Gitiles
Code Review Sign In
gerrit-public.fairphone.software / fp2-dev / platform / external / llvm / 80a75bfae980df96f969f1c05b0c4a80ce975240 / . / docs / LangRef.html
blob: 80817e28dcc7eb5af8ee28d20ae6589c23e59d4f [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>
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame^]29 <li><a href="#gc">Garbage Collector Names</a></li>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]30 <li><a href="#moduleasm">Module-Level Inline Assembly</a></li>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]31 <li><a href="#datalayout">Data Layout</a></li>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]32 </ol>
33 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]34 <li><a href="#typesystem">Type System</a>
35 <ol>
Robert Bocchino7b81c752006-02-17 21:18:08 +0000[diff] [blame]36 <li><a href="#t_primitive">Primitive Types</a>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]37 <ol>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]38 <li><a href="#t_classifications">Type Classifications</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]39 </ol>
40 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]41 <li><a href="#t_derived">Derived Types</a>
42 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]43 <li><a href="#t_array">Array Type</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]44 <li><a href="#t_function">Function Type</a></li>
45 <li><a href="#t_pointer">Pointer Type</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]46 <li><a href="#t_struct">Structure Type</a></li>
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]47 <li><a href="#t_pstruct">Packed Structure Type</a></li>
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]48 <li><a href="#t_vector">Vector Type</a></li>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]49 <li><a href="#t_opaque">Opaque Type</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]50 </ol>
51 </li>
52 </ol>
53 </li>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]54 <li><a href="#constants">Constants</a>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]55 <ol>
56 <li><a href="#simpleconstants">Simple Constants</a>
57 <li><a href="#aggregateconstants">Aggregate Constants</a>
58 <li><a href="#globalconstants">Global Variable and Function Addresses</a>
59 <li><a href="#undefvalues">Undefined Values</a>
60 <li><a href="#constantexprs">Constant Expressions</a>
61 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]62 </li>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]63 <li><a href="#othervalues">Other Values</a>
64 <ol>
65 <li><a href="#inlineasm">Inline Assembler Expressions</a>
66 </ol>
67 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]68 <li><a href="#instref">Instruction Reference</a>
69 <ol>
70 <li><a href="#terminators">Terminator Instructions</a>
71 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]72 <li><a href="#i_ret">'<tt>ret</tt>' Instruction</a></li>
73 <li><a href="#i_br">'<tt>br</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]74 <li><a href="#i_switch">'<tt>switch</tt>' Instruction</a></li>
75 <li><a href="#i_invoke">'<tt>invoke</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]76 <li><a href="#i_unwind">'<tt>unwind</tt>' Instruction</a></li>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]77 <li><a href="#i_unreachable">'<tt>unreachable</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]78 </ol>
79 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]80 <li><a href="#binaryops">Binary Operations</a>
81 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]82 <li><a href="#i_add">'<tt>add</tt>' Instruction</a></li>
83 <li><a href="#i_sub">'<tt>sub</tt>' Instruction</a></li>
84 <li><a href="#i_mul">'<tt>mul</tt>' Instruction</a></li>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]85 <li><a href="#i_udiv">'<tt>udiv</tt>' Instruction</a></li>
86 <li><a href="#i_sdiv">'<tt>sdiv</tt>' Instruction</a></li>
87 <li><a href="#i_fdiv">'<tt>fdiv</tt>' Instruction</a></li>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]88 <li><a href="#i_urem">'<tt>urem</tt>' Instruction</a></li>
89 <li><a href="#i_srem">'<tt>srem</tt>' Instruction</a></li>
90 <li><a href="#i_frem">'<tt>frem</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]91 </ol>
92 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]93 <li><a href="#bitwiseops">Bitwise Binary Operations</a>
94 <ol>
Reid Spencer8e11bf82007-02-02 13:57:07 +0000[diff] [blame]95 <li><a href="#i_shl">'<tt>shl</tt>' Instruction</a></li>
96 <li><a href="#i_lshr">'<tt>lshr</tt>' Instruction</a></li>
97 <li><a href="#i_ashr">'<tt>ashr</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]98 <li><a href="#i_and">'<tt>and</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]99 <li><a href="#i_or">'<tt>or</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]100 <li><a href="#i_xor">'<tt>xor</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]101 </ol>
102 </li>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]103 <li><a href="#vectorops">Vector Operations</a>
104 <ol>
105 <li><a href="#i_extractelement">'<tt>extractelement</tt>' Instruction</a></li>
106 <li><a href="#i_insertelement">'<tt>insertelement</tt>' Instruction</a></li>
107 <li><a href="#i_shufflevector">'<tt>shufflevector</tt>' Instruction</a></li>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]108 </ol>
109 </li>
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]110 <li><a href="#memoryops">Memory Access and Addressing Operations</a>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]111 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]112 <li><a href="#i_malloc">'<tt>malloc</tt>' Instruction</a></li>
113 <li><a href="#i_free">'<tt>free</tt>' Instruction</a></li>
114 <li><a href="#i_alloca">'<tt>alloca</tt>' Instruction</a></li>
Robert Bocchino7b81c752006-02-17 21:18:08 +0000[diff] [blame]115 <li><a href="#i_load">'<tt>load</tt>' Instruction</a></li>
116 <li><a href="#i_store">'<tt>store</tt>' Instruction</a></li>
117 <li><a href="#i_getelementptr">'<tt>getelementptr</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]118 </ol>
119 </li>
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]120 <li><a href="#convertops">Conversion Operations</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]121 <ol>
122 <li><a href="#i_trunc">'<tt>trunc .. to</tt>' Instruction</a></li>
123 <li><a href="#i_zext">'<tt>zext .. to</tt>' Instruction</a></li>
124 <li><a href="#i_sext">'<tt>sext .. to</tt>' Instruction</a></li>
125 <li><a href="#i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a></li>
126 <li><a href="#i_fpext">'<tt>fpext .. to</tt>' Instruction</a></li>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]127 <li><a href="#i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a></li>
128 <li><a href="#i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a></li>
129 <li><a href="#i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a></li>
130 <li><a href="#i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a></li>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]131 <li><a href="#i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a></li>
132 <li><a href="#i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a></li>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]133 <li><a href="#i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a></li>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]134 </ol>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]135 <li><a href="#otherops">Other Operations</a>
136 <ol>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]137 <li><a href="#i_icmp">'<tt>icmp</tt>' Instruction</a></li>
138 <li><a href="#i_fcmp">'<tt>fcmp</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]139 <li><a href="#i_phi">'<tt>phi</tt>' Instruction</a></li>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]140 <li><a href="#i_select">'<tt>select</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]141 <li><a href="#i_call">'<tt>call</tt>' Instruction</a></li>
Chris Lattnerfb6977d2006-01-13 23:26:01 +0000[diff] [blame]142 <li><a href="#i_va_arg">'<tt>va_arg</tt>' Instruction</a></li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]143 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]144 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]145 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]146 </li>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]147 <li><a href="#intrinsics">Intrinsic Functions</a>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]148 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]149 <li><a href="#int_varargs">Variable Argument Handling Intrinsics</a>
150 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]151 <li><a href="#int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a></li>
152 <li><a href="#int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a></li>
153 <li><a href="#int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]154 </ol>
155 </li>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]156 <li><a href="#int_gc">Accurate Garbage Collection Intrinsics</a>
157 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]158 <li><a href="#int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a></li>
159 <li><a href="#int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a></li>
160 <li><a href="#int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a></li>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]161 </ol>
162 </li>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]163 <li><a href="#int_codegen">Code Generator Intrinsics</a>
164 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]165 <li><a href="#int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a></li>
166 <li><a href="#int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a></li>
167 <li><a href="#int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a></li>
168 <li><a href="#int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a></li>
169 <li><a href="#int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a></li>
170 <li><a href="#int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a></li>
171 <li><a href="#int_readcyclecounter"><tt>llvm.readcyclecounter</tt>' Intrinsic</a></li>
John Criswell7123e272004-04-09 16:43:20 +0000[diff] [blame]172 </ol>
173 </li>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]174 <li><a href="#int_libc">Standard C Library Intrinsics</a>
175 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]176 <li><a href="#int_memcpy">'<tt>llvm.memcpy.*</tt>' Intrinsic</a></li>
177 <li><a href="#int_memmove">'<tt>llvm.memmove.*</tt>' Intrinsic</a></li>
178 <li><a href="#int_memset">'<tt>llvm.memset.*</tt>' Intrinsic</a></li>
179 <li><a href="#int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a></li>
180 <li><a href="#int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a></li>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]181 <li><a href="#int_sin">'<tt>llvm.sin.*</tt>' Intrinsic</a></li>
182 <li><a href="#int_cos">'<tt>llvm.cos.*</tt>' Intrinsic</a></li>
183 <li><a href="#int_pow">'<tt>llvm.pow.*</tt>' Intrinsic</a></li>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]184 </ol>
185 </li>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]186 <li><a href="#int_manip">Bit Manipulation Intrinsics</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]187 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]188 <li><a href="#int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a></li>
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]189 <li><a href="#int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic </a></li>
190 <li><a href="#int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic </a></li>
191 <li><a href="#int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic </a></li>
Reid Spencerf86037f2007-04-11 23:23:49 +0000[diff] [blame]192 <li><a href="#int_part_select">'<tt>llvm.part.select.*</tt>' Intrinsic </a></li>
193 <li><a href="#int_part_set">'<tt>llvm.part.set.*</tt>' Intrinsic </a></li>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]194 </ol>
195 </li>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]196 <li><a href="#int_debugger">Debugger intrinsics</a></li>
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]197 <li><a href="#int_eh">Exception Handling intrinsics</a></li>
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]198 <li><a href="#int_trampoline">Trampoline Intrinsic</a>
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]199 <ol>
200 <li><a href="#int_it">'<tt>llvm.init.trampoline</tt>' Intrinsic</a></li>
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]201 </ol>
202 </li>
Reid Spencer20677642007-07-20 19:59:11 +0000[diff] [blame]203 <li><a href="#int_general">General intrinsics</a>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]204 <ol>
Reid Spencer20677642007-07-20 19:59:11 +0000[diff] [blame]205 <li><a href="#int_var_annotation">
Tanya Lattner91d0b882007-09-22 00:01:26 +0000[diff] [blame]206 <tt>llvm.var.annotation</tt>' Intrinsic</a></li>
Reid Spencer20677642007-07-20 19:59:11 +0000[diff] [blame]207 </ol>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]208 <ol>
209 <li><a href="#int_annotation">
Tanya Lattner91d0b882007-09-22 00:01:26 +0000[diff] [blame]210 <tt>llvm.annotation.*</tt>' Intrinsic</a></li>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]211 </ol>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]212 </li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]213 </ol>
214 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]215</ol>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]216
217<div class="doc_author">
218 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>
219 and <a href="mailto:vadve@cs.uiuc.edu">Vikram Adve</a></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="abstract">Abstract </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 Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]227<p>This document is a reference manual for the LLVM assembly language.
228LLVM is an SSA based representation that provides type safety,
229low-level operations, flexibility, and the capability of representing
230'all' high-level languages cleanly. It is the common code
231representation used throughout all phases of the LLVM compilation
232strategy.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]233</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]234
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]235<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]236<div class="doc_section"> <a name="introduction">Introduction</a> </div>
237<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]238
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]239<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]240
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]241<p>The LLVM code representation is designed to be used in three
Gabor Greif04367bf2007-07-06 22:07:22 +0000[diff] [blame]242different forms: as an in-memory compiler IR, as an on-disk bitcode
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]243representation (suitable for fast loading by a Just-In-Time compiler),
244and as a human readable assembly language representation. This allows
245LLVM to provide a powerful intermediate representation for efficient
246compiler transformations and analysis, while providing a natural means
247to debug and visualize the transformations. The three different forms
248of LLVM are all equivalent. This document describes the human readable
249representation and notation.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]250
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]251<p>The LLVM representation aims to be light-weight and low-level
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]252while being expressive, typed, and extensible at the same time. It
253aims to be a "universal IR" of sorts, by being at a low enough level
254that high-level ideas may be cleanly mapped to it (similar to how
255microprocessors are "universal IR's", allowing many source languages to
256be mapped to them). By providing type information, LLVM can be used as
257the target of optimizations: for example, through pointer analysis, it
258can be proven that a C automatic variable is never accessed outside of
259the current function... allowing it to be promoted to a simple SSA
260value instead of a memory location.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]261
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]262</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]263
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]264<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]265<div class="doc_subsubsection"> <a name="wellformed">Well-Formedness</a> </div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]266
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]267<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]268
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]269<p>It is important to note that this document describes 'well formed'
270LLVM assembly language. There is a difference between what the parser
271accepts and what is considered 'well formed'. For example, the
272following instruction is syntactically okay, but not well formed:</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]273
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]274<div class="doc_code">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]275<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]276%x = <a href="#i_add">add</a> i32 1, %x
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]277</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]278</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]279
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]280<p>...because the definition of <tt>%x</tt> does not dominate all of
281its uses. The LLVM infrastructure provides a verification pass that may
282be used to verify that an LLVM module is well formed. This pass is
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]283automatically run by the parser after parsing input assembly and by
Gabor Greif04367bf2007-07-06 22:07:22 +0000[diff] [blame]284the optimizer before it outputs bitcode. The violations pointed out
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]285by the verifier pass indicate bugs in transformation passes or input to
286the parser.</p>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]287</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]288
Chris Lattnercc689392007-10-03 17:34:29 +0000[diff] [blame]289<!-- Describe the typesetting conventions here. -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]290
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]291<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]292<div class="doc_section"> <a name="identifiers">Identifiers</a> </div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]293<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]294
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]295<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]296
Reid Spencer2c452282007-08-07 14:34:28 +0000[diff] [blame]297 <p>LLVM identifiers come in two basic types: global and local. Global
298 identifiers (functions, global variables) begin with the @ character. Local
299 identifiers (register names, types) begin with the % character. Additionally,
300 there are three different formats for identifiers, for different purposes:
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]301
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]302<ol>
Reid Spencer2c452282007-08-07 14:34:28 +0000[diff] [blame]303 <li>Named values are represented as a string of characters with their prefix.
304 For example, %foo, @DivisionByZero, %a.really.long.identifier. The actual
305 regular expression used is '<tt>[%@][a-zA-Z$._][a-zA-Z$._0-9]*</tt>'.
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]306 Identifiers which require other characters in their names can be surrounded
Reid Spencer2c452282007-08-07 14:34:28 +0000[diff] [blame]307 with quotes. In this way, anything except a <tt>&quot;</tt> character can
308 be used in a named value.</li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]309
Reid Spencer2c452282007-08-07 14:34:28 +0000[diff] [blame]310 <li>Unnamed values are represented as an unsigned numeric value with their
311 prefix. For example, %12, @2, %44.</li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]312
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]313 <li>Constants, which are described in a <a href="#constants">section about
314 constants</a>, below.</li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]315</ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]316
Reid Spencer2c452282007-08-07 14:34:28 +0000[diff] [blame]317<p>LLVM requires that values start with a prefix for two reasons: Compilers
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]318don't need to worry about name clashes with reserved words, and the set of
319reserved words may be expanded in the future without penalty. Additionally,
320unnamed identifiers allow a compiler to quickly come up with a temporary
321variable without having to avoid symbol table conflicts.</p>
322
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]323<p>Reserved words in LLVM are very similar to reserved words in other
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]324languages. There are keywords for different opcodes
325('<tt><a href="#i_add">add</a></tt>',
326 '<tt><a href="#i_bitcast">bitcast</a></tt>',
327 '<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]328href="#t_void">void</a></tt>', '<tt><a href="#t_primitive">i32</a></tt>', etc...),
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]329and others. These reserved words cannot conflict with variable names, because
Reid Spencer2c452282007-08-07 14:34:28 +0000[diff] [blame]330none of them start with a prefix character ('%' or '@').</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]331
332<p>Here is an example of LLVM code to multiply the integer variable
333'<tt>%X</tt>' by 8:</p>
334
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]335<p>The easy way:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]336
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]337<div class="doc_code">
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]338<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]339%result = <a href="#i_mul">mul</a> i32 %X, 8
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]340</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]341</div>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]342
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]343<p>After strength reduction:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]344
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]345<div class="doc_code">
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]346<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]347%result = <a href="#i_shl">shl</a> i32 %X, i8 3
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]348</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]349</div>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]350
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]351<p>And the hard way:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]352
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]353<div class="doc_code">
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]354<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]355<a href="#i_add">add</a> i32 %X, %X <i>; yields {i32}:%0</i>
356<a href="#i_add">add</a> i32 %0, %0 <i>; yields {i32}:%1</i>
357%result = <a href="#i_add">add</a> i32 %1, %1
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]358</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]359</div>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]360
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]361<p>This last way of multiplying <tt>%X</tt> by 8 illustrates several
362important lexical features of LLVM:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]363
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]364<ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]365
366 <li>Comments are delimited with a '<tt>;</tt>' and go until the end of
367 line.</li>
368
369 <li>Unnamed temporaries are created when the result of a computation is not
370 assigned to a named value.</li>
371
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]372 <li>Unnamed temporaries are numbered sequentially</li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]373
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]374</ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]375
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]376<p>...and it also shows a convention that we follow in this document. When
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]377demonstrating instructions, we will follow an instruction with a comment that
378defines the type and name of value produced. Comments are shown in italic
379text.</p>
380
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]381</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]382
383<!-- *********************************************************************** -->
384<div class="doc_section"> <a name="highlevel">High Level Structure</a> </div>
385<!-- *********************************************************************** -->
386
387<!-- ======================================================================= -->
388<div class="doc_subsection"> <a name="modulestructure">Module Structure</a>
389</div>
390
391<div class="doc_text">
392
393<p>LLVM programs are composed of "Module"s, each of which is a
394translation unit of the input programs. Each module consists of
395functions, global variables, and symbol table entries. Modules may be
396combined together with the LLVM linker, which merges function (and
397global variable) definitions, resolves forward declarations, and merges
398symbol table entries. Here is an example of the "hello world" module:</p>
399
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]400<div class="doc_code">
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]401<pre><i>; Declare the string constant as a global constant...</i>
Chris Lattnera89e5f12007-06-12 17:00:26 +0000[diff] [blame]402<a href="#identifiers">@.LC0</a> = <a href="#linkage_internal">internal</a> <a
403 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]404
405<i>; External declaration of the puts function</i>
Chris Lattnera89e5f12007-06-12 17:00:26 +0000[diff] [blame]406<a href="#functionstructure">declare</a> i32 @puts(i8 *) <i>; i32(i8 *)* </i>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]407
408<i>; Definition of main function</i>
Chris Lattnera89e5f12007-06-12 17:00:26 +0000[diff] [blame]409define i32 @main() { <i>; i32()* </i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]410 <i>; Convert [13x i8 ]* to i8 *...</i>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]411 %cast210 = <a
Chris Lattner6c0955b2007-06-12 17:01:15 +0000[diff] [blame]412 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]413
414 <i>; Call puts function to write out the string to stdout...</i>
415 <a
Chris Lattnera89e5f12007-06-12 17:00:26 +0000[diff] [blame]416 href="#i_call">call</a> i32 @puts(i8 * %cast210) <i>; i32</i>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]417 <a
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]418 href="#i_ret">ret</a> i32 0<br>}<br>
419</pre>
420</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]421
422<p>This example is made up of a <a href="#globalvars">global variable</a>
423named "<tt>.LC0</tt>", an external declaration of the "<tt>puts</tt>"
424function, and a <a href="#functionstructure">function definition</a>
425for "<tt>main</tt>".</p>
426
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]427<p>In general, a module is made up of a list of global values,
428where both functions and global variables are global values. Global values are
429represented by a pointer to a memory location (in this case, a pointer to an
430array of char, and a pointer to a function), and have one of the following <a
431href="#linkage">linkage types</a>.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]432
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]433</div>
434
435<!-- ======================================================================= -->
436<div class="doc_subsection">
437 <a name="linkage">Linkage Types</a>
438</div>
439
440<div class="doc_text">
441
442<p>
443All Global Variables and Functions have one of the following types of linkage:
444</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]445
446<dl>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]447
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]448 <dt><tt><b><a name="linkage_internal">internal</a></b></tt> </dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]449
450 <dd>Global values with internal linkage are only directly accessible by
451 objects in the current module. In particular, linking code into a module with
452 an internal global value may cause the internal to be renamed as necessary to
453 avoid collisions. Because the symbol is internal to the module, all
454 references can be updated. This corresponds to the notion of the
Chris Lattner4887bd82007-01-14 06:51:48 +0000[diff] [blame]455 '<tt>static</tt>' keyword in C.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]456 </dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]457
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]458 <dt><tt><b><a name="linkage_linkonce">linkonce</a></b></tt>: </dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]459
Chris Lattner4887bd82007-01-14 06:51:48 +0000[diff] [blame]460 <dd>Globals with "<tt>linkonce</tt>" linkage are merged with other globals of
461 the same name when linkage occurs. This is typically used to implement
462 inline functions, templates, or other code which must be generated in each
463 translation unit that uses it. Unreferenced <tt>linkonce</tt> globals are
464 allowed to be discarded.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]465 </dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]466
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]467 <dt><tt><b><a name="linkage_weak">weak</a></b></tt>: </dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]468
469 <dd>"<tt>weak</tt>" linkage is exactly the same as <tt>linkonce</tt> linkage,
470 except that unreferenced <tt>weak</tt> globals may not be discarded. This is
Chris Lattner4887bd82007-01-14 06:51:48 +0000[diff] [blame]471 used for globals that may be emitted in multiple translation units, but that
472 are not guaranteed to be emitted into every translation unit that uses them.
473 One example of this are common globals in C, such as "<tt>int X;</tt>" at
474 global scope.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]475 </dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]476
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]477 <dt><tt><b><a name="linkage_appending">appending</a></b></tt>: </dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]478
479 <dd>"<tt>appending</tt>" linkage may only be applied to global variables of
480 pointer to array type. When two global variables with appending linkage are
481 linked together, the two global arrays are appended together. This is the
482 LLVM, typesafe, equivalent of having the system linker append together
483 "sections" with identical names when .o files are linked.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]484 </dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]485
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame]486 <dt><tt><b><a name="linkage_externweak">extern_weak</a></b></tt>: </dt>
487 <dd>The semantics of this linkage follow the ELF model: the symbol is weak
488 until linked, if not linked, the symbol becomes null instead of being an
489 undefined reference.
490 </dd>
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame]491
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]492 <dt><tt><b><a name="linkage_external">externally visible</a></b></tt>:</dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]493
494 <dd>If none of the above identifiers are used, the global is externally
495 visible, meaning that it participates in linkage and can be used to resolve
496 external symbol references.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]497 </dd>
Reid Spencerc8910842007-04-11 23:49:50 +0000[diff] [blame]498</dl>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]499
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]500 <p>
501 The next two types of linkage are targeted for Microsoft Windows platform
502 only. They are designed to support importing (exporting) symbols from (to)
503 DLLs.
504 </p>
505
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame]506 <dl>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]507 <dt><tt><b><a name="linkage_dllimport">dllimport</a></b></tt>: </dt>
508
509 <dd>"<tt>dllimport</tt>" linkage causes the compiler to reference a function
510 or variable via a global pointer to a pointer that is set up by the DLL
511 exporting the symbol. On Microsoft Windows targets, the pointer name is
512 formed by combining <code>_imp__</code> and the function or variable name.
513 </dd>
514
515 <dt><tt><b><a name="linkage_dllexport">dllexport</a></b></tt>: </dt>
516
517 <dd>"<tt>dllexport</tt>" linkage causes the compiler to provide a global
518 pointer to a pointer in a DLL, so that it can be referenced with the
519 <tt>dllimport</tt> attribute. On Microsoft Windows targets, the pointer
520 name is formed by combining <code>_imp__</code> and the function or variable
521 name.
522 </dd>
523
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]524</dl>
525
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame]526<p><a name="linkage_external"></a>For example, since the "<tt>.LC0</tt>"
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]527variable is defined to be internal, if another module defined a "<tt>.LC0</tt>"
528variable and was linked with this one, one of the two would be renamed,
529preventing a collision. Since "<tt>main</tt>" and "<tt>puts</tt>" are
530external (i.e., lacking any linkage declarations), they are accessible
Reid Spencerac8d2762007-01-05 00:59:10 +0000[diff] [blame]531outside of the current module.</p>
532<p>It is illegal for a function <i>declaration</i>
533to have any linkage type other than "externally visible", <tt>dllimport</tt>,
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame]534or <tt>extern_weak</tt>.</p>
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]535<p>Aliases can have only <tt>external</tt>, <tt>internal</tt> and <tt>weak</tt>
536linkages.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]537</div>
538
539<!-- ======================================================================= -->
540<div class="doc_subsection">
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]541 <a name="callingconv">Calling Conventions</a>
542</div>
543
544<div class="doc_text">
545
546<p>LLVM <a href="#functionstructure">functions</a>, <a href="#i_call">calls</a>
547and <a href="#i_invoke">invokes</a> can all have an optional calling convention
548specified for the call. The calling convention of any pair of dynamic
549caller/callee must match, or the behavior of the program is undefined. The
550following calling conventions are supported by LLVM, and more may be added in
551the future:</p>
552
553<dl>
554 <dt><b>"<tt>ccc</tt>" - The C calling convention</b>:</dt>
555
556 <dd>This calling convention (the default if no other calling convention is
557 specified) matches the target C calling conventions. This calling convention
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]558 supports varargs function calls and tolerates some mismatch in the declared
Reid Spencerc28d2bc2006-12-31 21:30:18 +0000[diff] [blame]559 prototype and implemented declaration of the function (as does normal C).
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]560 </dd>
561
562 <dt><b>"<tt>fastcc</tt>" - The fast calling convention</b>:</dt>
563
564 <dd>This calling convention attempts to make calls as fast as possible
565 (e.g. by passing things in registers). This calling convention allows the
566 target to use whatever tricks it wants to produce fast code for the target,
Chris Lattner8cdc5bc2005-05-06 23:08:23 +0000[diff] [blame]567 without having to conform to an externally specified ABI. Implementations of
568 this convention should allow arbitrary tail call optimization to be supported.
569 This calling convention does not support varargs and requires the prototype of
570 all callees to exactly match the prototype of the function definition.
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]571 </dd>
572
573 <dt><b>"<tt>coldcc</tt>" - The cold calling convention</b>:</dt>
574
575 <dd>This calling convention attempts to make code in the caller as efficient
576 as possible under the assumption that the call is not commonly executed. As
577 such, these calls often preserve all registers so that the call does not break
578 any live ranges in the caller side. This calling convention does not support
579 varargs and requires the prototype of all callees to exactly match the
580 prototype of the function definition.
