Gitiles
Code Review Sign In
gerrit-public.fairphone.software / fp2-dev / platform / external / llvm / 6d806e93cee502459fc15d039a16b9dc7e9cc30e / . / docs / LangRef.html
blob: 796aa563a7b27db5bc468c04415776a8fef74526 [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>
Anton Korobeynikovc6c98af2007-04-29 18:02:48 +0000[diff] [blame]27 <li><a href="#aliasstructure">Aliases</a>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]28 <li><a href="#paramattrs">Parameter Attributes</a></li>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]29 <li><a href="#moduleasm">Module-Level Inline Assembly</a></li>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]30 <li><a href="#datalayout">Data Layout</a></li>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]31 </ol>
32 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]33 <li><a href="#typesystem">Type System</a>
34 <ol>
Robert Bocchino7b81c752006-02-17 21:18:08 +0000[diff] [blame]35 <li><a href="#t_primitive">Primitive Types</a>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]36 <ol>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]37 <li><a href="#t_classifications">Type Classifications</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]38 </ol>
39 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]40 <li><a href="#t_derived">Derived Types</a>
41 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]42 <li><a href="#t_array">Array Type</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]43 <li><a href="#t_function">Function Type</a></li>
44 <li><a href="#t_pointer">Pointer Type</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]45 <li><a href="#t_struct">Structure Type</a></li>
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]46 <li><a href="#t_pstruct">Packed Structure Type</a></li>
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]47 <li><a href="#t_vector">Vector Type</a></li>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]48 <li><a href="#t_opaque">Opaque Type</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]49 </ol>
50 </li>
51 </ol>
52 </li>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]53 <li><a href="#constants">Constants</a>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]54 <ol>
55 <li><a href="#simpleconstants">Simple Constants</a>
56 <li><a href="#aggregateconstants">Aggregate Constants</a>
57 <li><a href="#globalconstants">Global Variable and Function Addresses</a>
58 <li><a href="#undefvalues">Undefined Values</a>
59 <li><a href="#constantexprs">Constant Expressions</a>
60 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]61 </li>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]62 <li><a href="#othervalues">Other Values</a>
63 <ol>
64 <li><a href="#inlineasm">Inline Assembler Expressions</a>
65 </ol>
66 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]67 <li><a href="#instref">Instruction Reference</a>
68 <ol>
69 <li><a href="#terminators">Terminator Instructions</a>
70 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]71 <li><a href="#i_ret">'<tt>ret</tt>' Instruction</a></li>
72 <li><a href="#i_br">'<tt>br</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]73 <li><a href="#i_switch">'<tt>switch</tt>' Instruction</a></li>
74 <li><a href="#i_invoke">'<tt>invoke</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]75 <li><a href="#i_unwind">'<tt>unwind</tt>' Instruction</a></li>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]76 <li><a href="#i_unreachable">'<tt>unreachable</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]77 </ol>
78 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]79 <li><a href="#binaryops">Binary Operations</a>
80 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]81 <li><a href="#i_add">'<tt>add</tt>' Instruction</a></li>
82 <li><a href="#i_sub">'<tt>sub</tt>' Instruction</a></li>
83 <li><a href="#i_mul">'<tt>mul</tt>' Instruction</a></li>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]84 <li><a href="#i_udiv">'<tt>udiv</tt>' Instruction</a></li>
85 <li><a href="#i_sdiv">'<tt>sdiv</tt>' Instruction</a></li>
86 <li><a href="#i_fdiv">'<tt>fdiv</tt>' Instruction</a></li>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]87 <li><a href="#i_urem">'<tt>urem</tt>' Instruction</a></li>
88 <li><a href="#i_srem">'<tt>srem</tt>' Instruction</a></li>
89 <li><a href="#i_frem">'<tt>frem</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]90 </ol>
91 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]92 <li><a href="#bitwiseops">Bitwise Binary Operations</a>
93 <ol>
Reid Spencer8e11bf82007-02-02 13:57:07 +0000[diff] [blame]94 <li><a href="#i_shl">'<tt>shl</tt>' Instruction</a></li>
95 <li><a href="#i_lshr">'<tt>lshr</tt>' Instruction</a></li>
96 <li><a href="#i_ashr">'<tt>ashr</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]97 <li><a href="#i_and">'<tt>and</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]98 <li><a href="#i_or">'<tt>or</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]99 <li><a href="#i_xor">'<tt>xor</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]100 </ol>
101 </li>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]102 <li><a href="#vectorops">Vector Operations</a>
103 <ol>
104 <li><a href="#i_extractelement">'<tt>extractelement</tt>' Instruction</a></li>
105 <li><a href="#i_insertelement">'<tt>insertelement</tt>' Instruction</a></li>
106 <li><a href="#i_shufflevector">'<tt>shufflevector</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>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]126 <li><a href="#i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a></li>
127 <li><a href="#i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a></li>
128 <li><a href="#i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a></li>
129 <li><a href="#i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a></li>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]130 <li><a href="#i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a></li>
131 <li><a href="#i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a></li>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]132 <li><a href="#i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a></li>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]133 </ol>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]134 <li><a href="#otherops">Other Operations</a>
135 <ol>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]136 <li><a href="#i_icmp">'<tt>icmp</tt>' Instruction</a></li>
137 <li><a href="#i_fcmp">'<tt>fcmp</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]138 <li><a href="#i_phi">'<tt>phi</tt>' Instruction</a></li>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]139 <li><a href="#i_select">'<tt>select</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]140 <li><a href="#i_call">'<tt>call</tt>' Instruction</a></li>
Chris Lattnerfb6977d2006-01-13 23:26:01 +0000[diff] [blame]141 <li><a href="#i_va_arg">'<tt>va_arg</tt>' Instruction</a></li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]142 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]143 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]144 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]145 </li>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]146 <li><a href="#intrinsics">Intrinsic Functions</a>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]147 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]148 <li><a href="#int_varargs">Variable Argument Handling Intrinsics</a>
149 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]150 <li><a href="#int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a></li>
151 <li><a href="#int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a></li>
152 <li><a href="#int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]153 </ol>
154 </li>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]155 <li><a href="#int_gc">Accurate Garbage Collection Intrinsics</a>
156 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]157 <li><a href="#int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a></li>
158 <li><a href="#int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a></li>
159 <li><a href="#int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a></li>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]160 </ol>
161 </li>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]162 <li><a href="#int_codegen">Code Generator Intrinsics</a>
163 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]164 <li><a href="#int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a></li>
165 <li><a href="#int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a></li>
166 <li><a href="#int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a></li>
167 <li><a href="#int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a></li>
168 <li><a href="#int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a></li>
169 <li><a href="#int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a></li>
170 <li><a href="#int_readcyclecounter"><tt>llvm.readcyclecounter</tt>' Intrinsic</a></li>
John Criswell7123e272004-04-09 16:43:20 +0000[diff] [blame]171 </ol>
172 </li>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]173 <li><a href="#int_libc">Standard C Library Intrinsics</a>
174 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]175 <li><a href="#int_memcpy">'<tt>llvm.memcpy.*</tt>' Intrinsic</a></li>
176 <li><a href="#int_memmove">'<tt>llvm.memmove.*</tt>' Intrinsic</a></li>
177 <li><a href="#int_memset">'<tt>llvm.memset.*</tt>' Intrinsic</a></li>
178 <li><a href="#int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a></li>
179 <li><a href="#int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a></li>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]180 </ol>
181 </li>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]182 <li><a href="#int_manip">Bit Manipulation Intrinsics</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]183 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]184 <li><a href="#int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a></li>
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]185 <li><a href="#int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic </a></li>
186 <li><a href="#int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic </a></li>
187 <li><a href="#int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic </a></li>
Reid Spencerf86037f2007-04-11 23:23:49 +0000[diff] [blame]188 <li><a href="#int_part_select">'<tt>llvm.part.select.*</tt>' Intrinsic </a></li>
189 <li><a href="#int_part_set">'<tt>llvm.part.set.*</tt>' Intrinsic </a></li>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]190 </ol>
191 </li>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]192 <li><a href="#int_debugger">Debugger intrinsics</a></li>
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]193 <li><a href="#int_eh">Exception Handling intrinsics</a></li>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame^]194 <li><a href="#int_general">General intrinsics</a></li>
195 <ol>
196 <li><a href="#int_var_annotation">'<tt>llvm.var.annotation</tt>'
197 Intrinsic</a></li>
198 </ol>
199 </li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]200 </ol>
201 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]202</ol>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]203
204<div class="doc_author">
205 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>
206 and <a href="mailto:vadve@cs.uiuc.edu">Vikram Adve</a></p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]207</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]208
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]209<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]210<div class="doc_section"> <a name="abstract">Abstract </a></div>
211<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]212
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]213<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]214<p>This document is a reference manual for the LLVM assembly language.
215LLVM is an SSA based representation that provides type safety,
216low-level operations, flexibility, and the capability of representing
217'all' high-level languages cleanly. It is the common code
218representation used throughout all phases of the LLVM compilation
219strategy.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]220</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]221
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]222<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]223<div class="doc_section"> <a name="introduction">Introduction</a> </div>
224<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]225
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]226<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]227
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]228<p>The LLVM code representation is designed to be used in three
229different forms: as an in-memory compiler IR, as an on-disk bytecode
230representation (suitable for fast loading by a Just-In-Time compiler),
231and as a human readable assembly language representation. This allows
232LLVM to provide a powerful intermediate representation for efficient
233compiler transformations and analysis, while providing a natural means
234to debug and visualize the transformations. The three different forms
235of LLVM are all equivalent. This document describes the human readable
236representation and notation.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]237
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]238<p>The LLVM representation aims to be light-weight and low-level
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]239while being expressive, typed, and extensible at the same time. It
240aims to be a "universal IR" of sorts, by being at a low enough level
241that high-level ideas may be cleanly mapped to it (similar to how
242microprocessors are "universal IR's", allowing many source languages to
243be mapped to them). By providing type information, LLVM can be used as
244the target of optimizations: for example, through pointer analysis, it
245can be proven that a C automatic variable is never accessed outside of
246the current function... allowing it to be promoted to a simple SSA
247value instead of a memory location.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]248
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]249</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]250
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]251<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]252<div class="doc_subsubsection"> <a name="wellformed">Well-Formedness</a> </div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]253
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]254<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]255
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]256<p>It is important to note that this document describes 'well formed'
257LLVM assembly language. There is a difference between what the parser
258accepts and what is considered 'well formed'. For example, the
259following instruction is syntactically okay, but not well formed:</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]260
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]261<div class="doc_code">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]262<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]263%x = <a href="#i_add">add</a> i32 1, %x
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]264</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]265</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]266
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]267<p>...because the definition of <tt>%x</tt> does not dominate all of
268its uses. The LLVM infrastructure provides a verification pass that may
269be used to verify that an LLVM module is well formed. This pass is
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]270automatically run by the parser after parsing input assembly and by
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]271the optimizer before it outputs bytecode. The violations pointed out
272by the verifier pass indicate bugs in transformation passes or input to
273the parser.</p>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]274</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]275
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]276<!-- Describe the typesetting conventions here. --> </div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]277
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]278<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]279<div class="doc_section"> <a name="identifiers">Identifiers</a> </div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]280<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]281
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]282<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]283
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]284<p>LLVM uses three different forms of identifiers, for different
285purposes:</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]286
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]287<ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]288 <li>Named values are represented as a string of characters with a '%' prefix.
289 For example, %foo, %DivisionByZero, %a.really.long.identifier. The actual
290 regular expression used is '<tt>%[a-zA-Z$._][a-zA-Z$._0-9]*</tt>'.
291 Identifiers which require other characters in their names can be surrounded
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame]292 with quotes. In this way, anything except a <tt>&quot;</tt> character can be used
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]293 in a name.</li>
294
295 <li>Unnamed values are represented as an unsigned numeric value with a '%'
296 prefix. For example, %12, %2, %44.</li>
297
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]298 <li>Constants, which are described in a <a href="#constants">section about
299 constants</a>, below.</li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]300</ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]301
302<p>LLVM requires that values start with a '%' sign for two reasons: Compilers
303don't need to worry about name clashes with reserved words, and the set of
304reserved words may be expanded in the future without penalty. Additionally,
305unnamed identifiers allow a compiler to quickly come up with a temporary
306variable without having to avoid symbol table conflicts.</p>
307
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]308<p>Reserved words in LLVM are very similar to reserved words in other
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]309languages. There are keywords for different opcodes
310('<tt><a href="#i_add">add</a></tt>',
311 '<tt><a href="#i_bitcast">bitcast</a></tt>',
312 '<tt><a href="#i_ret">ret</a></tt>', etc...), for primitive type names ('<tt><a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]313href="#t_void">void</a></tt>', '<tt><a href="#t_primitive">i32</a></tt>', etc...),
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]314and others. These reserved words cannot conflict with variable names, because
315none of them start with a '%' character.</p>
316
317<p>Here is an example of LLVM code to multiply the integer variable
318'<tt>%X</tt>' by 8:</p>
319
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]320<p>The easy way:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]321
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]322<div class="doc_code">
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]323<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]324%result = <a href="#i_mul">mul</a> i32 %X, 8
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]325</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]326</div>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]327
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]328<p>After strength reduction:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]329
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]330<div class="doc_code">
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]331<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]332%result = <a href="#i_shl">shl</a> i32 %X, i8 3
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]333</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]334</div>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]335
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]336<p>And the hard way:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]337
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]338<div class="doc_code">
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]339<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]340<a href="#i_add">add</a> i32 %X, %X <i>; yields {i32}:%0</i>
341<a href="#i_add">add</a> i32 %0, %0 <i>; yields {i32}:%1</i>
342%result = <a href="#i_add">add</a> i32 %1, %1
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]343</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]344</div>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]345
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]346<p>This last way of multiplying <tt>%X</tt> by 8 illustrates several
347important lexical features of LLVM:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]348
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]349<ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]350
351 <li>Comments are delimited with a '<tt>;</tt>' and go until the end of
352 line.</li>
353
354 <li>Unnamed temporaries are created when the result of a computation is not
355 assigned to a named value.</li>
356
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]357 <li>Unnamed temporaries are numbered sequentially</li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]358
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]359</ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]360
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]361<p>...and it also shows a convention that we follow in this document. When
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]362demonstrating instructions, we will follow an instruction with a comment that
363defines the type and name of value produced. Comments are shown in italic
364text.</p>
365
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]366</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]367
368<!-- *********************************************************************** -->
369<div class="doc_section"> <a name="highlevel">High Level Structure</a> </div>
370<!-- *********************************************************************** -->
371
372<!-- ======================================================================= -->
373<div class="doc_subsection"> <a name="modulestructure">Module Structure</a>
374</div>
375
376<div class="doc_text">
377
378<p>LLVM programs are composed of "Module"s, each of which is a
379translation unit of the input programs. Each module consists of
380functions, global variables, and symbol table entries. Modules may be
381combined together with the LLVM linker, which merges function (and
382global variable) definitions, resolves forward declarations, and merges
383symbol table entries. Here is an example of the "hello world" module:</p>
384
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]385<div class="doc_code">
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]386<pre><i>; Declare the string constant as a global constant...</i>
Chris Lattnera89e5f12007-06-12 17:00:26 +0000[diff] [blame]387<a href="#identifiers">@.LC0</a> = <a href="#linkage_internal">internal</a> <a
388 href="#globalvars">constant</a> <a href="#t_array">[13 x i8]</a> c"hello world\0A\00" <i>; [13 x i8]*</i>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]389
390<i>; External declaration of the puts function</i>
Chris Lattnera89e5f12007-06-12 17:00:26 +0000[diff] [blame]391<a href="#functionstructure">declare</a> i32 @puts(i8 *) <i>; i32(i8 *)* </i>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]392
393<i>; Definition of main function</i>
Chris Lattnera89e5f12007-06-12 17:00:26 +0000[diff] [blame]394define i32 @main() { <i>; i32()* </i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]395 <i>; Convert [13x i8 ]* to i8 *...</i>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]396 %cast210 = <a
Chris Lattner6c0955b2007-06-12 17:01:15 +0000[diff] [blame]397 href="#i_getelementptr">getelementptr</a> [13 x i8 ]* @.LC0, i64 0, i64 0 <i>; i8 *</i>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]398
399 <i>; Call puts function to write out the string to stdout...</i>
400 <a
Chris Lattnera89e5f12007-06-12 17:00:26 +0000[diff] [blame]401 href="#i_call">call</a> i32 @puts(i8 * %cast210) <i>; i32</i>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]402 <a
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]403 href="#i_ret">ret</a> i32 0<br>}<br>
404</pre>
405</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]406
407<p>This example is made up of a <a href="#globalvars">global variable</a>
408named "<tt>.LC0</tt>", an external declaration of the "<tt>puts</tt>"
409function, and a <a href="#functionstructure">function definition</a>
410for "<tt>main</tt>".</p>
411
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]412<p>In general, a module is made up of a list of global values,
413where both functions and global variables are global values. Global values are
414represented by a pointer to a memory location (in this case, a pointer to an
415array of char, and a pointer to a function), and have one of the following <a
416href="#linkage">linkage types</a>.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]417
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]418</div>
419
420<!-- ======================================================================= -->
421<div class="doc_subsection">
422 <a name="linkage">Linkage Types</a>
423</div>
424
425<div class="doc_text">
426
427<p>
428All Global Variables and Functions have one of the following types of linkage:
429</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]430
431<dl>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]432
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]433 <dt><tt><b><a name="linkage_internal">internal</a></b></tt> </dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]434
435 <dd>Global values with internal linkage are only directly accessible by
436 objects in the current module. In particular, linking code into a module with
437 an internal global value may cause the internal to be renamed as necessary to
438 avoid collisions. Because the symbol is internal to the module, all
439 references can be updated. This corresponds to the notion of the
Chris Lattner4887bd82007-01-14 06:51:48 +0000[diff] [blame]440 '<tt>static</tt>' keyword in C.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]441 </dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]442
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]443 <dt><tt><b><a name="linkage_linkonce">linkonce</a></b></tt>: </dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]444
Chris Lattner4887bd82007-01-14 06:51:48 +0000[diff] [blame]445 <dd>Globals with "<tt>linkonce</tt>" linkage are merged with other globals of
446 the same name when linkage occurs. This is typically used to implement
447 inline functions, templates, or other code which must be generated in each
448 translation unit that uses it. Unreferenced <tt>linkonce</tt> globals are
449 allowed to be discarded.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]450 </dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]451
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]452 <dt><tt><b><a name="linkage_weak">weak</a></b></tt>: </dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]453
454 <dd>"<tt>weak</tt>" linkage is exactly the same as <tt>linkonce</tt> linkage,
455 except that unreferenced <tt>weak</tt> globals may not be discarded. This is
Chris Lattner4887bd82007-01-14 06:51:48 +0000[diff] [blame]456 used for globals that may be emitted in multiple translation units, but that
457 are not guaranteed to be emitted into every translation unit that uses them.
458 One example of this are common globals in C, such as "<tt>int X;</tt>" at
459 global scope.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]460 </dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]461
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]462 <dt><tt><b><a name="linkage_appending">appending</a></b></tt>: </dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]463
464 <dd>"<tt>appending</tt>" linkage may only be applied to global variables of
465 pointer to array type. When two global variables with appending linkage are
466 linked together, the two global arrays are appended together. This is the
467 LLVM, typesafe, equivalent of having the system linker append together
468 "sections" with identical names when .o files are linked.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]469 </dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]470
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame]471 <dt><tt><b><a name="linkage_externweak">extern_weak</a></b></tt>: </dt>
472 <dd>The semantics of this linkage follow the ELF model: the symbol is weak
473 until linked, if not linked, the symbol becomes null instead of being an
474 undefined reference.
475 </dd>
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame]476
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]477 <dt><tt><b><a name="linkage_external">externally visible</a></b></tt>:</dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]478
479 <dd>If none of the above identifiers are used, the global is externally
480 visible, meaning that it participates in linkage and can be used to resolve
481 external symbol references.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]482 </dd>
Reid Spencerc8910842007-04-11 23:49:50 +0000[diff] [blame]483</dl>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]484
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]485 <p>
486 The next two types of linkage are targeted for Microsoft Windows platform
487 only. They are designed to support importing (exporting) symbols from (to)
488 DLLs.
489 </p>
490
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame]491 <dl>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]492 <dt><tt><b><a name="linkage_dllimport">dllimport</a></b></tt>: </dt>
493
494 <dd>"<tt>dllimport</tt>" linkage causes the compiler to reference a function
495 or variable via a global pointer to a pointer that is set up by the DLL
496 exporting the symbol. On Microsoft Windows targets, the pointer name is
497 formed by combining <code>_imp__</code> and the function or variable name.
498 </dd>
499
500 <dt><tt><b><a name="linkage_dllexport">dllexport</a></b></tt>: </dt>
501
502 <dd>"<tt>dllexport</tt>" linkage causes the compiler to provide a global
503 pointer to a pointer in a DLL, so that it can be referenced with the
504 <tt>dllimport</tt> attribute. On Microsoft Windows targets, the pointer
505 name is formed by combining <code>_imp__</code> and the function or variable
506 name.
507 </dd>
508
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]509</dl>
510
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame]511<p><a name="linkage_external"></a>For example, since the "<tt>.LC0</tt>"
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]512variable is defined to be internal, if another module defined a "<tt>.LC0</tt>"
513variable and was linked with this one, one of the two would be renamed,
514preventing a collision. Since "<tt>main</tt>" and "<tt>puts</tt>" are
515external (i.e., lacking any linkage declarations), they are accessible
Reid Spencerac8d2762007-01-05 00:59:10 +0000[diff] [blame]516outside of the current module.</p>
517<p>It is illegal for a function <i>declaration</i>
518to have any linkage type other than "externally visible", <tt>dllimport</tt>,
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame]519or <tt>extern_weak</tt>.</p>
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]520<p>Aliases can have only <tt>external</tt>, <tt>internal</tt> and <tt>weak</tt>
521linkages.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]522</div>
523
524<!-- ======================================================================= -->
525<div class="doc_subsection">
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]526 <a name="callingconv">Calling Conventions</a>
527</div>
528
529<div class="doc_text">
530
531<p>LLVM <a href="#functionstructure">functions</a>, <a href="#i_call">calls</a>
532and <a href="#i_invoke">invokes</a> can all have an optional calling convention
533specified for the call. The calling convention of any pair of dynamic
534caller/callee must match, or the behavior of the program is undefined. The
535following calling conventions are supported by LLVM, and more may be added in
536the future:</p>
537
538<dl>
539 <dt><b>"<tt>ccc</tt>" - The C calling convention</b>:</dt>
540
541 <dd>This calling convention (the default if no other calling convention is
542 specified) matches the target C calling conventions. This calling convention
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]543 supports varargs function calls and tolerates some mismatch in the declared
Reid Spencerc28d2bc2006-12-31 21:30:18 +0000[diff] [blame]544 prototype and implemented declaration of the function (as does normal C).
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]545 </dd>
546
547 <dt><b>"<tt>fastcc</tt>" - The fast calling convention</b>:</dt>
548
549 <dd>This calling convention attempts to make calls as fast as possible
550 (e.g. by passing things in registers). This calling convention allows the
551 target to use whatever tricks it wants to produce fast code for the target,
Chris Lattner8cdc5bc2005-05-06 23:08:23 +0000[diff] [blame]552 without having to conform to an externally specified ABI. Implementations of
553 this convention should allow arbitrary tail call optimization to be supported.
554 This calling convention does not support varargs and requires the prototype of
555 all callees to exactly match the prototype of the function definition.
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]556 </dd>
557
558 <dt><b>"<tt>coldcc</tt>" - The cold calling convention</b>:</dt>
559
560 <dd>This calling convention attempts to make code in the caller as efficient
561 as possible under the assumption that the call is not commonly executed. As
562 such, these calls often preserve all registers so that the call does not break
563 any live ranges in the caller side. This calling convention does not support
564 varargs and requires the prototype of all callees to exactly match the
565 prototype of the function definition.
566 </dd>
567
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]568 <dt><b>"<tt>cc &lt;<em>n</em>&gt;</tt>" - Numbered convention</b>:</dt>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]569
570 <dd>Any calling convention may be specified by number, allowing
571 target-specific calling conventions to be used. Target specific calling
572 conventions start at 64.
573 </dd>
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]574</dl>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]575
576<p>More calling conventions can be added/defined on an as-needed basis, to
577support pascal conventions or any other well-known target-independent
578convention.</p>
579
580</div>
581
582<!-- ======================================================================= -->
583<div class="doc_subsection">
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]584 <a name="visibility">Visibility Styles</a>
585</div>
586
587<div class="doc_text">
588
589<p>
590All Global Variables and Functions have one of the following visibility styles:
591</p>
592
593<dl>
594 <dt><b>"<tt>default</tt>" - Default style</b>:</dt>
595
596 <dd>On ELF, default visibility means that the declaration is visible to other
597 modules and, in shared libraries, means that the declared entity may be
598 overridden. On Darwin, default visibility means that the declaration is
599 visible to other modules. Default visibility corresponds to "external
600 linkage" in the language.
601 </dd>
602
603 <dt><b>"<tt>hidden</tt>" - Hidden style</b>:</dt>
604
605 <dd>Two declarations of an object with hidden visibility refer to the same
606 object if they are in the same shared object. Usually, hidden visibility
607 indicates that the symbol will not be placed into the dynamic symbol table,
608 so no other module (executable or shared library) can reference it
609 directly.
610 </dd>
611
Anton Korobeynikov6f9896f2007-04-29 18:35:00 +0000[diff] [blame]612 <dt><b>"<tt>protected</tt>" - Protected style</b>:</dt>
613
614 <dd>On ELF, protected visibility indicates that the symbol will be placed in
615 the dynamic symbol table, but that references within the defining module will
616 bind to the local symbol. That is, the symbol cannot be overridden by another
617 module.
