Gitiles
Code Review Sign In
gerrit-public.fairphone.software / fp2-dev / platform / external / llvm / 47ce179995662e2e8f09e8f8e16cdfcbaad06ae3 / . / docs / LangRef.html
blob: d02caf09c23107f715c07f2dc9879149d8e3356f [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>
Tanya Lattner09474292006-04-14 19:24:33 +0000[diff] [blame]104 <li><a href="#i_vsetint">'<tt>vsetint</tt>' Instruction</a></li>
105 <li><a href="#i_vsetfp">'<tt>vsetfp</tt>' Instruction</a></li>
106 <li><a href="#i_vselect">'<tt>vselect</tt>' Instruction</a></li>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]107 </ol>
108 </li>
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]109 <li><a href="#memoryops">Memory Access and Addressing Operations</a>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]110 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]111 <li><a href="#i_malloc">'<tt>malloc</tt>' Instruction</a></li>
112 <li><a href="#i_free">'<tt>free</tt>' Instruction</a></li>
113 <li><a href="#i_alloca">'<tt>alloca</tt>' Instruction</a></li>
Robert Bocchino7b81c752006-02-17 21:18:08 +0000[diff] [blame]114 <li><a href="#i_load">'<tt>load</tt>' Instruction</a></li>
115 <li><a href="#i_store">'<tt>store</tt>' Instruction</a></li>
116 <li><a href="#i_getelementptr">'<tt>getelementptr</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]117 </ol>
118 </li>
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]119 <li><a href="#convertops">Conversion Operations</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]120 <ol>
121 <li><a href="#i_trunc">'<tt>trunc .. to</tt>' Instruction</a></li>
122 <li><a href="#i_zext">'<tt>zext .. to</tt>' Instruction</a></li>
123 <li><a href="#i_sext">'<tt>sext .. to</tt>' Instruction</a></li>
124 <li><a href="#i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a></li>
125 <li><a href="#i_fpext">'<tt>fpext .. to</tt>' Instruction</a></li>
126 <li><a href="#i_fp2uint">'<tt>fp2uint .. to</tt>' Instruction</a></li>
127 <li><a href="#i_fp2sint">'<tt>fp2sint .. to</tt>' Instruction</a></li>
128 <li><a href="#i_uint2fp">'<tt>uint2fp .. to</tt>' Instruction</a></li>
129 <li><a href="#i_sint2fp">'<tt>sint2fp .. to</tt>' Instruction</a></li>
130 <li><a href="#i_bitconvert">'<tt>bitconvert .. to</tt>' Instruction</a></li>
131 </ol>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]132 <li><a href="#otherops">Other Operations</a>
133 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]134 <li><a href="#i_phi">'<tt>phi</tt>' Instruction</a></li>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]135 <li><a href="#i_select">'<tt>select</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]136 <li><a href="#i_call">'<tt>call</tt>' Instruction</a></li>
Chris Lattnerfb6977d2006-01-13 23:26:01 +0000[diff] [blame]137 <li><a href="#i_va_arg">'<tt>va_arg</tt>' Instruction</a></li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]138 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]139 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]140 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]141 </li>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]142 <li><a href="#intrinsics">Intrinsic Functions</a>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]143 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]144 <li><a href="#int_varargs">Variable Argument Handling Intrinsics</a>
145 <ol>
146 <li><a href="#i_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a></li>
147 <li><a href="#i_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a></li>
148 <li><a href="#i_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a></li>
149 </ol>
150 </li>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]151 <li><a href="#int_gc">Accurate Garbage Collection Intrinsics</a>
152 <ol>
153 <li><a href="#i_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a></li>
154 <li><a href="#i_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a></li>
155 <li><a href="#i_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a></li>
156 </ol>
157 </li>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]158 <li><a href="#int_codegen">Code Generator Intrinsics</a>
159 <ol>
160 <li><a href="#i_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a></li>
161 <li><a href="#i_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a></li>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]162 <li><a href="#i_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a></li>
163 <li><a href="#i_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a></li>
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]164 <li><a href="#i_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a></li>
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]165 <li><a href="#i_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a></li>
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]166 <li><a href="#i_readcyclecounter"><tt>llvm.readcyclecounter</tt>' Intrinsic</a></li>
John Criswell7123e272004-04-09 16:43:20 +0000[diff] [blame]167 </ol>
168 </li>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]169 <li><a href="#int_libc">Standard C Library Intrinsics</a>
170 <ol>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]171 <li><a href="#i_memcpy">'<tt>llvm.memcpy.*</tt>' Intrinsic</a></li>
172 <li><a href="#i_memmove">'<tt>llvm.memmove.*</tt>' Intrinsic</a></li>
173 <li><a href="#i_memset">'<tt>llvm.memset.*</tt>' Intrinsic</a></li>
Chris Lattnerec6cb612006-01-16 22:38:59 +0000[diff] [blame]174 <li><a href="#i_isunordered">'<tt>llvm.isunordered.*</tt>' Intrinsic</a></li>
175 <li><a href="#i_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a></li>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]176 <li><a href="#i_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a></li>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]177 </ol>
178 </li>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]179 <li><a href="#int_manip">Bit Manipulation Intrinsics</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]180 <ol>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]181 <li><a href="#i_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a></li>
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]182 <li><a href="#int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic </a></li>
183 <li><a href="#int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic </a></li>
184 <li><a href="#int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic </a></li>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]185 </ol>
186 </li>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]187 <li><a href="#int_debugger">Debugger intrinsics</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]188 </ol>
189 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]190</ol>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]191
192<div class="doc_author">
193 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>
194 and <a href="mailto:vadve@cs.uiuc.edu">Vikram Adve</a></p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]195</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]196
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]197<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]198<div class="doc_section"> <a name="abstract">Abstract </a></div>
199<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]200
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]201<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]202<p>This document is a reference manual for the LLVM assembly language.
203LLVM is an SSA based representation that provides type safety,
204low-level operations, flexibility, and the capability of representing
205'all' high-level languages cleanly. It is the common code
206representation used throughout all phases of the LLVM compilation
207strategy.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]208</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]209
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]210<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]211<div class="doc_section"> <a name="introduction">Introduction</a> </div>
212<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]213
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]214<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]215
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]216<p>The LLVM code representation is designed to be used in three
217different forms: as an in-memory compiler IR, as an on-disk bytecode
218representation (suitable for fast loading by a Just-In-Time compiler),
219and as a human readable assembly language representation. This allows
220LLVM to provide a powerful intermediate representation for efficient
221compiler transformations and analysis, while providing a natural means
222to debug and visualize the transformations. The three different forms
223of LLVM are all equivalent. This document describes the human readable
224representation and notation.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]225
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]226<p>The LLVM representation aims to be light-weight and low-level
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]227while being expressive, typed, and extensible at the same time. It
228aims to be a "universal IR" of sorts, by being at a low enough level
229that high-level ideas may be cleanly mapped to it (similar to how
230microprocessors are "universal IR's", allowing many source languages to
231be mapped to them). By providing type information, LLVM can be used as
232the target of optimizations: for example, through pointer analysis, it
233can be proven that a C automatic variable is never accessed outside of
234the current function... allowing it to be promoted to a simple SSA
235value instead of a memory location.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]236
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]237</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]238
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]239<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]240<div class="doc_subsubsection"> <a name="wellformed">Well-Formedness</a> </div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]241
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]242<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]243
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]244<p>It is important to note that this document describes 'well formed'
245LLVM assembly language. There is a difference between what the parser
246accepts and what is considered 'well formed'. For example, the
247following instruction is syntactically okay, but not well formed:</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]248
249<pre>
250 %x = <a href="#i_add">add</a> int 1, %x
251</pre>
252
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]253<p>...because the definition of <tt>%x</tt> does not dominate all of
254its uses. The LLVM infrastructure provides a verification pass that may
255be used to verify that an LLVM module is well formed. This pass is
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]256automatically run by the parser after parsing input assembly and by
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]257the optimizer before it outputs bytecode. The violations pointed out
258by the verifier pass indicate bugs in transformation passes or input to
259the parser.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]260
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]261<!-- Describe the typesetting conventions here. --> </div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]262
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]263<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]264<div class="doc_section"> <a name="identifiers">Identifiers</a> </div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]265<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]266
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]267<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]268
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]269<p>LLVM uses three different forms of identifiers, for different
270purposes:</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]271
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]272<ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]273 <li>Named values are represented as a string of characters with a '%' prefix.
274 For example, %foo, %DivisionByZero, %a.really.long.identifier. The actual
275 regular expression used is '<tt>%[a-zA-Z$._][a-zA-Z$._0-9]*</tt>'.
276 Identifiers which require other characters in their names can be surrounded
277 with quotes. In this way, anything except a <tt>"</tt> character can be used
278 in a name.</li>
279
280 <li>Unnamed values are represented as an unsigned numeric value with a '%'
281 prefix. For example, %12, %2, %44.</li>
282
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]283 <li>Constants, which are described in a <a href="#constants">section about
284 constants</a>, below.</li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]285</ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]286
287<p>LLVM requires that values start with a '%' sign for two reasons: Compilers
288don't need to worry about name clashes with reserved words, and the set of
289reserved words may be expanded in the future without penalty. Additionally,
290unnamed identifiers allow a compiler to quickly come up with a temporary
291variable without having to avoid symbol table conflicts.</p>
292
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]293<p>Reserved words in LLVM are very similar to reserved words in other
294languages. There are keywords for different opcodes ('<tt><a
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]295href="#i_add">add</a></tt>', '<tt><a href="#i_cast">cast</a></tt>', '<tt><a
296href="#i_ret">ret</a></tt>', etc...), for primitive type names ('<tt><a
297href="#t_void">void</a></tt>', '<tt><a href="#t_uint">uint</a></tt>', etc...),
298and others. These reserved words cannot conflict with variable names, because
299none of them start with a '%' character.</p>
300
301<p>Here is an example of LLVM code to multiply the integer variable
302'<tt>%X</tt>' by 8:</p>
303
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]304<p>The easy way:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]305
306<pre>
307 %result = <a href="#i_mul">mul</a> uint %X, 8
308</pre>
309
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]310<p>After strength reduction:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]311
312<pre>
313 %result = <a href="#i_shl">shl</a> uint %X, ubyte 3
314</pre>
315
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]316<p>And the hard way:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]317
318<pre>
319 <a href="#i_add">add</a> uint %X, %X <i>; yields {uint}:%0</i>
320 <a href="#i_add">add</a> uint %0, %0 <i>; yields {uint}:%1</i>
321 %result = <a href="#i_add">add</a> uint %1, %1
322</pre>
323
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]324<p>This last way of multiplying <tt>%X</tt> by 8 illustrates several
325important lexical features of LLVM:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]326
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]327<ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]328
329 <li>Comments are delimited with a '<tt>;</tt>' and go until the end of
330 line.</li>
331
332 <li>Unnamed temporaries are created when the result of a computation is not
333 assigned to a named value.</li>
334
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]335 <li>Unnamed temporaries are numbered sequentially</li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]336
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]337</ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]338
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]339<p>...and it also shows a convention that we follow in this document. When
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]340demonstrating instructions, we will follow an instruction with a comment that
341defines the type and name of value produced. Comments are shown in italic
342text.</p>
343
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]344</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]345
346<!-- *********************************************************************** -->
347<div class="doc_section"> <a name="highlevel">High Level Structure</a> </div>
348<!-- *********************************************************************** -->
349
350<!-- ======================================================================= -->
351<div class="doc_subsection"> <a name="modulestructure">Module Structure</a>
352</div>
353
354<div class="doc_text">
355
356<p>LLVM programs are composed of "Module"s, each of which is a
357translation unit of the input programs. Each module consists of
358functions, global variables, and symbol table entries. Modules may be
359combined together with the LLVM linker, which merges function (and
360global variable) definitions, resolves forward declarations, and merges
361symbol table entries. Here is an example of the "hello world" module:</p>
362
363<pre><i>; Declare the string constant as a global constant...</i>
364<a href="#identifiers">%.LC0</a> = <a href="#linkage_internal">internal</a> <a
365 href="#globalvars">constant</a> <a href="#t_array">[13 x sbyte]</a> c"hello world\0A\00" <i>; [13 x sbyte]*</i>
366
367<i>; External declaration of the puts function</i>
368<a href="#functionstructure">declare</a> int %puts(sbyte*) <i>; int(sbyte*)* </i>
369
Chris Lattner81c01f02006-06-13 03:05:47 +0000[diff] [blame]370<i>; Global variable / Function body section separator</i>
371implementation
372
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]373<i>; Definition of main function</i>
374int %main() { <i>; int()* </i>
375 <i>; Convert [13x sbyte]* to sbyte *...</i>
376 %cast210 = <a
377 href="#i_getelementptr">getelementptr</a> [13 x sbyte]* %.LC0, long 0, long 0 <i>; sbyte*</i>
378
379 <i>; Call puts function to write out the string to stdout...</i>
380 <a
381 href="#i_call">call</a> int %puts(sbyte* %cast210) <i>; int</i>
382 <a
383 href="#i_ret">ret</a> int 0<br>}<br></pre>
384
385<p>This example is made up of a <a href="#globalvars">global variable</a>
386named "<tt>.LC0</tt>", an external declaration of the "<tt>puts</tt>"
387function, and a <a href="#functionstructure">function definition</a>
388for "<tt>main</tt>".</p>
389
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]390<p>In general, a module is made up of a list of global values,
391where both functions and global variables are global values. Global values are
392represented by a pointer to a memory location (in this case, a pointer to an
393array of char, and a pointer to a function), and have one of the following <a
394href="#linkage">linkage types</a>.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]395
Chris Lattner81c01f02006-06-13 03:05:47 +0000[diff] [blame]396<p>Due to a limitation in the current LLVM assembly parser (it is limited by
397one-token lookahead), modules are split into two pieces by the "implementation"
398keyword. Global variable prototypes and definitions must occur before the
399keyword, and function definitions must occur after it. Function prototypes may
400occur either before or after it. In the future, the implementation keyword may
401become a noop, if the parser gets smarter.</p>
402
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]403</div>
404
405<!-- ======================================================================= -->
406<div class="doc_subsection">
407 <a name="linkage">Linkage Types</a>
408</div>
409
410<div class="doc_text">
411
412<p>
413All Global Variables and Functions have one of the following types of linkage:
414</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]415
416<dl>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]417
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]418 <dt><tt><b><a name="linkage_internal">internal</a></b></tt> </dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]419
420 <dd>Global values with internal linkage are only directly accessible by
421 objects in the current module. In particular, linking code into a module with
422 an internal global value may cause the internal to be renamed as necessary to
423 avoid collisions. Because the symbol is internal to the module, all
424 references can be updated. This corresponds to the notion of the
425 '<tt>static</tt>' keyword in C, or the idea of "anonymous namespaces" in C++.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]426 </dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]427
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]428 <dt><tt><b><a name="linkage_linkonce">linkonce</a></b></tt>: </dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]429
430 <dd>"<tt>linkonce</tt>" linkage is similar to <tt>internal</tt> linkage, with
431 the twist that linking together two modules defining the same
432 <tt>linkonce</tt> globals will cause one of the globals to be discarded. This
433 is typically used to implement inline functions. Unreferenced
434 <tt>linkonce</tt> globals are allowed to be discarded.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]435 </dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]436
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]437 <dt><tt><b><a name="linkage_weak">weak</a></b></tt>: </dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]438
439 <dd>"<tt>weak</tt>" linkage is exactly the same as <tt>linkonce</tt> linkage,
440 except that unreferenced <tt>weak</tt> globals may not be discarded. This is
441 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]442 </dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]443
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]444 <dt><tt><b><a name="linkage_appending">appending</a></b></tt>: </dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]445
446 <dd>"<tt>appending</tt>" linkage may only be applied to global variables of
447 pointer to array type. When two global variables with appending linkage are
448 linked together, the two global arrays are appended together. This is the
449 LLVM, typesafe, equivalent of having the system linker append together
450 "sections" with identical names when .o files are linked.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]451 </dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]452
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]453 <dt><tt><b><a name="linkage_external">externally visible</a></b></tt>:</dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]454
455 <dd>If none of the above identifiers are used, the global is externally
456 visible, meaning that it participates in linkage and can be used to resolve
457 external symbol references.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]458 </dd>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]459
460 <dt><tt><b><a name="linkage_externweak">extern_weak</a></b></tt>: </dt>
461
462 <dd>"<tt>extern_weak</tt>" TBD
463 </dd>
464
465 <p>
466 The next two types of linkage are targeted for Microsoft Windows platform
467 only. They are designed to support importing (exporting) symbols from (to)
468 DLLs.
469 </p>
470
471 <dt><tt><b><a name="linkage_dllimport">dllimport</a></b></tt>: </dt>
472
473 <dd>"<tt>dllimport</tt>" linkage causes the compiler to reference a function
474 or variable via a global pointer to a pointer that is set up by the DLL
475 exporting the symbol. On Microsoft Windows targets, the pointer name is
476 formed by combining <code>_imp__</code> and the function or variable name.
477 </dd>
478
479 <dt><tt><b><a name="linkage_dllexport">dllexport</a></b></tt>: </dt>
480
481 <dd>"<tt>dllexport</tt>" linkage causes the compiler to provide a global
482 pointer to a pointer in a DLL, so that it can be referenced with the
483 <tt>dllimport</tt> attribute. On Microsoft Windows targets, the pointer
484 name is formed by combining <code>_imp__</code> and the function or variable
485 name.
486 </dd>
487
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]488</dl>
489
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]490<p><a name="linkage_external">For example, since the "<tt>.LC0</tt>"
491variable is defined to be internal, if another module defined a "<tt>.LC0</tt>"
492variable and was linked with this one, one of the two would be renamed,
493preventing a collision. Since "<tt>main</tt>" and "<tt>puts</tt>" are
494external (i.e., lacking any linkage declarations), they are accessible
495outside of the current module. It is illegal for a function <i>declaration</i>
496to have any linkage type other than "externally visible".</a></p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]497
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]498</div>
499
500<!-- ======================================================================= -->
501<div class="doc_subsection">
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]502 <a name="callingconv">Calling Conventions</a>
503</div>
504
505<div class="doc_text">
506
507<p>LLVM <a href="#functionstructure">functions</a>, <a href="#i_call">calls</a>
508and <a href="#i_invoke">invokes</a> can all have an optional calling convention
509specified for the call. The calling convention of any pair of dynamic
510caller/callee must match, or the behavior of the program is undefined. The
511following calling conventions are supported by LLVM, and more may be added in
512the future:</p>
513
514<dl>
515 <dt><b>"<tt>ccc</tt>" - The C calling convention</b>:</dt>
516
517 <dd>This calling convention (the default if no other calling convention is
518 specified) matches the target C calling conventions. This calling convention
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]519 supports varargs function calls and tolerates some mismatch in the declared
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]520 prototype and implemented declaration of the function (as does normal C).
521 </dd>
522
Chris Lattner5710ce92006-05-19 21:15:36 +0000[diff] [blame]523 <dt><b>"<tt>csretcc</tt>" - The C struct return calling convention</b>:</dt>
524
525 <dd>This calling convention matches the target C calling conventions, except
526 that functions with this convention are required to take a pointer as their
527 first argument, and the return type of the function must be void. This is
528 used for C functions that return aggregates by-value. In this case, the
529 function has been transformed to take a pointer to the struct as the first
530 argument to the function. For targets where the ABI specifies specific
531 behavior for structure-return calls, the calling convention can be used to
532 distinguish between struct return functions and other functions that take a
533 pointer to a struct as the first argument.
534 </dd>
535
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]536 <dt><b>"<tt>fastcc</tt>" - The fast calling convention</b>:</dt>
537
538 <dd>This calling convention attempts to make calls as fast as possible
539 (e.g. by passing things in registers). This calling convention allows the
540 target to use whatever tricks it wants to produce fast code for the target,
Chris Lattner8cdc5bc2005-05-06 23:08:23 +0000[diff] [blame]541 without having to conform to an externally specified ABI. Implementations of
542 this convention should allow arbitrary tail call optimization to be supported.
543 This calling convention does not support varargs and requires the prototype of
544 all callees to exactly match the prototype of the function definition.
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]545 </dd>
546
547 <dt><b>"<tt>coldcc</tt>" - The cold calling convention</b>:</dt>
548
549 <dd>This calling convention attempts to make code in the caller as efficient
550 as possible under the assumption that the call is not commonly executed. As
551 such, these calls often preserve all registers so that the call does not break
552 any live ranges in the caller side. This calling convention does not support
553 varargs and requires the prototype of all callees to exactly match the
554 prototype of the function definition.
555 </dd>
556
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]557 <dt><b>"<tt>cc &lt;<em>n</em>&gt;</tt>" - Numbered convention</b>:</dt>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]558
559 <dd>Any calling convention may be specified by number, allowing
560 target-specific calling conventions to be used. Target specific calling
561 conventions start at 64.
562 </dd>
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]563</dl>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]564
565<p>More calling conventions can be added/defined on an as-needed basis, to
566support pascal conventions or any other well-known target-independent
567convention.</p>
568
569</div>
570
571<!-- ======================================================================= -->
572<div class="doc_subsection">
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]573 <a name="globalvars">Global Variables</a>
574</div>
575
576<div class="doc_text">
577
Chris Lattner3689a342005-02-12 19:30:21 +0000[diff] [blame]578<p>Global variables define regions of memory allocated at compilation time
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]579instead of run-time. Global variables may optionally be initialized, may have
580an explicit section to be placed in, and may
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]581have an optional explicit alignment specified. A
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]582variable may be defined as a global "constant," which indicates that the
Chris Lattner3689a342005-02-12 19:30:21 +0000[diff] [blame]583contents of the variable will <b>never</b> be modified (enabling better
584optimization, allowing the global data to be placed in the read-only section of
585an executable, etc). Note that variables that need runtime initialization
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]586cannot be marked "constant" as there is a store to the variable.</p>
Chris Lattner3689a342005-02-12 19:30:21 +0000[diff] [blame]587
588<p>
589LLVM explicitly allows <em>declarations</em> of global variables to be marked
590constant, even if the final definition of the global is not. This capability
591can be used to enable slightly better optimization of the program, but requires
592the language definition to guarantee that optimizations based on the
593'constantness' are valid for the translation units that do not include the
594definition.