581 </dd>
582
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]583 <dt><b>"<tt>cc &lt;<em>n</em>&gt;</tt>" - Numbered convention</b>:</dt>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]584
585 <dd>Any calling convention may be specified by number, allowing
586 target-specific calling conventions to be used. Target specific calling
587 conventions start at 64.
588 </dd>
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]589</dl>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]590
591<p>More calling conventions can be added/defined on an as-needed basis, to
592support pascal conventions or any other well-known target-independent
593convention.</p>
594
595</div>
596
597<!-- ======================================================================= -->
598<div class="doc_subsection">
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]599 <a name="visibility">Visibility Styles</a>
600</div>
601
602<div class="doc_text">
603
604<p>
605All Global Variables and Functions have one of the following visibility styles:
606</p>
607
608<dl>
609 <dt><b>"<tt>default</tt>" - Default style</b>:</dt>
610
611 <dd>On ELF, default visibility means that the declaration is visible to other
612 modules and, in shared libraries, means that the declared entity may be
613 overridden. On Darwin, default visibility means that the declaration is
614 visible to other modules. Default visibility corresponds to "external
615 linkage" in the language.
616 </dd>
617
618 <dt><b>"<tt>hidden</tt>" - Hidden style</b>:</dt>
619
620 <dd>Two declarations of an object with hidden visibility refer to the same
621 object if they are in the same shared object. Usually, hidden visibility
622 indicates that the symbol will not be placed into the dynamic symbol table,
623 so no other module (executable or shared library) can reference it
624 directly.
625 </dd>
626
Anton Korobeynikov6f9896f2007-04-29 18:35:00 +0000[diff] [blame]627 <dt><b>"<tt>protected</tt>" - Protected style</b>:</dt>
628
629 <dd>On ELF, protected visibility indicates that the symbol will be placed in
630 the dynamic symbol table, but that references within the defining module will
631 bind to the local symbol. That is, the symbol cannot be overridden by another
632 module.
633 </dd>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]634</dl>
635
636</div>
637
638<!-- ======================================================================= -->
639<div class="doc_subsection">
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]640 <a name="globalvars">Global Variables</a>
641</div>
642
643<div class="doc_text">
644
Chris Lattner3689a342005-02-12 19:30:21 +0000[diff] [blame]645<p>Global variables define regions of memory allocated at compilation time
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]646instead of run-time. Global variables may optionally be initialized, may have
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]647an explicit section to be placed in, and may have an optional explicit alignment
648specified. A variable may be defined as "thread_local", which means that it
649will not be shared by threads (each thread will have a separated copy of the
650variable). A variable may be defined as a global "constant," which indicates
651that the contents of the variable will <b>never</b> be modified (enabling better
Chris Lattner3689a342005-02-12 19:30:21 +0000[diff] [blame]652optimization, allowing the global data to be placed in the read-only section of
653an executable, etc). Note that variables that need runtime initialization
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]654cannot be marked "constant" as there is a store to the variable.</p>
Chris Lattner3689a342005-02-12 19:30:21 +0000[diff] [blame]655
656<p>
657LLVM explicitly allows <em>declarations</em> of global variables to be marked
658constant, even if the final definition of the global is not. This capability
659can be used to enable slightly better optimization of the program, but requires
660the language definition to guarantee that optimizations based on the
661'constantness' are valid for the translation units that do not include the
662definition.
663</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]664
665<p>As SSA values, global variables define pointer values that are in
666scope (i.e. they dominate) all basic blocks in the program. Global
667variables always define a pointer to their "content" type because they
668describe a region of memory, and all memory objects in LLVM are
669accessed through pointers.</p>
670
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]671<p>LLVM allows an explicit section to be specified for globals. If the target
672supports it, it will emit globals to the section specified.</p>
673
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]674<p>An explicit alignment may be specified for a global. If not present, or if
675the alignment is set to zero, the alignment of the global is set by the target
676to whatever it feels convenient. If an explicit alignment is specified, the
677global is forced to have at least that much alignment. All alignments must be
678a power of 2.</p>
679
Chris Lattner68027ea2007-01-14 00:27:09 +0000[diff] [blame]680<p>For example, the following defines a global with an initializer, section,
681 and alignment:</p>
682
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]683<div class="doc_code">
Chris Lattner68027ea2007-01-14 00:27:09 +0000[diff] [blame]684<pre>
Chris Lattner3e63a9d2007-07-13 20:01:46 +0000[diff] [blame]685@G = constant float 1.0, section "foo", align 4
Chris Lattner68027ea2007-01-14 00:27:09 +0000[diff] [blame]686</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]687</div>
Chris Lattner68027ea2007-01-14 00:27:09 +0000[diff] [blame]688
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]689</div>
690
691
692<!-- ======================================================================= -->
693<div class="doc_subsection">
694 <a name="functionstructure">Functions</a>
695</div>
696
697<div class="doc_text">
698
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]699<p>LLVM function definitions consist of the "<tt>define</tt>" keyord,
700an optional <a href="#linkage">linkage type</a>, an optional
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]701<a href="#visibility">visibility style</a>, an optional
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]702<a href="#callingconv">calling convention</a>, a return type, an optional
703<a href="#paramattrs">parameter attribute</a> for the return type, a function
704name, a (possibly empty) argument list (each with optional
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]705<a href="#paramattrs">parameter attributes</a>), an optional section, an
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame^]706optional alignment, an optional <a href="gc">garbage collector name</a>, an
707opening curly brace, a list of basic blocks, and a closing curly brace.
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]708
709LLVM function declarations consist of the "<tt>declare</tt>" keyword, an
710optional <a href="#linkage">linkage type</a>, an optional
711<a href="#visibility">visibility style</a>, an optional
712<a href="#callingconv">calling convention</a>, a return type, an optional
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]713<a href="#paramattrs">parameter attribute</a> for the return type, a function
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame^]714name, a possibly empty list of arguments, an optional alignment, and an optional
715<a href="gc">garbage collector name</a>.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]716
717<p>A function definition contains a list of basic blocks, forming the CFG for
718the function. Each basic block may optionally start with a label (giving the
719basic block a symbol table entry), contains a list of instructions, and ends
720with a <a href="#terminators">terminator</a> instruction (such as a branch or
721function return).</p>
722
Chris Lattner4a3c9012007-06-08 16:52:14 +0000[diff] [blame]723<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]724executed on entrance to the function, and it is not allowed to have predecessor
725basic blocks (i.e. there can not be any branches to the entry block of a
726function). Because the block can have no predecessors, it also cannot have any
727<a href="#i_phi">PHI nodes</a>.</p>
728
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]729<p>LLVM allows an explicit section to be specified for functions. If the target
730supports it, it will emit functions to the section specified.</p>
731
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]732<p>An explicit alignment may be specified for a function. If not present, or if
733the alignment is set to zero, the alignment of the function is set by the target
734to whatever it feels convenient. If an explicit alignment is specified, the
735function is forced to have at least that much alignment. All alignments must be
736a power of 2.</p>
737
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]738</div>
739
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]740
741<!-- ======================================================================= -->
742<div class="doc_subsection">
743 <a name="aliasstructure">Aliases</a>
744</div>
745<div class="doc_text">
746 <p>Aliases act as "second name" for the aliasee value (which can be either
Anton Korobeynikova80e1182007-04-28 13:45:00 +0000[diff] [blame]747 function or global variable or bitcast of global value). Aliases may have an
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]748 optional <a href="#linkage">linkage type</a>, and an
749 optional <a href="#visibility">visibility style</a>.</p>
750
751 <h5>Syntax:</h5>
752
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]753<div class="doc_code">
Bill Wendlingaac388b2007-05-29 09:42:13 +0000[diff] [blame]754<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]755@&lt;Name&gt; = [Linkage] [Visibility] alias &lt;AliaseeTy&gt; @&lt;Aliasee&gt;
Bill Wendlingaac388b2007-05-29 09:42:13 +0000[diff] [blame]756</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]757</div>
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]758
759</div>
760
761
762
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]763<!-- ======================================================================= -->
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]764<div class="doc_subsection"><a name="paramattrs">Parameter Attributes</a></div>
765<div class="doc_text">
766 <p>The return type and each parameter of a function type may have a set of
767 <i>parameter attributes</i> associated with them. Parameter attributes are
768 used to communicate additional information about the result or parameters of
Duncan Sandsdc024672007-11-27 13:23:08 +0000[diff] [blame]769 a function. Parameter attributes are considered to be part of the function,
770 not of the function type, so functions with different parameter attributes
771 can have the same function type.</p>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]772
Reid Spencer950e9f82007-01-15 18:27:39 +0000[diff] [blame]773 <p>Parameter attributes are simple keywords that follow the type specified. If
774 multiple parameter attributes are needed, they are space separated. For
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]775 example:</p>
776
777<div class="doc_code">
778<pre>
Duncan Sandsdc024672007-11-27 13:23:08 +0000[diff] [blame]779declare i32 @printf(i8* noalias , ...) nounwind
780declare i32 @atoi(i8*) nounwind readonly
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]781</pre>
782</div>
783
Duncan Sandsdc024672007-11-27 13:23:08 +0000[diff] [blame]784 <p>Note that any attributes for the function result (<tt>nounwind</tt>,
785 <tt>readonly</tt>) come immediately after the argument list.</p>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]786
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame]787 <p>Currently, only the following parameter attributes are defined:</p>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]788 <dl>
Reid Spencer9445e9a2007-07-19 23:13:04 +0000[diff] [blame]789 <dt><tt>zeroext</tt></dt>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]790 <dd>This indicates that the parameter should be zero extended just before
791 a call to this function.</dd>
Reid Spencer9445e9a2007-07-19 23:13:04 +0000[diff] [blame]792 <dt><tt>signext</tt></dt>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]793 <dd>This indicates that the parameter should be sign extended just before
794 a call to this function.</dd>
Anton Korobeynikov34d6dea2007-01-28 14:30:45 +0000[diff] [blame]795 <dt><tt>inreg</tt></dt>
796 <dd>This indicates that the parameter should be placed in register (if
Anton Korobeynikov66a8c8c2007-01-28 15:27:21 +0000[diff] [blame]797 possible) during assembling function call. Support for this attribute is
798 target-specific</dd>
Anton Korobeynikov34d6dea2007-01-28 14:30:45 +0000[diff] [blame]799 <dt><tt>sret</tt></dt>
Anton Korobeynikov66a8c8c2007-01-28 15:27:21 +0000[diff] [blame]800 <dd>This indicates that the parameter specifies the address of a structure
Reid Spencer67606122007-03-22 02:02:11 +0000[diff] [blame]801 that is the return value of the function in the source program.</dd>
Zhou Shengfebca342007-06-05 05:28:26 +0000[diff] [blame]802 <dt><tt>noalias</tt></dt>
803 <dd>This indicates that the parameter not alias any other object or any
804 other "noalias" objects during the function call.
Reid Spencer2dc52012007-03-22 02:18:56 +0000[diff] [blame]805 <dt><tt>noreturn</tt></dt>
806 <dd>This function attribute indicates that the function never returns. This
807 indicates to LLVM that every call to this function should be treated as if
808 an <tt>unreachable</tt> instruction immediately followed the call.</dd>
Reid Spencer67606122007-03-22 02:02:11 +0000[diff] [blame]809 <dt><tt>nounwind</tt></dt>
810 <dd>This function attribute indicates that the function type does not use
811 the unwind instruction and does not allow stack unwinding to propagate
812 through it.</dd>
Duncan Sands50f19f52007-07-27 19:57:41 +0000[diff] [blame]813 <dt><tt>nest</tt></dt>
814 <dd>This indicates that the parameter can be excised using the
815 <a href="#int_trampoline">trampoline intrinsics</a>.</dd>
Duncan Sandsed4a2f12007-11-22 20:23:04 +0000[diff] [blame]816 <dt><tt>readonly</tt></dt>
Duncan Sandsf04d5842007-11-14 21:14:02 +0000[diff] [blame]817 <dd>This function attribute indicates that the function has no side-effects
Duncan Sandsed4a2f12007-11-22 20:23:04 +0000[diff] [blame]818 except for producing a return value or throwing an exception. The value
819 returned must only depend on the function arguments and/or global variables.
820 It may use values obtained by dereferencing pointers.</dd>
821 <dt><tt>readnone</tt></dt>
822 <dd>A <tt>readnone</tt> function has the same restrictions as a <tt>readonly</tt>
Duncan Sandsf04d5842007-11-14 21:14:02 +0000[diff] [blame]823 function, but in addition it is not allowed to dereference any pointer arguments
824 or global variables.
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame]825 </dl>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]826
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]827</div>
828
829<!-- ======================================================================= -->
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]830<div class="doc_subsection">
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame^]831 <a name="gc">Garbage Collector Names</a>
832</div>
833
834<div class="doc_text">
835<p>Each function may specify a garbage collector name, which is simply a
836string.</p>
837
838<div class="doc_code"><pre
839>define void @f() gc "name" { ...</pre></div>
840
841<p>The compiler declares the supported values of <i>name</i>. Specifying a
842collector which will cause the compiler to alter its output in order to support
843the named garbage collection algorithm.</p>
844</div>
845
846<!-- ======================================================================= -->
847<div class="doc_subsection">
Chris Lattner1eeeb0c2006-04-08 04:40:53 +0000[diff] [blame]848 <a name="moduleasm">Module-Level Inline Assembly</a>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]849</div>
850
851<div class="doc_text">
852<p>
853Modules may contain "module-level inline asm" blocks, which corresponds to the
854GCC "file scope inline asm" blocks. These blocks are internally concatenated by
855LLVM and treated as a single unit, but may be separated in the .ll file if
856desired. The syntax is very simple:
857</p>
858
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]859<div class="doc_code">
860<pre>
861module asm "inline asm code goes here"
862module asm "more can go here"
863</pre>
864</div>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]865
866<p>The strings can contain any character by escaping non-printable characters.
867 The escape sequence used is simply "\xx" where "xx" is the two digit hex code
868 for the number.
869</p>
870
871<p>
872 The inline asm code is simply printed to the machine code .s file when
873 assembly code is generated.
874</p>
875</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]876
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]877<!-- ======================================================================= -->
878<div class="doc_subsection">
879 <a name="datalayout">Data Layout</a>
880</div>
881
882<div class="doc_text">
883<p>A module may specify a target specific data layout string that specifies how
Reid Spencerc8910842007-04-11 23:49:50 +0000[diff] [blame]884data is to be laid out in memory. The syntax for the data layout is simply:</p>
885<pre> target datalayout = "<i>layout specification</i>"</pre>
886<p>The <i>layout specification</i> consists of a list of specifications
887separated by the minus sign character ('-'). Each specification starts with a
888letter and may include other information after the letter to define some
889aspect of the data layout. The specifications accepted are as follows: </p>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]890<dl>
891 <dt><tt>E</tt></dt>
892 <dd>Specifies that the target lays out data in big-endian form. That is, the
893 bits with the most significance have the lowest address location.</dd>
894 <dt><tt>e</tt></dt>
895 <dd>Specifies that hte target lays out data in little-endian form. That is,
896 the bits with the least significance have the lowest address location.</dd>
897 <dt><tt>p:<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
898 <dd>This specifies the <i>size</i> of a pointer and its <i>abi</i> and
899 <i>preferred</i> alignments. All sizes are in bits. Specifying the <i>pref</i>
900 alignment is optional. If omitted, the preceding <tt>:</tt> should be omitted
901 too.</dd>
902 <dt><tt>i<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
903 <dd>This specifies the alignment for an integer type of a given bit
904 <i>size</i>. The value of <i>size</i> must be in the range [1,2^23).</dd>
905 <dt><tt>v<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
906 <dd>This specifies the alignment for a vector type of a given bit
907 <i>size</i>.</dd>
908 <dt><tt>f<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
909 <dd>This specifies the alignment for a floating point type of a given bit
910 <i>size</i>. The value of <i>size</i> must be either 32 (float) or 64
911 (double).</dd>
912 <dt><tt>a<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
913 <dd>This specifies the alignment for an aggregate type of a given bit
914 <i>size</i>.</dd>
915</dl>
916<p>When constructing the data layout for a given target, LLVM starts with a
917default set of specifications which are then (possibly) overriden by the
918specifications in the <tt>datalayout</tt> keyword. The default specifications
919are given in this list:</p>
920<ul>
921 <li><tt>E</tt> - big endian</li>
922 <li><tt>p:32:64:64</tt> - 32-bit pointers with 64-bit alignment</li>
923 <li><tt>i1:8:8</tt> - i1 is 8-bit (byte) aligned</li>
924 <li><tt>i8:8:8</tt> - i8 is 8-bit (byte) aligned</li>
925 <li><tt>i16:16:16</tt> - i16 is 16-bit aligned</li>
926 <li><tt>i32:32:32</tt> - i32 is 32-bit aligned</li>
927 <li><tt>i64:32:64</tt> - i64 has abi alignment of 32-bits but preferred
928 alignment of 64-bits</li>
929 <li><tt>f32:32:32</tt> - float is 32-bit aligned</li>
930 <li><tt>f64:64:64</tt> - double is 64-bit aligned</li>
931 <li><tt>v64:64:64</tt> - 64-bit vector is 64-bit aligned</li>
932 <li><tt>v128:128:128</tt> - 128-bit vector is 128-bit aligned</li>
933 <li><tt>a0:0:1</tt> - aggregates are 8-bit aligned</li>
934</ul>
935<p>When llvm is determining the alignment for a given type, it uses the
936following rules:
937<ol>
938 <li>If the type sought is an exact match for one of the specifications, that
939 specification is used.</li>
940 <li>If no match is found, and the type sought is an integer type, then the
941 smallest integer type that is larger than the bitwidth of the sought type is
942 used. If none of the specifications are larger than the bitwidth then the the
943 largest integer type is used. For example, given the default specifications
944 above, the i7 type will use the alignment of i8 (next largest) while both
945 i65 and i256 will use the alignment of i64 (largest specified).</li>
946 <li>If no match is found, and the type sought is a vector type, then the
947 largest vector type that is smaller than the sought vector type will be used
948 as a fall back. This happens because <128 x double> can be implemented in
949 terms of 64 <2 x double>, for example.</li>
950</ol>
951</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]952
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]953<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]954<div class="doc_section"> <a name="typesystem">Type System</a> </div>
955<!-- *********************************************************************** -->
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]956
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]957<div class="doc_text">
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]958
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]959<p>The LLVM type system is one of the most important features of the
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]960intermediate representation. Being typed enables a number of
961optimizations to be performed on the IR directly, without having to do
962extra analyses on the side before the transformation. A strong type
963system makes it easier to read the generated code and enables novel
964analyses and transformations that are not feasible to perform on normal
965three address code representations.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]966
967</div>
968
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]969<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]970<div class="doc_subsection"> <a name="t_primitive">Primitive Types</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]971<div class="doc_text">
John Criswell4457dc92004-04-09 16:48:45 +0000[diff] [blame]972<p>The primitive types are the fundamental building blocks of the LLVM
Chris Lattnerd4f6b172005-03-07 22:13:59 +0000[diff] [blame]973system. The current set of primitive types is as follows:</p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]974
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]975<table class="layout">
976 <tr class="layout">
977 <td class="left">
978 <table>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]979 <tbody>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]980 <tr><th>Type</th><th>Description</th></tr>
Duncan Sands8036ca42007-03-30 12:22:09 +0000[diff] [blame]981 <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]982 <tr><td><tt>label</tt></td><td>Branch destination</td></tr>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]983 </tbody>
984 </table>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]985 </td>
986 <td class="right">
987 <table>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]988 <tbody>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]989 <tr><th>Type</th><th>Description</th></tr>
Reid Spencer2b916312007-05-16 18:44:01 +0000[diff] [blame]990 <tr><td><tt>float</tt></td><td>32-bit floating point value</td></tr>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]991 <tr><td><tt>double</tt></td><td>64-bit floating point value</td></tr>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]992 </tbody>
993 </table>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]994 </td>
995 </tr>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]996</table>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]997</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]998
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]999<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1000<div class="doc_subsubsection"> <a name="t_classifications">Type
1001Classifications</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1002<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1003<p>These different primitive types fall into a few useful
1004classifications:</p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1005
1006<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1007 <tbody>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1008 <tr><th>Classification</th><th>Types</th></tr>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1009 <tr>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1010 <td><a name="t_integer">integer</a></td>
Reid Spencer2b916312007-05-16 18:44:01 +0000[diff] [blame]1011 <td><tt>i1, i2, i3, ... i8, ... i16, ... i32, ... i64, ... </tt></td>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1012 </tr>
1013 <tr>
1014 <td><a name="t_floating">floating point</a></td>
1015 <td><tt>float, double</tt></td>
1016 </tr>
1017 <tr>
1018 <td><a name="t_firstclass">first class</a></td>
Reid Spencer2b916312007-05-16 18:44:01 +0000[diff] [blame]1019 <td><tt>i1, ..., float, double, <br/>
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1020 <a href="#t_pointer">pointer</a>,<a href="#t_vector">vector</a></tt>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1021 </td>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1022 </tr>
1023 </tbody>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1024</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1025
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1026<p>The <a href="#t_firstclass">first class</a> types are perhaps the
1027most important. Values of these types are the only ones which can be
1028produced by instructions, passed as arguments, or used as operands to
1029instructions. This means that all structures and arrays must be
1030manipulated either by pointer or by component.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1031</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1032
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1033<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1034<div class="doc_subsection"> <a name="t_derived">Derived Types</a> </div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1035
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1036<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1037
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1038<p>The real power in LLVM comes from the derived types in the system.
1039This is what allows a programmer to represent arrays, functions,
1040pointers, and other useful types. Note that these derived types may be
1041recursive: For example, it is possible to have a two dimensional array.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1042
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1043</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1044
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1045<!-- _______________________________________________________________________ -->
Reid Spencer2b916312007-05-16 18:44:01 +0000[diff] [blame]1046<div class="doc_subsubsection"> <a name="t_integer">Integer Type</a> </div>
1047
1048<div class="doc_text">
1049
1050<h5>Overview:</h5>
1051<p>The integer type is a very simple derived type that simply specifies an
1052arbitrary bit width for the integer type desired. Any bit width from 1 bit to
10532^23-1 (about 8 million) can be specified.</p>
1054
1055<h5>Syntax:</h5>
1056
1057<pre>
1058 iN
1059</pre>
1060
1061<p>The number of bits the integer will occupy is specified by the <tt>N</tt>
1062value.</p>
1063
1064<h5>Examples:</h5>
1065<table class="layout">
1066 <tr class="layout">
1067 <td class="left">
1068 <tt>i1</tt><br/>
1069 <tt>i4</tt><br/>
1070 <tt>i8</tt><br/>
1071 <tt>i16</tt><br/>
1072 <tt>i32</tt><br/>
1073 <tt>i42</tt><br/>
1074 <tt>i64</tt><br/>
1075 <tt>i1942652</tt><br/>
1076 </td>
1077 <td class="left">
1078 A boolean integer of 1 bit<br/>
1079 A nibble sized integer of 4 bits.<br/>
1080 A byte sized integer of 8 bits.<br/>
1081 A half word sized integer of 16 bits.<br/>
1082 A word sized integer of 32 bits.<br/>
1083 An integer whose bit width is the answer. <br/>
1084 A double word sized integer of 64 bits.<br/>
1085 A really big integer of over 1 million bits.<br/>
1086 </td>
1087 </tr>
1088</table>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1089</div>
Reid Spencer2b916312007-05-16 18:44:01 +0000[diff] [blame]1090
1091<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1092<div class="doc_subsubsection"> <a name="t_array">Array Type</a> </div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1093
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1094<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1095
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1096<h5>Overview:</h5>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1097
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1098<p>The array type is a very simple derived type that arranges elements
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1099sequentially in memory. The array type requires a size (number of
1100elements) and an underlying data type.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1101
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1102<h5>Syntax:</h5>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1103
1104<pre>
1105 [&lt;# elements&gt; x &lt;elementtype&gt;]
1106</pre>
1107
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]1108<p>The number of elements is a constant integer value; elementtype may
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1109be any type with a size.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1110
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1111<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1112<table class="layout">
1113 <tr class="layout">
1114 <td class="left">
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1115 <tt>[40 x i32 ]</tt><br/>
1116 <tt>[41 x i32 ]</tt><br/>
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]1117 <tt>[40 x i8]</tt><br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1118 </td>
1119 <td class="left">
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]1120 Array of 40 32-bit integer values.<br/>
1121 Array of 41 32-bit integer values.<br/>
1122 Array of 40 8-bit integer values.<br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1123 </td>
1124 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1125</table>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1126<p>Here are some examples of multidimensional arrays:</p>
1127<table class="layout">
1128 <tr class="layout">
1129 <td class="left">
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1130 <tt>[3 x [4 x i32]]</tt><br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1131 <tt>[12 x [10 x float]]</tt><br/>
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]1132 <tt>[2 x [3 x [4 x i16]]]</tt><br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1133 </td>
1134 <td class="left">
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]1135 3x4 array of 32-bit integer values.<br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1136 12x10 array of single precision floating point values.<br/>
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]1137 2x3x4 array of 16-bit integer values.<br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1138 </td>
1139 </tr>
1140</table>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]1141
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]1142<p>Note that 'variable sized arrays' can be implemented in LLVM with a zero
1143length array. Normally, accesses past the end of an array are undefined in
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]1144LLVM (e.g. it is illegal to access the 5th element of a 3 element array).
1145As a special case, however, zero length arrays are recognized to be variable
1146length. This allows implementation of 'pascal style arrays' with the LLVM
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1147type "{ i32, [0 x float]}", for example.</p>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]1148
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1149</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1150
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1151<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1152<div class="doc_subsubsection"> <a name="t_function">Function Type</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1153<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1154<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1155<p>The function type can be thought of as a function signature. It
1156consists of a return type and a list of formal parameter types.
John Criswell009900b2003-11-25 21:45:46 +0000[diff] [blame]1157Function types are usually used to build virtual function tables
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1158(which are structures of pointers to functions), for indirect function
1159calls, and when defining a function.</p>
John Criswell009900b2003-11-25 21:45:46 +0000[diff] [blame]1160<p>
1161The return type of a function type cannot be an aggregate type.