618 </dd>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]619</dl>
620
621</div>
622
623<!-- ======================================================================= -->
624<div class="doc_subsection">
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]625 <a name="globalvars">Global Variables</a>
626</div>
627
628<div class="doc_text">
629
Chris Lattner3689a342005-02-12 19:30:21 +0000[diff] [blame]630<p>Global variables define regions of memory allocated at compilation time
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]631instead of run-time. Global variables may optionally be initialized, may have
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]632an explicit section to be placed in, and may have an optional explicit alignment
633specified. A variable may be defined as "thread_local", which means that it
634will not be shared by threads (each thread will have a separated copy of the
635variable). A variable may be defined as a global "constant," which indicates
636that the contents of the variable will <b>never</b> be modified (enabling better
Chris Lattner3689a342005-02-12 19:30:21 +0000[diff] [blame]637optimization, allowing the global data to be placed in the read-only section of
638an executable, etc). Note that variables that need runtime initialization
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]639cannot be marked "constant" as there is a store to the variable.</p>
Chris Lattner3689a342005-02-12 19:30:21 +0000[diff] [blame]640
641<p>
642LLVM explicitly allows <em>declarations</em> of global variables to be marked
643constant, even if the final definition of the global is not. This capability
644can be used to enable slightly better optimization of the program, but requires
645the language definition to guarantee that optimizations based on the
646'constantness' are valid for the translation units that do not include the
647definition.
648</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]649
650<p>As SSA values, global variables define pointer values that are in
651scope (i.e. they dominate) all basic blocks in the program. Global
652variables always define a pointer to their "content" type because they
653describe a region of memory, and all memory objects in LLVM are
654accessed through pointers.</p>
655
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]656<p>LLVM allows an explicit section to be specified for globals. If the target
657supports it, it will emit globals to the section specified.</p>
658
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]659<p>An explicit alignment may be specified for a global. If not present, or if
660the alignment is set to zero, the alignment of the global is set by the target
661to whatever it feels convenient. If an explicit alignment is specified, the
662global is forced to have at least that much alignment. All alignments must be
663a power of 2.</p>
664
Chris Lattner68027ea2007-01-14 00:27:09 +0000[diff] [blame]665<p>For example, the following defines a global with an initializer, section,
666 and alignment:</p>
667
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]668<div class="doc_code">
Chris Lattner68027ea2007-01-14 00:27:09 +0000[diff] [blame]669<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]670%G = constant float 1.0, section "foo", align 4
Chris Lattner68027ea2007-01-14 00:27:09 +0000[diff] [blame]671</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]672</div>
Chris Lattner68027ea2007-01-14 00:27:09 +0000[diff] [blame]673
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]674</div>
675
676
677<!-- ======================================================================= -->
678<div class="doc_subsection">
679 <a name="functionstructure">Functions</a>
680</div>
681
682<div class="doc_text">
683
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]684<p>LLVM function definitions consist of the "<tt>define</tt>" keyord,
685an optional <a href="#linkage">linkage type</a>, an optional
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]686<a href="#visibility">visibility style</a>, an optional
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]687<a href="#callingconv">calling convention</a>, a return type, an optional
688<a href="#paramattrs">parameter attribute</a> for the return type, a function
689name, a (possibly empty) argument list (each with optional
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]690<a href="#paramattrs">parameter attributes</a>), an optional section, an
691optional alignment, an opening curly brace, a list of basic blocks, and a
692closing curly brace.
693
694LLVM function declarations consist of the "<tt>declare</tt>" keyword, an
695optional <a href="#linkage">linkage type</a>, an optional
696<a href="#visibility">visibility style</a>, an optional
697<a href="#callingconv">calling convention</a>, a return type, an optional
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]698<a href="#paramattrs">parameter attribute</a> for the return type, a function
699name, a possibly empty list of arguments, and an optional alignment.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]700
701<p>A function definition contains a list of basic blocks, forming the CFG for
702the function. Each basic block may optionally start with a label (giving the
703basic block a symbol table entry), contains a list of instructions, and ends
704with a <a href="#terminators">terminator</a> instruction (such as a branch or
705function return).</p>
706
Chris Lattner4a3c9012007-06-08 16:52:14 +0000[diff] [blame]707<p>The first basic block in a function is special in two ways: it is immediately
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]708executed on entrance to the function, and it is not allowed to have predecessor
709basic blocks (i.e. there can not be any branches to the entry block of a
710function). Because the block can have no predecessors, it also cannot have any
711<a href="#i_phi">PHI nodes</a>.</p>
712
713<p>LLVM functions are identified by their name and type signature. Hence, two
714functions with the same name but different parameter lists or return values are
Chris Lattnerd4f6b172005-03-07 22:13:59 +0000[diff] [blame]715considered different functions, and LLVM will resolve references to each
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]716appropriately.</p>
717
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]718<p>LLVM allows an explicit section to be specified for functions. If the target
719supports it, it will emit functions to the section specified.</p>
720
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]721<p>An explicit alignment may be specified for a function. If not present, or if
722the alignment is set to zero, the alignment of the function is set by the target
723to whatever it feels convenient. If an explicit alignment is specified, the
724function is forced to have at least that much alignment. All alignments must be
725a power of 2.</p>
726
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]727</div>
728
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]729
730<!-- ======================================================================= -->
731<div class="doc_subsection">
732 <a name="aliasstructure">Aliases</a>
733</div>
734<div class="doc_text">
735 <p>Aliases act as "second name" for the aliasee value (which can be either
Anton Korobeynikova80e1182007-04-28 13:45:00 +0000[diff] [blame]736 function or global variable or bitcast of global value). Aliases may have an
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]737 optional <a href="#linkage">linkage type</a>, and an
738 optional <a href="#visibility">visibility style</a>.</p>
739
740 <h5>Syntax:</h5>
741
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]742<div class="doc_code">
Bill Wendlingaac388b2007-05-29 09:42:13 +0000[diff] [blame]743<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]744@&lt;Name&gt; = [Linkage] [Visibility] alias &lt;AliaseeTy&gt; @&lt;Aliasee&gt;
Bill Wendlingaac388b2007-05-29 09:42:13 +0000[diff] [blame]745</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]746</div>
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]747
748</div>
749
750
751
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]752<!-- ======================================================================= -->
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]753<div class="doc_subsection"><a name="paramattrs">Parameter Attributes</a></div>
754<div class="doc_text">
755 <p>The return type and each parameter of a function type may have a set of
756 <i>parameter attributes</i> associated with them. Parameter attributes are
757 used to communicate additional information about the result or parameters of
758 a function. Parameter attributes are considered to be part of the function
759 type so two functions types that differ only by the parameter attributes
760 are different function types.</p>
761
Reid Spencer950e9f82007-01-15 18:27:39 +0000[diff] [blame]762 <p>Parameter attributes are simple keywords that follow the type specified. If
763 multiple parameter attributes are needed, they are space separated. For
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]764 example:</p>
765
766<div class="doc_code">
767<pre>
768%someFunc = i16 (i8 sext %someParam) zext
769%someFunc = i16 (i8 zext %someParam) zext
770</pre>
771</div>
772
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame]773 <p>Note that the two function types above are unique because the parameter has
Reid Spencer950e9f82007-01-15 18:27:39 +0000[diff] [blame]774 a different attribute (sext in the first one, zext in the second). Also note
775 that the attribute for the function result (zext) comes immediately after the
776 argument list.</p>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]777
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame]778 <p>Currently, only the following parameter attributes are defined:</p>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]779 <dl>
Reid Spencer950e9f82007-01-15 18:27:39 +0000[diff] [blame]780 <dt><tt>zext</tt></dt>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]781 <dd>This indicates that the parameter should be zero extended just before
782 a call to this function.</dd>
Reid Spencer950e9f82007-01-15 18:27:39 +0000[diff] [blame]783 <dt><tt>sext</tt></dt>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]784 <dd>This indicates that the parameter should be sign extended just before
785 a call to this function.</dd>
Anton Korobeynikov34d6dea2007-01-28 14:30:45 +0000[diff] [blame]786 <dt><tt>inreg</tt></dt>
787 <dd>This indicates that the parameter should be placed in register (if
Anton Korobeynikov66a8c8c2007-01-28 15:27:21 +0000[diff] [blame]788 possible) during assembling function call. Support for this attribute is
789 target-specific</dd>
Anton Korobeynikov34d6dea2007-01-28 14:30:45 +0000[diff] [blame]790 <dt><tt>sret</tt></dt>
Anton Korobeynikov66a8c8c2007-01-28 15:27:21 +0000[diff] [blame]791 <dd>This indicates that the parameter specifies the address of a structure
Reid Spencer67606122007-03-22 02:02:11 +0000[diff] [blame]792 that is the return value of the function in the source program.</dd>
Zhou Shengfebca342007-06-05 05:28:26 +0000[diff] [blame]793 <dt><tt>noalias</tt></dt>
794 <dd>This indicates that the parameter not alias any other object or any
795 other "noalias" objects during the function call.
Reid Spencer2dc52012007-03-22 02:18:56 +0000[diff] [blame]796 <dt><tt>noreturn</tt></dt>
797 <dd>This function attribute indicates that the function never returns. This
798 indicates to LLVM that every call to this function should be treated as if
799 an <tt>unreachable</tt> instruction immediately followed the call.</dd>
Reid Spencer67606122007-03-22 02:02:11 +0000[diff] [blame]800 <dt><tt>nounwind</tt></dt>
801 <dd>This function attribute indicates that the function type does not use
802 the unwind instruction and does not allow stack unwinding to propagate
803 through it.</dd>
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame]804 </dl>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]805
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]806</div>
807
808<!-- ======================================================================= -->
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]809<div class="doc_subsection">
Chris Lattner1eeeb0c2006-04-08 04:40:53 +0000[diff] [blame]810 <a name="moduleasm">Module-Level Inline Assembly</a>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]811</div>
812
813<div class="doc_text">
814<p>
815Modules may contain "module-level inline asm" blocks, which corresponds to the
816GCC "file scope inline asm" blocks. These blocks are internally concatenated by
817LLVM and treated as a single unit, but may be separated in the .ll file if
818desired. The syntax is very simple:
819</p>
820
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]821<div class="doc_code">
822<pre>
823module asm "inline asm code goes here"
824module asm "more can go here"
825</pre>
826</div>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]827
828<p>The strings can contain any character by escaping non-printable characters.
829 The escape sequence used is simply "\xx" where "xx" is the two digit hex code
830 for the number.
831</p>
832
833<p>
834 The inline asm code is simply printed to the machine code .s file when
835 assembly code is generated.
836</p>
837</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]838
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]839<!-- ======================================================================= -->
840<div class="doc_subsection">
841 <a name="datalayout">Data Layout</a>
842</div>
843
844<div class="doc_text">
845<p>A module may specify a target specific data layout string that specifies how
Reid Spencerc8910842007-04-11 23:49:50 +0000[diff] [blame]846data is to be laid out in memory. The syntax for the data layout is simply:</p>
847<pre> target datalayout = "<i>layout specification</i>"</pre>
848<p>The <i>layout specification</i> consists of a list of specifications
849separated by the minus sign character ('-'). Each specification starts with a
850letter and may include other information after the letter to define some
851aspect of the data layout. The specifications accepted are as follows: </p>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]852<dl>
853 <dt><tt>E</tt></dt>
854 <dd>Specifies that the target lays out data in big-endian form. That is, the
855 bits with the most significance have the lowest address location.</dd>
856 <dt><tt>e</tt></dt>
857 <dd>Specifies that hte target lays out data in little-endian form. That is,
858 the bits with the least significance have the lowest address location.</dd>
859 <dt><tt>p:<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
860 <dd>This specifies the <i>size</i> of a pointer and its <i>abi</i> and
861 <i>preferred</i> alignments. All sizes are in bits. Specifying the <i>pref</i>
862 alignment is optional. If omitted, the preceding <tt>:</tt> should be omitted
863 too.</dd>
864 <dt><tt>i<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
865 <dd>This specifies the alignment for an integer type of a given bit
866 <i>size</i>. The value of <i>size</i> must be in the range [1,2^23).</dd>
867 <dt><tt>v<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
868 <dd>This specifies the alignment for a vector type of a given bit
869 <i>size</i>.</dd>
870 <dt><tt>f<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
871 <dd>This specifies the alignment for a floating point type of a given bit
872 <i>size</i>. The value of <i>size</i> must be either 32 (float) or 64
873 (double).</dd>
874 <dt><tt>a<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
875 <dd>This specifies the alignment for an aggregate type of a given bit
876 <i>size</i>.</dd>
877</dl>
878<p>When constructing the data layout for a given target, LLVM starts with a
879default set of specifications which are then (possibly) overriden by the
880specifications in the <tt>datalayout</tt> keyword. The default specifications
881are given in this list:</p>
882<ul>
883 <li><tt>E</tt> - big endian</li>
884 <li><tt>p:32:64:64</tt> - 32-bit pointers with 64-bit alignment</li>
885 <li><tt>i1:8:8</tt> - i1 is 8-bit (byte) aligned</li>
886 <li><tt>i8:8:8</tt> - i8 is 8-bit (byte) aligned</li>
887 <li><tt>i16:16:16</tt> - i16 is 16-bit aligned</li>
888 <li><tt>i32:32:32</tt> - i32 is 32-bit aligned</li>
889 <li><tt>i64:32:64</tt> - i64 has abi alignment of 32-bits but preferred
890 alignment of 64-bits</li>
891 <li><tt>f32:32:32</tt> - float is 32-bit aligned</li>
892 <li><tt>f64:64:64</tt> - double is 64-bit aligned</li>
893 <li><tt>v64:64:64</tt> - 64-bit vector is 64-bit aligned</li>
894 <li><tt>v128:128:128</tt> - 128-bit vector is 128-bit aligned</li>
895 <li><tt>a0:0:1</tt> - aggregates are 8-bit aligned</li>
896</ul>
897<p>When llvm is determining the alignment for a given type, it uses the
898following rules:
899<ol>
900 <li>If the type sought is an exact match for one of the specifications, that
901 specification is used.</li>
902 <li>If no match is found, and the type sought is an integer type, then the
903 smallest integer type that is larger than the bitwidth of the sought type is
904 used. If none of the specifications are larger than the bitwidth then the the
905 largest integer type is used. For example, given the default specifications
906 above, the i7 type will use the alignment of i8 (next largest) while both
907 i65 and i256 will use the alignment of i64 (largest specified).</li>
908 <li>If no match is found, and the type sought is a vector type, then the
909 largest vector type that is smaller than the sought vector type will be used
910 as a fall back. This happens because <128 x double> can be implemented in
911 terms of 64 <2 x double>, for example.</li>
912</ol>
913</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]914
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]915<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]916<div class="doc_section"> <a name="typesystem">Type System</a> </div>
917<!-- *********************************************************************** -->
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]918
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]919<div class="doc_text">
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]920
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]921<p>The LLVM type system is one of the most important features of the
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]922intermediate representation. Being typed enables a number of
923optimizations to be performed on the IR directly, without having to do
924extra analyses on the side before the transformation. A strong type
925system makes it easier to read the generated code and enables novel
926analyses and transformations that are not feasible to perform on normal
927three address code representations.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]928
929</div>
930
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]931<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]932<div class="doc_subsection"> <a name="t_primitive">Primitive Types</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]933<div class="doc_text">
John Criswell4457dc92004-04-09 16:48:45 +0000[diff] [blame]934<p>The primitive types are the fundamental building blocks of the LLVM
Chris Lattnerd4f6b172005-03-07 22:13:59 +0000[diff] [blame]935system. The current set of primitive types is as follows:</p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]936
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]937<table class="layout">
938 <tr class="layout">
939 <td class="left">
940 <table>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]941 <tbody>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]942 <tr><th>Type</th><th>Description</th></tr>
Duncan Sands8036ca42007-03-30 12:22:09 +0000[diff] [blame]943 <tr><td><tt><a name="t_void">void</a></tt></td><td>No value</td></tr>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]944 <tr><td><tt>label</tt></td><td>Branch destination</td></tr>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]945 </tbody>
946 </table>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]947 </td>
948 <td class="right">
949 <table>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]950 <tbody>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]951 <tr><th>Type</th><th>Description</th></tr>
Reid Spencer2b916312007-05-16 18:44:01 +0000[diff] [blame]952 <tr><td><tt>float</tt></td><td>32-bit floating point value</td></tr>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]953 <tr><td><tt>double</tt></td><td>64-bit floating point value</td></tr>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]954 </tbody>
955 </table>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]956 </td>
957 </tr>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]958</table>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]959</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]960
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]961<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]962<div class="doc_subsubsection"> <a name="t_classifications">Type
963Classifications</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]964<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]965<p>These different primitive types fall into a few useful
966classifications:</p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]967
968<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]969 <tbody>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]970 <tr><th>Classification</th><th>Types</th></tr>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]971 <tr>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]972 <td><a name="t_integer">integer</a></td>
Reid Spencer2b916312007-05-16 18:44:01 +0000[diff] [blame]973 <td><tt>i1, i2, i3, ... i8, ... i16, ... i32, ... i64, ... </tt></td>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]974 </tr>
975 <tr>
976 <td><a name="t_floating">floating point</a></td>
977 <td><tt>float, double</tt></td>
978 </tr>
979 <tr>
980 <td><a name="t_firstclass">first class</a></td>
Reid Spencer2b916312007-05-16 18:44:01 +0000[diff] [blame]981 <td><tt>i1, ..., float, double, <br/>
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]982 <a href="#t_pointer">pointer</a>,<a href="#t_vector">vector</a></tt>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]983 </td>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]984 </tr>
985 </tbody>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]986</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]987
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]988<p>The <a href="#t_firstclass">first class</a> types are perhaps the
989most important. Values of these types are the only ones which can be
990produced by instructions, passed as arguments, or used as operands to
991instructions. This means that all structures and arrays must be
992manipulated either by pointer or by component.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]993</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]994
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]995<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]996<div class="doc_subsection"> <a name="t_derived">Derived Types</a> </div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]997
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]998<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]999
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1000<p>The real power in LLVM comes from the derived types in the system.
1001This is what allows a programmer to represent arrays, functions,
1002pointers, and other useful types. Note that these derived types may be
1003recursive: For example, it is possible to have a two dimensional array.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1004
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1005</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1006
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1007<!-- _______________________________________________________________________ -->
Reid Spencer2b916312007-05-16 18:44:01 +0000[diff] [blame]1008<div class="doc_subsubsection"> <a name="t_integer">Integer Type</a> </div>
1009
1010<div class="doc_text">
1011
1012<h5>Overview:</h5>
1013<p>The integer type is a very simple derived type that simply specifies an
1014arbitrary bit width for the integer type desired. Any bit width from 1 bit to
10152^23-1 (about 8 million) can be specified.</p>
1016
1017<h5>Syntax:</h5>
1018
1019<pre>
1020 iN
1021</pre>
1022
1023<p>The number of bits the integer will occupy is specified by the <tt>N</tt>
1024value.</p>
1025
1026<h5>Examples:</h5>
1027<table class="layout">
1028 <tr class="layout">
1029 <td class="left">
1030 <tt>i1</tt><br/>
1031 <tt>i4</tt><br/>
1032 <tt>i8</tt><br/>
1033 <tt>i16</tt><br/>
1034 <tt>i32</tt><br/>
1035 <tt>i42</tt><br/>
1036 <tt>i64</tt><br/>
1037 <tt>i1942652</tt><br/>
1038 </td>
1039 <td class="left">
1040 A boolean integer of 1 bit<br/>
1041 A nibble sized integer of 4 bits.<br/>
1042 A byte sized integer of 8 bits.<br/>
1043 A half word sized integer of 16 bits.<br/>
1044 A word sized integer of 32 bits.<br/>
1045 An integer whose bit width is the answer. <br/>
1046 A double word sized integer of 64 bits.<br/>
1047 A really big integer of over 1 million bits.<br/>
1048 </td>
1049 </tr>
1050</table>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1051</div>
Reid Spencer2b916312007-05-16 18:44:01 +0000[diff] [blame]1052
1053<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1054<div class="doc_subsubsection"> <a name="t_array">Array Type</a> </div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1055
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1056<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1057
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1058<h5>Overview:</h5>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1059
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1060<p>The array type is a very simple derived type that arranges elements
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1061sequentially in memory. The array type requires a size (number of
1062elements) and an underlying data type.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1063
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1064<h5>Syntax:</h5>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1065
1066<pre>
1067 [&lt;# elements&gt; x &lt;elementtype&gt;]
1068</pre>
1069
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]1070<p>The number of elements is a constant integer value; elementtype may
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1071be any type with a size.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1072
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1073<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1074<table class="layout">
1075 <tr class="layout">
1076 <td class="left">
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1077 <tt>[40 x i32 ]</tt><br/>
1078 <tt>[41 x i32 ]</tt><br/>
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]1079 <tt>[40 x i8]</tt><br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1080 </td>
1081 <td class="left">
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]1082 Array of 40 32-bit integer values.<br/>
1083 Array of 41 32-bit integer values.<br/>
1084 Array of 40 8-bit integer values.<br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1085 </td>
1086 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1087</table>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1088<p>Here are some examples of multidimensional arrays:</p>
1089<table class="layout">
1090 <tr class="layout">
1091 <td class="left">
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1092 <tt>[3 x [4 x i32]]</tt><br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1093 <tt>[12 x [10 x float]]</tt><br/>
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]1094 <tt>[2 x [3 x [4 x i16]]]</tt><br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1095 </td>
1096 <td class="left">
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]1097 3x4 array of 32-bit integer values.<br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1098 12x10 array of single precision floating point values.<br/>
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]1099 2x3x4 array of 16-bit integer values.<br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1100 </td>
1101 </tr>
1102</table>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]1103
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]1104<p>Note that 'variable sized arrays' can be implemented in LLVM with a zero
1105length array. Normally, accesses past the end of an array are undefined in
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]1106LLVM (e.g. it is illegal to access the 5th element of a 3 element array).
1107As a special case, however, zero length arrays are recognized to be variable
1108length. This allows implementation of 'pascal style arrays' with the LLVM
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1109type "{ i32, [0 x float]}", for example.</p>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]1110
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1111</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1112
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1113<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1114<div class="doc_subsubsection"> <a name="t_function">Function Type</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1115<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1116<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1117<p>The function type can be thought of as a function signature. It
1118consists of a return type and a list of formal parameter types.
John Criswell009900b2003-11-25 21:45:46 +0000[diff] [blame]1119Function types are usually used to build virtual function tables
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1120(which are structures of pointers to functions), for indirect function
1121calls, and when defining a function.</p>
John Criswell009900b2003-11-25 21:45:46 +0000[diff] [blame]1122<p>
1123The return type of a function type cannot be an aggregate type.
1124</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1125<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1126<pre> &lt;returntype&gt; (&lt;parameter list&gt;)<br></pre>
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]1127<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]1128specifiers. Optionally, the parameter list may include a type <tt>...</tt>,
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]1129which indicates that the function takes a variable number of arguments.
1130Variable argument functions can access their arguments with the <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1131 href="#int_varargs">variable argument handling intrinsic</a> functions.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1132<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1133<table class="layout">
1134 <tr class="layout">
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]1135 <td class="left"><tt>i32 (i32)</tt></td>
1136 <td class="left">function taking an <tt>i32</tt>, returning an <tt>i32</tt>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1137 </td>
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]1138 </tr><tr class="layout">
Reid Spencer7bf214d2007-01-15 18:28:34 +0000[diff] [blame]1139 <td class="left"><tt>float&nbsp;(i16&nbsp;sext,&nbsp;i32&nbsp;*)&nbsp;*
Reid Spencerf17a0b72006-12-31 07:20:23 +0000[diff] [blame]1140 </tt></td>
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]1141 <td class="left"><a href="#t_pointer">Pointer</a> to a function that takes
1142 an <tt>i16</tt> that should be sign extended and a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1143 <a href="#t_pointer">pointer</a> to <tt>i32</tt>, returning
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]1144 <tt>float</tt>.