595</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]596
597<p>As SSA values, global variables define pointer values that are in
598scope (i.e. they dominate) all basic blocks in the program. Global
599variables always define a pointer to their "content" type because they
600describe a region of memory, and all memory objects in LLVM are
601accessed through pointers.</p>
602
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]603<p>LLVM allows an explicit section to be specified for globals. If the target
604supports it, it will emit globals to the section specified.</p>
605
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]606<p>An explicit alignment may be specified for a global. If not present, or if
607the alignment is set to zero, the alignment of the global is set by the target
608to whatever it feels convenient. If an explicit alignment is specified, the
609global is forced to have at least that much alignment. All alignments must be
610a power of 2.</p>
611
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]612</div>
613
614
615<!-- ======================================================================= -->
616<div class="doc_subsection">
617 <a name="functionstructure">Functions</a>
618</div>
619
620<div class="doc_text">
621
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]622<p>LLVM function definitions consist of an optional <a href="#linkage">linkage
623type</a>, an optional <a href="#callingconv">calling convention</a>, a return
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]624type, a function name, a (possibly empty) argument list, an optional section,
625an optional alignment, an opening curly brace,
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]626a list of basic blocks, and a closing curly brace. LLVM function declarations
627are defined with the "<tt>declare</tt>" keyword, an optional <a
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]628href="#callingconv">calling convention</a>, a return type, a function name,
629a possibly empty list of arguments, and an optional alignment.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]630
631<p>A function definition contains a list of basic blocks, forming the CFG for
632the function. Each basic block may optionally start with a label (giving the
633basic block a symbol table entry), contains a list of instructions, and ends
634with a <a href="#terminators">terminator</a> instruction (such as a branch or
635function return).</p>
636
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]637<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]638executed on entrance to the function, and it is not allowed to have predecessor
639basic blocks (i.e. there can not be any branches to the entry block of a
640function). Because the block can have no predecessors, it also cannot have any
641<a href="#i_phi">PHI nodes</a>.</p>
642
643<p>LLVM functions are identified by their name and type signature. Hence, two
644functions with the same name but different parameter lists or return values are
Chris Lattnerd4f6b172005-03-07 22:13:59 +0000[diff] [blame]645considered different functions, and LLVM will resolve references to each
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]646appropriately.</p>
647
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]648<p>LLVM allows an explicit section to be specified for functions. If the target
649supports it, it will emit functions to the section specified.</p>
650
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]651<p>An explicit alignment may be specified for a function. If not present, or if
652the alignment is set to zero, the alignment of the function is set by the target
653to whatever it feels convenient. If an explicit alignment is specified, the
654function is forced to have at least that much alignment. All alignments must be
655a power of 2.</p>
656
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]657</div>
658
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]659<!-- ======================================================================= -->
660<div class="doc_subsection">
Chris Lattner1eeeb0c2006-04-08 04:40:53 +0000[diff] [blame]661 <a name="moduleasm">Module-Level Inline Assembly</a>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]662</div>
663
664<div class="doc_text">
665<p>
666Modules may contain "module-level inline asm" blocks, which corresponds to the
667GCC "file scope inline asm" blocks. These blocks are internally concatenated by
668LLVM and treated as a single unit, but may be separated in the .ll file if
669desired. The syntax is very simple:
670</p>
671
672<div class="doc_code"><pre>
Chris Lattner52599e12006-01-24 00:37:20 +0000[diff] [blame]673 module asm "inline asm code goes here"
674 module asm "more can go here"
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]675</pre></div>
676
677<p>The strings can contain any character by escaping non-printable characters.
678 The escape sequence used is simply "\xx" where "xx" is the two digit hex code
679 for the number.
680</p>
681
682<p>
683 The inline asm code is simply printed to the machine code .s file when
684 assembly code is generated.
685</p>
686</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]687
688
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]689<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]690<div class="doc_section"> <a name="typesystem">Type System</a> </div>
691<!-- *********************************************************************** -->
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]692
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]693<div class="doc_text">
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]694
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]695<p>The LLVM type system is one of the most important features of the
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]696intermediate representation. Being typed enables a number of
697optimizations to be performed on the IR directly, without having to do
698extra analyses on the side before the transformation. A strong type
699system makes it easier to read the generated code and enables novel
700analyses and transformations that are not feasible to perform on normal
701three address code representations.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]702
703</div>
704
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]705<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]706<div class="doc_subsection"> <a name="t_primitive">Primitive Types</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]707<div class="doc_text">
John Criswell4457dc92004-04-09 16:48:45 +0000[diff] [blame]708<p>The primitive types are the fundamental building blocks of the LLVM
Chris Lattnerd4f6b172005-03-07 22:13:59 +0000[diff] [blame]709system. The current set of primitive types is as follows:</p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]710
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]711<table class="layout">
712 <tr class="layout">
713 <td class="left">
714 <table>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]715 <tbody>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]716 <tr><th>Type</th><th>Description</th></tr>
717 <tr><td><tt>void</tt></td><td>No value</td></tr>
Misha Brukmancfa87bc2005-04-22 18:02:52 +0000[diff] [blame]718 <tr><td><tt>ubyte</tt></td><td>Unsigned 8-bit value</td></tr>
719 <tr><td><tt>ushort</tt></td><td>Unsigned 16-bit value</td></tr>
720 <tr><td><tt>uint</tt></td><td>Unsigned 32-bit value</td></tr>
721 <tr><td><tt>ulong</tt></td><td>Unsigned 64-bit value</td></tr>
722 <tr><td><tt>float</tt></td><td>32-bit floating point value</td></tr>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]723 <tr><td><tt>label</tt></td><td>Branch destination</td></tr>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]724 </tbody>
725 </table>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]726 </td>
727 <td class="right">
728 <table>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]729 <tbody>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]730 <tr><th>Type</th><th>Description</th></tr>
731 <tr><td><tt>bool</tt></td><td>True or False value</td></tr>
Misha Brukmancfa87bc2005-04-22 18:02:52 +0000[diff] [blame]732 <tr><td><tt>sbyte</tt></td><td>Signed 8-bit value</td></tr>
733 <tr><td><tt>short</tt></td><td>Signed 16-bit value</td></tr>
734 <tr><td><tt>int</tt></td><td>Signed 32-bit value</td></tr>
735 <tr><td><tt>long</tt></td><td>Signed 64-bit value</td></tr>
736 <tr><td><tt>double</tt></td><td>64-bit floating point value</td></tr>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]737 </tbody>
738 </table>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]739 </td>
740 </tr>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]741</table>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]742</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]743
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]744<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]745<div class="doc_subsubsection"> <a name="t_classifications">Type
746Classifications</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]747<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]748<p>These different primitive types fall into a few useful
749classifications:</p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]750
751<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]752 <tbody>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]753 <tr><th>Classification</th><th>Types</th></tr>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]754 <tr>
755 <td><a name="t_signed">signed</a></td>
756 <td><tt>sbyte, short, int, long, float, double</tt></td>
757 </tr>
758 <tr>
759 <td><a name="t_unsigned">unsigned</a></td>
760 <td><tt>ubyte, ushort, uint, ulong</tt></td>
761 </tr>
762 <tr>
763 <td><a name="t_integer">integer</a></td>
764 <td><tt>ubyte, sbyte, ushort, short, uint, int, ulong, long</tt></td>
765 </tr>
766 <tr>
767 <td><a name="t_integral">integral</a></td>
Misha Brukmanc24b7582004-08-12 20:16:08 +0000[diff] [blame]768 <td><tt>bool, ubyte, sbyte, ushort, short, uint, int, ulong, long</tt>
769 </td>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]770 </tr>
771 <tr>
772 <td><a name="t_floating">floating point</a></td>
773 <td><tt>float, double</tt></td>
774 </tr>
775 <tr>
776 <td><a name="t_firstclass">first class</a></td>
Misha Brukmanc24b7582004-08-12 20:16:08 +0000[diff] [blame]777 <td><tt>bool, ubyte, sbyte, ushort, short, uint, int, ulong, long,<br>
778 float, double, <a href="#t_pointer">pointer</a>,
779 <a href="#t_packed">packed</a></tt></td>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]780 </tr>
781 </tbody>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]782</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]783
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]784<p>The <a href="#t_firstclass">first class</a> types are perhaps the
785most important. Values of these types are the only ones which can be
786produced by instructions, passed as arguments, or used as operands to
787instructions. This means that all structures and arrays must be
788manipulated either by pointer or by component.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]789</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]790
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]791<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]792<div class="doc_subsection"> <a name="t_derived">Derived Types</a> </div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]793
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]794<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]795
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]796<p>The real power in LLVM comes from the derived types in the system.
797This is what allows a programmer to represent arrays, functions,
798pointers, and other useful types. Note that these derived types may be
799recursive: For example, it is possible to have a two dimensional array.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]800
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]801</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]802
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]803<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]804<div class="doc_subsubsection"> <a name="t_array">Array Type</a> </div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]805
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]806<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]807
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]808<h5>Overview:</h5>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]809
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]810<p>The array type is a very simple derived type that arranges elements
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]811sequentially in memory. The array type requires a size (number of
812elements) and an underlying data type.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]813
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]814<h5>Syntax:</h5>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]815
816<pre>
817 [&lt;# elements&gt; x &lt;elementtype&gt;]
818</pre>
819
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]820<p>The number of elements is a constant integer value; elementtype may
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]821be any type with a size.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]822
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]823<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]824<table class="layout">
825 <tr class="layout">
826 <td class="left">
827 <tt>[40 x int ]</tt><br/>
828 <tt>[41 x int ]</tt><br/>
829 <tt>[40 x uint]</tt><br/>
830 </td>
831 <td class="left">
832 Array of 40 integer values.<br/>
833 Array of 41 integer values.<br/>
834 Array of 40 unsigned integer values.<br/>
835 </td>
836 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]837</table>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]838<p>Here are some examples of multidimensional arrays:</p>
839<table class="layout">
840 <tr class="layout">
841 <td class="left">
842 <tt>[3 x [4 x int]]</tt><br/>
843 <tt>[12 x [10 x float]]</tt><br/>
844 <tt>[2 x [3 x [4 x uint]]]</tt><br/>
845 </td>
846 <td class="left">
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]847 3x4 array of integer values.<br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]848 12x10 array of single precision floating point values.<br/>
849 2x3x4 array of unsigned integer values.<br/>
850 </td>
851 </tr>
852</table>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]853
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]854<p>Note that 'variable sized arrays' can be implemented in LLVM with a zero
855length array. Normally, accesses past the end of an array are undefined in
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]856LLVM (e.g. it is illegal to access the 5th element of a 3 element array).
857As a special case, however, zero length arrays are recognized to be variable
858length. This allows implementation of 'pascal style arrays' with the LLVM
859type "{ int, [0 x float]}", for example.</p>
860
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]861</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]862
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]863<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]864<div class="doc_subsubsection"> <a name="t_function">Function Type</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]865<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]866<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]867<p>The function type can be thought of as a function signature. It
868consists of a return type and a list of formal parameter types.
John Criswell009900b2003-11-25 21:45:46 +0000[diff] [blame]869Function types are usually used to build virtual function tables
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]870(which are structures of pointers to functions), for indirect function
871calls, and when defining a function.</p>
John Criswell009900b2003-11-25 21:45:46 +0000[diff] [blame]872<p>
873The return type of a function type cannot be an aggregate type.
874</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]875<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]876<pre> &lt;returntype&gt; (&lt;parameter list&gt;)<br></pre>
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]877<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]878specifiers. Optionally, the parameter list may include a type <tt>...</tt>,
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]879which indicates that the function takes a variable number of arguments.
880Variable argument functions can access their arguments with the <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]881 href="#int_varargs">variable argument handling intrinsic</a> functions.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]882<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]883<table class="layout">
884 <tr class="layout">
885 <td class="left">
886 <tt>int (int)</tt> <br/>
887 <tt>float (int, int *) *</tt><br/>
888 <tt>int (sbyte *, ...)</tt><br/>
889 </td>
890 <td class="left">
891 function taking an <tt>int</tt>, returning an <tt>int</tt><br/>
892 <a href="#t_pointer">Pointer</a> to a function that takes an
Misha Brukmanc24b7582004-08-12 20:16:08 +0000[diff] [blame]893 <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]894 returning <tt>float</tt>.<br/>
895 A vararg function that takes at least one <a href="#t_pointer">pointer</a>
896 to <tt>sbyte</tt> (signed char in C), which returns an integer. This is
897 the signature for <tt>printf</tt> in LLVM.<br/>
898 </td>
899 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]900</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]901
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]902</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]903<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]904<div class="doc_subsubsection"> <a name="t_struct">Structure Type</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]905<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]906<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]907<p>The structure type is used to represent a collection of data members
908together in memory. The packing of the field types is defined to match
909the ABI of the underlying processor. The elements of a structure may
910be any type that has a size.</p>
911<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
912and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
913field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
914instruction.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]915<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]916<pre> { &lt;type list&gt; }<br></pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]917<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]918<table class="layout">
919 <tr class="layout">
920 <td class="left">
921 <tt>{ int, int, int }</tt><br/>
922 <tt>{ float, int (int) * }</tt><br/>
923 </td>
924 <td class="left">
925 a triple of three <tt>int</tt> values<br/>
926 A pair, where the first element is a <tt>float</tt> and the second element
927 is a <a href="#t_pointer">pointer</a> to a <a href="#t_function">function</a>
928 that takes an <tt>int</tt>, returning an <tt>int</tt>.<br/>
929 </td>
930 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]931</table>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]932</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]933
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]934<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]935<div class="doc_subsubsection"> <a name="t_pointer">Pointer Type</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]936<div class="doc_text">
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]937<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]938<p>As in many languages, the pointer type represents a pointer or
939reference to another object, which must live in memory.</p>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]940<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]941<pre> &lt;type&gt; *<br></pre>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]942<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]943<table class="layout">
944 <tr class="layout">
945 <td class="left">
946 <tt>[4x int]*</tt><br/>
947 <tt>int (int *) *</tt><br/>
948 </td>
949 <td class="left">
950 A <a href="#t_pointer">pointer</a> to <a href="#t_array">array</a> of
951 four <tt>int</tt> values<br/>
952 A <a href="#t_pointer">pointer</a> to a <a
Chris Lattnera977c482005-02-19 02:22:14 +0000[diff] [blame]953 href="#t_function">function</a> that takes an <tt>int*</tt>, returning an
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]954 <tt>int</tt>.<br/>
955 </td>
956 </tr>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]957</table>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]958</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]959
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]960<!-- _______________________________________________________________________ -->
961<div class="doc_subsubsection"> <a name="t_packed">Packed Type</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]962<div class="doc_text">
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]963
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]964<h5>Overview:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]965
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]966<p>A packed type is a simple derived type that represents a vector
967of elements. Packed types are used when multiple primitive data
968are operated in parallel using a single instruction (SIMD).
969A packed type requires a size (number of
Chris Lattnerb8d172f2005-11-10 01:44:22 +0000[diff] [blame]970elements) and an underlying primitive data type. Vectors must have a power
971of two length (1, 2, 4, 8, 16 ...). Packed types are
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]972considered <a href="#t_firstclass">first class</a>.</p>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]973
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]974<h5>Syntax:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]975
976<pre>
977 &lt; &lt;# elements&gt; x &lt;elementtype&gt; &gt;
978</pre>
979
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]980<p>The number of elements is a constant integer value; elementtype may
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]981be any integral or floating point type.</p>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]982
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]983<h5>Examples:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]984
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]985<table class="layout">
986 <tr class="layout">
987 <td class="left">
988 <tt>&lt;4 x int&gt;</tt><br/>
989 <tt>&lt;8 x float&gt;</tt><br/>
990 <tt>&lt;2 x uint&gt;</tt><br/>
991 </td>
992 <td class="left">
993 Packed vector of 4 integer values.<br/>
994 Packed vector of 8 floating-point values.<br/>
995 Packed vector of 2 unsigned integer values.<br/>
996 </td>
997 </tr>
998</table>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]999</div>
1000
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1001<!-- _______________________________________________________________________ -->
1002<div class="doc_subsubsection"> <a name="t_opaque">Opaque Type</a> </div>
1003<div class="doc_text">
1004
1005<h5>Overview:</h5>
1006
1007<p>Opaque types are used to represent unknown types in the system. This
1008corresponds (for example) to the C notion of a foward declared structure type.
1009In LLVM, opaque types can eventually be resolved to any type (not just a
1010structure type).</p>
1011
1012<h5>Syntax:</h5>
1013
1014<pre>
1015 opaque
1016</pre>
1017
1018<h5>Examples:</h5>
1019
1020<table class="layout">
1021 <tr class="layout">
1022 <td class="left">
1023 <tt>opaque</tt>
1024 </td>
1025 <td class="left">
1026 An opaque type.<br/>
1027 </td>
1028 </tr>
1029</table>
1030</div>
1031
1032
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1033<!-- *********************************************************************** -->
1034<div class="doc_section"> <a name="constants">Constants</a> </div>
1035<!-- *********************************************************************** -->
1036
1037<div class="doc_text">
1038
1039<p>LLVM has several different basic types of constants. This section describes
1040them all and their syntax.</p>
1041
1042</div>
1043
1044<!-- ======================================================================= -->
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]1045<div class="doc_subsection"><a name="simpleconstants">Simple Constants</a></div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1046
1047<div class="doc_text">
1048
1049<dl>
1050 <dt><b>Boolean constants</b></dt>
1051
1052 <dd>The two strings '<tt>true</tt>' and '<tt>false</tt>' are both valid
1053 constants of the <tt><a href="#t_primitive">bool</a></tt> type.
1054 </dd>
1055
1056 <dt><b>Integer constants</b></dt>
1057
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]1058 <dd>Standard integers (such as '4') are constants of the <a
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1059 href="#t_integer">integer</a> type. Negative numbers may be used with signed
1060 integer types.
1061 </dd>
1062
1063 <dt><b>Floating point constants</b></dt>
1064
1065 <dd>Floating point constants use standard decimal notation (e.g. 123.421),
1066 exponential notation (e.g. 1.23421e+2), or a more precise hexadecimal
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1067 notation (see below). Floating point constants must have a <a
1068 href="#t_floating">floating point</a> type. </dd>
1069
1070 <dt><b>Null pointer constants</b></dt>
1071
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1072 <dd>The identifier '<tt>null</tt>' is recognized as a null pointer constant
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1073 and must be of <a href="#t_pointer">pointer type</a>.</dd>
1074
1075</dl>
1076
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1077<p>The one non-intuitive notation for constants is the optional hexadecimal form
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1078of floating point constants. For example, the form '<tt>double
10790x432ff973cafa8000</tt>' is equivalent to (but harder to read than) '<tt>double
10804.5e+15</tt>'. The only time hexadecimal floating point constants are required
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]1081(and the only time that they are generated by the disassembler) is when a
1082floating point constant must be emitted but it cannot be represented as a
1083decimal floating point number. For example, NaN's, infinities, and other
1084special values are represented in their IEEE hexadecimal format so that
1085assembly and disassembly do not cause any bits to change in the constants.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1086
1087</div>
1088
1089<!-- ======================================================================= -->
1090<div class="doc_subsection"><a name="aggregateconstants">Aggregate Constants</a>
1091</div>
1092
1093<div class="doc_text">
Chris Lattnerd4f6b172005-03-07 22:13:59 +0000[diff] [blame]1094<p>Aggregate constants arise from aggregation of simple constants
1095and smaller aggregate constants.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1096
1097<dl>
1098 <dt><b>Structure constants</b></dt>
1099
1100 <dd>Structure constants are represented with notation similar to structure
1101 type definitions (a comma separated list of elements, surrounded by braces
Chris Lattnerd4f6b172005-03-07 22:13:59 +0000[diff] [blame]1102 (<tt>{}</tt>)). For example: "<tt>{ int 4, float 17.0, int* %G }</tt>",
1103 where "<tt>%G</tt>" is declared as "<tt>%G = external global int</tt>". Structure constants
1104 must have <a href="#t_struct">structure type</a>, and the number and
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1105 types of elements must match those specified by the type.
1106 </dd>
1107
1108 <dt><b>Array constants</b></dt>
1109
1110 <dd>Array constants are represented with notation similar to array type
1111 definitions (a comma separated list of elements, surrounded by square brackets
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1112 (<tt>[]</tt>)). For example: "<tt>[ int 42, int 11, int 74 ]</tt>". Array
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1113 constants must have <a href="#t_array">array type</a>, and the number and
1114 types of elements must match those specified by the type.
1115 </dd>
1116
1117 <dt><b>Packed constants</b></dt>
1118
1119 <dd>Packed constants are represented with notation similar to packed type
1120 definitions (a comma separated list of elements, surrounded by
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1121 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]1122 int 11, int 74, int 100 &gt;</tt>". Packed constants must have <a
1123 href="#t_packed">packed type</a>, and the number and types of elements must
1124 match those specified by the type.
1125 </dd>
1126
1127 <dt><b>Zero initialization</b></dt>
1128
1129 <dd>The string '<tt>zeroinitializer</tt>' can be used to zero initialize a
1130 value to zero of <em>any</em> type, including scalar and aggregate types.
1131 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]1132 large arrays) and is always exactly equivalent to using explicit zero
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1133 initializers.
1134 </dd>
1135</dl>
1136
1137</div>
1138
1139<!-- ======================================================================= -->
1140<div class="doc_subsection">
1141 <a name="globalconstants">Global Variable and Function Addresses</a>
1142</div>
1143
1144<div class="doc_text">
1145
1146<p>The addresses of <a href="#globalvars">global variables</a> and <a
1147href="#functionstructure">functions</a> are always implicitly valid (link-time)
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1148constants. These constants are explicitly referenced when the <a
1149href="#identifiers">identifier for the global</a> is used and always have <a
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1150href="#t_pointer">pointer</a> type. For example, the following is a legal LLVM
1151file:</p>
1152
1153<pre>
1154 %X = global int 17
1155 %Y = global int 42
1156 %Z = global [2 x int*] [ int* %X, int* %Y ]
1157</pre>
1158
1159</div>
1160
1161<!-- ======================================================================= -->
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]1162<div class="doc_subsection"><a name="undefvalues">Undefined Values</a></div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1163<div class="doc_text">
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]1164 <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]1165 no specific value. Undefined values may be of any type and be used anywhere
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]1166 a constant is permitted.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1167
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]1168 <p>Undefined values indicate to the compiler that the program is well defined
1169 no matter what value is used, giving the compiler more freedom to optimize.