1162</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1163<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1164<pre> &lt;returntype&gt; (&lt;parameter list&gt;)<br></pre>
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]1165<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]1166specifiers. Optionally, the parameter list may include a type <tt>...</tt>,
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]1167which indicates that the function takes a variable number of arguments.
1168Variable argument functions can access their arguments with the <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1169 href="#int_varargs">variable argument handling intrinsic</a> functions.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1170<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1171<table class="layout">
1172 <tr class="layout">
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]1173 <td class="left"><tt>i32 (i32)</tt></td>
1174 <td class="left">function taking an <tt>i32</tt>, returning an <tt>i32</tt>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1175 </td>
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]1176 </tr><tr class="layout">
Reid Spencer9445e9a2007-07-19 23:13:04 +0000[diff] [blame]1177 <td class="left"><tt>float&nbsp;(i16&nbsp;signext,&nbsp;i32&nbsp;*)&nbsp;*
Reid Spencerf17a0b72006-12-31 07:20:23 +0000[diff] [blame]1178 </tt></td>
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]1179 <td class="left"><a href="#t_pointer">Pointer</a> to a function that takes
1180 an <tt>i16</tt> that should be sign extended and a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1181 <a href="#t_pointer">pointer</a> to <tt>i32</tt>, returning
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]1182 <tt>float</tt>.
1183 </td>
1184 </tr><tr class="layout">
1185 <td class="left"><tt>i32 (i8*, ...)</tt></td>
1186 <td class="left">A vararg function that takes at least one
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]1187 <a href="#t_pointer">pointer</a> to <tt>i8 </tt> (char in C),
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]1188 which returns an integer. This is the signature for <tt>printf</tt> in
1189 LLVM.
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1190 </td>
1191 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1192</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1193
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1194</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1195<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1196<div class="doc_subsubsection"> <a name="t_struct">Structure Type</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1197<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1198<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1199<p>The structure type is used to represent a collection of data members
1200together in memory. The packing of the field types is defined to match
1201the ABI of the underlying processor. The elements of a structure may
1202be any type that has a size.</p>
1203<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
1204and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
1205field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
1206instruction.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1207<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1208<pre> { &lt;type list&gt; }<br></pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1209<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1210<table class="layout">
1211 <tr class="layout">
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]1212 <td class="left"><tt>{ i32, i32, i32 }</tt></td>
1213 <td class="left">A triple of three <tt>i32</tt> values</td>
1214 </tr><tr class="layout">
1215 <td class="left"><tt>{&nbsp;float,&nbsp;i32&nbsp;(i32)&nbsp;*&nbsp;}</tt></td>
1216 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1217 second element is a <a href="#t_pointer">pointer</a> to a
1218 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1219 an <tt>i32</tt>.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1220 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1221</table>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1222</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1223
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1224<!-- _______________________________________________________________________ -->
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]1225<div class="doc_subsubsection"> <a name="t_pstruct">Packed Structure Type</a>
1226</div>
1227<div class="doc_text">
1228<h5>Overview:</h5>
1229<p>The packed structure type is used to represent a collection of data members
1230together in memory. There is no padding between fields. Further, the alignment
1231of a packed structure is 1 byte. The elements of a packed structure may
1232be any type that has a size.</p>
1233<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
1234and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
1235field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
1236instruction.</p>
1237<h5>Syntax:</h5>
1238<pre> &lt; { &lt;type list&gt; } &gt; <br></pre>
1239<h5>Examples:</h5>
1240<table class="layout">
1241 <tr class="layout">
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]1242 <td class="left"><tt>&lt; { i32, i32, i32 } &gt;</tt></td>
1243 <td class="left">A triple of three <tt>i32</tt> values</td>
1244 </tr><tr class="layout">
1245 <td class="left"><tt>&lt;&nbsp;{&nbsp;float,&nbsp;i32&nbsp;(i32)&nbsp;*&nbsp;}&nbsp;&gt;</tt></td>
1246 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1247 second element is a <a href="#t_pointer">pointer</a> to a
1248 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1249 an <tt>i32</tt>.</td>
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]1250 </tr>
1251</table>
1252</div>
1253
1254<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1255<div class="doc_subsubsection"> <a name="t_pointer">Pointer Type</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1256<div class="doc_text">
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1257<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1258<p>As in many languages, the pointer type represents a pointer or
1259reference to another object, which must live in memory.</p>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1260<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1261<pre> &lt;type&gt; *<br></pre>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1262<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1263<table class="layout">
1264 <tr class="layout">
1265 <td class="left">
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1266 <tt>[4x i32]*</tt><br/>
1267 <tt>i32 (i32 *) *</tt><br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1268 </td>
1269 <td class="left">
1270 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]1271 four <tt>i32</tt> values<br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1272 A <a href="#t_pointer">pointer</a> to a <a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1273 href="#t_function">function</a> that takes an <tt>i32*</tt>, returning an
1274 <tt>i32</tt>.<br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1275 </td>
1276 </tr>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1277</table>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1278</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1279
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1280<!-- _______________________________________________________________________ -->
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1281<div class="doc_subsubsection"> <a name="t_vector">Vector Type</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1282<div class="doc_text">
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1283
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1284<h5>Overview:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1285
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1286<p>A vector type is a simple derived type that represents a vector
1287of elements. Vector types are used when multiple primitive data
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1288are operated in parallel using a single instruction (SIMD).
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1289A vector type requires a size (number of
Chris Lattnerb8d172f2005-11-10 01:44:22 +0000[diff] [blame]1290elements) and an underlying primitive data type. Vectors must have a power
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1291of two length (1, 2, 4, 8, 16 ...). Vector types are
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1292considered <a href="#t_firstclass">first class</a>.</p>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1293
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1294<h5>Syntax:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1295
1296<pre>
1297 &lt; &lt;# elements&gt; x &lt;elementtype&gt; &gt;
1298</pre>
1299
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]1300<p>The number of elements is a constant integer value; elementtype may
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]1301be any integer or floating point type.</p>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1302
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1303<h5>Examples:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1304
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1305<table class="layout">
1306 <tr class="layout">
1307 <td class="left">
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1308 <tt>&lt;4 x i32&gt;</tt><br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1309 <tt>&lt;8 x float&gt;</tt><br/>
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]1310 <tt>&lt;2 x i64&gt;</tt><br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1311 </td>
1312 <td class="left">
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1313 Vector of 4 32-bit integer values.<br/>
1314 Vector of 8 floating-point values.<br/>
1315 Vector of 2 64-bit integer values.<br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1316 </td>
1317 </tr>
1318</table>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1319</div>
1320
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1321<!-- _______________________________________________________________________ -->
1322<div class="doc_subsubsection"> <a name="t_opaque">Opaque Type</a> </div>
1323<div class="doc_text">
1324
1325<h5>Overview:</h5>
1326
1327<p>Opaque types are used to represent unknown types in the system. This
Gordon Henriksen8ac04ff2007-10-14 00:34:53 +0000[diff] [blame]1328corresponds (for example) to the C notion of a forward declared structure type.
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1329In LLVM, opaque types can eventually be resolved to any type (not just a
1330structure type).</p>
1331
1332<h5>Syntax:</h5>
1333
1334<pre>
1335 opaque
1336</pre>
1337
1338<h5>Examples:</h5>
1339
1340<table class="layout">
1341 <tr class="layout">
1342 <td class="left">
1343 <tt>opaque</tt>
1344 </td>
1345 <td class="left">
1346 An opaque type.<br/>
1347 </td>
1348 </tr>
1349</table>
1350</div>
1351
1352
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1353<!-- *********************************************************************** -->
1354<div class="doc_section"> <a name="constants">Constants</a> </div>
1355<!-- *********************************************************************** -->
1356
1357<div class="doc_text">
1358
1359<p>LLVM has several different basic types of constants. This section describes
1360them all and their syntax.</p>
1361
1362</div>
1363
1364<!-- ======================================================================= -->
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]1365<div class="doc_subsection"><a name="simpleconstants">Simple Constants</a></div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1366
1367<div class="doc_text">
1368
1369<dl>
1370 <dt><b>Boolean constants</b></dt>
1371
1372 <dd>The two strings '<tt>true</tt>' and '<tt>false</tt>' are both valid
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]1373 constants of the <tt><a href="#t_primitive">i1</a></tt> type.
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1374 </dd>
1375
1376 <dt><b>Integer constants</b></dt>
1377
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]1378 <dd>Standard integers (such as '4') are constants of the <a
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]1379 href="#t_integer">integer</a> type. Negative numbers may be used with
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1380 integer types.
1381 </dd>
1382
1383 <dt><b>Floating point constants</b></dt>
1384
1385 <dd>Floating point constants use standard decimal notation (e.g. 123.421),
1386 exponential notation (e.g. 1.23421e+2), or a more precise hexadecimal
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1387 notation (see below). Floating point constants must have a <a
1388 href="#t_floating">floating point</a> type. </dd>
1389
1390 <dt><b>Null pointer constants</b></dt>
1391
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1392 <dd>The identifier '<tt>null</tt>' is recognized as a null pointer constant
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1393 and must be of <a href="#t_pointer">pointer type</a>.</dd>
1394
1395</dl>
1396
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1397<p>The one non-intuitive notation for constants is the optional hexadecimal form
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1398of floating point constants. For example, the form '<tt>double
13990x432ff973cafa8000</tt>' is equivalent to (but harder to read than) '<tt>double
14004.5e+15</tt>'. The only time hexadecimal floating point constants are required
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]1401(and the only time that they are generated by the disassembler) is when a
1402floating point constant must be emitted but it cannot be represented as a
1403decimal floating point number. For example, NaN's, infinities, and other
1404special values are represented in their IEEE hexadecimal format so that
1405assembly and disassembly do not cause any bits to change in the constants.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1406
1407</div>
1408
1409<!-- ======================================================================= -->
1410<div class="doc_subsection"><a name="aggregateconstants">Aggregate Constants</a>
1411</div>
1412
1413<div class="doc_text">
Chris Lattnerd4f6b172005-03-07 22:13:59 +0000[diff] [blame]1414<p>Aggregate constants arise from aggregation of simple constants
1415and smaller aggregate constants.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1416
1417<dl>
1418 <dt><b>Structure constants</b></dt>
1419
1420 <dd>Structure constants are represented with notation similar to structure
1421 type definitions (a comma separated list of elements, surrounded by braces
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1422 (<tt>{}</tt>)). For example: "<tt>{ i32 4, float 17.0, i32* %G }</tt>",
Chris Lattner3e63a9d2007-07-13 20:01:46 +0000[diff] [blame]1423 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]1424 must have <a href="#t_struct">structure type</a>, and the number and
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1425 types of elements must match those specified by the type.
1426 </dd>
1427
1428 <dt><b>Array constants</b></dt>
1429
1430 <dd>Array constants are represented with notation similar to array type
1431 definitions (a comma separated list of elements, surrounded by square brackets
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1432 (<tt>[]</tt>)). For example: "<tt>[ i32 42, i32 11, i32 74 ]</tt>". Array
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1433 constants must have <a href="#t_array">array type</a>, and the number and
1434 types of elements must match those specified by the type.
1435 </dd>
1436
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1437 <dt><b>Vector constants</b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1438
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1439 <dd>Vector constants are represented with notation similar to vector type
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1440 definitions (a comma separated list of elements, surrounded by
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1441 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]1442 i32 11, i32 74, i32 100 &gt;</tt>". Vector constants must have <a
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1443 href="#t_vector">vector type</a>, and the number and types of elements must
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1444 match those specified by the type.
1445 </dd>
1446
1447 <dt><b>Zero initialization</b></dt>
1448
1449 <dd>The string '<tt>zeroinitializer</tt>' can be used to zero initialize a
1450 value to zero of <em>any</em> type, including scalar and aggregate types.
1451 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]1452 large arrays) and is always exactly equivalent to using explicit zero
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1453 initializers.
1454 </dd>
1455</dl>
1456
1457</div>
1458
1459<!-- ======================================================================= -->
1460<div class="doc_subsection">
1461 <a name="globalconstants">Global Variable and Function Addresses</a>
1462</div>
1463
1464<div class="doc_text">
1465
1466<p>The addresses of <a href="#globalvars">global variables</a> and <a
1467href="#functionstructure">functions</a> are always implicitly valid (link-time)
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1468constants. These constants are explicitly referenced when the <a
1469href="#identifiers">identifier for the global</a> is used and always have <a
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1470href="#t_pointer">pointer</a> type. For example, the following is a legal LLVM
1471file:</p>
1472
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1473<div class="doc_code">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1474<pre>
Chris Lattnera18a4242007-06-06 18:28:13 +0000[diff] [blame]1475@X = global i32 17
1476@Y = global i32 42
1477@Z = global [2 x i32*] [ i32* @X, i32* @Y ]
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1478</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1479</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1480
1481</div>
1482
1483<!-- ======================================================================= -->
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]1484<div class="doc_subsection"><a name="undefvalues">Undefined Values</a></div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1485<div class="doc_text">
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]1486 <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]1487 no specific value. Undefined values may be of any type and be used anywhere
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]1488 a constant is permitted.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1489
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]1490 <p>Undefined values indicate to the compiler that the program is well defined
1491 no matter what value is used, giving the compiler more freedom to optimize.
1492 </p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1493</div>
1494
1495<!-- ======================================================================= -->
1496<div class="doc_subsection"><a name="constantexprs">Constant Expressions</a>
1497</div>
1498
1499<div class="doc_text">
1500
1501<p>Constant expressions are used to allow expressions involving other constants
1502to be used as constants. Constant expressions may be of any <a
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]1503href="#t_firstclass">first class</a> type and may involve any LLVM operation
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1504that does not have side effects (e.g. load and call are not supported). The
1505following is the syntax for constant expressions:</p>
1506
1507<dl>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1508 <dt><b><tt>trunc ( CST to TYPE )</tt></b></dt>
1509 <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]1510 than the bit size of TYPE. Both types must be integers.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1511
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1512 <dt><b><tt>zext ( CST to TYPE )</tt></b></dt>
1513 <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]1514 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]1515
1516 <dt><b><tt>sext ( CST to TYPE )</tt></b></dt>
1517 <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]1518 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]1519
1520 <dt><b><tt>fptrunc ( CST to TYPE )</tt></b></dt>
1521 <dd>Truncate a floating point constant to another floating point type. The
1522 size of CST must be larger than the size of TYPE. Both types must be
1523 floating point.</dd>
1524
1525 <dt><b><tt>fpext ( CST to TYPE )</tt></b></dt>
1526 <dd>Floating point extend a constant to another type. The size of CST must be
1527 smaller or equal to the size of TYPE. Both types must be floating point.</dd>
1528
Reid Spencer1539a1c2007-07-31 14:40:14 +0000[diff] [blame]1529 <dt><b><tt>fptoui ( CST to TYPE )</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1530 <dd>Convert a floating point constant to the corresponding unsigned integer
Nate Begemanb348d182007-11-17 03:58:34 +0000[diff] [blame]1531 constant. TYPE must be a scalar or vector integer type. CST must be of scalar
1532 or vector floating point type. Both CST and TYPE must be scalars, or vectors
1533 of the same number of elements. If the value won't fit in the integer type,
1534 the results are undefined.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1535
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]1536 <dt><b><tt>fptosi ( CST to TYPE )</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1537 <dd>Convert a floating point constant to the corresponding signed integer
Nate Begemanb348d182007-11-17 03:58:34 +0000[diff] [blame]1538 constant. TYPE must be a scalar or vector integer type. CST must be of scalar
1539 or vector floating point type. Both CST and TYPE must be scalars, or vectors
1540 of the same number of elements. If the value won't fit in the integer type,
1541 the results are undefined.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1542
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]1543 <dt><b><tt>uitofp ( CST to TYPE )</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1544 <dd>Convert an unsigned integer constant to the corresponding floating point
Nate Begemanb348d182007-11-17 03:58:34 +0000[diff] [blame]1545 constant. TYPE must be a scalar or vector floating point type. CST must be of
1546 scalar or vector integer type. Both CST and TYPE must be scalars, or vectors
1547 of the same number of elements. If the value won't fit in the floating point
1548 type, the results are undefined.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1549
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]1550 <dt><b><tt>sitofp ( CST to TYPE )</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1551 <dd>Convert a signed integer constant to the corresponding floating point
Nate Begemanb348d182007-11-17 03:58:34 +0000[diff] [blame]1552 constant. TYPE must be a scalar or vector floating point type. CST must be of
1553 scalar or vector integer type. Both CST and TYPE must be scalars, or vectors
1554 of the same number of elements. If the value won't fit in the floating point
1555 type, the results are undefined.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1556
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]1557 <dt><b><tt>ptrtoint ( CST to TYPE )</tt></b></dt>
1558 <dd>Convert a pointer typed constant to the corresponding integer constant
1559 TYPE must be an integer type. CST must be of pointer type. The CST value is
1560 zero extended, truncated, or unchanged to make it fit in TYPE.</dd>
1561
1562 <dt><b><tt>inttoptr ( CST to TYPE )</tt></b></dt>
1563 <dd>Convert a integer constant to a pointer constant. TYPE must be a
1564 pointer type. CST must be of integer type. The CST value is zero extended,
1565 truncated, or unchanged to make it fit in a pointer size. This one is
1566 <i>really</i> dangerous!</dd>
1567
1568 <dt><b><tt>bitcast ( CST to TYPE )</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1569 <dd>Convert a constant, CST, to another TYPE. The size of CST and TYPE must be
1570 identical (same number of bits). The conversion is done as if the CST value
1571 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]1572 with this operator, just the type. This can be used for conversion of
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1573 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]1574 pointers it is only valid to cast to another pointer type.
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1575 </dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1576
1577 <dt><b><tt>getelementptr ( CSTPTR, IDX0, IDX1, ... )</tt></b></dt>
1578
1579 <dd>Perform the <a href="#i_getelementptr">getelementptr operation</a> on
1580 constants. As with the <a href="#i_getelementptr">getelementptr</a>
1581 instruction, the index list may have zero or more indexes, which are required
1582 to make sense for the type of "CSTPTR".</dd>
1583
Robert Bocchino9fbe1452006-01-10 19:31:34 +0000[diff] [blame]1584 <dt><b><tt>select ( COND, VAL1, VAL2 )</tt></b></dt>
1585
1586 <dd>Perform the <a href="#i_select">select operation</a> on
Reid Spencer01c42592006-12-04 19:23:19 +0000[diff] [blame]1587 constants.</dd>
1588
1589 <dt><b><tt>icmp COND ( VAL1, VAL2 )</tt></b></dt>
1590 <dd>Performs the <a href="#i_icmp">icmp operation</a> on constants.</dd>
1591
1592 <dt><b><tt>fcmp COND ( VAL1, VAL2 )</tt></b></dt>
1593 <dd>Performs the <a href="#i_fcmp">fcmp operation</a> on constants.</dd>
Robert Bocchino9fbe1452006-01-10 19:31:34 +0000[diff] [blame]1594
1595 <dt><b><tt>extractelement ( VAL, IDX )</tt></b></dt>
1596
1597 <dd>Perform the <a href="#i_extractelement">extractelement
1598 operation</a> on constants.
1599
Robert Bocchino05ccd702006-01-15 20:48:27 +0000[diff] [blame]1600 <dt><b><tt>insertelement ( VAL, ELT, IDX )</tt></b></dt>
1601
1602 <dd>Perform the <a href="#i_insertelement">insertelement
Reid Spencer01c42592006-12-04 19:23:19 +0000[diff] [blame]1603 operation</a> on constants.</dd>
Robert Bocchino05ccd702006-01-15 20:48:27 +0000[diff] [blame]1604
Chris Lattnerc1989542006-04-08 00:13:41 +0000[diff] [blame]1605
1606 <dt><b><tt>shufflevector ( VEC1, VEC2, IDXMASK )</tt></b></dt>
1607
1608 <dd>Perform the <a href="#i_shufflevector">shufflevector
Reid Spencer01c42592006-12-04 19:23:19 +0000[diff] [blame]1609 operation</a> on constants.</dd>
Chris Lattnerc1989542006-04-08 00:13:41 +0000[diff] [blame]1610
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1611 <dt><b><tt>OPCODE ( LHS, RHS )</tt></b></dt>
1612
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]1613 <dd>Perform the specified operation of the LHS and RHS constants. OPCODE may
1614 be any of the <a href="#binaryops">binary</a> or <a href="#bitwiseops">bitwise
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1615 binary</a> operations. The constraints on operands are the same as those for
1616 the corresponding instruction (e.g. no bitwise operations on floating point
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]1617 values are allowed).</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1618</dl>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1619</div>
Chris Lattner9ee5d222004-03-08 16:49:10 +0000[diff] [blame]1620
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1621<!-- *********************************************************************** -->
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]1622<div class="doc_section"> <a name="othervalues">Other Values</a> </div>
1623<!-- *********************************************************************** -->
1624
1625<!-- ======================================================================= -->
1626<div class="doc_subsection">
1627<a name="inlineasm">Inline Assembler Expressions</a>
1628</div>
1629
1630<div class="doc_text">
1631
1632<p>
1633LLVM supports inline assembler expressions (as opposed to <a href="#moduleasm">
1634Module-Level Inline Assembly</a>) through the use of a special value. This
1635value represents the inline assembler as a string (containing the instructions
1636to emit), a list of operand constraints (stored as a string), and a flag that
1637indicates whether or not the inline asm expression has side effects. An example
1638inline assembler expression is:
1639</p>
1640
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1641<div class="doc_code">
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]1642<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1643i32 (i32) asm "bswap $0", "=r,r"
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]1644</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1645</div>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]1646
1647<p>
1648Inline assembler expressions may <b>only</b> be used as the callee operand of
1649a <a href="#i_call"><tt>call</tt> instruction</a>. Thus, typically we have:
1650</p>
1651
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1652<div class="doc_code">
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]1653<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1654%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]1655</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1656</div>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]1657
1658<p>
1659Inline asms with side effects not visible in the constraint list must be marked
1660as having side effects. This is done through the use of the
1661'<tt>sideeffect</tt>' keyword, like so:
1662</p>
1663
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1664<div class="doc_code">
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]1665<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1666call void asm sideeffect "eieio", ""()
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]1667</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1668</div>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]1669
1670<p>TODO: The format of the asm and constraints string still need to be
1671documented here. Constraints on what can be done (e.g. duplication, moving, etc
1672need to be documented).
1673</p>
1674
1675</div>
1676
1677<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1678<div class="doc_section"> <a name="instref">Instruction Reference</a> </div>
1679<!-- *********************************************************************** -->
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1680
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1681<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1682
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1683<p>The LLVM instruction set consists of several different
1684classifications of instructions: <a href="#terminators">terminator
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]1685instructions</a>, <a href="#binaryops">binary instructions</a>,
1686<a href="#bitwiseops">bitwise binary instructions</a>, <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1687 href="#memoryops">memory instructions</a>, and <a href="#otherops">other
1688instructions</a>.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1689
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1690</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1691
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1692<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1693<div class="doc_subsection"> <a name="terminators">Terminator
1694Instructions</a> </div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1695
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1696<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1697
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1698<p>As mentioned <a href="#functionstructure">previously</a>, every
1699basic block in a program ends with a "Terminator" instruction, which
1700indicates which block should be executed after the current block is
1701finished. These terminator instructions typically yield a '<tt>void</tt>'
1702value: they produce control flow, not values (the one exception being
1703the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction).</p>
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1704<p>There are six different terminator instructions: the '<a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1705 href="#i_ret"><tt>ret</tt></a>' instruction, the '<a href="#i_br"><tt>br</tt></a>'
1706instruction, the '<a href="#i_switch"><tt>switch</tt></a>' instruction,
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1707the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction, the '<a
1708 href="#i_unwind"><tt>unwind</tt></a>' instruction, and the '<a
1709 href="#i_unreachable"><tt>unreachable</tt></a>' instruction.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1710
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1711</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1712
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1713<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1714<div class="doc_subsubsection"> <a name="i_ret">'<tt>ret</tt>'
1715Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1716<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1717<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1718<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]1719 ret void <i>; Return from void function</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1720</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1721<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1722<p>The '<tt>ret</tt>' instruction is used to return control flow (and a
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]1723value) from a function back to the caller.</p>
John Criswell4457dc92004-04-09 16:48:45 +0000[diff] [blame]1724<p>There are two forms of the '<tt>ret</tt>' instruction: one that
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1725returns a value and then causes control flow, and one that just causes
1726control flow to occur.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1727<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1728<p>The '<tt>ret</tt>' instruction may return any '<a
1729 href="#t_firstclass">first class</a>' type. Notice that a function is
1730not <a href="#wellformed">well formed</a> if there exists a '<tt>ret</tt>'
1731instruction inside of the function that returns a value that does not
1732match the return type of the function.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1733<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1734<p>When the '<tt>ret</tt>' instruction is executed, control flow
1735returns back to the calling function's context. If the caller is a "<a
John Criswellfa081872004-06-25 15:16:57 +0000[diff] [blame]1736 href="#i_call"><tt>call</tt></a>" instruction, execution continues at
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1737the instruction after the call. If the caller was an "<a
1738 href="#i_invoke"><tt>invoke</tt></a>" instruction, execution continues
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]1739at the beginning of the "normal" destination block. If the instruction
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1740returns a value, that value shall set the call or invoke instruction's
1741return value.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1742<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1743<pre> ret i32 5 <i>; Return an integer value of 5</i>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1744 ret void <i>; Return from a void function</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1745</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1746</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1747<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1748<div class="doc_subsubsection"> <a name="i_br">'<tt>br</tt>' Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1749<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1750<h5>Syntax:</h5>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]1751<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]1752</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1753<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1754<p>The '<tt>br</tt>' instruction is used to cause control flow to
1755transfer to a different basic block in the current function. There are
1756two forms of this instruction, corresponding to a conditional branch
1757and an unconditional branch.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1758<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1759<p>The conditional branch form of the '<tt>br</tt>' instruction takes a
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]1760single '<tt>i1</tt>' value and two '<tt>label</tt>' values. The
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1761unconditional form of the '<tt>br</tt>' instruction takes a single
1762'<tt>label</tt>' value as a target.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1763<h5>Semantics:</h5>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]1764<p>Upon execution of a conditional '<tt>br</tt>' instruction, the '<tt>i1</tt>'
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1765argument is evaluated. If the value is <tt>true</tt>, control flows
1766to the '<tt>iftrue</tt>' <tt>label</tt> argument. If "cond" is <tt>false</tt>,
1767control flows to the '<tt>iffalse</tt>' <tt>label</tt> argument.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1768<h5>Example:</h5>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]1769<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]1770 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]1771</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1772<!-- _______________________________________________________________________ -->
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1773<div class="doc_subsubsection">
1774 <a name="i_switch">'<tt>switch</tt>' Instruction</a>
1775</div>
1776
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1777<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1778<h5>Syntax:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1779
1780<pre>
1781 switch &lt;intty&gt; &lt;value&gt;, label &lt;defaultdest&gt; [ &lt;intty&gt; &lt;val&gt;, label &lt;dest&gt; ... ]
1782</pre>
1783
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1784<h5>Overview:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1785
1786<p>The '<tt>switch</tt>' instruction is used to transfer control flow to one of
1787several different places. It is a generalization of the '<tt>br</tt>'
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1788instruction, allowing a branch to occur to one of many possible
1789destinations.</p>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1790
1791
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1792<h5>Arguments:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1793
1794<p>The '<tt>switch</tt>' instruction uses three parameters: an integer
1795comparison value '<tt>value</tt>', a default '<tt>label</tt>' destination, and
1796an array of pairs of comparison value constants and '<tt>label</tt>'s. The
1797table is not allowed to contain duplicate constant entries.</p>
1798
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1799<h5>Semantics:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1800
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1801<p>The <tt>switch</tt> instruction specifies a table of values and
1802destinations. When the '<tt>switch</tt>' instruction is executed, this
John Criswell84114752004-06-25 16:05:06 +0000[diff] [blame]1803table is searched for the given value. If the value is found, control flow is
1804transfered to the corresponding destination; otherwise, control flow is
1805transfered to the default destination.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1806
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1807<h5>Implementation:</h5>
1808
1809<p>Depending on properties of the target machine and the particular
1810<tt>switch</tt> instruction, this instruction may be code generated in different
John Criswell84114752004-06-25 16:05:06 +0000[diff] [blame]1811ways. For example, it could be generated as a series of chained conditional
1812branches or with a lookup table.</p>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1813
1814<h5>Example:</h5>
1815
1816<pre>
1817 <i>; Emulate a conditional br instruction</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]1818 %Val = <a href="#i_zext">zext</a> i1 %value to i32
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1819 switch i32 %Val, label %truedest [i32 0, label %falsedest ]
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1820
1821 <i>; Emulate an unconditional br instruction</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1822 switch i32 0, label %dest [ ]
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1823
1824 <i>; Implement a jump table:</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1825 switch i32 %val, label %otherwise [ i32 0, label %onzero
1826 i32 1, label %onone
1827 i32 2, label %ontwo ]
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1828</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1829</div>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1830
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1831<!-- _______________________________________________________________________ -->
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1832<div class="doc_subsubsection">
1833 <a name="i_invoke">'<tt>invoke</tt>' Instruction</a>
1834</div>
1835
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1836<div class="doc_text">
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1837
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1838<h5>Syntax:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1839
1840<pre>
1841 &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]1842 to label &lt;normal label&gt; unwind label &lt;exception label&gt;
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1843</pre>
1844
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]1845<h5>Overview:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1846
1847<p>The '<tt>invoke</tt>' instruction causes control to transfer to a specified
1848function, with the possibility of control flow transfer to either the
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]1849'<tt>normal</tt>' label or the
1850'<tt>exception</tt>' label. If the callee function returns with the
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1851"<tt><a href="#i_ret">ret</a></tt>" instruction, control flow will return to the
1852"normal" label. If the callee (or any indirect callees) returns with the "<a
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]1853href="#i_unwind"><tt>unwind</tt></a>" instruction, control is interrupted and
1854continued at the dynamically nearest "exception" label.</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1855
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1856<h5>Arguments:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1857
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1858<p>This instruction requires several arguments:</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1859
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1860<ol>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1861 <li>
Duncan Sands8036ca42007-03-30 12:22:09 +0000[diff] [blame]1862 The optional "cconv" marker indicates which <a href="#callingconv">calling
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1863 convention</a> the call should use. If none is specified, the call defaults
1864 to using C calling conventions.
