
John Criswell009900b2003-11-25 21:45:46 +0000[diff] [blame]869<p>
870The return type of a function type cannot be an aggregate type.
871</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]872<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]873<pre> &lt;returntype&gt; (&lt;parameter list&gt;)<br></pre>
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]874<p>...where '<tt>&lt;parameter list&gt;</tt>' is a comma-separated list of type
Misha Brukmanc24b7582004-08-12 20:16:08 +0000[diff] [blame]875specifiers. Optionally, the parameter list may include a type <tt>...</tt>,
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]876which indicates that the function takes a variable number of arguments.
877Variable argument functions can access their arguments with the <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]878 href="#int_varargs">variable argument handling intrinsic</a> functions.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]879<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]880<table class="layout">
881 <tr class="layout">
882 <td class="left">
883 <tt>int (int)</tt> <br/>
884 <tt>float (int, int *) *</tt><br/>
885 <tt>int (sbyte *, ...)</tt><br/>
886 </td>
887 <td class="left">
888 function taking an <tt>int</tt>, returning an <tt>int</tt><br/>
889 <a href="#t_pointer">Pointer</a> to a function that takes an
Misha Brukmanc24b7582004-08-12 20:16:08 +0000[diff] [blame]890 <tt>int</tt> and a <a href="#t_pointer">pointer</a> to <tt>int</tt>,
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]891 returning <tt>float</tt>.<br/>
892 A vararg function that takes at least one <a href="#t_pointer">pointer</a>
893 to <tt>sbyte</tt> (signed char in C), which returns an integer. This is
894 the signature for <tt>printf</tt> in LLVM.<br/>
895 </td>
896 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]897</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]898
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]899</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]900<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]901<div class="doc_subsubsection"> <a name="t_struct">Structure Type</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]902<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]903<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]904<p>The structure type is used to represent a collection of data members
905together in memory. The packing of the field types is defined to match
906the ABI of the underlying processor. The elements of a structure may
907be any type that has a size.</p>
908<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
909and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
910field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
911instruction.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]912<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]913<pre> { &lt;type list&gt; }<br></pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]914<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]915<table class="layout">
916 <tr class="layout">
917 <td class="left">
918 <tt>{ int, int, int }</tt><br/>
919 <tt>{ float, int (int) * }</tt><br/>
920 </td>
921 <td class="left">
922 a triple of three <tt>int</tt> values<br/>
923 A pair, where the first element is a <tt>float</tt> and the second element
924 is a <a href="#t_pointer">pointer</a> to a <a href="#t_function">function</a>
925 that takes an <tt>int</tt>, returning an <tt>int</tt>.<br/>
926 </td>
927 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]928</table>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]929</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]930
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]931<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]932<div class="doc_subsubsection"> <a name="t_pointer">Pointer Type</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]933<div class="doc_text">
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]934<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]935<p>As in many languages, the pointer type represents a pointer or
936reference to another object, which must live in memory.</p>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]937<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]938<pre> &lt;type&gt; *<br></pre>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]939<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]940<table class="layout">
941 <tr class="layout">
942 <td class="left">
943 <tt>[4x int]*</tt><br/>
944 <tt>int (int *) *</tt><br/>
945 </td>
946 <td class="left">
947 A <a href="#t_pointer">pointer</a> to <a href="#t_array">array</a> of
948 four <tt>int</tt> values<br/>
949 A <a href="#t_pointer">pointer</a> to a <a
Chris Lattnera977c482005-02-19 02:22:14 +0000[diff] [blame]950 href="#t_function">function</a> that takes an <tt>int*</tt>, returning an
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]951 <tt>int</tt>.<br/>
952 </td>
953 </tr>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]954</table>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]955</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]956
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]957<!-- _______________________________________________________________________ -->
958<div class="doc_subsubsection"> <a name="t_packed">Packed Type</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]959<div class="doc_text">
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]960
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]961<h5>Overview:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]962
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]963<p>A packed type is a simple derived type that represents a vector
964of elements. Packed types are used when multiple primitive data
965are operated in parallel using a single instruction (SIMD).
966A packed type requires a size (number of
Chris Lattnerb8d172f2005-11-10 01:44:22 +0000[diff] [blame]967elements) and an underlying primitive data type. Vectors must have a power
968of two length (1, 2, 4, 8, 16 ...). Packed types are
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]969considered <a href="#t_firstclass">first class</a>.</p>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]970
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]971<h5>Syntax:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]972
973<pre>
974 &lt; &lt;# elements&gt; x &lt;elementtype&gt; &gt;
975</pre>
976
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]977<p>The number of elements is a constant integer value; elementtype may
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]978be any integral or floating point type.</p>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]979
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]980<h5>Examples:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]981
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]982<table class="layout">
983 <tr class="layout">
984 <td class="left">
985 <tt>&lt;4 x int&gt;</tt><br/>
986 <tt>&lt;8 x float&gt;</tt><br/>
987 <tt>&lt;2 x uint&gt;</tt><br/>
988 </td>
989 <td class="left">
990 Packed vector of 4 integer values.<br/>
991 Packed vector of 8 floating-point values.<br/>
992 Packed vector of 2 unsigned integer values.<br/>
993 </td>
994 </tr>
995</table>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]996</div>
997
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]998<!-- _______________________________________________________________________ -->
999<div class="doc_subsubsection"> <a name="t_opaque">Opaque Type</a> </div>
1000<div class="doc_text">
1001
1002<h5>Overview:</h5>
1003
1004<p>Opaque types are used to represent unknown types in the system. This
1005corresponds (for example) to the C notion of a foward declared structure type.
1006In LLVM, opaque types can eventually be resolved to any type (not just a
1007structure type).</p>
1008
1009<h5>Syntax:</h5>
1010
1011<pre>
1012 opaque
1013</pre>
1014
1015<h5>Examples:</h5>
1016
1017<table class="layout">
1018 <tr class="layout">
1019 <td class="left">
1020 <tt>opaque</tt>
1021 </td>
1022 <td class="left">
1023 An opaque type.<br/>
1024 </td>
1025 </tr>
1026</table>
1027</div>
1028
1029
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1030<!-- *********************************************************************** -->
1031<div class="doc_section"> <a name="constants">Constants</a> </div>
1032<!-- *********************************************************************** -->
1033
1034<div class="doc_text">
1035
1036<p>LLVM has several different basic types of constants. This section describes
1037them all and their syntax.</p>
1038
1039</div>
1040
1041<!-- ======================================================================= -->
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]1042<div class="doc_subsection"><a name="simpleconstants">Simple Constants</a></div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1043
1044<div class="doc_text">
1045
1046<dl>
1047 <dt><b>Boolean constants</b></dt>
1048
1049 <dd>The two strings '<tt>true</tt>' and '<tt>false</tt>' are both valid
1050 constants of the <tt><a href="#t_primitive">bool</a></tt> type.
1051 </dd>
1052
1053 <dt><b>Integer constants</b></dt>
1054
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]1055 <dd>Standard integers (such as '4') are constants of the <a
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1056 href="#t_integer">integer</a> type. Negative numbers may be used with signed
1057 integer types.
1058 </dd>
1059
1060 <dt><b>Floating point constants</b></dt>
1061
1062 <dd>Floating point constants use standard decimal notation (e.g. 123.421),
1063 exponential notation (e.g. 1.23421e+2), or a more precise hexadecimal
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1064 notation (see below). Floating point constants must have a <a
1065 href="#t_floating">floating point</a> type. </dd>
1066
1067 <dt><b>Null pointer constants</b></dt>
1068
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1069 <dd>The identifier '<tt>null</tt>' is recognized as a null pointer constant
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1070 and must be of <a href="#t_pointer">pointer type</a>.</dd>
1071
1072</dl>
1073
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1074<p>The one non-intuitive notation for constants is the optional hexadecimal form
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1075of floating point constants. For example, the form '<tt>double
10760x432ff973cafa8000</tt>' is equivalent to (but harder to read than) '<tt>double
10774.5e+15</tt>'. The only time hexadecimal floating point constants are required
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]1078(and the only time that they are generated by the disassembler) is when a
1079floating point constant must be emitted but it cannot be represented as a
1080decimal floating point number. For example, NaN's, infinities, and other
1081special values are represented in their IEEE hexadecimal format so that
1082assembly and disassembly do not cause any bits to change in the constants.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1083
1084</div>
1085
1086<!-- ======================================================================= -->
1087<div class="doc_subsection"><a name="aggregateconstants">Aggregate Constants</a>
1088</div>
1089
1090<div class="doc_text">
Chris Lattnerd4f6b172005-03-07 22:13:59 +0000[diff] [blame]1091<p>Aggregate constants arise from aggregation of simple constants
1092and smaller aggregate constants.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1093
1094<dl>
1095 <dt><b>Structure constants</b></dt>
1096
1097 <dd>Structure constants are represented with notation similar to structure
1098 type definitions (a comma separated list of elements, surrounded by braces
Chris Lattnerd4f6b172005-03-07 22:13:59 +0000[diff] [blame]1099 (<tt>{}</tt>)). For example: "<tt>{ int 4, float 17.0, int* %G }</tt>",
1100 where "<tt>%G</tt>" is declared as "<tt>%G = external global int</tt>". Structure constants
1101 must have <a href="#t_struct">structure type</a>, and the number and
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1102 types of elements must match those specified by the type.
1103 </dd>
1104
1105 <dt><b>Array constants</b></dt>
1106
1107 <dd>Array constants are represented with notation similar to array type
1108 definitions (a comma separated list of elements, surrounded by square brackets
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1109 (<tt>[]</tt>)). For example: "<tt>[ int 42, int 11, int 74 ]</tt>". Array
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1110 constants must have <a href="#t_array">array type</a>, and the number and
1111 types of elements must match those specified by the type.
1112 </dd>
1113
1114 <dt><b>Packed constants</b></dt>
1115
1116 <dd>Packed constants are represented with notation similar to packed type
1117 definitions (a comma separated list of elements, surrounded by
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1118 less-than/greater-than's (<tt>&lt;&gt;</tt>)). For example: "<tt>&lt; int 42,
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1119 int 11, int 74, int 100 &gt;</tt>". Packed constants must have <a
1120 href="#t_packed">packed type</a>, and the number and types of elements must
1121 match those specified by the type.
1122 </dd>
1123
1124 <dt><b>Zero initialization</b></dt>
1125
1126 <dd>The string '<tt>zeroinitializer</tt>' can be used to zero initialize a
1127 value to zero of <em>any</em> type, including scalar and aggregate types.
1128 This is often used to avoid having to print large zero initializers (e.g. for
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]1129 large arrays) and is always exactly equivalent to using explicit zero
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1130 initializers.
1131 </dd>
1132</dl>
1133
1134</div>
1135
1136<!-- ======================================================================= -->
1137<div class="doc_subsection">
1138 <a name="globalconstants">Global Variable and Function Addresses</a>
1139</div>
1140
1141<div class="doc_text">
1142
1143<p>The addresses of <a href="#globalvars">global variables</a> and <a
1144href="#functionstructure">functions</a> are always implicitly valid (link-time)
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1145constants. These constants are explicitly referenced when the <a
1146href="#identifiers">identifier for the global</a> is used and always have <a
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1147href="#t_pointer">pointer</a> type. For example, the following is a legal LLVM
1148file:</p>
1149
1150<pre>
1151 %X = global int 17
1152 %Y = global int 42
1153 %Z = global [2 x int*] [ int* %X, int* %Y ]
1154</pre>
1155
1156</div>
1157
1158<!-- ======================================================================= -->
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]1159<div class="doc_subsection"><a name="undefvalues">Undefined Values</a></div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1160<div class="doc_text">
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]1161 <p>The string '<tt>undef</tt>' is recognized as a type-less constant that has
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]1162 no specific value. Undefined values may be of any type and be used anywhere
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]1163 a constant is permitted.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1164
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]1165 <p>Undefined values indicate to the compiler that the program is well defined
1166 no matter what value is used, giving the compiler more freedom to optimize.
1167 </p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1168</div>
1169
1170<!-- ======================================================================= -->
1171<div class="doc_subsection"><a name="constantexprs">Constant Expressions</a>
1172</div>
1173
1174<div class="doc_text">
1175
1176<p>Constant expressions are used to allow expressions involving other constants
1177to be used as constants. Constant expressions may be of any <a
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]1178href="#t_firstclass">first class</a> type and may involve any LLVM operation
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1179that does not have side effects (e.g. load and call are not supported). The
1180following is the syntax for constant expressions:</p>
1181
1182<dl>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1183 <dt><b><tt>trunc ( CST to TYPE )</tt></b></dt>
1184 <dd>Truncate a constant to another type. The bit size of CST must be larger
1185 than the bit size of TYPE. Both types must be integral.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1186
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1187 <dt><b><tt>zext ( CST to TYPE )</tt></b></dt>
1188 <dd>Zero extend a constant to another type. The bit size of CST must be
1189 smaller or equal to the bit size of TYPE. Both types must be integral.</dd>
1190
1191 <dt><b><tt>sext ( CST to TYPE )</tt></b></dt>
1192 <dd>Sign extend a constant to another type. The bit size of CST must be
1193 smaller or equal to the bit size of TYPE. Both types must be integral.</dd>
1194
1195 <dt><b><tt>fptrunc ( CST to TYPE )</tt></b></dt>
1196 <dd>Truncate a floating point constant to another floating point type. The
1197 size of CST must be larger than the size of TYPE. Both types must be
1198 floating point.</dd>
1199
1200 <dt><b><tt>fpext ( CST to TYPE )</tt></b></dt>
1201 <dd>Floating point extend a constant to another type. The size of CST must be
1202 smaller or equal to the size of TYPE. Both types must be floating point.</dd>
1203
1204 <dt><b><tt>fp2uint ( CST to TYPE )</tt></b></dt>
1205 <dd>Convert a floating point constant to the corresponding unsigned integer
1206 constant. TYPE must be an integer type. CST must be floating point. If the
1207 value won't fit in the integer type, the results are undefined.</dd>
1208
1209 <dt><b><tt>fp2sint ( CST to TYPE )</tt></b></dt>
1210 <dd>Convert a floating point constant to the corresponding signed integer
1211 constant. TYPE must be an integer type. CST must be floating point. If the
1212 value won't fit in the integer type, the results are undefined.</dd>
1213
1214 <dt><b><tt>uint2fp ( CST to TYPE )</tt></b></dt>
1215 <dd>Convert an unsigned integer constant to the corresponding floating point
1216 constant. TYPE must be floating point. CST must be of integer type. If the
1217 value won't fit in the floating point type, the results are undefined.</dd>
1218
1219 <dt><b><tt>sint2fp ( CST to TYPE )</tt></b></dt>
1220 <dd>Convert a signed integer constant to the corresponding floating point
1221 constant. TYPE must be floating point. CST must be of integer type. If the
1222 value won't fit in the floating point type, the results are undefined.</dd>
1223
1224 <dt><b><tt>bitconvert ( CST to TYPE )</tt></b></dt>
1225 <dd>Convert a constant, CST, to another TYPE. The size of CST and TYPE must be
1226 identical (same number of bits). The conversion is done as if the CST value
1227 was stored to memory and read back as TYPE. In other words, no bits change
1228 with this operator, just the type. This can be used for conversion of pointer
1229 and packed types to any other type, as long as they have the same bit width.
1230 </dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1231
1232 <dt><b><tt>getelementptr ( CSTPTR, IDX0, IDX1, ... )</tt></b></dt>
1233
1234 <dd>Perform the <a href="#i_getelementptr">getelementptr operation</a> on
1235 constants. As with the <a href="#i_getelementptr">getelementptr</a>
1236 instruction, the index list may have zero or more indexes, which are required
1237 to make sense for the type of "CSTPTR".</dd>
1238
Robert Bocchino9fbe1452006-01-10 19:31:34 +0000[diff] [blame]1239 <dt><b><tt>select ( COND, VAL1, VAL2 )</tt></b></dt>
1240
1241 <dd>Perform the <a href="#i_select">select operation</a> on
1242 constants.
1243
1244 <dt><b><tt>extractelement ( VAL, IDX )</tt></b></dt>
1245
1246 <dd>Perform the <a href="#i_extractelement">extractelement
1247 operation</a> on constants.
1248
Robert Bocchino05ccd702006-01-15 20:48:27 +0000[diff] [blame]1249 <dt><b><tt>insertelement ( VAL, ELT, IDX )</tt></b></dt>
1250
1251 <dd>Perform the <a href="#i_insertelement">insertelement
1252 operation</a> on constants.
1253
Chris Lattnerc1989542006-04-08 00:13:41 +0000[diff] [blame]1254
1255 <dt><b><tt>shufflevector ( VEC1, VEC2, IDXMASK )</tt></b></dt>
1256
1257 <dd>Perform the <a href="#i_shufflevector">shufflevector
1258 operation</a> on constants.