1170 </p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1171</div>
1172
1173<!-- ======================================================================= -->
1174<div class="doc_subsection"><a name="constantexprs">Constant Expressions</a>
1175</div>
1176
1177<div class="doc_text">
1178
1179<p>Constant expressions are used to allow expressions involving other constants
1180to be used as constants. Constant expressions may be of any <a
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]1181href="#t_firstclass">first class</a> type and may involve any LLVM operation
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1182that does not have side effects (e.g. load and call are not supported). The
1183following is the syntax for constant expressions:</p>
1184
1185<dl>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1186 <dt><b><tt>trunc ( CST to TYPE )</tt></b></dt>
1187 <dd>Truncate a constant to another type. The bit size of CST must be larger
1188 than the bit size of TYPE. Both types must be integral.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1189
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1190 <dt><b><tt>zext ( CST to TYPE )</tt></b></dt>
1191 <dd>Zero extend a constant to another type. The bit size of CST must be
1192 smaller or equal to the bit size of TYPE. Both types must be integral.</dd>
1193
1194 <dt><b><tt>sext ( CST to TYPE )</tt></b></dt>
1195 <dd>Sign extend a constant to another type. The bit size of CST must be
1196 smaller or equal to the bit size of TYPE. Both types must be integral.</dd>
1197
1198 <dt><b><tt>fptrunc ( CST to TYPE )</tt></b></dt>
1199 <dd>Truncate a floating point constant to another floating point type. The
1200 size of CST must be larger than the size of TYPE. Both types must be
1201 floating point.</dd>
1202
1203 <dt><b><tt>fpext ( CST to TYPE )</tt></b></dt>
1204 <dd>Floating point extend a constant to another type. The size of CST must be
1205 smaller or equal to the size of TYPE. Both types must be floating point.</dd>
1206
1207 <dt><b><tt>fp2uint ( CST to TYPE )</tt></b></dt>
1208 <dd>Convert a floating point constant to the corresponding unsigned integer
1209 constant. TYPE must be an integer type. CST must be floating point. If the
1210 value won't fit in the integer type, the results are undefined.</dd>
1211
1212 <dt><b><tt>fp2sint ( CST to TYPE )</tt></b></dt>
1213 <dd>Convert a floating point constant to the corresponding signed integer
1214 constant. TYPE must be an integer type. CST must be floating point. If the
1215 value won't fit in the integer type, the results are undefined.</dd>
1216
1217 <dt><b><tt>uint2fp ( CST to TYPE )</tt></b></dt>
1218 <dd>Convert an unsigned integer constant to the corresponding floating point
1219 constant. TYPE must be floating point. CST must be of integer type. If the
1220 value won't fit in the floating point type, the results are undefined.</dd>
1221
1222 <dt><b><tt>sint2fp ( CST to TYPE )</tt></b></dt>
1223 <dd>Convert a signed integer constant to the corresponding floating point
1224 constant. TYPE must be floating point. CST must be of integer type. If the
1225 value won't fit in the floating point type, the results are undefined.</dd>
1226
1227 <dt><b><tt>bitconvert ( CST to TYPE )</tt></b></dt>
1228 <dd>Convert a constant, CST, to another TYPE. The size of CST and TYPE must be
1229 identical (same number of bits). The conversion is done as if the CST value
1230 was stored to memory and read back as TYPE. In other words, no bits change
1231 with this operator, just the type. This can be used for conversion of pointer
1232 and packed types to any other type, as long as they have the same bit width.
1233 </dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1234
1235 <dt><b><tt>getelementptr ( CSTPTR, IDX0, IDX1, ... )</tt></b></dt>
1236
1237 <dd>Perform the <a href="#i_getelementptr">getelementptr operation</a> on
1238 constants. As with the <a href="#i_getelementptr">getelementptr</a>
1239 instruction, the index list may have zero or more indexes, which are required
1240 to make sense for the type of "CSTPTR".</dd>
1241
Robert Bocchino9fbe1452006-01-10 19:31:34 +0000[diff] [blame]1242 <dt><b><tt>select ( COND, VAL1, VAL2 )</tt></b></dt>
1243
1244 <dd>Perform the <a href="#i_select">select operation</a> on
1245 constants.
1246
1247 <dt><b><tt>extractelement ( VAL, IDX )</tt></b></dt>
1248
1249 <dd>Perform the <a href="#i_extractelement">extractelement
1250 operation</a> on constants.
1251
Robert Bocchino05ccd702006-01-15 20:48:27 +0000[diff] [blame]1252 <dt><b><tt>insertelement ( VAL, ELT, IDX )</tt></b></dt>
1253
1254 <dd>Perform the <a href="#i_insertelement">insertelement
1255 operation</a> on constants.
1256
Chris Lattnerc1989542006-04-08 00:13:41 +0000[diff] [blame]1257
1258 <dt><b><tt>shufflevector ( VEC1, VEC2, IDXMASK )</tt></b></dt>
1259
1260 <dd>Perform the <a href="#i_shufflevector">shufflevector
1261 operation</a> on constants.
1262
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1263 <dt><b><tt>OPCODE ( LHS, RHS )</tt></b></dt>
1264
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]1265 <dd>Perform the specified operation of the LHS and RHS constants. OPCODE may
1266 be any of the <a href="#binaryops">binary</a> or <a href="#bitwiseops">bitwise
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1267 binary</a> operations. The constraints on operands are the same as those for
1268 the corresponding instruction (e.g. no bitwise operations on floating point
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]1269 values are allowed).</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1270</dl>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1271</div>
Chris Lattner9ee5d222004-03-08 16:49:10 +0000[diff] [blame]1272
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1273<!-- *********************************************************************** -->
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]1274<div class="doc_section"> <a name="othervalues">Other Values</a> </div>
1275<!-- *********************************************************************** -->
1276
1277<!-- ======================================================================= -->
1278<div class="doc_subsection">
1279<a name="inlineasm">Inline Assembler Expressions</a>
1280</div>
1281
1282<div class="doc_text">
1283
1284<p>
1285LLVM supports inline assembler expressions (as opposed to <a href="#moduleasm">
1286Module-Level Inline Assembly</a>) through the use of a special value. This
1287value represents the inline assembler as a string (containing the instructions
1288to emit), a list of operand constraints (stored as a string), and a flag that
1289indicates whether or not the inline asm expression has side effects. An example
1290inline assembler expression is:
1291</p>
1292
1293<pre>
1294 int(int) asm "bswap $0", "=r,r"
1295</pre>
1296
1297<p>
1298Inline assembler expressions may <b>only</b> be used as the callee operand of
1299a <a href="#i_call"><tt>call</tt> instruction</a>. Thus, typically we have:
1300</p>
1301
1302<pre>
1303 %X = call int asm "<a href="#i_bswap">bswap</a> $0", "=r,r"(int %Y)
1304</pre>
1305
1306<p>
1307Inline asms with side effects not visible in the constraint list must be marked
1308as having side effects. This is done through the use of the
1309'<tt>sideeffect</tt>' keyword, like so:
1310</p>
1311
1312<pre>
1313 call void asm sideeffect "eieio", ""()
1314</pre>
1315
1316<p>TODO: The format of the asm and constraints string still need to be
1317documented here. Constraints on what can be done (e.g. duplication, moving, etc
1318need to be documented).
1319</p>
1320
1321</div>
1322
1323<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1324<div class="doc_section"> <a name="instref">Instruction Reference</a> </div>
1325<!-- *********************************************************************** -->
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1326
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1327<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1328
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1329<p>The LLVM instruction set consists of several different
1330classifications of instructions: <a href="#terminators">terminator
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]1331instructions</a>, <a href="#binaryops">binary instructions</a>,
1332<a href="#bitwiseops">bitwise binary instructions</a>, <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1333 href="#memoryops">memory instructions</a>, and <a href="#otherops">other
1334instructions</a>.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1335
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1336</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1337
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1338<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1339<div class="doc_subsection"> <a name="terminators">Terminator
1340Instructions</a> </div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1341
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1342<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1343
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1344<p>As mentioned <a href="#functionstructure">previously</a>, every
1345basic block in a program ends with a "Terminator" instruction, which
1346indicates which block should be executed after the current block is
1347finished. These terminator instructions typically yield a '<tt>void</tt>'
1348value: they produce control flow, not values (the one exception being
1349the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction).</p>
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1350<p>There are six different terminator instructions: the '<a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1351 href="#i_ret"><tt>ret</tt></a>' instruction, the '<a href="#i_br"><tt>br</tt></a>'
1352instruction, the '<a href="#i_switch"><tt>switch</tt></a>' instruction,
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1353the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction, the '<a
1354 href="#i_unwind"><tt>unwind</tt></a>' instruction, and the '<a
1355 href="#i_unreachable"><tt>unreachable</tt></a>' instruction.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1356
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1357</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1358
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1359<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1360<div class="doc_subsubsection"> <a name="i_ret">'<tt>ret</tt>'
1361Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1362<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1363<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1364<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]1365 ret void <i>; Return from void function</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1366</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1367<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1368<p>The '<tt>ret</tt>' instruction is used to return control flow (and a
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]1369value) from a function back to the caller.</p>
John Criswell4457dc92004-04-09 16:48:45 +0000[diff] [blame]1370<p>There are two forms of the '<tt>ret</tt>' instruction: one that
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1371returns a value and then causes control flow, and one that just causes
1372control flow to occur.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1373<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1374<p>The '<tt>ret</tt>' instruction may return any '<a
1375 href="#t_firstclass">first class</a>' type. Notice that a function is
1376not <a href="#wellformed">well formed</a> if there exists a '<tt>ret</tt>'
1377instruction inside of the function that returns a value that does not
1378match the return type of the function.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1379<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1380<p>When the '<tt>ret</tt>' instruction is executed, control flow
1381returns back to the calling function's context. If the caller is a "<a
John Criswellfa081872004-06-25 15:16:57 +0000[diff] [blame]1382 href="#i_call"><tt>call</tt></a>" instruction, execution continues at
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1383the instruction after the call. If the caller was an "<a
1384 href="#i_invoke"><tt>invoke</tt></a>" instruction, execution continues
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]1385at the beginning of the "normal" destination block. If the instruction
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1386returns a value, that value shall set the call or invoke instruction's
1387return value.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1388<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1389<pre> ret int 5 <i>; Return an integer value of 5</i>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1390 ret void <i>; Return from a void function</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1391</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1392</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1393<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1394<div class="doc_subsubsection"> <a name="i_br">'<tt>br</tt>' Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1395<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1396<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1397<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]1398</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1399<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1400<p>The '<tt>br</tt>' instruction is used to cause control flow to
1401transfer to a different basic block in the current function. There are
1402two forms of this instruction, corresponding to a conditional branch
1403and an unconditional branch.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1404<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1405<p>The conditional branch form of the '<tt>br</tt>' instruction takes a
1406single '<tt>bool</tt>' value and two '<tt>label</tt>' values. The
1407unconditional form of the '<tt>br</tt>' instruction takes a single '<tt>label</tt>'
1408value as a target.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1409<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1410<p>Upon execution of a conditional '<tt>br</tt>' instruction, the '<tt>bool</tt>'
1411argument is evaluated. If the value is <tt>true</tt>, control flows
1412to the '<tt>iftrue</tt>' <tt>label</tt> argument. If "cond" is <tt>false</tt>,
1413control flows to the '<tt>iffalse</tt>' <tt>label</tt> argument.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1414<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1415<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
1416 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]1417</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1418<!-- _______________________________________________________________________ -->
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1419<div class="doc_subsubsection">
1420 <a name="i_switch">'<tt>switch</tt>' Instruction</a>
1421</div>
1422
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1423<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1424<h5>Syntax:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1425
1426<pre>
1427 switch &lt;intty&gt; &lt;value&gt;, label &lt;defaultdest&gt; [ &lt;intty&gt; &lt;val&gt;, label &lt;dest&gt; ... ]
1428</pre>
1429
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1430<h5>Overview:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1431
1432<p>The '<tt>switch</tt>' instruction is used to transfer control flow to one of
1433several different places. It is a generalization of the '<tt>br</tt>'
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1434instruction, allowing a branch to occur to one of many possible
1435destinations.</p>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1436
1437
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1438<h5>Arguments:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1439
1440<p>The '<tt>switch</tt>' instruction uses three parameters: an integer
1441comparison value '<tt>value</tt>', a default '<tt>label</tt>' destination, and
1442an array of pairs of comparison value constants and '<tt>label</tt>'s. The
1443table is not allowed to contain duplicate constant entries.</p>
1444
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1445<h5>Semantics:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1446
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1447<p>The <tt>switch</tt> instruction specifies a table of values and
1448destinations. When the '<tt>switch</tt>' instruction is executed, this
John Criswell84114752004-06-25 16:05:06 +0000[diff] [blame]1449table is searched for the given value. If the value is found, control flow is
1450transfered to the corresponding destination; otherwise, control flow is
1451transfered to the default destination.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1452
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1453<h5>Implementation:</h5>
1454
1455<p>Depending on properties of the target machine and the particular
1456<tt>switch</tt> instruction, this instruction may be code generated in different
John Criswell84114752004-06-25 16:05:06 +0000[diff] [blame]1457ways. For example, it could be generated as a series of chained conditional
1458branches or with a lookup table.</p>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1459
1460<h5>Example:</h5>
1461
1462<pre>
1463 <i>; Emulate a conditional br instruction</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1464 %Val = <a href="#i_zext">zext</a> bool %value to int
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1465 switch int %Val, label %truedest [int 0, label %falsedest ]
1466
1467 <i>; Emulate an unconditional br instruction</i>
1468 switch uint 0, label %dest [ ]
1469
1470 <i>; Implement a jump table:</i>
1471 switch uint %val, label %otherwise [ uint 0, label %onzero
1472 uint 1, label %onone
1473 uint 2, label %ontwo ]
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1474</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1475</div>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1476
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1477<!-- _______________________________________________________________________ -->
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1478<div class="doc_subsubsection">
1479 <a name="i_invoke">'<tt>invoke</tt>' Instruction</a>
1480</div>
1481
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1482<div class="doc_text">
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1483
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1484<h5>Syntax:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1485
1486<pre>
1487 &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]1488 to label &lt;normal label&gt; unwind label &lt;exception label&gt;
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1489</pre>
1490
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]1491<h5>Overview:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1492
1493<p>The '<tt>invoke</tt>' instruction causes control to transfer to a specified
1494function, with the possibility of control flow transfer to either the
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]1495'<tt>normal</tt>' label or the
1496'<tt>exception</tt>' label. If the callee function returns with the
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1497"<tt><a href="#i_ret">ret</a></tt>" instruction, control flow will return to the
1498"normal" label. If the callee (or any indirect callees) returns with the "<a
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]1499href="#i_unwind"><tt>unwind</tt></a>" instruction, control is interrupted and
1500continued at the dynamically nearest "exception" label.</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1501
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1502<h5>Arguments:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1503
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1504<p>This instruction requires several arguments:</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1505
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1506<ol>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1507 <li>
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]1508 The optional "cconv" marker indicates which <a href="callingconv">calling
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1509 convention</a> the call should use. If none is specified, the call defaults
1510 to using C calling conventions.
1511 </li>
1512 <li>'<tt>ptr to function ty</tt>': shall be the signature of the pointer to
1513 function value being invoked. In most cases, this is a direct function
1514 invocation, but indirect <tt>invoke</tt>s are just as possible, branching off
1515 an arbitrary pointer to function value.
1516 </li>
1517
1518 <li>'<tt>function ptr val</tt>': An LLVM value containing a pointer to a
1519 function to be invoked. </li>
1520
1521 <li>'<tt>function args</tt>': argument list whose types match the function
1522 signature argument types. If the function signature indicates the function
1523 accepts a variable number of arguments, the extra arguments can be
1524 specified. </li>
1525
1526 <li>'<tt>normal label</tt>': the label reached when the called function
1527 executes a '<tt><a href="#i_ret">ret</a></tt>' instruction. </li>
1528
1529 <li>'<tt>exception label</tt>': the label reached when a callee returns with
1530 the <a href="#i_unwind"><tt>unwind</tt></a> instruction. </li>
1531
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1532</ol>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1533
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1534<h5>Semantics:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1535
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1536<p>This instruction is designed to operate as a standard '<tt><a
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1537href="#i_call">call</a></tt>' instruction in most regards. The primary
1538difference is that it establishes an association with a label, which is used by
1539the runtime library to unwind the stack.</p>
1540
1541<p>This instruction is used in languages with destructors to ensure that proper
1542cleanup is performed in the case of either a <tt>longjmp</tt> or a thrown
1543exception. Additionally, this is important for implementation of
1544'<tt>catch</tt>' clauses in high-level languages that support them.</p>
1545
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1546<h5>Example:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1547<pre>
1548 %retval = invoke int %Test(int 15) to label %Continue
Chris Lattner76b8a332006-05-14 18:23:06 +0000[diff] [blame]1549 unwind label %TestCleanup <i>; {int}:retval set</i>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1550 %retval = invoke <a href="#callingconv">coldcc</a> int %Test(int 15) to label %Continue
Chris Lattner76b8a332006-05-14 18:23:06 +0000[diff] [blame]1551 unwind label %TestCleanup <i>; {int}:retval set</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1552</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1553</div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1554
1555
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]1556<!-- _______________________________________________________________________ -->
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1557
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1558<div class="doc_subsubsection"> <a name="i_unwind">'<tt>unwind</tt>'
1559Instruction</a> </div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1560
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1561<div class="doc_text">
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1562
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]1563<h5>Syntax:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1564<pre>
1565 unwind
1566</pre>
1567
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]1568<h5>Overview:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1569
1570<p>The '<tt>unwind</tt>' instruction unwinds the stack, continuing control flow
1571at the first callee in the dynamic call stack which used an <a
1572href="#i_invoke"><tt>invoke</tt></a> instruction to perform the call. This is
1573primarily used to implement exception handling.</p>
1574
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]1575<h5>Semantics:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1576
1577<p>The '<tt>unwind</tt>' intrinsic causes execution of the current function to
1578immediately halt. The dynamic call stack is then searched for the first <a
1579href="#i_invoke"><tt>invoke</tt></a> instruction on the call stack. Once found,
1580execution continues at the "exceptional" destination block specified by the
1581<tt>invoke</tt> instruction. If there is no <tt>invoke</tt> instruction in the
1582dynamic call chain, undefined behavior results.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1583</div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1584
1585<!-- _______________________________________________________________________ -->
1586
1587<div class="doc_subsubsection"> <a name="i_unreachable">'<tt>unreachable</tt>'
1588Instruction</a> </div>
1589
1590<div class="doc_text">
1591
1592<h5>Syntax:</h5>
1593<pre>
1594 unreachable
1595</pre>
1596
1597<h5>Overview:</h5>
1598
1599<p>The '<tt>unreachable</tt>' instruction has no defined semantics. This
1600instruction is used to inform the optimizer that a particular portion of the
1601code is not reachable. This can be used to indicate that the code after a
1602no-return function cannot be reached, and other facts.</p>
1603
1604<h5>Semantics:</h5>
1605
1606<p>The '<tt>unreachable</tt>' instruction has no defined semantics.</p>
1607</div>
1608
1609
1610
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1611<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1612<div class="doc_subsection"> <a name="binaryops">Binary Operations</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1613<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1614<p>Binary operators are used to do most of the computation in a
1615program. They require two operands, execute an operation on them, and
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1616produce a single value. The operands might represent
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1617multiple data, as is the case with the <a href="#t_packed">packed</a> data type.
1618The result value of a binary operator is not
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1619necessarily the same type as its operands.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1620<p>There are several different binary operators:</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1621</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1622<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1623<div class="doc_subsubsection"> <a name="i_add">'<tt>add</tt>'
1624Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1625<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1626<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1627<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]1628</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1629<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1630<p>The '<tt>add</tt>' instruction returns the sum of its two operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1631<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1632<p>The two arguments to the '<tt>add</tt>' instruction must be either <a
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1633 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a> values.
1634 This instruction can also take <a href="#t_packed">packed</a> versions of the values.
1635Both arguments must have identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1636<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1637<p>The value produced is the integer or floating point sum of the two
1638operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1639<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1640<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]1641</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1642</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1643<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1644<div class="doc_subsubsection"> <a name="i_sub">'<tt>sub</tt>'
1645Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1646<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1647<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1648<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]1649</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1650<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1651<p>The '<tt>sub</tt>' instruction returns the difference of its two
1652operands.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1653<p>Note that the '<tt>sub</tt>' instruction is used to represent the '<tt>neg</tt>'
1654instruction present in most other intermediate representations.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1655<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1656<p>The two arguments to the '<tt>sub</tt>' instruction must be either <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1657 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1658values.
1659This instruction can also take <a href="#t_packed">packed</a> versions of the values.
1660Both arguments must have identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1661<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1662<p>The value produced is the integer or floating point difference of
1663the two operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1664<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1665<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]1666 &lt;result&gt; = sub int 0, %val <i>; yields {int}:result = -%var</i>
1667</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1668</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1669<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1670<div class="doc_subsubsection"> <a name="i_mul">'<tt>mul</tt>'
1671Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1672<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1673<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1674<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]1675</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1676<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1677<p>The '<tt>mul</tt>' instruction returns the product of its two
1678operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1679<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1680<p>The two arguments to the '<tt>mul</tt>' instruction must be either <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1681 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1682values.
1683This instruction can also take <a href="#t_packed">packed</a> versions of the values.