1145 </td>
1146 </tr><tr class="layout">
1147 <td class="left"><tt>i32 (i8*, ...)</tt></td>
1148 <td class="left">A vararg function that takes at least one
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]1149 <a href="#t_pointer">pointer</a> to <tt>i8 </tt> (char in C),
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]1150 which returns an integer. This is the signature for <tt>printf</tt> in
1151 LLVM.
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1152 </td>
1153 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1154</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1155
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1156</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1157<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1158<div class="doc_subsubsection"> <a name="t_struct">Structure Type</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1159<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1160<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1161<p>The structure type is used to represent a collection of data members
1162together in memory. The packing of the field types is defined to match
1163the ABI of the underlying processor. The elements of a structure may
1164be any type that has a size.</p>
1165<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
1166and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
1167field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
1168instruction.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1169<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1170<pre> { &lt;type list&gt; }<br></pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1171<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1172<table class="layout">
1173 <tr class="layout">
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]1174 <td class="left"><tt>{ i32, i32, i32 }</tt></td>
1175 <td class="left">A triple of three <tt>i32</tt> values</td>
1176 </tr><tr class="layout">
1177 <td class="left"><tt>{&nbsp;float,&nbsp;i32&nbsp;(i32)&nbsp;*&nbsp;}</tt></td>
1178 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1179 second element is a <a href="#t_pointer">pointer</a> to a
1180 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1181 an <tt>i32</tt>.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1182 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1183</table>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1184</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1185
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1186<!-- _______________________________________________________________________ -->
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]1187<div class="doc_subsubsection"> <a name="t_pstruct">Packed Structure Type</a>
1188</div>
1189<div class="doc_text">
1190<h5>Overview:</h5>
1191<p>The packed structure type is used to represent a collection of data members
1192together in memory. There is no padding between fields. Further, the alignment
1193of a packed structure is 1 byte. The elements of a packed structure may
1194be any type that has a size.</p>
1195<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
1196and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
1197field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
1198instruction.</p>
1199<h5>Syntax:</h5>
1200<pre> &lt; { &lt;type list&gt; } &gt; <br></pre>
1201<h5>Examples:</h5>
1202<table class="layout">
1203 <tr class="layout">
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]1204 <td class="left"><tt>&lt; { i32, i32, i32 } &gt;</tt></td>
1205 <td class="left">A triple of three <tt>i32</tt> values</td>
1206 </tr><tr class="layout">
1207 <td class="left"><tt>&lt;&nbsp;{&nbsp;float,&nbsp;i32&nbsp;(i32)&nbsp;*&nbsp;}&nbsp;&gt;</tt></td>
1208 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1209 second element is a <a href="#t_pointer">pointer</a> to a
1210 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1211 an <tt>i32</tt>.</td>
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]1212 </tr>
1213</table>
1214</div>
1215
1216<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1217<div class="doc_subsubsection"> <a name="t_pointer">Pointer Type</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1218<div class="doc_text">
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1219<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1220<p>As in many languages, the pointer type represents a pointer or
1221reference to another object, which must live in memory.</p>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1222<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1223<pre> &lt;type&gt; *<br></pre>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1224<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1225<table class="layout">
1226 <tr class="layout">
1227 <td class="left">
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1228 <tt>[4x i32]*</tt><br/>
1229 <tt>i32 (i32 *) *</tt><br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1230 </td>
1231 <td class="left">
1232 A <a href="#t_pointer">pointer</a> to <a href="#t_array">array</a> of
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1233 four <tt>i32</tt> values<br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1234 A <a href="#t_pointer">pointer</a> to a <a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1235 href="#t_function">function</a> that takes an <tt>i32*</tt>, returning an
1236 <tt>i32</tt>.<br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1237 </td>
1238 </tr>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1239</table>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1240</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1241
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1242<!-- _______________________________________________________________________ -->
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1243<div class="doc_subsubsection"> <a name="t_vector">Vector Type</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1244<div class="doc_text">
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1245
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1246<h5>Overview:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1247
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1248<p>A vector type is a simple derived type that represents a vector
1249of elements. Vector types are used when multiple primitive data
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1250are operated in parallel using a single instruction (SIMD).
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1251A vector type requires a size (number of
Chris Lattnerb8d172f2005-11-10 01:44:22 +0000[diff] [blame]1252elements) and an underlying primitive data type. Vectors must have a power
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1253of two length (1, 2, 4, 8, 16 ...). Vector types are
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1254considered <a href="#t_firstclass">first class</a>.</p>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1255
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1256<h5>Syntax:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1257
1258<pre>
1259 &lt; &lt;# elements&gt; x &lt;elementtype&gt; &gt;
1260</pre>
1261
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]1262<p>The number of elements is a constant integer value; elementtype may
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]1263be any integer or floating point type.</p>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1264
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1265<h5>Examples:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1266
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1267<table class="layout">
1268 <tr class="layout">
1269 <td class="left">
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1270 <tt>&lt;4 x i32&gt;</tt><br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1271 <tt>&lt;8 x float&gt;</tt><br/>
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]1272 <tt>&lt;2 x i64&gt;</tt><br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1273 </td>
1274 <td class="left">
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1275 Vector of 4 32-bit integer values.<br/>
1276 Vector of 8 floating-point values.<br/>
1277 Vector of 2 64-bit integer values.<br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1278 </td>
1279 </tr>
1280</table>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1281</div>
1282
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1283<!-- _______________________________________________________________________ -->
1284<div class="doc_subsubsection"> <a name="t_opaque">Opaque Type</a> </div>
1285<div class="doc_text">
1286
1287<h5>Overview:</h5>
1288
1289<p>Opaque types are used to represent unknown types in the system. This
1290corresponds (for example) to the C notion of a foward declared structure type.
1291In LLVM, opaque types can eventually be resolved to any type (not just a
1292structure type).</p>
1293
1294<h5>Syntax:</h5>
1295
1296<pre>
1297 opaque
1298</pre>
1299
1300<h5>Examples:</h5>
1301
1302<table class="layout">
1303 <tr class="layout">
1304 <td class="left">
1305 <tt>opaque</tt>
1306 </td>
1307 <td class="left">
1308 An opaque type.<br/>
1309 </td>
1310 </tr>
1311</table>
1312</div>
1313
1314
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1315<!-- *********************************************************************** -->
1316<div class="doc_section"> <a name="constants">Constants</a> </div>
1317<!-- *********************************************************************** -->
1318
1319<div class="doc_text">
1320
1321<p>LLVM has several different basic types of constants. This section describes
1322them all and their syntax.</p>
1323
1324</div>
1325
1326<!-- ======================================================================= -->
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]1327<div class="doc_subsection"><a name="simpleconstants">Simple Constants</a></div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1328
1329<div class="doc_text">
1330
1331<dl>
1332 <dt><b>Boolean constants</b></dt>
1333
1334 <dd>The two strings '<tt>true</tt>' and '<tt>false</tt>' are both valid
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]1335 constants of the <tt><a href="#t_primitive">i1</a></tt> type.
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1336 </dd>
1337
1338 <dt><b>Integer constants</b></dt>
1339
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]1340 <dd>Standard integers (such as '4') are constants of the <a
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]1341 href="#t_integer">integer</a> type. Negative numbers may be used with
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1342 integer types.
1343 </dd>
1344
1345 <dt><b>Floating point constants</b></dt>
1346
1347 <dd>Floating point constants use standard decimal notation (e.g. 123.421),
1348 exponential notation (e.g. 1.23421e+2), or a more precise hexadecimal
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1349 notation (see below). Floating point constants must have a <a
1350 href="#t_floating">floating point</a> type. </dd>
1351
1352 <dt><b>Null pointer constants</b></dt>
1353
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1354 <dd>The identifier '<tt>null</tt>' is recognized as a null pointer constant
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1355 and must be of <a href="#t_pointer">pointer type</a>.</dd>
1356
1357</dl>
1358
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1359<p>The one non-intuitive notation for constants is the optional hexadecimal form
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1360of floating point constants. For example, the form '<tt>double
13610x432ff973cafa8000</tt>' is equivalent to (but harder to read than) '<tt>double
13624.5e+15</tt>'. The only time hexadecimal floating point constants are required
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]1363(and the only time that they are generated by the disassembler) is when a
1364floating point constant must be emitted but it cannot be represented as a
1365decimal floating point number. For example, NaN's, infinities, and other
1366special values are represented in their IEEE hexadecimal format so that
1367assembly and disassembly do not cause any bits to change in the constants.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1368
1369</div>
1370
1371<!-- ======================================================================= -->
1372<div class="doc_subsection"><a name="aggregateconstants">Aggregate Constants</a>
1373</div>
1374
1375<div class="doc_text">
Chris Lattnerd4f6b172005-03-07 22:13:59 +0000[diff] [blame]1376<p>Aggregate constants arise from aggregation of simple constants
1377and smaller aggregate constants.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1378
1379<dl>
1380 <dt><b>Structure constants</b></dt>
1381
1382 <dd>Structure constants are represented with notation similar to structure
1383 type definitions (a comma separated list of elements, surrounded by braces
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1384 (<tt>{}</tt>)). For example: "<tt>{ i32 4, float 17.0, i32* %G }</tt>",
1385 where "<tt>%G</tt>" is declared as "<tt>%G = external global i32</tt>". Structure constants
Chris Lattnerd4f6b172005-03-07 22:13:59 +0000[diff] [blame]1386 must have <a href="#t_struct">structure type</a>, and the number and
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1387 types of elements must match those specified by the type.
1388 </dd>
1389
1390 <dt><b>Array constants</b></dt>
1391
1392 <dd>Array constants are represented with notation similar to array type
1393 definitions (a comma separated list of elements, surrounded by square brackets
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1394 (<tt>[]</tt>)). For example: "<tt>[ i32 42, i32 11, i32 74 ]</tt>". Array
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1395 constants must have <a href="#t_array">array type</a>, and the number and
1396 types of elements must match those specified by the type.
1397 </dd>
1398
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1399 <dt><b>Vector constants</b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1400
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1401 <dd>Vector constants are represented with notation similar to vector type
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1402 definitions (a comma separated list of elements, surrounded by
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1403 less-than/greater-than's (<tt>&lt;&gt;</tt>)). For example: "<tt>&lt; i32 42,
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]1404 i32 11, i32 74, i32 100 &gt;</tt>". Vector constants must have <a
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1405 href="#t_vector">vector type</a>, and the number and types of elements must
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1406 match those specified by the type.
1407 </dd>
1408
1409 <dt><b>Zero initialization</b></dt>
1410
1411 <dd>The string '<tt>zeroinitializer</tt>' can be used to zero initialize a
1412 value to zero of <em>any</em> type, including scalar and aggregate types.
1413 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]1414 large arrays) and is always exactly equivalent to using explicit zero
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1415 initializers.
1416 </dd>
1417</dl>
1418
1419</div>
1420
1421<!-- ======================================================================= -->
1422<div class="doc_subsection">
1423 <a name="globalconstants">Global Variable and Function Addresses</a>
1424</div>
1425
1426<div class="doc_text">
1427
1428<p>The addresses of <a href="#globalvars">global variables</a> and <a
1429href="#functionstructure">functions</a> are always implicitly valid (link-time)
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1430constants. These constants are explicitly referenced when the <a
1431href="#identifiers">identifier for the global</a> is used and always have <a
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1432href="#t_pointer">pointer</a> type. For example, the following is a legal LLVM
1433file:</p>
1434
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1435<div class="doc_code">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1436<pre>
Chris Lattnera18a4242007-06-06 18:28:13 +0000[diff] [blame]1437@X = global i32 17
1438@Y = global i32 42
1439@Z = global [2 x i32*] [ i32* @X, i32* @Y ]
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1440</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1441</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1442
1443</div>
1444
1445<!-- ======================================================================= -->
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]1446<div class="doc_subsection"><a name="undefvalues">Undefined Values</a></div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1447<div class="doc_text">
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]1448 <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]1449 no specific value. Undefined values may be of any type and be used anywhere
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]1450 a constant is permitted.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1451
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]1452 <p>Undefined values indicate to the compiler that the program is well defined
1453 no matter what value is used, giving the compiler more freedom to optimize.
1454 </p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1455</div>
1456
1457<!-- ======================================================================= -->
1458<div class="doc_subsection"><a name="constantexprs">Constant Expressions</a>
1459</div>
1460
1461<div class="doc_text">
1462
1463<p>Constant expressions are used to allow expressions involving other constants
1464to be used as constants. Constant expressions may be of any <a
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]1465href="#t_firstclass">first class</a> type and may involve any LLVM operation
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1466that does not have side effects (e.g. load and call are not supported). The
1467following is the syntax for constant expressions:</p>
1468
1469<dl>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1470 <dt><b><tt>trunc ( CST to TYPE )</tt></b></dt>
1471 <dd>Truncate a constant to another type. The bit size of CST must be larger
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]1472 than the bit size of TYPE. Both types must be integers.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1473
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1474 <dt><b><tt>zext ( CST to TYPE )</tt></b></dt>
1475 <dd>Zero extend a constant to another type. The bit size of CST must be
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]1476 smaller or equal to the bit size of TYPE. Both types must be integers.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1477
1478 <dt><b><tt>sext ( CST to TYPE )</tt></b></dt>
1479 <dd>Sign extend a constant to another type. The bit size of CST must be
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]1480 smaller or equal to the bit size of TYPE. Both types must be integers.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1481
1482 <dt><b><tt>fptrunc ( CST to TYPE )</tt></b></dt>
1483 <dd>Truncate a floating point constant to another floating point type. The
1484 size of CST must be larger than the size of TYPE. Both types must be
1485 floating point.</dd>
1486
1487 <dt><b><tt>fpext ( CST to TYPE )</tt></b></dt>
1488 <dd>Floating point extend a constant to another type. The size of CST must be
1489 smaller or equal to the size of TYPE. Both types must be floating point.</dd>
1490
1491 <dt><b><tt>fp2uint ( CST to TYPE )</tt></b></dt>
1492 <dd>Convert a floating point constant to the corresponding unsigned integer
1493 constant. TYPE must be an integer type. CST must be floating point. If the
1494 value won't fit in the integer type, the results are undefined.</dd>
1495
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]1496 <dt><b><tt>fptosi ( CST to TYPE )</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1497 <dd>Convert a floating point constant to the corresponding signed integer
1498 constant. TYPE must be an integer type. CST must be floating point. If the
1499 value won't fit in the integer type, the results are undefined.</dd>
1500
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]1501 <dt><b><tt>uitofp ( CST to TYPE )</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1502 <dd>Convert an unsigned integer constant to the corresponding floating point
1503 constant. TYPE must be floating point. CST must be of integer type. If the
Jeff Cohencb757312007-04-22 14:56:37 +0000[diff] [blame]1504 value won't fit in the floating point type, the results are undefined.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1505
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]1506 <dt><b><tt>sitofp ( CST to TYPE )</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1507 <dd>Convert a signed integer constant to the corresponding floating point
1508 constant. TYPE must be floating point. CST must be of integer type. If the
Jeff Cohencb757312007-04-22 14:56:37 +0000[diff] [blame]1509 value won't fit in the floating point type, the results are undefined.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1510
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]1511 <dt><b><tt>ptrtoint ( CST to TYPE )</tt></b></dt>
1512 <dd>Convert a pointer typed constant to the corresponding integer constant
1513 TYPE must be an integer type. CST must be of pointer type. The CST value is
1514 zero extended, truncated, or unchanged to make it fit in TYPE.</dd>
1515
1516 <dt><b><tt>inttoptr ( CST to TYPE )</tt></b></dt>
1517 <dd>Convert a integer constant to a pointer constant. TYPE must be a
1518 pointer type. CST must be of integer type. The CST value is zero extended,
1519 truncated, or unchanged to make it fit in a pointer size. This one is
1520 <i>really</i> dangerous!</dd>
1521
1522 <dt><b><tt>bitcast ( CST to TYPE )</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1523 <dd>Convert a constant, CST, to another TYPE. The size of CST and TYPE must be
1524 identical (same number of bits). The conversion is done as if the CST value
1525 was stored to memory and read back as TYPE. In other words, no bits change
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]1526 with this operator, just the type. This can be used for conversion of
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1527 vector types to any other type, as long as they have the same bit width. For
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]1528 pointers it is only valid to cast to another pointer type.
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1529 </dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1530
1531 <dt><b><tt>getelementptr ( CSTPTR, IDX0, IDX1, ... )</tt></b></dt>
1532
1533 <dd>Perform the <a href="#i_getelementptr">getelementptr operation</a> on
1534 constants. As with the <a href="#i_getelementptr">getelementptr</a>
1535 instruction, the index list may have zero or more indexes, which are required
1536 to make sense for the type of "CSTPTR".</dd>
1537
Robert Bocchino9fbe1452006-01-10 19:31:34 +0000[diff] [blame]1538 <dt><b><tt>select ( COND, VAL1, VAL2 )</tt></b></dt>
1539
1540 <dd>Perform the <a href="#i_select">select operation</a> on
Reid Spencer01c42592006-12-04 19:23:19 +0000[diff] [blame]1541 constants.</dd>
1542
1543 <dt><b><tt>icmp COND ( VAL1, VAL2 )</tt></b></dt>
1544 <dd>Performs the <a href="#i_icmp">icmp operation</a> on constants.</dd>
1545
1546 <dt><b><tt>fcmp COND ( VAL1, VAL2 )</tt></b></dt>
1547 <dd>Performs the <a href="#i_fcmp">fcmp operation</a> on constants.</dd>
Robert Bocchino9fbe1452006-01-10 19:31:34 +0000[diff] [blame]1548
1549 <dt><b><tt>extractelement ( VAL, IDX )</tt></b></dt>
1550
1551 <dd>Perform the <a href="#i_extractelement">extractelement
1552 operation</a> on constants.
1553
Robert Bocchino05ccd702006-01-15 20:48:27 +0000[diff] [blame]1554 <dt><b><tt>insertelement ( VAL, ELT, IDX )</tt></b></dt>
1555
1556 <dd>Perform the <a href="#i_insertelement">insertelement
Reid Spencer01c42592006-12-04 19:23:19 +0000[diff] [blame]1557 operation</a> on constants.</dd>
Robert Bocchino05ccd702006-01-15 20:48:27 +0000[diff] [blame]1558
Chris Lattnerc1989542006-04-08 00:13:41 +0000[diff] [blame]1559
1560 <dt><b><tt>shufflevector ( VEC1, VEC2, IDXMASK )</tt></b></dt>
1561
1562 <dd>Perform the <a href="#i_shufflevector">shufflevector
Reid Spencer01c42592006-12-04 19:23:19 +0000[diff] [blame]1563 operation</a> on constants.</dd>
Chris Lattnerc1989542006-04-08 00:13:41 +0000[diff] [blame]1564
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1565 <dt><b><tt>OPCODE ( LHS, RHS )</tt></b></dt>
1566
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]1567 <dd>Perform the specified operation of the LHS and RHS constants. OPCODE may
1568 be any of the <a href="#binaryops">binary</a> or <a href="#bitwiseops">bitwise
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1569 binary</a> operations. The constraints on operands are the same as those for
1570 the corresponding instruction (e.g. no bitwise operations on floating point
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]1571 values are allowed).</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1572</dl>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1573</div>
Chris Lattner9ee5d222004-03-08 16:49:10 +0000[diff] [blame]1574
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1575<!-- *********************************************************************** -->
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]1576<div class="doc_section"> <a name="othervalues">Other Values</a> </div>
1577<!-- *********************************************************************** -->
1578
1579<!-- ======================================================================= -->
1580<div class="doc_subsection">
1581<a name="inlineasm">Inline Assembler Expressions</a>
1582</div>
1583
1584<div class="doc_text">
1585
1586<p>
1587LLVM supports inline assembler expressions (as opposed to <a href="#moduleasm">
1588Module-Level Inline Assembly</a>) through the use of a special value. This
1589value represents the inline assembler as a string (containing the instructions
1590to emit), a list of operand constraints (stored as a string), and a flag that
1591indicates whether or not the inline asm expression has side effects. An example
1592inline assembler expression is:
1593</p>
1594
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1595<div class="doc_code">
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]1596<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1597i32 (i32) asm "bswap $0", "=r,r"
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]1598</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1599</div>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]1600
1601<p>
1602Inline assembler expressions may <b>only</b> be used as the callee operand of
1603a <a href="#i_call"><tt>call</tt> instruction</a>. Thus, typically we have:
1604</p>
1605
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1606<div class="doc_code">
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]1607<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1608%X = call i32 asm "<a href="#int_bswap">bswap</a> $0", "=r,r"(i32 %Y)
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]1609</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1610</div>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]1611
1612<p>
1613Inline asms with side effects not visible in the constraint list must be marked
1614as having side effects. This is done through the use of the
1615'<tt>sideeffect</tt>' keyword, like so:
1616</p>
1617
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1618<div class="doc_code">
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]1619<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1620call void asm sideeffect "eieio", ""()
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]1621</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1622</div>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]1623
1624<p>TODO: The format of the asm and constraints string still need to be
1625documented here. Constraints on what can be done (e.g. duplication, moving, etc
1626need to be documented).
1627</p>
1628
1629</div>
1630
1631<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1632<div class="doc_section"> <a name="instref">Instruction Reference</a> </div>
1633<!-- *********************************************************************** -->
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1634
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1635<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1636
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1637<p>The LLVM instruction set consists of several different
1638classifications of instructions: <a href="#terminators">terminator
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]1639instructions</a>, <a href="#binaryops">binary instructions</a>,
1640<a href="#bitwiseops">bitwise binary instructions</a>, <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1641 href="#memoryops">memory instructions</a>, and <a href="#otherops">other
1642instructions</a>.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1643
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1644</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1645
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1646<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1647<div class="doc_subsection"> <a name="terminators">Terminator
1648Instructions</a> </div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1649
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1650<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1651
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1652<p>As mentioned <a href="#functionstructure">previously</a>, every
1653basic block in a program ends with a "Terminator" instruction, which
1654indicates which block should be executed after the current block is
1655finished. These terminator instructions typically yield a '<tt>void</tt>'
1656value: they produce control flow, not values (the one exception being
1657the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction).</p>
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1658<p>There are six different terminator instructions: the '<a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1659 href="#i_ret"><tt>ret</tt></a>' instruction, the '<a href="#i_br"><tt>br</tt></a>'
1660instruction, the '<a href="#i_switch"><tt>switch</tt></a>' instruction,
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1661the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction, the '<a
1662 href="#i_unwind"><tt>unwind</tt></a>' instruction, and the '<a
1663 href="#i_unreachable"><tt>unreachable</tt></a>' instruction.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1664
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1665</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1666
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1667<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1668<div class="doc_subsubsection"> <a name="i_ret">'<tt>ret</tt>'
1669Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1670<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1671<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1672<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]1673 ret void <i>; Return from void function</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1674</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1675<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1676<p>The '<tt>ret</tt>' instruction is used to return control flow (and a
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]1677value) from a function back to the caller.</p>
John Criswell4457dc92004-04-09 16:48:45 +0000[diff] [blame]1678<p>There are two forms of the '<tt>ret</tt>' instruction: one that
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1679returns a value and then causes control flow, and one that just causes
1680control flow to occur.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1681<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1682<p>The '<tt>ret</tt>' instruction may return any '<a
1683 href="#t_firstclass">first class</a>' type. Notice that a function is
1684not <a href="#wellformed">well formed</a> if there exists a '<tt>ret</tt>'
1685instruction inside of the function that returns a value that does not
1686match the return type of the function.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1687<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1688<p>When the '<tt>ret</tt>' instruction is executed, control flow
1689returns back to the calling function's context. If the caller is a "<a
John Criswellfa081872004-06-25 15:16:57 +0000[diff] [blame]1690 href="#i_call"><tt>call</tt></a>" instruction, execution continues at
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1691the instruction after the call. If the caller was an "<a
1692 href="#i_invoke"><tt>invoke</tt></a>" instruction, execution continues
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]1693at the beginning of the "normal" destination block. If the instruction
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1694returns a value, that value shall set the call or invoke instruction's
1695return value.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1696<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1697<pre> ret i32 5 <i>; Return an integer value of 5</i>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1698 ret void <i>; Return from a void function</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1699</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1700</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1701<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1702<div class="doc_subsubsection"> <a name="i_br">'<tt>br</tt>' Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1703<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1704<h5>Syntax:</h5>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]1705<pre> br i1 &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]1706</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1707<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1708<p>The '<tt>br</tt>' instruction is used to cause control flow to
1709transfer to a different basic block in the current function. There are
1710two forms of this instruction, corresponding to a conditional branch
1711and an unconditional branch.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1712<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1713<p>The conditional branch form of the '<tt>br</tt>' instruction takes a
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]1714single '<tt>i1</tt>' value and two '<tt>label</tt>' values. The
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1715unconditional form of the '<tt>br</tt>' instruction takes a single
1716'<tt>label</tt>' value as a target.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1717<h5>Semantics:</h5>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]1718<p>Upon execution of a conditional '<tt>br</tt>' instruction, the '<tt>i1</tt>'
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1719argument is evaluated. If the value is <tt>true</tt>, control flows
1720to the '<tt>iftrue</tt>' <tt>label</tt> argument. If "cond" is <tt>false</tt>,
1721control flows to the '<tt>iffalse</tt>' <tt>label</tt> argument.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1722<h5>Example:</h5>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]1723<pre>Test:<br> %cond = <a href="#i_icmp">icmp</a> eq, i32 %a, %b<br> br i1 %cond, label %IfEqual, label %IfUnequal<br>IfEqual:<br> <a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1724 href="#i_ret">ret</a> i32 1<br>IfUnequal:<br> <a href="#i_ret">ret</a> i32 0<br></pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1725</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1726<!-- _______________________________________________________________________ -->
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1727<div class="doc_subsubsection">
1728 <a name="i_switch">'<tt>switch</tt>' Instruction</a>
1729</div>
1730
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1731<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1732<h5>Syntax:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1733
1734<pre>
1735 switch &lt;intty&gt; &lt;value&gt;, label &lt;defaultdest&gt; [ &lt;intty&gt; &lt;val&gt;, label &lt;dest&gt; ... ]
1736</pre>
1737
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1738<h5>Overview:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1739
1740<p>The '<tt>switch</tt>' instruction is used to transfer control flow to one of
1741several different places. It is a generalization of the '<tt>br</tt>'
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1742instruction, allowing a branch to occur to one of many possible
1743destinations.</p>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1744
1745
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1746<h5>Arguments:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1747
1748<p>The '<tt>switch</tt>' instruction uses three parameters: an integer
1749comparison value '<tt>value</tt>', a default '<tt>label</tt>' destination, and
1750an array of pairs of comparison value constants and '<tt>label</tt>'s. The
1751table is not allowed to contain duplicate constant entries.</p>
1752
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1753<h5>Semantics:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1754
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1755<p>The <tt>switch</tt> instruction specifies a table of values and
1756destinations. When the '<tt>switch</tt>' instruction is executed, this
John Criswell84114752004-06-25 16:05:06 +0000[diff] [blame]1757table is searched for the given value. If the value is found, control flow is
1758transfered to the corresponding destination; otherwise, control flow is
1759transfered to the default destination.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1760
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1761<h5>Implementation:</h5>
1762
1763<p>Depending on properties of the target machine and the particular
1764<tt>switch</tt> instruction, this instruction may be code generated in different
John Criswell84114752004-06-25 16:05:06 +0000[diff] [blame]1765ways. For example, it could be generated as a series of chained conditional
1766branches or with a lookup table.</p>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1767
1768<h5>Example:</h5>
1769
1770<pre>
1771 <i>; Emulate a conditional br instruction</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]1772 %Val = <a href="#i_zext">zext</a> i1 %value to i32
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1773 switch i32 %Val, label %truedest [i32 0, label %falsedest ]
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1774
1775 <i>; Emulate an unconditional br instruction</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1776 switch i32 0, label %dest [ ]
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1777
1778 <i>; Implement a jump table:</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1779 switch i32 %val, label %otherwise [ i32 0, label %onzero
1780 i32 1, label %onone
1781 i32 2, label %ontwo ]
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1782</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1783</div>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1784
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1785<!-- _______________________________________________________________________ -->
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1786<div class="doc_subsubsection">
1787 <a name="i_invoke">'<tt>invoke</tt>' Instruction</a>
1788</div>
1789
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1790<div class="doc_text">
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1791
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1792<h5>Syntax:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1793
1794<pre>
1795 &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]1796 to label &lt;normal label&gt; unwind label &lt;exception label&gt;
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1797</pre>
1798
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]1799<h5>Overview:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1800
1801<p>The '<tt>invoke</tt>' instruction causes control to transfer to a specified
1802function, with the possibility of control flow transfer to either the
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]1803'<tt>normal</tt>' label or the
1804'<tt>exception</tt>' label. If the callee function returns with the
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1805"<tt><a href="#i_ret">ret</a></tt>" instruction, control flow will return to the
1806"normal" label. If the callee (or any indirect callees) returns with the "<a
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]1807href="#i_unwind"><tt>unwind</tt></a>" instruction, control is interrupted and
1808continued at the dynamically nearest "exception" label.</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1809
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1810<h5>Arguments:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1811
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1812<p>This instruction requires several arguments:</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1813
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1814<ol>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1815 <li>
Duncan Sands8036ca42007-03-30 12:22:09 +0000[diff] [blame]1816 The optional "cconv" marker indicates which <a href="#callingconv">calling
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1817 convention</a> the call should use. If none is specified, the call defaults
1818 to using C calling conventions.