1259
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1260 <dt><b><tt>OPCODE ( LHS, RHS )</tt></b></dt>
1261
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]1262 <dd>Perform the specified operation of the LHS and RHS constants. OPCODE may
1263 be any of the <a href="#binaryops">binary</a> or <a href="#bitwiseops">bitwise
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1264 binary</a> operations. The constraints on operands are the same as those for
1265 the corresponding instruction (e.g. no bitwise operations on floating point
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]1266 values are allowed).</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1267</dl>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1268</div>
Chris Lattner9ee5d222004-03-08 16:49:10 +0000[diff] [blame]1269
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1270<!-- *********************************************************************** -->
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]1271<div class="doc_section"> <a name="othervalues">Other Values</a> </div>
1272<!-- *********************************************************************** -->
1273
1274<!-- ======================================================================= -->
1275<div class="doc_subsection">
1276<a name="inlineasm">Inline Assembler Expressions</a>
1277</div>
1278
1279<div class="doc_text">
1280
1281<p>
1282LLVM supports inline assembler expressions (as opposed to <a href="#moduleasm">
1283Module-Level Inline Assembly</a>) through the use of a special value. This
1284value represents the inline assembler as a string (containing the instructions
1285to emit), a list of operand constraints (stored as a string), and a flag that
1286indicates whether or not the inline asm expression has side effects. An example
1287inline assembler expression is:
1288</p>
1289
1290<pre>
1291 int(int) asm "bswap $0", "=r,r"
1292</pre>
1293
1294<p>
1295Inline assembler expressions may <b>only</b> be used as the callee operand of
1296a <a href="#i_call"><tt>call</tt> instruction</a>. Thus, typically we have:
1297</p>
1298
1299<pre>
1300 %X = call int asm "<a href="#i_bswap">bswap</a> $0", "=r,r"(int %Y)
1301</pre>
1302
1303<p>
1304Inline asms with side effects not visible in the constraint list must be marked
1305as having side effects. This is done through the use of the
1306'<tt>sideeffect</tt>' keyword, like so:
1307</p>
1308
1309<pre>
1310 call void asm sideeffect "eieio", ""()
1311</pre>
1312
1313<p>TODO: The format of the asm and constraints string still need to be
1314documented here. Constraints on what can be done (e.g. duplication, moving, etc
1315need to be documented).
1316</p>
1317
1318</div>
1319
1320<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1321<div class="doc_section"> <a name="instref">Instruction Reference</a> </div>
1322<!-- *********************************************************************** -->
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1323
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1324<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1325
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1326<p>The LLVM instruction set consists of several different
1327classifications of instructions: <a href="#terminators">terminator
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]1328instructions</a>, <a href="#binaryops">binary instructions</a>,
1329<a href="#bitwiseops">bitwise binary instructions</a>, <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1330 href="#memoryops">memory instructions</a>, and <a href="#otherops">other
1331instructions</a>.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1332
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1333</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1334
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1335<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1336<div class="doc_subsection"> <a name="terminators">Terminator
1337Instructions</a> </div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1338
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1339<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1340
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1341<p>As mentioned <a href="#functionstructure">previously</a>, every
1342basic block in a program ends with a "Terminator" instruction, which
1343indicates which block should be executed after the current block is
1344finished. These terminator instructions typically yield a '<tt>void</tt>'
1345value: they produce control flow, not values (the one exception being
1346the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction).</p>
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1347<p>There are six different terminator instructions: the '<a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1348 href="#i_ret"><tt>ret</tt></a>' instruction, the '<a href="#i_br"><tt>br</tt></a>'
1349instruction, the '<a href="#i_switch"><tt>switch</tt></a>' instruction,
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1350the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction, the '<a
1351 href="#i_unwind"><tt>unwind</tt></a>' instruction, and the '<a
1352 href="#i_unreachable"><tt>unreachable</tt></a>' instruction.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1353
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1354</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1355
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1356<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1357<div class="doc_subsubsection"> <a name="i_ret">'<tt>ret</tt>'
1358Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1359<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1360<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1361<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]1362 ret void <i>; Return from void function</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1363</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1364<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1365<p>The '<tt>ret</tt>' instruction is used to return control flow (and a
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]1366value) from a function back to the caller.</p>
John Criswell4457dc92004-04-09 16:48:45 +0000[diff] [blame]1367<p>There are two forms of the '<tt>ret</tt>' instruction: one that
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1368returns a value and then causes control flow, and one that just causes
1369control flow to occur.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1370<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1371<p>The '<tt>ret</tt>' instruction may return any '<a
1372 href="#t_firstclass">first class</a>' type. Notice that a function is
1373not <a href="#wellformed">well formed</a> if there exists a '<tt>ret</tt>'
1374instruction inside of the function that returns a value that does not
1375match the return type of the function.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1376<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1377<p>When the '<tt>ret</tt>' instruction is executed, control flow
1378returns back to the calling function's context. If the caller is a "<a
John Criswellfa081872004-06-25 15:16:57 +0000[diff] [blame]1379 href="#i_call"><tt>call</tt></a>" instruction, execution continues at
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1380the instruction after the call. If the caller was an "<a
1381 href="#i_invoke"><tt>invoke</tt></a>" instruction, execution continues
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]1382at the beginning of the "normal" destination block. If the instruction
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1383returns a value, that value shall set the call or invoke instruction's
1384return value.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1385<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1386<pre> ret int 5 <i>; Return an integer value of 5</i>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1387 ret void <i>; Return from a void function</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1388</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1389</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1390<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1391<div class="doc_subsubsection"> <a name="i_br">'<tt>br</tt>' Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1392<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1393<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1394<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]1395</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1396<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1397<p>The '<tt>br</tt>' instruction is used to cause control flow to
1398transfer to a different basic block in the current function. There are
1399two forms of this instruction, corresponding to a conditional branch
1400and an unconditional branch.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1401<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1402<p>The conditional branch form of the '<tt>br</tt>' instruction takes a
1403single '<tt>bool</tt>' value and two '<tt>label</tt>' values. The
1404unconditional form of the '<tt>br</tt>' instruction takes a single '<tt>label</tt>'
1405value as a target.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1406<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1407<p>Upon execution of a conditional '<tt>br</tt>' instruction, the '<tt>bool</tt>'
1408argument is evaluated. If the value is <tt>true</tt>, control flows
1409to the '<tt>iftrue</tt>' <tt>label</tt> argument. If "cond" is <tt>false</tt>,
1410control flows to the '<tt>iffalse</tt>' <tt>label</tt> argument.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1411<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1412<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
1413 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]1414</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1415<!-- _______________________________________________________________________ -->
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1416<div class="doc_subsubsection">
1417 <a name="i_switch">'<tt>switch</tt>' Instruction</a>
1418</div>
1419
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1420<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1421<h5>Syntax:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1422
1423<pre>
1424 switch &lt;intty&gt; &lt;value&gt;, label &lt;defaultdest&gt; [ &lt;intty&gt; &lt;val&gt;, label &lt;dest&gt; ... ]
1425</pre>
1426
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1427<h5>Overview:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1428
1429<p>The '<tt>switch</tt>' instruction is used to transfer control flow to one of
1430several different places. It is a generalization of the '<tt>br</tt>'
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1431instruction, allowing a branch to occur to one of many possible
1432destinations.</p>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1433
1434
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1435<h5>Arguments:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1436
1437<p>The '<tt>switch</tt>' instruction uses three parameters: an integer
1438comparison value '<tt>value</tt>', a default '<tt>label</tt>' destination, and
1439an array of pairs of comparison value constants and '<tt>label</tt>'s. The
1440table is not allowed to contain duplicate constant entries.</p>
1441
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1442<h5>Semantics:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1443
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1444<p>The <tt>switch</tt> instruction specifies a table of values and
1445destinations. When the '<tt>switch</tt>' instruction is executed, this
John Criswell84114752004-06-25 16:05:06 +0000[diff] [blame]1446table is searched for the given value. If the value is found, control flow is
1447transfered to the corresponding destination; otherwise, control flow is
1448transfered to the default destination.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1449
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1450<h5>Implementation:</h5>
1451
1452<p>Depending on properties of the target machine and the particular
1453<tt>switch</tt> instruction, this instruction may be code generated in different
John Criswell84114752004-06-25 16:05:06 +0000[diff] [blame]1454ways. For example, it could be generated as a series of chained conditional
1455branches or with a lookup table.</p>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1456
1457<h5>Example:</h5>
1458
1459<pre>
1460 <i>; Emulate a conditional br instruction</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1461 %Val = <a href="#i_zext">zext</a> bool %value to int
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1462 switch int %Val, label %truedest [int 0, label %falsedest ]
1463
1464 <i>; Emulate an unconditional br instruction</i>
1465 switch uint 0, label %dest [ ]
1466
1467 <i>; Implement a jump table:</i>
1468 switch uint %val, label %otherwise [ uint 0, label %onzero
1469 uint 1, label %onone
1470 uint 2, label %ontwo ]
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1471</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1472</div>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1473
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1474<!-- _______________________________________________________________________ -->
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1475<div class="doc_subsubsection">
1476 <a name="i_invoke">'<tt>invoke</tt>' Instruction</a>
1477</div>
1478
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1479<div class="doc_text">
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1480
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1481<h5>Syntax:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1482
1483<pre>
1484 &lt;result&gt; = invoke [<a href="#callingconv">cconv</a>] &lt;ptr to function ty&gt; %&lt;function ptr val&gt;(&lt;function args&gt;)
Chris Lattner76b8a332006-05-14 18:23:06 +0000[diff] [blame]1485 to label &lt;normal label&gt; unwind label &lt;exception label&gt;
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1486</pre>
1487
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]1488<h5>Overview:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1489
1490<p>The '<tt>invoke</tt>' instruction causes control to transfer to a specified
1491function, with the possibility of control flow transfer to either the
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]1492'<tt>normal</tt>' label or the
1493'<tt>exception</tt>' label. If the callee function returns with the
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1494"<tt><a href="#i_ret">ret</a></tt>" instruction, control flow will return to the
1495"normal" label. If the callee (or any indirect callees) returns with the "<a
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]1496href="#i_unwind"><tt>unwind</tt></a>" instruction, control is interrupted and
1497continued at the dynamically nearest "exception" label.</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1498
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1499<h5>Arguments:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1500
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1501<p>This instruction requires several arguments:</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1502
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1503<ol>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1504 <li>
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]1505 The optional "cconv" marker indicates which <a href="callingconv">calling
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1506 convention</a> the call should use. If none is specified, the call defaults
1507 to using C calling conventions.
1508 </li>
1509 <li>'<tt>ptr to function ty</tt>': shall be the signature of the pointer to
1510 function value being invoked. In most cases, this is a direct function
1511 invocation, but indirect <tt>invoke</tt>s are just as possible, branching off
1512 an arbitrary pointer to function value.
1513 </li>
1514
1515 <li>'<tt>function ptr val</tt>': An LLVM value containing a pointer to a
1516 function to be invoked. </li>
1517
1518 <li>'<tt>function args</tt>': argument list whose types match the function
1519 signature argument types. If the function signature indicates the function
1520 accepts a variable number of arguments, the extra arguments can be
1521 specified. </li>
1522
1523 <li>'<tt>normal label</tt>': the label reached when the called function
1524 executes a '<tt><a href="#i_ret">ret</a></tt>' instruction. </li>
1525
1526 <li>'<tt>exception label</tt>': the label reached when a callee returns with
1527 the <a href="#i_unwind"><tt>unwind</tt></a> instruction. </li>
1528
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1529</ol>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1530
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1531<h5>Semantics:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1532
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1533<p>This instruction is designed to operate as a standard '<tt><a
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1534href="#i_call">call</a></tt>' instruction in most regards. The primary
1535difference is that it establishes an association with a label, which is used by
1536the runtime library to unwind the stack.</p>
1537
1538<p>This instruction is used in languages with destructors to ensure that proper
1539cleanup is performed in the case of either a <tt>longjmp</tt> or a thrown
1540exception. Additionally, this is important for implementation of
1541'<tt>catch</tt>' clauses in high-level languages that support them.</p>
1542
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1543<h5>Example:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1544<pre>
1545 %retval = invoke int %Test(int 15) to label %Continue
Chris Lattner76b8a332006-05-14 18:23:06 +0000[diff] [blame]1546 unwind label %TestCleanup <i>; {int}:retval set</i>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1547 %retval = invoke <a href="#callingconv">coldcc</a> int %Test(int 15) to label %Continue
Chris Lattner76b8a332006-05-14 18:23:06 +0000[diff] [blame]1548 unwind label %TestCleanup <i>; {int}:retval set</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1549</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1550</div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1551
1552
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]1553<!-- _______________________________________________________________________ -->
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1554
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1555<div class="doc_subsubsection"> <a name="i_unwind">'<tt>unwind</tt>'
1556Instruction</a> </div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1557
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1558<div class="doc_text">
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1559
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]1560<h5>Syntax:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1561<pre>
1562 unwind
1563</pre>
1564
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]1565<h5>Overview:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1566
1567<p>The '<tt>unwind</tt>' instruction unwinds the stack, continuing control flow
1568at the first callee in the dynamic call stack which used an <a
1569href="#i_invoke"><tt>invoke</tt></a> instruction to perform the call. This is
1570primarily used to implement exception handling.</p>
1571
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]1572<h5>Semantics:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1573
1574<p>The '<tt>unwind</tt>' intrinsic causes execution of the current function to
1575immediately halt. The dynamic call stack is then searched for the first <a
1576href="#i_invoke"><tt>invoke</tt></a> instruction on the call stack. Once found,
1577execution continues at the "exceptional" destination block specified by the
1578<tt>invoke</tt> instruction. If there is no <tt>invoke</tt> instruction in the
1579dynamic call chain, undefined behavior results.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1580</div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1581
1582<!-- _______________________________________________________________________ -->
1583
1584<div class="doc_subsubsection"> <a name="i_unreachable">'<tt>unreachable</tt>'
1585Instruction</a> </div>
1586
1587<div class="doc_text">
1588
1589<h5>Syntax:</h5>
1590<pre>
1591 unreachable
1592</pre>
1593
1594<h5>Overview:</h5>
1595
1596<p>The '<tt>unreachable</tt>' instruction has no defined semantics. This
1597instruction is used to inform the optimizer that a particular portion of the
1598code is not reachable. This can be used to indicate that the code after a
1599no-return function cannot be reached, and other facts.</p>
1600
1601<h5>Semantics:</h5>
1602
1603<p>The '<tt>unreachable</tt>' instruction has no defined semantics.</p>
1604</div>
1605
1606
1607
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1608<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1609<div class="doc_subsection"> <a name="binaryops">Binary Operations</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1610<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1611<p>Binary operators are used to do most of the computation in a
1612program. They require two operands, execute an operation on them, and
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1613produce a single value. The operands might represent
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1614multiple data, as is the case with the <a href="#t_packed">packed</a> data type.
1615The result value of a binary operator is not
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1616necessarily the same type as its operands.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1617<p>There are several different binary operators:</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1618</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1619<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1620<div class="doc_subsubsection"> <a name="i_add">'<tt>add</tt>'
1621Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1622<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1623<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1624<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]1625</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1626<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1627<p>The '<tt>add</tt>' instruction returns the sum of its two operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1628<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1629<p>The two arguments to the '<tt>add</tt>' instruction must be either <a
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1630 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a> values.
1631 This instruction can also take <a href="#t_packed">packed</a> versions of the values.
1632Both arguments must have identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1633<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1634<p>The value produced is the integer or floating point sum of the two
1635operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1636<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1637<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]1638</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1639</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1640<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1641<div class="doc_subsubsection"> <a name="i_sub">'<tt>sub</tt>'
1642Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1643<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1644<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1645<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]1646</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1647<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1648<p>The '<tt>sub</tt>' instruction returns the difference of its two
1649operands.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1650<p>Note that the '<tt>sub</tt>' instruction is used to represent the '<tt>neg</tt>'
1651instruction present in most other intermediate representations.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1652<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1653<p>The two arguments to the '<tt>sub</tt>' instruction must be either <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1654 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1655values.
1656This instruction can also take <a href="#t_packed">packed</a> versions of the values.
1657Both arguments must have identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1658<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1659<p>The value produced is the integer or floating point difference of
1660the two operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1661<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1662<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]1663 &lt;result&gt; = sub int 0, %val <i>; yields {int}:result = -%var</i>
1664</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1665</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1666<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1667<div class="doc_subsubsection"> <a name="i_mul">'<tt>mul</tt>'
1668Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1669<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1670<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1671<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]1672</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1673<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1674<p>The '<tt>mul</tt>' instruction returns the product of its two
1675operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1676<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1677<p>The two arguments to the '<tt>mul</tt>' instruction must be either <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1678 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1679values.
1680This instruction can also take <a href="#t_packed">packed</a> versions of the values.