1684Both arguments must have identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1685<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1686<p>The value produced is the integer or floating point product of the
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1687two operands.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1688<p>There is no signed vs unsigned multiplication. The appropriate
1689action is taken based on the type of the operand.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1690<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1691<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]1692</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1693</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1694<!-- _______________________________________________________________________ -->
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]1695<div class="doc_subsubsection"> <a name="i_udiv">'<tt>udiv</tt>' Instruction
1696</a></div>
1697<div class="doc_text">
1698<h5>Syntax:</h5>
1699<pre> &lt;result&gt; = udiv &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
1700</pre>
1701<h5>Overview:</h5>
1702<p>The '<tt>udiv</tt>' instruction returns the quotient of its two
1703operands.</p>
1704<h5>Arguments:</h5>
1705<p>The two arguments to the '<tt>udiv</tt>' instruction must be
1706<a href="#t_integer">integer</a> values. Both arguments must have identical
1707types. This instruction can also take <a href="#t_packed">packed</a> versions
1708of the values in which case the elements must be integers.</p>
1709<h5>Semantics:</h5>
1710<p>The value produced is the unsigned integer quotient of the two operands. This
1711instruction always performs an unsigned division operation, regardless of
1712whether the arguments are unsigned or not.</p>
1713<h5>Example:</h5>
1714<pre> &lt;result&gt; = udiv uint 4, %var <i>; yields {uint}:result = 4 / %var</i>
1715</pre>
1716</div>
1717<!-- _______________________________________________________________________ -->
1718<div class="doc_subsubsection"> <a name="i_sdiv">'<tt>sdiv</tt>' Instruction
1719</a> </div>
1720<div class="doc_text">
1721<h5>Syntax:</h5>
1722<pre> &lt;result&gt; = sdiv &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
1723</pre>
1724<h5>Overview:</h5>
1725<p>The '<tt>sdiv</tt>' instruction returns the quotient of its two
1726operands.</p>
1727<h5>Arguments:</h5>
1728<p>The two arguments to the '<tt>sdiv</tt>' instruction must be
1729<a href="#t_integer">integer</a> values. Both arguments must have identical
1730types. This instruction can also take <a href="#t_packed">packed</a> versions
1731of the values in which case the elements must be integers.</p>
1732<h5>Semantics:</h5>
1733<p>The value produced is the signed integer quotient of the two operands. This
1734instruction always performs a signed division operation, regardless of whether
1735the arguments are signed or not.</p>
1736<h5>Example:</h5>
1737<pre> &lt;result&gt; = sdiv int 4, %var <i>; yields {int}:result = 4 / %var</i>
1738</pre>
1739</div>
1740<!-- _______________________________________________________________________ -->
1741<div class="doc_subsubsection"> <a name="i_fdiv">'<tt>fdiv</tt>'
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1742Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1743<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1744<h5>Syntax:</h5>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]1745<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]1746</pre>
1747<h5>Overview:</h5>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]1748<p>The '<tt>fdiv</tt>' instruction returns the quotient of its two
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1749operands.</p>
1750<h5>Arguments:</h5>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]1751<p>The two arguments to the '<tt>div</tt>' instruction must be
1752<a href="#t_floating">floating point</a> values. Both arguments must have
1753identical types. This instruction can also take <a href="#t_packed">packed</a>
1754versions of the values in which case the elements must be floating point.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1755<h5>Semantics:</h5>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]1756<p>The value produced is the floating point quotient of the two operands.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1757<h5>Example:</h5>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]1758<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]1759</pre>
1760</div>
1761<!-- _______________________________________________________________________ -->
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]1762<div class="doc_subsubsection"> <a name="i_urem">'<tt>urem</tt>' Instruction</a>
1763</div>
1764<div class="doc_text">
1765<h5>Syntax:</h5>
1766<pre> &lt;result&gt; = urem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
1767</pre>
1768<h5>Overview:</h5>
1769<p>The '<tt>urem</tt>' instruction returns the remainder from the
1770unsigned division of its two arguments.</p>
1771<h5>Arguments:</h5>
1772<p>The two arguments to the '<tt>urem</tt>' instruction must be
1773<a href="#t_integer">integer</a> values. Both arguments must have identical
1774types.</p>
1775<h5>Semantics:</h5>
1776<p>This instruction returns the unsigned integer <i>remainder</i> of a division.
1777This instruction always performs an unsigned division to get the remainder,
1778regardless of whether the arguments are unsigned or not.</p>
1779<h5>Example:</h5>
1780<pre> &lt;result&gt; = urem uint 4, %var <i>; yields {uint}:result = 4 % %var</i>
1781</pre>
1782
1783</div>
1784<!-- _______________________________________________________________________ -->
1785<div class="doc_subsubsection"> <a name="i_srem">'<tt>srem</tt>'
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1786Instruction</a> </div>
1787<div class="doc_text">
1788<h5>Syntax:</h5>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]1789<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]1790</pre>
1791<h5>Overview:</h5>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]1792<p>The '<tt>srem</tt>' instruction returns the remainder from the
1793signed division of its two operands.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1794<h5>Arguments:</h5>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]1795<p>The two arguments to the '<tt>srem</tt>' instruction must be
1796<a href="#t_integer">integer</a> values. Both arguments must have identical
1797types.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1798<h5>Semantics:</h5>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]1799<p>This instruction returns the <i>remainder</i> of a division (where the result
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1800has the same sign as the divisor), not the <i>modulus</i> (where the
1801result has the same sign as the dividend) of a value. For more
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]1802information about the difference, see <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1803 href="http://mathforum.org/dr.math/problems/anne.4.28.99.html">The
1804Math Forum</a>.</p>
1805<h5>Example:</h5>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]1806<pre> &lt;result&gt; = srem int 4, %var <i>; yields {int}:result = 4 % %var</i>
1807</pre>
1808
1809</div>
1810<!-- _______________________________________________________________________ -->
1811<div class="doc_subsubsection"> <a name="i_frem">'<tt>frem</tt>'
1812Instruction</a> </div>
1813<div class="doc_text">
1814<h5>Syntax:</h5>
1815<pre> &lt;result&gt; = frem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
1816</pre>
1817<h5>Overview:</h5>
1818<p>The '<tt>frem</tt>' instruction returns the remainder from the
1819division of its two operands.</p>
1820<h5>Arguments:</h5>
1821<p>The two arguments to the '<tt>frem</tt>' instruction must be
1822<a href="#t_floating">floating point</a> values. Both arguments must have
1823identical types.</p>
1824<h5>Semantics:</h5>
1825<p>This instruction returns the <i>remainder</i> of a division.</p>
1826<h5>Example:</h5>
1827<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]1828</pre>
Robert Bocchino7b81c752006-02-17 21:18:08 +0000[diff] [blame]1829
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1830</div>
1831<!-- _______________________________________________________________________ -->
1832<div class="doc_subsubsection"> <a name="i_setcc">'<tt>set<i>cc</i></tt>'
1833Instructions</a> </div>
1834<div class="doc_text">
1835<h5>Syntax:</h5>
1836<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]1837 &lt;result&gt; = setne &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {bool}:result</i>
1838 &lt;result&gt; = setlt &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {bool}:result</i>
1839 &lt;result&gt; = setgt &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {bool}:result</i>
1840 &lt;result&gt; = setle &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {bool}:result</i>
1841 &lt;result&gt; = setge &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {bool}:result</i>
1842</pre>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1843<h5>Overview:</h5>
1844<p>The '<tt>set<i>cc</i></tt>' family of instructions returns a boolean
1845value based on a comparison of their two operands.</p>
1846<h5>Arguments:</h5>
1847<p>The two arguments to the '<tt>set<i>cc</i></tt>' instructions must
1848be of <a href="#t_firstclass">first class</a> type (it is not possible
1849to compare '<tt>label</tt>'s, '<tt>array</tt>'s, '<tt>structure</tt>'
1850or '<tt>void</tt>' values, etc...). Both arguments must have identical
1851types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1852<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1853<p>The '<tt>seteq</tt>' instruction yields a <tt>true</tt> '<tt>bool</tt>'
1854value if both operands are equal.<br>
1855The '<tt>setne</tt>' instruction yields a <tt>true</tt> '<tt>bool</tt>'
1856value if both operands are unequal.<br>
1857The '<tt>setlt</tt>' instruction yields a <tt>true</tt> '<tt>bool</tt>'
1858value if the first operand is less than the second operand.<br>
1859The '<tt>setgt</tt>' instruction yields a <tt>true</tt> '<tt>bool</tt>'
1860value if the first operand is greater than the second operand.<br>
1861The '<tt>setle</tt>' instruction yields a <tt>true</tt> '<tt>bool</tt>'
1862value if the first operand is less than or equal to the second operand.<br>
1863The '<tt>setge</tt>' instruction yields a <tt>true</tt> '<tt>bool</tt>'
1864value if the first operand is greater than or equal to the second
1865operand.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1866<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1867<pre> &lt;result&gt; = seteq int 4, 5 <i>; yields {bool}:result = false</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1868 &lt;result&gt; = setne float 4, 5 <i>; yields {bool}:result = true</i>
1869 &lt;result&gt; = setlt uint 4, 5 <i>; yields {bool}:result = true</i>
1870 &lt;result&gt; = setgt sbyte 4, 5 <i>; yields {bool}:result = false</i>
1871 &lt;result&gt; = setle sbyte 4, 5 <i>; yields {bool}:result = true</i>
1872 &lt;result&gt; = setge sbyte 4, 5 <i>; yields {bool}:result = false</i>
1873</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1874</div>
Robert Bocchino7b81c752006-02-17 21:18:08 +0000[diff] [blame]1875
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1876<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1877<div class="doc_subsection"> <a name="bitwiseops">Bitwise Binary
1878Operations</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1879<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1880<p>Bitwise binary operators are used to do various forms of
1881bit-twiddling in a program. They are generally very efficient
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1882instructions and can commonly be strength reduced from other
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1883instructions. They require two operands, execute an operation on them,
1884and produce a single value. The resulting value of the bitwise binary
1885operators is always the same type as its first operand.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1886</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1887<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1888<div class="doc_subsubsection"> <a name="i_and">'<tt>and</tt>'
1889Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1890<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1891<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1892<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]1893</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1894<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1895<p>The '<tt>and</tt>' instruction returns the bitwise logical and of
1896its two operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1897<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1898<p>The two arguments to the '<tt>and</tt>' instruction must be <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1899 href="#t_integral">integral</a> values. Both arguments must have
1900identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1901<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1902<p>The truth table used for the '<tt>and</tt>' instruction is:</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1903<p> </p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1904<div style="align: center">
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1905<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1906 <tbody>
1907 <tr>
1908 <td>In0</td>
1909 <td>In1</td>
1910 <td>Out</td>
1911 </tr>
1912 <tr>
1913 <td>0</td>
1914 <td>0</td>
1915 <td>0</td>
1916 </tr>
1917 <tr>
1918 <td>0</td>
1919 <td>1</td>
1920 <td>0</td>
1921 </tr>
1922 <tr>
1923 <td>1</td>
1924 <td>0</td>
1925 <td>0</td>
1926 </tr>
1927 <tr>
1928 <td>1</td>
1929 <td>1</td>
1930 <td>1</td>
1931 </tr>
1932 </tbody>
1933</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1934</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1935<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1936<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]1937 &lt;result&gt; = and int 15, 40 <i>; yields {int}:result = 8</i>
1938 &lt;result&gt; = and int 4, 8 <i>; yields {int}:result = 0</i>
1939</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1940</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1941<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1942<div class="doc_subsubsection"> <a name="i_or">'<tt>or</tt>' Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1943<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1944<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1945<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]1946</pre>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1947<h5>Overview:</h5>
1948<p>The '<tt>or</tt>' instruction returns the bitwise logical inclusive
1949or of its two operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1950<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1951<p>The two arguments to the '<tt>or</tt>' instruction must be <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1952 href="#t_integral">integral</a> values. Both arguments must have
1953identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1954<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1955<p>The truth table used for the '<tt>or</tt>' instruction is:</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1956<p> </p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1957<div style="align: center">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1958<table border="1" cellspacing="0" cellpadding="4">
1959 <tbody>
1960 <tr>
1961 <td>In0</td>
1962 <td>In1</td>
1963 <td>Out</td>
1964 </tr>
1965 <tr>
1966 <td>0</td>
1967 <td>0</td>
1968 <td>0</td>
1969 </tr>
1970 <tr>
1971 <td>0</td>
1972 <td>1</td>
1973 <td>1</td>
1974 </tr>
1975 <tr>
1976 <td>1</td>
1977 <td>0</td>
1978 <td>1</td>
1979 </tr>
1980 <tr>
1981 <td>1</td>
1982 <td>1</td>
1983 <td>1</td>
1984 </tr>
1985 </tbody>
1986</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1987</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1988<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1989<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]1990 &lt;result&gt; = or int 15, 40 <i>; yields {int}:result = 47</i>
1991 &lt;result&gt; = or int 4, 8 <i>; yields {int}:result = 12</i>
1992</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1993</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1994<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1995<div class="doc_subsubsection"> <a name="i_xor">'<tt>xor</tt>'
1996Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1997<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1998<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1999<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]2000</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2001<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2002<p>The '<tt>xor</tt>' instruction returns the bitwise logical exclusive
2003or of its two operands. The <tt>xor</tt> is used to implement the
2004"one's complement" operation, which is the "~" operator in C.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2005<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2006<p>The two arguments to the '<tt>xor</tt>' instruction must be <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2007 href="#t_integral">integral</a> values. Both arguments must have
2008identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2009<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2010<p>The truth table used for the '<tt>xor</tt>' instruction is:</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2011<p> </p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]2012<div style="align: center">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2013<table border="1" cellspacing="0" cellpadding="4">
2014 <tbody>
2015 <tr>
2016 <td>In0</td>
2017 <td>In1</td>
2018 <td>Out</td>
2019 </tr>
2020 <tr>
2021 <td>0</td>
2022 <td>0</td>
2023 <td>0</td>
2024 </tr>
2025 <tr>
2026 <td>0</td>
2027 <td>1</td>
2028 <td>1</td>
2029 </tr>
2030 <tr>
2031 <td>1</td>
2032 <td>0</td>
2033 <td>1</td>
2034 </tr>
2035 <tr>
2036 <td>1</td>
2037 <td>1</td>
2038 <td>0</td>
2039 </tr>
2040 </tbody>
2041</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]2042</div>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2043<p> </p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2044<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2045<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]2046 &lt;result&gt; = xor int 15, 40 <i>; yields {int}:result = 39</i>
2047 &lt;result&gt; = xor int 4, 8 <i>; yields {int}:result = 12</i>
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]2048 &lt;result&gt; = xor int %V, -1 <i>; yields {int}:result = ~%V</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2049</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2050</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2051<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2052<div class="doc_subsubsection"> <a name="i_shl">'<tt>shl</tt>'
2053Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2054<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2055<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2056<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]2057</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2058<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2059<p>The '<tt>shl</tt>' instruction returns the first operand shifted to
2060the left a specified number of bits.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2061<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2062<p>The first argument to the '<tt>shl</tt>' instruction must be an <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2063 href="#t_integer">integer</a> type. The second argument must be an '<tt>ubyte</tt>'
2064type.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2065<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2066<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]2067<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2068<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]2069 &lt;result&gt; = shl int 4, ubyte 2 <i>; yields {int}:result = 16</i>
2070 &lt;result&gt; = shl int 1, ubyte 10 <i>; yields {int}:result = 1024</i>
2071</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2072</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2073<!-- _______________________________________________________________________ -->
Reid Spencer3822ff52006-11-08 06:47:33 +0000[diff] [blame]2074<div class="doc_subsubsection"> <a name="i_lshr">'<tt>lshr</tt>'
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2075Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2076<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2077<h5>Syntax:</h5>
Reid Spencer3822ff52006-11-08 06:47:33 +0000[diff] [blame]2078<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]2079</pre>
Reid Spencer3822ff52006-11-08 06:47:33 +0000[diff] [blame]2080
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2081<h5>Overview:</h5>
Reid Spencer3822ff52006-11-08 06:47:33 +0000[diff] [blame]2082<p>The '<tt>lshr</tt>' instruction (logical shift right) returns the first
2083operand shifted to the right a specified number of bits.</p>
2084
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2085<h5>Arguments:</h5>
Reid Spencer3822ff52006-11-08 06:47:33 +0000[diff] [blame]2086<p>The first argument to the '<tt>lshr</tt>' instruction must be an <a
2087 href="#t_integer">integer</a> type. The second argument must be an '<tt>ubyte</tt>' type.</p>
2088
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2089<h5>Semantics:</h5>
Reid Spencer3822ff52006-11-08 06:47:33 +0000[diff] [blame]2090<p>This instruction always performs a logical shift right operation, regardless
2091of whether the arguments are unsigned or not. The <tt>var2</tt> most significant
2092bits will be filled with zero bits after the shift.</p>
2093
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2094<h5>Example:</h5>
Reid Spencer3822ff52006-11-08 06:47:33 +0000[diff] [blame]2095<pre>
2096 &lt;result&gt; = lshr uint 4, ubyte 1 <i>; yields {uint}:result = 2</i>
2097 &lt;result&gt; = lshr int 4, ubyte 2 <i>; yields {uint}:result = 1</i>
2098 &lt;result&gt; = lshr sbyte 4, ubyte 3 <i>; yields {sbyte}:result = 0</i>
2099 &lt;result&gt; = lshr sbyte -2, ubyte 1 <i>; yields {sbyte}:result = 0x7FFFFFFF </i>
2100</pre>
2101</div>
2102
2103<!-- ======================================================================= -->
2104<div class="doc_subsubsection"> <a name="i_ashr">'<tt>ashr</tt>'
2105Instruction</a> </div>
2106<div class="doc_text">
2107
2108<h5>Syntax:</h5>
2109<pre> &lt;result&gt; = ashr &lt;ty&gt; &lt;var1&gt;, ubyte &lt;var2&gt; <i>; yields {ty}:result</i>
2110</pre>
2111
2112<h5>Overview:</h5>
2113<p>The '<tt>ashr</tt>' instruction (arithmetic shift right) returns the first
2114operand shifted to the right a specified number of bits.</p>
2115
2116<h5>Arguments:</h5>
2117<p>The first argument to the '<tt>ashr</tt>' instruction must be an
2118<a href="#t_integer">integer</a> type. The second argument must be an
2119'<tt>ubyte</tt>' type.</p>
2120
2121<h5>Semantics:</h5>
2122<p>This instruction always performs an arithmetic shift right operation,
2123regardless of whether the arguments are signed or not. The <tt>var2</tt> most
2124significant bits will be filled with the sign bit of <tt>var1</tt>.</p>
2125
2126<h5>Example:</h5>
2127<pre>
2128 &lt;result&gt; = ashr uint 4, ubyte 1 <i>; yields {uint}:result = 2</i>
2129 &lt;result&gt; = ashr int 4, ubyte 2 <i>; yields {int}:result = 1</i>
2130 &lt;result&gt; = ashr ubyte 4, ubyte 3 <i>; yields {ubyte}:result = 0</i>
2131 &lt;result&gt; = ashr sbyte -2, ubyte 1 <i>; yields {sbyte}:result = -1</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2132</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2133</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2134
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2135<!-- ======================================================================= -->
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2136<div class="doc_subsection">
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2137 <a name="vectorops">Vector Operations</a>
2138</div>
2139
2140<div class="doc_text">
2141
2142<p>LLVM supports several instructions to represent vector operations in a
2143target-independent manner. This instructions cover the element-access and
2144vector-specific operations needed to process vectors effectively. While LLVM
2145does directly support these vector operations, many sophisticated algorithms
2146will want to use target-specific intrinsics to take full advantage of a specific
2147target.</p>
2148
2149</div>
2150
2151<!-- _______________________________________________________________________ -->
2152<div class="doc_subsubsection">
2153 <a name="i_extractelement">'<tt>extractelement</tt>' Instruction</a>
2154</div>
2155
2156<div class="doc_text">
2157
2158<h5>Syntax:</h5>
2159
2160<pre>
2161 &lt;result&gt; = extractelement &lt;n x &lt;ty&gt;&gt; &lt;val&gt;, uint &lt;idx&gt; <i>; yields &lt;ty&gt;</i>
2162</pre>
2163
2164<h5>Overview:</h5>
2165
2166<p>
2167The '<tt>extractelement</tt>' instruction extracts a single scalar
2168element from a packed vector at a specified index.
2169</p>
2170
2171
2172<h5>Arguments:</h5>
2173
2174<p>
2175The first operand of an '<tt>extractelement</tt>' instruction is a
2176value of <a href="#t_packed">packed</a> type. The second operand is
2177an index indicating the position from which to extract the element.
2178The index may be a variable.</p>
2179
2180<h5>Semantics:</h5>
2181
2182<p>
2183The result is a scalar of the same type as the element type of
2184<tt>val</tt>. Its value is the value at position <tt>idx</tt> of
2185<tt>val</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
2186results are undefined.
2187</p>
2188
2189<h5>Example:</h5>
2190
2191<pre>
2192 %result = extractelement &lt;4 x int&gt; %vec, uint 0 <i>; yields int</i>
2193</pre>
2194</div>
2195
2196
2197<!-- _______________________________________________________________________ -->
2198<div class="doc_subsubsection">
2199 <a name="i_insertelement">'<tt>insertelement</tt>' Instruction</a>
2200</div>
2201
2202<div class="doc_text">
2203
2204<h5>Syntax:</h5>
2205
2206<pre>
2207 &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>
2208</pre>
2209
2210<h5>Overview:</h5>
2211
2212<p>
2213The '<tt>insertelement</tt>' instruction inserts a scalar
2214element into a packed vector at a specified index.
2215</p>
2216
2217
2218<h5>Arguments:</h5>
2219
2220<p>
2221The first operand of an '<tt>insertelement</tt>' instruction is a
2222value of <a href="#t_packed">packed</a> type. The second operand is a
2223scalar value whose type must equal the element type of the first
2224operand. The third operand is an index indicating the position at
2225which to insert the value. The index may be a variable.</p>
2226
2227<h5>Semantics:</h5>
2228
2229<p>
2230The result is a packed vector of the same type as <tt>val</tt>. Its
2231element values are those of <tt>val</tt> except at position
2232<tt>idx</tt>, where it gets the value <tt>elt</tt>. If <tt>idx</tt>
2233exceeds the length of <tt>val</tt>, the results are undefined.
2234</p>
2235
2236<h5>Example:</h5>
2237
2238<pre>
2239 %result = insertelement &lt;4 x int&gt; %vec, int 1, uint 0 <i>; yields &lt;4 x int&gt;</i>
2240</pre>
2241</div>
2242
2243<!-- _______________________________________________________________________ -->
2244<div class="doc_subsubsection">
2245 <a name="i_shufflevector">'<tt>shufflevector</tt>' Instruction</a>
2246</div>
2247
2248<div class="doc_text">
2249
2250<h5>Syntax:</h5>
2251
2252<pre>
2253 &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>
2254</pre>
2255
2256<h5>Overview:</h5>
2257
2258<p>
2259The '<tt>shufflevector</tt>' instruction constructs a permutation of elements
2260from two input vectors, returning a vector of the same type.
2261</p>
2262
2263<h5>Arguments:</h5>
2264
2265<p>
2266The first two operands of a '<tt>shufflevector</tt>' instruction are vectors
2267with types that match each other and types that match the result of the
2268instruction. The third argument is a shuffle mask, which has the same number
2269of elements as the other vector type, but whose element type is always 'uint'.