1819 </li>
1820 <li>'<tt>ptr to function ty</tt>': shall be the signature of the pointer to
1821 function value being invoked. In most cases, this is a direct function
1822 invocation, but indirect <tt>invoke</tt>s are just as possible, branching off
1823 an arbitrary pointer to function value.
1824 </li>
1825
1826 <li>'<tt>function ptr val</tt>': An LLVM value containing a pointer to a
1827 function to be invoked. </li>
1828
1829 <li>'<tt>function args</tt>': argument list whose types match the function
1830 signature argument types. If the function signature indicates the function
1831 accepts a variable number of arguments, the extra arguments can be
1832 specified. </li>
1833
1834 <li>'<tt>normal label</tt>': the label reached when the called function
1835 executes a '<tt><a href="#i_ret">ret</a></tt>' instruction. </li>
1836
1837 <li>'<tt>exception label</tt>': the label reached when a callee returns with
1838 the <a href="#i_unwind"><tt>unwind</tt></a> instruction. </li>
1839
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1840</ol>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1841
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1842<h5>Semantics:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1843
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1844<p>This instruction is designed to operate as a standard '<tt><a
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1845href="#i_call">call</a></tt>' instruction in most regards. The primary
1846difference is that it establishes an association with a label, which is used by
1847the runtime library to unwind the stack.</p>
1848
1849<p>This instruction is used in languages with destructors to ensure that proper
1850cleanup is performed in the case of either a <tt>longjmp</tt> or a thrown
1851exception. Additionally, this is important for implementation of
1852'<tt>catch</tt>' clauses in high-level languages that support them.</p>
1853
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1854<h5>Example:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1855<pre>
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]1856 %retval = invoke i32 %Test(i32 15) to label %Continue
1857 unwind label %TestCleanup <i>; {i32}:retval set</i>
1858 %retval = invoke <a href="#callingconv">coldcc</a> i32 %Test(i32 15) to label %Continue
1859 unwind label %TestCleanup <i>; {i32}:retval set</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1860</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1861</div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1862
1863
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]1864<!-- _______________________________________________________________________ -->
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1865
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1866<div class="doc_subsubsection"> <a name="i_unwind">'<tt>unwind</tt>'
1867Instruction</a> </div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1868
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1869<div class="doc_text">
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1870
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]1871<h5>Syntax:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1872<pre>
1873 unwind
1874</pre>
1875
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]1876<h5>Overview:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1877
1878<p>The '<tt>unwind</tt>' instruction unwinds the stack, continuing control flow
1879at the first callee in the dynamic call stack which used an <a
1880href="#i_invoke"><tt>invoke</tt></a> instruction to perform the call. This is
1881primarily used to implement exception handling.</p>
1882
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]1883<h5>Semantics:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1884
1885<p>The '<tt>unwind</tt>' intrinsic causes execution of the current function to
1886immediately halt. The dynamic call stack is then searched for the first <a
1887href="#i_invoke"><tt>invoke</tt></a> instruction on the call stack. Once found,
1888execution continues at the "exceptional" destination block specified by the
1889<tt>invoke</tt> instruction. If there is no <tt>invoke</tt> instruction in the
1890dynamic call chain, undefined behavior results.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1891</div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1892
1893<!-- _______________________________________________________________________ -->
1894
1895<div class="doc_subsubsection"> <a name="i_unreachable">'<tt>unreachable</tt>'
1896Instruction</a> </div>
1897
1898<div class="doc_text">
1899
1900<h5>Syntax:</h5>
1901<pre>
1902 unreachable
1903</pre>
1904
1905<h5>Overview:</h5>
1906
1907<p>The '<tt>unreachable</tt>' instruction has no defined semantics. This
1908instruction is used to inform the optimizer that a particular portion of the
1909code is not reachable. This can be used to indicate that the code after a
1910no-return function cannot be reached, and other facts.</p>
1911
1912<h5>Semantics:</h5>
1913
1914<p>The '<tt>unreachable</tt>' instruction has no defined semantics.</p>
1915</div>
1916
1917
1918
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1919<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1920<div class="doc_subsection"> <a name="binaryops">Binary Operations</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1921<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1922<p>Binary operators are used to do most of the computation in a
1923program. They require two operands, execute an operation on them, and
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1924produce a single value. The operands might represent
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1925multiple data, as is the case with the <a href="#t_vector">vector</a> data type.
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1926The result value of a binary operator is not
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1927necessarily the same type as its operands.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1928<p>There are several different binary operators:</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1929</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1930<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1931<div class="doc_subsubsection"> <a name="i_add">'<tt>add</tt>'
1932Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1933<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1934<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1935<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]1936</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1937<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1938<p>The '<tt>add</tt>' instruction returns the sum of its two operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1939<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1940<p>The two arguments to the '<tt>add</tt>' instruction must be either <a
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1941 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a> values.
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1942 This instruction can also take <a href="#t_vector">vector</a> versions of the values.
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1943Both arguments must have identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1944<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1945<p>The value produced is the integer or floating point sum of the two
1946operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1947<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1948<pre> &lt;result&gt; = add i32 4, %var <i>; yields {i32}:result = 4 + %var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1949</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1950</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1951<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1952<div class="doc_subsubsection"> <a name="i_sub">'<tt>sub</tt>'
1953Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1954<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1955<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1956<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]1957</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1958<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1959<p>The '<tt>sub</tt>' instruction returns the difference of its two
1960operands.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1961<p>Note that the '<tt>sub</tt>' instruction is used to represent the '<tt>neg</tt>'
1962instruction present in most other intermediate representations.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1963<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1964<p>The two arguments to the '<tt>sub</tt>' instruction must be either <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1965 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1966values.
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1967This instruction can also take <a href="#t_vector">vector</a> versions of the values.
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1968Both arguments must have identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1969<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1970<p>The value produced is the integer or floating point difference of
1971the two operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1972<h5>Example:</h5>
Bill Wendlingaac388b2007-05-29 09:42:13 +0000[diff] [blame]1973<pre>
1974 &lt;result&gt; = sub i32 4, %var <i>; yields {i32}:result = 4 - %var</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1975 &lt;result&gt; = sub i32 0, %val <i>; yields {i32}:result = -%var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1976</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1977</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1978<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1979<div class="doc_subsubsection"> <a name="i_mul">'<tt>mul</tt>'
1980Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1981<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1982<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1983<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]1984</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1985<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1986<p>The '<tt>mul</tt>' instruction returns the product of its two
1987operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1988<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1989<p>The two arguments to the '<tt>mul</tt>' instruction must be either <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1990 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1991values.
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1992This instruction can also take <a href="#t_vector">vector</a> versions of the values.
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1993Both arguments must have identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1994<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1995<p>The value produced is the integer or floating point product of the
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1996two operands.</p>
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]1997<p>Because the operands are the same width, the result of an integer
1998multiplication is the same whether the operands should be deemed unsigned or
1999signed.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2000<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2001<pre> &lt;result&gt; = mul i32 4, %var <i>; yields {i32}:result = 4 * %var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2002</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2003</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2004<!-- _______________________________________________________________________ -->
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2005<div class="doc_subsubsection"> <a name="i_udiv">'<tt>udiv</tt>' Instruction
2006</a></div>
2007<div class="doc_text">
2008<h5>Syntax:</h5>
2009<pre> &lt;result&gt; = udiv &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2010</pre>
2011<h5>Overview:</h5>
2012<p>The '<tt>udiv</tt>' instruction returns the quotient of its two
2013operands.</p>
2014<h5>Arguments:</h5>
2015<p>The two arguments to the '<tt>udiv</tt>' instruction must be
2016<a href="#t_integer">integer</a> values. Both arguments must have identical
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]2017types. This instruction can also take <a href="#t_vector">vector</a> versions
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2018of the values in which case the elements must be integers.</p>
2019<h5>Semantics:</h5>
2020<p>The value produced is the unsigned integer quotient of the two operands. This
2021instruction always performs an unsigned division operation, regardless of
2022whether the arguments are unsigned or not.</p>
2023<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2024<pre> &lt;result&gt; = udiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2025</pre>
2026</div>
2027<!-- _______________________________________________________________________ -->
2028<div class="doc_subsubsection"> <a name="i_sdiv">'<tt>sdiv</tt>' Instruction
2029</a> </div>
2030<div class="doc_text">
2031<h5>Syntax:</h5>
2032<pre> &lt;result&gt; = sdiv &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2033</pre>
2034<h5>Overview:</h5>
2035<p>The '<tt>sdiv</tt>' instruction returns the quotient of its two
2036operands.</p>
2037<h5>Arguments:</h5>
2038<p>The two arguments to the '<tt>sdiv</tt>' instruction must be
2039<a href="#t_integer">integer</a> values. Both arguments must have identical
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]2040types. This instruction can also take <a href="#t_vector">vector</a> versions
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2041of the values in which case the elements must be integers.</p>
2042<h5>Semantics:</h5>
2043<p>The value produced is the signed integer quotient of the two operands. This
2044instruction always performs a signed division operation, regardless of whether
2045the arguments are signed or not.</p>
2046<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2047<pre> &lt;result&gt; = sdiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2048</pre>
2049</div>
2050<!-- _______________________________________________________________________ -->
2051<div class="doc_subsubsection"> <a name="i_fdiv">'<tt>fdiv</tt>'
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2052Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2053<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2054<h5>Syntax:</h5>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2055<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]2056</pre>
2057<h5>Overview:</h5>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2058<p>The '<tt>fdiv</tt>' instruction returns the quotient of its two
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2059operands.</p>
2060<h5>Arguments:</h5>
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]2061<p>The two arguments to the '<tt>fdiv</tt>' instruction must be
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2062<a href="#t_floating">floating point</a> values. Both arguments must have
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]2063identical types. This instruction can also take <a href="#t_vector">vector</a>
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]2064versions of floating point values.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2065<h5>Semantics:</h5>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2066<p>The value produced is the floating point quotient of the two operands.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2067<h5>Example:</h5>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2068<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]2069</pre>
2070</div>
2071<!-- _______________________________________________________________________ -->
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]2072<div class="doc_subsubsection"> <a name="i_urem">'<tt>urem</tt>' Instruction</a>
2073</div>
2074<div class="doc_text">
2075<h5>Syntax:</h5>
2076<pre> &lt;result&gt; = urem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2077</pre>
2078<h5>Overview:</h5>
2079<p>The '<tt>urem</tt>' instruction returns the remainder from the
2080unsigned division of its two arguments.</p>
2081<h5>Arguments:</h5>
2082<p>The two arguments to the '<tt>urem</tt>' instruction must be
2083<a href="#t_integer">integer</a> values. Both arguments must have identical
2084types.</p>
2085<h5>Semantics:</h5>
2086<p>This instruction returns the unsigned integer <i>remainder</i> of a division.
2087This instruction always performs an unsigned division to get the remainder,
2088regardless of whether the arguments are unsigned or not.</p>
2089<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2090<pre> &lt;result&gt; = urem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]2091</pre>
2092
2093</div>
2094<!-- _______________________________________________________________________ -->
2095<div class="doc_subsubsection"> <a name="i_srem">'<tt>srem</tt>'
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2096Instruction</a> </div>
2097<div class="doc_text">
2098<h5>Syntax:</h5>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]2099<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]2100</pre>
2101<h5>Overview:</h5>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]2102<p>The '<tt>srem</tt>' instruction returns the remainder from the
2103signed division of its two operands.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2104<h5>Arguments:</h5>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]2105<p>The two arguments to the '<tt>srem</tt>' instruction must be
2106<a href="#t_integer">integer</a> values. Both arguments must have identical
2107types.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2108<h5>Semantics:</h5>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]2109<p>This instruction returns the <i>remainder</i> of a division (where the result
Reid Spencerc9fdfc82007-03-24 22:23:39 +0000[diff] [blame]2110has the same sign as the dividend, <tt>var1</tt>), not the <i>modulo</i>
2111operator (where the result has the same sign as the divisor, <tt>var2</tt>) of
2112a value. For more information about the difference, see <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2113 href="http://mathforum.org/dr.math/problems/anne.4.28.99.html">The
Reid Spencerc9fdfc82007-03-24 22:23:39 +0000[diff] [blame]2114Math Forum</a>. For a table of how this is implemented in various languages,
Reid Spencer64f5c6c2007-03-24 22:40:44 +0000[diff] [blame]2115please see <a href="http://en.wikipedia.org/wiki/Modulo_operation">
Reid Spencerc9fdfc82007-03-24 22:23:39 +0000[diff] [blame]2116Wikipedia: modulo operation</a>.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2117<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2118<pre> &lt;result&gt; = srem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]2119</pre>
2120
2121</div>
2122<!-- _______________________________________________________________________ -->
2123<div class="doc_subsubsection"> <a name="i_frem">'<tt>frem</tt>'
2124Instruction</a> </div>
2125<div class="doc_text">
2126<h5>Syntax:</h5>
2127<pre> &lt;result&gt; = frem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2128</pre>
2129<h5>Overview:</h5>
2130<p>The '<tt>frem</tt>' instruction returns the remainder from the
2131division of its two operands.</p>
2132<h5>Arguments:</h5>
2133<p>The two arguments to the '<tt>frem</tt>' instruction must be
2134<a href="#t_floating">floating point</a> values. Both arguments must have
2135identical types.</p>
2136<h5>Semantics:</h5>
2137<p>This instruction returns the <i>remainder</i> of a division.</p>
2138<h5>Example:</h5>
2139<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]2140</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2141</div>
Robert Bocchino7b81c752006-02-17 21:18:08 +0000[diff] [blame]2142
Reid Spencer8e11bf82007-02-02 13:57:07 +0000[diff] [blame]2143<!-- ======================================================================= -->
2144<div class="doc_subsection"> <a name="bitwiseops">Bitwise Binary
2145Operations</a> </div>
2146<div class="doc_text">
2147<p>Bitwise binary operators are used to do various forms of
2148bit-twiddling in a program. They are generally very efficient
2149instructions and can commonly be strength reduced from other
2150instructions. They require two operands, execute an operation on them,
2151and produce a single value. The resulting value of the bitwise binary
2152operators is always the same type as its first operand.</p>
2153</div>
2154
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2155<!-- _______________________________________________________________________ -->
2156<div class="doc_subsubsection"> <a name="i_shl">'<tt>shl</tt>'
2157Instruction</a> </div>
2158<div class="doc_text">
2159<h5>Syntax:</h5>
2160<pre> &lt;result&gt; = shl &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2161</pre>
2162<h5>Overview:</h5>
2163<p>The '<tt>shl</tt>' instruction returns the first operand shifted to
2164the left a specified number of bits.</p>
2165<h5>Arguments:</h5>
2166<p>Both arguments to the '<tt>shl</tt>' instruction must be the same <a
2167 href="#t_integer">integer</a> type.</p>
2168<h5>Semantics:</h5>
2169<p>The value produced is <tt>var1</tt> * 2<sup><tt>var2</tt></sup>.</p>
2170<h5>Example:</h5><pre>
2171 &lt;result&gt; = shl i32 4, %var <i>; yields {i32}: 4 &lt;&lt; %var</i>
2172 &lt;result&gt; = shl i32 4, 2 <i>; yields {i32}: 16</i>
2173 &lt;result&gt; = shl i32 1, 10 <i>; yields {i32}: 1024</i>
2174</pre>
2175</div>
2176<!-- _______________________________________________________________________ -->
2177<div class="doc_subsubsection"> <a name="i_lshr">'<tt>lshr</tt>'
2178Instruction</a> </div>
2179<div class="doc_text">
2180<h5>Syntax:</h5>
2181<pre> &lt;result&gt; = lshr &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2182</pre>
2183
2184<h5>Overview:</h5>
2185<p>The '<tt>lshr</tt>' instruction (logical shift right) returns the first
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]2186operand shifted to the right a specified number of bits with zero fill.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2187
2188<h5>Arguments:</h5>
2189<p>Both arguments to the '<tt>lshr</tt>' instruction must be the same
2190<a href="#t_integer">integer</a> type.</p>
2191
2192<h5>Semantics:</h5>
2193<p>This instruction always performs a logical shift right operation. The most
2194significant bits of the result will be filled with zero bits after the
2195shift.</p>
2196
2197<h5>Example:</h5>
2198<pre>
2199 &lt;result&gt; = lshr i32 4, 1 <i>; yields {i32}:result = 2</i>
2200 &lt;result&gt; = lshr i32 4, 2 <i>; yields {i32}:result = 1</i>
2201 &lt;result&gt; = lshr i8 4, 3 <i>; yields {i8}:result = 0</i>
2202 &lt;result&gt; = lshr i8 -2, 1 <i>; yields {i8}:result = 0x7FFFFFFF </i>
2203</pre>
2204</div>
2205
Reid Spencer8e11bf82007-02-02 13:57:07 +0000[diff] [blame]2206<!-- _______________________________________________________________________ -->
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2207<div class="doc_subsubsection"> <a name="i_ashr">'<tt>ashr</tt>'
2208Instruction</a> </div>
2209<div class="doc_text">
2210
2211<h5>Syntax:</h5>
2212<pre> &lt;result&gt; = ashr &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2213</pre>
2214
2215<h5>Overview:</h5>
2216<p>The '<tt>ashr</tt>' instruction (arithmetic shift right) returns the first
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]2217operand shifted to the right a specified number of bits with sign extension.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2218
2219<h5>Arguments:</h5>
2220<p>Both arguments to the '<tt>ashr</tt>' instruction must be the same
2221<a href="#t_integer">integer</a> type.</p>
2222
2223<h5>Semantics:</h5>
2224<p>This instruction always performs an arithmetic shift right operation,
2225The most significant bits of the result will be filled with the sign bit
2226of <tt>var1</tt>.</p>
2227
2228<h5>Example:</h5>
2229<pre>
2230 &lt;result&gt; = ashr i32 4, 1 <i>; yields {i32}:result = 2</i>
2231 &lt;result&gt; = ashr i32 4, 2 <i>; yields {i32}:result = 1</i>
2232 &lt;result&gt; = ashr i8 4, 3 <i>; yields {i8}:result = 0</i>
2233 &lt;result&gt; = ashr i8 -2, 1 <i>; yields {i8}:result = -1</i>
2234</pre>
2235</div>
2236
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2237<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2238<div class="doc_subsubsection"> <a name="i_and">'<tt>and</tt>'
2239Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2240<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2241<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2242<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]2243</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2244<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2245<p>The '<tt>and</tt>' instruction returns the bitwise logical and of
2246its two operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2247<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2248<p>The two arguments to the '<tt>and</tt>' instruction must be <a
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]2249 href="#t_integer">integer</a> values. Both arguments must have
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2250identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2251<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2252<p>The truth table used for the '<tt>and</tt>' instruction is:</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2253<p> </p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]2254<div style="align: center">
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2255<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2256 <tbody>
2257 <tr>
2258 <td>In0</td>
2259 <td>In1</td>
2260 <td>Out</td>
2261 </tr>
2262 <tr>
2263 <td>0</td>
2264 <td>0</td>
2265 <td>0</td>
2266 </tr>
2267 <tr>
2268 <td>0</td>
2269 <td>1</td>
2270 <td>0</td>
2271 </tr>
2272 <tr>
2273 <td>1</td>
2274 <td>0</td>
2275 <td>0</td>
2276 </tr>
2277 <tr>
2278 <td>1</td>
2279 <td>1</td>
2280 <td>1</td>
2281 </tr>
2282 </tbody>
2283</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]2284</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2285<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2286<pre> &lt;result&gt; = and i32 4, %var <i>; yields {i32}:result = 4 &amp; %var</i>
2287 &lt;result&gt; = and i32 15, 40 <i>; yields {i32}:result = 8</i>
2288 &lt;result&gt; = and i32 4, 8 <i>; yields {i32}:result = 0</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2289</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2290</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2291<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2292<div class="doc_subsubsection"> <a name="i_or">'<tt>or</tt>' Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2293<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2294<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2295<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]2296</pre>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2297<h5>Overview:</h5>
2298<p>The '<tt>or</tt>' instruction returns the bitwise logical inclusive
2299or of its two operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2300<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2301<p>The two arguments to the '<tt>or</tt>' instruction must be <a
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]2302 href="#t_integer">integer</a> values. Both arguments must have
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2303identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2304<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2305<p>The truth table used for the '<tt>or</tt>' instruction is:</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2306<p> </p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]2307<div style="align: center">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2308<table border="1" cellspacing="0" cellpadding="4">
2309 <tbody>
2310 <tr>
2311 <td>In0</td>
2312 <td>In1</td>
2313 <td>Out</td>
2314 </tr>
2315 <tr>
2316 <td>0</td>
2317 <td>0</td>
2318 <td>0</td>
2319 </tr>
2320 <tr>
2321 <td>0</td>
2322 <td>1</td>
2323 <td>1</td>
2324 </tr>
2325 <tr>
2326 <td>1</td>
2327 <td>0</td>
2328 <td>1</td>
2329 </tr>
2330 <tr>
2331 <td>1</td>
2332 <td>1</td>
2333 <td>1</td>
2334 </tr>
2335 </tbody>
2336</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]2337</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2338<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2339<pre> &lt;result&gt; = or i32 4, %var <i>; yields {i32}:result = 4 | %var</i>
2340 &lt;result&gt; = or i32 15, 40 <i>; yields {i32}:result = 47</i>
2341 &lt;result&gt; = or i32 4, 8 <i>; yields {i32}:result = 12</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2342</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2343</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2344<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2345<div class="doc_subsubsection"> <a name="i_xor">'<tt>xor</tt>'
2346Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2347<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2348<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2349<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]2350</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2351<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2352<p>The '<tt>xor</tt>' instruction returns the bitwise logical exclusive
2353or of its two operands. The <tt>xor</tt> is used to implement the
2354"one's complement" operation, which is the "~" operator in C.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2355<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2356<p>The two arguments to the '<tt>xor</tt>' instruction must be <a
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]2357 href="#t_integer">integer</a> values. Both arguments must have
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2358identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2359<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2360<p>The truth table used for the '<tt>xor</tt>' instruction is:</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2361<p> </p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]2362<div style="align: center">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2363<table border="1" cellspacing="0" cellpadding="4">
2364 <tbody>
2365 <tr>
2366 <td>In0</td>
2367 <td>In1</td>
2368 <td>Out</td>
2369 </tr>
2370 <tr>
2371 <td>0</td>
2372 <td>0</td>
2373 <td>0</td>
2374 </tr>
2375 <tr>
2376 <td>0</td>
2377 <td>1</td>
2378 <td>1</td>
2379 </tr>
2380 <tr>
2381 <td>1</td>
2382 <td>0</td>
2383 <td>1</td>
2384 </tr>
2385 <tr>
2386 <td>1</td>
2387 <td>1</td>
2388 <td>0</td>
2389 </tr>
2390 </tbody>
2391</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]2392</div>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2393<p> </p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2394<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2395<pre> &lt;result&gt; = xor i32 4, %var <i>; yields {i32}:result = 4 ^ %var</i>
2396 &lt;result&gt; = xor i32 15, 40 <i>; yields {i32}:result = 39</i>
2397 &lt;result&gt; = xor i32 4, 8 <i>; yields {i32}:result = 12</i>
2398 &lt;result&gt; = xor i32 %V, -1 <i>; yields {i32}:result = ~%V</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2399</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2400</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2401
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2402<!-- ======================================================================= -->
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2403<div class="doc_subsection">
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2404 <a name="vectorops">Vector Operations</a>
2405</div>
2406
2407<div class="doc_text">
2408
2409<p>LLVM supports several instructions to represent vector operations in a
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]2410target-independent manner. These instructions cover the element-access and
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2411vector-specific operations needed to process vectors effectively. While LLVM
2412does directly support these vector operations, many sophisticated algorithms
2413will want to use target-specific intrinsics to take full advantage of a specific
2414target.</p>
2415
2416</div>
2417
2418<!-- _______________________________________________________________________ -->
2419<div class="doc_subsubsection">
2420 <a name="i_extractelement">'<tt>extractelement</tt>' Instruction</a>
2421</div>
2422
2423<div class="doc_text">
2424
2425<h5>Syntax:</h5>
2426
2427<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2428 &lt;result&gt; = extractelement &lt;n x &lt;ty&gt;&gt; &lt;val&gt;, i32 &lt;idx&gt; <i>; yields &lt;ty&gt;</i>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2429</pre>
2430
2431<h5>Overview:</h5>
2432
2433<p>
2434The '<tt>extractelement</tt>' instruction extracts a single scalar
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]2435element from a vector at a specified index.