1681Both arguments must have identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1682<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1683<p>The value produced is the integer or floating point product of the
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1684two operands.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1685<p>There is no signed vs unsigned multiplication. The appropriate
1686action is taken based on the type of the operand.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1687<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1688<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]1689</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1690</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1691<!-- _______________________________________________________________________ -->
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]1692<div class="doc_subsubsection"> <a name="i_udiv">'<tt>udiv</tt>' Instruction
1693</a></div>
1694<div class="doc_text">
1695<h5>Syntax:</h5>
1696<pre> &lt;result&gt; = udiv &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
1697</pre>
1698<h5>Overview:</h5>
1699<p>The '<tt>udiv</tt>' instruction returns the quotient of its two
1700operands.</p>
1701<h5>Arguments:</h5>
1702<p>The two arguments to the '<tt>udiv</tt>' instruction must be
1703<a href="#t_integer">integer</a> values. Both arguments must have identical
1704types. This instruction can also take <a href="#t_packed">packed</a> versions
1705of the values in which case the elements must be integers.</p>
1706<h5>Semantics:</h5>
1707<p>The value produced is the unsigned integer quotient of the two operands. This
1708instruction always performs an unsigned division operation, regardless of
1709whether the arguments are unsigned or not.</p>
1710<h5>Example:</h5>
1711<pre> &lt;result&gt; = udiv uint 4, %var <i>; yields {uint}:result = 4 / %var</i>
1712</pre>
1713</div>
1714<!-- _______________________________________________________________________ -->
1715<div class="doc_subsubsection"> <a name="i_sdiv">'<tt>sdiv</tt>' Instruction
1716</a> </div>
1717<div class="doc_text">
1718<h5>Syntax:</h5>
1719<pre> &lt;result&gt; = sdiv &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
1720</pre>
1721<h5>Overview:</h5>
1722<p>The '<tt>sdiv</tt>' instruction returns the quotient of its two
1723operands.</p>
1724<h5>Arguments:</h5>
1725<p>The two arguments to the '<tt>sdiv</tt>' instruction must be
1726<a href="#t_integer">integer</a> values. Both arguments must have identical
1727types. This instruction can also take <a href="#t_packed">packed</a> versions
1728of the values in which case the elements must be integers.</p>
1729<h5>Semantics:</h5>
1730<p>The value produced is the signed integer quotient of the two operands. This
1731instruction always performs a signed division operation, regardless of whether
1732the arguments are signed or not.</p>
1733<h5>Example:</h5>
1734<pre> &lt;result&gt; = sdiv int 4, %var <i>; yields {int}:result = 4 / %var</i>
1735</pre>
1736</div>
1737<!-- _______________________________________________________________________ -->
1738<div class="doc_subsubsection"> <a name="i_fdiv">'<tt>fdiv</tt>'
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1739Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1740<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1741<h5>Syntax:</h5>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]1742<pre> &lt;result&gt; = fdiv &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1743</pre>
1744<h5>Overview:</h5>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]1745<p>The '<tt>fdiv</tt>' instruction returns the quotient of its two
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1746operands.</p>
1747<h5>Arguments:</h5>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]1748<p>The two arguments to the '<tt>div</tt>' instruction must be
1749<a href="#t_floating">floating point</a> values. Both arguments must have
1750identical types. This instruction can also take <a href="#t_packed">packed</a>
1751versions of the values in which case the elements must be floating point.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1752<h5>Semantics:</h5>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]1753<p>The value produced is the floating point quotient of the two operands.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1754<h5>Example:</h5>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]1755<pre> &lt;result&gt; = fdiv float 4.0, %var <i>; yields {float}:result = 4.0 / %var</i>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1756</pre>
1757</div>
1758<!-- _______________________________________________________________________ -->
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]1759<div class="doc_subsubsection"> <a name="i_urem">'<tt>urem</tt>' Instruction</a>
1760</div>
1761<div class="doc_text">
1762<h5>Syntax:</h5>
1763<pre> &lt;result&gt; = urem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
1764</pre>
1765<h5>Overview:</h5>
1766<p>The '<tt>urem</tt>' instruction returns the remainder from the
1767unsigned division of its two arguments.</p>
1768<h5>Arguments:</h5>
1769<p>The two arguments to the '<tt>urem</tt>' instruction must be
1770<a href="#t_integer">integer</a> values. Both arguments must have identical
1771types.</p>
1772<h5>Semantics:</h5>
1773<p>This instruction returns the unsigned integer <i>remainder</i> of a division.
1774This instruction always performs an unsigned division to get the remainder,
1775regardless of whether the arguments are unsigned or not.</p>
1776<h5>Example:</h5>
1777<pre> &lt;result&gt; = urem uint 4, %var <i>; yields {uint}:result = 4 % %var</i>
1778</pre>
1779
1780</div>
1781<!-- _______________________________________________________________________ -->
1782<div class="doc_subsubsection"> <a name="i_srem">'<tt>srem</tt>'
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1783Instruction</a> </div>
1784<div class="doc_text">
1785<h5>Syntax:</h5>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]1786<pre> &lt;result&gt; = srem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1787</pre>
1788<h5>Overview:</h5>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]1789<p>The '<tt>srem</tt>' instruction returns the remainder from the
1790signed division of its two operands.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1791<h5>Arguments:</h5>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]1792<p>The two arguments to the '<tt>srem</tt>' instruction must be
1793<a href="#t_integer">integer</a> values. Both arguments must have identical
1794types.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1795<h5>Semantics:</h5>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]1796<p>This instruction returns the <i>remainder</i> of a division (where the result
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1797has the same sign as the divisor), not the <i>modulus</i> (where the
1798result has the same sign as the dividend) of a value. For more
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]1799information about the difference, see <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1800 href="http://mathforum.org/dr.math/problems/anne.4.28.99.html">The
1801Math Forum</a>.</p>
1802<h5>Example:</h5>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]1803<pre> &lt;result&gt; = srem int 4, %var <i>; yields {int}:result = 4 % %var</i>
1804</pre>
1805
1806</div>
1807<!-- _______________________________________________________________________ -->
1808<div class="doc_subsubsection"> <a name="i_frem">'<tt>frem</tt>'
1809Instruction</a> </div>
1810<div class="doc_text">
1811<h5>Syntax:</h5>
1812<pre> &lt;result&gt; = frem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
1813</pre>
1814<h5>Overview:</h5>
1815<p>The '<tt>frem</tt>' instruction returns the remainder from the
1816division of its two operands.</p>
1817<h5>Arguments:</h5>
1818<p>The two arguments to the '<tt>frem</tt>' instruction must be
1819<a href="#t_floating">floating point</a> values. Both arguments must have
1820identical types.</p>
1821<h5>Semantics:</h5>
1822<p>This instruction returns the <i>remainder</i> of a division.</p>
1823<h5>Example:</h5>
1824<pre> &lt;result&gt; = frem float 4.0, %var <i>; yields {float}:result = 4.0 % %var</i>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1825</pre>
Robert Bocchino7b81c752006-02-17 21:18:08 +0000[diff] [blame]1826
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1827</div>
1828<!-- _______________________________________________________________________ -->
1829<div class="doc_subsubsection"> <a name="i_setcc">'<tt>set<i>cc</i></tt>'
1830Instructions</a> </div>
1831<div class="doc_text">
1832<h5>Syntax:</h5>
1833<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]1834 &lt;result&gt; = setne &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {bool}:result</i>
1835 &lt;result&gt; = setlt &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {bool}:result</i>
1836 &lt;result&gt; = setgt &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {bool}:result</i>
1837 &lt;result&gt; = setle &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {bool}:result</i>
1838 &lt;result&gt; = setge &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {bool}:result</i>
1839</pre>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1840<h5>Overview:</h5>
1841<p>The '<tt>set<i>cc</i></tt>' family of instructions returns a boolean
1842value based on a comparison of their two operands.</p>
1843<h5>Arguments:</h5>
1844<p>The two arguments to the '<tt>set<i>cc</i></tt>' instructions must
1845be of <a href="#t_firstclass">first class</a> type (it is not possible
1846to compare '<tt>label</tt>'s, '<tt>array</tt>'s, '<tt>structure</tt>'
1847or '<tt>void</tt>' values, etc...). Both arguments must have identical
1848types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1849<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1850<p>The '<tt>seteq</tt>' instruction yields a <tt>true</tt> '<tt>bool</tt>'
1851value if both operands are equal.<br>
1852The '<tt>setne</tt>' instruction yields a <tt>true</tt> '<tt>bool</tt>'
1853value if both operands are unequal.<br>
1854The '<tt>setlt</tt>' instruction yields a <tt>true</tt> '<tt>bool</tt>'
1855value if the first operand is less than the second operand.<br>
1856The '<tt>setgt</tt>' instruction yields a <tt>true</tt> '<tt>bool</tt>'
1857value if the first operand is greater than the second operand.<br>
1858The '<tt>setle</tt>' instruction yields a <tt>true</tt> '<tt>bool</tt>'
1859value if the first operand is less than or equal to the second operand.<br>
1860The '<tt>setge</tt>' instruction yields a <tt>true</tt> '<tt>bool</tt>'
1861value if the first operand is greater than or equal to the second
1862operand.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1863<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1864<pre> &lt;result&gt; = seteq int 4, 5 <i>; yields {bool}:result = false</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1865 &lt;result&gt; = setne float 4, 5 <i>; yields {bool}:result = true</i>
1866 &lt;result&gt; = setlt uint 4, 5 <i>; yields {bool}:result = true</i>
1867 &lt;result&gt; = setgt sbyte 4, 5 <i>; yields {bool}:result = false</i>
1868 &lt;result&gt; = setle sbyte 4, 5 <i>; yields {bool}:result = true</i>
1869 &lt;result&gt; = setge sbyte 4, 5 <i>; yields {bool}:result = false</i>
1870</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1871</div>
Robert Bocchino7b81c752006-02-17 21:18:08 +0000[diff] [blame]1872
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1873<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1874<div class="doc_subsection"> <a name="bitwiseops">Bitwise Binary
1875Operations</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1876<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1877<p>Bitwise binary operators are used to do various forms of
1878bit-twiddling in a program. They are generally very efficient
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1879instructions and can commonly be strength reduced from other
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1880instructions. They require two operands, execute an operation on them,
1881and produce a single value. The resulting value of the bitwise binary
1882operators is always the same type as its first operand.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1883</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1884<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1885<div class="doc_subsubsection"> <a name="i_and">'<tt>and</tt>'
1886Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1887<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1888<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1889<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]1890</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1891<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1892<p>The '<tt>and</tt>' instruction returns the bitwise logical and of
1893its two operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1894<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1895<p>The two arguments to the '<tt>and</tt>' instruction must be <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1896 href="#t_integral">integral</a> values. Both arguments must have
1897identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1898<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1899<p>The truth table used for the '<tt>and</tt>' instruction is:</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1900<p> </p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1901<div style="align: center">
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1902<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1903 <tbody>
1904 <tr>
1905 <td>In0</td>
1906 <td>In1</td>
1907 <td>Out</td>
1908 </tr>
1909 <tr>
1910 <td>0</td>
1911 <td>0</td>
1912 <td>0</td>
1913 </tr>
1914 <tr>
1915 <td>0</td>
1916 <td>1</td>
1917 <td>0</td>
1918 </tr>
1919 <tr>
1920 <td>1</td>
1921 <td>0</td>
1922 <td>0</td>
1923 </tr>
1924 <tr>
1925 <td>1</td>
1926 <td>1</td>
1927 <td>1</td>
1928 </tr>
1929 </tbody>
1930</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1931</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1932<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1933<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]1934 &lt;result&gt; = and int 15, 40 <i>; yields {int}:result = 8</i>
1935 &lt;result&gt; = and int 4, 8 <i>; yields {int}:result = 0</i>
1936</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1937</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1938<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1939<div class="doc_subsubsection"> <a name="i_or">'<tt>or</tt>' Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1940<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1941<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1942<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]1943</pre>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1944<h5>Overview:</h5>
1945<p>The '<tt>or</tt>' instruction returns the bitwise logical inclusive
1946or of its two operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1947<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1948<p>The two arguments to the '<tt>or</tt>' instruction must be <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1949 href="#t_integral">integral</a> values. Both arguments must have
1950identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1951<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1952<p>The truth table used for the '<tt>or</tt>' instruction is:</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1953<p> </p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1954<div style="align: center">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1955<table border="1" cellspacing="0" cellpadding="4">
1956 <tbody>
1957 <tr>
1958 <td>In0</td>
1959 <td>In1</td>
1960 <td>Out</td>
1961 </tr>
1962 <tr>
1963 <td>0</td>
1964 <td>0</td>
1965 <td>0</td>
1966 </tr>
1967 <tr>
1968 <td>0</td>
1969 <td>1</td>
1970 <td>1</td>
1971 </tr>
1972 <tr>
1973 <td>1</td>
1974 <td>0</td>
1975 <td>1</td>
1976 </tr>
1977 <tr>
1978 <td>1</td>
1979 <td>1</td>
1980 <td>1</td>
1981 </tr>
1982 </tbody>
1983</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1984</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1985<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1986<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]1987 &lt;result&gt; = or int 15, 40 <i>; yields {int}:result = 47</i>
1988 &lt;result&gt; = or int 4, 8 <i>; yields {int}:result = 12</i>
1989</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1990</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1991<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1992<div class="doc_subsubsection"> <a name="i_xor">'<tt>xor</tt>'
1993Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1994<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1995<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1996<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]1997</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1998<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1999<p>The '<tt>xor</tt>' instruction returns the bitwise logical exclusive
2000or of its two operands. The <tt>xor</tt> is used to implement the
2001"one's complement" operation, which is the "~" operator in C.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2002<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2003<p>The two arguments to the '<tt>xor</tt>' instruction must be <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2004 href="#t_integral">integral</a> values. Both arguments must have
2005identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2006<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2007<p>The truth table used for the '<tt>xor</tt>' instruction is:</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2008<p> </p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]2009<div style="align: center">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2010<table border="1" cellspacing="0" cellpadding="4">
2011 <tbody>
2012 <tr>
2013 <td>In0</td>
2014 <td>In1</td>
2015 <td>Out</td>
2016 </tr>
2017 <tr>
2018 <td>0</td>
2019 <td>0</td>
2020 <td>0</td>
2021 </tr>
2022 <tr>
2023 <td>0</td>
2024 <td>1</td>
2025 <td>1</td>
2026 </tr>
2027 <tr>
2028 <td>1</td>
2029 <td>0</td>
2030 <td>1</td>
2031 </tr>
2032 <tr>
2033 <td>1</td>
2034 <td>1</td>
2035 <td>0</td>
2036 </tr>
2037 </tbody>
2038</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]2039</div>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2040<p> </p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2041<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2042<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]2043 &lt;result&gt; = xor int 15, 40 <i>; yields {int}:result = 39</i>
2044 &lt;result&gt; = xor int 4, 8 <i>; yields {int}:result = 12</i>
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]2045 &lt;result&gt; = xor int %V, -1 <i>; yields {int}:result = ~%V</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2046</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2047</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2048<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2049<div class="doc_subsubsection"> <a name="i_shl">'<tt>shl</tt>'
2050Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2051<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2052<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2053<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]2054</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2055<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2056<p>The '<tt>shl</tt>' instruction returns the first operand shifted to
2057the left a specified number of bits.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2058<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2059<p>The first argument to the '<tt>shl</tt>' instruction must be an <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2060 href="#t_integer">integer</a> type. The second argument must be an '<tt>ubyte</tt>'
2061type.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2062<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2063<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]2064<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2065<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]2066 &lt;result&gt; = shl int 4, ubyte 2 <i>; yields {int}:result = 16</i>
2067 &lt;result&gt; = shl int 1, ubyte 10 <i>; yields {int}:result = 1024</i>
2068</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2069</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2070<!-- _______________________________________________________________________ -->
Reid Spencer3822ff52006-11-08 06:47:33 +0000[diff] [blame]2071<div class="doc_subsubsection"> <a name="i_lshr">'<tt>lshr</tt>'
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2072Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2073<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2074<h5>Syntax:</h5>
Reid Spencer3822ff52006-11-08 06:47:33 +0000[diff] [blame]2075<pre> &lt;result&gt; = lshr &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]2076</pre>
Reid Spencer3822ff52006-11-08 06:47:33 +0000[diff] [blame]2077
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2078<h5>Overview:</h5>
Reid Spencer3822ff52006-11-08 06:47:33 +0000[diff] [blame]2079<p>The '<tt>lshr</tt>' instruction (logical shift right) returns the first
2080operand shifted to the right a specified number of bits.</p>
2081
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2082<h5>Arguments:</h5>
Reid Spencer3822ff52006-11-08 06:47:33 +0000[diff] [blame]2083<p>The first argument to the '<tt>lshr</tt>' instruction must be an <a
2084 href="#t_integer">integer</a> type. The second argument must be an '<tt>ubyte</tt>' type.</p>
2085
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2086<h5>Semantics:</h5>
Reid Spencer3822ff52006-11-08 06:47:33 +0000[diff] [blame]2087<p>This instruction always performs a logical shift right operation, regardless
2088of whether the arguments are unsigned or not. The <tt>var2</tt> most significant
2089bits will be filled with zero bits after the shift.</p>
2090
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2091<h5>Example:</h5>
Reid Spencer3822ff52006-11-08 06:47:33 +0000[diff] [blame]2092<pre>
2093 &lt;result&gt; = lshr uint 4, ubyte 1 <i>; yields {uint}:result = 2</i>
2094 &lt;result&gt; = lshr int 4, ubyte 2 <i>; yields {uint}:result = 1</i>
2095 &lt;result&gt; = lshr sbyte 4, ubyte 3 <i>; yields {sbyte}:result = 0</i>
2096 &lt;result&gt; = lshr sbyte -2, ubyte 1 <i>; yields {sbyte}:result = 0x7FFFFFFF </i>
2097</pre>
2098</div>
2099
2100<!-- ======================================================================= -->
2101<div class="doc_subsubsection"> <a name="i_ashr">'<tt>ashr</tt>'
2102Instruction</a> </div>
2103<div class="doc_text">
2104
2105<h5>Syntax:</h5>
2106<pre> &lt;result&gt; = ashr &lt;ty&gt; &lt;var1&gt;, ubyte &lt;var2&gt; <i>; yields {ty}:result</i>
2107</pre>
2108
2109<h5>Overview:</h5>
2110<p>The '<tt>ashr</tt>' instruction (arithmetic shift right) returns the first
2111operand shifted to the right a specified number of bits.</p>
2112
2113<h5>Arguments:</h5>
2114<p>The first argument to the '<tt>ashr</tt>' instruction must be an
2115<a href="#t_integer">integer</a> type. The second argument must be an
2116'<tt>ubyte</tt>' type.</p>
2117
2118<h5>Semantics:</h5>
2119<p>This instruction always performs an arithmetic shift right operation,
2120regardless of whether the arguments are signed or not. The <tt>var2</tt> most
2121significant bits will be filled with the sign bit of <tt>var1</tt>.</p>
2122
2123<h5>Example:</h5>
2124<pre>
2125 &lt;result&gt; = ashr uint 4, ubyte 1 <i>; yields {uint}:result = 2</i>
2126 &lt;result&gt; = ashr int 4, ubyte 2 <i>; yields {int}:result = 1</i>
2127 &lt;result&gt; = ashr ubyte 4, ubyte 3 <i>; yields {ubyte}:result = 0</i>
2128 &lt;result&gt; = ashr sbyte -2, ubyte 1 <i>; yields {sbyte}:result = -1</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2129</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2130</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2131
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2132<!-- ======================================================================= -->
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2133<div class="doc_subsection">
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2134 <a name="vectorops">Vector Operations</a>
2135</div>
2136
2137<div class="doc_text">
2138
2139<p>LLVM supports several instructions to represent vector operations in a
2140target-independent manner. This instructions cover the element-access and
2141vector-specific operations needed to process vectors effectively. While LLVM
2142does directly support these vector operations, many sophisticated algorithms
2143will want to use target-specific intrinsics to take full advantage of a specific
2144target.</p>
2145
2146</div>
2147
2148<!-- _______________________________________________________________________ -->
2149<div class="doc_subsubsection">
2150 <a name="i_extractelement">'<tt>extractelement</tt>' Instruction</a>
2151</div>
2152
2153<div class="doc_text">
2154
2155<h5>Syntax:</h5>
2156
2157<pre>
2158 &lt;result&gt; = extractelement &lt;n x &lt;ty&gt;&gt; &lt;val&gt;, uint &lt;idx&gt; <i>; yields &lt;ty&gt;</i>
2159</pre>
2160
2161<h5>Overview:</h5>
2162
2163<p>
2164The '<tt>extractelement</tt>' instruction extracts a single scalar
2165element from a packed vector at a specified index.