2270</p>
2271
2272<p>
2273The shuffle mask operand is required to be a constant vector with either
2274constant integer or undef values.
2275</p>
2276
2277<h5>Semantics:</h5>
2278
2279<p>
2280The elements of the two input vectors are numbered from left to right across
2281both of the vectors. The shuffle mask operand specifies, for each element of
2282the result vector, which element of the two input registers the result element
2283gets. The element selector may be undef (meaning "don't care") and the second
2284operand may be undef if performing a shuffle from only one vector.
2285</p>
2286
2287<h5>Example:</h5>
2288
2289<pre>
2290 %result = shufflevector &lt;4 x int&gt; %v1, &lt;4 x int&gt; %v2,
2291 &lt;4 x uint&gt; &lt;uint 0, uint 4, uint 1, uint 5&gt; <i>; yields &lt;4 x int&gt;</i>
2292 %result = shufflevector &lt;4 x int&gt; %v1, &lt;4 x int&gt; undef,
2293 &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.
2294</pre>
2295</div>
2296
Tanya Lattner09474292006-04-14 19:24:33 +0000[diff] [blame]2297
2298<!-- _______________________________________________________________________ -->
2299<div class="doc_subsubsection"> <a name="i_vsetint">'<tt>vsetint</tt>'
2300Instruction</a> </div>
2301<div class="doc_text">
2302<h5>Syntax:</h5>
2303<pre>&lt;result&gt; = vsetint &lt;op&gt;, &lt;n x &lt;ty&gt;&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields &lt;n x bool&gt;</i>
2304</pre>
2305
2306<h5>Overview:</h5>
2307
2308<p>The '<tt>vsetint</tt>' instruction takes two integer vectors and
2309returns a vector of boolean values representing, at each position, the
2310result of the comparison between the values at that position in the
2311two operands.</p>
2312
2313<h5>Arguments:</h5>
2314
2315<p>The arguments to a '<tt>vsetint</tt>' instruction are a comparison
2316operation and two value arguments. The value arguments must be of <a
2317href="#t_integral">integral</a> <a href="#t_packed">packed</a> type,
2318and they must have identical types. The operation argument must be
2319one of <tt>eq</tt>, <tt>ne</tt>, <tt>slt</tt>, <tt>sgt</tt>,
2320<tt>sle</tt>, <tt>sge</tt>, <tt>ult</tt>, <tt>ugt</tt>, <tt>ule</tt>,
2321<tt>uge</tt>, <tt>true</tt>, and <tt>false</tt>. The result is a
2322packed <tt>bool</tt> value with the same length as each operand.</p>
2323
2324<h5>Semantics:</h5>
2325
2326<p>The following table shows the semantics of '<tt>vsetint</tt>'. For
2327each position of the result, the comparison is done on the
2328corresponding positions of the two value arguments. Note that the
2329signedness of the comparison depends on the comparison opcode and
2330<i>not</i> on the signedness of the value operands. E.g., <tt>vsetint
2331slt <4 x unsigned> %x, %y</tt> does an elementwise <i>signed</i>
2332comparison of <tt>%x</tt> and <tt>%y</tt>.</p>
2333
2334<table border="1" cellspacing="0" cellpadding="4">
2335 <tbody>
2336 <tr><th>Operation</th><th>Result is true iff</th><th>Comparison is</th></tr>
2337 <tr><td><tt>eq</tt></td><td>var1 == var2</td><td>--</td></tr>
2338 <tr><td><tt>ne</tt></td><td>var1 != var2</td><td>--</td></tr>
2339 <tr><td><tt>slt</tt></td><td>var1 &lt; var2</td><td>signed</td></tr>
2340 <tr><td><tt>sgt</tt></td><td>var1 &gt; var2</td><td>signed</td></tr>
2341 <tr><td><tt>sle</tt></td><td>var1 &lt;= var2</td><td>signed</td></tr>
2342 <tr><td><tt>sge</tt></td><td>var1 &gt;= var2</td><td>signed</td></tr>
2343 <tr><td><tt>ult</tt></td><td>var1 &lt; var2</td><td>unsigned</td></tr>
2344 <tr><td><tt>ugt</tt></td><td>var1 &gt; var2</td><td>unsigned</td></tr>
2345 <tr><td><tt>ule</tt></td><td>var1 &lt;= var2</td><td>unsigned</td></tr>
2346 <tr><td><tt>uge</tt></td><td>var1 &gt;= var2</td><td>unsigned</td></tr>
2347 <tr><td><tt>true</tt></td><td>always</td><td>--</td></tr>
2348 <tr><td><tt>false</tt></td><td>never</td><td>--</td></tr>
2349 </tbody>
2350</table>
2351
2352<h5>Example:</h5>
2353<pre> &lt;result&gt; = vsetint eq &lt;2 x int&gt; &lt;int 0, int 1&gt;, &lt;int 1, int 0&gt; <i>; yields {&lt;2 x bool&gt;}:result = false, false</i>
2354 &lt;result&gt; = vsetint ne &lt;2 x int&gt; &lt;int 0, int 1&gt;, &lt;int 1, int 0&gt; <i>; yields {&lt;2 x bool&gt;}:result = true, true</i>
2355 &lt;result&gt; = vsetint slt &lt;2 x int&gt; &lt;int 0, int 1&gt;, &lt;int 1, int 0&gt; <i>; yields {&lt;2 x bool&gt;}:result = true, false</i>
2356 &lt;result&gt; = vsetint sgt &lt;2 x int&gt; &lt;int 0, int 1&gt;, &lt;int 1, int 0&gt; <i>; yields {&lt;2 x bool&gt;}:result = false, true</i>
2357 &lt;result&gt; = vsetint sle &lt;2 x int&gt; &lt;int 0, int 1&gt;, &lt;int 1, int 0&gt; <i>; yields {&lt;2 x bool&gt;}:result = true, false</i>
2358 &lt;result&gt; = vsetint sge &lt;2 x int&gt; &lt;int 0, int 1&gt;, &lt;int 1, int 0&gt; <i>; yields {&lt;2 x bool&gt;}:result = false, true</i>
2359</pre>
2360</div>
2361
2362<!-- _______________________________________________________________________ -->
2363<div class="doc_subsubsection"> <a name="i_vsetfp">'<tt>vsetfp</tt>'
2364Instruction</a> </div>
2365<div class="doc_text">
2366<h5>Syntax:</h5>
2367<pre>&lt;result&gt; = vsetfp &lt;op&gt;, &lt;n x &lt;ty&gt;&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields &lt;n x bool&gt;</i>
2368</pre>
2369
2370<h5>Overview:</h5>
2371
2372<p>The '<tt>vsetfp</tt>' instruction takes two floating point vector
2373arguments and returns a vector of boolean values representing, at each
2374position, the result of the comparison between the values at that
2375position in the two operands.</p>
2376
2377<h5>Arguments:</h5>
2378
2379<p>The arguments to a '<tt>vsetfp</tt>' instruction are a comparison
2380operation and two value arguments. The value arguments must be of <a
2381href="t_floating">floating point</a> <a href="#t_packed">packed</a>
2382type, and they must have identical types. The operation argument must
2383be one of <tt>eq</tt>, <tt>ne</tt>, <tt>lt</tt>, <tt>gt</tt>,
2384<tt>le</tt>, <tt>ge</tt>, <tt>oeq</tt>, <tt>one</tt>, <tt>olt</tt>,
2385<tt>ogt</tt>, <tt>ole</tt>, <tt>oge</tt>, <tt>ueq</tt>, <tt>une</tt>,
2386<tt>ult</tt>, <tt>ugt</tt>, <tt>ule</tt>, <tt>uge</tt>, <tt>o</tt>,
2387<tt>u</tt>, <tt>true</tt>, and <tt>false</tt>. The result is a packed
2388<tt>bool</tt> value with the same length as each operand.</p>
2389
2390<h5>Semantics:</h5>
2391
2392<p>The following table shows the semantics of '<tt>vsetfp</tt>' for
2393floating point types. If either operand is a floating point Not a
2394Number (NaN) value, the operation is unordered, and the value in the
2395first column below is produced at that position. Otherwise, the
2396operation is ordered, and the value in the second column is
2397produced.</p>
2398
2399<table border="1" cellspacing="0" cellpadding="4">
2400 <tbody>
2401 <tr><th>Operation</th><th>If unordered<th>Otherwise true iff</th></tr>
2402 <tr><td><tt>eq</tt></td><td>undefined</td><td>var1 == var2</td></tr>
2403 <tr><td><tt>ne</tt></td><td>undefined</td><td>var1 != var2</td></tr>
2404 <tr><td><tt>lt</tt></td><td>undefined</td><td>var1 &lt; var2</td></tr>
2405 <tr><td><tt>gt</tt></td><td>undefined</td><td>var1 &gt; var2</td></tr>
2406 <tr><td><tt>le</tt></td><td>undefined</td><td>var1 &lt;= var2</td></tr>
2407 <tr><td><tt>ge</tt></td><td>undefined</td><td>var1 &gt;= var2</td></tr>
2408 <tr><td><tt>oeq</tt></td><td>false</td><td>var1 == var2</td></tr>
2409 <tr><td><tt>one</tt></td><td>false</td><td>var1 != var2</td></tr>
2410 <tr><td><tt>olt</tt></td><td>false</td><td>var1 &lt; var2</td></tr>
2411 <tr><td><tt>ogt</tt></td><td>false</td><td>var1 &gt; var2</td></tr>
2412 <tr><td><tt>ole</tt></td><td>false</td><td>var1 &lt;= var2</td></tr>
2413 <tr><td><tt>oge</tt></td><td>false</td><td>var1 &gt;= var2</td></tr>
2414 <tr><td><tt>ueq</tt></td><td>true</td><td>var1 == var2</td></tr>
2415 <tr><td><tt>une</tt></td><td>true</td><td>var1 != var2</td></tr>
2416 <tr><td><tt>ult</tt></td><td>true</td><td>var1 &lt; var2</td></tr>
2417 <tr><td><tt>ugt</tt></td><td>true</td><td>var1 &gt; var2</td></tr>
2418 <tr><td><tt>ule</tt></td><td>true</td><td>var1 &lt;= var2</td></tr>
2419 <tr><td><tt>uge</tt></td><td>true</td><td>var1 &gt;= var2</td></tr>
2420 <tr><td><tt>o</tt></td><td>false</td><td>always</td></tr>
2421 <tr><td><tt>u</tt></td><td>true</td><td>never</td></tr>
2422 <tr><td><tt>true</tt></td><td>true</td><td>always</td></tr>
2423 <tr><td><tt>false</tt></td><td>false</td><td>never</td></tr>
2424 </tbody>
2425</table>
2426
2427<h5>Example:</h5>
2428<pre> &lt;result&gt; = vsetfp eq &lt;2 x float&gt; &lt;float 0.0, float 1.0&gt;, &lt;float 1.0, float 0.0&gt; <i>; yields {&lt;2 x bool&gt;}:result = false, false</i>
2429 &lt;result&gt; = vsetfp ne &lt;2 x float&gt; &lt;float 0.0, float 1.0&gt;, &lt;float 1.0, float 0.0&gt; <i>; yields {&lt;2 x bool&gt;}:result = true, true</i>
2430 &lt;result&gt; = vsetfp lt &lt;2 x float&gt; &lt;float 0.0, float 1.0&gt;, &lt;float 1.0, float 0.0&gt; <i>; yields {&lt;2 x bool&gt;}:result = true, false</i>
2431 &lt;result&gt; = vsetfp gt &lt;2 x float&gt; &lt;float 0.0, float 1.0&gt;, &lt;float 1.0, float 0.0&gt; <i>; yields {&lt;2 x bool&gt;}:result = false, true</i>
2432 &lt;result&gt; = vsetfp le &lt;2 x float&gt; &lt;float 0.0, float 1.0&gt;, &lt;float 1.0, float 0.0&gt; <i>; yields {&lt;2 x bool&gt;}:result = true, false</i>
2433 &lt;result&gt; = vsetfp ge &lt;2 x float&gt; &lt;float 0.0, float 1.0&gt;, &lt;float 1.0, float 0.0&gt; <i>; yields {&lt;2 x bool&gt;}:result = false, true</i>
2434</pre>
2435</div>
2436
2437<!-- _______________________________________________________________________ -->
2438<div class="doc_subsubsection">
2439 <a name="i_vselect">'<tt>vselect</tt>' Instruction</a>
2440</div>
2441
2442<div class="doc_text">
2443
2444<h5>Syntax:</h5>
2445
2446<pre>
2447 &lt;result&gt; = vselect &lt;n x bool&gt; &lt;cond&gt;, &lt;n x &lt;ty&gt;&gt; &lt;val1&gt;, &lt;n x &lt;ty&gt;&gt; &lt;val2&gt; <i>; yields &lt;n x &lt;ty&gt;&gt;</i>
2448</pre>
2449
2450<h5>Overview:</h5>
2451
2452<p>
2453The '<tt>vselect</tt>' instruction chooses one value at each position
2454of a vector based on a condition.
2455</p>
2456
2457
2458<h5>Arguments:</h5>
2459
2460<p>
2461The '<tt>vselect</tt>' instruction requires a <a
2462href="#t_packed">packed</a> <tt>bool</tt> value indicating the
2463condition at each vector position, and two values of the same packed
2464type. All three operands must have the same length. The type of the
2465result is the same as the type of the two value operands.</p>
2466
2467<h5>Semantics:</h5>
2468
2469<p>
2470At each position where the <tt>bool</tt> vector is true, that position
2471of the result gets its value from the first value argument; otherwise,
2472it gets its value from the second value argument.
2473</p>
2474
2475<h5>Example:</h5>
2476
2477<pre>
2478 %X = vselect bool &lt;2 x bool&gt; &lt;bool true, bool false&gt;, &lt;2 x ubyte&gt; &lt;ubyte 17, ubyte 17&gt;,
2479 &lt;2 x ubyte&gt; &lt;ubyte 42, ubyte 42&gt; <i>; yields &lt;2 x ubyte&gt;:17, 42</i>
2480</pre>
2481</div>
2482
2483
2484
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2485<!-- ======================================================================= -->
2486<div class="doc_subsection">
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]2487 <a name="memoryops">Memory Access and Addressing Operations</a>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2488</div>
2489
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2490<div class="doc_text">
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2491
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2492<p>A key design point of an SSA-based representation is how it
2493represents memory. In LLVM, no memory locations are in SSA form, which
2494makes things very simple. This section describes how to read, write,
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]2495allocate, and free memory in LLVM.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2496
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2497</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2498
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2499<!-- _______________________________________________________________________ -->
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2500<div class="doc_subsubsection">
2501 <a name="i_malloc">'<tt>malloc</tt>' Instruction</a>
2502</div>
2503
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2504<div class="doc_text">
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2505
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2506<h5>Syntax:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2507
2508<pre>
2509 &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]2510</pre>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2511
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2512<h5>Overview:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2513
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2514<p>The '<tt>malloc</tt>' instruction allocates memory from the system
2515heap and returns a pointer to it.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2516
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2517<h5>Arguments:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2518
2519<p>The '<tt>malloc</tt>' instruction allocates
2520<tt>sizeof(&lt;type&gt;)*NumElements</tt>
John Criswell6e4ca612004-02-24 16:13:56 +0000[diff] [blame]2521bytes of memory from the operating system and returns a pointer of the
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2522appropriate type to the program. If "NumElements" is specified, it is the
2523number of elements allocated. If an alignment is specified, the value result
2524of the allocation is guaranteed to be aligned to at least that boundary. If
2525not specified, or if zero, the target can choose to align the allocation on any
2526convenient boundary.</p>
2527
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2528<p>'<tt>type</tt>' must be a sized type.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2529
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2530<h5>Semantics:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2531
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2532<p>Memory is allocated using the system "<tt>malloc</tt>" function, and
2533a pointer is returned.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2534
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2535<h5>Example:</h5>
2536
2537<pre>
2538 %array = malloc [4 x ubyte ] <i>; yields {[%4 x ubyte]*}:array</i>
2539
2540 %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]2541 %array1 = malloc ubyte, uint 4 <i>; yields {ubyte*}:array1</i>
2542 %array2 = malloc [12 x ubyte], uint %size <i>; yields {[12 x ubyte]*}:array2</i>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2543 %array3 = malloc int, uint 4, align 1024 <i>; yields {int*}:array3</i>
2544 %array4 = malloc int, align 1024 <i>; yields {int*}:array4</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2545</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2546</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2547
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2548<!-- _______________________________________________________________________ -->
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2549<div class="doc_subsubsection">
2550 <a name="i_free">'<tt>free</tt>' Instruction</a>
2551</div>
2552
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2553<div class="doc_text">
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2554
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2555<h5>Syntax:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2556
2557<pre>
2558 free &lt;type&gt; &lt;value&gt; <i>; yields {void}</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2559</pre>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2560
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2561<h5>Overview:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2562
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2563<p>The '<tt>free</tt>' instruction returns memory back to the unused
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]2564memory heap to be reallocated in the future.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2565
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2566<h5>Arguments:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2567
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2568<p>'<tt>value</tt>' shall be a pointer value that points to a value
2569that was allocated with the '<tt><a href="#i_malloc">malloc</a></tt>'
2570instruction.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2571
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2572<h5>Semantics:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2573
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]2574<p>Access to the memory pointed to by the pointer is no longer defined
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2575after this instruction executes.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2576
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2577<h5>Example:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2578
2579<pre>
2580 %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]2581 free [4 x ubyte]* %array
2582</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2583</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2584
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2585<!-- _______________________________________________________________________ -->
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2586<div class="doc_subsubsection">
2587 <a name="i_alloca">'<tt>alloca</tt>' Instruction</a>
2588</div>
2589
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2590<div class="doc_text">
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2591
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2592<h5>Syntax:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2593
2594<pre>
2595 &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]2596</pre>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2597
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2598<h5>Overview:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2599
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2600<p>The '<tt>alloca</tt>' instruction allocates memory on the current
2601stack frame of the procedure that is live until the current function
2602returns to its caller.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2603
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2604<h5>Arguments:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2605
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]2606<p>The '<tt>alloca</tt>' instruction allocates <tt>sizeof(&lt;type&gt;)*NumElements</tt>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2607bytes of memory on the runtime stack, returning a pointer of the
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2608appropriate type to the program. If "NumElements" is specified, it is the
2609number of elements allocated. If an alignment is specified, the value result
2610of the allocation is guaranteed to be aligned to at least that boundary. If
2611not specified, or if zero, the target can choose to align the allocation on any
2612convenient boundary.</p>
2613
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2614<p>'<tt>type</tt>' may be any sized type.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2615
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2616<h5>Semantics:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2617
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]2618<p>Memory is allocated; a pointer is returned. '<tt>alloca</tt>'d
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2619memory is automatically released when the function returns. The '<tt>alloca</tt>'
2620instruction is commonly used to represent automatic variables that must
2621have an address available. When the function returns (either with the <tt><a
John Criswelldae2e932005-05-12 16:55:34 +0000[diff] [blame]2622 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]2623instructions), the memory is reclaimed.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2624
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2625<h5>Example:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2626
2627<pre>
2628 %ptr = alloca int <i>; yields {int*}:ptr</i>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2629 %ptr = alloca int, uint 4 <i>; yields {int*}:ptr</i>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2630 %ptr = alloca int, uint 4, align 1024 <i>; yields {int*}:ptr</i>
2631 %ptr = alloca int, align 1024 <i>; yields {int*}:ptr</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2632</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2633</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2634
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2635<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2636<div class="doc_subsubsection"> <a name="i_load">'<tt>load</tt>'
2637Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2638<div class="doc_text">
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2639<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2640<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]2641<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2642<p>The '<tt>load</tt>' instruction is used to read from memory.</p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2643<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2644<p>The argument to the '<tt>load</tt>' instruction specifies the memory
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]2645address from which to load. The pointer must point to a <a
Chris Lattnere53e5082004-06-03 22:57:15 +0000[diff] [blame]2646 href="#t_firstclass">first class</a> type. If the <tt>load</tt> is
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]2647marked as <tt>volatile</tt>, then the optimizer is not allowed to modify
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2648the number or order of execution of this <tt>load</tt> with other
2649volatile <tt>load</tt> and <tt><a href="#i_store">store</a></tt>
2650instructions. </p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2651<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2652<p>The location of memory pointed to is loaded.</p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2653<h5>Examples:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2654<pre> %ptr = <a href="#i_alloca">alloca</a> int <i>; yields {int*}:ptr</i>
2655 <a
2656 href="#i_store">store</a> int 3, int* %ptr <i>; yields {void}</i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2657 %val = load int* %ptr <i>; yields {int}:val = int 3</i>
2658</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2659</div>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2660<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2661<div class="doc_subsubsection"> <a name="i_store">'<tt>store</tt>'
2662Instruction</a> </div>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2663<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2664<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]2665 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]2666</pre>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2667<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2668<p>The '<tt>store</tt>' instruction is used to write to memory.</p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2669<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2670<p>There are two arguments to the '<tt>store</tt>' instruction: a value
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]2671to 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]2672operand must be a pointer to the type of the '<tt>&lt;value&gt;</tt>'
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]2673operand. If the <tt>store</tt> is marked as <tt>volatile</tt>, then the
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2674optimizer is not allowed to modify the number or order of execution of
2675this <tt>store</tt> with other volatile <tt>load</tt> and <tt><a
2676 href="#i_store">store</a></tt> instructions.</p>
2677<h5>Semantics:</h5>
2678<p>The contents of memory are updated to contain '<tt>&lt;value&gt;</tt>'
2679at the location specified by the '<tt>&lt;pointer&gt;</tt>' operand.</p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2680<h5>Example:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2681<pre> %ptr = <a href="#i_alloca">alloca</a> int <i>; yields {int*}:ptr</i>
2682 <a
2683 href="#i_store">store</a> int 3, int* %ptr <i>; yields {void}</i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2684 %val = load int* %ptr <i>; yields {int}:val = int 3</i>
2685</pre>
Reid Spencer47ce1792006-11-09 21:15:49 +0000[diff] [blame^]2686</div>
2687
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2688<!-- _______________________________________________________________________ -->
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2689<div class="doc_subsubsection">
2690 <a name="i_getelementptr">'<tt>getelementptr</tt>' Instruction</a>
2691</div>
2692
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2693<div class="doc_text">
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2694<h5>Syntax:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2695<pre>
2696 &lt;result&gt; = getelementptr &lt;ty&gt;* &lt;ptrval&gt;{, &lt;ty&gt; &lt;idx&gt;}*
2697</pre>
2698
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2699<h5>Overview:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2700
2701<p>
2702The '<tt>getelementptr</tt>' instruction is used to get the address of a
2703subelement of an aggregate data structure.</p>
2704
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2705<h5>Arguments:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2706
2707<p>This instruction takes a list of integer constants that indicate what
2708elements of the aggregate object to index to. The actual types of the arguments
2709provided depend on the type of the first pointer argument. The
2710'<tt>getelementptr</tt>' instruction is used to index down through the type
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]2711levels of a structure or to a specific index in an array. When indexing into a
2712structure, only <tt>uint</tt>
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]2713integer constants are allowed. When indexing into an array or pointer,
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2714<tt>int</tt> and <tt>long</tt> indexes are allowed of any sign.</p>
2715
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2716<p>For example, let's consider a C code fragment and how it gets
2717compiled to LLVM:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2718
2719<pre>
2720 struct RT {
2721 char A;
2722 int B[10][20];
2723 char C;
2724 };
2725 struct ST {
2726 int X;
2727 double Y;
2728 struct RT Z;
2729 };
2730
2731 int *foo(struct ST *s) {
2732 return &amp;s[1].Z.B[5][13];
2733 }
2734</pre>
2735
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2736<p>The LLVM code generated by the GCC frontend is:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2737
2738<pre>
2739 %RT = type { sbyte, [10 x [20 x int]], sbyte }
2740 %ST = type { int, double, %RT }
2741
Brian Gaeke7283e7c2004-07-02 21:08:14 +0000[diff] [blame]2742 implementation
2743
2744 int* %foo(%ST* %s) {
2745 entry:
2746 %reg = getelementptr %ST* %s, int 1, uint 2, uint 1, int 5, int 13
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2747 ret int* %reg
2748 }
2749</pre>
2750
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2751<h5>Semantics:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2752
2753<p>The index types specified for the '<tt>getelementptr</tt>' instruction depend
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]2754on the pointer type that is being indexed into. <a href="#t_pointer">Pointer</a>
Chris Lattnere53e5082004-06-03 22:57:15 +0000[diff] [blame]2755and <a href="#t_array">array</a> types require <tt>uint</tt>, <tt>int</tt>,
2756<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]2757types require <tt>uint</tt> <b>constants</b>.</p>
2758
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2759<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]2760type, which is a pointer, yielding a '<tt>%ST</tt>' = '<tt>{ int, double, %RT
2761}</tt>' type, a structure. The second index indexes into the third element of
2762the structure, yielding a '<tt>%RT</tt>' = '<tt>{ sbyte, [10 x [20 x int]],
2763sbyte }</tt>' type, another structure. The third index indexes into the second
2764element of the structure, yielding a '<tt>[10 x [20 x int]]</tt>' type, an
2765array. The two dimensions of the array are subscripted into, yielding an
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]2766'<tt>int</tt>' type. The '<tt>getelementptr</tt>' instruction returns a pointer
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2767to this element, thus computing a value of '<tt>int*</tt>' type.</p>
2768
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2769<p>Note that it is perfectly legal to index partially through a
2770structure, returning a pointer to an inner element. Because of this,
2771the LLVM code for the given testcase is equivalent to:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2772
2773<pre>
Chris Lattnerd4f6b172005-03-07 22:13:59 +0000[diff] [blame]2774 int* %foo(%ST* %s) {
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2775 %t1 = getelementptr %ST* %s, int 1 <i>; yields %ST*:%t1</i>
2776 %t2 = getelementptr %ST* %t1, int 0, uint 2 <i>; yields %RT*:%t2</i>
2777 %t3 = getelementptr %RT* %t2, int 0, uint 1 <i>; yields [10 x [20 x int]]*:%t3</i>
2778 %t4 = getelementptr [10 x [20 x int]]* %t3, int 0, int 5 <i>; yields [20 x int]*:%t4</i>
2779 %t5 = getelementptr [20 x int]* %t4, int 0, int 13 <i>; yields int*:%t5</i>
2780 ret int* %t5
2781 }
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]2782</pre>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]2783
2784<p>Note that it is undefined to access an array out of bounds: array and
2785pointer indexes must always be within the defined bounds of the array type.