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2436</p>
2437
2438
2439<h5>Arguments:</h5>
2440
2441<p>
2442The first operand of an '<tt>extractelement</tt>' instruction is a
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]2443value of <a href="#t_vector">vector</a> type. The second operand is
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2444an index indicating the position from which to extract the element.
2445The index may be a variable.</p>
2446
2447<h5>Semantics:</h5>
2448
2449<p>
2450The result is a scalar of the same type as the element type of
2451<tt>val</tt>. Its value is the value at position <tt>idx</tt> of
2452<tt>val</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
2453results are undefined.
2454</p>
2455
2456<h5>Example:</h5>
2457
2458<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2459 %result = extractelement &lt;4 x i32&gt; %vec, i32 0 <i>; yields i32</i>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2460</pre>
2461</div>
2462
2463
2464<!-- _______________________________________________________________________ -->
2465<div class="doc_subsubsection">
2466 <a name="i_insertelement">'<tt>insertelement</tt>' Instruction</a>
2467</div>
2468
2469<div class="doc_text">
2470
2471<h5>Syntax:</h5>
2472
2473<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2474 &lt;result&gt; = insertelement &lt;n x &lt;ty&gt;&gt; &lt;val&gt;, &lt;ty&gt; &lt;elt&gt, i32 &lt;idx&gt; <i>; yields &lt;n x &lt;ty&gt;&gt;</i>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2475</pre>
2476
2477<h5>Overview:</h5>
2478
2479<p>
2480The '<tt>insertelement</tt>' instruction inserts a scalar
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]2481element into a vector at a specified index.
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2482</p>
2483
2484
2485<h5>Arguments:</h5>
2486
2487<p>
2488The first operand of an '<tt>insertelement</tt>' instruction is a
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]2489value of <a href="#t_vector">vector</a> type. The second operand is a
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2490scalar value whose type must equal the element type of the first
2491operand. The third operand is an index indicating the position at
2492which to insert the value. The index may be a variable.</p>
2493
2494<h5>Semantics:</h5>
2495
2496<p>
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]2497The result is a vector of the same type as <tt>val</tt>. Its
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2498element values are those of <tt>val</tt> except at position
2499<tt>idx</tt>, where it gets the value <tt>elt</tt>. If <tt>idx</tt>
2500exceeds the length of <tt>val</tt>, the results are undefined.
2501</p>
2502
2503<h5>Example:</h5>
2504
2505<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2506 %result = insertelement &lt;4 x i32&gt; %vec, i32 1, i32 0 <i>; yields &lt;4 x i32&gt;</i>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2507</pre>
2508</div>
2509
2510<!-- _______________________________________________________________________ -->
2511<div class="doc_subsubsection">
2512 <a name="i_shufflevector">'<tt>shufflevector</tt>' Instruction</a>
2513</div>
2514
2515<div class="doc_text">
2516
2517<h5>Syntax:</h5>
2518
2519<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2520 &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 i32&gt; &lt;mask&gt; <i>; yields &lt;n x &lt;ty&gt;&gt;</i>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2521</pre>
2522
2523<h5>Overview:</h5>
2524
2525<p>
2526The '<tt>shufflevector</tt>' instruction constructs a permutation of elements
2527from two input vectors, returning a vector of the same type.
2528</p>
2529
2530<h5>Arguments:</h5>
2531
2532<p>
2533The first two operands of a '<tt>shufflevector</tt>' instruction are vectors
2534with types that match each other and types that match the result of the
2535instruction. The third argument is a shuffle mask, which has the same number
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2536of elements as the other vector type, but whose element type is always 'i32'.
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2537</p>
2538
2539<p>
2540The shuffle mask operand is required to be a constant vector with either
2541constant integer or undef values.
2542</p>
2543
2544<h5>Semantics:</h5>
2545
2546<p>
2547The elements of the two input vectors are numbered from left to right across
2548both of the vectors. The shuffle mask operand specifies, for each element of
2549the result vector, which element of the two input registers the result element
2550gets. The element selector may be undef (meaning "don't care") and the second
2551operand may be undef if performing a shuffle from only one vector.
2552</p>
2553
2554<h5>Example:</h5>
2555
2556<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2557 %result = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; %v2,
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]2558 &lt;4 x i32&gt; &lt;i32 0, i32 4, i32 1, i32 5&gt; <i>; yields &lt;4 x i32&gt;</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2559 %result = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; undef,
2560 &lt;4 x i32&gt; &lt;i32 0, i32 1, i32 2, i32 3&gt; <i>; yields &lt;4 x i32&gt;</i> - Identity shuffle.
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2561</pre>
2562</div>
2563
Tanya Lattner09474292006-04-14 19:24:33 +0000[diff] [blame]2564
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2565<!-- ======================================================================= -->
2566<div class="doc_subsection">
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]2567 <a name="memoryops">Memory Access and Addressing Operations</a>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2568</div>
2569
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2570<div class="doc_text">
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2571
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2572<p>A key design point of an SSA-based representation is how it
2573represents memory. In LLVM, no memory locations are in SSA form, which
2574makes things very simple. This section describes how to read, write,
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]2575allocate, and free memory in LLVM.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2576
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2577</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2578
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2579<!-- _______________________________________________________________________ -->
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2580<div class="doc_subsubsection">
2581 <a name="i_malloc">'<tt>malloc</tt>' Instruction</a>
2582</div>
2583
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2584<div class="doc_text">
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2585
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2586<h5>Syntax:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2587
2588<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2589 &lt;result&gt; = malloc &lt;type&gt;[, i32 &lt;NumElements&gt;][, align &lt;alignment&gt;] <i>; yields {type*}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2590</pre>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2591
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2592<h5>Overview:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2593
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2594<p>The '<tt>malloc</tt>' instruction allocates memory from the system
2595heap and returns a pointer to it.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2596
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2597<h5>Arguments:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2598
2599<p>The '<tt>malloc</tt>' instruction allocates
2600<tt>sizeof(&lt;type&gt;)*NumElements</tt>
John Criswell6e4ca612004-02-24 16:13:56 +0000[diff] [blame]2601bytes of memory from the operating system and returns a pointer of the
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2602appropriate type to the program. If "NumElements" is specified, it is the
2603number of elements allocated. If an alignment is specified, the value result
2604of the allocation is guaranteed to be aligned to at least that boundary. If
2605not specified, or if zero, the target can choose to align the allocation on any
2606convenient boundary.</p>
2607
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2608<p>'<tt>type</tt>' must be a sized type.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2609
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2610<h5>Semantics:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2611
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2612<p>Memory is allocated using the system "<tt>malloc</tt>" function, and
2613a pointer is returned.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2614
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2615<h5>Example:</h5>
2616
2617<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2618 %array = malloc [4 x i8 ] <i>; yields {[%4 x i8]*}:array</i>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2619
Bill Wendlingaac388b2007-05-29 09:42:13 +0000[diff] [blame]2620 %size = <a href="#i_add">add</a> i32 2, 2 <i>; yields {i32}:size = i32 4</i>
2621 %array1 = malloc i8, i32 4 <i>; yields {i8*}:array1</i>
2622 %array2 = malloc [12 x i8], i32 %size <i>; yields {[12 x i8]*}:array2</i>
2623 %array3 = malloc i32, i32 4, align 1024 <i>; yields {i32*}:array3</i>
2624 %array4 = malloc i32, align 1024 <i>; yields {i32*}:array4</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2625</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2626</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2627
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2628<!-- _______________________________________________________________________ -->
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2629<div class="doc_subsubsection">
2630 <a name="i_free">'<tt>free</tt>' Instruction</a>
2631</div>
2632
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2633<div class="doc_text">
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2634
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2635<h5>Syntax:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2636
2637<pre>
2638 free &lt;type&gt; &lt;value&gt; <i>; yields {void}</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2639</pre>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2640
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2641<h5>Overview:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2642
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2643<p>The '<tt>free</tt>' instruction returns memory back to the unused
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]2644memory heap to be reallocated in the future.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2645
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2646<h5>Arguments:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2647
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2648<p>'<tt>value</tt>' shall be a pointer value that points to a value
2649that was allocated with the '<tt><a href="#i_malloc">malloc</a></tt>'
2650instruction.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2651
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2652<h5>Semantics:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2653
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]2654<p>Access to the memory pointed to by the pointer is no longer defined
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2655after this instruction executes.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2656
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2657<h5>Example:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2658
2659<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2660 %array = <a href="#i_malloc">malloc</a> [4 x i8] <i>; yields {[4 x i8]*}:array</i>
2661 free [4 x i8]* %array
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2662</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2663</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2664
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2665<!-- _______________________________________________________________________ -->
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2666<div class="doc_subsubsection">
2667 <a name="i_alloca">'<tt>alloca</tt>' Instruction</a>
2668</div>
2669
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2670<div class="doc_text">
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2671
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2672<h5>Syntax:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2673
2674<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2675 &lt;result&gt; = alloca &lt;type&gt;[, i32 &lt;NumElements&gt;][, align &lt;alignment&gt;] <i>; yields {type*}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2676</pre>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2677
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2678<h5>Overview:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2679
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]2680<p>The '<tt>alloca</tt>' instruction allocates memory on the stack frame of the
2681currently executing function, to be automatically released when this function
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2682returns to its caller.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2683
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2684<h5>Arguments:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2685
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]2686<p>The '<tt>alloca</tt>' instruction allocates <tt>sizeof(&lt;type&gt;)*NumElements</tt>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2687bytes of memory on the runtime stack, returning a pointer of the
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2688appropriate type to the program. If "NumElements" is specified, it is the
2689number of elements allocated. If an alignment is specified, the value result
2690of the allocation is guaranteed to be aligned to at least that boundary. If
2691not specified, or if zero, the target can choose to align the allocation on any
2692convenient boundary.</p>
2693
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2694<p>'<tt>type</tt>' may be any sized type.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2695
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2696<h5>Semantics:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2697
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]2698<p>Memory is allocated; a pointer is returned. '<tt>alloca</tt>'d
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2699memory is automatically released when the function returns. The '<tt>alloca</tt>'
2700instruction is commonly used to represent automatic variables that must
2701have an address available. When the function returns (either with the <tt><a
John Criswelldae2e932005-05-12 16:55:34 +0000[diff] [blame]2702 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]2703instructions), the memory is reclaimed.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2704
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2705<h5>Example:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2706
2707<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2708 %ptr = alloca i32 <i>; yields {i32*}:ptr</i>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2709 %ptr = alloca i32, i32 4 <i>; yields {i32*}:ptr</i>
2710 %ptr = alloca i32, i32 4, align 1024 <i>; yields {i32*}:ptr</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2711 %ptr = alloca i32, align 1024 <i>; yields {i32*}:ptr</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2712</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2713</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2714
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2715<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2716<div class="doc_subsubsection"> <a name="i_load">'<tt>load</tt>'
2717Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2718<div class="doc_text">
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2719<h5>Syntax:</h5>
Christopher Lamb2330e4d2007-04-21 08:16:25 +0000[diff] [blame]2720<pre> &lt;result&gt; = load &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;]<br> &lt;result&gt; = volatile load &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;]<br></pre>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2721<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2722<p>The '<tt>load</tt>' instruction is used to read from memory.</p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2723<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2724<p>The argument to the '<tt>load</tt>' instruction specifies the memory
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]2725address from which to load. The pointer must point to a <a
Chris Lattnere53e5082004-06-03 22:57:15 +0000[diff] [blame]2726 href="#t_firstclass">first class</a> type. If the <tt>load</tt> is
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]2727marked as <tt>volatile</tt>, then the optimizer is not allowed to modify
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2728the number or order of execution of this <tt>load</tt> with other
2729volatile <tt>load</tt> and <tt><a href="#i_store">store</a></tt>
2730instructions. </p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2731<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2732<p>The location of memory pointed to is loaded.</p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2733<h5>Examples:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2734<pre> %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2735 <a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2736 href="#i_store">store</a> i32 3, i32* %ptr <i>; yields {void}</i>
2737 %val = load i32* %ptr <i>; yields {i32}:val = i32 3</i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2738</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2739</div>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2740<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2741<div class="doc_subsubsection"> <a name="i_store">'<tt>store</tt>'
2742Instruction</a> </div>
Reid Spencer035ab572006-11-09 21:18:01 +0000[diff] [blame]2743<div class="doc_text">
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2744<h5>Syntax:</h5>
Christopher Lamb2330e4d2007-04-21 08:16:25 +0000[diff] [blame]2745<pre> store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;] <i>; yields {void}</i>
2746 volatile store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;] <i>; yields {void}</i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2747</pre>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2748<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2749<p>The '<tt>store</tt>' instruction is used to write to memory.</p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2750<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2751<p>There are two arguments to the '<tt>store</tt>' instruction: a value
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]2752to store and an address at which to store it. The type of the '<tt>&lt;pointer&gt;</tt>'
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2753operand must be a pointer to the type of the '<tt>&lt;value&gt;</tt>'
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]2754operand. If the <tt>store</tt> is marked as <tt>volatile</tt>, then the
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2755optimizer is not allowed to modify the number or order of execution of
2756this <tt>store</tt> with other volatile <tt>load</tt> and <tt><a
2757 href="#i_store">store</a></tt> instructions.</p>
2758<h5>Semantics:</h5>
2759<p>The contents of memory are updated to contain '<tt>&lt;value&gt;</tt>'
2760at the location specified by the '<tt>&lt;pointer&gt;</tt>' operand.</p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2761<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2762<pre> %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2763 <a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2764 href="#i_store">store</a> i32 3, i32* %ptr <i>; yields {void}</i>
2765 %val = load i32* %ptr <i>; yields {i32}:val = i32 3</i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2766</pre>
Reid Spencer47ce1792006-11-09 21:15:49 +0000[diff] [blame]2767</div>
2768
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2769<!-- _______________________________________________________________________ -->
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2770<div class="doc_subsubsection">
2771 <a name="i_getelementptr">'<tt>getelementptr</tt>' Instruction</a>
2772</div>
2773
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2774<div class="doc_text">
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2775<h5>Syntax:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2776<pre>
2777 &lt;result&gt; = getelementptr &lt;ty&gt;* &lt;ptrval&gt;{, &lt;ty&gt; &lt;idx&gt;}*
2778</pre>
2779
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2780<h5>Overview:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2781
2782<p>
2783The '<tt>getelementptr</tt>' instruction is used to get the address of a
2784subelement of an aggregate data structure.</p>
2785
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2786<h5>Arguments:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2787
Reid Spencer85f5b5b2006-12-04 21:29:24 +0000[diff] [blame]2788<p>This instruction takes a list of integer operands that indicate what
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2789elements of the aggregate object to index to. The actual types of the arguments
2790provided depend on the type of the first pointer argument. The
2791'<tt>getelementptr</tt>' instruction is used to index down through the type
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]2792levels of a structure or to a specific index in an array. When indexing into a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2793structure, only <tt>i32</tt> integer constants are allowed. When indexing
Reid Spencer85f5b5b2006-12-04 21:29:24 +0000[diff] [blame]2794into an array or pointer, only integers of 32 or 64 bits are allowed, and will
2795be sign extended to 64-bit values.</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2796
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2797<p>For example, let's consider a C code fragment and how it gets
2798compiled to LLVM:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2799
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2800<div class="doc_code">
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2801<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2802struct RT {
2803 char A;
Chris Lattnercabc8462007-05-29 15:43:56 +0000[diff] [blame]2804 int B[10][20];
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2805 char C;
2806};
2807struct ST {
Chris Lattnercabc8462007-05-29 15:43:56 +0000[diff] [blame]2808 int X;
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2809 double Y;
2810 struct RT Z;
2811};
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2812
Chris Lattnercabc8462007-05-29 15:43:56 +0000[diff] [blame]2813int *foo(struct ST *s) {
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2814 return &amp;s[1].Z.B[5][13];
2815}
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2816</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2817</div>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2818
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2819<p>The LLVM code generated by the GCC frontend is:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2820
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2821<div class="doc_code">
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2822<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2823%RT = type { i8 , [10 x [20 x i32]], i8 }
2824%ST = type { i32, double, %RT }
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2825
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2826define i32* %foo(%ST* %s) {
2827entry:
2828 %reg = getelementptr %ST* %s, i32 1, i32 2, i32 1, i32 5, i32 13
2829 ret i32* %reg
2830}
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2831</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2832</div>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2833
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2834<h5>Semantics:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2835
2836<p>The index types specified for the '<tt>getelementptr</tt>' instruction depend
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]2837on the pointer type that is being indexed into. <a href="#t_pointer">Pointer</a>
Reid Spencer85f5b5b2006-12-04 21:29:24 +0000[diff] [blame]2838and <a href="#t_array">array</a> types can use a 32-bit or 64-bit
Reid Spencer42ddd842006-12-03 16:53:48 +0000[diff] [blame]2839<a href="#t_integer">integer</a> type but the value will always be sign extended
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]2840to 64-bits. <a href="#t_struct">Structure</a> types require <tt>i32</tt>
Reid Spencer42ddd842006-12-03 16:53:48 +0000[diff] [blame]2841<b>constants</b>.</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2842
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2843<p>In the example above, the first index is indexing into the '<tt>%ST*</tt>'
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2844type, which is a pointer, yielding a '<tt>%ST</tt>' = '<tt>{ i32, double, %RT
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2845}</tt>' type, a structure. The second index indexes into the third element of
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2846the structure, yielding a '<tt>%RT</tt>' = '<tt>{ i8 , [10 x [20 x i32]],
2847i8 }</tt>' type, another structure. The third index indexes into the second
2848element of the structure, yielding a '<tt>[10 x [20 x i32]]</tt>' type, an
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2849array. The two dimensions of the array are subscripted into, yielding an
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2850'<tt>i32</tt>' type. The '<tt>getelementptr</tt>' instruction returns a pointer
2851to this element, thus computing a value of '<tt>i32*</tt>' type.</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2852
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2853<p>Note that it is perfectly legal to index partially through a
2854structure, returning a pointer to an inner element. Because of this,
2855the LLVM code for the given testcase is equivalent to:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2856
2857<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2858 define i32* %foo(%ST* %s) {
2859 %t1 = getelementptr %ST* %s, i32 1 <i>; yields %ST*:%t1</i>
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]2860 %t2 = getelementptr %ST* %t1, i32 0, i32 2 <i>; yields %RT*:%t2</i>
2861 %t3 = getelementptr %RT* %t2, i32 0, i32 1 <i>; yields [10 x [20 x i32]]*:%t3</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2862 %t4 = getelementptr [10 x [20 x i32]]* %t3, i32 0, i32 5 <i>; yields [20 x i32]*:%t4</i>
2863 %t5 = getelementptr [20 x i32]* %t4, i32 0, i32 13 <i>; yields i32*:%t5</i>
2864 ret i32* %t5
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2865 }
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]2866</pre>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]2867
2868<p>Note that it is undefined to access an array out of bounds: array and
2869pointer indexes must always be within the defined bounds of the array type.
2870The one exception for this rules is zero length arrays. These arrays are
2871defined to be accessible as variable length arrays, which requires access
2872beyond the zero'th element.</p>
2873
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]2874<p>The getelementptr instruction is often confusing. For some more insight
2875into how it works, see <a href="GetElementPtr.html">the getelementptr
2876FAQ</a>.</p>
2877
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2878<h5>Example:</h5>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]2879
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2880<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2881 <i>; yields [12 x i8]*:aptr</i>
2882 %aptr = getelementptr {i32, [12 x i8]}* %sptr, i64 0, i32 1
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2883</pre>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2884</div>
Reid Spencer47ce1792006-11-09 21:15:49 +0000[diff] [blame]2885
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2886<!-- ======================================================================= -->
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]2887<div class="doc_subsection"> <a name="convertops">Conversion Operations</a>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2888</div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2889<div class="doc_text">
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]2890<p>The instructions in this category are the conversion instructions (casting)
2891which all take a single operand and a type. They perform various bit conversions
2892on the operand.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2893</div>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]2894
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]2895<!-- _______________________________________________________________________ -->
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]2896<div class="doc_subsubsection">
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2897 <a name="i_trunc">'<tt>trunc .. to</tt>' Instruction</a>
2898</div>
2899<div class="doc_text">
2900
2901<h5>Syntax:</h5>
2902<pre>
2903 &lt;result&gt; = trunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2904</pre>
2905
2906<h5>Overview:</h5>
2907<p>
2908The '<tt>trunc</tt>' instruction truncates its operand to the type <tt>ty2</tt>.
2909</p>
2910
2911<h5>Arguments:</h5>
2912<p>
2913The '<tt>trunc</tt>' instruction takes a <tt>value</tt> to trunc, which must
2914be an <a href="#t_integer">integer</a> type, and a type that specifies the size
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]2915and type of the result, which must be an <a href="#t_integer">integer</a>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2916type. The bit size of <tt>value</tt> must be larger than the bit size of
2917<tt>ty2</tt>. Equal sized types are not allowed.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2918
2919<h5>Semantics:</h5>
2920<p>
2921The '<tt>trunc</tt>' instruction truncates the high order bits in <tt>value</tt>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2922and converts the remaining bits to <tt>ty2</tt>. Since the source size must be
2923larger than the destination size, <tt>trunc</tt> cannot be a <i>no-op cast</i>.
2924It will always truncate bits.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2925
2926<h5>Example:</h5>
2927<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2928 %X = trunc i32 257 to i8 <i>; yields i8:1</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]2929 %Y = trunc i32 123 to i1 <i>; yields i1:true</i>
2930 %Y = trunc i32 122 to i1 <i>; yields i1:false</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2931</pre>
2932</div>
2933
2934<!-- _______________________________________________________________________ -->
2935<div class="doc_subsubsection">
2936 <a name="i_zext">'<tt>zext .. to</tt>' Instruction</a>
2937</div>
2938<div class="doc_text">
2939
2940<h5>Syntax:</h5>
2941<pre>
2942 &lt;result&gt; = zext &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>zext</tt>' instruction zero extends its operand to type
2947<tt>ty2</tt>.</p>
2948
2949
2950<h5>Arguments:</h5>
2951<p>The '<tt>zext</tt>' instruction takes a value to cast, which must be of
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]2952<a href="#t_integer">integer</a> type, and a type to cast it to, which must
2953also be of <a href="#t_integer">integer</a> type. The bit size of the
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2954<tt>value</tt> must be smaller than the bit size of the destination type,
2955<tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2956
2957<h5>Semantics:</h5>
2958<p>The <tt>zext</tt> fills the high order bits of the <tt>value</tt> with zero
Chris Lattnerd1d25172007-05-24 19:13:27 +0000[diff] [blame]2959bits until it reaches the size of the destination type, <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2960
Reid Spencerb5929522007-01-12 15:46:11 +0000[diff] [blame]2961<p>When zero extending from i1, the result will always be either 0 or 1.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2962
2963<h5>Example:</h5>
2964<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2965 %X = zext i32 257 to i64 <i>; yields i64:257</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]2966 %Y = zext i1 true to i32 <i>; yields i32:1</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2967</pre>
2968</div>
2969
2970<!-- _______________________________________________________________________ -->
2971<div class="doc_subsubsection">
2972 <a name="i_sext">'<tt>sext .. to</tt>' Instruction</a>
2973</div>
2974<div class="doc_text">
2975
2976<h5>Syntax:</h5>
2977<pre>
2978 &lt;result&gt; = sext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2979</pre>
2980
2981<h5>Overview:</h5>
2982<p>The '<tt>sext</tt>' sign extends <tt>value</tt> to the type <tt>ty2</tt>.</p>
2983
2984<h5>Arguments:</h5>
2985<p>
2986The '<tt>sext</tt>' instruction takes a value to cast, which must be of
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]2987<a href="#t_integer">integer</a> type, and a type to cast it to, which must
2988also be of <a href="#t_integer">integer</a> type. The bit size of the
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2989<tt>value</tt> must be smaller than the bit size of the destination type,
2990<tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2991
2992<h5>Semantics:</h5>
2993<p>
2994The '<tt>sext</tt>' instruction performs a sign extension by copying the sign
2995bit (highest order bit) of the <tt>value</tt> until it reaches the bit size of
Chris Lattnerd1d25172007-05-24 19:13:27 +0000[diff] [blame]2996the type <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2997
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]2998<p>When sign extending from i1, the extension always results in -1 or 0.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2999
3000<h5>Example:</h5>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3001<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3002 %X = sext i8 -1 to i16 <i>; yields i16 :65535</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3003 %Y = sext i1 true to i32 <i>; yields i32:-1</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3004</pre>
3005</div>
3006
3007<!-- _______________________________________________________________________ -->
3008<div class="doc_subsubsection">
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]3009 <a name="i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a>
3010</div>
3011
3012<div class="doc_text">
3013
3014<h5>Syntax:</h5>
3015
3016<pre>
3017 &lt;result&gt; = fptrunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3018</pre>
3019
3020<h5>Overview:</h5>
3021<p>The '<tt>fptrunc</tt>' instruction truncates <tt>value</tt> to type
3022<tt>ty2</tt>.</p>
3023
3024
3025<h5>Arguments:</h5>
3026<p>The '<tt>fptrunc</tt>' instruction takes a <a href="#t_floating">floating
3027 point</a> value to cast and a <a href="#t_floating">floating point</a> type to
3028cast it to. The size of <tt>value</tt> must be larger than the size of
3029<tt>ty2</tt>. This implies that <tt>fptrunc</tt> cannot be used to make a
3030<i>no-op cast</i>.</p>
3031
3032<h5>Semantics:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3033<p> The '<tt>fptrunc</tt>' instruction truncates a <tt>value</tt> from a larger
3034<a href="#t_floating">floating point</a> type to a smaller
3035<a href="#t_floating">floating point</a> type. If the value cannot fit within
3036the destination type, <tt>ty2</tt>, then the results are undefined.</p>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]3037
3038<h5>Example:</h5>
3039<pre>
3040 %X = fptrunc double 123.0 to float <i>; yields float:123.0</i>
3041 %Y = fptrunc double 1.0E+300 to float <i>; yields undefined</i>
3042</pre>
3043</div>
3044
3045<!-- _______________________________________________________________________ -->
3046<div class="doc_subsubsection">
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3047 <a name="i_fpext">'<tt>fpext .. to</tt>' Instruction</a>
3048</div>
3049<div class="doc_text">
3050
3051<h5>Syntax:</h5>
3052<pre>
3053 &lt;result&gt; = fpext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3054</pre>
3055
3056<h5>Overview:</h5>
3057<p>The '<tt>fpext</tt>' extends a floating point <tt>value</tt> to a larger
3058floating point value.</p>
3059
3060<h5>Arguments:</h5>
3061<p>The '<tt>fpext</tt>' instruction takes a
3062<a href="#t_floating">floating point</a> <tt>value</tt> to cast,
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3063and a <a href="#t_floating">floating point</a> type to cast it to. The source
3064type must be smaller than the destination type.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3065
3066<h5>Semantics:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3067<p>The '<tt>fpext</tt>' instruction extends the <tt>value</tt> from a smaller
Duncan Sands8036ca42007-03-30 12:22:09 +0000[diff] [blame]3068<a href="#t_floating">floating point</a> type to a larger
3069<a href="#t_floating">floating point</a> type. The <tt>fpext</tt> cannot be
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3070used to make a <i>no-op cast</i> because it always changes bits. Use
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]3071<tt>bitcast</tt> to make a <i>no-op cast</i> for a floating point cast.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3072
3073<h5>Example:</h5>
3074<pre>
3075 %X = fpext float 3.1415 to double <i>; yields double:3.1415</i>
3076 %Y = fpext float 1.0 to float <i>; yields float:1.0 (no-op)</i>
3077</pre>
3078</div>
3079
3080<!-- _______________________________________________________________________ -->
3081<div class="doc_subsubsection">
Reid Spencer24d6da52007-01-21 00:29:26 +0000[diff] [blame]3082 <a name="i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3083</div>
3084<div class="doc_text">
3085
3086<h5>Syntax:</h5>
3087<pre>
3088 &lt;result&gt; = fp2uint &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3089</pre>
3090
3091<h5>Overview:</h5>
3092<p>The '<tt>fp2uint</tt>' converts a floating point <tt>value</tt> to its
3093unsigned integer equivalent of type <tt>ty2</tt>.