1865 </li>
1866 <li>'<tt>ptr to function ty</tt>': shall be the signature of the pointer to
1867 function value being invoked. In most cases, this is a direct function
1868 invocation, but indirect <tt>invoke</tt>s are just as possible, branching off
1869 an arbitrary pointer to function value.
1870 </li>
1871
1872 <li>'<tt>function ptr val</tt>': An LLVM value containing a pointer to a
1873 function to be invoked. </li>
1874
1875 <li>'<tt>function args</tt>': argument list whose types match the function
1876 signature argument types. If the function signature indicates the function
1877 accepts a variable number of arguments, the extra arguments can be
1878 specified. </li>
1879
1880 <li>'<tt>normal label</tt>': the label reached when the called function
1881 executes a '<tt><a href="#i_ret">ret</a></tt>' instruction. </li>
1882
1883 <li>'<tt>exception label</tt>': the label reached when a callee returns with
1884 the <a href="#i_unwind"><tt>unwind</tt></a> instruction. </li>
1885
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1886</ol>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1887
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1888<h5>Semantics:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1889
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1890<p>This instruction is designed to operate as a standard '<tt><a
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1891href="#i_call">call</a></tt>' instruction in most regards. The primary
1892difference is that it establishes an association with a label, which is used by
1893the runtime library to unwind the stack.</p>
1894
1895<p>This instruction is used in languages with destructors to ensure that proper
1896cleanup is performed in the case of either a <tt>longjmp</tt> or a thrown
1897exception. Additionally, this is important for implementation of
1898'<tt>catch</tt>' clauses in high-level languages that support them.</p>
1899
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1900<h5>Example:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1901<pre>
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]1902 %retval = invoke i32 %Test(i32 15) to label %Continue
1903 unwind label %TestCleanup <i>; {i32}:retval set</i>
1904 %retval = invoke <a href="#callingconv">coldcc</a> i32 %Test(i32 15) to label %Continue
1905 unwind label %TestCleanup <i>; {i32}:retval set</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1906</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1907</div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1908
1909
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]1910<!-- _______________________________________________________________________ -->
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1911
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1912<div class="doc_subsubsection"> <a name="i_unwind">'<tt>unwind</tt>'
1913Instruction</a> </div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1914
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1915<div class="doc_text">
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1916
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]1917<h5>Syntax:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1918<pre>
1919 unwind
1920</pre>
1921
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]1922<h5>Overview:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1923
1924<p>The '<tt>unwind</tt>' instruction unwinds the stack, continuing control flow
1925at the first callee in the dynamic call stack which used an <a
1926href="#i_invoke"><tt>invoke</tt></a> instruction to perform the call. This is
1927primarily used to implement exception handling.</p>
1928
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]1929<h5>Semantics:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1930
1931<p>The '<tt>unwind</tt>' intrinsic causes execution of the current function to
1932immediately halt. The dynamic call stack is then searched for the first <a
1933href="#i_invoke"><tt>invoke</tt></a> instruction on the call stack. Once found,
1934execution continues at the "exceptional" destination block specified by the
1935<tt>invoke</tt> instruction. If there is no <tt>invoke</tt> instruction in the
1936dynamic call chain, undefined behavior results.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1937</div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1938
1939<!-- _______________________________________________________________________ -->
1940
1941<div class="doc_subsubsection"> <a name="i_unreachable">'<tt>unreachable</tt>'
1942Instruction</a> </div>
1943
1944<div class="doc_text">
1945
1946<h5>Syntax:</h5>
1947<pre>
1948 unreachable
1949</pre>
1950
1951<h5>Overview:</h5>
1952
1953<p>The '<tt>unreachable</tt>' instruction has no defined semantics. This
1954instruction is used to inform the optimizer that a particular portion of the
1955code is not reachable. This can be used to indicate that the code after a
1956no-return function cannot be reached, and other facts.</p>
1957
1958<h5>Semantics:</h5>
1959
1960<p>The '<tt>unreachable</tt>' instruction has no defined semantics.</p>
1961</div>
1962
1963
1964
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1965<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1966<div class="doc_subsection"> <a name="binaryops">Binary Operations</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1967<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1968<p>Binary operators are used to do most of the computation in a
1969program. They require two operands, execute an operation on them, and
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1970produce a single value. The operands might represent
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1971multiple 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]1972The result value of a binary operator is not
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1973necessarily the same type as its operands.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1974<p>There are several different binary operators:</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1975</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1976<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1977<div class="doc_subsubsection"> <a name="i_add">'<tt>add</tt>'
1978Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1979<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1980<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1981<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]1982</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1983<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1984<p>The '<tt>add</tt>' instruction returns the sum of its two operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1985<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1986<p>The two arguments to the '<tt>add</tt>' instruction must be either <a
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1987 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a> values.
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1988 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]1989Both arguments must have identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1990<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1991<p>The value produced is the integer or floating point sum of the two
1992operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1993<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1994<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]1995</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1996</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1997<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1998<div class="doc_subsubsection"> <a name="i_sub">'<tt>sub</tt>'
1999Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2000<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2001<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2002<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]2003</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2004<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2005<p>The '<tt>sub</tt>' instruction returns the difference of its two
2006operands.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2007<p>Note that the '<tt>sub</tt>' instruction is used to represent the '<tt>neg</tt>'
2008instruction present in most other intermediate representations.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2009<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2010<p>The two arguments to the '<tt>sub</tt>' instruction must be either <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2011 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]2012values.
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]2013This instruction can also take <a href="#t_vector">vector</a> versions of the values.
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]2014Both arguments must have identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2015<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2016<p>The value produced is the integer or floating point difference of
2017the two operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2018<h5>Example:</h5>
Bill Wendlingaac388b2007-05-29 09:42:13 +0000[diff] [blame]2019<pre>
2020 &lt;result&gt; = sub i32 4, %var <i>; yields {i32}:result = 4 - %var</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2021 &lt;result&gt; = sub i32 0, %val <i>; yields {i32}:result = -%var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2022</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2023</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2024<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2025<div class="doc_subsubsection"> <a name="i_mul">'<tt>mul</tt>'
2026Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2027<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2028<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2029<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]2030</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2031<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2032<p>The '<tt>mul</tt>' instruction returns the product of its two
2033operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2034<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2035<p>The two arguments to the '<tt>mul</tt>' instruction must be either <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2036 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]2037values.
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]2038This instruction can also take <a href="#t_vector">vector</a> versions of the values.
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]2039Both arguments must have identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2040<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2041<p>The value produced is the integer or floating point product of the
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2042two operands.</p>
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]2043<p>Because the operands are the same width, the result of an integer
2044multiplication is the same whether the operands should be deemed unsigned or
2045signed.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2046<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2047<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]2048</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2049</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2050<!-- _______________________________________________________________________ -->
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2051<div class="doc_subsubsection"> <a name="i_udiv">'<tt>udiv</tt>' Instruction
2052</a></div>
2053<div class="doc_text">
2054<h5>Syntax:</h5>
2055<pre> &lt;result&gt; = udiv &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2056</pre>
2057<h5>Overview:</h5>
2058<p>The '<tt>udiv</tt>' instruction returns the quotient of its two
2059operands.</p>
2060<h5>Arguments:</h5>
2061<p>The two arguments to the '<tt>udiv</tt>' instruction must be
2062<a href="#t_integer">integer</a> values. Both arguments must have identical
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]2063types. This instruction can also take <a href="#t_vector">vector</a> versions
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2064of the values in which case the elements must be integers.</p>
2065<h5>Semantics:</h5>
2066<p>The value produced is the unsigned integer quotient of the two operands. This
2067instruction always performs an unsigned division operation, regardless of
2068whether the arguments are unsigned or not.</p>
2069<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2070<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]2071</pre>
2072</div>
2073<!-- _______________________________________________________________________ -->
2074<div class="doc_subsubsection"> <a name="i_sdiv">'<tt>sdiv</tt>' Instruction
2075</a> </div>
2076<div class="doc_text">
2077<h5>Syntax:</h5>
2078<pre> &lt;result&gt; = sdiv &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2079</pre>
2080<h5>Overview:</h5>
2081<p>The '<tt>sdiv</tt>' instruction returns the quotient of its two
2082operands.</p>
2083<h5>Arguments:</h5>
2084<p>The two arguments to the '<tt>sdiv</tt>' instruction must be
2085<a href="#t_integer">integer</a> values. Both arguments must have identical
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]2086types. This instruction can also take <a href="#t_vector">vector</a> versions
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2087of the values in which case the elements must be integers.</p>
2088<h5>Semantics:</h5>
2089<p>The value produced is the signed integer quotient of the two operands. This
2090instruction always performs a signed division operation, regardless of whether
2091the arguments are signed or not.</p>
2092<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2093<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]2094</pre>
2095</div>
2096<!-- _______________________________________________________________________ -->
2097<div class="doc_subsubsection"> <a name="i_fdiv">'<tt>fdiv</tt>'
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2098Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2099<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2100<h5>Syntax:</h5>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2101<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]2102</pre>
2103<h5>Overview:</h5>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2104<p>The '<tt>fdiv</tt>' instruction returns the quotient of its two
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2105operands.</p>
2106<h5>Arguments:</h5>
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]2107<p>The two arguments to the '<tt>fdiv</tt>' instruction must be
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2108<a href="#t_floating">floating point</a> values. Both arguments must have
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]2109identical types. This instruction can also take <a href="#t_vector">vector</a>
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]2110versions of floating point values.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2111<h5>Semantics:</h5>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2112<p>The value produced is the floating point quotient of the two operands.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2113<h5>Example:</h5>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2114<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]2115</pre>
2116</div>
2117<!-- _______________________________________________________________________ -->
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]2118<div class="doc_subsubsection"> <a name="i_urem">'<tt>urem</tt>' Instruction</a>
2119</div>
2120<div class="doc_text">
2121<h5>Syntax:</h5>
2122<pre> &lt;result&gt; = urem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2123</pre>
2124<h5>Overview:</h5>
2125<p>The '<tt>urem</tt>' instruction returns the remainder from the
2126unsigned division of its two arguments.</p>
2127<h5>Arguments:</h5>
2128<p>The two arguments to the '<tt>urem</tt>' instruction must be
2129<a href="#t_integer">integer</a> values. Both arguments must have identical
Dan Gohman80176312007-11-05 23:35:22 +0000[diff] [blame]2130types. This instruction can also take <a href="#t_vector">vector</a> versions
2131of the values in which case the elements must be integers.</p>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]2132<h5>Semantics:</h5>
2133<p>This instruction returns the unsigned integer <i>remainder</i> of a division.
2134This instruction always performs an unsigned division to get the remainder,
2135regardless of whether the arguments are unsigned or not.</p>
2136<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2137<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]2138</pre>
2139
2140</div>
2141<!-- _______________________________________________________________________ -->
2142<div class="doc_subsubsection"> <a name="i_srem">'<tt>srem</tt>'
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2143Instruction</a> </div>
2144<div class="doc_text">
2145<h5>Syntax:</h5>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]2146<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]2147</pre>
2148<h5>Overview:</h5>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]2149<p>The '<tt>srem</tt>' instruction returns the remainder from the
Dan Gohman80176312007-11-05 23:35:22 +0000[diff] [blame]2150signed division of its two operands. This instruction can also take
2151<a href="#t_vector">vector</a> versions of the values in which case
2152the elements must be integers.</p>
2153</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2154<h5>Arguments:</h5>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]2155<p>The two arguments to the '<tt>srem</tt>' instruction must be
2156<a href="#t_integer">integer</a> values. Both arguments must have identical
2157types.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2158<h5>Semantics:</h5>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]2159<p>This instruction returns the <i>remainder</i> of a division (where the result
Reid Spencerc9fdfc82007-03-24 22:23:39 +0000[diff] [blame]2160has the same sign as the dividend, <tt>var1</tt>), not the <i>modulo</i>
2161operator (where the result has the same sign as the divisor, <tt>var2</tt>) of
2162a value. For more information about the difference, see <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2163 href="http://mathforum.org/dr.math/problems/anne.4.28.99.html">The
Reid Spencerc9fdfc82007-03-24 22:23:39 +0000[diff] [blame]2164Math Forum</a>. For a table of how this is implemented in various languages,
Reid Spencer64f5c6c2007-03-24 22:40:44 +0000[diff] [blame]2165please see <a href="http://en.wikipedia.org/wiki/Modulo_operation">
Reid Spencerc9fdfc82007-03-24 22:23:39 +0000[diff] [blame]2166Wikipedia: modulo operation</a>.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2167<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2168<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]2169</pre>
2170
2171</div>
2172<!-- _______________________________________________________________________ -->
2173<div class="doc_subsubsection"> <a name="i_frem">'<tt>frem</tt>'
2174Instruction</a> </div>
2175<div class="doc_text">
2176<h5>Syntax:</h5>
2177<pre> &lt;result&gt; = frem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2178</pre>
2179<h5>Overview:</h5>
2180<p>The '<tt>frem</tt>' instruction returns the remainder from the
2181division of its two operands.</p>
2182<h5>Arguments:</h5>
2183<p>The two arguments to the '<tt>frem</tt>' instruction must be
2184<a href="#t_floating">floating point</a> values. Both arguments must have
Dan Gohman80176312007-11-05 23:35:22 +0000[diff] [blame]2185identical types. This instruction can also take <a href="#t_vector">vector</a>
2186versions of floating point values.</p>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]2187<h5>Semantics:</h5>
2188<p>This instruction returns the <i>remainder</i> of a division.</p>
2189<h5>Example:</h5>
2190<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]2191</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2192</div>
Robert Bocchino7b81c752006-02-17 21:18:08 +0000[diff] [blame]2193
Reid Spencer8e11bf82007-02-02 13:57:07 +0000[diff] [blame]2194<!-- ======================================================================= -->
2195<div class="doc_subsection"> <a name="bitwiseops">Bitwise Binary
2196Operations</a> </div>
2197<div class="doc_text">
2198<p>Bitwise binary operators are used to do various forms of
2199bit-twiddling in a program. They are generally very efficient
2200instructions and can commonly be strength reduced from other
2201instructions. They require two operands, execute an operation on them,
2202and produce a single value. The resulting value of the bitwise binary
2203operators is always the same type as its first operand.</p>
2204</div>
2205
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2206<!-- _______________________________________________________________________ -->
2207<div class="doc_subsubsection"> <a name="i_shl">'<tt>shl</tt>'
2208Instruction</a> </div>
2209<div class="doc_text">
2210<h5>Syntax:</h5>
2211<pre> &lt;result&gt; = shl &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2212</pre>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]2213
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2214<h5>Overview:</h5>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]2215
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2216<p>The '<tt>shl</tt>' instruction returns the first operand shifted to
2217the left a specified number of bits.</p>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]2218
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2219<h5>Arguments:</h5>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]2220
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2221<p>Both arguments to the '<tt>shl</tt>' instruction must be the same <a
2222 href="#t_integer">integer</a> type.</p>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]2223
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2224<h5>Semantics:</h5>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]2225
2226<p>The value produced is <tt>var1</tt> * 2<sup><tt>var2</tt></sup>. If
2227<tt>var2</tt> is (statically or dynamically) equal to or larger than the number
2228of bits in <tt>var1</tt>, the result is undefined.</p>
2229
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2230<h5>Example:</h5><pre>
2231 &lt;result&gt; = shl i32 4, %var <i>; yields {i32}: 4 &lt;&lt; %var</i>
2232 &lt;result&gt; = shl i32 4, 2 <i>; yields {i32}: 16</i>
2233 &lt;result&gt; = shl i32 1, 10 <i>; yields {i32}: 1024</i>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]2234 &lt;result&gt; = shl i32 1, 32 <i>; undefined</i>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2235</pre>
2236</div>
2237<!-- _______________________________________________________________________ -->
2238<div class="doc_subsubsection"> <a name="i_lshr">'<tt>lshr</tt>'
2239Instruction</a> </div>
2240<div class="doc_text">
2241<h5>Syntax:</h5>
2242<pre> &lt;result&gt; = lshr &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2243</pre>
2244
2245<h5>Overview:</h5>
2246<p>The '<tt>lshr</tt>' instruction (logical shift right) returns the first
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]2247operand shifted to the right a specified number of bits with zero fill.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2248
2249<h5>Arguments:</h5>
2250<p>Both arguments to the '<tt>lshr</tt>' instruction must be the same
2251<a href="#t_integer">integer</a> type.</p>
2252
2253<h5>Semantics:</h5>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]2254
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2255<p>This instruction always performs a logical shift right operation. The most
2256significant bits of the result will be filled with zero bits after the
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]2257shift. If <tt>var2</tt> is (statically or dynamically) equal to or larger than
2258the number of bits in <tt>var1</tt>, the result is undefined.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2259
2260<h5>Example:</h5>
2261<pre>
2262 &lt;result&gt; = lshr i32 4, 1 <i>; yields {i32}:result = 2</i>
2263 &lt;result&gt; = lshr i32 4, 2 <i>; yields {i32}:result = 1</i>
2264 &lt;result&gt; = lshr i8 4, 3 <i>; yields {i8}:result = 0</i>
2265 &lt;result&gt; = lshr i8 -2, 1 <i>; yields {i8}:result = 0x7FFFFFFF </i>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]2266 &lt;result&gt; = lshr i32 1, 32 <i>; undefined</i>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2267</pre>
2268</div>
2269
Reid Spencer8e11bf82007-02-02 13:57:07 +0000[diff] [blame]2270<!-- _______________________________________________________________________ -->
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2271<div class="doc_subsubsection"> <a name="i_ashr">'<tt>ashr</tt>'
2272Instruction</a> </div>
2273<div class="doc_text">
2274
2275<h5>Syntax:</h5>
2276<pre> &lt;result&gt; = ashr &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2277</pre>
2278
2279<h5>Overview:</h5>
2280<p>The '<tt>ashr</tt>' instruction (arithmetic shift right) returns the first
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]2281operand shifted to the right a specified number of bits with sign extension.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2282
2283<h5>Arguments:</h5>
2284<p>Both arguments to the '<tt>ashr</tt>' instruction must be the same
2285<a href="#t_integer">integer</a> type.</p>
2286
2287<h5>Semantics:</h5>
2288<p>This instruction always performs an arithmetic shift right operation,
2289The most significant bits of the result will be filled with the sign bit
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]2290of <tt>var1</tt>. If <tt>var2</tt> is (statically or dynamically) equal to or
2291larger than the number of bits in <tt>var1</tt>, the result is undefined.
2292</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2293
2294<h5>Example:</h5>
2295<pre>
2296 &lt;result&gt; = ashr i32 4, 1 <i>; yields {i32}:result = 2</i>
2297 &lt;result&gt; = ashr i32 4, 2 <i>; yields {i32}:result = 1</i>
2298 &lt;result&gt; = ashr i8 4, 3 <i>; yields {i8}:result = 0</i>
2299 &lt;result&gt; = ashr i8 -2, 1 <i>; yields {i8}:result = -1</i>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]2300 &lt;result&gt; = ashr i32 1, 32 <i>; undefined</i>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2301</pre>
2302</div>
2303
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2304<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2305<div class="doc_subsubsection"> <a name="i_and">'<tt>and</tt>'
2306Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2307<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2308<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2309<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]2310</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2311<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2312<p>The '<tt>and</tt>' instruction returns the bitwise logical and of
2313its two operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2314<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2315<p>The two arguments to the '<tt>and</tt>' instruction must be <a
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]2316 href="#t_integer">integer</a> values. Both arguments must have
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2317identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2318<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2319<p>The truth table used for the '<tt>and</tt>' instruction is:</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2320<p> </p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]2321<div style="align: center">
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2322<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2323 <tbody>
2324 <tr>
2325 <td>In0</td>
2326 <td>In1</td>
2327 <td>Out</td>
2328 </tr>
2329 <tr>
2330 <td>0</td>
2331 <td>0</td>
2332 <td>0</td>
2333 </tr>
2334 <tr>
2335 <td>0</td>
2336 <td>1</td>
2337 <td>0</td>
2338 </tr>
2339 <tr>
2340 <td>1</td>
2341 <td>0</td>
2342 <td>0</td>
2343 </tr>
2344 <tr>
2345 <td>1</td>
2346 <td>1</td>
2347 <td>1</td>
2348 </tr>
2349 </tbody>
2350</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]2351</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2352<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2353<pre> &lt;result&gt; = and i32 4, %var <i>; yields {i32}:result = 4 &amp; %var</i>
2354 &lt;result&gt; = and i32 15, 40 <i>; yields {i32}:result = 8</i>
2355 &lt;result&gt; = and i32 4, 8 <i>; yields {i32}:result = 0</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2356</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2357</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2358<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2359<div class="doc_subsubsection"> <a name="i_or">'<tt>or</tt>' Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2360<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2361<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2362<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]2363</pre>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2364<h5>Overview:</h5>
2365<p>The '<tt>or</tt>' instruction returns the bitwise logical inclusive
2366or of its two operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2367<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2368<p>The two arguments to the '<tt>or</tt>' instruction must be <a
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]2369 href="#t_integer">integer</a> values. Both arguments must have
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2370identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2371<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2372<p>The truth table used for the '<tt>or</tt>' instruction is:</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2373<p> </p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]2374<div style="align: center">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2375<table border="1" cellspacing="0" cellpadding="4">
2376 <tbody>
2377 <tr>
2378 <td>In0</td>
2379 <td>In1</td>
2380 <td>Out</td>
2381 </tr>
2382 <tr>
2383 <td>0</td>
2384 <td>0</td>
2385 <td>0</td>
2386 </tr>
2387 <tr>
2388 <td>0</td>
2389 <td>1</td>
2390 <td>1</td>
2391 </tr>
2392 <tr>
2393 <td>1</td>
2394 <td>0</td>
2395 <td>1</td>
2396 </tr>
2397 <tr>
2398 <td>1</td>
2399 <td>1</td>
2400 <td>1</td>
2401 </tr>
2402 </tbody>
2403</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]2404</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2405<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2406<pre> &lt;result&gt; = or i32 4, %var <i>; yields {i32}:result = 4 | %var</i>
2407 &lt;result&gt; = or i32 15, 40 <i>; yields {i32}:result = 47</i>
2408 &lt;result&gt; = or i32 4, 8 <i>; yields {i32}:result = 12</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2409</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2410</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2411<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2412<div class="doc_subsubsection"> <a name="i_xor">'<tt>xor</tt>'
2413Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2414<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2415<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2416<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]2417</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2418<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2419<p>The '<tt>xor</tt>' instruction returns the bitwise logical exclusive
2420or of its two operands. The <tt>xor</tt> is used to implement the
2421"one's complement" operation, which is the "~" operator in C.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2422<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2423<p>The two arguments to the '<tt>xor</tt>' instruction must be <a
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]2424 href="#t_integer">integer</a> values. Both arguments must have
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2425identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2426<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2427<p>The truth table used for the '<tt>xor</tt>' instruction is:</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2428<p> </p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]2429<div style="align: center">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2430<table border="1" cellspacing="0" cellpadding="4">
2431 <tbody>
2432 <tr>
2433 <td>In0</td>
2434 <td>In1</td>
2435 <td>Out</td>
2436 </tr>
2437 <tr>
2438 <td>0</td>
2439 <td>0</td>
2440 <td>0</td>
2441 </tr>
2442 <tr>
2443 <td>0</td>
2444 <td>1</td>
2445 <td>1</td>
2446 </tr>
2447 <tr>
2448 <td>1</td>
2449 <td>0</td>
2450 <td>1</td>
2451 </tr>
2452 <tr>
2453 <td>1</td>
2454 <td>1</td>
2455 <td>0</td>
2456 </tr>
2457 </tbody>
2458</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]2459</div>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2460<p> </p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2461<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2462<pre> &lt;result&gt; = xor i32 4, %var <i>; yields {i32}:result = 4 ^ %var</i>
2463 &lt;result&gt; = xor i32 15, 40 <i>; yields {i32}:result = 39</i>
2464 &lt;result&gt; = xor i32 4, 8 <i>; yields {i32}:result = 12</i>
2465 &lt;result&gt; = xor i32 %V, -1 <i>; yields {i32}:result = ~%V</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2466</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2467</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2468
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2469<!-- ======================================================================= -->
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2470<div class="doc_subsection">
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2471 <a name="vectorops">Vector Operations</a>
2472</div>
2473
2474<div class="doc_text">
2475
2476<p>LLVM supports several instructions to represent vector operations in a
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]2477target-independent manner. These instructions cover the element-access and
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2478vector-specific operations needed to process vectors effectively. While LLVM
2479does directly support these vector operations, many sophisticated algorithms
2480will want to use target-specific intrinsics to take full advantage of a specific
2481target.</p>
2482
2483</div>
2484
2485<!-- _______________________________________________________________________ -->
2486<div class="doc_subsubsection">
2487 <a name="i_extractelement">'<tt>extractelement</tt>' Instruction</a>
2488</div>
2489
2490<div class="doc_text">
2491
2492<h5>Syntax:</h5>
2493
2494<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2495 &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]2496</pre>
2497
2498<h5>Overview:</h5>
2499
2500<p>
2501The '<tt>extractelement</tt>' instruction extracts a single scalar
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]2502element from a vector at a specified index.