2786The one exception for this rules is zero length arrays. These arrays are
2787defined to be accessible as variable length arrays, which requires access
2788beyond the zero'th element.</p>
2789
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]2790<p>The getelementptr instruction is often confusing. For some more insight
2791into how it works, see <a href="GetElementPtr.html">the getelementptr
2792FAQ</a>.</p>
2793
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2794<h5>Example:</h5>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]2795
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2796<pre>
2797 <i>; yields [12 x ubyte]*:aptr</i>
2798 %aptr = getelementptr {int, [12 x ubyte]}* %sptr, long 0, uint 1
2799</pre>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2800</div>
Reid Spencer47ce1792006-11-09 21:15:49 +0000[diff] [blame^]2801
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2802<!-- ======================================================================= -->
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]2803<div class="doc_subsection"> <a name="convertops">Conversion Operations</a>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2804</div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2805<div class="doc_text">
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]2806<p>The instructions in this category are the conversion instructions (casting)
2807which all take a single operand and a type. They perform various bit conversions
2808on the operand.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2809</div>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]2810
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]2811<!-- _______________________________________________________________________ -->
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]2812<div class="doc_subsubsection">
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2813 <a name="i_trunc">'<tt>trunc .. to</tt>' Instruction</a>
2814</div>
2815<div class="doc_text">
2816
2817<h5>Syntax:</h5>
2818<pre>
2819 &lt;result&gt; = trunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2820</pre>
2821
2822<h5>Overview:</h5>
2823<p>
2824The '<tt>trunc</tt>' instruction truncates its operand to the type <tt>ty2</tt>.
2825</p>
2826
2827<h5>Arguments:</h5>
2828<p>
2829The '<tt>trunc</tt>' instruction takes a <tt>value</tt> to trunc, which must
2830be an <a href="#t_integer">integer</a> type, and a type that specifies the size
2831and type of the result, which must be an <a href="#t_integral">integral</a>
2832type.</p>
2833
2834<h5>Semantics:</h5>
2835<p>
2836The '<tt>trunc</tt>' instruction truncates the high order bits in <tt>value</tt>
2837and converts the reamining bits to <tt>ty2</tt>. The bit size of <tt>value</tt>
2838must be larger than the bit size of <tt>ty2</tt>. Equal sized types are not
2839allowed. This implies that a <tt>trunc</tt> cannot be a <i>no-op cast</i>. It
2840will always truncate bits.</p>
2841
2842<p>When truncating to bool, the truncation is done as a comparison against
2843zero. If the <tt>value</tt> was zero, the bool result will be <tt>false</tt>.
2844If the <tt>value</tt> was non-zero, the bool result will be <tt>true</tt>.</p>
2845
2846<h5>Example:</h5>
2847<pre>
2848 %X = trunc int 257 to ubyte <i>; yields ubyte:1</i>
2849 %Y = trunc int 123 to bool <i>; yields bool:true</i>
2850</pre>
2851</div>
2852
2853<!-- _______________________________________________________________________ -->
2854<div class="doc_subsubsection">
2855 <a name="i_zext">'<tt>zext .. to</tt>' Instruction</a>
2856</div>
2857<div class="doc_text">
2858
2859<h5>Syntax:</h5>
2860<pre>
2861 &lt;result&gt; = zext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2862</pre>
2863
2864<h5>Overview:</h5>
2865<p>The '<tt>zext</tt>' instruction zero extends its operand to type
2866<tt>ty2</tt>.</p>
2867
2868
2869<h5>Arguments:</h5>
2870<p>The '<tt>zext</tt>' instruction takes a value to cast, which must be of
2871<a href="#t_integral">integral</a> type, and a type to cast it to, which must
2872also be of <a href="#t_integral">integral</a> type. The bit size of the
2873<tt>value</tt> must be smaller than or equal to the bit size of the
2874destination type, <tt>ty2</tt>.</p>
2875
2876<h5>Semantics:</h5>
2877<p>The <tt>zext</tt> fills the high order bits of the <tt>value</tt> with zero
2878bits until it reaches the size of the destination type, <tt>ty2</tt>. When the
2879the operand and the type are the same size, no bit filling is done and the
2880cast is considered a <i>no-op cast</i> because no bits change (only the type
2881changes).</p>
2882
2883<p>When zero extending to bool, the extension 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>
2886
2887<h5>Example:</h5>
2888<pre>
2889 %X = zext int 257 to ulong <i>; yields ulong:257</i>
2890 %Y = zext bool true to int <i>; yields int:1</i>
2891</pre>
2892</div>
2893
2894<!-- _______________________________________________________________________ -->
2895<div class="doc_subsubsection">
2896 <a name="i_sext">'<tt>sext .. to</tt>' Instruction</a>
2897</div>
2898<div class="doc_text">
2899
2900<h5>Syntax:</h5>
2901<pre>
2902 &lt;result&gt; = sext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2903</pre>
2904
2905<h5>Overview:</h5>
2906<p>The '<tt>sext</tt>' sign extends <tt>value</tt> to the type <tt>ty2</tt>.</p>
2907
2908<h5>Arguments:</h5>
2909<p>
2910The '<tt>sext</tt>' instruction takes a value to cast, which must be of
2911<a href="#t_integral">integral</a> type, and a type to cast it to, which must
2912also be of <a href="#t_integral">integral</a> type.</p>
2913
2914<h5>Semantics:</h5>
2915<p>
2916The '<tt>sext</tt>' instruction performs a sign extension by copying the sign
2917bit (highest order bit) of the <tt>value</tt> until it reaches the bit size of
2918the type <tt>ty2</tt>. When the the operand and the type are the same size,
2919no bit filling is done and the cast is considered a <i>no-op cast</i> because
2920no bits change (only the type changes).</p>
2921
2922<p>When sign extending to bool, the extension is done as a comparison against
2923zero. If the <tt>value</tt> was zero, the bool result will be <tt>false</tt>.
2924If the <tt>value</tt> was non-zero, the bool result will be <tt>true</tt>.</p>
2925
2926<h5>Example:</h5>
2927
2928<pre>
2929 %X = sext sbyte -1 to ushort <i>; yields ushort:65535</i>
2930 %Y = sext bool true to int <i>; yields int:-1</i>
2931</pre>
2932</div>
2933
2934<!-- _______________________________________________________________________ -->
2935<div class="doc_subsubsection">
2936 <a name="i_fpext">'<tt>fpext .. to</tt>' Instruction</a>
2937</div>
2938<div class="doc_text">
2939
2940<h5>Syntax:</h5>
2941<pre>
2942 &lt;result&gt; = fpext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2943</pre>
2944
2945<h5>Overview:</h5>
2946<p>The '<tt>fpext</tt>' extends a floating point <tt>value</tt> to a larger
2947floating point value.</p>
2948
2949<h5>Arguments:</h5>
2950<p>The '<tt>fpext</tt>' instruction takes a
2951<a href="#t_floating">floating point</a> <tt>value</tt> to cast,
2952and a <a href="#t_floating">floating point</a> type to cast it to.</p>
2953
2954<h5>Semantics:</h5>
2955<p>The '<tt>fpext</tt>' instruction extends the <tt>value</tt> from one floating
2956point type to another. If the type of the <tt>value</tt> and <tt>ty2</tt> are
2957the same, the instruction is considered a <i>no-op cast</i> because no bits
2958change.</p>
2959
2960<h5>Example:</h5>
2961<pre>
2962 %X = fpext float 3.1415 to double <i>; yields double:3.1415</i>
2963 %Y = fpext float 1.0 to float <i>; yields float:1.0 (no-op)</i>
2964</pre>
2965</div>
2966
2967<!-- _______________________________________________________________________ -->
2968<div class="doc_subsubsection">
2969 <a name="i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]2970</div>
2971
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2972<div class="doc_text">
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]2973
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]2974<h5>Syntax:</h5>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]2975
2976<pre>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2977 &lt;result&gt; = fptrunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]2978</pre>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]2979
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]2980<h5>Overview:</h5>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2981<p>The '<tt>fptrunc</tt>' instruction truncates <tt>value</tt> to type
2982<tt>ty2</tt>.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]2983
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2984
2985<h5>Arguments:</h5>
2986<p>The '<tt>fptrunc</tt>' instruction takes a <a href="#t_floating">floating
2987 point</a> value to cast and a <a href="#t_floating">floating point</a> type to
2988cast it to. The size of <tt>value</tt> must be larger than the size of
2989<tt>ty2</a>. This implies that <tt>fptrunc</tt> cannot be used to make a
2990<i>no-op cast</i>.</p>
2991
2992<h5>Semantics:</h5>
2993<p> The '<tt>fptrunc</tt>' instruction converts a
2994<a href="#t_floating">floating point</a> value from a larger type to a smaller
2995type. If the value cannot fit within the destination type, <tt>ty2</tt>, then
2996the results are undefined.</p>
2997
2998<h5>Example:</h5>
2999<pre>
3000 %X = fptrunc double 123.0 to float <i>; yields float:123.0</i>
3001 %Y = fptrunc double 1.0E+300 to float <i>; yields undefined</i>
3002</pre>
3003</div>
3004
3005<!-- _______________________________________________________________________ -->
3006<div class="doc_subsubsection">
3007 <a name="i_fp2uint">'<tt>fp2uint .. to</tt>' Instruction</a>
3008</div>
3009<div class="doc_text">
3010
3011<h5>Syntax:</h5>
3012<pre>
3013 &lt;result&gt; = fp2uint &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3014</pre>
3015
3016<h5>Overview:</h5>
3017<p>The '<tt>fp2uint</tt>' converts a floating point <tt>value</tt> to its
3018unsigned integer equivalent of type <tt>ty2</tt>.
3019</p>
3020
3021<h5>Arguments:</h5>
3022<p>The '<tt>fp2uint</tt>' instruction takes a value to cast, which must be a
3023<a href="#t_floating">floating point</a> value, and a type to cast it to, which
3024must be an <a href="#t_integral">integral</a> type.</p>
3025
3026<h5>Semantics:</h5>
3027<p> The '<tt>fp2uint</tt>' instruction converts its
3028<a href="#t_floating">floating point</a> operand into the nearest (rounding
3029towards zero) unsigned integer value. If the value cannot fit in <tt>ty2</tt>,
3030the results are undefined.</p>
3031
3032<p>When converting to bool, the conversion is done as a comparison against
3033zero. If the <tt>value</tt> was zero, the bool result will be <tt>false</tt>.
3034If the <tt>value</tt> was non-zero, the bool result will be <tt>true</tt>.</p>
3035
3036<h5>Example:</h5>
3037<pre>
3038 %X = fp2uint double 123.0 to int <i>; yields int:123</i>
3039 %Y = fp2uint float 1.0E+300 to bool <i>; yields bool:true</i>
3040 %X = fp2uint float 1.04E+17 to ubyte <i>; yields undefined:1</i>
3041</pre>
3042</div>
3043
3044<!-- _______________________________________________________________________ -->
3045<div class="doc_subsubsection">
3046 <a name="i_fp2sint">'<tt>fp2sint .. to</tt>' Instruction</a>
3047</div>
3048<div class="doc_text">
3049
3050<h5>Syntax:</h5>
3051<pre>
3052 &lt;result&gt; = fp2sint &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3053</pre>
3054
3055<h5>Overview:</h5>
3056<p>The '<tt>fp2sint</tt>' instruction converts
3057<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]3058</p>
3059
3060
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]3061<h5>Arguments:</h5>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3062<p> The '<tt>fp2sint</tt>' instruction takes a value to cast, which must be a
3063<a href="#t_floating">floating point</a> value, and a type to cast it to, which
3064must also be an <a href="#t_integral">integral</a> type.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3065
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]3066<h5>Semantics:</h5>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3067<p>The '<tt>fp2sint</tt>' instruction converts its
3068<a href="#t_floating">floating point</a> operand into the nearest (rounding
3069towards zero) signed integer value. If the value cannot fit in <tt>ty2</tt>,
3070the results are undefined.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3071
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3072<p>When converting to bool, the conversion is done as a comparison against
3073zero. If the <tt>value</tt> was zero, the bool result will be <tt>false</tt>.
3074If 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]3075
Chris Lattner33ba0d92001-07-09 00:26:23 +0000[diff] [blame]3076<h5>Example:</h5>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3077<pre>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3078 %X = fp2sint double -123.0 to int <i>; yields int:-123</i>
3079 %Y = fp2sint float 1.0E-247 to bool <i>; yields bool:true</i>
3080 %X = fp2sint float 1.04E+17 to sbyte <i>; yields undefined:1</i>
3081</pre>
3082</div>
3083
3084<!-- _______________________________________________________________________ -->
3085<div class="doc_subsubsection">
3086 <a name="i_uint2fp">'<tt>uint2fp .. to</tt>' Instruction</a>
3087</div>
3088<div class="doc_text">
3089
3090<h5>Syntax:</h5>
3091<pre>
3092 &lt;result&gt; = uint2fp &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3093</pre>
3094
3095<h5>Overview:</h5>
3096<p>The '<tt>uint2fp</tt>' instruction regards <tt>value</tt> as an unsigned
3097integer and converts that value to the <tt>ty2</tt> type.</p>
3098
3099
3100<h5>Arguments:</h5>
3101<p>The '<tt>uint2fp</tt>' instruction takes a value to cast, which must be an
3102<a href="#t_integral">integral</a> value, and a type to cast it to, which must
3103be a <a href="#t_floating">floating point</a> type.</p>
3104
3105<h5>Semantics:</h5>
3106<p>The '<tt>uint2fp</tt>' instruction interprets its operand as an unsigned
3107integer quantity and converts it to the corresponding floating point value. If
3108the value cannot fit in the floating point value, the results are undefined.</p>
3109
3110
3111<h5>Example:</h5>
3112<pre>
3113 %X = uint2fp int 257 to float <i>; yields float:257.0</i>
3114 %Y = uint2fp sbyte -1 to double <i>; yields double:255.0</i>
3115</pre>
3116</div>
3117
3118<!-- _______________________________________________________________________ -->
3119<div class="doc_subsubsection">
3120 <a name="i_sint2fp">'<tt>sint2fp .. to</tt>' Instruction</a>
3121</div>
3122<div class="doc_text">
3123
3124<h5>Syntax:</h5>
3125<pre>
3126 &lt;result&gt; = sint2fp &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3127</pre>
3128
3129<h5>Overview:</h5>
3130<p>The '<tt>sint2fp</tt>' instruction regards <tt>value</tt> as a signed
3131integer and converts that value to the <tt>ty2</tt> type.</p>
3132
3133<h5>Arguments:</h5>
3134<p>The '<tt>sint2fp</tt>' instruction takes a value to cast, which must be an
3135<a href="#t_integral">integral</a> value, and a type to cast it to, which must be
3136a <a href="#t_floating">floating point</a> type.</p>
3137
3138<h5>Semantics:</h5>
3139<p>The '<tt>sint2fp</tt>' instruction interprets its operand as a signed
3140integer quantity and converts it to the corresponding floating point value. If
3141the value cannot fit in the floating point value, the results are undefined.</p>
3142
3143<h5>Example:</h5>
3144<pre>
3145 %X = sint2fp int 257 to float <i>; yields float:257.0</i>
3146 %Y = sint2fp sbyte -1 to double <i>; yields double:-1.0</i>
3147</pre>
3148</div>
3149
3150<!-- _______________________________________________________________________ -->
3151<div class="doc_subsubsection">
3152 <a name="i_bitconvert">'<tt>bitconvert .. to</tt>' Instruction</a>
3153</div>
3154<div class="doc_text">
3155
3156<h5>Syntax:</h5>
3157<pre>
3158 &lt;result&gt; = bitconvert &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3159</pre>
3160
3161<h5>Overview:</h5>
3162<p>The '<tt>bitconvert</tt>' instruction converts <tt>value</tt> to type
3163<tt>ty2</tt> without changing any bits.</p>
3164
3165<h5>Arguments:</h5>
3166<p>The '<tt>bitconvert</tt>' instruction takes a value to cast, which must be
3167a first class value, and a type to cast it to, which must also be a <a
3168 href="#t_firstclass">first class</a> type. The bit sizes of <tt>value</tt>
3169and the destination type, <tt>ty2</tt>, must be identical.</p>
3170
3171<h5>Semantics:</h5>
3172<p>The '<tt>bitconvert</tt>' instruction converts <tt>value</tt> to type
3173<tt>ty2</tt> as if the value had been stored to memory and read back as type
3174<tt>ty2</tt>. That is, no bits are changed during the conversion. The
3175<tt>bitconvert</tt> instruction may be used to construct <i>no-op casts</i> that
3176the <tt>zext, sext, and fpext</tt> instructions do not permit.</p>
3177
3178<h5>Example:</h5>
3179<pre>
3180 %X = bitconvert ubyte 255 to sbyte <i>; yields sbyte:-1</i>
3181 %Y = bitconvert uint* %x to uint <i>; yields uint:%x</i>
Chris Lattner33ba0d92001-07-09 00:26:23 +0000[diff] [blame]3182</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3183</div>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3184
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]3185<!-- ======================================================================= -->
3186<div class="doc_subsection"> <a name="otherops">Other Operations</a> </div>
3187<div class="doc_text">
3188<p>The instructions in this category are the "miscellaneous"
3189instructions, which defy better classification.</p>
3190</div>
3191<!-- _______________________________________________________________________ -->
3192<div class="doc_subsubsection"> <a name="i_phi">'<tt>phi</tt>'
3193Instruction</a> </div>
3194<div class="doc_text">
3195<h5>Syntax:</h5>
3196<pre> &lt;result&gt; = phi &lt;ty&gt; [ &lt;val0&gt;, &lt;label0&gt;], ...<br></pre>
3197<h5>Overview:</h5>
3198<p>The '<tt>phi</tt>' instruction is used to implement the &#966; node in
3199the SSA graph representing the function.</p>
3200<h5>Arguments:</h5>
3201<p>The type of the incoming values are specified with the first type
3202field. After this, the '<tt>phi</tt>' instruction takes a list of pairs
3203as arguments, with one pair for each predecessor basic block of the
3204current block. Only values of <a href="#t_firstclass">first class</a>
3205type may be used as the value arguments to the PHI node. Only labels
3206may be used as the label arguments.</p>
3207<p>There must be no non-phi instructions between the start of a basic
3208block and the PHI instructions: i.e. PHI instructions must be first in
3209a basic block.</p>
3210<h5>Semantics:</h5>
3211<p>At runtime, the '<tt>phi</tt>' instruction logically takes on the
3212value specified by the parameter, depending on which basic block we
3213came from in the last <a href="#terminators">terminator</a> instruction.</p>
3214<h5>Example:</h5>
3215<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>
3216</div>
3217
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3218<!-- _______________________________________________________________________ -->
3219<div class="doc_subsubsection">
3220 <a name="i_select">'<tt>select</tt>' Instruction</a>
3221</div>
3222
3223<div class="doc_text">
3224
3225<h5>Syntax:</h5>
3226
3227<pre>
3228 &lt;result&gt; = select bool &lt;cond&gt;, &lt;ty&gt; &lt;val1&gt;, &lt;ty&gt; &lt;val2&gt; <i>; yields ty</i>
3229</pre>
3230
3231<h5>Overview:</h5>
3232
3233<p>
3234The '<tt>select</tt>' instruction is used to choose one value based on a
3235condition, without branching.