3094</p>
3095
3096<h5>Arguments:</h5>
3097<p>The '<tt>fp2uint</tt>' instruction takes a value to cast, which must be a
3098<a href="#t_floating">floating point</a> value, and a type to cast it to, which
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]3099must be an <a href="#t_integer">integer</a> type.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3100
3101<h5>Semantics:</h5>
3102<p> The '<tt>fp2uint</tt>' instruction converts its
3103<a href="#t_floating">floating point</a> operand into the nearest (rounding
3104towards zero) unsigned integer value. If the value cannot fit in <tt>ty2</tt>,
3105the results are undefined.</p>
3106
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3107<p>When converting to i1, the conversion is done as a comparison against
3108zero. If the <tt>value</tt> was zero, the i1 result will be <tt>false</tt>.
3109If the <tt>value</tt> was non-zero, the i1 result will be <tt>true</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3110
3111<h5>Example:</h5>
3112<pre>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3113 %X = fp2uint double 123.0 to i32 <i>; yields i32:123</i>
3114 %Y = fp2uint float 1.0E+300 to i1 <i>; yields i1:true</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3115 %X = fp2uint float 1.04E+17 to i8 <i>; yields undefined:1</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3116</pre>
3117</div>
3118
3119<!-- _______________________________________________________________________ -->
3120<div class="doc_subsubsection">
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3121 <a name="i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3122</div>
3123<div class="doc_text">
3124
3125<h5>Syntax:</h5>
3126<pre>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3127 &lt;result&gt; = fptosi &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3128</pre>
3129
3130<h5>Overview:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3131<p>The '<tt>fptosi</tt>' instruction converts
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3132<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]3133</p>
3134
3135
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]3136<h5>Arguments:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3137<p> The '<tt>fptosi</tt>' instruction takes a value to cast, which must be a
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3138<a href="#t_floating">floating point</a> value, and a type to cast it to, which
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]3139must also be an <a href="#t_integer">integer</a> type.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3140
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]3141<h5>Semantics:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3142<p>The '<tt>fptosi</tt>' instruction converts its
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3143<a href="#t_floating">floating point</a> operand into the nearest (rounding
3144towards zero) signed integer value. If the value cannot fit in <tt>ty2</tt>,
3145the results are undefined.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3146
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3147<p>When converting to i1, the conversion is done as a comparison against
3148zero. If the <tt>value</tt> was zero, the i1 result will be <tt>false</tt>.
3149If the <tt>value</tt> was non-zero, the i1 result will be <tt>true</tt>.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3150
Chris Lattner33ba0d92001-07-09 00:26:23 +0000[diff] [blame]3151<h5>Example:</h5>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3152<pre>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3153 %X = fptosi double -123.0 to i32 <i>; yields i32:-123</i>
3154 %Y = fptosi float 1.0E-247 to i1 <i>; yields i1:true</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3155 %X = fptosi float 1.04E+17 to i8 <i>; yields undefined:1</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3156</pre>
3157</div>
3158
3159<!-- _______________________________________________________________________ -->
3160<div class="doc_subsubsection">
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3161 <a name="i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3162</div>
3163<div class="doc_text">
3164
3165<h5>Syntax:</h5>
3166<pre>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3167 &lt;result&gt; = uitofp &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3168</pre>
3169
3170<h5>Overview:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3171<p>The '<tt>uitofp</tt>' instruction regards <tt>value</tt> as an unsigned
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3172integer and converts that value to the <tt>ty2</tt> type.</p>
3173
3174
3175<h5>Arguments:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3176<p>The '<tt>uitofp</tt>' instruction takes a value to cast, which must be an
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]3177<a href="#t_integer">integer</a> value, and a type to cast it to, which must
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3178be a <a href="#t_floating">floating point</a> type.</p>
3179
3180<h5>Semantics:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3181<p>The '<tt>uitofp</tt>' instruction interprets its operand as an unsigned
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3182integer quantity and converts it to the corresponding floating point value. If
Jeff Cohencb757312007-04-22 14:56:37 +0000[diff] [blame]3183the value cannot fit in the floating point value, the results are undefined.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3184
3185
3186<h5>Example:</h5>
3187<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3188 %X = uitofp i32 257 to float <i>; yields float:257.0</i>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3189 %Y = uitofp i8 -1 to double <i>; yields double:255.0</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3190</pre>
3191</div>
3192
3193<!-- _______________________________________________________________________ -->
3194<div class="doc_subsubsection">
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3195 <a name="i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3196</div>
3197<div class="doc_text">
3198
3199<h5>Syntax:</h5>
3200<pre>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3201 &lt;result&gt; = sitofp &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3202</pre>
3203
3204<h5>Overview:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3205<p>The '<tt>sitofp</tt>' instruction regards <tt>value</tt> as a signed
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3206integer and converts that value to the <tt>ty2</tt> type.</p>
3207
3208<h5>Arguments:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3209<p>The '<tt>sitofp</tt>' instruction takes a value to cast, which must be an
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]3210<a href="#t_integer">integer</a> value, and a type to cast it to, which must be
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3211a <a href="#t_floating">floating point</a> type.</p>
3212
3213<h5>Semantics:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3214<p>The '<tt>sitofp</tt>' instruction interprets its operand as a signed
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3215integer quantity and converts it to the corresponding floating point value. If
Jeff Cohencb757312007-04-22 14:56:37 +0000[diff] [blame]3216the value cannot fit in the floating point value, the results are undefined.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3217
3218<h5>Example:</h5>
3219<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3220 %X = sitofp i32 257 to float <i>; yields float:257.0</i>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3221 %Y = sitofp i8 -1 to double <i>; yields double:-1.0</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3222</pre>
3223</div>
3224
3225<!-- _______________________________________________________________________ -->
3226<div class="doc_subsubsection">
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]3227 <a name="i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a>
3228</div>
3229<div class="doc_text">
3230
3231<h5>Syntax:</h5>
3232<pre>
3233 &lt;result&gt; = ptrtoint &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3234</pre>
3235
3236<h5>Overview:</h5>
3237<p>The '<tt>ptrtoint</tt>' instruction converts the pointer <tt>value</tt> to
3238the integer type <tt>ty2</tt>.</p>
3239
3240<h5>Arguments:</h5>
3241<p>The '<tt>ptrtoint</tt>' instruction takes a <tt>value</tt> to cast, which
Duncan Sands8036ca42007-03-30 12:22:09 +0000[diff] [blame]3242must be a <a href="#t_pointer">pointer</a> value, and a type to cast it to
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]3243<tt>ty2</tt>, which must be an <a href="#t_integer">integer</a> type.
3244
3245<h5>Semantics:</h5>
3246<p>The '<tt>ptrtoint</tt>' instruction converts <tt>value</tt> to integer type
3247<tt>ty2</tt> by interpreting the pointer value as an integer and either
3248truncating or zero extending that value to the size of the integer type. If
3249<tt>value</tt> is smaller than <tt>ty2</tt> then a zero extension is done. If
3250<tt>value</tt> is larger than <tt>ty2</tt> then a truncation is done. If they
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3251are the same size, then nothing is done (<i>no-op cast</i>) other than a type
3252change.</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]3253
3254<h5>Example:</h5>
3255<pre>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3256 %X = ptrtoint i32* %X to i8 <i>; yields truncation on 32-bit architecture</i>
3257 %Y = ptrtoint i32* %x to i64 <i>; yields zero extension on 32-bit architecture</i>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]3258</pre>
3259</div>
3260
3261<!-- _______________________________________________________________________ -->
3262<div class="doc_subsubsection">
3263 <a name="i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a>
3264</div>
3265<div class="doc_text">
3266
3267<h5>Syntax:</h5>
3268<pre>
3269 &lt;result&gt; = inttoptr &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3270</pre>
3271
3272<h5>Overview:</h5>
3273<p>The '<tt>inttoptr</tt>' instruction converts an integer <tt>value</tt> to
3274a pointer type, <tt>ty2</tt>.</p>
3275
3276<h5>Arguments:</h5>
Duncan Sands8036ca42007-03-30 12:22:09 +0000[diff] [blame]3277<p>The '<tt>inttoptr</tt>' instruction takes an <a href="#t_integer">integer</a>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]3278value to cast, and a type to cast it to, which must be a
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame]3279<a href="#t_pointer">pointer</a> type.
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]3280
3281<h5>Semantics:</h5>
3282<p>The '<tt>inttoptr</tt>' instruction converts <tt>value</tt> to type
3283<tt>ty2</tt> by applying either a zero extension or a truncation depending on
3284the size of the integer <tt>value</tt>. If <tt>value</tt> is larger than the
3285size of a pointer then a truncation is done. If <tt>value</tt> is smaller than
3286the size of a pointer then a zero extension is done. If they are the same size,
3287nothing is done (<i>no-op cast</i>).</p>
3288
3289<h5>Example:</h5>
3290<pre>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3291 %X = inttoptr i32 255 to i32* <i>; yields zero extension on 64-bit architecture</i>
3292 %X = inttoptr i32 255 to i32* <i>; yields no-op on 32-bit architecture</i>
3293 %Y = inttoptr i64 0 to i32* <i>; yields truncation on 32-bit architecture</i>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]3294</pre>
3295</div>
3296
3297<!-- _______________________________________________________________________ -->
3298<div class="doc_subsubsection">
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]3299 <a name="i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3300</div>
3301<div class="doc_text">
3302
3303<h5>Syntax:</h5>
3304<pre>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]3305 &lt;result&gt; = bitcast &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3306</pre>
3307
3308<h5>Overview:</h5>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]3309<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3310<tt>ty2</tt> without changing any bits.</p>
3311
3312<h5>Arguments:</h5>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]3313<p>The '<tt>bitcast</tt>' instruction takes a value to cast, which must be
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3314a first class value, and a type to cast it to, which must also be a <a
3315 href="#t_firstclass">first class</a> type. The bit sizes of <tt>value</tt>
Reid Spencer19b569f2007-01-09 20:08:58 +0000[diff] [blame]3316and the destination type, <tt>ty2</tt>, must be identical. If the source
3317type is a pointer, the destination type must also be a pointer.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3318
3319<h5>Semantics:</h5>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]3320<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]3321<tt>ty2</tt>. It is always a <i>no-op cast</i> because no bits change with
3322this conversion. The conversion is done as if the <tt>value</tt> had been
3323stored to memory and read back as type <tt>ty2</tt>. Pointer types may only be
3324converted to other pointer types with this instruction. To convert pointers to
3325other types, use the <a href="#i_inttoptr">inttoptr</a> or
3326<a href="#i_ptrtoint">ptrtoint</a> instructions first.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3327
3328<h5>Example:</h5>
3329<pre>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3330 %X = bitcast i8 255 to i8 <i>; yields i8 :-1</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3331 %Y = bitcast i32* %x to sint* <i>; yields sint*:%x</i>
3332 %Z = bitcast <2xint> %V to i64; <i>; yields i64: %V</i>
Chris Lattner33ba0d92001-07-09 00:26:23 +0000[diff] [blame]3333</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3334</div>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3335
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]3336<!-- ======================================================================= -->
3337<div class="doc_subsection"> <a name="otherops">Other Operations</a> </div>
3338<div class="doc_text">
3339<p>The instructions in this category are the "miscellaneous"
3340instructions, which defy better classification.</p>
3341</div>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3342
3343<!-- _______________________________________________________________________ -->
3344<div class="doc_subsubsection"><a name="i_icmp">'<tt>icmp</tt>' Instruction</a>
3345</div>
3346<div class="doc_text">
3347<h5>Syntax:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3348<pre> &lt;result&gt; = icmp &lt;cond&gt; &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {i1}:result</i>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3349</pre>
3350<h5>Overview:</h5>
3351<p>The '<tt>icmp</tt>' instruction returns a boolean value based on comparison
3352of its two integer operands.</p>
3353<h5>Arguments:</h5>
3354<p>The '<tt>icmp</tt>' instruction takes three operands. The first operand is
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3355the condition code indicating the kind of comparison to perform. It is not
3356a value, just a keyword. The possible condition code are:
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3357<ol>
3358 <li><tt>eq</tt>: equal</li>
3359 <li><tt>ne</tt>: not equal </li>
3360 <li><tt>ugt</tt>: unsigned greater than</li>
3361 <li><tt>uge</tt>: unsigned greater or equal</li>
3362 <li><tt>ult</tt>: unsigned less than</li>
3363 <li><tt>ule</tt>: unsigned less or equal</li>
3364 <li><tt>sgt</tt>: signed greater than</li>
3365 <li><tt>sge</tt>: signed greater or equal</li>
3366 <li><tt>slt</tt>: signed less than</li>
3367 <li><tt>sle</tt>: signed less or equal</li>
3368</ol>
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]3369<p>The remaining two arguments must be <a href="#t_integer">integer</a> or
Reid Spencer350f8aa2007-01-04 05:19:58 +0000[diff] [blame]3370<a href="#t_pointer">pointer</a> typed. They must also be identical types.</p>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3371<h5>Semantics:</h5>
3372<p>The '<tt>icmp</tt>' compares <tt>var1</tt> and <tt>var2</tt> according to
3373the condition code given as <tt>cond</tt>. The comparison performed always
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3374yields a <a href="#t_primitive">i1</a> result, as follows:
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3375<ol>
3376 <li><tt>eq</tt>: yields <tt>true</tt> if the operands are equal,
3377 <tt>false</tt> otherwise. No sign interpretation is necessary or performed.
3378 </li>
3379 <li><tt>ne</tt>: yields <tt>true</tt> if the operands are unequal,
3380 <tt>false</tt> otherwise. No sign interpretation is necessary or performed.
3381 <li><tt>ugt</tt>: interprets the operands as unsigned values and yields
3382 <tt>true</tt> if <tt>var1</tt> is greater than <tt>var2</tt>.</li>
3383 <li><tt>uge</tt>: interprets the operands as unsigned values and yields
3384 <tt>true</tt> if <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
3385 <li><tt>ult</tt>: interprets the operands as unsigned values and yields
3386 <tt>true</tt> if <tt>var1</tt> is less than <tt>var2</tt>.</li>
3387 <li><tt>ule</tt>: interprets the operands as unsigned values and yields
3388 <tt>true</tt> if <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
3389 <li><tt>sgt</tt>: interprets the operands as signed values and yields
3390 <tt>true</tt> if <tt>var1</tt> is greater than <tt>var2</tt>.</li>
3391 <li><tt>sge</tt>: interprets the operands as signed values and yields
3392 <tt>true</tt> if <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
3393 <li><tt>slt</tt>: interprets the operands as signed values and yields
3394 <tt>true</tt> if <tt>var1</tt> is less than <tt>var2</tt>.</li>
3395 <li><tt>sle</tt>: interprets the operands as signed values and yields
3396 <tt>true</tt> if <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3397</ol>
3398<p>If the operands are <a href="#t_pointer">pointer</a> typed, the pointer
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3399values are compared as if they were integers.</p>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3400
3401<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3402<pre> &lt;result&gt; = icmp eq i32 4, 5 <i>; yields: result=false</i>
3403 &lt;result&gt; = icmp ne float* %X, %X <i>; yields: result=false</i>
3404 &lt;result&gt; = icmp ult i16 4, 5 <i>; yields: result=true</i>
3405 &lt;result&gt; = icmp sgt i16 4, 5 <i>; yields: result=false</i>
3406 &lt;result&gt; = icmp ule i16 -4, 5 <i>; yields: result=false</i>
3407 &lt;result&gt; = icmp sge i16 4, 5 <i>; yields: result=false</i>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3408</pre>
3409</div>
3410
3411<!-- _______________________________________________________________________ -->
3412<div class="doc_subsubsection"><a name="i_fcmp">'<tt>fcmp</tt>' Instruction</a>
3413</div>
3414<div class="doc_text">
3415<h5>Syntax:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3416<pre> &lt;result&gt; = fcmp &lt;cond&gt; &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {i1}:result</i>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3417</pre>
3418<h5>Overview:</h5>
3419<p>The '<tt>fcmp</tt>' instruction returns a boolean value based on comparison
3420of its floating point operands.</p>
3421<h5>Arguments:</h5>
3422<p>The '<tt>fcmp</tt>' instruction takes three operands. The first operand is
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3423the condition code indicating the kind of comparison to perform. It is not
3424a value, just a keyword. The possible condition code are:
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3425<ol>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3426 <li><tt>false</tt>: no comparison, always returns false</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3427 <li><tt>oeq</tt>: ordered and equal</li>
3428 <li><tt>ogt</tt>: ordered and greater than </li>
3429 <li><tt>oge</tt>: ordered and greater than or equal</li>
3430 <li><tt>olt</tt>: ordered and less than </li>
3431 <li><tt>ole</tt>: ordered and less than or equal</li>
3432 <li><tt>one</tt>: ordered and not equal</li>
3433 <li><tt>ord</tt>: ordered (no nans)</li>
3434 <li><tt>ueq</tt>: unordered or equal</li>
3435 <li><tt>ugt</tt>: unordered or greater than </li>
3436 <li><tt>uge</tt>: unordered or greater than or equal</li>
3437 <li><tt>ult</tt>: unordered or less than </li>
3438 <li><tt>ule</tt>: unordered or less than or equal</li>
3439 <li><tt>une</tt>: unordered or not equal</li>
3440 <li><tt>uno</tt>: unordered (either nans)</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3441 <li><tt>true</tt>: no comparison, always returns true</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3442</ol>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3443<p><i>Ordered</i> means that neither operand is a QNAN while
Reid Spencer93a49852006-12-06 07:08:07 +0000[diff] [blame]3444<i>unordered</i> means that either operand may be a QNAN.</p>
Reid Spencer350f8aa2007-01-04 05:19:58 +0000[diff] [blame]3445<p>The <tt>val1</tt> and <tt>val2</tt> arguments must be
3446<a href="#t_floating">floating point</a> typed. They must have identical
3447types.</p>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3448<h5>Semantics:</h5>
3449<p>The '<tt>fcmp</tt>' compares <tt>var1</tt> and <tt>var2</tt> according to
3450the condition code given as <tt>cond</tt>. The comparison performed always
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3451yields a <a href="#t_primitive">i1</a> result, as follows:
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3452<ol>
3453 <li><tt>false</tt>: always yields <tt>false</tt>, regardless of operands.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3454 <li><tt>oeq</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3455 <tt>var1</tt> is equal to <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3456 <li><tt>ogt</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3457 <tt>var1</tt> is greather than <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3458 <li><tt>oge</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3459 <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3460 <li><tt>olt</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3461 <tt>var1</tt> is less than <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3462 <li><tt>ole</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3463 <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3464 <li><tt>one</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3465 <tt>var1</tt> is not equal to <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3466 <li><tt>ord</tt>: yields <tt>true</tt> if both operands are not a QNAN.</li>
3467 <li><tt>ueq</tt>: yields <tt>true</tt> if either operand is a QNAN or
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3468 <tt>var1</tt> is equal to <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3469 <li><tt>ugt</tt>: yields <tt>true</tt> if either operand is a QNAN or
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3470 <tt>var1</tt> is greater than <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3471 <li><tt>uge</tt>: yields <tt>true</tt> if either operand is a QNAN or
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3472 <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3473 <li><tt>ult</tt>: yields <tt>true</tt> if either operand is a QNAN or
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3474 <tt>var1</tt> is less than <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3475 <li><tt>ule</tt>: yields <tt>true</tt> if either operand is a QNAN or
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3476 <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3477 <li><tt>une</tt>: yields <tt>true</tt> if either operand is a QNAN or
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3478 <tt>var1</tt> is not equal to <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3479 <li><tt>uno</tt>: yields <tt>true</tt> if either operand is a QNAN.</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3480 <li><tt>true</tt>: always yields <tt>true</tt>, regardless of operands.</li>
3481</ol>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3482
3483<h5>Example:</h5>
3484<pre> &lt;result&gt; = fcmp oeq float 4.0, 5.0 <i>; yields: result=false</i>
3485 &lt;result&gt; = icmp one float 4.0, 5.0 <i>; yields: result=true</i>
3486 &lt;result&gt; = icmp olt float 4.0, 5.0 <i>; yields: result=true</i>
3487 &lt;result&gt; = icmp ueq double 1.0, 2.0 <i>; yields: result=false</i>
3488</pre>
3489</div>
3490
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]3491<!-- _______________________________________________________________________ -->
3492<div class="doc_subsubsection"> <a name="i_phi">'<tt>phi</tt>'
3493Instruction</a> </div>
3494<div class="doc_text">
3495<h5>Syntax:</h5>
3496<pre> &lt;result&gt; = phi &lt;ty&gt; [ &lt;val0&gt;, &lt;label0&gt;], ...<br></pre>
3497<h5>Overview:</h5>
3498<p>The '<tt>phi</tt>' instruction is used to implement the &#966; node in
3499the SSA graph representing the function.</p>
3500<h5>Arguments:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3501<p>The type of the incoming values is specified with the first type
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]3502field. After this, the '<tt>phi</tt>' instruction takes a list of pairs
3503as arguments, with one pair for each predecessor basic block of the
3504current block. Only values of <a href="#t_firstclass">first class</a>
3505type may be used as the value arguments to the PHI node. Only labels
3506may be used as the label arguments.</p>
3507<p>There must be no non-phi instructions between the start of a basic
3508block and the PHI instructions: i.e. PHI instructions must be first in
3509a basic block.</p>
3510<h5>Semantics:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3511<p>At runtime, the '<tt>phi</tt>' instruction logically takes on the value
3512specified by the pair corresponding to the predecessor basic block that executed
3513just prior to the current block.</p>
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]3514<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3515<pre>Loop: ; Infinite loop that counts from 0 on up...<br> %indvar = phi i32 [ 0, %LoopHeader ], [ %nextindvar, %Loop ]<br> %nextindvar = add i32 %indvar, 1<br> br label %Loop<br></pre>
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]3516</div>
3517
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3518<!-- _______________________________________________________________________ -->
3519<div class="doc_subsubsection">
3520 <a name="i_select">'<tt>select</tt>' Instruction</a>
3521</div>
3522
3523<div class="doc_text">
3524
3525<h5>Syntax:</h5>
3526
3527<pre>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3528 &lt;result&gt; = select i1 &lt;cond&gt;, &lt;ty&gt; &lt;val1&gt;, &lt;ty&gt; &lt;val2&gt; <i>; yields ty</i>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3529</pre>
3530
3531<h5>Overview:</h5>
3532
3533<p>
3534The '<tt>select</tt>' instruction is used to choose one value based on a
3535condition, without branching.
3536</p>
3537
3538
3539<h5>Arguments:</h5>
3540
3541<p>
3542The '<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.