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2503</p>
2504
2505
2506<h5>Arguments:</h5>
2507
2508<p>
2509The first operand of an '<tt>extractelement</tt>' instruction is a
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]2510value of <a href="#t_vector">vector</a> type. The second operand is
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2511an index indicating the position from which to extract the element.
2512The index may be a variable.</p>
2513
2514<h5>Semantics:</h5>
2515
2516<p>
2517The result is a scalar of the same type as the element type of
2518<tt>val</tt>. Its value is the value at position <tt>idx</tt> of
2519<tt>val</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
2520results are undefined.
2521</p>
2522
2523<h5>Example:</h5>
2524
2525<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2526 %result = extractelement &lt;4 x i32&gt; %vec, i32 0 <i>; yields i32</i>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2527</pre>
2528</div>
2529
2530
2531<!-- _______________________________________________________________________ -->
2532<div class="doc_subsubsection">
2533 <a name="i_insertelement">'<tt>insertelement</tt>' Instruction</a>
2534</div>
2535
2536<div class="doc_text">
2537
2538<h5>Syntax:</h5>
2539
2540<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2541 &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]2542</pre>
2543
2544<h5>Overview:</h5>
2545
2546<p>
2547The '<tt>insertelement</tt>' instruction inserts a scalar
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]2548element into a vector at a specified index.
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2549</p>
2550
2551
2552<h5>Arguments:</h5>
2553
2554<p>
2555The first operand of an '<tt>insertelement</tt>' instruction is a
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]2556value of <a href="#t_vector">vector</a> type. The second operand is a
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2557scalar value whose type must equal the element type of the first
2558operand. The third operand is an index indicating the position at
2559which to insert the value. The index may be a variable.</p>
2560
2561<h5>Semantics:</h5>
2562
2563<p>
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]2564The result is a vector of the same type as <tt>val</tt>. Its
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2565element values are those of <tt>val</tt> except at position
2566<tt>idx</tt>, where it gets the value <tt>elt</tt>. If <tt>idx</tt>
2567exceeds the length of <tt>val</tt>, the results are undefined.
2568</p>
2569
2570<h5>Example:</h5>
2571
2572<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2573 %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]2574</pre>
2575</div>
2576
2577<!-- _______________________________________________________________________ -->
2578<div class="doc_subsubsection">
2579 <a name="i_shufflevector">'<tt>shufflevector</tt>' Instruction</a>
2580</div>
2581
2582<div class="doc_text">
2583
2584<h5>Syntax:</h5>
2585
2586<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2587 &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]2588</pre>
2589
2590<h5>Overview:</h5>
2591
2592<p>
2593The '<tt>shufflevector</tt>' instruction constructs a permutation of elements
2594from two input vectors, returning a vector of the same type.
2595</p>
2596
2597<h5>Arguments:</h5>
2598
2599<p>
2600The first two operands of a '<tt>shufflevector</tt>' instruction are vectors
2601with types that match each other and types that match the result of the
2602instruction. The third argument is a shuffle mask, which has the same number
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2603of elements as the other vector type, but whose element type is always 'i32'.
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2604</p>
2605
2606<p>
2607The shuffle mask operand is required to be a constant vector with either
2608constant integer or undef values.
2609</p>
2610
2611<h5>Semantics:</h5>
2612
2613<p>
2614The elements of the two input vectors are numbered from left to right across
2615both of the vectors. The shuffle mask operand specifies, for each element of
2616the result vector, which element of the two input registers the result element
2617gets. The element selector may be undef (meaning "don't care") and the second
2618operand may be undef if performing a shuffle from only one vector.
2619</p>
2620
2621<h5>Example:</h5>
2622
2623<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2624 %result = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; %v2,
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]2625 &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]2626 %result = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; undef,
2627 &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]2628</pre>
2629</div>
2630
Tanya Lattner09474292006-04-14 19:24:33 +0000[diff] [blame]2631
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2632<!-- ======================================================================= -->
2633<div class="doc_subsection">
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]2634 <a name="memoryops">Memory Access and Addressing Operations</a>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2635</div>
2636
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2637<div class="doc_text">
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2638
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2639<p>A key design point of an SSA-based representation is how it
2640represents memory. In LLVM, no memory locations are in SSA form, which
2641makes things very simple. This section describes how to read, write,
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]2642allocate, and free memory in LLVM.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2643
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2644</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2645
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2646<!-- _______________________________________________________________________ -->
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2647<div class="doc_subsubsection">
2648 <a name="i_malloc">'<tt>malloc</tt>' Instruction</a>
2649</div>
2650
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2651<div class="doc_text">
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2652
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2653<h5>Syntax:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2654
2655<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2656 &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]2657</pre>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2658
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2659<h5>Overview:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2660
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2661<p>The '<tt>malloc</tt>' instruction allocates memory from the system
2662heap and returns a pointer to it.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2663
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2664<h5>Arguments:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2665
2666<p>The '<tt>malloc</tt>' instruction allocates
2667<tt>sizeof(&lt;type&gt;)*NumElements</tt>
John Criswell6e4ca612004-02-24 16:13:56 +0000[diff] [blame]2668bytes of memory from the operating system and returns a pointer of the
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2669appropriate type to the program. If "NumElements" is specified, it is the
2670number of elements allocated. If an alignment is specified, the value result
2671of the allocation is guaranteed to be aligned to at least that boundary. If
2672not specified, or if zero, the target can choose to align the allocation on any
2673convenient boundary.</p>
2674
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2675<p>'<tt>type</tt>' must be a sized type.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2676
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2677<h5>Semantics:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2678
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2679<p>Memory is allocated using the system "<tt>malloc</tt>" function, and
2680a pointer is returned.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2681
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2682<h5>Example:</h5>
2683
2684<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2685 %array = malloc [4 x i8 ] <i>; yields {[%4 x i8]*}:array</i>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2686
Bill Wendlingaac388b2007-05-29 09:42:13 +0000[diff] [blame]2687 %size = <a href="#i_add">add</a> i32 2, 2 <i>; yields {i32}:size = i32 4</i>
2688 %array1 = malloc i8, i32 4 <i>; yields {i8*}:array1</i>
2689 %array2 = malloc [12 x i8], i32 %size <i>; yields {[12 x i8]*}:array2</i>
2690 %array3 = malloc i32, i32 4, align 1024 <i>; yields {i32*}:array3</i>
2691 %array4 = malloc i32, align 1024 <i>; yields {i32*}:array4</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2692</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2693</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2694
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2695<!-- _______________________________________________________________________ -->
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2696<div class="doc_subsubsection">
2697 <a name="i_free">'<tt>free</tt>' Instruction</a>
2698</div>
2699
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2700<div class="doc_text">
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2701
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2702<h5>Syntax:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2703
2704<pre>
2705 free &lt;type&gt; &lt;value&gt; <i>; yields {void}</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2706</pre>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2707
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2708<h5>Overview:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2709
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2710<p>The '<tt>free</tt>' instruction returns memory back to the unused
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]2711memory heap to be reallocated in the future.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2712
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2713<h5>Arguments:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2714
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2715<p>'<tt>value</tt>' shall be a pointer value that points to a value
2716that was allocated with the '<tt><a href="#i_malloc">malloc</a></tt>'
2717instruction.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2718
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2719<h5>Semantics:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2720
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]2721<p>Access to the memory pointed to by the pointer is no longer defined
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2722after this instruction executes.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2723
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2724<h5>Example:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2725
2726<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2727 %array = <a href="#i_malloc">malloc</a> [4 x i8] <i>; yields {[4 x i8]*}:array</i>
2728 free [4 x i8]* %array
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2729</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2730</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2731
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2732<!-- _______________________________________________________________________ -->
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2733<div class="doc_subsubsection">
2734 <a name="i_alloca">'<tt>alloca</tt>' Instruction</a>
2735</div>
2736
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2737<div class="doc_text">
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2738
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2739<h5>Syntax:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2740
2741<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2742 &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]2743</pre>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2744
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2745<h5>Overview:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2746
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]2747<p>The '<tt>alloca</tt>' instruction allocates memory on the stack frame of the
2748currently executing function, to be automatically released when this function
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2749returns to its caller.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2750
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2751<h5>Arguments:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2752
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]2753<p>The '<tt>alloca</tt>' instruction allocates <tt>sizeof(&lt;type&gt;)*NumElements</tt>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2754bytes of memory on the runtime stack, returning a pointer of the
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2755appropriate type to the program. If "NumElements" is specified, it is the
2756number of elements allocated. If an alignment is specified, the value result
2757of the allocation is guaranteed to be aligned to at least that boundary. If
2758not specified, or if zero, the target can choose to align the allocation on any
2759convenient boundary.</p>
2760
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2761<p>'<tt>type</tt>' may be any sized type.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2762
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2763<h5>Semantics:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2764
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]2765<p>Memory is allocated; a pointer is returned. '<tt>alloca</tt>'d
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2766memory is automatically released when the function returns. The '<tt>alloca</tt>'
2767instruction is commonly used to represent automatic variables that must
2768have an address available. When the function returns (either with the <tt><a
John Criswelldae2e932005-05-12 16:55:34 +0000[diff] [blame]2769 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]2770instructions), the memory is reclaimed.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2771
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2772<h5>Example:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2773
2774<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2775 %ptr = alloca i32 <i>; yields {i32*}:ptr</i>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2776 %ptr = alloca i32, i32 4 <i>; yields {i32*}:ptr</i>
2777 %ptr = alloca i32, i32 4, align 1024 <i>; yields {i32*}:ptr</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2778 %ptr = alloca i32, align 1024 <i>; yields {i32*}:ptr</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2779</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2780</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2781
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2782<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2783<div class="doc_subsubsection"> <a name="i_load">'<tt>load</tt>'
2784Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2785<div class="doc_text">
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2786<h5>Syntax:</h5>
Christopher Lamb2330e4d2007-04-21 08:16:25 +0000[diff] [blame]2787<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]2788<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2789<p>The '<tt>load</tt>' instruction is used to read from memory.</p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2790<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2791<p>The argument to the '<tt>load</tt>' instruction specifies the memory
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]2792address from which to load. The pointer must point to a <a
Chris Lattnere53e5082004-06-03 22:57:15 +0000[diff] [blame]2793 href="#t_firstclass">first class</a> type. If the <tt>load</tt> is
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]2794marked as <tt>volatile</tt>, then the optimizer is not allowed to modify
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2795the number or order of execution of this <tt>load</tt> with other
2796volatile <tt>load</tt> and <tt><a href="#i_store">store</a></tt>
2797instructions. </p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2798<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2799<p>The location of memory pointed to is loaded.</p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2800<h5>Examples:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2801<pre> %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2802 <a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2803 href="#i_store">store</a> i32 3, i32* %ptr <i>; yields {void}</i>
2804 %val = load i32* %ptr <i>; yields {i32}:val = i32 3</i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2805</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2806</div>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2807<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2808<div class="doc_subsubsection"> <a name="i_store">'<tt>store</tt>'
2809Instruction</a> </div>
Reid Spencer035ab572006-11-09 21:18:01 +0000[diff] [blame]2810<div class="doc_text">
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2811<h5>Syntax:</h5>
Christopher Lamb2330e4d2007-04-21 08:16:25 +0000[diff] [blame]2812<pre> store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;] <i>; yields {void}</i>
2813 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]2814</pre>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2815<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2816<p>The '<tt>store</tt>' instruction is used to write to memory.</p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2817<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2818<p>There are two arguments to the '<tt>store</tt>' instruction: a value
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]2819to 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]2820operand must be a pointer to the type of the '<tt>&lt;value&gt;</tt>'
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]2821operand. If the <tt>store</tt> is marked as <tt>volatile</tt>, then the
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2822optimizer is not allowed to modify the number or order of execution of
2823this <tt>store</tt> with other volatile <tt>load</tt> and <tt><a
2824 href="#i_store">store</a></tt> instructions.</p>
2825<h5>Semantics:</h5>
2826<p>The contents of memory are updated to contain '<tt>&lt;value&gt;</tt>'
2827at the location specified by the '<tt>&lt;pointer&gt;</tt>' operand.</p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2828<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2829<pre> %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
Bill Wendling8c6c72d2007-10-22 05:10:05 +0000[diff] [blame]2830 store i32 3, i32* %ptr <i>; yields {void}</i>
2831 %val = <a href="#i_load">load</a> i32* %ptr <i>; yields {i32}:val = i32 3</i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2832</pre>
Reid Spencer47ce1792006-11-09 21:15:49 +0000[diff] [blame]2833</div>
2834
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2835<!-- _______________________________________________________________________ -->
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2836<div class="doc_subsubsection">
2837 <a name="i_getelementptr">'<tt>getelementptr</tt>' Instruction</a>
2838</div>
2839
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2840<div class="doc_text">
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2841<h5>Syntax:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2842<pre>
2843 &lt;result&gt; = getelementptr &lt;ty&gt;* &lt;ptrval&gt;{, &lt;ty&gt; &lt;idx&gt;}*
2844</pre>
2845
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2846<h5>Overview:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2847
2848<p>
2849The '<tt>getelementptr</tt>' instruction is used to get the address of a
2850subelement of an aggregate data structure.</p>
2851
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2852<h5>Arguments:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2853
Reid Spencer85f5b5b2006-12-04 21:29:24 +0000[diff] [blame]2854<p>This instruction takes a list of integer operands that indicate what
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2855elements of the aggregate object to index to. The actual types of the arguments
2856provided depend on the type of the first pointer argument. The
2857'<tt>getelementptr</tt>' instruction is used to index down through the type
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]2858levels 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]2859structure, only <tt>i32</tt> integer constants are allowed. When indexing
Reid Spencer85f5b5b2006-12-04 21:29:24 +0000[diff] [blame]2860into an array or pointer, only integers of 32 or 64 bits are allowed, and will
2861be sign extended to 64-bit values.</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2862
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2863<p>For example, let's consider a C code fragment and how it gets
2864compiled to LLVM:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2865
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2866<div class="doc_code">
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2867<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2868struct RT {
2869 char A;
Chris Lattnercabc8462007-05-29 15:43:56 +0000[diff] [blame]2870 int B[10][20];
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2871 char C;
2872};
2873struct ST {
Chris Lattnercabc8462007-05-29 15:43:56 +0000[diff] [blame]2874 int X;
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2875 double Y;
2876 struct RT Z;
2877};
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2878
Chris Lattnercabc8462007-05-29 15:43:56 +0000[diff] [blame]2879int *foo(struct ST *s) {
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2880 return &amp;s[1].Z.B[5][13];
2881}
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2882</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2883</div>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2884
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2885<p>The LLVM code generated by the GCC frontend is:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2886
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2887<div class="doc_code">
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2888<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2889%RT = type { i8 , [10 x [20 x i32]], i8 }
2890%ST = type { i32, double, %RT }
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2891
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2892define i32* %foo(%ST* %s) {
2893entry:
2894 %reg = getelementptr %ST* %s, i32 1, i32 2, i32 1, i32 5, i32 13
2895 ret i32* %reg
2896}
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2897</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2898</div>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2899
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2900<h5>Semantics:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2901
2902<p>The index types specified for the '<tt>getelementptr</tt>' instruction depend
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]2903on the pointer type that is being indexed into. <a href="#t_pointer">Pointer</a>
Reid Spencer85f5b5b2006-12-04 21:29:24 +0000[diff] [blame]2904and <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]2905<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]2906to 64-bits. <a href="#t_struct">Structure</a> types require <tt>i32</tt>
Reid Spencer42ddd842006-12-03 16:53:48 +0000[diff] [blame]2907<b>constants</b>.</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2908
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2909<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]2910type, which is a pointer, yielding a '<tt>%ST</tt>' = '<tt>{ i32, double, %RT
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2911}</tt>' type, a structure. The second index indexes into the third element of
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2912the structure, yielding a '<tt>%RT</tt>' = '<tt>{ i8 , [10 x [20 x i32]],
2913i8 }</tt>' type, another structure. The third index indexes into the second
2914element of the structure, yielding a '<tt>[10 x [20 x i32]]</tt>' type, an
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2915array. The two dimensions of the array are subscripted into, yielding an
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2916'<tt>i32</tt>' type. The '<tt>getelementptr</tt>' instruction returns a pointer
2917to this element, thus computing a value of '<tt>i32*</tt>' type.</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2918
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2919<p>Note that it is perfectly legal to index partially through a
2920structure, returning a pointer to an inner element. Because of this,
2921the LLVM code for the given testcase is equivalent to:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2922
2923<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2924 define i32* %foo(%ST* %s) {
2925 %t1 = getelementptr %ST* %s, i32 1 <i>; yields %ST*:%t1</i>
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]2926 %t2 = getelementptr %ST* %t1, i32 0, i32 2 <i>; yields %RT*:%t2</i>
2927 %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]2928 %t4 = getelementptr [10 x [20 x i32]]* %t3, i32 0, i32 5 <i>; yields [20 x i32]*:%t4</i>
2929 %t5 = getelementptr [20 x i32]* %t4, i32 0, i32 13 <i>; yields i32*:%t5</i>
2930 ret i32* %t5
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2931 }
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]2932</pre>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]2933
2934<p>Note that it is undefined to access an array out of bounds: array and
2935pointer indexes must always be within the defined bounds of the array type.
2936The one exception for this rules is zero length arrays. These arrays are
2937defined to be accessible as variable length arrays, which requires access
2938beyond the zero'th element.</p>
2939
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]2940<p>The getelementptr instruction is often confusing. For some more insight
2941into how it works, see <a href="GetElementPtr.html">the getelementptr
2942FAQ</a>.</p>
2943
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2944<h5>Example:</h5>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]2945
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2946<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2947 <i>; yields [12 x i8]*:aptr</i>
2948 %aptr = getelementptr {i32, [12 x i8]}* %sptr, i64 0, i32 1
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2949</pre>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2950</div>
Reid Spencer47ce1792006-11-09 21:15:49 +0000[diff] [blame]2951
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2952<!-- ======================================================================= -->
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]2953<div class="doc_subsection"> <a name="convertops">Conversion Operations</a>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2954</div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2955<div class="doc_text">
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]2956<p>The instructions in this category are the conversion instructions (casting)
2957which all take a single operand and a type. They perform various bit conversions
2958on the operand.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2959</div>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]2960
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]2961<!-- _______________________________________________________________________ -->
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]2962<div class="doc_subsubsection">
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2963 <a name="i_trunc">'<tt>trunc .. to</tt>' Instruction</a>
2964</div>
2965<div class="doc_text">
2966
2967<h5>Syntax:</h5>
2968<pre>
2969 &lt;result&gt; = trunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2970</pre>
2971
2972<h5>Overview:</h5>
2973<p>
2974The '<tt>trunc</tt>' instruction truncates its operand to the type <tt>ty2</tt>.
2975</p>
2976
2977<h5>Arguments:</h5>
2978<p>
2979The '<tt>trunc</tt>' instruction takes a <tt>value</tt> to trunc, which must
2980be 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]2981and type of the result, which must be an <a href="#t_integer">integer</a>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2982type. The bit size of <tt>value</tt> must be larger than the bit size of
2983<tt>ty2</tt>. Equal sized types are not allowed.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2984
2985<h5>Semantics:</h5>
2986<p>
2987The '<tt>trunc</tt>' instruction truncates the high order bits in <tt>value</tt>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2988and converts the remaining bits to <tt>ty2</tt>. Since the source size must be
2989larger than the destination size, <tt>trunc</tt> cannot be a <i>no-op cast</i>.
2990It will always truncate bits.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2991
2992<h5>Example:</h5>
2993<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2994 %X = trunc i32 257 to i8 <i>; yields i8:1</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]2995 %Y = trunc i32 123 to i1 <i>; yields i1:true</i>
2996 %Y = trunc i32 122 to i1 <i>; yields i1:false</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2997</pre>
2998</div>
2999
3000<!-- _______________________________________________________________________ -->
3001<div class="doc_subsubsection">
3002 <a name="i_zext">'<tt>zext .. to</tt>' Instruction</a>
3003</div>
3004<div class="doc_text">
3005
3006<h5>Syntax:</h5>
3007<pre>
3008 &lt;result&gt; = zext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3009</pre>
3010
3011<h5>Overview:</h5>
3012<p>The '<tt>zext</tt>' instruction zero extends its operand to type
3013<tt>ty2</tt>.</p>
3014
3015
3016<h5>Arguments:</h5>
3017<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]3018<a href="#t_integer">integer</a> type, and a type to cast it to, which must
3019also be of <a href="#t_integer">integer</a> type. The bit size of the
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3020<tt>value</tt> must be smaller than the bit size of the destination type,
3021<tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3022
3023<h5>Semantics:</h5>
3024<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]3025bits until it reaches the size of the destination type, <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3026
Reid Spencerb5929522007-01-12 15:46:11 +0000[diff] [blame]3027<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]3028
3029<h5>Example:</h5>
3030<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3031 %X = zext i32 257 to i64 <i>; yields i64:257</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3032 %Y = zext i1 true to i32 <i>; yields i32:1</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3033</pre>
3034</div>
3035
3036<!-- _______________________________________________________________________ -->
3037<div class="doc_subsubsection">
3038 <a name="i_sext">'<tt>sext .. to</tt>' Instruction</a>
3039</div>
3040<div class="doc_text">
3041
3042<h5>Syntax:</h5>
3043<pre>
3044 &lt;result&gt; = sext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3045</pre>
3046
3047<h5>Overview:</h5>
3048<p>The '<tt>sext</tt>' sign extends <tt>value</tt> to the type <tt>ty2</tt>.</p>
3049
3050<h5>Arguments:</h5>
3051<p>
3052The '<tt>sext</tt>' instruction takes a value to cast, which must be of
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]3053<a href="#t_integer">integer</a> type, and a type to cast it to, which must
3054also be of <a href="#t_integer">integer</a> type. The bit size of the
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3055<tt>value</tt> must be smaller than the bit size of the destination type,
3056<tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3057
3058<h5>Semantics:</h5>
3059<p>
3060The '<tt>sext</tt>' instruction performs a sign extension by copying the sign
3061bit (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]3062the type <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3063
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3064<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]3065
3066<h5>Example:</h5>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3067<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3068 %X = sext i8 -1 to i16 <i>; yields i16 :65535</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3069 %Y = sext i1 true to i32 <i>; yields i32:-1</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3070</pre>
3071</div>
3072
3073<!-- _______________________________________________________________________ -->
3074<div class="doc_subsubsection">
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]3075 <a name="i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a>
3076</div>
3077
3078<div class="doc_text">
3079
3080<h5>Syntax:</h5>
3081
3082<pre>
3083 &lt;result&gt; = fptrunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3084</pre>
3085
3086<h5>Overview:</h5>
3087<p>The '<tt>fptrunc</tt>' instruction truncates <tt>value</tt> to type
3088<tt>ty2</tt>.</p>
3089
3090
3091<h5>Arguments:</h5>
3092<p>The '<tt>fptrunc</tt>' instruction takes a <a href="#t_floating">floating
3093 point</a> value to cast and a <a href="#t_floating">floating point</a> type to
3094cast it to. The size of <tt>value</tt> must be larger than the size of
3095<tt>ty2</tt>. This implies that <tt>fptrunc</tt> cannot be used to make a
3096<i>no-op cast</i>.</p>
3097
3098<h5>Semantics:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3099<p> The '<tt>fptrunc</tt>' instruction truncates a <tt>value</tt> from a larger
3100<a href="#t_floating">floating point</a> type to a smaller
3101<a href="#t_floating">floating point</a> type. If the value cannot fit within
3102the destination type, <tt>ty2</tt>, then the results are undefined.</p>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]3103
3104<h5>Example:</h5>
3105<pre>
3106 %X = fptrunc double 123.0 to float <i>; yields float:123.0</i>
3107 %Y = fptrunc double 1.0E+300 to float <i>; yields undefined</i>
3108</pre>
3109</div>
3110
3111<!-- _______________________________________________________________________ -->
3112<div class="doc_subsubsection">
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3113 <a name="i_fpext">'<tt>fpext .. to</tt>' Instruction</a>
3114</div>
3115<div class="doc_text">
3116
3117<h5>Syntax:</h5>
3118<pre>
3119 &lt;result&gt; = fpext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3120</pre>
3121
3122<h5>Overview:</h5>
3123<p>The '<tt>fpext</tt>' extends a floating point <tt>value</tt> to a larger
3124floating point value.</p>
3125
3126<h5>Arguments:</h5>
3127<p>The '<tt>fpext</tt>' instruction takes a
3128<a href="#t_floating">floating point</a> <tt>value</tt> to cast,
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3129and a <a href="#t_floating">floating point</a> type to cast it to. The source
3130type must be smaller than the destination type.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3131
3132<h5>Semantics:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3133<p>The '<tt>fpext</tt>' instruction extends the <tt>value</tt> from a smaller
Duncan Sands8036ca42007-03-30 12:22:09 +0000[diff] [blame]3134<a href="#t_floating">floating point</a> type to a larger
3135<a href="#t_floating">floating point</a> type. The <tt>fpext</tt> cannot be
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3136used to make a <i>no-op cast</i> because it always changes bits. Use
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]3137<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]3138
3139<h5>Example:</h5>
3140<pre>
3141 %X = fpext float 3.1415 to double <i>; yields double:3.1415</i>
3142 %Y = fpext float 1.0 to float <i>; yields float:1.0 (no-op)</i>
3143</pre>
3144</div>
3145
3146<!-- _______________________________________________________________________ -->
3147<div class="doc_subsubsection">
Reid Spencer24d6da52007-01-21 00:29:26 +0000[diff] [blame]3148 <a name="i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3149</div>
3150<div class="doc_text">
3151
3152<h5>Syntax:</h5>
3153<pre>
Reid Spencer1539a1c2007-07-31 14:40:14 +0000[diff] [blame]3154 &lt;result&gt; = fptoui &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3155</pre>
3156
3157<h5>Overview:</h5>
Reid Spencer1539a1c2007-07-31 14:40:14 +0000[diff] [blame]3158<p>The '<tt>fptoui</tt>' converts a floating point <tt>value</tt> to its
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3159unsigned integer equivalent of type <tt>ty2</tt>.