3236</p>
3237
3238
3239<h5>Arguments:</h5>
3240
3241<p>
3242The '<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.
3243</p>
3244
3245<h5>Semantics:</h5>
3246
3247<p>
3248If the boolean condition evaluates to true, the instruction returns the first
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3249value argument; otherwise, it returns the second value argument.
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3250</p>
3251
3252<h5>Example:</h5>
3253
3254<pre>
3255 %X = select bool true, ubyte 17, ubyte 42 <i>; yields ubyte:17</i>
3256</pre>
3257</div>
3258
Robert Bocchino05ccd702006-01-15 20:48:27 +0000[diff] [blame]3259
3260<!-- _______________________________________________________________________ -->
3261<div class="doc_subsubsection">
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3262 <a name="i_call">'<tt>call</tt>' Instruction</a>
3263</div>
3264
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3265<div class="doc_text">
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3266
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3267<h5>Syntax:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3268<pre>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3269 &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]3270</pre>
3271
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3272<h5>Overview:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3273
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3274<p>The '<tt>call</tt>' instruction represents a simple function call.</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3275
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3276<h5>Arguments:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3277
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3278<p>This instruction requires several arguments:</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3279
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]3280<ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3281 <li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3282 <p>The optional "tail" marker indicates whether the callee function accesses
3283 any allocas or varargs in the caller. If the "tail" marker is present, the
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3284 function call is eligible for tail call optimization. Note that calls may
3285 be marked "tail" even if they do not occur before a <a
3286 href="#i_ret"><tt>ret</tt></a> instruction.
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3287 </li>
3288 <li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3289 <p>The optional "cconv" marker indicates which <a href="callingconv">calling
3290 convention</a> the call should use. If none is specified, the call defaults
3291 to using C calling conventions.
3292 </li>
3293 <li>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3294 <p>'<tt>ty</tt>': shall be the signature of the pointer to function value
3295 being invoked. The argument types must match the types implied by this
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3296 signature. This type can be omitted if the function is not varargs and
3297 if the function type does not return a pointer to a function.</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3298 </li>
3299 <li>
3300 <p>'<tt>fnptrval</tt>': An LLVM value containing a pointer to a function to
3301 be invoked. In most cases, this is a direct function invocation, but
3302 indirect <tt>call</tt>s are just as possible, calling an arbitrary pointer
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3303 to function value.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3304 </li>
3305 <li>
3306 <p>'<tt>function args</tt>': argument list whose types match the
Reid Spencera7e302a2005-05-01 22:22:57 +0000[diff] [blame]3307 function signature argument types. All arguments must be of
3308 <a href="#t_firstclass">first class</a> type. If the function signature
3309 indicates the function accepts a variable number of arguments, the extra
3310 arguments can be specified.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3311 </li>
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]3312</ol>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3313
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3314<h5>Semantics:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3315
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3316<p>The '<tt>call</tt>' instruction is used to cause control flow to
3317transfer to a specified function, with its incoming arguments bound to
3318the specified values. Upon a '<tt><a href="#i_ret">ret</a></tt>'
3319instruction in the called function, control flow continues with the
3320instruction after the function call, and the return value of the
3321function is bound to the result argument. This is a simpler case of
3322the <a href="#i_invoke">invoke</a> instruction.</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3323
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3324<h5>Example:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3325
3326<pre>
3327 %retval = call int %test(int %argc)
3328 call int(sbyte*, ...) *%printf(sbyte* %msg, int 12, sbyte 42);
3329 %X = tail call int %foo()
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3330 %Y = tail call <a href="#callingconv">fastcc</a> int %foo()
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3331</pre>
3332
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3333</div>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3334
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3335<!-- _______________________________________________________________________ -->
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3336<div class="doc_subsubsection">
Chris Lattnerfb6977d2006-01-13 23:26:01 +0000[diff] [blame]3337 <a name="i_va_arg">'<tt>va_arg</tt>' Instruction</a>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3338</div>
3339
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3340<div class="doc_text">
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3341
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]3342<h5>Syntax:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3343
3344<pre>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3345 &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]3346</pre>
3347
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]3348<h5>Overview:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3349
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3350<p>The '<tt>va_arg</tt>' instruction is used to access arguments passed through
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3351the "variable argument" area of a function call. It is used to implement the
3352<tt>va_arg</tt> macro in C.</p>
3353
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]3354<h5>Arguments:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3355
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3356<p>This instruction takes a <tt>va_list*</tt> value and the type of
3357the argument. It returns a value of the specified argument type and
Jeff Cohen25d4f7e2005-11-11 02:15:27 +0000[diff] [blame]3358increments the <tt>va_list</tt> to point to the next argument. Again, the
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3359actual type of <tt>va_list</tt> is target specific.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3360
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]3361<h5>Semantics:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3362
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3363<p>The '<tt>va_arg</tt>' instruction loads an argument of the specified
3364type from the specified <tt>va_list</tt> and causes the
3365<tt>va_list</tt> to point to the next argument. For more information,
3366see the variable argument handling <a href="#int_varargs">Intrinsic
3367Functions</a>.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3368
3369<p>It is legal for this instruction to be called in a function which does not
3370take a variable number of arguments, for example, the <tt>vfprintf</tt>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3371function.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3372
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3373<p><tt>va_arg</tt> is an LLVM instruction instead of an <a
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3374href="#intrinsics">intrinsic function</a> because it takes a type as an
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3375argument.</p>
3376
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]3377<h5>Example:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3378
3379<p>See the <a href="#int_varargs">variable argument processing</a> section.</p>
3380
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3381</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3382
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3383<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3384<div class="doc_section"> <a name="intrinsics">Intrinsic Functions</a> </div>
3385<!-- *********************************************************************** -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3386
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3387<div class="doc_text">
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3388
3389<p>LLVM supports the notion of an "intrinsic function". These functions have
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3390well known names and semantics and are required to follow certain
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3391restrictions. Overall, these instructions represent an extension mechanism for
3392the LLVM language that does not require changing all of the transformations in
3393LLVM to add to the language (or the bytecode reader/writer, the parser,
3394etc...).</p>
3395
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3396<p>Intrinsic function names must all start with an "<tt>llvm.</tt>" prefix. This
3397prefix is reserved in LLVM for intrinsic names; thus, functions may not be named
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3398this. Intrinsic functions must always be external functions: you cannot define
3399the body of intrinsic functions. Intrinsic functions may only be used in call
3400or invoke instructions: it is illegal to take the address of an intrinsic
3401function. Additionally, because intrinsic functions are part of the LLVM
3402language, it is required that they all be documented here if any are added.</p>
3403
3404
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3405<p>To learn how to add an intrinsic function, please see the <a
Chris Lattner590cff32005-05-11 03:35:57 +0000[diff] [blame]3406href="ExtendingLLVM.html">Extending LLVM Guide</a>.
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3407</p>
3408
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3409</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3410
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3411<!-- ======================================================================= -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3412<div class="doc_subsection">
3413 <a name="int_varargs">Variable Argument Handling Intrinsics</a>
3414</div>
3415
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3416<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3417
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3418<p>Variable argument support is defined in LLVM with the <a
Chris Lattnerfb6977d2006-01-13 23:26:01 +0000[diff] [blame]3419 href="#i_va_arg"><tt>va_arg</tt></a> instruction and these three
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3420intrinsic functions. These functions are related to the similarly
3421named macros defined in the <tt>&lt;stdarg.h&gt;</tt> header file.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3422
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3423<p>All of these functions operate on arguments that use a
3424target-specific value type "<tt>va_list</tt>". The LLVM assembly
3425language reference manual does not define what this type is, so all
3426transformations should be prepared to handle intrinsics with any type
3427used.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3428
Chris Lattner374ab302006-05-15 17:26:46 +0000[diff] [blame]3429<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]3430instruction and the variable argument handling intrinsic functions are
3431used.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3432
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3433<pre>
3434int %test(int %X, ...) {
3435 ; Initialize variable argument processing
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3436 %ap = alloca sbyte*
3437 call void %<a href="#i_va_start">llvm.va_start</a>(sbyte** %ap)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3438
3439 ; Read a single integer argument
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3440 %tmp = va_arg sbyte** %ap, int
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3441
3442 ; Demonstrate usage of llvm.va_copy and llvm.va_end
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3443 %aq = alloca sbyte*
Andrew Lenharthd0a4c622005-06-22 20:38:11 +0000[diff] [blame]3444 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]3445 call void %<a href="#i_va_end">llvm.va_end</a>(sbyte** %aq)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3446
3447 ; Stop processing of arguments.
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3448 call void %<a href="#i_va_end">llvm.va_end</a>(sbyte** %ap)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3449 ret int %tmp
3450}
3451</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3452</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3453
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3454<!-- _______________________________________________________________________ -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3455<div class="doc_subsubsection">
3456 <a name="i_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a>
3457</div>
3458
3459
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3460<div class="doc_text">
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3461<h5>Syntax:</h5>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3462<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]3463<h5>Overview:</h5>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3464<P>The '<tt>llvm.va_start</tt>' intrinsic initializes
3465<tt>*&lt;arglist&gt;</tt> for subsequent use by <tt><a
3466href="#i_va_arg">va_arg</a></tt>.</p>
3467
3468<h5>Arguments:</h5>
3469
3470<P>The argument is a pointer to a <tt>va_list</tt> element to initialize.</p>
3471
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3472<h5>Semantics:</h5>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3473
3474<P>The '<tt>llvm.va_start</tt>' intrinsic works just like the <tt>va_start</tt>
3475macro available in C. In a target-dependent way, it initializes the
3476<tt>va_list</tt> element the argument points to, so that the next call to
3477<tt>va_arg</tt> will produce the first variable argument passed to the function.
3478Unlike the C <tt>va_start</tt> macro, this intrinsic does not need to know the
3479last argument of the function, the compiler can figure that out.</p>
3480
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3481</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3482
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3483<!-- _______________________________________________________________________ -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3484<div class="doc_subsubsection">
3485 <a name="i_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a>
3486</div>
3487
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3488<div class="doc_text">
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3489<h5>Syntax:</h5>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3490<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]3491<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3492<p>The '<tt>llvm.va_end</tt>' intrinsic destroys <tt>&lt;arglist&gt;</tt>
3493which has been initialized previously with <tt><a href="#i_va_start">llvm.va_start</a></tt>
3494or <tt><a href="#i_va_copy">llvm.va_copy</a></tt>.</p>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3495<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3496<p>The argument is a <tt>va_list</tt> to destroy.</p>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3497<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3498<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]3499macro available in C. In a target-dependent way, it destroys the <tt>va_list</tt>.
3500Calls to <a href="#i_va_start"><tt>llvm.va_start</tt></a> and <a
3501 href="#i_va_copy"><tt>llvm.va_copy</tt></a> must be matched exactly
3502with calls to <tt>llvm.va_end</tt>.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3503</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3504
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3505<!-- _______________________________________________________________________ -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3506<div class="doc_subsubsection">
3507 <a name="i_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a>
3508</div>
3509
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3510<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3511
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3512<h5>Syntax:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3513
3514<pre>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3515 declare void %llvm.va_copy(&lt;va_list&gt;* &lt;destarglist&gt;,
Andrew Lenharthd0a4c622005-06-22 20:38:11 +0000[diff] [blame]3516 &lt;va_list&gt;* &lt;srcarglist&gt;)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3517</pre>
3518
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3519<h5>Overview:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3520
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3521<p>The '<tt>llvm.va_copy</tt>' intrinsic copies the current argument position from
3522the source argument list to the destination argument list.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3523
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3524<h5>Arguments:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3525
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3526<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]3527The 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]3528
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3529
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3530<h5>Semantics:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3531
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3532<p>The '<tt>llvm.va_copy</tt>' intrinsic works just like the <tt>va_copy</tt> macro
3533available in C. In a target-dependent way, it copies the source
3534<tt>va_list</tt> element into the destination list. This intrinsic is necessary
3535because 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]3536arbitrarily complex and require memory allocation, for example.</p>
3537
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3538</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3539
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3540<!-- ======================================================================= -->
3541<div class="doc_subsection">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3542 <a name="int_gc">Accurate Garbage Collection Intrinsics</a>
3543</div>
3544
3545<div class="doc_text">
3546
3547<p>
3548LLVM support for <a href="GarbageCollection.html">Accurate Garbage
3549Collection</a> requires the implementation and generation of these intrinsics.
3550These intrinsics allow identification of <a href="#i_gcroot">GC roots on the
3551stack</a>, as well as garbage collector implementations that require <a
3552href="#i_gcread">read</a> and <a href="#i_gcwrite">write</a> barriers.
3553Front-ends for type-safe garbage collected languages should generate these
3554intrinsics to make use of the LLVM garbage collectors. For more details, see <a
3555href="GarbageCollection.html">Accurate Garbage Collection with LLVM</a>.
3556</p>
3557</div>
3558
3559<!-- _______________________________________________________________________ -->
3560<div class="doc_subsubsection">
3561 <a name="i_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a>
3562</div>
3563
3564<div class="doc_text">
3565
3566<h5>Syntax:</h5>
3567
3568<pre>
Reid Spencera8d451e2005-04-26 20:50:44 +0000[diff] [blame]3569 declare void %llvm.gcroot(&lt;ty&gt;** %ptrloc, &lt;ty2&gt;* %metadata)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3570</pre>
3571
3572<h5>Overview:</h5>
3573
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]3574<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]3575the code generator, and allows some metadata to be associated with it.</p>
3576
3577<h5>Arguments:</h5>
3578
3579<p>The first argument specifies the address of a stack object that contains the
3580root pointer. The second pointer (which must be either a constant or a global
3581value address) contains the meta-data to be associated with the root.</p>
3582
3583<h5>Semantics:</h5>
3584
3585<p>At runtime, a call to this intrinsics stores a null pointer into the "ptrloc"
3586location. At compile-time, the code generator generates information to allow
3587the runtime to find the pointer at GC safe points.
3588</p>
3589
3590</div>
3591
3592
3593<!-- _______________________________________________________________________ -->
3594<div class="doc_subsubsection">
3595 <a name="i_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a>
3596</div>
3597
3598<div class="doc_text">
3599
3600<h5>Syntax:</h5>
3601
3602<pre>
Chris Lattner80626e92006-03-14 20:02:51 +0000[diff] [blame]3603 declare sbyte* %llvm.gcread(sbyte* %ObjPtr, sbyte** %Ptr)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3604</pre>
3605
3606<h5>Overview:</h5>
3607
3608<p>The '<tt>llvm.gcread</tt>' intrinsic identifies reads of references from heap
3609locations, allowing garbage collector implementations that require read
3610barriers.</p>
3611
3612<h5>Arguments:</h5>
3613
Chris Lattner80626e92006-03-14 20:02:51 +0000[diff] [blame]3614<p>The second argument is the address to read from, which should be an address
3615allocated from the garbage collector. The first object is a pointer to the
3616start of the referenced object, if needed by the language runtime (otherwise
3617null).</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3618
3619<h5>Semantics:</h5>
3620
3621<p>The '<tt>llvm.gcread</tt>' intrinsic has the same semantics as a load
3622instruction, but may be replaced with substantially more complex code by the
3623garbage collector runtime, as needed.</p>
3624
3625</div>
3626
3627
3628<!-- _______________________________________________________________________ -->
3629<div class="doc_subsubsection">
3630 <a name="i_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a>
3631</div>
3632
3633<div class="doc_text">
3634
3635<h5>Syntax:</h5>
3636
3637<pre>
Chris Lattner80626e92006-03-14 20:02:51 +0000[diff] [blame]3638 declare void %llvm.gcwrite(sbyte* %P1, sbyte* %Obj, sbyte** %P2)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3639</pre>
3640
3641<h5>Overview:</h5>
3642
3643<p>The '<tt>llvm.gcwrite</tt>' intrinsic identifies writes of references to heap
3644locations, allowing garbage collector implementations that require write
3645barriers (such as generational or reference counting collectors).</p>
3646
3647<h5>Arguments:</h5>
3648
Chris Lattner80626e92006-03-14 20:02:51 +0000[diff] [blame]3649<p>The first argument is the reference to store, the second is the start of the
3650object to store it to, and the third is the address of the field of Obj to
3651store to. If the runtime does not require a pointer to the object, Obj may be
3652null.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3653
3654<h5>Semantics:</h5>
3655
3656<p>The '<tt>llvm.gcwrite</tt>' intrinsic has the same semantics as a store
3657instruction, but may be replaced with substantially more complex code by the
3658garbage collector runtime, as needed.</p>
3659
3660</div>
3661
3662
3663
3664<!-- ======================================================================= -->
3665<div class="doc_subsection">
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3666 <a name="int_codegen">Code Generator Intrinsics</a>
3667</div>
3668
3669<div class="doc_text">
3670<p>
3671These intrinsics are provided by LLVM to expose special features that may only
3672be implemented with code generator support.
3673</p>
3674
3675</div>
3676
3677<!-- _______________________________________________________________________ -->
3678<div class="doc_subsubsection">
3679 <a name="i_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a>
3680</div>
3681
3682<div class="doc_text">
3683
3684<h5>Syntax:</h5>
3685<pre>
Chris Lattnerfcf39d42006-01-13 01:20:27 +0000[diff] [blame]3686 declare sbyte *%llvm.returnaddress(uint &lt;level&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3687</pre>
3688
3689<h5>Overview:</h5>
3690
3691<p>
Chris Lattner32b5d712006-10-15 20:05:59 +0000[diff] [blame]3692The '<tt>llvm.returnaddress</tt>' intrinsic attempts to compute a
3693target-specific value indicating the return address of the current function
3694or one of its callers.
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3695</p>
3696
3697<h5>Arguments:</h5>
3698
3699<p>
3700The argument to this intrinsic indicates which function to return the address
3701for. Zero indicates the calling function, one indicates its caller, etc. The
3702argument is <b>required</b> to be a constant integer value.
3703</p>
3704
3705<h5>Semantics:</h5>
3706
3707<p>
3708The '<tt>llvm.returnaddress</tt>' intrinsic either returns a pointer indicating
3709the return address of the specified call frame, or zero if it cannot be
3710identified. The value returned by this intrinsic is likely to be incorrect or 0
3711for arguments other than zero, so it should only be used for debugging purposes.
3712</p>
3713
3714<p>
3715Note that calling this intrinsic does not prevent function inlining or other
Chris Lattnerb40bb382005-03-07 20:30:51 +0000[diff] [blame]3716aggressive transformations, so the value returned may not be that of the obvious
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3717source-language caller.
3718</p>
3719</div>
3720
3721
3722<!-- _______________________________________________________________________ -->
3723<div class="doc_subsubsection">
3724 <a name="i_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a>
3725</div>
3726
3727<div class="doc_text">
3728
3729<h5>Syntax:</h5>
3730<pre>
Chris Lattnerfcf39d42006-01-13 01:20:27 +0000[diff] [blame]3731 declare sbyte *%llvm.frameaddress(uint &lt;level&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3732</pre>
3733
3734<h5>Overview:</h5>
3735
3736<p>
Chris Lattner32b5d712006-10-15 20:05:59 +0000[diff] [blame]3737The '<tt>llvm.frameaddress</tt>' intrinsic attempts to return the
3738target-specific frame pointer value for the specified stack frame.
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3739</p>
3740
3741<h5>Arguments:</h5>
3742
3743<p>
3744The argument to this intrinsic indicates which function to return the frame
3745pointer for. Zero indicates the calling function, one indicates its caller,
3746etc. The argument is <b>required</b> to be a constant integer value.
3747</p>
3748
3749<h5>Semantics:</h5>
3750
3751<p>
3752The '<tt>llvm.frameaddress</tt>' intrinsic either returns a pointer indicating
3753the frame address of the specified call frame, or zero if it cannot be
3754identified. The value returned by this intrinsic is likely to be incorrect or 0
3755for arguments other than zero, so it should only be used for debugging purposes.
3756</p>
3757
3758<p>
3759Note that calling this intrinsic does not prevent function inlining or other
Chris Lattnerb40bb382005-03-07 20:30:51 +0000[diff] [blame]3760aggressive transformations, so the value returned may not be that of the obvious
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3761source-language caller.
3762</p>
3763</div>
3764
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]3765<!-- _______________________________________________________________________ -->
3766<div class="doc_subsubsection">
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]3767 <a name="i_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a>
3768</div>
3769
3770<div class="doc_text">
3771
3772<h5>Syntax:</h5>
3773<pre>
3774 declare sbyte *%llvm.stacksave()
3775</pre>
3776
3777<h5>Overview:</h5>
3778
3779<p>
3780The '<tt>llvm.stacksave</tt>' intrinsic is used to remember the current state of
3781the function stack, for use with <a href="#i_stackrestore">
3782<tt>llvm.stackrestore</tt></a>. This is useful for implementing language
3783features like scoped automatic variable sized arrays in C99.
3784</p>
3785
3786<h5>Semantics:</h5>
3787
3788<p>
3789This intrinsic returns a opaque pointer value that can be passed to <a
3790href="#i_stackrestore"><tt>llvm.stackrestore</tt></a>. When an
3791<tt>llvm.stackrestore</tt> intrinsic is executed with a value saved from
3792<tt>llvm.stacksave</tt>, it effectively restores the state of the stack to the
3793state it was in when the <tt>llvm.stacksave</tt> intrinsic executed. In
3794practice, this pops any <a href="#i_alloca">alloca</a> blocks from the stack
3795that were allocated after the <tt>llvm.stacksave</tt> was executed.