2166</p>
2167
2168
2169<h5>Arguments:</h5>
2170
2171<p>
2172The first operand of an '<tt>extractelement</tt>' instruction is a
2173value of <a href="#t_packed">packed</a> type. The second operand is
2174an index indicating the position from which to extract the element.
2175The index may be a variable.</p>
2176
2177<h5>Semantics:</h5>
2178
2179<p>
2180The result is a scalar of the same type as the element type of
2181<tt>val</tt>. Its value is the value at position <tt>idx</tt> of
2182<tt>val</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
2183results are undefined.
2184</p>
2185
2186<h5>Example:</h5>
2187
2188<pre>
2189 %result = extractelement &lt;4 x int&gt; %vec, uint 0 <i>; yields int</i>
2190</pre>
2191</div>
2192
2193
2194<!-- _______________________________________________________________________ -->
2195<div class="doc_subsubsection">
2196 <a name="i_insertelement">'<tt>insertelement</tt>' Instruction</a>
2197</div>
2198
2199<div class="doc_text">
2200
2201<h5>Syntax:</h5>
2202
2203<pre>
2204 &lt;result&gt; = insertelement &lt;n x &lt;ty&gt;&gt; &lt;val&gt;, &lt;ty&gt; &lt;elt&gt, uint &lt;idx&gt; <i>; yields &lt;n x &lt;ty&gt;&gt;</i>
2205</pre>
2206
2207<h5>Overview:</h5>
2208
2209<p>
2210The '<tt>insertelement</tt>' instruction inserts a scalar
2211element into a packed vector at a specified index.
2212</p>
2213
2214
2215<h5>Arguments:</h5>
2216
2217<p>
2218The first operand of an '<tt>insertelement</tt>' instruction is a
2219value of <a href="#t_packed">packed</a> type. The second operand is a
2220scalar value whose type must equal the element type of the first
2221operand. The third operand is an index indicating the position at
2222which to insert the value. The index may be a variable.</p>
2223
2224<h5>Semantics:</h5>
2225
2226<p>
2227The result is a packed vector of the same type as <tt>val</tt>. Its
2228element values are those of <tt>val</tt> except at position
2229<tt>idx</tt>, where it gets the value <tt>elt</tt>. If <tt>idx</tt>
2230exceeds the length of <tt>val</tt>, the results are undefined.
2231</p>
2232
2233<h5>Example:</h5>
2234
2235<pre>
2236 %result = insertelement &lt;4 x int&gt; %vec, int 1, uint 0 <i>; yields &lt;4 x int&gt;</i>
2237</pre>
2238</div>
2239
2240<!-- _______________________________________________________________________ -->
2241<div class="doc_subsubsection">
2242 <a name="i_shufflevector">'<tt>shufflevector</tt>' Instruction</a>
2243</div>
2244
2245<div class="doc_text">
2246
2247<h5>Syntax:</h5>
2248
2249<pre>
2250 &lt;result&gt; = shufflevector &lt;n x &lt;ty&gt;&gt; &lt;v1&gt;, &lt;n x &lt;ty&gt;&gt; &lt;v2&gt;, &lt;n x uint&gt; &lt;mask&gt; <i>; yields &lt;n x &lt;ty&gt;&gt;</i>
2251</pre>
2252
2253<h5>Overview:</h5>
2254
2255<p>
2256The '<tt>shufflevector</tt>' instruction constructs a permutation of elements
2257from two input vectors, returning a vector of the same type.
2258</p>
2259
2260<h5>Arguments:</h5>
2261
2262<p>
2263The first two operands of a '<tt>shufflevector</tt>' instruction are vectors
2264with types that match each other and types that match the result of the
2265instruction. The third argument is a shuffle mask, which has the same number
2266of elements as the other vector type, but whose element type is always 'uint'.
2267</p>
2268
2269<p>
2270The shuffle mask operand is required to be a constant vector with either
2271constant integer or undef values.
2272</p>
2273
2274<h5>Semantics:</h5>
2275
2276<p>
2277The elements of the two input vectors are numbered from left to right across
2278both of the vectors. The shuffle mask operand specifies, for each element of
2279the result vector, which element of the two input registers the result element
2280gets. The element selector may be undef (meaning "don't care") and the second
2281operand may be undef if performing a shuffle from only one vector.
2282</p>
2283
2284<h5>Example:</h5>
2285
2286<pre>
2287 %result = shufflevector &lt;4 x int&gt; %v1, &lt;4 x int&gt; %v2,
2288 &lt;4 x uint&gt; &lt;uint 0, uint 4, uint 1, uint 5&gt; <i>; yields &lt;4 x int&gt;</i>
2289 %result = shufflevector &lt;4 x int&gt; %v1, &lt;4 x int&gt; undef,
2290 &lt;4 x uint&gt; &lt;uint 0, uint 1, uint 2, uint 3&gt; <i>; yields &lt;4 x int&gt;</i> - Identity shuffle.
2291</pre>
2292</div>
2293
Tanya Lattner09474292006-04-14 19:24:33 +0000[diff] [blame]2294
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2295<!-- ======================================================================= -->
2296<div class="doc_subsection">
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]2297 <a name="memoryops">Memory Access and Addressing Operations</a>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2298</div>
2299
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2300<div class="doc_text">
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2301
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2302<p>A key design point of an SSA-based representation is how it
2303represents memory. In LLVM, no memory locations are in SSA form, which
2304makes things very simple. This section describes how to read, write,
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]2305allocate, and free memory in LLVM.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2306
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2307</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2308
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2309<!-- _______________________________________________________________________ -->
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2310<div class="doc_subsubsection">
2311 <a name="i_malloc">'<tt>malloc</tt>' Instruction</a>
2312</div>
2313
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2314<div class="doc_text">
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2315
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2316<h5>Syntax:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2317
2318<pre>
2319 &lt;result&gt; = malloc &lt;type&gt;[, uint &lt;NumElements&gt;][, align &lt;alignment&gt;] <i>; yields {type*}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2320</pre>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2321
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2322<h5>Overview:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2323
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2324<p>The '<tt>malloc</tt>' instruction allocates memory from the system
2325heap and returns a pointer to it.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2326
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2327<h5>Arguments:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2328
2329<p>The '<tt>malloc</tt>' instruction allocates
2330<tt>sizeof(&lt;type&gt;)*NumElements</tt>
John Criswell6e4ca612004-02-24 16:13:56 +0000[diff] [blame]2331bytes of memory from the operating system and returns a pointer of the
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2332appropriate type to the program. If "NumElements" is specified, it is the
2333number of elements allocated. If an alignment is specified, the value result
2334of the allocation is guaranteed to be aligned to at least that boundary. If
2335not specified, or if zero, the target can choose to align the allocation on any
2336convenient boundary.</p>
2337
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2338<p>'<tt>type</tt>' must be a sized type.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2339
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2340<h5>Semantics:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2341
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2342<p>Memory is allocated using the system "<tt>malloc</tt>" function, and
2343a pointer is returned.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2344
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2345<h5>Example:</h5>
2346
2347<pre>
2348 %array = malloc [4 x ubyte ] <i>; yields {[%4 x ubyte]*}:array</i>
2349
2350 %size = <a 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]2351 %array1 = malloc ubyte, uint 4 <i>; yields {ubyte*}:array1</i>
2352 %array2 = malloc [12 x ubyte], uint %size <i>; yields {[12 x ubyte]*}:array2</i>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2353 %array3 = malloc int, uint 4, align 1024 <i>; yields {int*}:array3</i>
2354 %array4 = malloc int, align 1024 <i>; yields {int*}:array4</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2355</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2356</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2357
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2358<!-- _______________________________________________________________________ -->
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2359<div class="doc_subsubsection">
2360 <a name="i_free">'<tt>free</tt>' Instruction</a>
2361</div>
2362
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2363<div class="doc_text">
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2364
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2365<h5>Syntax:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2366
2367<pre>
2368 free &lt;type&gt; &lt;value&gt; <i>; yields {void}</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2369</pre>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2370
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2371<h5>Overview:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2372
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2373<p>The '<tt>free</tt>' instruction returns memory back to the unused
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]2374memory heap to be reallocated in the future.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2375
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2376<h5>Arguments:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2377
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2378<p>'<tt>value</tt>' shall be a pointer value that points to a value
2379that was allocated with the '<tt><a href="#i_malloc">malloc</a></tt>'
2380instruction.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2381
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2382<h5>Semantics:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2383
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]2384<p>Access to the memory pointed to by the pointer is no longer defined
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2385after this instruction executes.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2386
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2387<h5>Example:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2388
2389<pre>
2390 %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]2391 free [4 x ubyte]* %array
2392</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2393</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2394
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2395<!-- _______________________________________________________________________ -->
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2396<div class="doc_subsubsection">
2397 <a name="i_alloca">'<tt>alloca</tt>' Instruction</a>
2398</div>
2399
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2400<div class="doc_text">
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2401
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2402<h5>Syntax:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2403
2404<pre>
2405 &lt;result&gt; = alloca &lt;type&gt;[, uint &lt;NumElements&gt;][, align &lt;alignment&gt;] <i>; yields {type*}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2406</pre>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2407
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2408<h5>Overview:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2409
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2410<p>The '<tt>alloca</tt>' instruction allocates memory on the current
2411stack frame of the procedure that is live until the current function
2412returns to its caller.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2413
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2414<h5>Arguments:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2415
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]2416<p>The '<tt>alloca</tt>' instruction allocates <tt>sizeof(&lt;type&gt;)*NumElements</tt>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2417bytes of memory on the runtime stack, returning a pointer of the
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2418appropriate type to the program. If "NumElements" is specified, it is the
2419number of elements allocated. If an alignment is specified, the value result
2420of the allocation is guaranteed to be aligned to at least that boundary. If
2421not specified, or if zero, the target can choose to align the allocation on any
2422convenient boundary.</p>
2423
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2424<p>'<tt>type</tt>' may be any sized type.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2425
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2426<h5>Semantics:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2427
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]2428<p>Memory is allocated; a pointer is returned. '<tt>alloca</tt>'d
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2429memory is automatically released when the function returns. The '<tt>alloca</tt>'
2430instruction is commonly used to represent automatic variables that must
2431have an address available. When the function returns (either with the <tt><a
John Criswelldae2e932005-05-12 16:55:34 +0000[diff] [blame]2432 href="#i_ret">ret</a></tt> or <tt><a href="#i_unwind">unwind</a></tt>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2433instructions), the memory is reclaimed.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2434
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2435<h5>Example:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2436
2437<pre>
2438 %ptr = alloca int <i>; yields {int*}:ptr</i>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2439 %ptr = alloca int, uint 4 <i>; yields {int*}:ptr</i>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2440 %ptr = alloca int, uint 4, align 1024 <i>; yields {int*}:ptr</i>
2441 %ptr = alloca int, align 1024 <i>; yields {int*}:ptr</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2442</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2443</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2444
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2445<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2446<div class="doc_subsubsection"> <a name="i_load">'<tt>load</tt>'
2447Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2448<div class="doc_text">
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2449<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2450<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]2451<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2452<p>The '<tt>load</tt>' instruction is used to read from memory.</p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2453<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2454<p>The argument to the '<tt>load</tt>' instruction specifies the memory
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]2455address from which to load. The pointer must point to a <a
Chris Lattnere53e5082004-06-03 22:57:15 +0000[diff] [blame]2456 href="#t_firstclass">first class</a> type. If the <tt>load</tt> is
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]2457marked as <tt>volatile</tt>, then the optimizer is not allowed to modify
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2458the number or order of execution of this <tt>load</tt> with other
2459volatile <tt>load</tt> and <tt><a href="#i_store">store</a></tt>
2460instructions. </p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2461<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2462<p>The location of memory pointed to is loaded.</p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2463<h5>Examples:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2464<pre> %ptr = <a href="#i_alloca">alloca</a> int <i>; yields {int*}:ptr</i>
2465 <a
2466 href="#i_store">store</a> int 3, int* %ptr <i>; yields {void}</i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2467 %val = load int* %ptr <i>; yields {int}:val = int 3</i>
2468</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2469</div>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2470<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2471<div class="doc_subsubsection"> <a name="i_store">'<tt>store</tt>'
2472Instruction</a> </div>
Reid Spencer035ab572006-11-09 21:18:01 +0000[diff] [blame]2473<div class="doc_text">
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2474<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2475<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]2476 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]2477</pre>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2478<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2479<p>The '<tt>store</tt>' instruction is used to write to memory.</p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2480<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2481<p>There are two arguments to the '<tt>store</tt>' instruction: a value
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]2482to store and an address in which to store it. The type of the '<tt>&lt;pointer&gt;</tt>'
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2483operand must be a pointer to the type of the '<tt>&lt;value&gt;</tt>'
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]2484operand. If the <tt>store</tt> is marked as <tt>volatile</tt>, then the
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2485optimizer is not allowed to modify the number or order of execution of
2486this <tt>store</tt> with other volatile <tt>load</tt> and <tt><a
2487 href="#i_store">store</a></tt> instructions.</p>
2488<h5>Semantics:</h5>
2489<p>The contents of memory are updated to contain '<tt>&lt;value&gt;</tt>'
2490at the location specified by the '<tt>&lt;pointer&gt;</tt>' operand.</p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2491<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2492<pre> %ptr = <a href="#i_alloca">alloca</a> int <i>; yields {int*}:ptr</i>
2493 <a
2494 href="#i_store">store</a> int 3, int* %ptr <i>; yields {void}</i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2495 %val = load int* %ptr <i>; yields {int}:val = int 3</i>
2496</pre>
Reid Spencer47ce1792006-11-09 21:15:49 +0000[diff] [blame]2497</div>
2498
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2499<!-- _______________________________________________________________________ -->
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2500<div class="doc_subsubsection">
2501 <a name="i_getelementptr">'<tt>getelementptr</tt>' Instruction</a>
2502</div>
2503
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2504<div class="doc_text">
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2505<h5>Syntax:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2506<pre>
2507 &lt;result&gt; = getelementptr &lt;ty&gt;* &lt;ptrval&gt;{, &lt;ty&gt; &lt;idx&gt;}*
2508</pre>
2509
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2510<h5>Overview:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2511
2512<p>
2513The '<tt>getelementptr</tt>' instruction is used to get the address of a
2514subelement of an aggregate data structure.</p>
2515
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2516<h5>Arguments:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2517
2518<p>This instruction takes a list of integer constants that indicate what
2519elements of the aggregate object to index to. The actual types of the arguments
2520provided depend on the type of the first pointer argument. The
2521'<tt>getelementptr</tt>' instruction is used to index down through the type
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]2522levels of a structure or to a specific index in an array. When indexing into a
2523structure, only <tt>uint</tt>
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]2524integer constants are allowed. When indexing into an array or pointer,
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2525<tt>int</tt> and <tt>long</tt> indexes are allowed of any sign.</p>
2526
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2527<p>For example, let's consider a C code fragment and how it gets
2528compiled to LLVM:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2529
2530<pre>
2531 struct RT {
2532 char A;
2533 int B[10][20];
2534 char C;
2535 };
2536 struct ST {
2537 int X;
2538 double Y;
2539 struct RT Z;
2540 };
2541
2542 int *foo(struct ST *s) {
2543 return &amp;s[1].Z.B[5][13];
2544 }
2545</pre>
2546
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2547<p>The LLVM code generated by the GCC frontend is:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2548
2549<pre>
2550 %RT = type { sbyte, [10 x [20 x int]], sbyte }
2551 %ST = type { int, double, %RT }
2552
Brian Gaeke7283e7c2004-07-02 21:08:14 +0000[diff] [blame]2553 implementation
2554
2555 int* %foo(%ST* %s) {
2556 entry:
2557 %reg = getelementptr %ST* %s, int 1, uint 2, uint 1, int 5, int 13
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2558 ret int* %reg
2559 }
2560</pre>
2561
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2562<h5>Semantics:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2563
2564<p>The index types specified for the '<tt>getelementptr</tt>' instruction depend
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]2565on the pointer type that is being indexed into. <a href="#t_pointer">Pointer</a>
Chris Lattnere53e5082004-06-03 22:57:15 +0000[diff] [blame]2566and <a href="#t_array">array</a> types require <tt>uint</tt>, <tt>int</tt>,
2567<tt>ulong</tt>, or <tt>long</tt> values, and <a href="#t_struct">structure</a>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2568types require <tt>uint</tt> <b>constants</b>.</p>
2569
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2570<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]2571type, which is a pointer, yielding a '<tt>%ST</tt>' = '<tt>{ int, double, %RT
2572}</tt>' type, a structure. The second index indexes into the third element of
2573the structure, yielding a '<tt>%RT</tt>' = '<tt>{ sbyte, [10 x [20 x int]],
2574sbyte }</tt>' type, another structure. The third index indexes into the second
2575element of the structure, yielding a '<tt>[10 x [20 x int]]</tt>' type, an
2576array. The two dimensions of the array are subscripted into, yielding an
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]2577'<tt>int</tt>' type. The '<tt>getelementptr</tt>' instruction returns a pointer
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2578to this element, thus computing a value of '<tt>int*</tt>' type.</p>
2579
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2580<p>Note that it is perfectly legal to index partially through a
2581structure, returning a pointer to an inner element. Because of this,
2582the LLVM code for the given testcase is equivalent to:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2583
2584<pre>
Chris Lattnerd4f6b172005-03-07 22:13:59 +0000[diff] [blame]2585 int* %foo(%ST* %s) {
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2586 %t1 = getelementptr %ST* %s, int 1 <i>; yields %ST*:%t1</i>
2587 %t2 = getelementptr %ST* %t1, int 0, uint 2 <i>; yields %RT*:%t2</i>
2588 %t3 = getelementptr %RT* %t2, int 0, uint 1 <i>; yields [10 x [20 x int]]*:%t3</i>
2589 %t4 = getelementptr [10 x [20 x int]]* %t3, int 0, int 5 <i>; yields [20 x int]*:%t4</i>
2590 %t5 = getelementptr [20 x int]* %t4, int 0, int 13 <i>; yields int*:%t5</i>
2591 ret int* %t5
2592 }
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]2593</pre>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]2594
2595<p>Note that it is undefined to access an array out of bounds: array and
2596pointer indexes must always be within the defined bounds of the array type.