3160</p>
3161
3162<h5>Arguments:</h5>
Reid Spencer1539a1c2007-07-31 14:40:14 +0000[diff] [blame]3163<p>The '<tt>fptoui</tt>' instruction takes a value to cast, which must be a
Nate Begemanb348d182007-11-17 03:58:34 +0000[diff] [blame]3164scalar or vector <a href="#t_floating">floating point</a> value, and a type
3165to cast it to <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a>
3166type. If <tt>ty</tt> is a vector floating point type, <tt>ty2</tt> must be a
3167vector integer type with the same number of elements as <tt>ty</tt></p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3168
3169<h5>Semantics:</h5>
Reid Spencer1539a1c2007-07-31 14:40:14 +0000[diff] [blame]3170<p> The '<tt>fptoui</tt>' instruction converts its
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3171<a href="#t_floating">floating point</a> operand into the nearest (rounding
3172towards zero) unsigned integer value. If the value cannot fit in <tt>ty2</tt>,
3173the results are undefined.</p>
3174
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3175<h5>Example:</h5>
3176<pre>
Reid Spencer1539a1c2007-07-31 14:40:14 +0000[diff] [blame]3177 %X = fptoui double 123.0 to i32 <i>; yields i32:123</i>
Chris Lattner88519042007-09-22 03:17:52 +0000[diff] [blame]3178 %Y = fptoui float 1.0E+300 to i1 <i>; yields undefined:1</i>
Reid Spencer1539a1c2007-07-31 14:40:14 +0000[diff] [blame]3179 %X = fptoui float 1.04E+17 to i8 <i>; yields undefined:1</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3180</pre>
3181</div>
3182
3183<!-- _______________________________________________________________________ -->
3184<div class="doc_subsubsection">
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3185 <a name="i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3186</div>
3187<div class="doc_text">
3188
3189<h5>Syntax:</h5>
3190<pre>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3191 &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]3192</pre>
3193
3194<h5>Overview:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3195<p>The '<tt>fptosi</tt>' instruction converts
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3196<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]3197</p>
3198
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]3199<h5>Arguments:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3200<p> The '<tt>fptosi</tt>' instruction takes a value to cast, which must be a
Nate Begemanb348d182007-11-17 03:58:34 +0000[diff] [blame]3201scalar or vector <a href="#t_floating">floating point</a> value, and a type
3202to cast it to <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a>
3203type. If <tt>ty</tt> is a vector floating point type, <tt>ty2</tt> must be a
3204vector integer type with the same number of elements as <tt>ty</tt></p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3205
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]3206<h5>Semantics:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3207<p>The '<tt>fptosi</tt>' instruction converts its
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3208<a href="#t_floating">floating point</a> operand into the nearest (rounding
3209towards zero) signed integer value. If the value cannot fit in <tt>ty2</tt>,
3210the results are undefined.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3211
Chris Lattner33ba0d92001-07-09 00:26:23 +0000[diff] [blame]3212<h5>Example:</h5>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3213<pre>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3214 %X = fptosi double -123.0 to i32 <i>; yields i32:-123</i>
Chris Lattner88519042007-09-22 03:17:52 +0000[diff] [blame]3215 %Y = fptosi float 1.0E-247 to i1 <i>; yields undefined:1</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3216 %X = fptosi float 1.04E+17 to i8 <i>; yields undefined:1</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3217</pre>
3218</div>
3219
3220<!-- _______________________________________________________________________ -->
3221<div class="doc_subsubsection">
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3222 <a name="i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3223</div>
3224<div class="doc_text">
3225
3226<h5>Syntax:</h5>
3227<pre>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3228 &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]3229</pre>
3230
3231<h5>Overview:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3232<p>The '<tt>uitofp</tt>' instruction regards <tt>value</tt> as an unsigned
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3233integer and converts that value to the <tt>ty2</tt> type.</p>
3234
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3235<h5>Arguments:</h5>
Nate Begemanb348d182007-11-17 03:58:34 +0000[diff] [blame]3236<p>The '<tt>uitofp</tt>' instruction takes a value to cast, which must be a
3237scalar or vector <a href="#t_integer">integer</a> value, and a type to cast it
3238to <tt>ty2</tt>, which must be an <a href="#t_floating">floating point</a>
3239type. If <tt>ty</tt> is a vector integer type, <tt>ty2</tt> must be a vector
3240floating point type with the same number of elements as <tt>ty</tt></p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3241
3242<h5>Semantics:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3243<p>The '<tt>uitofp</tt>' instruction interprets its operand as an unsigned
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3244integer quantity and converts it to the corresponding floating point value. If
Jeff Cohencb757312007-04-22 14:56:37 +0000[diff] [blame]3245the value cannot fit in the floating point value, the results are undefined.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3246
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3247<h5>Example:</h5>
3248<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3249 %X = uitofp i32 257 to float <i>; yields float:257.0</i>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3250 %Y = uitofp i8 -1 to double <i>; yields double:255.0</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3251</pre>
3252</div>
3253
3254<!-- _______________________________________________________________________ -->
3255<div class="doc_subsubsection">
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3256 <a name="i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3257</div>
3258<div class="doc_text">
3259
3260<h5>Syntax:</h5>
3261<pre>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3262 &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]3263</pre>
3264
3265<h5>Overview:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3266<p>The '<tt>sitofp</tt>' instruction regards <tt>value</tt> as a signed
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3267integer and converts that value to the <tt>ty2</tt> type.</p>
3268
3269<h5>Arguments:</h5>
Nate Begemanb348d182007-11-17 03:58:34 +0000[diff] [blame]3270<p>The '<tt>sitofp</tt>' instruction takes a value to cast, which must be a
3271scalar or vector <a href="#t_integer">integer</a> value, and a type to cast it
3272to <tt>ty2</tt>, which must be an <a href="#t_floating">floating point</a>
3273type. If <tt>ty</tt> is a vector integer type, <tt>ty2</tt> must be a vector
3274floating point type with the same number of elements as <tt>ty</tt></p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3275
3276<h5>Semantics:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3277<p>The '<tt>sitofp</tt>' instruction interprets its operand as a signed
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3278integer quantity and converts it to the corresponding floating point value. If
Jeff Cohencb757312007-04-22 14:56:37 +0000[diff] [blame]3279the value cannot fit in the floating point value, the results are undefined.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3280
3281<h5>Example:</h5>
3282<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3283 %X = sitofp i32 257 to float <i>; yields float:257.0</i>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3284 %Y = sitofp i8 -1 to double <i>; yields double:-1.0</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3285</pre>
3286</div>
3287
3288<!-- _______________________________________________________________________ -->
3289<div class="doc_subsubsection">
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]3290 <a name="i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a>
3291</div>
3292<div class="doc_text">
3293
3294<h5>Syntax:</h5>
3295<pre>
3296 &lt;result&gt; = ptrtoint &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3297</pre>
3298
3299<h5>Overview:</h5>
3300<p>The '<tt>ptrtoint</tt>' instruction converts the pointer <tt>value</tt> to
3301the integer type <tt>ty2</tt>.</p>
3302
3303<h5>Arguments:</h5>
3304<p>The '<tt>ptrtoint</tt>' instruction takes a <tt>value</tt> to cast, which
Duncan Sands8036ca42007-03-30 12:22:09 +0000[diff] [blame]3305must 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]3306<tt>ty2</tt>, which must be an <a href="#t_integer">integer</a> type.
3307
3308<h5>Semantics:</h5>
3309<p>The '<tt>ptrtoint</tt>' instruction converts <tt>value</tt> to integer type
3310<tt>ty2</tt> by interpreting the pointer value as an integer and either
3311truncating or zero extending that value to the size of the integer type. If
3312<tt>value</tt> is smaller than <tt>ty2</tt> then a zero extension is done. If
3313<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]3314are the same size, then nothing is done (<i>no-op cast</i>) other than a type
3315change.</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]3316
3317<h5>Example:</h5>
3318<pre>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3319 %X = ptrtoint i32* %X to i8 <i>; yields truncation on 32-bit architecture</i>
3320 %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]3321</pre>
3322</div>
3323
3324<!-- _______________________________________________________________________ -->
3325<div class="doc_subsubsection">
3326 <a name="i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a>
3327</div>
3328<div class="doc_text">
3329
3330<h5>Syntax:</h5>
3331<pre>
3332 &lt;result&gt; = inttoptr &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3333</pre>
3334
3335<h5>Overview:</h5>
3336<p>The '<tt>inttoptr</tt>' instruction converts an integer <tt>value</tt> to
3337a pointer type, <tt>ty2</tt>.</p>
3338
3339<h5>Arguments:</h5>
Duncan Sands8036ca42007-03-30 12:22:09 +0000[diff] [blame]3340<p>The '<tt>inttoptr</tt>' instruction takes an <a href="#t_integer">integer</a>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]3341value to cast, and a type to cast it to, which must be a
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame]3342<a href="#t_pointer">pointer</a> type.
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]3343
3344<h5>Semantics:</h5>
3345<p>The '<tt>inttoptr</tt>' instruction converts <tt>value</tt> to type
3346<tt>ty2</tt> by applying either a zero extension or a truncation depending on
3347the size of the integer <tt>value</tt>. If <tt>value</tt> is larger than the
3348size of a pointer then a truncation is done. If <tt>value</tt> is smaller than
3349the size of a pointer then a zero extension is done. If they are the same size,
3350nothing is done (<i>no-op cast</i>).</p>
3351
3352<h5>Example:</h5>
3353<pre>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3354 %X = inttoptr i32 255 to i32* <i>; yields zero extension on 64-bit architecture</i>
3355 %X = inttoptr i32 255 to i32* <i>; yields no-op on 32-bit architecture</i>
3356 %Y = inttoptr i64 0 to i32* <i>; yields truncation on 32-bit architecture</i>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]3357</pre>
3358</div>
3359
3360<!-- _______________________________________________________________________ -->
3361<div class="doc_subsubsection">
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]3362 <a name="i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3363</div>
3364<div class="doc_text">
3365
3366<h5>Syntax:</h5>
3367<pre>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]3368 &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]3369</pre>
3370
3371<h5>Overview:</h5>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]3372<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3373<tt>ty2</tt> without changing any bits.</p>
3374
3375<h5>Arguments:</h5>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]3376<p>The '<tt>bitcast</tt>' instruction takes a value to cast, which must be
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3377a first class value, and a type to cast it to, which must also be a <a
3378 href="#t_firstclass">first class</a> type. The bit sizes of <tt>value</tt>
Reid Spencer19b569f2007-01-09 20:08:58 +0000[diff] [blame]3379and the destination type, <tt>ty2</tt>, must be identical. If the source
3380type is a pointer, the destination type must also be a pointer.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3381
3382<h5>Semantics:</h5>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]3383<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]3384<tt>ty2</tt>. It is always a <i>no-op cast</i> because no bits change with
3385this conversion. The conversion is done as if the <tt>value</tt> had been
3386stored to memory and read back as type <tt>ty2</tt>. Pointer types may only be
3387converted to other pointer types with this instruction. To convert pointers to
3388other types, use the <a href="#i_inttoptr">inttoptr</a> or
3389<a href="#i_ptrtoint">ptrtoint</a> instructions first.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3390
3391<h5>Example:</h5>
3392<pre>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3393 %X = bitcast i8 255 to i8 <i>; yields i8 :-1</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3394 %Y = bitcast i32* %x to sint* <i>; yields sint*:%x</i>
3395 %Z = bitcast <2xint> %V to i64; <i>; yields i64: %V</i>
Chris Lattner33ba0d92001-07-09 00:26:23 +0000[diff] [blame]3396</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3397</div>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3398
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]3399<!-- ======================================================================= -->
3400<div class="doc_subsection"> <a name="otherops">Other Operations</a> </div>
3401<div class="doc_text">
3402<p>The instructions in this category are the "miscellaneous"
3403instructions, which defy better classification.</p>
3404</div>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3405
3406<!-- _______________________________________________________________________ -->
3407<div class="doc_subsubsection"><a name="i_icmp">'<tt>icmp</tt>' Instruction</a>
3408</div>
3409<div class="doc_text">
3410<h5>Syntax:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3411<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]3412</pre>
3413<h5>Overview:</h5>
3414<p>The '<tt>icmp</tt>' instruction returns a boolean value based on comparison
3415of its two integer operands.</p>
3416<h5>Arguments:</h5>
3417<p>The '<tt>icmp</tt>' instruction takes three operands. The first operand is
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3418the condition code indicating the kind of comparison to perform. It is not
3419a value, just a keyword. The possible condition code are:
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3420<ol>
3421 <li><tt>eq</tt>: equal</li>
3422 <li><tt>ne</tt>: not equal </li>
3423 <li><tt>ugt</tt>: unsigned greater than</li>
3424 <li><tt>uge</tt>: unsigned greater or equal</li>
3425 <li><tt>ult</tt>: unsigned less than</li>
3426 <li><tt>ule</tt>: unsigned less or equal</li>
3427 <li><tt>sgt</tt>: signed greater than</li>
3428 <li><tt>sge</tt>: signed greater or equal</li>
3429 <li><tt>slt</tt>: signed less than</li>
3430 <li><tt>sle</tt>: signed less or equal</li>
3431</ol>
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]3432<p>The remaining two arguments must be <a href="#t_integer">integer</a> or
Reid Spencer350f8aa2007-01-04 05:19:58 +0000[diff] [blame]3433<a href="#t_pointer">pointer</a> typed. They must also be identical types.</p>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3434<h5>Semantics:</h5>
3435<p>The '<tt>icmp</tt>' compares <tt>var1</tt> and <tt>var2</tt> according to
3436the condition code given as <tt>cond</tt>. The comparison performed always
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3437yields a <a href="#t_primitive">i1</a> result, as follows:
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3438<ol>
3439 <li><tt>eq</tt>: yields <tt>true</tt> if the operands are equal,
3440 <tt>false</tt> otherwise. No sign interpretation is necessary or performed.
3441 </li>
3442 <li><tt>ne</tt>: yields <tt>true</tt> if the operands are unequal,
3443 <tt>false</tt> otherwise. No sign interpretation is necessary or performed.
3444 <li><tt>ugt</tt>: interprets the operands as unsigned values and yields
3445 <tt>true</tt> if <tt>var1</tt> is greater than <tt>var2</tt>.</li>
3446 <li><tt>uge</tt>: interprets the operands as unsigned values and yields
3447 <tt>true</tt> if <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
3448 <li><tt>ult</tt>: interprets the operands as unsigned values and yields
3449 <tt>true</tt> if <tt>var1</tt> is less than <tt>var2</tt>.</li>
3450 <li><tt>ule</tt>: interprets the operands as unsigned values and yields
3451 <tt>true</tt> if <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
3452 <li><tt>sgt</tt>: interprets the operands as signed values and yields
3453 <tt>true</tt> if <tt>var1</tt> is greater than <tt>var2</tt>.</li>
3454 <li><tt>sge</tt>: interprets the operands as signed values and yields
3455 <tt>true</tt> if <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
3456 <li><tt>slt</tt>: interprets the operands as signed values and yields
3457 <tt>true</tt> if <tt>var1</tt> is less than <tt>var2</tt>.</li>
3458 <li><tt>sle</tt>: interprets the operands as signed values and yields
3459 <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]3460</ol>
3461<p>If the operands are <a href="#t_pointer">pointer</a> typed, the pointer
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3462values are compared as if they were integers.</p>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3463
3464<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3465<pre> &lt;result&gt; = icmp eq i32 4, 5 <i>; yields: result=false</i>
3466 &lt;result&gt; = icmp ne float* %X, %X <i>; yields: result=false</i>
3467 &lt;result&gt; = icmp ult i16 4, 5 <i>; yields: result=true</i>
3468 &lt;result&gt; = icmp sgt i16 4, 5 <i>; yields: result=false</i>
3469 &lt;result&gt; = icmp ule i16 -4, 5 <i>; yields: result=false</i>
3470 &lt;result&gt; = icmp sge i16 4, 5 <i>; yields: result=false</i>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3471</pre>
3472</div>
3473
3474<!-- _______________________________________________________________________ -->
3475<div class="doc_subsubsection"><a name="i_fcmp">'<tt>fcmp</tt>' Instruction</a>
3476</div>
3477<div class="doc_text">
3478<h5>Syntax:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3479<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]3480</pre>
3481<h5>Overview:</h5>
3482<p>The '<tt>fcmp</tt>' instruction returns a boolean value based on comparison
3483of its floating point operands.</p>
3484<h5>Arguments:</h5>
3485<p>The '<tt>fcmp</tt>' instruction takes three operands. The first operand is
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3486the condition code indicating the kind of comparison to perform. It is not
3487a value, just a keyword. The possible condition code are:
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3488<ol>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3489 <li><tt>false</tt>: no comparison, always returns false</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3490 <li><tt>oeq</tt>: ordered and equal</li>
3491 <li><tt>ogt</tt>: ordered and greater than </li>
3492 <li><tt>oge</tt>: ordered and greater than or equal</li>
3493 <li><tt>olt</tt>: ordered and less than </li>
3494 <li><tt>ole</tt>: ordered and less than or equal</li>
3495 <li><tt>one</tt>: ordered and not equal</li>
3496 <li><tt>ord</tt>: ordered (no nans)</li>
3497 <li><tt>ueq</tt>: unordered or equal</li>
3498 <li><tt>ugt</tt>: unordered or greater than </li>
3499 <li><tt>uge</tt>: unordered or greater than or equal</li>
3500 <li><tt>ult</tt>: unordered or less than </li>
3501 <li><tt>ule</tt>: unordered or less than or equal</li>
3502 <li><tt>une</tt>: unordered or not equal</li>
3503 <li><tt>uno</tt>: unordered (either nans)</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3504 <li><tt>true</tt>: no comparison, always returns true</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3505</ol>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3506<p><i>Ordered</i> means that neither operand is a QNAN while
Reid Spencer93a49852006-12-06 07:08:07 +0000[diff] [blame]3507<i>unordered</i> means that either operand may be a QNAN.</p>
Reid Spencer350f8aa2007-01-04 05:19:58 +0000[diff] [blame]3508<p>The <tt>val1</tt> and <tt>val2</tt> arguments must be
3509<a href="#t_floating">floating point</a> typed. They must have identical
3510types.</p>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3511<h5>Semantics:</h5>
3512<p>The '<tt>fcmp</tt>' compares <tt>var1</tt> and <tt>var2</tt> according to
3513the condition code given as <tt>cond</tt>. The comparison performed always
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3514yields a <a href="#t_primitive">i1</a> result, as follows:
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3515<ol>
3516 <li><tt>false</tt>: always yields <tt>false</tt>, regardless of operands.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3517 <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]3518 <tt>var1</tt> is equal to <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3519 <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]3520 <tt>var1</tt> is greather than <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3521 <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]3522 <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3523 <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]3524 <tt>var1</tt> is less than <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3525 <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]3526 <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3527 <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]3528 <tt>var1</tt> is not equal to <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3529 <li><tt>ord</tt>: yields <tt>true</tt> if both operands are not a QNAN.</li>
3530 <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]3531 <tt>var1</tt> is equal to <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3532 <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]3533 <tt>var1</tt> is greater than <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3534 <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]3535 <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3536 <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]3537 <tt>var1</tt> is less than <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3538 <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]3539 <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3540 <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]3541 <tt>var1</tt> is not equal to <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3542 <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]3543 <li><tt>true</tt>: always yields <tt>true</tt>, regardless of operands.</li>
3544</ol>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3545
3546<h5>Example:</h5>
3547<pre> &lt;result&gt; = fcmp oeq float 4.0, 5.0 <i>; yields: result=false</i>
3548 &lt;result&gt; = icmp one float 4.0, 5.0 <i>; yields: result=true</i>
3549 &lt;result&gt; = icmp olt float 4.0, 5.0 <i>; yields: result=true</i>
3550 &lt;result&gt; = icmp ueq double 1.0, 2.0 <i>; yields: result=false</i>
3551</pre>
3552</div>
3553
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]3554<!-- _______________________________________________________________________ -->
3555<div class="doc_subsubsection"> <a name="i_phi">'<tt>phi</tt>'
3556Instruction</a> </div>
3557<div class="doc_text">
3558<h5>Syntax:</h5>
3559<pre> &lt;result&gt; = phi &lt;ty&gt; [ &lt;val0&gt;, &lt;label0&gt;], ...<br></pre>
3560<h5>Overview:</h5>
3561<p>The '<tt>phi</tt>' instruction is used to implement the &#966; node in
3562the SSA graph representing the function.</p>
3563<h5>Arguments:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3564<p>The type of the incoming values is specified with the first type
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]3565field. After this, the '<tt>phi</tt>' instruction takes a list of pairs
3566as arguments, with one pair for each predecessor basic block of the
3567current block. Only values of <a href="#t_firstclass">first class</a>
3568type may be used as the value arguments to the PHI node. Only labels
3569may be used as the label arguments.</p>
3570<p>There must be no non-phi instructions between the start of a basic
3571block and the PHI instructions: i.e. PHI instructions must be first in
3572a basic block.</p>
3573<h5>Semantics:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3574<p>At runtime, the '<tt>phi</tt>' instruction logically takes on the value
3575specified by the pair corresponding to the predecessor basic block that executed
3576just prior to the current block.</p>
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]3577<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3578<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]3579</div>
3580
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3581<!-- _______________________________________________________________________ -->
3582<div class="doc_subsubsection">
3583 <a name="i_select">'<tt>select</tt>' Instruction</a>
3584</div>
3585
3586<div class="doc_text">
3587
3588<h5>Syntax:</h5>
3589
3590<pre>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3591 &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]3592</pre>
3593
3594<h5>Overview:</h5>
3595
3596<p>
3597The '<tt>select</tt>' instruction is used to choose one value based on a
3598condition, without branching.
3599</p>
3600
3601
3602<h5>Arguments:</h5>
3603
3604<p>
3605The '<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.
3606</p>
3607
3608<h5>Semantics:</h5>
3609
3610<p>
3611If the boolean condition evaluates to true, the instruction returns the first
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3612value argument; otherwise, it returns the second value argument.
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3613</p>
3614
3615<h5>Example:</h5>
3616
3617<pre>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3618 %X = select i1 true, i8 17, i8 42 <i>; yields i8:17</i>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3619</pre>
3620</div>
3621
Robert Bocchino05ccd702006-01-15 20:48:27 +0000[diff] [blame]3622
3623<!-- _______________________________________________________________________ -->
3624<div class="doc_subsubsection">
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3625 <a name="i_call">'<tt>call</tt>' Instruction</a>
3626</div>
3627
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3628<div class="doc_text">
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3629
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3630<h5>Syntax:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3631<pre>
Nick Lewyckydb7e3c92007-09-08 13:57:50 +0000[diff] [blame]3632 &lt;result&gt; = [tail] call [<a href="#callingconv">cconv</a>] &lt;ty&gt; [&lt;fnty&gt;*] &lt;fnptrval&gt;(&lt;param list&gt;)
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3633</pre>
3634
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3635<h5>Overview:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3636
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3637<p>The '<tt>call</tt>' instruction represents a simple function call.</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3638
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3639<h5>Arguments:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3640
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3641<p>This instruction requires several arguments:</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3642
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]3643<ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3644 <li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3645 <p>The optional "tail" marker indicates whether the callee function accesses
3646 any allocas or varargs in the caller. If the "tail" marker is present, the
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3647 function call is eligible for tail call optimization. Note that calls may
3648 be marked "tail" even if they do not occur before a <a
3649 href="#i_ret"><tt>ret</tt></a> instruction.
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3650 </li>
3651 <li>
Duncan Sands8036ca42007-03-30 12:22:09 +0000[diff] [blame]3652 <p>The optional "cconv" marker indicates which <a href="#callingconv">calling
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3653 convention</a> the call should use. If none is specified, the call defaults
3654 to using C calling conventions.