3796</p>
3797
3798</div>
3799
3800<!-- _______________________________________________________________________ -->
3801<div class="doc_subsubsection">
3802 <a name="i_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a>
3803</div>
3804
3805<div class="doc_text">
3806
3807<h5>Syntax:</h5>
3808<pre>
3809 declare void %llvm.stackrestore(sbyte* %ptr)
3810</pre>
3811
3812<h5>Overview:</h5>
3813
3814<p>
3815The '<tt>llvm.stackrestore</tt>' intrinsic is used to restore the state of
3816the function stack to the state it was in when the corresponding <a
3817href="#llvm.stacksave"><tt>llvm.stacksave</tt></a> intrinsic executed. This is
3818useful for implementing language features like scoped automatic variable sized
3819arrays in C99.
3820</p>
3821
3822<h5>Semantics:</h5>
3823
3824<p>
3825See the description for <a href="#i_stacksave"><tt>llvm.stacksave</tt></a>.
3826</p>
3827
3828</div>
3829
3830
3831<!-- _______________________________________________________________________ -->
3832<div class="doc_subsubsection">
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]3833 <a name="i_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a>
3834</div>
3835
3836<div class="doc_text">
3837
3838<h5>Syntax:</h5>
3839<pre>
Reid Spencera8d451e2005-04-26 20:50:44 +0000[diff] [blame]3840 declare void %llvm.prefetch(sbyte * &lt;address&gt;,
3841 uint &lt;rw&gt;, uint &lt;locality&gt;)
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]3842</pre>
3843
3844<h5>Overview:</h5>
3845
3846
3847<p>
3848The '<tt>llvm.prefetch</tt>' intrinsic is a hint to the code generator to insert
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3849a prefetch instruction if supported; otherwise, it is a noop. Prefetches have
3850no
3851effect on the behavior of the program but can change its performance
Chris Lattner2a615362005-02-28 19:47:14 +0000[diff] [blame]3852characteristics.
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]3853</p>
3854
3855<h5>Arguments:</h5>
3856
3857<p>
3858<tt>address</tt> is the address to be prefetched, <tt>rw</tt> is the specifier
3859determining if the fetch should be for a read (0) or write (1), and
3860<tt>locality</tt> is a temporal locality specifier ranging from (0) - no
Chris Lattneraeffb4a2005-03-07 20:31:38 +0000[diff] [blame]3861locality, to (3) - extremely local keep in cache. The <tt>rw</tt> and
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]3862<tt>locality</tt> arguments must be constant integers.
3863</p>
3864
3865<h5>Semantics:</h5>
3866
3867<p>
3868This intrinsic does not modify the behavior of the program. In particular,
3869prefetches cannot trap and do not produce a value. On targets that support this
3870intrinsic, the prefetch can provide hints to the processor cache for better
3871performance.
3872</p>
3873
3874</div>
3875
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]3876<!-- _______________________________________________________________________ -->
3877<div class="doc_subsubsection">
3878 <a name="i_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a>
3879</div>
3880
3881<div class="doc_text">
3882
3883<h5>Syntax:</h5>
3884<pre>
Reid Spencera8d451e2005-04-26 20:50:44 +0000[diff] [blame]3885 declare void %llvm.pcmarker( uint &lt;id&gt; )
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]3886</pre>
3887
3888<h5>Overview:</h5>
3889
3890
3891<p>
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3892The '<tt>llvm.pcmarker</tt>' intrinsic is a method to export a Program Counter
3893(PC) in a region of
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]3894code to simulators and other tools. The method is target specific, but it is
3895expected that the marker will use exported symbols to transmit the PC of the marker.
Jeff Cohen25d4f7e2005-11-11 02:15:27 +0000[diff] [blame]3896The marker makes no guarantees that it will remain with any specific instruction
Chris Lattnerd07c3f42005-11-15 06:07:55 +0000[diff] [blame]3897after optimizations. It is possible that the presence of a marker will inhibit
Chris Lattnerb3e7afd2006-03-24 07:16:10 +0000[diff] [blame]3898optimizations. The intended use is to be inserted after optimizations to allow
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3899correlations of simulation runs.
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]3900</p>
3901
3902<h5>Arguments:</h5>
3903
3904<p>
3905<tt>id</tt> is a numerical id identifying the marker.
3906</p>
3907
3908<h5>Semantics:</h5>
3909
3910<p>
3911This intrinsic does not modify the behavior of the program. Backends that do not
3912support this intrinisic may ignore it.
3913</p>
3914
3915</div>
3916
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]3917<!-- _______________________________________________________________________ -->
3918<div class="doc_subsubsection">
3919 <a name="i_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a>
3920</div>
3921
3922<div class="doc_text">
3923
3924<h5>Syntax:</h5>
3925<pre>
3926 declare ulong %llvm.readcyclecounter( )
3927</pre>
3928
3929<h5>Overview:</h5>
3930
3931
3932<p>
3933The '<tt>llvm.readcyclecounter</tt>' intrinsic provides access to the cycle
3934counter register (or similar low latency, high accuracy clocks) on those targets
3935that support it. On X86, it should map to RDTSC. On Alpha, it should map to RPCC.
3936As the backing counters overflow quickly (on the order of 9 seconds on alpha), this
3937should only be used for small timings.
3938</p>
3939
3940<h5>Semantics:</h5>
3941
3942<p>
3943When directly supported, reading the cycle counter should not modify any memory.
3944Implementations are allowed to either return a application specific value or a
3945system wide value. On backends without support, this is lowered to a constant 0.
3946</p>
3947
3948</div>
3949
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3950<!-- ======================================================================= -->
3951<div class="doc_subsection">
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3952 <a name="int_libc">Standard C Library Intrinsics</a>
3953</div>
3954
3955<div class="doc_text">
3956<p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3957LLVM provides intrinsics for a few important standard C library functions.
3958These intrinsics allow source-language front-ends to pass information about the
3959alignment of the pointer arguments to the code generator, providing opportunity
3960for more efficient code generation.
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3961</p>
3962
3963</div>
3964
3965<!-- _______________________________________________________________________ -->
3966<div class="doc_subsubsection">
3967 <a name="i_memcpy">'<tt>llvm.memcpy</tt>' Intrinsic</a>
3968</div>
3969
3970<div class="doc_text">
3971
3972<h5>Syntax:</h5>
3973<pre>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]3974 declare void %llvm.memcpy.i32(sbyte* &lt;dest&gt;, sbyte* &lt;src&gt;,
3975 uint &lt;len&gt;, uint &lt;align&gt;)
3976 declare void %llvm.memcpy.i64(sbyte* &lt;dest&gt;, sbyte* &lt;src&gt;,
3977 ulong &lt;len&gt;, uint &lt;align&gt;)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3978</pre>
3979
3980<h5>Overview:</h5>
3981
3982<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]3983The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3984location to the destination location.
3985</p>
3986
3987<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]3988Note that, unlike the standard libc function, the <tt>llvm.memcpy.*</tt>
3989intrinsics do not return a value, and takes an extra alignment argument.
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3990</p>
3991
3992<h5>Arguments:</h5>
3993
3994<p>
3995The first argument is a pointer to the destination, the second is a pointer to
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]3996the source. The third argument is an integer argument
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3997specifying the number of bytes to copy, and the fourth argument is the alignment
3998of the source and destination locations.
3999</p>
4000
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]4001<p>
4002If 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]4003the caller guarantees that both the source and destination pointers are aligned
4004to that boundary.
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]4005</p>
4006
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4007<h5>Semantics:</h5>
4008
4009<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4010The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4011location to the destination location, which are not allowed to overlap. It
4012copies "len" bytes of memory over. If the argument is known to be aligned to
4013some boundary, this can be specified as the fourth argument, otherwise it should
4014be set to 0 or 1.
4015</p>
4016</div>
4017
4018
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]4019<!-- _______________________________________________________________________ -->
4020<div class="doc_subsubsection">
4021 <a name="i_memmove">'<tt>llvm.memmove</tt>' Intrinsic</a>
4022</div>
4023
4024<div class="doc_text">
4025
4026<h5>Syntax:</h5>
4027<pre>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4028 declare void %llvm.memmove.i32(sbyte* &lt;dest&gt;, sbyte* &lt;src&gt;,
4029 uint &lt;len&gt;, uint &lt;align&gt;)
4030 declare void %llvm.memmove.i64(sbyte* &lt;dest&gt;, sbyte* &lt;src&gt;,
4031 ulong &lt;len&gt;, uint &lt;align&gt;)
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]4032</pre>
4033
4034<h5>Overview:</h5>
4035
4036<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4037The '<tt>llvm.memmove.*</tt>' intrinsics move a block of memory from the source
4038location to the destination location. It is similar to the
4039'<tt>llvm.memcmp</tt>' intrinsic but allows the two memory locations to overlap.
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]4040</p>
4041
4042<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4043Note that, unlike the standard libc function, the <tt>llvm.memmove.*</tt>
4044intrinsics do not return a value, and takes an extra alignment argument.
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]4045</p>
4046
4047<h5>Arguments:</h5>
4048
4049<p>
4050The first argument is a pointer to the destination, the second is a pointer to
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4051the source. The third argument is an integer argument
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]4052specifying the number of bytes to copy, and the fourth argument is the alignment
4053of the source and destination locations.
4054</p>
4055
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]4056<p>
4057If 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]4058the caller guarantees that the source and destination pointers are aligned to
4059that boundary.
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]4060</p>
4061
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]4062<h5>Semantics:</h5>
4063
4064<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4065The '<tt>llvm.memmove.*</tt>' intrinsics copy a block of memory from the source
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]4066location to the destination location, which may overlap. It
4067copies "len" bytes of memory over. If the argument is known to be aligned to
4068some boundary, this can be specified as the fourth argument, otherwise it should
4069be set to 0 or 1.
4070</p>
4071</div>
4072
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]4073
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4074<!-- _______________________________________________________________________ -->
4075<div class="doc_subsubsection">
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4076 <a name="i_memset">'<tt>llvm.memset.*</tt>' Intrinsics</a>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4077</div>
4078
4079<div class="doc_text">
4080
4081<h5>Syntax:</h5>
4082<pre>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4083 declare void %llvm.memset.i32(sbyte* &lt;dest&gt;, ubyte &lt;val&gt;,
4084 uint &lt;len&gt;, uint &lt;align&gt;)
4085 declare void %llvm.memset.i64(sbyte* &lt;dest&gt;, ubyte &lt;val&gt;,
4086 ulong &lt;len&gt;, uint &lt;align&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4087</pre>
4088
4089<h5>Overview:</h5>
4090
4091<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4092The '<tt>llvm.memset.*</tt>' intrinsics fill a block of memory with a particular
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4093byte value.
4094</p>
4095
4096<p>
4097Note that, unlike the standard libc function, the <tt>llvm.memset</tt> intrinsic
4098does not return a value, and takes an extra alignment argument.
4099</p>
4100
4101<h5>Arguments:</h5>
4102
4103<p>
4104The first argument is a pointer to the destination to fill, the second is the
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4105byte value to fill it with, the third argument is an integer
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4106argument specifying the number of bytes to fill, and the fourth argument is the
4107known alignment of destination location.
4108</p>
4109
4110<p>
4111If 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]4112the caller guarantees that the destination pointer is aligned to that boundary.
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4113</p>
4114
4115<h5>Semantics:</h5>
4116
4117<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4118The '<tt>llvm.memset.*</tt>' intrinsics fill "len" bytes of memory starting at
4119the
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4120destination location. If the argument is known to be aligned to some boundary,
4121this can be specified as the fourth argument, otherwise it should be set to 0 or
41221.
4123</p>
4124</div>
4125
4126
Chris Lattner32006282004-06-11 02:28:03 +0000[diff] [blame]4127<!-- _______________________________________________________________________ -->
4128<div class="doc_subsubsection">
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]4129 <a name="i_isunordered">'<tt>llvm.isunordered.*</tt>' Intrinsic</a>
Alkis Evlogimenos26bbe932004-06-13 01:16:15 +0000[diff] [blame]4130</div>
4131
4132<div class="doc_text">
4133
4134<h5>Syntax:</h5>
4135<pre>
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]4136 declare bool %llvm.isunordered.f32(float Val1, float Val2)
4137 declare bool %llvm.isunordered.f64(double Val1, double Val2)
Alkis Evlogimenos26bbe932004-06-13 01:16:15 +0000[diff] [blame]4138</pre>
4139
4140<h5>Overview:</h5>
4141
4142<p>
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]4143The '<tt>llvm.isunordered</tt>' intrinsics return true if either or both of the
Alkis Evlogimenos26bbe932004-06-13 01:16:15 +0000[diff] [blame]4144specified floating point values is a NAN.
4145</p>
4146
4147<h5>Arguments:</h5>
4148
4149<p>
4150The arguments are floating point numbers of the same type.
4151</p>
4152
4153<h5>Semantics:</h5>
4154
4155<p>
4156If either or both of the arguments is a SNAN or QNAN, it returns true, otherwise
4157false.
4158</p>
4159</div>
4160
4161
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]4162<!-- _______________________________________________________________________ -->
4163<div class="doc_subsubsection">
Chris Lattnerec6cb612006-01-16 22:38:59 +0000[diff] [blame]4164 <a name="i_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a>
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]4165</div>
4166
4167<div class="doc_text">
4168
4169<h5>Syntax:</h5>
4170<pre>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]4171 declare float %llvm.sqrt.f32(float %Val)
4172 declare double %llvm.sqrt.f64(double %Val)
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]4173</pre>
4174
4175<h5>Overview:</h5>
4176
4177<p>
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]4178The '<tt>llvm.sqrt</tt>' intrinsics return the sqrt of the specified operand,
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]4179returning the same value as the libm '<tt>sqrt</tt>' function would. Unlike
4180<tt>sqrt</tt> in libm, however, <tt>llvm.sqrt</tt> has undefined behavior for
4181negative numbers (which allows for better optimization).
4182</p>
4183
4184<h5>Arguments:</h5>
4185
4186<p>
4187The argument and return value are floating point numbers of the same type.
4188</p>
4189
4190<h5>Semantics:</h5>
4191
4192<p>
4193This function returns the sqrt of the specified operand if it is a positive
4194floating point number.
4195</p>
4196</div>
4197
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]4198<!-- _______________________________________________________________________ -->
4199<div class="doc_subsubsection">
4200 <a name="i_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a>
4201</div>
4202
4203<div class="doc_text">
4204
4205<h5>Syntax:</h5>
4206<pre>
4207 declare float %llvm.powi.f32(float %Val, int %power)
4208 declare double %llvm.powi.f64(double %Val, int %power)
4209</pre>
4210
4211<h5>Overview:</h5>
4212
4213<p>
4214The '<tt>llvm.powi.*</tt>' intrinsics return the first operand raised to the
4215specified (positive or negative) power. The order of evaluation of
4216multiplications is not defined.
4217</p>
4218
4219<h5>Arguments:</h5>
4220
4221<p>
4222The second argument is an integer power, and the first is a value to raise to
4223that power.
4224</p>
4225
4226<h5>Semantics:</h5>
4227
4228<p>
4229This function returns the first value raised to the second power with an
4230unspecified sequence of rounding operations.</p>
4231</div>
4232
4233
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4234<!-- ======================================================================= -->
4235<div class="doc_subsection">
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]4236 <a name="int_manip">Bit Manipulation Intrinsics</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4237</div>
4238
4239<div class="doc_text">
4240<p>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]4241LLVM provides intrinsics for a few important bit manipulation operations.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4242These allow efficient code generation for some algorithms.
4243</p>
4244
4245</div>
4246
4247<!-- _______________________________________________________________________ -->
4248<div class="doc_subsubsection">
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]4249 <a name="i_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a>
4250</div>
4251
4252<div class="doc_text">
4253
4254<h5>Syntax:</h5>
4255<pre>
Chris Lattnerec6cb612006-01-16 22:38:59 +0000[diff] [blame]4256 declare ushort %llvm.bswap.i16(ushort &lt;id&gt;)
4257 declare uint %llvm.bswap.i32(uint &lt;id&gt;)
4258 declare ulong %llvm.bswap.i64(ulong &lt;id&gt;)
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]4259</pre>
4260
4261<h5>Overview:</h5>
4262
4263<p>
4264The '<tt>llvm.bwsap</tt>' family of intrinsics is used to byteswap a 16, 32 or
426564 bit quantity. These are useful for performing operations on data that is not
4266in the target's native byte order.
4267</p>
4268
4269<h5>Semantics:</h5>
4270
4271<p>
Chris Lattnerec6cb612006-01-16 22:38:59 +0000[diff] [blame]4272The <tt>llvm.bswap.16</tt> intrinsic returns a ushort value that has the high and low
4273byte of the input ushort swapped. Similarly, the <tt>llvm.bswap.i32</tt> intrinsic
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]4274returns a uint value that has the four bytes of the input uint swapped, so that
4275if 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]4276bytes 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]4277to 64 bits.
4278</p>
4279
4280</div>
4281
4282<!-- _______________________________________________________________________ -->
4283<div class="doc_subsubsection">
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]4284 <a name="int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4285</div>
4286
4287<div class="doc_text">
4288
4289<h5>Syntax:</h5>
4290<pre>
Chris Lattnerec6cb612006-01-16 22:38:59 +0000[diff] [blame]4291 declare ubyte %llvm.ctpop.i8 (ubyte &lt;src&gt;)
4292 declare ushort %llvm.ctpop.i16(ushort &lt;src&gt;)
4293 declare uint %llvm.ctpop.i32(uint &lt;src&gt;)
4294 declare ulong %llvm.ctpop.i64(ulong &lt;src&gt;)
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4295</pre>
4296
4297<h5>Overview:</h5>
4298
4299<p>
Chris Lattnerec6cb612006-01-16 22:38:59 +0000[diff] [blame]4300The '<tt>llvm.ctpop</tt>' family of intrinsics counts the number of bits set in a
4301value.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4302</p>
4303
4304<h5>Arguments:</h5>
4305
4306<p>
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]4307The only argument is the value to be counted. The argument may be of any
Chris Lattnerec6cb612006-01-16 22:38:59 +0000[diff] [blame]4308unsigned integer type. The return type must match the argument type.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4309</p>
4310
4311<h5>Semantics:</h5>
4312
4313<p>
4314The '<tt>llvm.ctpop</tt>' intrinsic counts the 1's in a variable.
4315</p>
4316</div>
4317
4318<!-- _______________________________________________________________________ -->
4319<div class="doc_subsubsection">
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]4320 <a name="int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4321</div>
4322
4323<div class="doc_text">
4324
4325<h5>Syntax:</h5>
4326<pre>
Chris Lattnerec6cb612006-01-16 22:38:59 +0000[diff] [blame]4327 declare ubyte %llvm.ctlz.i8 (ubyte &lt;src&gt;)
4328 declare ushort %llvm.ctlz.i16(ushort &lt;src&gt;)
4329 declare uint %llvm.ctlz.i32(uint &lt;src&gt;)
4330 declare ulong %llvm.ctlz.i64(ulong &lt;src&gt;)
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4331</pre>
4332
4333<h5>Overview:</h5>
4334
4335<p>
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]4336The '<tt>llvm.ctlz</tt>' family of intrinsic functions counts the number of
4337leading zeros in a variable.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4338</p>
4339
4340<h5>Arguments:</h5>
4341
4342<p>
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]4343The only argument is the value to be counted. The argument may be of any
Chris Lattnerec6cb612006-01-16 22:38:59 +0000[diff] [blame]4344unsigned integer type. The return type must match the argument type.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4345</p>
4346
4347<h5>Semantics:</h5>
4348
4349<p>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4350The '<tt>llvm.ctlz</tt>' intrinsic counts the leading (most significant) zeros
4351in 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]4352of src. For example, <tt>llvm.ctlz(int 2) = 30</tt>.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4353</p>
4354</div>
Chris Lattner32006282004-06-11 02:28:03 +0000[diff] [blame]4355
4356
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4357
4358<!-- _______________________________________________________________________ -->
4359<div class="doc_subsubsection">
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]4360 <a name="int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic</a>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4361</div>
4362
4363<div class="doc_text">
4364
4365<h5>Syntax:</h5>
4366<pre>
Chris Lattnerec6cb612006-01-16 22:38:59 +0000[diff] [blame]4367 declare ubyte %llvm.cttz.i8 (ubyte &lt;src&gt;)
4368 declare ushort %llvm.cttz.i16(ushort &lt;src&gt;)
4369 declare uint %llvm.cttz.i32(uint &lt;src&gt;)
4370 declare ulong %llvm.cttz.i64(ulong &lt;src&gt;)
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4371</pre>
4372
4373<h5>Overview:</h5>
4374
4375<p>
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]4376The '<tt>llvm.cttz</tt>' family of intrinsic functions counts the number of
4377trailing zeros.
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4378</p>
4379
4380<h5>Arguments:</h5>
4381
4382<p>
4383The only argument is the value to be counted. The argument may be of any
Chris Lattnerec6cb612006-01-16 22:38:59 +0000[diff] [blame]4384unsigned integer type. The return type must match the argument type.
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4385</p>
4386
4387<h5>Semantics:</h5>
4388
4389<p>
4390The '<tt>llvm.cttz</tt>' intrinsic counts the trailing (least significant) zeros
4391in a variable. If the src == 0 then the result is the size in bits of the type
4392of src. For example, <tt>llvm.cttz(2) = 1</tt>.
4393</p>
4394</div>
4395
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]4396<!-- ======================================================================= -->
4397<div class="doc_subsection">
4398 <a name="int_debugger">Debugger Intrinsics</a>
4399</div>
4400
4401<div class="doc_text">
4402<p>
4403The LLVM debugger intrinsics (which all start with <tt>llvm.dbg.</tt> prefix),
4404are described in the <a
4405href="SourceLevelDebugging.html#format_common_intrinsics">LLVM Source Level
4406Debugging</a> document.
4407</p>
4408</div>
4409
4410
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4411<!-- *********************************************************************** -->
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4412<hr>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]4413<address>
4414 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
4415 src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
4416 <a href="http://validator.w3.org/check/referer"><img
4417 src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!" /></a>
4418
4419 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
Reid Spencer05fe4b02006-03-14 05:39:39 +0000[diff] [blame]4420 <a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]4421 Last modified: $Date$
4422</address>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4423</body>
4424</html>