2597The one exception for this rules is zero length arrays. These arrays are
2598defined to be accessible as variable length arrays, which requires access
2599beyond the zero'th element.</p>
2600
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]2601<p>The getelementptr instruction is often confusing. For some more insight
2602into how it works, see <a href="GetElementPtr.html">the getelementptr
2603FAQ</a>.</p>
2604
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2605<h5>Example:</h5>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]2606
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2607<pre>
2608 <i>; yields [12 x ubyte]*:aptr</i>
2609 %aptr = getelementptr {int, [12 x ubyte]}* %sptr, long 0, uint 1
2610</pre>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2611</div>
Reid Spencer47ce1792006-11-09 21:15:49 +0000[diff] [blame]2612
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2613<!-- ======================================================================= -->
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]2614<div class="doc_subsection"> <a name="convertops">Conversion Operations</a>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2615</div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2616<div class="doc_text">
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]2617<p>The instructions in this category are the conversion instructions (casting)
2618which all take a single operand and a type. They perform various bit conversions
2619on the operand.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2620</div>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]2621
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]2622<!-- _______________________________________________________________________ -->
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]2623<div class="doc_subsubsection">
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2624 <a name="i_trunc">'<tt>trunc .. to</tt>' Instruction</a>
2625</div>
2626<div class="doc_text">
2627
2628<h5>Syntax:</h5>
2629<pre>
2630 &lt;result&gt; = trunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2631</pre>
2632
2633<h5>Overview:</h5>
2634<p>
2635The '<tt>trunc</tt>' instruction truncates its operand to the type <tt>ty2</tt>.
2636</p>
2637
2638<h5>Arguments:</h5>
2639<p>
2640The '<tt>trunc</tt>' instruction takes a <tt>value</tt> to trunc, which must
2641be an <a href="#t_integer">integer</a> type, and a type that specifies the size
2642and type of the result, which must be an <a href="#t_integral">integral</a>
2643type.</p>
2644
2645<h5>Semantics:</h5>
2646<p>
2647The '<tt>trunc</tt>' instruction truncates the high order bits in <tt>value</tt>
2648and converts the reamining bits to <tt>ty2</tt>. The bit size of <tt>value</tt>
2649must be larger than the bit size of <tt>ty2</tt>. Equal sized types are not
2650allowed. This implies that a <tt>trunc</tt> cannot be a <i>no-op cast</i>. It
2651will always truncate bits.</p>
2652
2653<p>When truncating to bool, the truncation is done as a comparison against
2654zero. If the <tt>value</tt> was zero, the bool result will be <tt>false</tt>.
2655If the <tt>value</tt> was non-zero, the bool result will be <tt>true</tt>.</p>
2656
2657<h5>Example:</h5>
2658<pre>
2659 %X = trunc int 257 to ubyte <i>; yields ubyte:1</i>
2660 %Y = trunc int 123 to bool <i>; yields bool:true</i>
2661</pre>
2662</div>
2663
2664<!-- _______________________________________________________________________ -->
2665<div class="doc_subsubsection">
2666 <a name="i_zext">'<tt>zext .. to</tt>' Instruction</a>
2667</div>
2668<div class="doc_text">
2669
2670<h5>Syntax:</h5>
2671<pre>
2672 &lt;result&gt; = zext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2673</pre>
2674
2675<h5>Overview:</h5>
2676<p>The '<tt>zext</tt>' instruction zero extends its operand to type
2677<tt>ty2</tt>.</p>
2678
2679
2680<h5>Arguments:</h5>
2681<p>The '<tt>zext</tt>' instruction takes a value to cast, which must be of
2682<a href="#t_integral">integral</a> type, and a type to cast it to, which must
2683also be of <a href="#t_integral">integral</a> type. The bit size of the
2684<tt>value</tt> must be smaller than or equal to the bit size of the
2685destination type, <tt>ty2</tt>.</p>
2686
2687<h5>Semantics:</h5>
2688<p>The <tt>zext</tt> fills the high order bits of the <tt>value</tt> with zero
2689bits until it reaches the size of the destination type, <tt>ty2</tt>. When the
2690the operand and the type are the same size, no bit filling is done and the
2691cast is considered a <i>no-op cast</i> because no bits change (only the type
2692changes).</p>
2693
2694<p>When zero extending to bool, the extension is done as a comparison against
2695zero. If the <tt>value</tt> was zero, the bool result will be <tt>false</tt>.
2696If the <tt>value</tt> was non-zero, the bool result will be <tt>true</tt>.</p>
2697
2698<h5>Example:</h5>
2699<pre>
2700 %X = zext int 257 to ulong <i>; yields ulong:257</i>
2701 %Y = zext bool true to int <i>; yields int:1</i>
2702</pre>
2703</div>
2704
2705<!-- _______________________________________________________________________ -->
2706<div class="doc_subsubsection">
2707 <a name="i_sext">'<tt>sext .. to</tt>' Instruction</a>
2708</div>
2709<div class="doc_text">
2710
2711<h5>Syntax:</h5>
2712<pre>
2713 &lt;result&gt; = sext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2714</pre>
2715
2716<h5>Overview:</h5>
2717<p>The '<tt>sext</tt>' sign extends <tt>value</tt> to the type <tt>ty2</tt>.</p>
2718
2719<h5>Arguments:</h5>
2720<p>
2721The '<tt>sext</tt>' instruction takes a value to cast, which must be of
2722<a href="#t_integral">integral</a> type, and a type to cast it to, which must
2723also be of <a href="#t_integral">integral</a> type.</p>
2724
2725<h5>Semantics:</h5>
2726<p>
2727The '<tt>sext</tt>' instruction performs a sign extension by copying the sign
2728bit (highest order bit) of the <tt>value</tt> until it reaches the bit size of
2729the type <tt>ty2</tt>. When the the operand and the type are the same size,
2730no bit filling is done and the cast is considered a <i>no-op cast</i> because
2731no bits change (only the type changes).</p>
2732
2733<p>When sign extending to bool, the extension is done as a comparison against
2734zero. If the <tt>value</tt> was zero, the bool result will be <tt>false</tt>.
2735If the <tt>value</tt> was non-zero, the bool result will be <tt>true</tt>.</p>
2736
2737<h5>Example:</h5>
2738
2739<pre>
2740 %X = sext sbyte -1 to ushort <i>; yields ushort:65535</i>
2741 %Y = sext bool true to int <i>; yields int:-1</i>
2742</pre>
2743</div>
2744
2745<!-- _______________________________________________________________________ -->
2746<div class="doc_subsubsection">
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame^]2747 <a name="i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a>
2748</div>
2749
2750<div class="doc_text">
2751
2752<h5>Syntax:</h5>
2753
2754<pre>
2755 &lt;result&gt; = fptrunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2756</pre>
2757
2758<h5>Overview:</h5>
2759<p>The '<tt>fptrunc</tt>' instruction truncates <tt>value</tt> to type
2760<tt>ty2</tt>.</p>
2761
2762
2763<h5>Arguments:</h5>
2764<p>The '<tt>fptrunc</tt>' instruction takes a <a href="#t_floating">floating
2765 point</a> value to cast and a <a href="#t_floating">floating point</a> type to
2766cast it to. The size of <tt>value</tt> must be larger than the size of
2767<tt>ty2</tt>. This implies that <tt>fptrunc</tt> cannot be used to make a
2768<i>no-op cast</i>.</p>
2769
2770<h5>Semantics:</h5>
2771<p> The '<tt>fptrunc</tt>' instruction converts a
2772<a href="#t_floating">floating point</a> value from a larger type to a smaller
2773type. If the value cannot fit within the destination type, <tt>ty2</tt>, then
2774the results are undefined.</p>
2775
2776<h5>Example:</h5>
2777<pre>
2778 %X = fptrunc double 123.0 to float <i>; yields float:123.0</i>
2779 %Y = fptrunc double 1.0E+300 to float <i>; yields undefined</i>
2780</pre>
2781</div>
2782
2783<!-- _______________________________________________________________________ -->
2784<div class="doc_subsubsection">
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2785 <a name="i_fpext">'<tt>fpext .. to</tt>' Instruction</a>
2786</div>
2787<div class="doc_text">
2788
2789<h5>Syntax:</h5>
2790<pre>
2791 &lt;result&gt; = fpext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2792</pre>
2793
2794<h5>Overview:</h5>
2795<p>The '<tt>fpext</tt>' extends a floating point <tt>value</tt> to a larger
2796floating point value.</p>
2797
2798<h5>Arguments:</h5>
2799<p>The '<tt>fpext</tt>' instruction takes a
2800<a href="#t_floating">floating point</a> <tt>value</tt> to cast,
2801and a <a href="#t_floating">floating point</a> type to cast it to.</p>
2802
2803<h5>Semantics:</h5>
2804<p>The '<tt>fpext</tt>' instruction extends the <tt>value</tt> from one floating
2805point type to another. If the type of the <tt>value</tt> and <tt>ty2</tt> are
2806the same, the instruction is considered a <i>no-op cast</i> because no bits
2807change.</p>
2808
2809<h5>Example:</h5>
2810<pre>
2811 %X = fpext float 3.1415 to double <i>; yields double:3.1415</i>
2812 %Y = fpext float 1.0 to float <i>; yields float:1.0 (no-op)</i>
2813</pre>
2814</div>
2815
2816<!-- _______________________________________________________________________ -->
2817<div class="doc_subsubsection">
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2818 <a name="i_fp2uint">'<tt>fp2uint .. to</tt>' Instruction</a>
2819</div>
2820<div class="doc_text">
2821
2822<h5>Syntax:</h5>
2823<pre>
2824 &lt;result&gt; = fp2uint &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2825</pre>
2826
2827<h5>Overview:</h5>
2828<p>The '<tt>fp2uint</tt>' converts a floating point <tt>value</tt> to its
2829unsigned integer equivalent of type <tt>ty2</tt>.
2830</p>
2831
2832<h5>Arguments:</h5>
2833<p>The '<tt>fp2uint</tt>' instruction takes a value to cast, which must be a
2834<a href="#t_floating">floating point</a> value, and a type to cast it to, which
2835must be an <a href="#t_integral">integral</a> type.</p>
2836
2837<h5>Semantics:</h5>
2838<p> The '<tt>fp2uint</tt>' instruction converts its
2839<a href="#t_floating">floating point</a> operand into the nearest (rounding
2840towards zero) unsigned integer value. If the value cannot fit in <tt>ty2</tt>,
2841the results are undefined.</p>
2842
2843<p>When converting to bool, the conversion is done as a comparison against
2844zero. If the <tt>value</tt> was zero, the bool result will be <tt>false</tt>.
2845If the <tt>value</tt> was non-zero, the bool result will be <tt>true</tt>.</p>
2846
2847<h5>Example:</h5>
2848<pre>
2849 %X = fp2uint double 123.0 to int <i>; yields int:123</i>
2850 %Y = fp2uint float 1.0E+300 to bool <i>; yields bool:true</i>
2851 %X = fp2uint float 1.04E+17 to ubyte <i>; yields undefined:1</i>
2852</pre>
2853</div>
2854
2855<!-- _______________________________________________________________________ -->
2856<div class="doc_subsubsection">
2857 <a name="i_fp2sint">'<tt>fp2sint .. to</tt>' Instruction</a>
2858</div>
2859<div class="doc_text">
2860
2861<h5>Syntax:</h5>
2862<pre>
2863 &lt;result&gt; = fp2sint &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2864</pre>
2865
2866<h5>Overview:</h5>
2867<p>The '<tt>fp2sint</tt>' instruction converts
2868<a href="#t_floating">floating point</a> <tt>value</tt> to type <tt>ty2</tt>.
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]2869</p>
2870
2871
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]2872<h5>Arguments:</h5>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2873<p> The '<tt>fp2sint</tt>' instruction takes a value to cast, which must be a
2874<a href="#t_floating">floating point</a> value, and a type to cast it to, which
2875must also be an <a href="#t_integral">integral</a> type.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]2876
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]2877<h5>Semantics:</h5>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2878<p>The '<tt>fp2sint</tt>' instruction converts its
2879<a href="#t_floating">floating point</a> operand into the nearest (rounding
2880towards zero) signed integer value. If the value cannot fit in <tt>ty2</tt>,
2881the results are undefined.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]2882
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2883<p>When converting to bool, the conversion is done as a comparison against
2884zero. If the <tt>value</tt> was zero, the bool result will be <tt>false</tt>.