3543</p>
3544
3545<h5>Semantics:</h5>
3546
3547<p>
3548If the boolean condition evaluates to true, the instruction returns the first
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3549value argument; otherwise, it returns the second value argument.
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3550</p>
3551
3552<h5>Example:</h5>
3553
3554<pre>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3555 %X = select i1 true, i8 17, i8 42 <i>; yields i8:17</i>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3556</pre>
3557</div>
3558
Robert Bocchino05ccd702006-01-15 20:48:27 +0000[diff] [blame]3559
3560<!-- _______________________________________________________________________ -->
3561<div class="doc_subsubsection">
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3562 <a name="i_call">'<tt>call</tt>' Instruction</a>
3563</div>
3564
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3565<div class="doc_text">
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3566
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3567<h5>Syntax:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3568<pre>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3569 &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]3570</pre>
3571
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3572<h5>Overview:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3573
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3574<p>The '<tt>call</tt>' instruction represents a simple function call.</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3575
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3576<h5>Arguments:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3577
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3578<p>This instruction requires several arguments:</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3579
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]3580<ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3581 <li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3582 <p>The optional "tail" marker indicates whether the callee function accesses
3583 any allocas or varargs in the caller. If the "tail" marker is present, the
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3584 function call is eligible for tail call optimization. Note that calls may
3585 be marked "tail" even if they do not occur before a <a
3586 href="#i_ret"><tt>ret</tt></a> instruction.
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3587 </li>
3588 <li>
Duncan Sands8036ca42007-03-30 12:22:09 +0000[diff] [blame]3589 <p>The optional "cconv" marker indicates which <a href="#callingconv">calling
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3590 convention</a> the call should use. If none is specified, the call defaults
3591 to using C calling conventions.
3592 </li>
3593 <li>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3594 <p>'<tt>ty</tt>': shall be the signature of the pointer to function value
3595 being invoked. The argument types must match the types implied by this
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3596 signature. This type can be omitted if the function is not varargs and
3597 if the function type does not return a pointer to a function.</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3598 </li>
3599 <li>
3600 <p>'<tt>fnptrval</tt>': An LLVM value containing a pointer to a function to
3601 be invoked. In most cases, this is a direct function invocation, but
3602 indirect <tt>call</tt>s are just as possible, calling an arbitrary pointer
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3603 to function value.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3604 </li>
3605 <li>
3606 <p>'<tt>function args</tt>': argument list whose types match the
Reid Spencera7e302a2005-05-01 22:22:57 +0000[diff] [blame]3607 function signature argument types. All arguments must be of
3608 <a href="#t_firstclass">first class</a> type. If the function signature
3609 indicates the function accepts a variable number of arguments, the extra
3610 arguments can be specified.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3611 </li>
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]3612</ol>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3613
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3614<h5>Semantics:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3615
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3616<p>The '<tt>call</tt>' instruction is used to cause control flow to
3617transfer to a specified function, with its incoming arguments bound to
3618the specified values. Upon a '<tt><a href="#i_ret">ret</a></tt>'
3619instruction in the called function, control flow continues with the
3620instruction after the function call, and the return value of the
3621function is bound to the result argument. This is a simpler case of
3622the <a href="#i_invoke">invoke</a> instruction.</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3623
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3624<h5>Example:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3625
3626<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3627 %retval = call i32 %test(i32 %argc)
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3628 call i32(i8 *, ...) *%printf(i8 * %msg, i32 12, i8 42);
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3629 %X = tail call i32 %foo()
3630 %Y = tail call <a href="#callingconv">fastcc</a> i32 %foo()
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3631</pre>
3632
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3633</div>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3634
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3635<!-- _______________________________________________________________________ -->
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3636<div class="doc_subsubsection">
Chris Lattnerfb6977d2006-01-13 23:26:01 +0000[diff] [blame]3637 <a name="i_va_arg">'<tt>va_arg</tt>' Instruction</a>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3638</div>
3639
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3640<div class="doc_text">
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3641
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]3642<h5>Syntax:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3643
3644<pre>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3645 &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]3646</pre>
3647
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]3648<h5>Overview:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3649
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3650<p>The '<tt>va_arg</tt>' instruction is used to access arguments passed through
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3651the "variable argument" area of a function call. It is used to implement the
3652<tt>va_arg</tt> macro in C.</p>
3653
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]3654<h5>Arguments:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3655
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3656<p>This instruction takes a <tt>va_list*</tt> value and the type of
3657the argument. It returns a value of the specified argument type and
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3658increments the <tt>va_list</tt> to point to the next argument. The
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3659actual type of <tt>va_list</tt> is target specific.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3660
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]3661<h5>Semantics:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3662
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3663<p>The '<tt>va_arg</tt>' instruction loads an argument of the specified
3664type from the specified <tt>va_list</tt> and causes the
3665<tt>va_list</tt> to point to the next argument. For more information,
3666see the variable argument handling <a href="#int_varargs">Intrinsic
3667Functions</a>.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3668
3669<p>It is legal for this instruction to be called in a function which does not
3670take a variable number of arguments, for example, the <tt>vfprintf</tt>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3671function.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3672
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3673<p><tt>va_arg</tt> is an LLVM instruction instead of an <a
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3674href="#intrinsics">intrinsic function</a> because it takes a type as an
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3675argument.</p>
3676
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]3677<h5>Example:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3678
3679<p>See the <a href="#int_varargs">variable argument processing</a> section.</p>
3680
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3681</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3682
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3683<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3684<div class="doc_section"> <a name="intrinsics">Intrinsic Functions</a> </div>
3685<!-- *********************************************************************** -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3686
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3687<div class="doc_text">
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3688
3689<p>LLVM supports the notion of an "intrinsic function". These functions have
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]3690well known names and semantics and are required to follow certain restrictions.
3691Overall, these intrinsics represent an extension mechanism for the LLVM
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3692language that does not require changing all of the transformations in LLVM when
3693adding to the language (or the bytecode reader/writer, the parser, etc...).</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3694
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3695<p>Intrinsic function names must all start with an "<tt>llvm.</tt>" prefix. This
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3696prefix is reserved in LLVM for intrinsic names; thus, function names may not
3697begin with this prefix. Intrinsic functions must always be external functions:
3698you cannot define the body of intrinsic functions. Intrinsic functions may
3699only be used in call or invoke instructions: it is illegal to take the address
3700of an intrinsic function. Additionally, because intrinsic functions are part
3701of the LLVM language, it is required if any are added that they be documented
3702here.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3703
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3704<p>Some intrinsic functions can be overloaded, i.e., the intrinsic represents
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]3705a family of functions that perform the same operation but on different data
3706types. This is most frequent with the integer types. Since LLVM can represent
3707over 8 million different integer types, there is a way to declare an intrinsic
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3708that can be overloaded based on its arguments. Such an intrinsic will have the
3709names of its argument types encoded into its function name, each
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]3710preceded by a period. For example, the <tt>llvm.ctpop</tt> function can take an
3711integer of any width. This leads to a family of functions such as
3712<tt>i32 @llvm.ctpop.i8(i8 %val)</tt> and <tt>i32 @llvm.ctpop.i29(i29 %val)</tt>.
3713</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3714
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]3715
3716<p>To learn how to add an intrinsic function, please see the
3717<a href="ExtendingLLVM.html">Extending LLVM Guide</a>.
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3718</p>
3719
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3720</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3721
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3722<!-- ======================================================================= -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3723<div class="doc_subsection">
3724 <a name="int_varargs">Variable Argument Handling Intrinsics</a>
3725</div>
3726
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3727<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3728
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3729<p>Variable argument support is defined in LLVM with the <a
Chris Lattnerfb6977d2006-01-13 23:26:01 +0000[diff] [blame]3730 href="#i_va_arg"><tt>va_arg</tt></a> instruction and these three
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3731intrinsic functions. These functions are related to the similarly
3732named macros defined in the <tt>&lt;stdarg.h&gt;</tt> header file.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3733
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3734<p>All of these functions operate on arguments that use a
3735target-specific value type "<tt>va_list</tt>". The LLVM assembly
3736language reference manual does not define what this type is, so all
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3737transformations should be prepared to handle these functions regardless of
3738the type used.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3739
Chris Lattner374ab302006-05-15 17:26:46 +0000[diff] [blame]3740<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]3741instruction and the variable argument handling intrinsic functions are
3742used.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3743
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]3744<div class="doc_code">
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3745<pre>
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]3746define i32 @test(i32 %X, ...) {
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3747 ; Initialize variable argument processing
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3748 %ap = alloca i8*
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]3749 %ap2 = bitcast i8** %ap to i8*
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]3750 call void @llvm.va_start(i8* %ap2)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3751
3752 ; Read a single integer argument
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3753 %tmp = va_arg i8** %ap, i32
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3754
3755 ; Demonstrate usage of llvm.va_copy and llvm.va_end
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3756 %aq = alloca i8*
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]3757 %aq2 = bitcast i8** %aq to i8*
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3758 call void @llvm.va_copy(i8* %aq2, i8* %ap2)
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]3759 call void @llvm.va_end(i8* %aq2)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3760
3761 ; Stop processing of arguments.
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]3762 call void @llvm.va_end(i8* %ap2)
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3763 ret i32 %tmp
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3764}
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]3765
3766declare void @llvm.va_start(i8*)
3767declare void @llvm.va_copy(i8*, i8*)
3768declare void @llvm.va_end(i8*)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3769</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3770</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3771
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]3772</div>
3773
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3774<!-- _______________________________________________________________________ -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3775<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]3776 <a name="int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3777</div>
3778
3779
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3780<div class="doc_text">
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3781<h5>Syntax:</h5>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]3782<pre> declare void %llvm.va_start(i8* &lt;arglist&gt;)<br></pre>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3783<h5>Overview:</h5>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3784<P>The '<tt>llvm.va_start</tt>' intrinsic initializes
3785<tt>*&lt;arglist&gt;</tt> for subsequent use by <tt><a
3786href="#i_va_arg">va_arg</a></tt>.</p>
3787
3788<h5>Arguments:</h5>
3789
3790<P>The argument is a pointer to a <tt>va_list</tt> element to initialize.</p>
3791
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3792<h5>Semantics:</h5>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3793
3794<P>The '<tt>llvm.va_start</tt>' intrinsic works just like the <tt>va_start</tt>
3795macro available in C. In a target-dependent way, it initializes the
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3796<tt>va_list</tt> element to which the argument points, so that the next call to
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3797<tt>va_arg</tt> will produce the first variable argument passed to the function.
3798Unlike the C <tt>va_start</tt> macro, this intrinsic does not need to know the
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3799last argument of the function as the compiler can figure that out.</p>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3800
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3801</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3802
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3803<!-- _______________________________________________________________________ -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3804<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]3805 <a name="int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3806</div>
3807
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3808<div class="doc_text">
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3809<h5>Syntax:</h5>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]3810<pre> declare void @llvm.va_end(i8* &lt;arglist&gt;)<br></pre>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3811<h5>Overview:</h5>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]3812
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3813<p>The '<tt>llvm.va_end</tt>' intrinsic destroys <tt>*&lt;arglist&gt;</tt>,
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]3814which has been initialized previously with <tt><a href="#int_va_start">llvm.va_start</a></tt>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3815or <tt><a href="#i_va_copy">llvm.va_copy</a></tt>.</p>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]3816
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3817<h5>Arguments:</h5>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]3818
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3819<p>The argument is a pointer to a <tt>va_list</tt> to destroy.</p>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]3820
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3821<h5>Semantics:</h5>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]3822
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3823<p>The '<tt>llvm.va_end</tt>' intrinsic works just like the <tt>va_end</tt>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3824macro available in C. In a target-dependent way, it destroys the
3825<tt>va_list</tt> element to which the argument points. Calls to <a
3826href="#int_va_start"><tt>llvm.va_start</tt></a> and <a href="#int_va_copy">
3827<tt>llvm.va_copy</tt></a> must be matched exactly with calls to
3828<tt>llvm.va_end</tt>.</p>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]3829
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3830</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3831
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3832<!-- _______________________________________________________________________ -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3833<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]3834 <a name="int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3835</div>
3836
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3837<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3838
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3839<h5>Syntax:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3840
3841<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]3842 declare void @llvm.va_copy(i8* &lt;destarglist&gt;, i8* &lt;srcarglist&gt;)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3843</pre>
3844
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3845<h5>Overview:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3846
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3847<p>The '<tt>llvm.va_copy</tt>' intrinsic copies the current argument position
3848from the source argument list to the destination argument list.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3849
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3850<h5>Arguments:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3851
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3852<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]3853The 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]3854
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3855
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3856<h5>Semantics:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3857
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3858<p>The '<tt>llvm.va_copy</tt>' intrinsic works just like the <tt>va_copy</tt>
3859macro available in C. In a target-dependent way, it copies the source
3860<tt>va_list</tt> element into the destination <tt>va_list</tt> element. This
3861intrinsic is necessary because the <tt><a href="#int_va_start">
3862llvm.va_start</a></tt> intrinsic may be arbitrarily complex and require, for
3863example, memory allocation.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3864
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3865</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3866
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3867<!-- ======================================================================= -->
3868<div class="doc_subsection">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3869 <a name="int_gc">Accurate Garbage Collection Intrinsics</a>
3870</div>
3871
3872<div class="doc_text">
3873
3874<p>
3875LLVM support for <a href="GarbageCollection.html">Accurate Garbage
3876Collection</a> requires the implementation and generation of these intrinsics.
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]3877These intrinsics allow identification of <a href="#int_gcroot">GC roots on the
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3878stack</a>, as well as garbage collector implementations that require <a
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]3879href="#int_gcread">read</a> and <a href="#int_gcwrite">write</a> barriers.
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3880Front-ends for type-safe garbage collected languages should generate these
3881intrinsics to make use of the LLVM garbage collectors. For more details, see <a
3882href="GarbageCollection.html">Accurate Garbage Collection with LLVM</a>.
3883</p>
3884</div>
3885
3886<!-- _______________________________________________________________________ -->
3887<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]3888 <a name="int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3889</div>
3890
3891<div class="doc_text">
3892
3893<h5>Syntax:</h5>
3894
3895<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]3896 declare void @llvm.gcroot(&lt;ty&gt;** %ptrloc, &lt;ty2&gt;* %metadata)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3897</pre>
3898
3899<h5>Overview:</h5>
3900
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]3901<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]3902the code generator, and allows some metadata to be associated with it.</p>
3903
3904<h5>Arguments:</h5>
3905
3906<p>The first argument specifies the address of a stack object that contains the
3907root pointer. The second pointer (which must be either a constant or a global
3908value address) contains the meta-data to be associated with the root.</p>
3909
3910<h5>Semantics:</h5>
3911
3912<p>At runtime, a call to this intrinsics stores a null pointer into the "ptrloc"
3913location. At compile-time, the code generator generates information to allow
3914the runtime to find the pointer at GC safe points.
3915</p>
3916
3917</div>
3918
3919
3920<!-- _______________________________________________________________________ -->
3921<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]3922 <a name="int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3923</div>
3924
3925<div class="doc_text">
3926
3927<h5>Syntax:</h5>
3928
3929<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]3930 declare i8 * @llvm.gcread(i8 * %ObjPtr, i8 ** %Ptr)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3931</pre>
3932
3933<h5>Overview:</h5>
3934
3935<p>The '<tt>llvm.gcread</tt>' intrinsic identifies reads of references from heap
3936locations, allowing garbage collector implementations that require read
3937barriers.</p>
3938
3939<h5>Arguments:</h5>
3940
Chris Lattner80626e92006-03-14 20:02:51 +0000[diff] [blame]3941<p>The second argument is the address to read from, which should be an address
3942allocated from the garbage collector. The first object is a pointer to the
3943start of the referenced object, if needed by the language runtime (otherwise
3944null).</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3945
3946<h5>Semantics:</h5>
3947
3948<p>The '<tt>llvm.gcread</tt>' intrinsic has the same semantics as a load
3949instruction, but may be replaced with substantially more complex code by the
3950garbage collector runtime, as needed.</p>
3951
3952</div>
3953
3954
3955<!-- _______________________________________________________________________ -->
3956<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]3957 <a name="int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3958</div>
3959
3960<div class="doc_text">
3961
3962<h5>Syntax:</h5>
3963
3964<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]3965 declare void @llvm.gcwrite(i8 * %P1, i8 * %Obj, i8 ** %P2)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3966</pre>
3967
3968<h5>Overview:</h5>
3969
3970<p>The '<tt>llvm.gcwrite</tt>' intrinsic identifies writes of references to heap
3971locations, allowing garbage collector implementations that require write
3972barriers (such as generational or reference counting collectors).</p>
3973
3974<h5>Arguments:</h5>
3975
Chris Lattner80626e92006-03-14 20:02:51 +0000[diff] [blame]3976<p>The first argument is the reference to store, the second is the start of the
3977object to store it to, and the third is the address of the field of Obj to
3978store to. If the runtime does not require a pointer to the object, Obj may be
3979null.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3980
3981<h5>Semantics:</h5>
3982
3983<p>The '<tt>llvm.gcwrite</tt>' intrinsic has the same semantics as a store
3984instruction, but may be replaced with substantially more complex code by the
3985garbage collector runtime, as needed.</p>
3986
3987</div>
3988
3989
3990
3991<!-- ======================================================================= -->
3992<div class="doc_subsection">
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3993 <a name="int_codegen">Code Generator Intrinsics</a>
3994</div>
3995
3996<div class="doc_text">
3997<p>
3998These intrinsics are provided by LLVM to expose special features that may only
3999be implemented with code generator support.
4000</p>
4001
4002</div>
4003
4004<!-- _______________________________________________________________________ -->
4005<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4006 <a name="int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4007</div>
4008
4009<div class="doc_text">
4010
4011<h5>Syntax:</h5>
4012<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4013 declare i8 *@llvm.returnaddress(i32 &lt;level&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4014</pre>
4015
4016<h5>Overview:</h5>
4017
4018<p>
Chris Lattner32b5d712006-10-15 20:05:59 +0000[diff] [blame]4019The '<tt>llvm.returnaddress</tt>' intrinsic attempts to compute a
4020target-specific value indicating the return address of the current function
4021or one of its callers.
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4022</p>
4023
4024<h5>Arguments:</h5>
4025
4026<p>
4027The argument to this intrinsic indicates which function to return the address
4028for. Zero indicates the calling function, one indicates its caller, etc. The
4029argument is <b>required</b> to be a constant integer value.
4030</p>
4031
4032<h5>Semantics:</h5>
4033
4034<p>
4035The '<tt>llvm.returnaddress</tt>' intrinsic either returns a pointer indicating
4036the return address of the specified call frame, or zero if it cannot be
4037identified. The value returned by this intrinsic is likely to be incorrect or 0
4038for arguments other than zero, so it should only be used for debugging purposes.
4039</p>
4040
4041<p>
4042Note that calling this intrinsic does not prevent function inlining or other
Chris Lattnerb40bb382005-03-07 20:30:51 +0000[diff] [blame]4043aggressive transformations, so the value returned may not be that of the obvious
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4044source-language caller.
4045</p>
4046</div>
4047
4048
4049<!-- _______________________________________________________________________ -->
4050<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4051 <a name="int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4052</div>
4053
4054<div class="doc_text">
4055
4056<h5>Syntax:</h5>
4057<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4058 declare i8 *@llvm.frameaddress(i32 &lt;level&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4059</pre>
4060
4061<h5>Overview:</h5>
4062
4063<p>
Chris Lattner32b5d712006-10-15 20:05:59 +0000[diff] [blame]4064The '<tt>llvm.frameaddress</tt>' intrinsic attempts to return the
4065target-specific frame pointer value for the specified stack frame.
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4066</p>
4067
4068<h5>Arguments:</h5>
4069
4070<p>
4071The argument to this intrinsic indicates which function to return the frame
4072pointer for. Zero indicates the calling function, one indicates its caller,
4073etc. The argument is <b>required</b> to be a constant integer value.
4074</p>
4075
4076<h5>Semantics:</h5>
4077
4078<p>
4079The '<tt>llvm.frameaddress</tt>' intrinsic either returns a pointer indicating
4080the frame address of the specified call frame, or zero if it cannot be
4081identified. The value returned by this intrinsic is likely to be incorrect or 0
4082for arguments other than zero, so it should only be used for debugging purposes.
4083</p>
4084
4085<p>
4086Note that calling this intrinsic does not prevent function inlining or other
Chris Lattnerb40bb382005-03-07 20:30:51 +0000[diff] [blame]4087aggressive transformations, so the value returned may not be that of the obvious
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4088source-language caller.
4089</p>
4090</div>
4091
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]4092<!-- _______________________________________________________________________ -->
4093<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4094 <a name="int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]4095</div>
4096
4097<div class="doc_text">
4098
4099<h5>Syntax:</h5>
4100<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4101 declare i8 *@llvm.stacksave()
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]4102</pre>
4103
4104<h5>Overview:</h5>
4105
4106<p>
4107The '<tt>llvm.stacksave</tt>' intrinsic is used to remember the current state of
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4108the function stack, for use with <a href="#int_stackrestore">
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]4109<tt>llvm.stackrestore</tt></a>. This is useful for implementing language
4110features like scoped automatic variable sized arrays in C99.
4111</p>
4112
4113<h5>Semantics:</h5>
4114
4115<p>
4116This intrinsic returns a opaque pointer value that can be passed to <a
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4117href="#int_stackrestore"><tt>llvm.stackrestore</tt></a>. When an
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]4118<tt>llvm.stackrestore</tt> intrinsic is executed with a value saved from
4119<tt>llvm.stacksave</tt>, it effectively restores the state of the stack to the
4120state it was in when the <tt>llvm.stacksave</tt> intrinsic executed. In
4121practice, this pops any <a href="#i_alloca">alloca</a> blocks from the stack
4122that were allocated after the <tt>llvm.stacksave</tt> was executed.
4123</p>
4124
4125</div>
4126
4127<!-- _______________________________________________________________________ -->
4128<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4129 <a name="int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]4130</div>
4131
4132<div class="doc_text">
4133
4134<h5>Syntax:</h5>
4135<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4136 declare void @llvm.stackrestore(i8 * %ptr)
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]4137</pre>
4138
4139<h5>Overview:</h5>
4140
4141<p>
4142The '<tt>llvm.stackrestore</tt>' intrinsic is used to restore the state of
4143the function stack to the state it was in when the corresponding <a
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4144href="#int_stacksave"><tt>llvm.stacksave</tt></a> intrinsic executed. This is
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]4145useful for implementing language features like scoped automatic variable sized
4146arrays in C99.
4147</p>
4148
4149<h5>Semantics:</h5>
4150
4151<p>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4152See the description for <a href="#int_stacksave"><tt>llvm.stacksave</tt></a>.
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]4153</p>
4154
4155</div>
4156
4157
4158<!-- _______________________________________________________________________ -->
4159<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4160 <a name="int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a>
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]4161</div>
4162
4163<div class="doc_text">
4164
4165<h5>Syntax:</h5>
4166<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4167 declare void @llvm.prefetch(i8 * &lt;address&gt;,
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4168 i32 &lt;rw&gt;, i32 &lt;locality&gt;)
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]4169</pre>
4170
4171<h5>Overview:</h5>
4172
4173
4174<p>
4175The '<tt>llvm.prefetch</tt>' intrinsic is a hint to the code generator to insert
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]4176a prefetch instruction if supported; otherwise, it is a noop. Prefetches have
4177no
4178effect on the behavior of the program but can change its performance
Chris Lattner2a615362005-02-28 19:47:14 +0000[diff] [blame]4179characteristics.
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]4180</p>
4181
4182<h5>Arguments:</h5>
4183
4184<p>
4185<tt>address</tt> is the address to be prefetched, <tt>rw</tt> is the specifier
4186determining if the fetch should be for a read (0) or write (1), and
4187<tt>locality</tt> is a temporal locality specifier ranging from (0) - no
Chris Lattneraeffb4a2005-03-07 20:31:38 +0000[diff] [blame]4188locality, to (3) - extremely local keep in cache. The <tt>rw</tt> and
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]4189<tt>locality</tt> arguments must be constant integers.
4190</p>
4191
4192<h5>Semantics:</h5>
4193
4194<p>
4195This intrinsic does not modify the behavior of the program. In particular,
4196prefetches cannot trap and do not produce a value. On targets that support this
4197intrinsic, the prefetch can provide hints to the processor cache for better
4198performance.
4199</p>
4200
4201</div>
4202
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]4203<!-- _______________________________________________________________________ -->
4204<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4205 <a name="int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a>
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]4206</div>
4207
4208<div class="doc_text">
4209
4210<h5>Syntax:</h5>
4211<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4212 declare void @llvm.pcmarker( i32 &lt;id&gt; )
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]4213</pre>
4214
4215<h5>Overview:</h5>
4216
4217
4218<p>
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]4219The '<tt>llvm.pcmarker</tt>' intrinsic is a method to export a Program Counter
4220(PC) in a region of
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]4221code to simulators and other tools. The method is target specific, but it is
4222expected that the marker will use exported symbols to transmit the PC of the marker.
Jeff Cohen25d4f7e2005-11-11 02:15:27 +0000[diff] [blame]4223The marker makes no guarantees that it will remain with any specific instruction
Chris Lattnerd07c3f42005-11-15 06:07:55 +0000[diff] [blame]4224after optimizations. It is possible that the presence of a marker will inhibit
Chris Lattnerb3e7afd2006-03-24 07:16:10 +0000[diff] [blame]4225optimizations. The intended use is to be inserted after optimizations to allow
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]4226correlations of simulation runs.
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]4227</p>
4228
4229<h5>Arguments:</h5>
4230
4231<p>
4232<tt>id</tt> is a numerical id identifying the marker.
4233</p>
4234
4235<h5>Semantics:</h5>
4236
4237<p>
4238This intrinsic does not modify the behavior of the program. Backends that do not
4239support this intrinisic may ignore it.