3655 </li>
3656 <li>
Nick Lewyckydb7e3c92007-09-08 13:57:50 +0000[diff] [blame]3657 <p>'<tt>ty</tt>': the type of the call instruction itself which is also
3658 the type of the return value. Functions that return no value are marked
3659 <tt><a href="#t_void">void</a></tt>.</p>
3660 </li>
3661 <li>
3662 <p>'<tt>fnty</tt>': shall be the signature of the pointer to function
3663 value being invoked. The argument types must match the types implied by
3664 this signature. This type can be omitted if the function is not varargs
3665 and if the function type does not return a pointer to a function.</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3666 </li>
3667 <li>
3668 <p>'<tt>fnptrval</tt>': An LLVM value containing a pointer to a function to
3669 be invoked. In most cases, this is a direct function invocation, but
3670 indirect <tt>call</tt>s are just as possible, calling an arbitrary pointer
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3671 to function value.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3672 </li>
3673 <li>
3674 <p>'<tt>function args</tt>': argument list whose types match the
Reid Spencera7e302a2005-05-01 22:22:57 +0000[diff] [blame]3675 function signature argument types. All arguments must be of
3676 <a href="#t_firstclass">first class</a> type. If the function signature
3677 indicates the function accepts a variable number of arguments, the extra
3678 arguments can be specified.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3679 </li>
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]3680</ol>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3681
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3682<h5>Semantics:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3683
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3684<p>The '<tt>call</tt>' instruction is used to cause control flow to
3685transfer to a specified function, with its incoming arguments bound to
3686the specified values. Upon a '<tt><a href="#i_ret">ret</a></tt>'
3687instruction in the called function, control flow continues with the
3688instruction after the function call, and the return value of the
3689function is bound to the result argument. This is a simpler case of
3690the <a href="#i_invoke">invoke</a> instruction.</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3691
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3692<h5>Example:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3693
3694<pre>
Nick Lewyckydb7e3c92007-09-08 13:57:50 +0000[diff] [blame]3695 %retval = call i32 @test(i32 %argc)
3696 call i32 (i8 *, ...)* @printf(i8 * %msg, i32 12, i8 42);
3697 %X = tail call i32 @foo()
3698 %Y = tail call <a href="#callingconv">fastcc</a> i32 @foo()
3699 %Z = call void %foo(i8 97 signext)
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3700</pre>
3701
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3702</div>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3703
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3704<!-- _______________________________________________________________________ -->
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3705<div class="doc_subsubsection">
Chris Lattnerfb6977d2006-01-13 23:26:01 +0000[diff] [blame]3706 <a name="i_va_arg">'<tt>va_arg</tt>' Instruction</a>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3707</div>
3708
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3709<div class="doc_text">
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3710
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]3711<h5>Syntax:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3712
3713<pre>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3714 &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]3715</pre>
3716
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]3717<h5>Overview:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3718
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3719<p>The '<tt>va_arg</tt>' instruction is used to access arguments passed through
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3720the "variable argument" area of a function call. It is used to implement the
3721<tt>va_arg</tt> macro in C.</p>
3722
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]3723<h5>Arguments:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3724
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3725<p>This instruction takes a <tt>va_list*</tt> value and the type of
3726the argument. It returns a value of the specified argument type and
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3727increments the <tt>va_list</tt> to point to the next argument. The
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3728actual type of <tt>va_list</tt> is target specific.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3729
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]3730<h5>Semantics:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3731
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3732<p>The '<tt>va_arg</tt>' instruction loads an argument of the specified
3733type from the specified <tt>va_list</tt> and causes the
3734<tt>va_list</tt> to point to the next argument. For more information,
3735see the variable argument handling <a href="#int_varargs">Intrinsic
3736Functions</a>.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3737
3738<p>It is legal for this instruction to be called in a function which does not
3739take a variable number of arguments, for example, the <tt>vfprintf</tt>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3740function.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3741
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3742<p><tt>va_arg</tt> is an LLVM instruction instead of an <a
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3743href="#intrinsics">intrinsic function</a> because it takes a type as an
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3744argument.</p>
3745
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]3746<h5>Example:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3747
3748<p>See the <a href="#int_varargs">variable argument processing</a> section.</p>
3749
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3750</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3751
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3752<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3753<div class="doc_section"> <a name="intrinsics">Intrinsic Functions</a> </div>
3754<!-- *********************************************************************** -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3755
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3756<div class="doc_text">
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3757
3758<p>LLVM supports the notion of an "intrinsic function". These functions have
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]3759well known names and semantics and are required to follow certain restrictions.
3760Overall, these intrinsics represent an extension mechanism for the LLVM
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3761language that does not require changing all of the transformations in LLVM when
Gabor Greif04367bf2007-07-06 22:07:22 +0000[diff] [blame]3762adding to the language (or the bitcode reader/writer, the parser, etc...).</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3763
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3764<p>Intrinsic function names must all start with an "<tt>llvm.</tt>" prefix. This
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3765prefix is reserved in LLVM for intrinsic names; thus, function names may not
3766begin with this prefix. Intrinsic functions must always be external functions:
3767you cannot define the body of intrinsic functions. Intrinsic functions may
3768only be used in call or invoke instructions: it is illegal to take the address
3769of an intrinsic function. Additionally, because intrinsic functions are part
3770of the LLVM language, it is required if any are added that they be documented
3771here.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3772
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]3773<p>Some intrinsic functions can be overloaded, i.e., the intrinsic represents
3774a family of functions that perform the same operation but on different data
3775types. Because LLVM can represent over 8 million different integer types,
3776overloading is used commonly to allow an intrinsic function to operate on any
3777integer type. One or more of the argument types or the result type can be
3778overloaded to accept any integer type. Argument types may also be defined as
3779exactly matching a previous argument's type or the result type. This allows an
3780intrinsic function which accepts multiple arguments, but needs all of them to
3781be of the same type, to only be overloaded with respect to a single argument or
3782the result.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3783
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]3784<p>Overloaded intrinsics will have the names of its overloaded argument types
3785encoded into its function name, each preceded by a period. Only those types
3786which are overloaded result in a name suffix. Arguments whose type is matched
3787against another type do not. For example, the <tt>llvm.ctpop</tt> function can
3788take an integer of any width and returns an integer of exactly the same integer
3789width. This leads to a family of functions such as
3790<tt>i8 @llvm.ctpop.i8(i8 %val)</tt> and <tt>i29 @llvm.ctpop.i29(i29 %val)</tt>.
3791Only one type, the return type, is overloaded, and only one type suffix is
3792required. Because the argument's type is matched against the return type, it
3793does not require its own name suffix.</p>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]3794
3795<p>To learn how to add an intrinsic function, please see the
3796<a href="ExtendingLLVM.html">Extending LLVM Guide</a>.
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3797</p>
3798
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3799</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3800
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3801<!-- ======================================================================= -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3802<div class="doc_subsection">
3803 <a name="int_varargs">Variable Argument Handling Intrinsics</a>
3804</div>
3805
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3806<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3807
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3808<p>Variable argument support is defined in LLVM with the <a
Chris Lattnerfb6977d2006-01-13 23:26:01 +0000[diff] [blame]3809 href="#i_va_arg"><tt>va_arg</tt></a> instruction and these three
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3810intrinsic functions. These functions are related to the similarly
3811named macros defined in the <tt>&lt;stdarg.h&gt;</tt> header file.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3812
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3813<p>All of these functions operate on arguments that use a
3814target-specific value type "<tt>va_list</tt>". The LLVM assembly
3815language reference manual does not define what this type is, so all
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3816transformations should be prepared to handle these functions regardless of
3817the type used.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3818
Chris Lattner374ab302006-05-15 17:26:46 +0000[diff] [blame]3819<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]3820instruction and the variable argument handling intrinsic functions are
3821used.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3822
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]3823<div class="doc_code">
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3824<pre>
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]3825define i32 @test(i32 %X, ...) {
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3826 ; Initialize variable argument processing
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3827 %ap = alloca i8*
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]3828 %ap2 = bitcast i8** %ap to i8*
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]3829 call void @llvm.va_start(i8* %ap2)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3830
3831 ; Read a single integer argument
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3832 %tmp = va_arg i8** %ap, i32
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3833
3834 ; Demonstrate usage of llvm.va_copy and llvm.va_end
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3835 %aq = alloca i8*
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]3836 %aq2 = bitcast i8** %aq to i8*
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3837 call void @llvm.va_copy(i8* %aq2, i8* %ap2)
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]3838 call void @llvm.va_end(i8* %aq2)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3839
3840 ; Stop processing of arguments.
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]3841 call void @llvm.va_end(i8* %ap2)
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3842 ret i32 %tmp
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3843}
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]3844
3845declare void @llvm.va_start(i8*)
3846declare void @llvm.va_copy(i8*, i8*)
3847declare void @llvm.va_end(i8*)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3848</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3849</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3850
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]3851</div>
3852
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3853<!-- _______________________________________________________________________ -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3854<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]3855 <a name="int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3856</div>
3857
3858
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3859<div class="doc_text">
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3860<h5>Syntax:</h5>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]3861<pre> declare void %llvm.va_start(i8* &lt;arglist&gt;)<br></pre>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3862<h5>Overview:</h5>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3863<P>The '<tt>llvm.va_start</tt>' intrinsic initializes
3864<tt>*&lt;arglist&gt;</tt> for subsequent use by <tt><a
3865href="#i_va_arg">va_arg</a></tt>.</p>
3866
3867<h5>Arguments:</h5>
3868
3869<P>The argument is a pointer to a <tt>va_list</tt> element to initialize.</p>
3870
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3871<h5>Semantics:</h5>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3872
3873<P>The '<tt>llvm.va_start</tt>' intrinsic works just like the <tt>va_start</tt>
3874macro available in C. In a target-dependent way, it initializes the
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3875<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]3876<tt>va_arg</tt> will produce the first variable argument passed to the function.
3877Unlike 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]3878last argument of the function as the compiler can figure that out.</p>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3879
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3880</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3881
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3882<!-- _______________________________________________________________________ -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3883<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]3884 <a name="int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3885</div>
3886
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3887<div class="doc_text">
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3888<h5>Syntax:</h5>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]3889<pre> declare void @llvm.va_end(i8* &lt;arglist&gt;)<br></pre>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3890<h5>Overview:</h5>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]3891
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3892<p>The '<tt>llvm.va_end</tt>' intrinsic destroys <tt>*&lt;arglist&gt;</tt>,
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]3893which 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]3894or <tt><a href="#i_va_copy">llvm.va_copy</a></tt>.</p>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]3895
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3896<h5>Arguments:</h5>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]3897
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3898<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]3899
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3900<h5>Semantics:</h5>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]3901
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3902<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]3903macro available in C. In a target-dependent way, it destroys the
3904<tt>va_list</tt> element to which the argument points. Calls to <a
3905href="#int_va_start"><tt>llvm.va_start</tt></a> and <a href="#int_va_copy">
3906<tt>llvm.va_copy</tt></a> must be matched exactly with calls to
3907<tt>llvm.va_end</tt>.</p>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]3908
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3909</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3910
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3911<!-- _______________________________________________________________________ -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3912<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]3913 <a name="int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3914</div>
3915
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3916<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3917
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3918<h5>Syntax:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3919
3920<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]3921 declare void @llvm.va_copy(i8* &lt;destarglist&gt;, i8* &lt;srcarglist&gt;)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3922</pre>
3923
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3924<h5>Overview:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3925
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3926<p>The '<tt>llvm.va_copy</tt>' intrinsic copies the current argument position
3927from the source argument list to the destination argument list.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3928
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3929<h5>Arguments:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3930
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3931<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]3932The 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]3933
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3934
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3935<h5>Semantics:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3936
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]3937<p>The '<tt>llvm.va_copy</tt>' intrinsic works just like the <tt>va_copy</tt>
3938macro available in C. In a target-dependent way, it copies the source
3939<tt>va_list</tt> element into the destination <tt>va_list</tt> element. This
3940intrinsic is necessary because the <tt><a href="#int_va_start">
3941llvm.va_start</a></tt> intrinsic may be arbitrarily complex and require, for
3942example, memory allocation.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3943
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3944</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3945
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3946<!-- ======================================================================= -->
3947<div class="doc_subsection">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3948 <a name="int_gc">Accurate Garbage Collection Intrinsics</a>
3949</div>
3950
3951<div class="doc_text">
3952
3953<p>
3954LLVM support for <a href="GarbageCollection.html">Accurate Garbage
3955Collection</a> requires the implementation and generation of these intrinsics.
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]3956These intrinsics allow identification of <a href="#int_gcroot">GC roots on the
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3957stack</a>, as well as garbage collector implementations that require <a
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]3958href="#int_gcread">read</a> and <a href="#int_gcwrite">write</a> barriers.
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3959Front-ends for type-safe garbage collected languages should generate these
3960intrinsics to make use of the LLVM garbage collectors. For more details, see <a
3961href="GarbageCollection.html">Accurate Garbage Collection with LLVM</a>.
3962</p>
3963</div>
3964
3965<!-- _______________________________________________________________________ -->
3966<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]3967 <a name="int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3968</div>
3969
3970<div class="doc_text">
3971
3972<h5>Syntax:</h5>
3973
3974<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]3975 declare void @llvm.gcroot(i8** %ptrloc, i8* %metadata)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3976</pre>
3977
3978<h5>Overview:</h5>
3979
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]3980<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]3981the code generator, and allows some metadata to be associated with it.</p>
3982
3983<h5>Arguments:</h5>
3984
3985<p>The first argument specifies the address of a stack object that contains the
3986root pointer. The second pointer (which must be either a constant or a global
3987value address) contains the meta-data to be associated with the root.</p>
3988
3989<h5>Semantics:</h5>
3990
3991<p>At runtime, a call to this intrinsics stores a null pointer into the "ptrloc"
3992location. At compile-time, the code generator generates information to allow
3993the runtime to find the pointer at GC safe points.
3994</p>
3995
3996</div>
3997
3998
3999<!-- _______________________________________________________________________ -->
4000<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4001 <a name="int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4002</div>
4003
4004<div class="doc_text">
4005
4006<h5>Syntax:</h5>
4007
4008<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]4009 declare i8* @llvm.gcread(i8* %ObjPtr, i8** %Ptr)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4010</pre>
4011
4012<h5>Overview:</h5>
4013
4014<p>The '<tt>llvm.gcread</tt>' intrinsic identifies reads of references from heap
4015locations, allowing garbage collector implementations that require read
4016barriers.</p>
4017
4018<h5>Arguments:</h5>
4019
Chris Lattner80626e92006-03-14 20:02:51 +0000[diff] [blame]4020<p>The second argument is the address to read from, which should be an address
4021allocated from the garbage collector. The first object is a pointer to the
4022start of the referenced object, if needed by the language runtime (otherwise
4023null).</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4024
4025<h5>Semantics:</h5>
4026
4027<p>The '<tt>llvm.gcread</tt>' intrinsic has the same semantics as a load
4028instruction, but may be replaced with substantially more complex code by the
4029garbage collector runtime, as needed.</p>
4030
4031</div>
4032
4033
4034<!-- _______________________________________________________________________ -->
4035<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4036 <a name="int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4037</div>
4038
4039<div class="doc_text">
4040
4041<h5>Syntax:</h5>
4042
4043<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]4044 declare void @llvm.gcwrite(i8* %P1, i8* %Obj, i8** %P2)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4045</pre>
4046
4047<h5>Overview:</h5>
4048
4049<p>The '<tt>llvm.gcwrite</tt>' intrinsic identifies writes of references to heap
4050locations, allowing garbage collector implementations that require write
4051barriers (such as generational or reference counting collectors).</p>
4052
4053<h5>Arguments:</h5>
4054
Chris Lattner80626e92006-03-14 20:02:51 +0000[diff] [blame]4055<p>The first argument is the reference to store, the second is the start of the
4056object to store it to, and the third is the address of the field of Obj to
4057store to. If the runtime does not require a pointer to the object, Obj may be
4058null.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4059
4060<h5>Semantics:</h5>
4061
4062<p>The '<tt>llvm.gcwrite</tt>' intrinsic has the same semantics as a store
4063instruction, but may be replaced with substantially more complex code by the
4064garbage collector runtime, as needed.</p>
4065
4066</div>
4067
4068
4069
4070<!-- ======================================================================= -->
4071<div class="doc_subsection">
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4072 <a name="int_codegen">Code Generator Intrinsics</a>
4073</div>
4074
4075<div class="doc_text">
4076<p>
4077These intrinsics are provided by LLVM to expose special features that may only
4078be implemented with code generator support.
4079</p>
4080
4081</div>
4082
4083<!-- _______________________________________________________________________ -->
4084<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4085 <a name="int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4086</div>
4087
4088<div class="doc_text">
4089
4090<h5>Syntax:</h5>
4091<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4092 declare i8 *@llvm.returnaddress(i32 &lt;level&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4093</pre>
4094
4095<h5>Overview:</h5>
4096
4097<p>
Chris Lattner32b5d712006-10-15 20:05:59 +0000[diff] [blame]4098The '<tt>llvm.returnaddress</tt>' intrinsic attempts to compute a
4099target-specific value indicating the return address of the current function
4100or one of its callers.
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4101</p>
4102
4103<h5>Arguments:</h5>
4104
4105<p>
4106The argument to this intrinsic indicates which function to return the address
4107for. Zero indicates the calling function, one indicates its caller, etc. The
4108argument is <b>required</b> to be a constant integer value.
4109</p>
4110
4111<h5>Semantics:</h5>
4112
4113<p>
4114The '<tt>llvm.returnaddress</tt>' intrinsic either returns a pointer indicating
4115the return address of the specified call frame, or zero if it cannot be
4116identified. The value returned by this intrinsic is likely to be incorrect or 0
4117for arguments other than zero, so it should only be used for debugging purposes.
4118</p>
4119
4120<p>
4121Note that calling this intrinsic does not prevent function inlining or other
Chris Lattnerb40bb382005-03-07 20:30:51 +0000[diff] [blame]4122aggressive transformations, so the value returned may not be that of the obvious
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4123source-language caller.
4124</p>
4125</div>
4126
4127
4128<!-- _______________________________________________________________________ -->
4129<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4130 <a name="int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4131</div>
4132
4133<div class="doc_text">
4134
4135<h5>Syntax:</h5>
4136<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]4137 declare i8 *@llvm.frameaddress(i32 &lt;level&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4138</pre>
4139
4140<h5>Overview:</h5>
4141
4142<p>
Chris Lattner32b5d712006-10-15 20:05:59 +0000[diff] [blame]4143The '<tt>llvm.frameaddress</tt>' intrinsic attempts to return the
4144target-specific frame pointer value for the specified stack frame.
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4145</p>
4146
4147<h5>Arguments:</h5>
4148
4149<p>
4150The argument to this intrinsic indicates which function to return the frame
4151pointer for. Zero indicates the calling function, one indicates its caller,
4152etc. The argument is <b>required</b> to be a constant integer value.
4153</p>
4154
4155<h5>Semantics:</h5>
4156
4157<p>
4158The '<tt>llvm.frameaddress</tt>' intrinsic either returns a pointer indicating
4159the frame address of the specified call frame, or zero if it cannot be
4160identified. The value returned by this intrinsic is likely to be incorrect or 0
4161for arguments other than zero, so it should only be used for debugging purposes.
4162</p>
4163
4164<p>
4165Note that calling this intrinsic does not prevent function inlining or other
Chris Lattnerb40bb382005-03-07 20:30:51 +0000[diff] [blame]4166aggressive transformations, so the value returned may not be that of the obvious
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4167source-language caller.
4168</p>
4169</div>
4170
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]4171<!-- _______________________________________________________________________ -->
4172<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4173 <a name="int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]4174</div>
4175
4176<div class="doc_text">
4177
4178<h5>Syntax:</h5>
4179<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]4180 declare i8 *@llvm.stacksave()
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]4181</pre>
4182
4183<h5>Overview:</h5>
4184
4185<p>
4186The '<tt>llvm.stacksave</tt>' intrinsic is used to remember the current state of
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4187the function stack, for use with <a href="#int_stackrestore">
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]4188<tt>llvm.stackrestore</tt></a>. This is useful for implementing language
4189features like scoped automatic variable sized arrays in C99.
4190</p>
4191
4192<h5>Semantics:</h5>
4193
4194<p>
4195This intrinsic returns a opaque pointer value that can be passed to <a
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4196href="#int_stackrestore"><tt>llvm.stackrestore</tt></a>. When an
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]4197<tt>llvm.stackrestore</tt> intrinsic is executed with a value saved from
4198<tt>llvm.stacksave</tt>, it effectively restores the state of the stack to the
4199state it was in when the <tt>llvm.stacksave</tt> intrinsic executed. In
4200practice, this pops any <a href="#i_alloca">alloca</a> blocks from the stack
4201that were allocated after the <tt>llvm.stacksave</tt> was executed.
4202</p>
4203
4204</div>
4205
4206<!-- _______________________________________________________________________ -->
4207<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4208 <a name="int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]4209</div>
4210
4211<div class="doc_text">
4212
4213<h5>Syntax:</h5>
4214<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4215 declare void @llvm.stackrestore(i8 * %ptr)
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]4216</pre>
4217
4218<h5>Overview:</h5>
4219
4220<p>
4221The '<tt>llvm.stackrestore</tt>' intrinsic is used to restore the state of
4222the function stack to the state it was in when the corresponding <a
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4223href="#int_stacksave"><tt>llvm.stacksave</tt></a> intrinsic executed. This is
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]4224useful for implementing language features like scoped automatic variable sized
4225arrays in C99.
4226</p>
4227
4228<h5>Semantics:</h5>
4229
4230<p>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4231See the description for <a href="#int_stacksave"><tt>llvm.stacksave</tt></a>.
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]4232</p>
4233
4234</div>
4235
4236
4237<!-- _______________________________________________________________________ -->
4238<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4239 <a name="int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a>
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]4240</div>
4241
4242<div class="doc_text">
4243
4244<h5>Syntax:</h5>
4245<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]4246 declare void @llvm.prefetch(i8* &lt;address&gt;, i32 &lt;rw&gt;, i32 &lt;locality&gt;)
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]4247</pre>
4248
4249<h5>Overview:</h5>
4250
4251
4252<p>
4253The '<tt>llvm.prefetch</tt>' intrinsic is a hint to the code generator to insert
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]4254a prefetch instruction if supported; otherwise, it is a noop. Prefetches have
4255no
4256effect on the behavior of the program but can change its performance
Chris Lattner2a615362005-02-28 19:47:14 +0000[diff] [blame]4257characteristics.
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]4258</p>
4259
4260<h5>Arguments:</h5>
4261
4262<p>
4263<tt>address</tt> is the address to be prefetched, <tt>rw</tt> is the specifier
4264determining if the fetch should be for a read (0) or write (1), and
4265<tt>locality</tt> is a temporal locality specifier ranging from (0) - no
Chris Lattneraeffb4a2005-03-07 20:31:38 +0000[diff] [blame]4266locality, to (3) - extremely local keep in cache. The <tt>rw</tt> and
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]4267<tt>locality</tt> arguments must be constant integers.
4268</p>
4269
4270<h5>Semantics:</h5>
4271
4272<p>
4273This intrinsic does not modify the behavior of the program. In particular,
4274prefetches cannot trap and do not produce a value. On targets that support this
4275intrinsic, the prefetch can provide hints to the processor cache for better
4276performance.
4277</p>
4278
4279</div>
4280
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]4281<!-- _______________________________________________________________________ -->
4282<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4283 <a name="int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a>
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]4284</div>
4285
4286<div class="doc_text">
4287
4288<h5>Syntax:</h5>
4289<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]4290 declare void @llvm.pcmarker(i32 &lt;id&gt;)
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]4291</pre>
4292
4293<h5>Overview:</h5>
4294
4295
4296<p>
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]4297The '<tt>llvm.pcmarker</tt>' intrinsic is a method to export a Program Counter
4298(PC) in a region of
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]4299code to simulators and other tools. The method is target specific, but it is
4300expected that the marker will use exported symbols to transmit the PC of the marker.
Jeff Cohen25d4f7e2005-11-11 02:15:27 +0000[diff] [blame]4301The marker makes no guarantees that it will remain with any specific instruction
Chris Lattnerd07c3f42005-11-15 06:07:55 +0000[diff] [blame]4302after optimizations. It is possible that the presence of a marker will inhibit
Chris Lattnerb3e7afd2006-03-24 07:16:10 +0000[diff] [blame]4303optimizations. The intended use is to be inserted after optimizations to allow
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]4304correlations of simulation runs.
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]4305</p>
4306
4307<h5>Arguments:</h5>
4308
4309<p>
4310<tt>id</tt> is a numerical id identifying the marker.
4311</p>
4312
4313<h5>Semantics:</h5>
4314
4315<p>
4316This intrinsic does not modify the behavior of the program. Backends that do not
4317support this intrinisic may ignore it.
4318</p>
4319
4320</div>
4321
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]4322<!-- _______________________________________________________________________ -->
4323<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4324 <a name="int_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a>
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]4325</div>
4326
4327<div class="doc_text">
4328
4329<h5>Syntax:</h5>
4330<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4331 declare i64 @llvm.readcyclecounter( )
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]4332</pre>
4333
4334<h5>Overview:</h5>
4335
4336
4337<p>
4338The '<tt>llvm.readcyclecounter</tt>' intrinsic provides access to the cycle
4339counter register (or similar low latency, high accuracy clocks) on those targets
4340that support it. On X86, it should map to RDTSC. On Alpha, it should map to RPCC.
4341As the backing counters overflow quickly (on the order of 9 seconds on alpha), this
4342should only be used for small timings.
4343</p>
4344
4345<h5>Semantics:</h5>
4346
4347<p>
4348When directly supported, reading the cycle counter should not modify any memory.
4349Implementations are allowed to either return a application specific value or a
4350system wide value. On backends without support, this is lowered to a constant 0.
4351</p>
4352
4353</div>
4354
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4355<!-- ======================================================================= -->
4356<div class="doc_subsection">
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4357 <a name="int_libc">Standard C Library Intrinsics</a>
4358</div>
4359
4360<div class="doc_text">
4361<p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4362LLVM provides intrinsics for a few important standard C library functions.
4363These intrinsics allow source-language front-ends to pass information about the
4364alignment of the pointer arguments to the code generator, providing opportunity
4365for more efficient code generation.
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4366</p>
4367
4368</div>
4369
4370<!-- _______________________________________________________________________ -->
4371<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4372 <a name="int_memcpy">'<tt>llvm.memcpy</tt>' Intrinsic</a>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4373</div>
4374
4375<div class="doc_text">
4376
4377<h5>Syntax:</h5>
4378<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4379 declare void @llvm.memcpy.i32(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4380 i32 &lt;len&gt;, i32 &lt;align&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4381 declare void @llvm.memcpy.i64(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4382 i64 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4383</pre>
4384
4385<h5>Overview:</h5>
4386
4387<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4388The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4389location to the destination location.
4390</p>
4391
4392<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4393Note that, unlike the standard libc function, the <tt>llvm.memcpy.*</tt>
4394intrinsics do not return a value, and takes an extra alignment argument.