2885If the <tt>value</tt> was non-zero, the bool result will be <tt>true</tt>.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]2886
Chris Lattner33ba0d92001-07-09 00:26:23 +0000[diff] [blame]2887<h5>Example:</h5>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]2888<pre>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2889 %X = fp2sint double -123.0 to int <i>; yields int:-123</i>
2890 %Y = fp2sint float 1.0E-247 to bool <i>; yields bool:true</i>
2891 %X = fp2sint float 1.04E+17 to sbyte <i>; yields undefined:1</i>
2892</pre>
2893</div>
2894
2895<!-- _______________________________________________________________________ -->
2896<div class="doc_subsubsection">
2897 <a name="i_uint2fp">'<tt>uint2fp .. to</tt>' Instruction</a>
2898</div>
2899<div class="doc_text">
2900
2901<h5>Syntax:</h5>
2902<pre>
2903 &lt;result&gt; = uint2fp &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2904</pre>
2905
2906<h5>Overview:</h5>
2907<p>The '<tt>uint2fp</tt>' instruction regards <tt>value</tt> as an unsigned
2908integer and converts that value to the <tt>ty2</tt> type.</p>
2909
2910
2911<h5>Arguments:</h5>
2912<p>The '<tt>uint2fp</tt>' instruction takes a value to cast, which must be an
2913<a href="#t_integral">integral</a> value, and a type to cast it to, which must
2914be a <a href="#t_floating">floating point</a> type.</p>
2915
2916<h5>Semantics:</h5>
2917<p>The '<tt>uint2fp</tt>' instruction interprets its operand as an unsigned
2918integer quantity and converts it to the corresponding floating point value. If
2919the value cannot fit in the floating point value, the results are undefined.</p>
2920
2921
2922<h5>Example:</h5>
2923<pre>
2924 %X = uint2fp int 257 to float <i>; yields float:257.0</i>
2925 %Y = uint2fp sbyte -1 to double <i>; yields double:255.0</i>
2926</pre>
2927</div>
2928
2929<!-- _______________________________________________________________________ -->
2930<div class="doc_subsubsection">
2931 <a name="i_sint2fp">'<tt>sint2fp .. to</tt>' Instruction</a>
2932</div>
2933<div class="doc_text">
2934
2935<h5>Syntax:</h5>
2936<pre>
2937 &lt;result&gt; = sint2fp &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2938</pre>
2939
2940<h5>Overview:</h5>
2941<p>The '<tt>sint2fp</tt>' instruction regards <tt>value</tt> as a signed
2942integer and converts that value to the <tt>ty2</tt> type.</p>
2943
2944<h5>Arguments:</h5>
2945<p>The '<tt>sint2fp</tt>' instruction takes a value to cast, which must be an
2946<a href="#t_integral">integral</a> value, and a type to cast it to, which must be
2947a <a href="#t_floating">floating point</a> type.</p>
2948
2949<h5>Semantics:</h5>
2950<p>The '<tt>sint2fp</tt>' instruction interprets its operand as a signed
2951integer quantity and converts it to the corresponding floating point value. If
2952the value cannot fit in the floating point value, the results are undefined.</p>
2953
2954<h5>Example:</h5>
2955<pre>
2956 %X = sint2fp int 257 to float <i>; yields float:257.0</i>
2957 %Y = sint2fp sbyte -1 to double <i>; yields double:-1.0</i>
2958</pre>
2959</div>
2960
2961<!-- _______________________________________________________________________ -->
2962<div class="doc_subsubsection">
2963 <a name="i_bitconvert">'<tt>bitconvert .. to</tt>' Instruction</a>
2964</div>
2965<div class="doc_text">
2966
2967<h5>Syntax:</h5>
2968<pre>
2969 &lt;result&gt; = bitconvert &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2970</pre>
2971
2972<h5>Overview:</h5>
2973<p>The '<tt>bitconvert</tt>' instruction converts <tt>value</tt> to type
2974<tt>ty2</tt> without changing any bits.</p>
2975
2976<h5>Arguments:</h5>
2977<p>The '<tt>bitconvert</tt>' instruction takes a value to cast, which must be
2978a first class value, and a type to cast it to, which must also be a <a
2979 href="#t_firstclass">first class</a> type. The bit sizes of <tt>value</tt>
2980and the destination type, <tt>ty2</tt>, must be identical.</p>
2981
2982<h5>Semantics:</h5>
2983<p>The '<tt>bitconvert</tt>' instruction converts <tt>value</tt> to type
2984<tt>ty2</tt> as if the value had been stored to memory and read back as type
2985<tt>ty2</tt>. That is, no bits are changed during the conversion. The
2986<tt>bitconvert</tt> instruction may be used to construct <i>no-op casts</i> that
2987the <tt>zext, sext, and fpext</tt> instructions do not permit.</p>
2988
2989<h5>Example:</h5>
2990<pre>
2991 %X = bitconvert ubyte 255 to sbyte <i>; yields sbyte:-1</i>
2992 %Y = bitconvert uint* %x to uint <i>; yields uint:%x</i>
Chris Lattner33ba0d92001-07-09 00:26:23 +0000[diff] [blame]2993</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2994</div>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]2995
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]2996<!-- ======================================================================= -->
2997<div class="doc_subsection"> <a name="otherops">Other Operations</a> </div>
2998<div class="doc_text">
2999<p>The instructions in this category are the "miscellaneous"
3000instructions, which defy better classification.</p>
3001</div>
3002<!-- _______________________________________________________________________ -->
3003<div class="doc_subsubsection"> <a name="i_phi">'<tt>phi</tt>'
3004Instruction</a> </div>
3005<div class="doc_text">
3006<h5>Syntax:</h5>
3007<pre> &lt;result&gt; = phi &lt;ty&gt; [ &lt;val0&gt;, &lt;label0&gt;], ...<br></pre>
3008<h5>Overview:</h5>
3009<p>The '<tt>phi</tt>' instruction is used to implement the &#966; node in
3010the SSA graph representing the function.</p>
3011<h5>Arguments:</h5>
3012<p>The type of the incoming values are specified with the first type
3013field. After this, the '<tt>phi</tt>' instruction takes a list of pairs
3014as arguments, with one pair for each predecessor basic block of the
3015current block. Only values of <a href="#t_firstclass">first class</a>
3016type may be used as the value arguments to the PHI node. Only labels
3017may be used as the label arguments.</p>
3018<p>There must be no non-phi instructions between the start of a basic
3019block and the PHI instructions: i.e. PHI instructions must be first in
3020a basic block.</p>
3021<h5>Semantics:</h5>
3022<p>At runtime, the '<tt>phi</tt>' instruction logically takes on the
3023value specified by the parameter, depending on which basic block we
3024came from in the last <a href="#terminators">terminator</a> instruction.</p>
3025<h5>Example:</h5>
3026<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>
3027</div>
3028
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3029<!-- _______________________________________________________________________ -->
3030<div class="doc_subsubsection">
3031 <a name="i_select">'<tt>select</tt>' Instruction</a>
3032</div>
3033
3034<div class="doc_text">
3035
3036<h5>Syntax:</h5>
3037
3038<pre>
3039 &lt;result&gt; = select bool &lt;cond&gt;, &lt;ty&gt; &lt;val1&gt;, &lt;ty&gt; &lt;val2&gt; <i>; yields ty</i>
3040</pre>
3041
3042<h5>Overview:</h5>
3043
3044<p>
3045The '<tt>select</tt>' instruction is used to choose one value based on a
3046condition, without branching.
3047</p>
3048
3049
3050<h5>Arguments:</h5>
3051
3052<p>
3053The '<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.
3054</p>
3055
3056<h5>Semantics:</h5>
3057
3058<p>
3059If the boolean condition evaluates to true, the instruction returns the first
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3060value argument; otherwise, it returns the second value argument.
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3061</p>
3062
3063<h5>Example:</h5>
3064
3065<pre>
3066 %X = select bool true, ubyte 17, ubyte 42 <i>; yields ubyte:17</i>
3067</pre>
3068</div>
3069
Robert Bocchino05ccd702006-01-15 20:48:27 +0000[diff] [blame]3070
3071<!-- _______________________________________________________________________ -->
3072<div class="doc_subsubsection">
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3073 <a name="i_call">'<tt>call</tt>' Instruction</a>
3074</div>
3075
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3076<div class="doc_text">
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3077
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3078<h5>Syntax:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3079<pre>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3080 &lt;result&gt; = [tail] call [<a href="#callingconv">cconv</a>] &lt;ty&gt;* &lt;fnptrval&gt;(&lt;param list&gt;)
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3081</pre>
3082
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3083<h5>Overview:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3084
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3085<p>The '<tt>call</tt>' instruction represents a simple function call.</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3086
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3087<h5>Arguments:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3088
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3089<p>This instruction requires several arguments:</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3090
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]3091<ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3092 <li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3093 <p>The optional "tail" marker indicates whether the callee function accesses
3094 any allocas or varargs in the caller. If the "tail" marker is present, the
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3095 function call is eligible for tail call optimization. Note that calls may
3096 be marked "tail" even if they do not occur before a <a
3097 href="#i_ret"><tt>ret</tt></a> instruction.
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3098 </li>
3099 <li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3100 <p>The optional "cconv" marker indicates which <a href="callingconv">calling
3101 convention</a> the call should use. If none is specified, the call defaults
3102 to using C calling conventions.
3103 </li>
3104 <li>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3105 <p>'<tt>ty</tt>': shall be the signature of the pointer to function value
3106 being invoked. The argument types must match the types implied by this
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3107 signature. This type can be omitted if the function is not varargs and
3108 if the function type does not return a pointer to a function.</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3109 </li>
3110 <li>
3111 <p>'<tt>fnptrval</tt>': An LLVM value containing a pointer to a function to
3112 be invoked. In most cases, this is a direct function invocation, but
3113 indirect <tt>call</tt>s are just as possible, calling an arbitrary pointer
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3114 to function value.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3115 </li>
3116 <li>
3117 <p>'<tt>function args</tt>': argument list whose types match the
Reid Spencera7e302a2005-05-01 22:22:57 +0000[diff] [blame]3118 function signature argument types. All arguments must be of
3119 <a href="#t_firstclass">first class</a> type. If the function signature
3120 indicates the function accepts a variable number of arguments, the extra
3121 arguments can be specified.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3122 </li>
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]3123</ol>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3124
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3125<h5>Semantics:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3126
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3127<p>The '<tt>call</tt>' instruction is used to cause control flow to
3128transfer to a specified function, with its incoming arguments bound to
3129the specified values. Upon a '<tt><a href="#i_ret">ret</a></tt>'
3130instruction in the called function, control flow continues with the
3131instruction after the function call, and the return value of the
3132function is bound to the result argument. This is a simpler case of
3133the <a href="#i_invoke">invoke</a> instruction.</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3134
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3135<h5>Example:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3136
3137<pre>
3138 %retval = call int %test(int %argc)
3139 call int(sbyte*, ...) *%printf(sbyte* %msg, int 12, sbyte 42);
3140 %X = tail call int %foo()
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3141 %Y = tail call <a href="#callingconv">fastcc</a> int %foo()
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3142</pre>
3143
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3144</div>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3145
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3146<!-- _______________________________________________________________________ -->
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3147<div class="doc_subsubsection">
Chris Lattnerfb6977d2006-01-13 23:26:01 +0000[diff] [blame]3148 <a name="i_va_arg">'<tt>va_arg</tt>' Instruction</a>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3149</div>
3150
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3151<div class="doc_text">
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3152
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]3153<h5>Syntax:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3154
3155<pre>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3156 &lt;resultval&gt; = va_arg &lt;va_list*&gt; &lt;arglist&gt;, &lt;argty&gt;
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3157</pre>
3158
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]3159<h5>Overview:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3160
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3161<p>The '<tt>va_arg</tt>' instruction is used to access arguments passed through
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3162the "variable argument" area of a function call. It is used to implement the
3163<tt>va_arg</tt> macro in C.</p>
3164
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]3165<h5>Arguments:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3166
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3167<p>This instruction takes a <tt>va_list*</tt> value and the type of
3168the argument. It returns a value of the specified argument type and
Jeff Cohen25d4f7e2005-11-11 02:15:27 +0000[diff] [blame]3169increments the <tt>va_list</tt> to point to the next argument. Again, the
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3170actual type of <tt>va_list</tt> is target specific.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3171
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]3172<h5>Semantics:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3173
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3174<p>The '<tt>va_arg</tt>' instruction loads an argument of the specified
3175type from the specified <tt>va_list</tt> and causes the
3176<tt>va_list</tt> to point to the next argument. For more information,
3177see the variable argument handling <a href="#int_varargs">Intrinsic
3178Functions</a>.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3179
3180<p>It is legal for this instruction to be called in a function which does not
3181take a variable number of arguments, for example, the <tt>vfprintf</tt>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3182function.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3183
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3184<p><tt>va_arg</tt> is an LLVM instruction instead of an <a
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3185href="#intrinsics">intrinsic function</a> because it takes a type as an
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3186argument.</p>
3187
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]3188<h5>Example:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3189
3190<p>See the <a href="#int_varargs">variable argument processing</a> section.</p>
3191
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3192</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3193
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3194<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3195<div class="doc_section"> <a name="intrinsics">Intrinsic Functions</a> </div>
3196<!-- *********************************************************************** -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3197
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3198<div class="doc_text">
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3199
3200<p>LLVM supports the notion of an "intrinsic function". These functions have
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3201well known names and semantics and are required to follow certain
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3202restrictions. Overall, these instructions represent an extension mechanism for
3203the LLVM language that does not require changing all of the transformations in
3204LLVM to add to the language (or the bytecode reader/writer, the parser,
3205etc...).</p>
3206
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3207<p>Intrinsic function names must all start with an "<tt>llvm.</tt>" prefix. This
3208prefix is reserved in LLVM for intrinsic names; thus, functions may not be named
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3209this. Intrinsic functions must always be external functions: you cannot define
3210the body of intrinsic functions. Intrinsic functions may only be used in call
3211or invoke instructions: it is illegal to take the address of an intrinsic
3212function. Additionally, because intrinsic functions are part of the LLVM
3213language, it is required that they all be documented here if any are added.</p>
3214
3215
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3216<p>To learn how to add an intrinsic function, please see the <a
Chris Lattner590cff32005-05-11 03:35:57 +0000[diff] [blame]3217href="ExtendingLLVM.html">Extending LLVM Guide</a>.
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3218</p>
3219
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3220</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3221
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3222<!-- ======================================================================= -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3223<div class="doc_subsection">
3224 <a name="int_varargs">Variable Argument Handling Intrinsics</a>
3225</div>
3226
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3227<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3228
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3229<p>Variable argument support is defined in LLVM with the <a
Chris Lattnerfb6977d2006-01-13 23:26:01 +0000[diff] [blame]3230 href="#i_va_arg"><tt>va_arg</tt></a> instruction and these three
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3231intrinsic functions. These functions are related to the similarly
3232named macros defined in the <tt>&lt;stdarg.h&gt;</tt> header file.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3233
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3234<p>All of these functions operate on arguments that use a
3235target-specific value type "<tt>va_list</tt>". The LLVM assembly
3236language reference manual does not define what this type is, so all
3237transformations should be prepared to handle intrinsics with any type
3238used.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3239
Chris Lattner374ab302006-05-15 17:26:46 +0000[diff] [blame]3240<p>This example shows how the <a href="#i_va_arg"><tt>va_arg</tt></a>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3241instruction and the variable argument handling intrinsic functions are
3242used.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3243
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3244<pre>
3245int %test(int %X, ...) {
3246 ; Initialize variable argument processing
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3247 %ap = alloca sbyte*
3248 call void %<a href="#i_va_start">llvm.va_start</a>(sbyte** %ap)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3249
3250 ; Read a single integer argument
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3251 %tmp = va_arg sbyte** %ap, int
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3252
3253 ; Demonstrate usage of llvm.va_copy and llvm.va_end
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3254 %aq = alloca sbyte*
Andrew Lenharthd0a4c622005-06-22 20:38:11 +0000[diff] [blame]3255 call void %<a href="#i_va_copy">llvm.va_copy</a>(sbyte** %aq, sbyte** %ap)
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3256 call void %<a href="#i_va_end">llvm.va_end</a>(sbyte** %aq)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3257
3258 ; Stop processing of arguments.
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3259 call void %<a href="#i_va_end">llvm.va_end</a>(sbyte** %ap)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3260 ret int %tmp
3261}
3262</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3263</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3264
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3265<!-- _______________________________________________________________________ -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3266<div class="doc_subsubsection">
3267 <a name="i_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a>
3268</div>
3269
3270
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3271<div class="doc_text">
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3272<h5>Syntax:</h5>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3273<pre> declare void %llvm.va_start(&lt;va_list&gt;* &lt;arglist&gt;)<br></pre>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3274<h5>Overview:</h5>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3275<P>The '<tt>llvm.va_start</tt>' intrinsic initializes
3276<tt>*&lt;arglist&gt;</tt> for subsequent use by <tt><a
3277href="#i_va_arg">va_arg</a></tt>.</p>
3278
3279<h5>Arguments:</h5>
3280
3281<P>The argument is a pointer to a <tt>va_list</tt> element to initialize.</p>
3282
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3283<h5>Semantics:</h5>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3284
3285<P>The '<tt>llvm.va_start</tt>' intrinsic works just like the <tt>va_start</tt>
3286macro available in C. In a target-dependent way, it initializes the
3287<tt>va_list</tt> element the argument points to, so that the next call to
3288<tt>va_arg</tt> will produce the first variable argument passed to the function.
3289Unlike the C <tt>va_start</tt> macro, this intrinsic does not need to know the
3290last argument of the function, the compiler can figure that out.</p>
3291
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3292</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3293
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3294<!-- _______________________________________________________________________ -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3295<div class="doc_subsubsection">
3296 <a name="i_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a>
3297</div>
3298
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3299<div class="doc_text">
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3300<h5>Syntax:</h5>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3301<pre> declare void %llvm.va_end(&lt;va_list*&gt; &lt;arglist&gt;)<br></pre>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3302<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3303<p>The '<tt>llvm.va_end</tt>' intrinsic destroys <tt>&lt;arglist&gt;</tt>
3304which has been initialized previously with <tt><a href="#i_va_start">llvm.va_start</a></tt>
3305or <tt><a href="#i_va_copy">llvm.va_copy</a></tt>.</p>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3306<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3307<p>The argument is a <tt>va_list</tt> to destroy.</p>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3308<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3309<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]3310macro available in C. In a target-dependent way, it destroys the <tt>va_list</tt>.