4240</p>
4241
4242</div>
4243
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]4244<!-- _______________________________________________________________________ -->
4245<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4246 <a name="int_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a>
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]4247</div>
4248
4249<div class="doc_text">
4250
4251<h5>Syntax:</h5>
4252<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4253 declare i64 @llvm.readcyclecounter( )
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]4254</pre>
4255
4256<h5>Overview:</h5>
4257
4258
4259<p>
4260The '<tt>llvm.readcyclecounter</tt>' intrinsic provides access to the cycle
4261counter register (or similar low latency, high accuracy clocks) on those targets
4262that support it. On X86, it should map to RDTSC. On Alpha, it should map to RPCC.
4263As the backing counters overflow quickly (on the order of 9 seconds on alpha), this
4264should only be used for small timings.
4265</p>
4266
4267<h5>Semantics:</h5>
4268
4269<p>
4270When directly supported, reading the cycle counter should not modify any memory.
4271Implementations are allowed to either return a application specific value or a
4272system wide value. On backends without support, this is lowered to a constant 0.
4273</p>
4274
4275</div>
4276
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4277<!-- ======================================================================= -->
4278<div class="doc_subsection">
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4279 <a name="int_libc">Standard C Library Intrinsics</a>
4280</div>
4281
4282<div class="doc_text">
4283<p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4284LLVM provides intrinsics for a few important standard C library functions.
4285These intrinsics allow source-language front-ends to pass information about the
4286alignment of the pointer arguments to the code generator, providing opportunity
4287for more efficient code generation.
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4288</p>
4289
4290</div>
4291
4292<!-- _______________________________________________________________________ -->
4293<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4294 <a name="int_memcpy">'<tt>llvm.memcpy</tt>' Intrinsic</a>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4295</div>
4296
4297<div class="doc_text">
4298
4299<h5>Syntax:</h5>
4300<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4301 declare void @llvm.memcpy.i32(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4302 i32 &lt;len&gt;, i32 &lt;align&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4303 declare void @llvm.memcpy.i64(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4304 i64 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4305</pre>
4306
4307<h5>Overview:</h5>
4308
4309<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4310The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4311location to the destination location.
4312</p>
4313
4314<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4315Note that, unlike the standard libc function, the <tt>llvm.memcpy.*</tt>
4316intrinsics do not return a value, and takes an extra alignment argument.
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4317</p>
4318
4319<h5>Arguments:</h5>
4320
4321<p>
4322The first argument is a pointer to the destination, the second is a pointer to
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4323the source. The third argument is an integer argument
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4324specifying the number of bytes to copy, and the fourth argument is the alignment
4325of the source and destination locations.
4326</p>
4327
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]4328<p>
4329If 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]4330the caller guarantees that both the source and destination pointers are aligned
4331to that boundary.
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]4332</p>
4333
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4334<h5>Semantics:</h5>
4335
4336<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4337The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4338location to the destination location, which are not allowed to overlap. It
4339copies "len" bytes of memory over. If the argument is known to be aligned to
4340some boundary, this can be specified as the fourth argument, otherwise it should
4341be set to 0 or 1.
4342</p>
4343</div>
4344
4345
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]4346<!-- _______________________________________________________________________ -->
4347<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4348 <a name="int_memmove">'<tt>llvm.memmove</tt>' Intrinsic</a>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]4349</div>
4350
4351<div class="doc_text">
4352
4353<h5>Syntax:</h5>
4354<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4355 declare void @llvm.memmove.i32(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4356 i32 &lt;len&gt;, i32 &lt;align&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4357 declare void @llvm.memmove.i64(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4358 i64 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]4359</pre>
4360
4361<h5>Overview:</h5>
4362
4363<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4364The '<tt>llvm.memmove.*</tt>' intrinsics move a block of memory from the source
4365location to the destination location. It is similar to the
4366'<tt>llvm.memcmp</tt>' intrinsic but allows the two memory locations to overlap.
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]4367</p>
4368
4369<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4370Note that, unlike the standard libc function, the <tt>llvm.memmove.*</tt>
4371intrinsics do not return a value, and takes an extra alignment argument.
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]4372</p>
4373
4374<h5>Arguments:</h5>
4375
4376<p>
4377The first argument is a pointer to the destination, the second is a pointer to
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4378the source. The third argument is an integer argument
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]4379specifying the number of bytes to copy, and the fourth argument is the alignment
4380of the source and destination locations.
4381</p>
4382
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]4383<p>
4384If 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]4385the caller guarantees that the source and destination pointers are aligned to
4386that boundary.
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]4387</p>
4388
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]4389<h5>Semantics:</h5>
4390
4391<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4392The '<tt>llvm.memmove.*</tt>' intrinsics copy a block of memory from the source
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]4393location to the destination location, which may overlap. It
4394copies "len" bytes of memory over. If the argument is known to be aligned to
4395some boundary, this can be specified as the fourth argument, otherwise it should
4396be set to 0 or 1.
4397</p>
4398</div>
4399
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]4400
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4401<!-- _______________________________________________________________________ -->
4402<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4403 <a name="int_memset">'<tt>llvm.memset.*</tt>' Intrinsics</a>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4404</div>
4405
4406<div class="doc_text">
4407
4408<h5>Syntax:</h5>
4409<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4410 declare void @llvm.memset.i32(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4411 i32 &lt;len&gt;, i32 &lt;align&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4412 declare void @llvm.memset.i64(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4413 i64 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4414</pre>
4415
4416<h5>Overview:</h5>
4417
4418<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4419The '<tt>llvm.memset.*</tt>' intrinsics fill a block of memory with a particular
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4420byte value.
4421</p>
4422
4423<p>
4424Note that, unlike the standard libc function, the <tt>llvm.memset</tt> intrinsic
4425does not return a value, and takes an extra alignment argument.
4426</p>
4427
4428<h5>Arguments:</h5>
4429
4430<p>
4431The first argument is a pointer to the destination to fill, the second is the
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4432byte value to fill it with, the third argument is an integer
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4433argument specifying the number of bytes to fill, and the fourth argument is the
4434known alignment of destination location.
4435</p>
4436
4437<p>
4438If 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]4439the caller guarantees that the destination pointer is aligned to that boundary.
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4440</p>
4441
4442<h5>Semantics:</h5>
4443
4444<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4445The '<tt>llvm.memset.*</tt>' intrinsics fill "len" bytes of memory starting at
4446the
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4447destination location. If the argument is known to be aligned to some boundary,
4448this can be specified as the fourth argument, otherwise it should be set to 0 or
44491.
4450</p>
4451</div>
4452
4453
Chris Lattner32006282004-06-11 02:28:03 +0000[diff] [blame]4454<!-- _______________________________________________________________________ -->
4455<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4456 <a name="int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a>
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]4457</div>
4458
4459<div class="doc_text">
4460
4461<h5>Syntax:</h5>
4462<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4463 declare float @llvm.sqrt.f32(float %Val)
4464 declare double @llvm.sqrt.f64(double %Val)
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]4465</pre>
4466
4467<h5>Overview:</h5>
4468
4469<p>
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]4470The '<tt>llvm.sqrt</tt>' intrinsics return the sqrt of the specified operand,
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]4471returning the same value as the libm '<tt>sqrt</tt>' function would. Unlike
4472<tt>sqrt</tt> in libm, however, <tt>llvm.sqrt</tt> has undefined behavior for
4473negative numbers (which allows for better optimization).
4474</p>
4475
4476<h5>Arguments:</h5>
4477
4478<p>
4479The argument and return value are floating point numbers of the same type.
4480</p>
4481
4482<h5>Semantics:</h5>
4483
4484<p>
4485This function returns the sqrt of the specified operand if it is a positive
4486floating point number.
4487</p>
4488</div>
4489
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]4490<!-- _______________________________________________________________________ -->
4491<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4492 <a name="int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]4493</div>
4494
4495<div class="doc_text">
4496
4497<h5>Syntax:</h5>
4498<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4499 declare float @llvm.powi.f32(float %Val, i32 %power)
4500 declare double @llvm.powi.f64(double %Val, i32 %power)
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]4501</pre>
4502
4503<h5>Overview:</h5>
4504
4505<p>
4506The '<tt>llvm.powi.*</tt>' intrinsics return the first operand raised to the
4507specified (positive or negative) power. The order of evaluation of
4508multiplications is not defined.
4509</p>
4510
4511<h5>Arguments:</h5>
4512
4513<p>
4514The second argument is an integer power, and the first is a value to raise to
4515that power.
4516</p>
4517
4518<h5>Semantics:</h5>
4519
4520<p>
4521This function returns the first value raised to the second power with an
4522unspecified sequence of rounding operations.</p>
4523</div>
4524
4525
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4526<!-- ======================================================================= -->
4527<div class="doc_subsection">
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]4528 <a name="int_manip">Bit Manipulation Intrinsics</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4529</div>
4530
4531<div class="doc_text">
4532<p>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]4533LLVM provides intrinsics for a few important bit manipulation operations.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4534These allow efficient code generation for some algorithms.
4535</p>
4536
4537</div>
4538
4539<!-- _______________________________________________________________________ -->
4540<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4541 <a name="int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]4542</div>
4543
4544<div class="doc_text">
4545
4546<h5>Syntax:</h5>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]4547<p>This is an overloaded intrinsic function. You can use bswap on any integer
4548type that is an even number of bytes (i.e. BitWidth % 16 == 0). Note the suffix
4549that includes the type for the result and the operand.
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]4550<pre>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]4551 declare i16 @llvm.bswap.i16.i16(i16 &lt;id&gt;)
4552 declare i32 @llvm.bswap.i32.i32(i32 &lt;id&gt;)
Reid Spencer543ab1d2007-04-02 00:19:52 +0000[diff] [blame]4553 declare i64 @llvm.bswap.i64.i64(i64 &lt;id&gt;)
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]4554</pre>
4555
4556<h5>Overview:</h5>
4557
4558<p>
Reid Spencer338ea092007-04-02 02:25:19 +0000[diff] [blame]4559The '<tt>llvm.bswap</tt>' family of intrinsics is used to byte swap integer
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]4560values with an even number of bytes (positive multiple of 16 bits). These are
4561useful for performing operations on data that is not in the target's native
4562byte order.
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]4563</p>
4564
4565<h5>Semantics:</h5>
4566
4567<p>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]4568The <tt>llvm.bswap.16.i16</tt> intrinsic returns an i16 value that has the high
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4569and low byte of the input i16 swapped. Similarly, the <tt>llvm.bswap.i32</tt>
4570intrinsic returns an i32 value that has the four bytes of the input i32
4571swapped, so that if the input bytes are numbered 0, 1, 2, 3 then the returned
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]4572i32 will have its bytes in 3, 2, 1, 0 order. The <tt>llvm.bswap.i48.i48</tt>,
4573<tt>llvm.bswap.i64.i64</tt> and other intrinsics extend this concept to
4574additional even-byte lengths (6 bytes, 8 bytes and more, respectively).
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]4575</p>
4576
4577</div>
4578
4579<!-- _______________________________________________________________________ -->
4580<div class="doc_subsubsection">
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]4581 <a name="int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4582</div>
4583
4584<div class="doc_text">
4585
4586<h5>Syntax:</h5>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]4587<p>This is an overloaded intrinsic. You can use llvm.ctpop on any integer bit
4588width. Not all targets support all bit widths however.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4589<pre>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]4590 declare i32 @llvm.ctpop.i8 (i8 &lt;src&gt;)
4591 declare i32 @llvm.ctpop.i16(i16 &lt;src&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4592 declare i32 @llvm.ctpop.i32(i32 &lt;src&gt;)
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]4593 declare i32 @llvm.ctpop.i64(i64 &lt;src&gt;)
4594 declare i32 @llvm.ctpop.i256(i256 &lt;src&gt;)
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4595</pre>
4596
4597<h5>Overview:</h5>
4598
4599<p>
Chris Lattnerec6cb612006-01-16 22:38:59 +0000[diff] [blame]4600The '<tt>llvm.ctpop</tt>' family of intrinsics counts the number of bits set in a
4601value.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4602</p>
4603
4604<h5>Arguments:</h5>
4605
4606<p>
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]4607The only argument is the value to be counted. The argument may be of any
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]4608integer type. The return type must match the argument type.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4609</p>
4610
4611<h5>Semantics:</h5>
4612
4613<p>
4614The '<tt>llvm.ctpop</tt>' intrinsic counts the 1's in a variable.
4615</p>
4616</div>
4617
4618<!-- _______________________________________________________________________ -->
4619<div class="doc_subsubsection">
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]4620 <a name="int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4621</div>
4622
4623<div class="doc_text">
4624
4625<h5>Syntax:</h5>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]4626<p>This is an overloaded intrinsic. You can use <tt>llvm.ctlz</tt> on any
4627integer bit width. Not all targets support all bit widths however.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4628<pre>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]4629 declare i32 @llvm.ctlz.i8 (i8 &lt;src&gt;)
4630 declare i32 @llvm.ctlz.i16(i16 &lt;src&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4631 declare i32 @llvm.ctlz.i32(i32 &lt;src&gt;)
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]4632 declare i32 @llvm.ctlz.i64(i64 &lt;src&gt;)
4633 declare i32 @llvm.ctlz.i256(i256 &lt;src&gt;)
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4634</pre>
4635
4636<h5>Overview:</h5>
4637
4638<p>
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]4639The '<tt>llvm.ctlz</tt>' family of intrinsic functions counts the number of
4640leading zeros in a variable.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4641</p>
4642
4643<h5>Arguments:</h5>
4644
4645<p>
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]4646The only argument is the value to be counted. The argument may be of any
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]4647integer type. The return type must match the argument type.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4648</p>
4649
4650<h5>Semantics:</h5>
4651
4652<p>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4653The '<tt>llvm.ctlz</tt>' intrinsic counts the leading (most significant) zeros
4654in a variable. If the src == 0 then the result is the size in bits of the type
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4655of src. For example, <tt>llvm.ctlz(i32 2) = 30</tt>.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4656</p>
4657</div>
Chris Lattner32006282004-06-11 02:28:03 +0000[diff] [blame]4658
4659
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4660
4661<!-- _______________________________________________________________________ -->
4662<div class="doc_subsubsection">
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]4663 <a name="int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic</a>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4664</div>
4665
4666<div class="doc_text">
4667
4668<h5>Syntax:</h5>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]4669<p>This is an overloaded intrinsic. You can use <tt>llvm.cttz</tt> on any
4670integer bit width. Not all targets support all bit widths however.
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4671<pre>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]4672 declare i32 @llvm.cttz.i8 (i8 &lt;src&gt;)
4673 declare i32 @llvm.cttz.i16(i16 &lt;src&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4674 declare i32 @llvm.cttz.i32(i32 &lt;src&gt;)
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]4675 declare i32 @llvm.cttz.i64(i64 &lt;src&gt;)
4676 declare i32 @llvm.cttz.i256(i256 &lt;src&gt;)
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4677</pre>
4678
4679<h5>Overview:</h5>
4680
4681<p>
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]4682The '<tt>llvm.cttz</tt>' family of intrinsic functions counts the number of
4683trailing zeros.
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4684</p>
4685
4686<h5>Arguments:</h5>
4687
4688<p>
4689The only argument is the value to be counted. The argument may be of any
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]4690integer type. The return type must match the argument type.
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4691</p>
4692
4693<h5>Semantics:</h5>
4694
4695<p>
4696The '<tt>llvm.cttz</tt>' intrinsic counts the trailing (least significant) zeros
4697in a variable. If the src == 0 then the result is the size in bits of the type
4698of src. For example, <tt>llvm.cttz(2) = 1</tt>.
4699</p>
4700</div>
4701
Reid Spencer497d93e2007-04-01 08:27:01 +0000[diff] [blame]4702<!-- _______________________________________________________________________ -->
4703<div class="doc_subsubsection">
Reid Spencerbeacf662007-04-10 02:51:31 +0000[diff] [blame]4704 <a name="int_part_select">'<tt>llvm.part.select.*</tt>' Intrinsic</a>
Reid Spencera13ba7d2007-04-01 19:00:37 +0000[diff] [blame]4705</div>
4706
4707<div class="doc_text">
4708
4709<h5>Syntax:</h5>
Reid Spencerbeacf662007-04-10 02:51:31 +0000[diff] [blame]4710<p>This is an overloaded intrinsic. You can use <tt>llvm.part.select</tt>
Reid Spencera13ba7d2007-04-01 19:00:37 +0000[diff] [blame]4711on any integer bit width.
4712<pre>
Reid Spencerbeacf662007-04-10 02:51:31 +0000[diff] [blame]4713 declare i17 @llvm.part.select.i17.i17 (i17 %val, i32 %loBit, i32 %hiBit)
4714 declare i29 @llvm.part.select.i29.i29 (i29 %val, i32 %loBit, i32 %hiBit)
Reid Spencera13ba7d2007-04-01 19:00:37 +0000[diff] [blame]4715</pre>
4716
4717<h5>Overview:</h5>
Reid Spencerbeacf662007-04-10 02:51:31 +0000[diff] [blame]4718<p>The '<tt>llvm.part.select</tt>' family of intrinsic functions selects a
Reid Spencera13ba7d2007-04-01 19:00:37 +0000[diff] [blame]4719range of bits from an integer value and returns them in the same bit width as
4720the original value.</p>
4721
4722<h5>Arguments:</h5>
4723<p>The first argument, <tt>%val</tt> and the result may be integer types of
4724any bit width but they must have the same bit width. The second and third
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4725arguments must be <tt>i32</tt> type since they specify only a bit index.</p>
Reid Spencera13ba7d2007-04-01 19:00:37 +0000[diff] [blame]4726
4727<h5>Semantics:</h5>
Reid Spencerbeacf662007-04-10 02:51:31 +0000[diff] [blame]4728<p>The operation of the '<tt>llvm.part.select</tt>' intrinsic has two modes
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4729of operation: forwards and reverse. If <tt>%loBit</tt> is greater than
4730<tt>%hiBits</tt> then the intrinsic operates in reverse mode. Otherwise it
4731operates in forward mode.</p>
4732<p>In forward mode, this intrinsic is the equivalent of shifting <tt>%val</tt>
4733right by <tt>%loBit</tt> bits and then ANDing it with a mask with
Reid Spencera13ba7d2007-04-01 19:00:37 +0000[diff] [blame]4734only the <tt>%hiBit - %loBit</tt> bits set, as follows:</p>
4735<ol>
4736 <li>The <tt>%val</tt> is shifted right (LSHR) by the number of bits specified
4737 by <tt>%loBits</tt>. This normalizes the value to the low order bits.</li>
4738 <li>The <tt>%loBits</tt> value is subtracted from the <tt>%hiBits</tt> value
4739 to determine the number of bits to retain.</li>
4740 <li>A mask of the retained bits is created by shifting a -1 value.</li>
4741 <li>The mask is ANDed with <tt>%val</tt> to produce the result.
4742</ol>
Reid Spencerd6a85b52007-05-14 16:14:57 +0000[diff] [blame]4743<p>In reverse mode, a similar computation is made except that the bits are
4744returned in the reverse order. So, for example, if <tt>X</tt> has the value
4745<tt>i16 0x0ACF (101011001111)</tt> and we apply
4746<tt>part.select(i16 X, 8, 3)</tt> to it, we get back the value
4747<tt>i16 0x0026 (000000100110)</tt>.</p>
Reid Spencera13ba7d2007-04-01 19:00:37 +0000[diff] [blame]4748</div>
4749
Reid Spencerf86037f2007-04-11 23:23:49 +0000[diff] [blame]4750<div class="doc_subsubsection">
4751 <a name="int_part_set">'<tt>llvm.part.set.*</tt>' Intrinsic</a>
4752</div>
4753
4754<div class="doc_text">
4755
4756<h5>Syntax:</h5>
4757<p>This is an overloaded intrinsic. You can use <tt>llvm.part.set</tt>
4758on any integer bit width.
4759<pre>
4760 declare i17 @llvm.part.set.i17.i17.i9 (i17 %val, i9 %repl, i32 %lo, i32 %hi)
4761 declare i29 @llvm.part.set.i29.i29.i9 (i29 %val, i9 %repl, i32 %lo, i32 %hi)
4762</pre>
4763
4764<h5>Overview:</h5>
4765<p>The '<tt>llvm.part.set</tt>' family of intrinsic functions replaces a range
4766of bits in an integer value with another integer value. It returns the integer
4767with the replaced bits.</p>
4768
4769<h5>Arguments:</h5>
4770<p>The first argument, <tt>%val</tt> and the result may be integer types of
4771any bit width but they must have the same bit width. <tt>%val</tt> is the value
4772whose bits will be replaced. The second argument, <tt>%repl</tt> may be an
4773integer of any bit width. The third and fourth arguments must be <tt>i32</tt>
4774type since they specify only a bit index.</p>
4775
4776<h5>Semantics:</h5>
4777<p>The operation of the '<tt>llvm.part.set</tt>' intrinsic has two modes
4778of operation: forwards and reverse. If <tt>%lo</tt> is greater than
4779<tt>%hi</tt> then the intrinsic operates in reverse mode. Otherwise it
4780operates in forward mode.</p>
4781<p>For both modes, the <tt>%repl</tt> value is prepared for use by either
4782truncating it down to the size of the replacement area or zero extending it
4783up to that size.</p>
4784<p>In forward mode, the bits between <tt>%lo</tt> and <tt>%hi</tt> (inclusive)
4785are replaced with corresponding bits from <tt>%repl</tt>. That is the 0th bit
4786in <tt>%repl</tt> replaces the <tt>%lo</tt>th bit in <tt>%val</tt> and etc. up
4787to the <tt>%hi</tt>th bit.
Reid Spencerc6749c42007-05-14 16:50:20 +0000[diff] [blame]4788<p>In reverse mode, a similar computation is made except that the bits are
4789reversed. That is, the <tt>0</tt>th bit in <tt>%repl</tt> replaces the
4790<tt>%hi</tt> bit in <tt>%val</tt> and etc. down to the <tt>%lo</tt>th bit.
Reid Spencerf86037f2007-04-11 23:23:49 +0000[diff] [blame]4791<h5>Examples:</h5>
4792<pre>
Reid Spencerf0dbf642007-04-12 01:03:03 +0000[diff] [blame]4793 llvm.part.set(0xFFFF, 0, 4, 7) -&gt; 0xFF0F
Reid Spencerc6749c42007-05-14 16:50:20 +0000[diff] [blame]4794 llvm.part.set(0xFFFF, 0, 7, 4) -&gt; 0xFF0F
4795 llvm.part.set(0xFFFF, 1, 7, 4) -&gt; 0xFF8F
4796 llvm.part.set(0xFFFF, F, 8, 3) -&gt; 0xFFE7
Reid Spencerf0dbf642007-04-12 01:03:03 +0000[diff] [blame]4797 llvm.part.set(0xFFFF, 0, 3, 8) -&gt; 0xFE07
Reid Spencerc8910842007-04-11 23:49:50 +0000[diff] [blame]4798</pre>
Reid Spencerf86037f2007-04-11 23:23:49 +0000[diff] [blame]4799</div>
4800
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]4801<!-- ======================================================================= -->
4802<div class="doc_subsection">
4803 <a name="int_debugger">Debugger Intrinsics</a>
4804</div>
4805
4806<div class="doc_text">
4807<p>
4808The LLVM debugger intrinsics (which all start with <tt>llvm.dbg.</tt> prefix),
4809are described in the <a
4810href="SourceLevelDebugging.html#format_common_intrinsics">LLVM Source Level
4811Debugging</a> document.
4812</p>
4813</div>
4814
4815
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]4816<!-- ======================================================================= -->
4817<div class="doc_subsection">
4818 <a name="int_eh">Exception Handling Intrinsics</a>
4819</div>
4820
4821<div class="doc_text">
4822<p> The LLVM exception handling intrinsics (which all start with
4823<tt>llvm.eh.</tt> prefix), are described in the <a
4824href="ExceptionHandling.html#format_common_intrinsics">LLVM Exception
4825Handling</a> document. </p>
4826</div>
4827
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame^]4828<!-- ======================================================================= -->
4829<div class="doc_subsection">
4830 <a name="int_general">General Intrinsics</a>
4831</div>
4832
4833<div class="doc_text">
4834<p> This class of intrinsics is designed to be generic and has
4835no specific purpose. </p>
4836</div>
4837
4838<!-- _______________________________________________________________________ -->
4839<div class="doc_subsubsection">
4840 <a name="int_var_annotation">'<tt>llvm.var.annotation</tt>' Intrinsic</a>
4841</div>
4842
4843<div class="doc_text">
4844
4845<h5>Syntax:</h5>
4846<pre>
4847 declare void @llvm.var.annotation(i8* &lt;val&gt;, i8* &lt;str&gt;)
4848</pre>
4849
4850<h5>Overview:</h5>
4851
4852<p>
4853The '<tt>llvm.var.annotation</tt>' intrinsic
4854</p>
4855
4856<h5>Arguments:</h5>
4857
4858<p>
4859The first argument is a pointer to a value, and the second is a pointer to a
4860global string.
4861</p>
4862
4863<h5>Semantics:</h5>
4864
4865<p>
4866This intrinsic allows annotation of local variables with arbitrary strings.
4867This can be useful for special purpose optimizations that want to look for these
4868 annotations. These have no other defined use, they are ignored by code
4869 generation and optimization.
4870</div>
4871
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]4872
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4873<!-- *********************************************************************** -->
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4874<hr>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]4875<address>
4876 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
4877 src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
4878 <a href="http://validator.w3.org/check/referer"><img
4879 src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!" /></a>
4880
4881 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
Reid Spencer05fe4b02006-03-14 05:39:39 +0000[diff] [blame]4882 <a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]4883 Last modified: $Date$
4884</address>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4885</body>
4886</html>