3311Calls to <a href="#i_va_start"><tt>llvm.va_start</tt></a> and <a
3312 href="#i_va_copy"><tt>llvm.va_copy</tt></a> must be matched exactly
3313with calls to <tt>llvm.va_end</tt>.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3314</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3315
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3316<!-- _______________________________________________________________________ -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3317<div class="doc_subsubsection">
3318 <a name="i_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a>
3319</div>
3320
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3321<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3322
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3323<h5>Syntax:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3324
3325<pre>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3326 declare void %llvm.va_copy(&lt;va_list&gt;* &lt;destarglist&gt;,
Andrew Lenharthd0a4c622005-06-22 20:38:11 +0000[diff] [blame]3327 &lt;va_list&gt;* &lt;srcarglist&gt;)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3328</pre>
3329
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3330<h5>Overview:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3331
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3332<p>The '<tt>llvm.va_copy</tt>' intrinsic copies the current argument position from
3333the source argument list to the destination argument list.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3334
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3335<h5>Arguments:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3336
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3337<p>The first argument is a pointer to a <tt>va_list</tt> element to initialize.
Andrew Lenharthd0a4c622005-06-22 20:38:11 +0000[diff] [blame]3338The second argument is a pointer to a <tt>va_list</tt> element to copy from.</p>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3339
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3340
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3341<h5>Semantics:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3342
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3343<p>The '<tt>llvm.va_copy</tt>' intrinsic works just like the <tt>va_copy</tt> macro
3344available in C. In a target-dependent way, it copies the source
3345<tt>va_list</tt> element into the destination list. This intrinsic is necessary
3346because the <tt><a href="i_va_begin">llvm.va_begin</a></tt> intrinsic may be
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3347arbitrarily complex and require memory allocation, for example.</p>
3348
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3349</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3350
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3351<!-- ======================================================================= -->
3352<div class="doc_subsection">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3353 <a name="int_gc">Accurate Garbage Collection Intrinsics</a>
3354</div>
3355
3356<div class="doc_text">
3357
3358<p>
3359LLVM support for <a href="GarbageCollection.html">Accurate Garbage
3360Collection</a> requires the implementation and generation of these intrinsics.
3361These intrinsics allow identification of <a href="#i_gcroot">GC roots on the
3362stack</a>, as well as garbage collector implementations that require <a
3363href="#i_gcread">read</a> and <a href="#i_gcwrite">write</a> barriers.
3364Front-ends for type-safe garbage collected languages should generate these
3365intrinsics to make use of the LLVM garbage collectors. For more details, see <a
3366href="GarbageCollection.html">Accurate Garbage Collection with LLVM</a>.
3367</p>
3368</div>
3369
3370<!-- _______________________________________________________________________ -->
3371<div class="doc_subsubsection">
3372 <a name="i_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a>
3373</div>
3374
3375<div class="doc_text">
3376
3377<h5>Syntax:</h5>
3378
3379<pre>
Reid Spencera8d451e2005-04-26 20:50:44 +0000[diff] [blame]3380 declare void %llvm.gcroot(&lt;ty&gt;** %ptrloc, &lt;ty2&gt;* %metadata)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3381</pre>
3382
3383<h5>Overview:</h5>
3384
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]3385<p>The '<tt>llvm.gcroot</tt>' intrinsic declares the existence of a GC root to
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3386the code generator, and allows some metadata to be associated with it.</p>
3387
3388<h5>Arguments:</h5>
3389
3390<p>The first argument specifies the address of a stack object that contains the
3391root pointer. The second pointer (which must be either a constant or a global
3392value address) contains the meta-data to be associated with the root.</p>
3393
3394<h5>Semantics:</h5>
3395
3396<p>At runtime, a call to this intrinsics stores a null pointer into the "ptrloc"
3397location. At compile-time, the code generator generates information to allow
3398the runtime to find the pointer at GC safe points.
3399</p>
3400
3401</div>
3402
3403
3404<!-- _______________________________________________________________________ -->
3405<div class="doc_subsubsection">
3406 <a name="i_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a>
3407</div>
3408
3409<div class="doc_text">
3410
3411<h5>Syntax:</h5>
3412
3413<pre>
Chris Lattner80626e92006-03-14 20:02:51 +0000[diff] [blame]3414 declare sbyte* %llvm.gcread(sbyte* %ObjPtr, sbyte** %Ptr)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3415</pre>
3416
3417<h5>Overview:</h5>
3418
3419<p>The '<tt>llvm.gcread</tt>' intrinsic identifies reads of references from heap
3420locations, allowing garbage collector implementations that require read
3421barriers.</p>
3422
3423<h5>Arguments:</h5>
3424
Chris Lattner80626e92006-03-14 20:02:51 +0000[diff] [blame]3425<p>The second argument is the address to read from, which should be an address
3426allocated from the garbage collector. The first object is a pointer to the
3427start of the referenced object, if needed by the language runtime (otherwise
3428null).</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3429
3430<h5>Semantics:</h5>
3431
3432<p>The '<tt>llvm.gcread</tt>' intrinsic has the same semantics as a load
3433instruction, but may be replaced with substantially more complex code by the
3434garbage collector runtime, as needed.</p>
3435
3436</div>
3437
3438
3439<!-- _______________________________________________________________________ -->
3440<div class="doc_subsubsection">
3441 <a name="i_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a>
3442</div>
3443
3444<div class="doc_text">
3445
3446<h5>Syntax:</h5>
3447
3448<pre>
Chris Lattner80626e92006-03-14 20:02:51 +0000[diff] [blame]3449 declare void %llvm.gcwrite(sbyte* %P1, sbyte* %Obj, sbyte** %P2)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3450</pre>
3451
3452<h5>Overview:</h5>
3453
3454<p>The '<tt>llvm.gcwrite</tt>' intrinsic identifies writes of references to heap
3455locations, allowing garbage collector implementations that require write
3456barriers (such as generational or reference counting collectors).</p>
3457
3458<h5>Arguments:</h5>
3459
Chris Lattner80626e92006-03-14 20:02:51 +0000[diff] [blame]3460<p>The first argument is the reference to store, the second is the start of the
3461object to store it to, and the third is the address of the field of Obj to
3462store to. If the runtime does not require a pointer to the object, Obj may be
3463null.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3464
3465<h5>Semantics:</h5>
3466
3467<p>The '<tt>llvm.gcwrite</tt>' intrinsic has the same semantics as a store
3468instruction, but may be replaced with substantially more complex code by the
3469garbage collector runtime, as needed.</p>
3470
3471</div>
3472
3473
3474
3475<!-- ======================================================================= -->
3476<div class="doc_subsection">
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3477 <a name="int_codegen">Code Generator Intrinsics</a>
3478</div>
3479
3480<div class="doc_text">
3481<p>
3482These intrinsics are provided by LLVM to expose special features that may only
3483be implemented with code generator support.
3484</p>
3485
3486</div>
3487
3488<!-- _______________________________________________________________________ -->
3489<div class="doc_subsubsection">
3490 <a name="i_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a>
3491</div>
3492
3493<div class="doc_text">
3494
3495<h5>Syntax:</h5>
3496<pre>
Chris Lattnerfcf39d42006-01-13 01:20:27 +0000[diff] [blame]3497 declare sbyte *%llvm.returnaddress(uint &lt;level&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3498</pre>
3499
3500<h5>Overview:</h5>
3501
3502<p>
Chris Lattner32b5d712006-10-15 20:05:59 +0000[diff] [blame]3503The '<tt>llvm.returnaddress</tt>' intrinsic attempts to compute a
3504target-specific value indicating the return address of the current function
3505or one of its callers.
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3506</p>
3507
3508<h5>Arguments:</h5>
3509
3510<p>
3511The argument to this intrinsic indicates which function to return the address
3512for. Zero indicates the calling function, one indicates its caller, etc. The
3513argument is <b>required</b> to be a constant integer value.
3514</p>
3515
3516<h5>Semantics:</h5>
3517
3518<p>
3519The '<tt>llvm.returnaddress</tt>' intrinsic either returns a pointer indicating
3520the return address of the specified call frame, or zero if it cannot be
3521identified. The value returned by this intrinsic is likely to be incorrect or 0
3522for arguments other than zero, so it should only be used for debugging purposes.
3523</p>
3524
3525<p>
3526Note that calling this intrinsic does not prevent function inlining or other
Chris Lattnerb40bb382005-03-07 20:30:51 +0000[diff] [blame]3527aggressive transformations, so the value returned may not be that of the obvious
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3528source-language caller.
3529</p>
3530</div>
3531
3532
3533<!-- _______________________________________________________________________ -->
3534<div class="doc_subsubsection">
3535 <a name="i_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a>
3536</div>
3537
3538<div class="doc_text">
3539
3540<h5>Syntax:</h5>
3541<pre>
Chris Lattnerfcf39d42006-01-13 01:20:27 +0000[diff] [blame]3542 declare sbyte *%llvm.frameaddress(uint &lt;level&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3543</pre>
3544
3545<h5>Overview:</h5>
3546
3547<p>
Chris Lattner32b5d712006-10-15 20:05:59 +0000[diff] [blame]3548The '<tt>llvm.frameaddress</tt>' intrinsic attempts to return the
3549target-specific frame pointer value for the specified stack frame.
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3550</p>
3551
3552<h5>Arguments:</h5>
3553
3554<p>
3555The argument to this intrinsic indicates which function to return the frame
3556pointer for. Zero indicates the calling function, one indicates its caller,
3557etc. The argument is <b>required</b> to be a constant integer value.
3558</p>
3559
3560<h5>Semantics:</h5>
3561
3562<p>
3563The '<tt>llvm.frameaddress</tt>' intrinsic either returns a pointer indicating
3564the frame address of the specified call frame, or zero if it cannot be
3565identified. The value returned by this intrinsic is likely to be incorrect or 0
3566for arguments other than zero, so it should only be used for debugging purposes.
3567</p>
3568
3569<p>
3570Note that calling this intrinsic does not prevent function inlining or other
Chris Lattnerb40bb382005-03-07 20:30:51 +0000[diff] [blame]3571aggressive transformations, so the value returned may not be that of the obvious
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3572source-language caller.
3573</p>
3574</div>
3575
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]3576<!-- _______________________________________________________________________ -->
3577<div class="doc_subsubsection">
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]3578 <a name="i_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a>
3579</div>
3580
3581<div class="doc_text">
3582
3583<h5>Syntax:</h5>
3584<pre>
3585 declare sbyte *%llvm.stacksave()
3586</pre>
3587
3588<h5>Overview:</h5>
3589
3590<p>
3591The '<tt>llvm.stacksave</tt>' intrinsic is used to remember the current state of
3592the function stack, for use with <a href="#i_stackrestore">
3593<tt>llvm.stackrestore</tt></a>. This is useful for implementing language
3594features like scoped automatic variable sized arrays in C99.
3595</p>
3596
3597<h5>Semantics:</h5>
3598
3599<p>
3600This intrinsic returns a opaque pointer value that can be passed to <a
3601href="#i_stackrestore"><tt>llvm.stackrestore</tt></a>. When an
3602<tt>llvm.stackrestore</tt> intrinsic is executed with a value saved from
3603<tt>llvm.stacksave</tt>, it effectively restores the state of the stack to the
3604state it was in when the <tt>llvm.stacksave</tt> intrinsic executed. In
3605practice, this pops any <a href="#i_alloca">alloca</a> blocks from the stack
3606that were allocated after the <tt>llvm.stacksave</tt> was executed.
3607</p>
3608
3609</div>
3610
3611<!-- _______________________________________________________________________ -->
3612<div class="doc_subsubsection">
3613 <a name="i_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a>
3614</div>
3615
3616<div class="doc_text">
3617
3618<h5>Syntax:</h5>
3619<pre>
3620 declare void %llvm.stackrestore(sbyte* %ptr)
3621</pre>
3622
3623<h5>Overview:</h5>
3624
3625<p>
3626The '<tt>llvm.stackrestore</tt>' intrinsic is used to restore the state of
3627the function stack to the state it was in when the corresponding <a
3628href="#llvm.stacksave"><tt>llvm.stacksave</tt></a> intrinsic executed. This is
3629useful for implementing language features like scoped automatic variable sized
3630arrays in C99.
3631</p>
3632
3633<h5>Semantics:</h5>
3634
3635<p>
3636See the description for <a href="#i_stacksave"><tt>llvm.stacksave</tt></a>.
3637</p>
3638
3639</div>
3640
3641
3642<!-- _______________________________________________________________________ -->
3643<div class="doc_subsubsection">
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]3644 <a name="i_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a>
3645</div>
3646
3647<div class="doc_text">
3648
3649<h5>Syntax:</h5>
3650<pre>
Reid Spencera8d451e2005-04-26 20:50:44 +0000[diff] [blame]3651 declare void %llvm.prefetch(sbyte * &lt;address&gt;,
3652 uint &lt;rw&gt;, uint &lt;locality&gt;)
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]3653</pre>
3654
3655<h5>Overview:</h5>
3656
3657
3658<p>
3659The '<tt>llvm.prefetch</tt>' intrinsic is a hint to the code generator to insert
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3660a prefetch instruction if supported; otherwise, it is a noop. Prefetches have
3661no
3662effect on the behavior of the program but can change its performance
Chris Lattner2a615362005-02-28 19:47:14 +0000[diff] [blame]3663characteristics.
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]3664</p>
3665
3666<h5>Arguments:</h5>
3667
3668<p>
3669<tt>address</tt> is the address to be prefetched, <tt>rw</tt> is the specifier
3670determining if the fetch should be for a read (0) or write (1), and
3671<tt>locality</tt> is a temporal locality specifier ranging from (0) - no
Chris Lattneraeffb4a2005-03-07 20:31:38 +0000[diff] [blame]3672locality, to (3) - extremely local keep in cache. The <tt>rw</tt> and
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]3673<tt>locality</tt> arguments must be constant integers.
3674</p>
3675
3676<h5>Semantics:</h5>
3677
3678<p>
3679This intrinsic does not modify the behavior of the program. In particular,
3680prefetches cannot trap and do not produce a value. On targets that support this
3681intrinsic, the prefetch can provide hints to the processor cache for better
3682performance.
3683</p>
3684
3685</div>
3686
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]3687<!-- _______________________________________________________________________ -->
3688<div class="doc_subsubsection">
3689 <a name="i_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a>
3690</div>
3691
3692<div class="doc_text">
3693
3694<h5>Syntax:</h5>
3695<pre>
Reid Spencera8d451e2005-04-26 20:50:44 +0000[diff] [blame]3696 declare void %llvm.pcmarker( uint &lt;id&gt; )
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]3697</pre>
3698
3699<h5>Overview:</h5>
3700
3701
3702<p>
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3703The '<tt>llvm.pcmarker</tt>' intrinsic is a method to export a Program Counter
3704(PC) in a region of
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]3705code to simulators and other tools. The method is target specific, but it is
3706expected that the marker will use exported symbols to transmit the PC of the marker.
Jeff Cohen25d4f7e2005-11-11 02:15:27 +0000[diff] [blame]3707The marker makes no guarantees that it will remain with any specific instruction
Chris Lattnerd07c3f42005-11-15 06:07:55 +0000[diff] [blame]3708after optimizations. It is possible that the presence of a marker will inhibit
Chris Lattnerb3e7afd2006-03-24 07:16:10 +0000[diff] [blame]3709optimizations. The intended use is to be inserted after optimizations to allow
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3710correlations of simulation runs.
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]3711</p>
3712
3713<h5>Arguments:</h5>
3714
3715<p>
3716<tt>id</tt> is a numerical id identifying the marker.
3717</p>
3718
3719<h5>Semantics:</h5>
3720
3721<p>
3722This intrinsic does not modify the behavior of the program. Backends that do not
3723support this intrinisic may ignore it.
3724</p>
3725
3726</div>
3727
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]3728<!-- _______________________________________________________________________ -->
3729<div class="doc_subsubsection">
3730 <a name="i_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a>
3731</div>
3732
3733<div class="doc_text">
3734
3735<h5>Syntax:</h5>
3736<pre>
3737 declare ulong %llvm.readcyclecounter( )
3738</pre>
3739
3740<h5>Overview:</h5>
3741
3742
3743<p>
3744The '<tt>llvm.readcyclecounter</tt>' intrinsic provides access to the cycle
3745counter register (or similar low latency, high accuracy clocks) on those targets
3746that support it. On X86, it should map to RDTSC. On Alpha, it should map to RPCC.
3747As the backing counters overflow quickly (on the order of 9 seconds on alpha), this
3748should only be used for small timings.
3749</p>
3750
3751<h5>Semantics:</h5>
3752
3753<p>
3754When directly supported, reading the cycle counter should not modify any memory.
3755Implementations are allowed to either return a application specific value or a
3756system wide value. On backends without support, this is lowered to a constant 0.
3757</p>
3758
3759</div>
3760
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3761<!-- ======================================================================= -->
3762<div class="doc_subsection">
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3763 <a name="int_libc">Standard C Library Intrinsics</a>
3764</div>
3765
3766<div class="doc_text">
3767<p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3768LLVM provides intrinsics for a few important standard C library functions.
3769These intrinsics allow source-language front-ends to pass information about the
3770alignment of the pointer arguments to the code generator, providing opportunity
3771for more efficient code generation.
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3772</p>
3773
3774</div>
3775
3776<!-- _______________________________________________________________________ -->
3777<div class="doc_subsubsection">
3778 <a name="i_memcpy">'<tt>llvm.memcpy</tt>' Intrinsic</a>
3779</div>
3780
3781<div class="doc_text">
3782
3783<h5>Syntax:</h5>
3784<pre>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]3785 declare void %llvm.memcpy.i32(sbyte* &lt;dest&gt;, sbyte* &lt;src&gt;,
3786 uint &lt;len&gt;, uint &lt;align&gt;)
3787 declare void %llvm.memcpy.i64(sbyte* &lt;dest&gt;, sbyte* &lt;src&gt;,
3788 ulong &lt;len&gt;, uint &lt;align&gt;)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3789</pre>
3790
3791<h5>Overview:</h5>
3792
3793<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]3794The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3795location to the destination location.
3796</p>
3797
3798<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]3799Note that, unlike the standard libc function, the <tt>llvm.memcpy.*</tt>
3800intrinsics do not return a value, and takes an extra alignment argument.
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3801</p>
3802
3803<h5>Arguments:</h5>
3804
3805<p>
3806The first argument is a pointer to the destination, the second is a pointer to
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]3807the source. The third argument is an integer argument
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3808specifying the number of bytes to copy, and the fourth argument is the alignment
3809of the source and destination locations.
3810</p>
3811
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]3812<p>
3813If the call to this intrinisic has an alignment value that is not 0 or 1, then
Chris Lattnerf0afc2c2006-03-04 00:02:10 +0000[diff] [blame]3814the caller guarantees that both the source and destination pointers are aligned
3815to that boundary.
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]3816</p>
3817
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3818<h5>Semantics:</h5>
3819
3820<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]3821The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3822location to the destination location, which are not allowed to overlap. It
3823copies "len" bytes of memory over. If the argument is known to be aligned to
3824some boundary, this can be specified as the fourth argument, otherwise it should
3825be set to 0 or 1.
3826</p>
3827</div>
3828
3829
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]3830<!-- _______________________________________________________________________ -->
3831<div class="doc_subsubsection">
3832 <a name="i_memmove">'<tt>llvm.memmove</tt>' Intrinsic</a>
3833</div>
3834
3835<div class="doc_text">
3836
3837<h5>Syntax:</h5>
3838<pre>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]3839 declare void %llvm.memmove.i32(sbyte* &lt;dest&gt;, sbyte* &lt;src&gt;,
3840 uint &lt;len&gt;, uint &lt;align&gt;)
3841 declare void %llvm.memmove.i64(sbyte* &lt;dest&gt;, sbyte* &lt;src&gt;,
3842 ulong &lt;len&gt;, uint &lt;align&gt;)
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]3843</pre>
3844
3845<h5>Overview:</h5>
3846
3847<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]3848The '<tt>llvm.memmove.*</tt>' intrinsics move a block of memory from the source
3849location to the destination location. It is similar to the
3850'<tt>llvm.memcmp</tt>' intrinsic but allows the two memory locations to overlap.
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]3851</p>
3852
3853<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]3854Note that, unlike the standard libc function, the <tt>llvm.memmove.*</tt>
3855intrinsics do not return a value, and takes an extra alignment argument.
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]3856</p>
3857
3858<h5>Arguments:</h5>
3859
3860<p>
3861The first argument is a pointer to the destination, the second is a pointer to
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]3862the source. The third argument is an integer argument
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]3863specifying the number of bytes to copy, and the fourth argument is the alignment
3864of the source and destination locations.
3865</p>
3866
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]3867<p>
3868If the call to this intrinisic has an alignment value that is not 0 or 1, then
Chris Lattnerf0afc2c2006-03-04 00:02:10 +0000[diff] [blame]3869the caller guarantees that the source and destination pointers are aligned to
3870that boundary.
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]3871</p>
3872
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]3873<h5>Semantics:</h5>
3874
3875<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]3876The '<tt>llvm.memmove.*</tt>' intrinsics copy a block of memory from the source
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]3877location to the destination location, which may overlap. It
3878copies "len" bytes of memory over. If the argument is known to be aligned to
3879some boundary, this can be specified as the fourth argument, otherwise it should
3880be set to 0 or 1.
3881</p>
3882</div>
3883
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3884
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3885<!-- _______________________________________________________________________ -->
3886<div class="doc_subsubsection">
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]3887 <a name="i_memset">'<tt>llvm.memset.*</tt>' Intrinsics</a>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3888</div>
3889
3890<div class="doc_text">
3891
3892<h5>Syntax:</h5>
3893<pre>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]3894 declare void %llvm.memset.i32(sbyte* &lt;dest&gt;, ubyte &lt;val&gt;,
3895 uint &lt;len&gt;, uint &lt;align&gt;)
3896 declare void %llvm.memset.i64(sbyte* &lt;dest&gt;, ubyte &lt;val&gt;,
3897 ulong &lt;len&gt;, uint &lt;align&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3898</pre>
3899
3900<h5>Overview:</h5>
3901
3902<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]3903The '<tt>llvm.memset.*</tt>' intrinsics fill a block of memory with a particular
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3904byte value.
3905</p>
3906
3907<p>
3908Note that, unlike the standard libc function, the <tt>llvm.memset</tt> intrinsic
3909does not return a value, and takes an extra alignment argument.
3910</p>
3911
3912<h5>Arguments:</h5>
3913
3914<p>
3915The first argument is a pointer to the destination to fill, the second is the
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]3916byte value to fill it with, the third argument is an integer
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3917argument specifying the number of bytes to fill, and the fourth argument is the
3918known alignment of destination location.
3919</p>
3920
3921<p>
3922If the call to this intrinisic has an alignment value that is not 0 or 1, then
Chris Lattnerf0afc2c2006-03-04 00:02:10 +0000[diff] [blame]3923the caller guarantees that the destination pointer is aligned to that boundary.
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3924</p>
3925
3926<h5>Semantics:</h5>
3927
3928<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]3929The '<tt>llvm.memset.*</tt>' intrinsics fill "len" bytes of memory starting at
3930the
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3931destination location. If the argument is known to be aligned to some boundary,
3932this can be specified as the fourth argument, otherwise it should be set to 0 or
39331.
3934</p>
3935</div>
3936
3937
Chris Lattner32006282004-06-11 02:28:03 +0000[diff] [blame]3938<!-- _______________________________________________________________________ -->
3939<div class="doc_subsubsection">
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]3940 <a name="i_isunordered">'<tt>llvm.isunordered.*</tt>' Intrinsic</a>
Alkis Evlogimenos26bbe932004-06-13 01:16:15 +0000[diff] [blame]3941</div>
3942
3943<div class="doc_text">
3944
3945<h5>Syntax:</h5>
3946<pre>
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]3947 declare bool %llvm.isunordered.f32(float Val1, float Val2)
3948 declare bool %llvm.isunordered.f64(double Val1, double Val2)
Alkis Evlogimenos26bbe932004-06-13 01:16:15 +0000[diff] [blame]3949</pre>
3950
3951<h5>Overview:</h5>
3952
3953<p>
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]3954The '<tt>llvm.isunordered</tt>' intrinsics return true if either or both of the
Alkis Evlogimenos26bbe932004-06-13 01:16:15 +0000[diff] [blame]3955specified floating point values is a NAN.
3956</p>
3957
3958<h5>Arguments:</h5>
3959
3960<p>
3961The arguments are floating point numbers of the same type.
3962</p>
3963
3964<h5>Semantics:</h5>
3965
3966<p>
3967If either or both of the arguments is a SNAN or QNAN, it returns true, otherwise
3968false.
3969</p>
3970</div>
3971
3972
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]3973<!-- _______________________________________________________________________ -->
3974<div class="doc_subsubsection">
Chris Lattnerec6cb612006-01-16 22:38:59 +0000[diff] [blame]3975 <a name="i_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a>
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]3976</div>
3977
3978<div class="doc_text">
3979
3980<h5>Syntax:</h5>
3981<pre>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]3982 declare float %llvm.sqrt.f32(float %Val)
3983 declare double %llvm.sqrt.f64(double %Val)
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]3984</pre>
3985
3986<h5>Overview:</h5>
3987
3988<p>
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]3989The '<tt>llvm.sqrt</tt>' intrinsics return the sqrt of the specified operand,
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]3990returning the same value as the libm '<tt>sqrt</tt>' function would. Unlike
3991<tt>sqrt</tt> in libm, however, <tt>llvm.sqrt</tt> has undefined behavior for
3992negative numbers (which allows for better optimization).
3993</p>
3994
3995<h5>Arguments:</h5>
3996
3997<p>
3998The argument and return value are floating point numbers of the same type.
3999</p>
4000
4001<h5>Semantics:</h5>
4002
4003<p>
4004This function returns the sqrt of the specified operand if it is a positive
4005floating point number.
4006</p>
4007</div>
4008
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]4009<!-- _______________________________________________________________________ -->
4010<div class="doc_subsubsection">
4011 <a name="i_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a>
4012</div>
4013
4014<div class="doc_text">
4015
4016<h5>Syntax:</h5>
4017<pre>
4018 declare float %llvm.powi.f32(float %Val, int %power)
4019 declare double %llvm.powi.f64(double %Val, int %power)
4020</pre>
4021
4022<h5>Overview:</h5>
4023
4024<p>
4025The '<tt>llvm.powi.*</tt>' intrinsics return the first operand raised to the
4026specified (positive or negative) power. The order of evaluation of
4027multiplications is not defined.
4028</p>
4029
4030<h5>Arguments:</h5>
4031
4032<p>
4033The second argument is an integer power, and the first is a value to raise to
4034that power.
4035</p>
4036
4037<h5>Semantics:</h5>
4038
4039<p>
4040This function returns the first value raised to the second power with an
4041unspecified sequence of rounding operations.</p>
4042</div>
4043
4044
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4045<!-- ======================================================================= -->
4046<div class="doc_subsection">
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]4047 <a name="int_manip">Bit Manipulation Intrinsics</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4048</div>
4049
4050<div class="doc_text">
4051<p>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]4052LLVM provides intrinsics for a few important bit manipulation operations.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4053These allow efficient code generation for some algorithms.
4054</p>
4055
4056</div>
4057
4058<!-- _______________________________________________________________________ -->
4059<div class="doc_subsubsection">
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]4060 <a name="i_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a>
4061</div>
4062
4063<div class="doc_text">
4064
4065<h5>Syntax:</h5>
4066<pre>
Chris Lattnerec6cb612006-01-16 22:38:59 +0000[diff] [blame]4067 declare ushort %llvm.bswap.i16(ushort &lt;id&gt;)
4068 declare uint %llvm.bswap.i32(uint &lt;id&gt;)
4069 declare ulong %llvm.bswap.i64(ulong &lt;id&gt;)
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]4070</pre>
4071
4072<h5>Overview:</h5>
4073
4074<p>
4075The '<tt>llvm.bwsap</tt>' family of intrinsics is used to byteswap a 16, 32 or
407664 bit quantity. These are useful for performing operations on data that is not
4077in the target's native byte order.
4078</p>
4079
4080<h5>Semantics:</h5>
4081
4082<p>
Chris Lattnerec6cb612006-01-16 22:38:59 +0000[diff] [blame]4083The <tt>llvm.bswap.16</tt> intrinsic returns a ushort value that has the high and low
4084byte of the input ushort swapped. Similarly, the <tt>llvm.bswap.i32</tt> intrinsic
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]4085returns a uint value that has the four bytes of the input uint swapped, so that
4086if the input bytes are numbered 0, 1, 2, 3 then the returned uint will have its
Chris Lattnerec6cb612006-01-16 22:38:59 +0000[diff] [blame]4087bytes in 3, 2, 1, 0 order. The <tt>llvm.bswap.i64</tt> intrinsic extends this concept
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]4088to 64 bits.
4089</p>
4090
4091</div>
4092
4093<!-- _______________________________________________________________________ -->
4094<div class="doc_subsubsection">
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]4095 <a name="int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4096</div>
4097
4098<div class="doc_text">
4099
4100<h5>Syntax:</h5>
4101<pre>
Chris Lattnerec6cb612006-01-16 22:38:59 +0000[diff] [blame]4102 declare ubyte %llvm.ctpop.i8 (ubyte &lt;src&gt;)
4103 declare ushort %llvm.ctpop.i16(ushort &lt;src&gt;)
4104 declare uint %llvm.ctpop.i32(uint &lt;src&gt;)
4105 declare ulong %llvm.ctpop.i64(ulong &lt;src&gt;)
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4106</pre>
4107
4108<h5>Overview:</h5>
4109
4110<p>
Chris Lattnerec6cb612006-01-16 22:38:59 +0000[diff] [blame]4111The '<tt>llvm.ctpop</tt>' family of intrinsics counts the number of bits set in a
4112value.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4113</p>
4114
4115<h5>Arguments:</h5>
4116
4117<p>
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]4118The only argument is the value to be counted. The argument may be of any
Chris Lattnerec6cb612006-01-16 22:38:59 +0000[diff] [blame]4119unsigned integer type. The return type must match the argument type.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4120</p>
4121
4122<h5>Semantics:</h5>
4123
4124<p>
4125The '<tt>llvm.ctpop</tt>' intrinsic counts the 1's in a variable.
4126</p>
4127</div>
4128
4129<!-- _______________________________________________________________________ -->
4130<div class="doc_subsubsection">
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]4131 <a name="int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4132</div>
4133
4134<div class="doc_text">
4135
4136<h5>Syntax:</h5>
4137<pre>
Chris Lattnerec6cb612006-01-16 22:38:59 +0000[diff] [blame]4138 declare ubyte %llvm.ctlz.i8 (ubyte &lt;src&gt;)
4139 declare ushort %llvm.ctlz.i16(ushort &lt;src&gt;)
4140 declare uint %llvm.ctlz.i32(uint &lt;src&gt;)
4141 declare ulong %llvm.ctlz.i64(ulong &lt;src&gt;)
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4142</pre>
4143
4144<h5>Overview:</h5>
4145
4146<p>
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]4147The '<tt>llvm.ctlz</tt>' family of intrinsic functions counts the number of
4148leading zeros in a variable.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4149</p>
4150
4151<h5>Arguments:</h5>
4152
4153<p>
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]4154The only argument is the value to be counted. The argument may be of any
Chris Lattnerec6cb612006-01-16 22:38:59 +0000[diff] [blame]4155unsigned integer type. The return type must match the argument type.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4156</p>
4157
4158<h5>Semantics:</h5>
4159
4160<p>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4161The '<tt>llvm.ctlz</tt>' intrinsic counts the leading (most significant) zeros
4162in a variable. If the src == 0 then the result is the size in bits of the type
Chris Lattner99d3c272006-04-21 21:37:40 +0000[diff] [blame]4163of src. For example, <tt>llvm.ctlz(int 2) = 30</tt>.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4164</p>
4165</div>
Chris Lattner32006282004-06-11 02:28:03 +0000[diff] [blame]4166
4167
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4168
4169<!-- _______________________________________________________________________ -->
4170<div class="doc_subsubsection">
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]4171 <a name="int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic</a>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4172</div>
4173
4174<div class="doc_text">
4175
4176<h5>Syntax:</h5>
4177<pre>
Chris Lattnerec6cb612006-01-16 22:38:59 +0000[diff] [blame]4178 declare ubyte %llvm.cttz.i8 (ubyte &lt;src&gt;)
4179 declare ushort %llvm.cttz.i16(ushort &lt;src&gt;)
4180 declare uint %llvm.cttz.i32(uint &lt;src&gt;)
4181 declare ulong %llvm.cttz.i64(ulong &lt;src&gt;)
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4182</pre>
4183
4184<h5>Overview:</h5>
4185
4186<p>
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]4187The '<tt>llvm.cttz</tt>' family of intrinsic functions counts the number of
4188trailing zeros.
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4189</p>
4190
4191<h5>Arguments:</h5>
4192
4193<p>
4194The only argument is the value to be counted. The argument may be of any
Chris Lattnerec6cb612006-01-16 22:38:59 +0000[diff] [blame]4195unsigned integer type. The return type must match the argument type.
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4196</p>
4197
4198<h5>Semantics:</h5>
4199
4200<p>
4201The '<tt>llvm.cttz</tt>' intrinsic counts the trailing (least significant) zeros
4202in a variable. If the src == 0 then the result is the size in bits of the type
4203of src. For example, <tt>llvm.cttz(2) = 1</tt>.
4204</p>
4205</div>
4206
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]4207<!-- ======================================================================= -->
4208<div class="doc_subsection">
4209 <a name="int_debugger">Debugger Intrinsics</a>
4210</div>
4211
4212<div class="doc_text">
4213<p>
4214The LLVM debugger intrinsics (which all start with <tt>llvm.dbg.</tt> prefix),
4215are described in the <a
4216href="SourceLevelDebugging.html#format_common_intrinsics">LLVM Source Level
4217Debugging</a> document.
4218</p>
4219</div>
4220
4221
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4222<!-- *********************************************************************** -->
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4223<hr>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]4224<address>
4225 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
4226 src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
4227 <a href="http://validator.w3.org/check/referer"><img
4228 src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!" /></a>
4229
4230 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
Reid Spencer05fe4b02006-03-14 05:39:39 +0000[diff] [blame]4231 <a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]4232 Last modified: $Date$
4233</address>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4234</body>
4235</html>