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4395</p>
4396
4397<h5>Arguments:</h5>
4398
4399<p>
4400The first argument is a pointer to the destination, the second is a pointer to
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4401the source. The third argument is an integer argument
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4402specifying the number of bytes to copy, and the fourth argument is the alignment
4403of the source and destination locations.
4404</p>
4405
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]4406<p>
4407If 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]4408the caller guarantees that both the source and destination pointers are aligned
4409to that boundary.
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]4410</p>
4411
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4412<h5>Semantics:</h5>
4413
4414<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4415The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4416location to the destination location, which are not allowed to overlap. It
4417copies "len" bytes of memory over. If the argument is known to be aligned to
4418some boundary, this can be specified as the fourth argument, otherwise it should
4419be set to 0 or 1.
4420</p>
4421</div>
4422
4423
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]4424<!-- _______________________________________________________________________ -->
4425<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4426 <a name="int_memmove">'<tt>llvm.memmove</tt>' Intrinsic</a>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]4427</div>
4428
4429<div class="doc_text">
4430
4431<h5>Syntax:</h5>
4432<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4433 declare void @llvm.memmove.i32(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4434 i32 &lt;len&gt;, i32 &lt;align&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4435 declare void @llvm.memmove.i64(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4436 i64 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]4437</pre>
4438
4439<h5>Overview:</h5>
4440
4441<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4442The '<tt>llvm.memmove.*</tt>' intrinsics move a block of memory from the source
4443location to the destination location. It is similar to the
4444'<tt>llvm.memcmp</tt>' intrinsic but allows the two memory locations to overlap.
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]4445</p>
4446
4447<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4448Note that, unlike the standard libc function, the <tt>llvm.memmove.*</tt>
4449intrinsics do not return a value, and takes an extra alignment argument.
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]4450</p>
4451
4452<h5>Arguments:</h5>
4453
4454<p>
4455The first argument is a pointer to the destination, the second is a pointer to
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4456the source. The third argument is an integer argument
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]4457specifying the number of bytes to copy, and the fourth argument is the alignment
4458of the source and destination locations.
4459</p>
4460
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]4461<p>
4462If 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]4463the caller guarantees that the source and destination pointers are aligned to
4464that boundary.
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]4465</p>
4466
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]4467<h5>Semantics:</h5>
4468
4469<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4470The '<tt>llvm.memmove.*</tt>' intrinsics copy a block of memory from the source
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]4471location to the destination location, which may overlap. It
4472copies "len" bytes of memory over. If the argument is known to be aligned to
4473some boundary, this can be specified as the fourth argument, otherwise it should
4474be set to 0 or 1.
4475</p>
4476</div>
4477
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]4478
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4479<!-- _______________________________________________________________________ -->
4480<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4481 <a name="int_memset">'<tt>llvm.memset.*</tt>' Intrinsics</a>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4482</div>
4483
4484<div class="doc_text">
4485
4486<h5>Syntax:</h5>
4487<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4488 declare void @llvm.memset.i32(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4489 i32 &lt;len&gt;, i32 &lt;align&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4490 declare void @llvm.memset.i64(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4491 i64 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4492</pre>
4493
4494<h5>Overview:</h5>
4495
4496<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4497The '<tt>llvm.memset.*</tt>' intrinsics fill a block of memory with a particular
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4498byte value.
4499</p>
4500
4501<p>
4502Note that, unlike the standard libc function, the <tt>llvm.memset</tt> intrinsic
4503does not return a value, and takes an extra alignment argument.
4504</p>
4505
4506<h5>Arguments:</h5>
4507
4508<p>
4509The first argument is a pointer to the destination to fill, the second is the
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4510byte value to fill it with, the third argument is an integer
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4511argument specifying the number of bytes to fill, and the fourth argument is the
4512known alignment of destination location.
4513</p>
4514
4515<p>
4516If 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]4517the caller guarantees that the destination pointer is aligned to that boundary.
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4518</p>
4519
4520<h5>Semantics:</h5>
4521
4522<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4523The '<tt>llvm.memset.*</tt>' intrinsics fill "len" bytes of memory starting at
4524the
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4525destination location. If the argument is known to be aligned to some boundary,
4526this can be specified as the fourth argument, otherwise it should be set to 0 or
45271.
4528</p>
4529</div>
4530
4531
Chris Lattner32006282004-06-11 02:28:03 +0000[diff] [blame]4532<!-- _______________________________________________________________________ -->
4533<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4534 <a name="int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a>
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]4535</div>
4536
4537<div class="doc_text">
4538
4539<h5>Syntax:</h5>
Dale Johannesen408f9c12007-10-02 17:47:38 +0000[diff] [blame]4540<p>This is an overloaded intrinsic. You can use <tt>llvm.sqrt</tt> on any
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]4541floating point or vector of floating point type. Not all targets support all
4542types however.
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]4543<pre>
Dale Johannesen408f9c12007-10-02 17:47:38 +0000[diff] [blame]4544 declare float @llvm.sqrt.f32(float %Val)
4545 declare double @llvm.sqrt.f64(double %Val)
4546 declare x86_fp80 @llvm.sqrt.f80(x86_fp80 %Val)
4547 declare fp128 @llvm.sqrt.f128(fp128 %Val)
4548 declare ppc_fp128 @llvm.sqrt.ppcf128(ppc_fp128 %Val)
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]4549</pre>
4550
4551<h5>Overview:</h5>
4552
4553<p>
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]4554The '<tt>llvm.sqrt</tt>' intrinsics return the sqrt of the specified operand,
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]4555returning the same value as the libm '<tt>sqrt</tt>' functions would. Unlike
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]4556<tt>sqrt</tt> in libm, however, <tt>llvm.sqrt</tt> has undefined behavior for
4557negative numbers (which allows for better optimization).
4558</p>
4559
4560<h5>Arguments:</h5>
4561
4562<p>
4563The argument and return value are floating point numbers of the same type.
4564</p>
4565
4566<h5>Semantics:</h5>
4567
4568<p>
Dan Gohmand6257fe2007-07-16 14:37:41 +0000[diff] [blame]4569This function returns the sqrt of the specified operand if it is a nonnegative
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]4570floating point number.
4571</p>
4572</div>
4573
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]4574<!-- _______________________________________________________________________ -->
4575<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4576 <a name="int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]4577</div>
4578
4579<div class="doc_text">
4580
4581<h5>Syntax:</h5>
Dale Johannesen408f9c12007-10-02 17:47:38 +0000[diff] [blame]4582<p>This is an overloaded intrinsic. You can use <tt>llvm.powi</tt> on any
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]4583floating point or vector of floating point type. Not all targets support all
4584types however.
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]4585<pre>
Dale Johannesen408f9c12007-10-02 17:47:38 +0000[diff] [blame]4586 declare float @llvm.powi.f32(float %Val, i32 %power)
4587 declare double @llvm.powi.f64(double %Val, i32 %power)
4588 declare x86_fp80 @llvm.powi.f80(x86_fp80 %Val, i32 %power)
4589 declare fp128 @llvm.powi.f128(fp128 %Val, i32 %power)
4590 declare ppc_fp128 @llvm.powi.ppcf128(ppc_fp128 %Val, i32 %power)
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]4591</pre>
4592
4593<h5>Overview:</h5>
4594
4595<p>
4596The '<tt>llvm.powi.*</tt>' intrinsics return the first operand raised to the
4597specified (positive or negative) power. The order of evaluation of
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]4598multiplications is not defined. When a vector of floating point type is
4599used, the second argument remains a scalar integer value.
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]4600</p>
4601
4602<h5>Arguments:</h5>
4603
4604<p>
4605The second argument is an integer power, and the first is a value to raise to
4606that power.
4607</p>
4608
4609<h5>Semantics:</h5>
4610
4611<p>
4612This function returns the first value raised to the second power with an
4613unspecified sequence of rounding operations.</p>
4614</div>
4615
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]4616<!-- _______________________________________________________________________ -->
4617<div class="doc_subsubsection">
4618 <a name="int_sin">'<tt>llvm.sin.*</tt>' Intrinsic</a>
4619</div>
4620
4621<div class="doc_text">
4622
4623<h5>Syntax:</h5>
4624<p>This is an overloaded intrinsic. You can use <tt>llvm.sin</tt> on any
4625floating point or vector of floating point type. Not all targets support all
4626types however.
4627<pre>
4628 declare float @llvm.sin.f32(float %Val)
4629 declare double @llvm.sin.f64(double %Val)
4630 declare x86_fp80 @llvm.sin.f80(x86_fp80 %Val)
4631 declare fp128 @llvm.sin.f128(fp128 %Val)
4632 declare ppc_fp128 @llvm.sin.ppcf128(ppc_fp128 %Val)
4633</pre>
4634
4635<h5>Overview:</h5>
4636
4637<p>
4638The '<tt>llvm.sin.*</tt>' intrinsics return the sine of the operand.
4639</p>
4640
4641<h5>Arguments:</h5>
4642
4643<p>
4644The argument and return value are floating point numbers of the same type.
4645</p>
4646
4647<h5>Semantics:</h5>
4648
4649<p>
4650This function returns the sine of the specified operand, returning the
4651same values as the libm <tt>sin</tt> functions would, and handles error
Dan Gohmanba83b7e2007-10-17 18:05:13 +0000[diff] [blame]4652conditions in the same way.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]4653</div>
4654
4655<!-- _______________________________________________________________________ -->
4656<div class="doc_subsubsection">
4657 <a name="int_cos">'<tt>llvm.cos.*</tt>' Intrinsic</a>
4658</div>
4659
4660<div class="doc_text">
4661
4662<h5>Syntax:</h5>
4663<p>This is an overloaded intrinsic. You can use <tt>llvm.cos</tt> on any
4664floating point or vector of floating point type. Not all targets support all
4665types however.
4666<pre>
4667 declare float @llvm.cos.f32(float %Val)
4668 declare double @llvm.cos.f64(double %Val)
4669 declare x86_fp80 @llvm.cos.f80(x86_fp80 %Val)
4670 declare fp128 @llvm.cos.f128(fp128 %Val)
4671 declare ppc_fp128 @llvm.cos.ppcf128(ppc_fp128 %Val)
4672</pre>
4673
4674<h5>Overview:</h5>
4675
4676<p>
4677The '<tt>llvm.cos.*</tt>' intrinsics return the cosine of the operand.
4678</p>
4679
4680<h5>Arguments:</h5>
4681
4682<p>
4683The argument and return value are floating point numbers of the same type.
4684</p>
4685
4686<h5>Semantics:</h5>
4687
4688<p>
4689This function returns the cosine of the specified operand, returning the
4690same values as the libm <tt>cos</tt> functions would, and handles error
Dan Gohmanba83b7e2007-10-17 18:05:13 +0000[diff] [blame]4691conditions in the same way.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]4692</div>
4693
4694<!-- _______________________________________________________________________ -->
4695<div class="doc_subsubsection">
4696 <a name="int_pow">'<tt>llvm.pow.*</tt>' Intrinsic</a>
4697</div>
4698
4699<div class="doc_text">
4700
4701<h5>Syntax:</h5>
4702<p>This is an overloaded intrinsic. You can use <tt>llvm.pow</tt> on any
4703floating point or vector of floating point type. Not all targets support all
4704types however.
4705<pre>
4706 declare float @llvm.pow.f32(float %Val, float %Power)
4707 declare double @llvm.pow.f64(double %Val, double %Power)
4708 declare x86_fp80 @llvm.pow.f80(x86_fp80 %Val, x86_fp80 %Power)
4709 declare fp128 @llvm.pow.f128(fp128 %Val, fp128 %Power)
4710 declare ppc_fp128 @llvm.pow.ppcf128(ppc_fp128 %Val, ppc_fp128 Power)
4711</pre>
4712
4713<h5>Overview:</h5>
4714
4715<p>
4716The '<tt>llvm.pow.*</tt>' intrinsics return the first operand raised to the
4717specified (positive or negative) power.
4718</p>
4719
4720<h5>Arguments:</h5>
4721
4722<p>
4723The second argument is a floating point power, and the first is a value to
4724raise to that power.
4725</p>
4726
4727<h5>Semantics:</h5>
4728
4729<p>
4730This function returns the first value raised to the second power,
4731returning the
4732same values as the libm <tt>pow</tt> functions would, and handles error
Dan Gohmanba83b7e2007-10-17 18:05:13 +0000[diff] [blame]4733conditions in the same way.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]4734</div>
4735
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]4736
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4737<!-- ======================================================================= -->
4738<div class="doc_subsection">
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]4739 <a name="int_manip">Bit Manipulation Intrinsics</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4740</div>
4741
4742<div class="doc_text">
4743<p>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]4744LLVM provides intrinsics for a few important bit manipulation operations.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4745These allow efficient code generation for some algorithms.
4746</p>
4747
4748</div>
4749
4750<!-- _______________________________________________________________________ -->
4751<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4752 <a name="int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]4753</div>
4754
4755<div class="doc_text">
4756
4757<h5>Syntax:</h5>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]4758<p>This is an overloaded intrinsic function. You can use bswap on any integer
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]4759type that is an even number of bytes (i.e. BitWidth % 16 == 0).
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]4760<pre>
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]4761 declare i16 @llvm.bswap.i16(i16 &lt;id&gt;)
4762 declare i32 @llvm.bswap.i32(i32 &lt;id&gt;)
4763 declare i64 @llvm.bswap.i64(i64 &lt;id&gt;)
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]4764</pre>
4765
4766<h5>Overview:</h5>
4767
4768<p>
Reid Spencer338ea092007-04-02 02:25:19 +0000[diff] [blame]4769The '<tt>llvm.bswap</tt>' family of intrinsics is used to byte swap integer
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]4770values with an even number of bytes (positive multiple of 16 bits). These are
4771useful for performing operations on data that is not in the target's native
4772byte order.
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]4773</p>
4774
4775<h5>Semantics:</h5>
4776
4777<p>
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]4778The <tt>llvm.bswap.i16</tt> intrinsic returns an i16 value that has the high
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4779and low byte of the input i16 swapped. Similarly, the <tt>llvm.bswap.i32</tt>
4780intrinsic returns an i32 value that has the four bytes of the input i32
4781swapped, so that if the input bytes are numbered 0, 1, 2, 3 then the returned
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]4782i32 will have its bytes in 3, 2, 1, 0 order. The <tt>llvm.bswap.i48</tt>,
4783<tt>llvm.bswap.i64</tt> and other intrinsics extend this concept to
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]4784additional even-byte lengths (6 bytes, 8 bytes and more, respectively).
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]4785</p>
4786
4787</div>
4788
4789<!-- _______________________________________________________________________ -->
4790<div class="doc_subsubsection">
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]4791 <a name="int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4792</div>
4793
4794<div class="doc_text">
4795
4796<h5>Syntax:</h5>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]4797<p>This is an overloaded intrinsic. You can use llvm.ctpop on any integer bit
4798width. Not all targets support all bit widths however.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4799<pre>
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]4800 declare i8 @llvm.ctpop.i8 (i8 &lt;src&gt;)
4801 declare i16 @llvm.ctpop.i16(i16 &lt;src&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4802 declare i32 @llvm.ctpop.i32(i32 &lt;src&gt;)
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]4803 declare i64 @llvm.ctpop.i64(i64 &lt;src&gt;)
4804 declare i256 @llvm.ctpop.i256(i256 &lt;src&gt;)
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4805</pre>
4806
4807<h5>Overview:</h5>
4808
4809<p>
Chris Lattnerec6cb612006-01-16 22:38:59 +0000[diff] [blame]4810The '<tt>llvm.ctpop</tt>' family of intrinsics counts the number of bits set in a
4811value.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4812</p>
4813
4814<h5>Arguments:</h5>
4815
4816<p>
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]4817The only argument is the value to be counted. The argument may be of any
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]4818integer type. The return type must match the argument type.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4819</p>
4820
4821<h5>Semantics:</h5>
4822
4823<p>
4824The '<tt>llvm.ctpop</tt>' intrinsic counts the 1's in a variable.
4825</p>
4826</div>
4827
4828<!-- _______________________________________________________________________ -->
4829<div class="doc_subsubsection">
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]4830 <a name="int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4831</div>
4832
4833<div class="doc_text">
4834
4835<h5>Syntax:</h5>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]4836<p>This is an overloaded intrinsic. You can use <tt>llvm.ctlz</tt> on any
4837integer bit width. Not all targets support all bit widths however.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4838<pre>
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]4839 declare i8 @llvm.ctlz.i8 (i8 &lt;src&gt;)
4840 declare i16 @llvm.ctlz.i16(i16 &lt;src&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4841 declare i32 @llvm.ctlz.i32(i32 &lt;src&gt;)
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]4842 declare i64 @llvm.ctlz.i64(i64 &lt;src&gt;)
4843 declare i256 @llvm.ctlz.i256(i256 &lt;src&gt;)
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4844</pre>
4845
4846<h5>Overview:</h5>
4847
4848<p>
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]4849The '<tt>llvm.ctlz</tt>' family of intrinsic functions counts the number of
4850leading zeros in a variable.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4851</p>
4852
4853<h5>Arguments:</h5>
4854
4855<p>
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]4856The only argument is the value to be counted. The argument may be of any
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]4857integer type. The return type must match the argument type.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4858</p>
4859
4860<h5>Semantics:</h5>
4861
4862<p>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4863The '<tt>llvm.ctlz</tt>' intrinsic counts the leading (most significant) zeros
4864in 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]4865of src. For example, <tt>llvm.ctlz(i32 2) = 30</tt>.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4866</p>
4867</div>
Chris Lattner32006282004-06-11 02:28:03 +0000[diff] [blame]4868
4869
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4870
4871<!-- _______________________________________________________________________ -->
4872<div class="doc_subsubsection">
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]4873 <a name="int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic</a>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4874</div>
4875
4876<div class="doc_text">
4877
4878<h5>Syntax:</h5>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]4879<p>This is an overloaded intrinsic. You can use <tt>llvm.cttz</tt> on any
4880integer bit width. Not all targets support all bit widths however.
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4881<pre>
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]4882 declare i8 @llvm.cttz.i8 (i8 &lt;src&gt;)
4883 declare i16 @llvm.cttz.i16(i16 &lt;src&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4884 declare i32 @llvm.cttz.i32(i32 &lt;src&gt;)
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]4885 declare i64 @llvm.cttz.i64(i64 &lt;src&gt;)
4886 declare i256 @llvm.cttz.i256(i256 &lt;src&gt;)
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4887</pre>
4888
4889<h5>Overview:</h5>
4890
4891<p>
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]4892The '<tt>llvm.cttz</tt>' family of intrinsic functions counts the number of
4893trailing zeros.
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4894</p>
4895
4896<h5>Arguments:</h5>
4897
4898<p>
4899The only argument is the value to be counted. The argument may be of any
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]4900integer type. The return type must match the argument type.
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4901</p>
4902
4903<h5>Semantics:</h5>
4904
4905<p>
4906The '<tt>llvm.cttz</tt>' intrinsic counts the trailing (least significant) zeros
4907in a variable. If the src == 0 then the result is the size in bits of the type
4908of src. For example, <tt>llvm.cttz(2) = 1</tt>.
4909</p>
4910</div>
4911
Reid Spencer497d93e2007-04-01 08:27:01 +0000[diff] [blame]4912<!-- _______________________________________________________________________ -->
4913<div class="doc_subsubsection">
Reid Spencerbeacf662007-04-10 02:51:31 +0000[diff] [blame]4914 <a name="int_part_select">'<tt>llvm.part.select.*</tt>' Intrinsic</a>
Reid Spencera13ba7d2007-04-01 19:00:37 +0000[diff] [blame]4915</div>
4916
4917<div class="doc_text">
4918
4919<h5>Syntax:</h5>
Reid Spencerbeacf662007-04-10 02:51:31 +0000[diff] [blame]4920<p>This is an overloaded intrinsic. You can use <tt>llvm.part.select</tt>
Reid Spencera13ba7d2007-04-01 19:00:37 +0000[diff] [blame]4921on any integer bit width.
4922<pre>
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]4923 declare i17 @llvm.part.select.i17 (i17 %val, i32 %loBit, i32 %hiBit)
4924 declare i29 @llvm.part.select.i29 (i29 %val, i32 %loBit, i32 %hiBit)
Reid Spencera13ba7d2007-04-01 19:00:37 +0000[diff] [blame]4925</pre>
4926
4927<h5>Overview:</h5>
Reid Spencerbeacf662007-04-10 02:51:31 +0000[diff] [blame]4928<p>The '<tt>llvm.part.select</tt>' family of intrinsic functions selects a
Reid Spencera13ba7d2007-04-01 19:00:37 +0000[diff] [blame]4929range of bits from an integer value and returns them in the same bit width as
4930the original value.</p>
4931
4932<h5>Arguments:</h5>
4933<p>The first argument, <tt>%val</tt> and the result may be integer types of
4934any bit width but they must have the same bit width. The second and third
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4935arguments must be <tt>i32</tt> type since they specify only a bit index.</p>
Reid Spencera13ba7d2007-04-01 19:00:37 +0000[diff] [blame]4936
4937<h5>Semantics:</h5>
Reid Spencerbeacf662007-04-10 02:51:31 +0000[diff] [blame]4938<p>The operation of the '<tt>llvm.part.select</tt>' intrinsic has two modes
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4939of operation: forwards and reverse. If <tt>%loBit</tt> is greater than
4940<tt>%hiBits</tt> then the intrinsic operates in reverse mode. Otherwise it
4941operates in forward mode.</p>
4942<p>In forward mode, this intrinsic is the equivalent of shifting <tt>%val</tt>
4943right by <tt>%loBit</tt> bits and then ANDing it with a mask with
Reid Spencera13ba7d2007-04-01 19:00:37 +0000[diff] [blame]4944only the <tt>%hiBit - %loBit</tt> bits set, as follows:</p>
4945<ol>
4946 <li>The <tt>%val</tt> is shifted right (LSHR) by the number of bits specified
4947 by <tt>%loBits</tt>. This normalizes the value to the low order bits.</li>
4948 <li>The <tt>%loBits</tt> value is subtracted from the <tt>%hiBits</tt> value
4949 to determine the number of bits to retain.</li>
4950 <li>A mask of the retained bits is created by shifting a -1 value.</li>
4951 <li>The mask is ANDed with <tt>%val</tt> to produce the result.
4952</ol>
Reid Spencerd6a85b52007-05-14 16:14:57 +0000[diff] [blame]4953<p>In reverse mode, a similar computation is made except that the bits are
4954returned in the reverse order. So, for example, if <tt>X</tt> has the value
4955<tt>i16 0x0ACF (101011001111)</tt> and we apply
4956<tt>part.select(i16 X, 8, 3)</tt> to it, we get back the value
4957<tt>i16 0x0026 (000000100110)</tt>.</p>
Reid Spencera13ba7d2007-04-01 19:00:37 +0000[diff] [blame]4958</div>
4959
Reid Spencerf86037f2007-04-11 23:23:49 +0000[diff] [blame]4960<div class="doc_subsubsection">
4961 <a name="int_part_set">'<tt>llvm.part.set.*</tt>' Intrinsic</a>
4962</div>
4963
4964<div class="doc_text">
4965
4966<h5>Syntax:</h5>
4967<p>This is an overloaded intrinsic. You can use <tt>llvm.part.set</tt>
4968on any integer bit width.
4969<pre>
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]4970 declare i17 @llvm.part.set.i17.i9 (i17 %val, i9 %repl, i32 %lo, i32 %hi)
4971 declare i29 @llvm.part.set.i29.i9 (i29 %val, i9 %repl, i32 %lo, i32 %hi)
Reid Spencerf86037f2007-04-11 23:23:49 +0000[diff] [blame]4972</pre>
4973
4974<h5>Overview:</h5>
4975<p>The '<tt>llvm.part.set</tt>' family of intrinsic functions replaces a range
4976of bits in an integer value with another integer value. It returns the integer
4977with the replaced bits.</p>
4978
4979<h5>Arguments:</h5>
4980<p>The first argument, <tt>%val</tt> and the result may be integer types of
4981any bit width but they must have the same bit width. <tt>%val</tt> is the value
4982whose bits will be replaced. The second argument, <tt>%repl</tt> may be an
4983integer of any bit width. The third and fourth arguments must be <tt>i32</tt>
4984type since they specify only a bit index.</p>
4985
4986<h5>Semantics:</h5>
4987<p>The operation of the '<tt>llvm.part.set</tt>' intrinsic has two modes
4988of operation: forwards and reverse. If <tt>%lo</tt> is greater than
4989<tt>%hi</tt> then the intrinsic operates in reverse mode. Otherwise it
4990operates in forward mode.</p>
4991<p>For both modes, the <tt>%repl</tt> value is prepared for use by either
4992truncating it down to the size of the replacement area or zero extending it
4993up to that size.</p>
4994<p>In forward mode, the bits between <tt>%lo</tt> and <tt>%hi</tt> (inclusive)
4995are replaced with corresponding bits from <tt>%repl</tt>. That is the 0th bit
4996in <tt>%repl</tt> replaces the <tt>%lo</tt>th bit in <tt>%val</tt> and etc. up
4997to the <tt>%hi</tt>th bit.
Reid Spencerc6749c42007-05-14 16:50:20 +0000[diff] [blame]4998<p>In reverse mode, a similar computation is made except that the bits are
4999reversed. That is, the <tt>0</tt>th bit in <tt>%repl</tt> replaces the
5000<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]5001<h5>Examples:</h5>
5002<pre>
Reid Spencerf0dbf642007-04-12 01:03:03 +0000[diff] [blame]5003 llvm.part.set(0xFFFF, 0, 4, 7) -&gt; 0xFF0F
Reid Spencerc6749c42007-05-14 16:50:20 +0000[diff] [blame]5004 llvm.part.set(0xFFFF, 0, 7, 4) -&gt; 0xFF0F
5005 llvm.part.set(0xFFFF, 1, 7, 4) -&gt; 0xFF8F
5006 llvm.part.set(0xFFFF, F, 8, 3) -&gt; 0xFFE7
Reid Spencerf0dbf642007-04-12 01:03:03 +0000[diff] [blame]5007 llvm.part.set(0xFFFF, 0, 3, 8) -&gt; 0xFE07
Reid Spencerc8910842007-04-11 23:49:50 +0000[diff] [blame]5008</pre>
Reid Spencerf86037f2007-04-11 23:23:49 +0000[diff] [blame]5009</div>
5010
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5011<!-- ======================================================================= -->
5012<div class="doc_subsection">
5013 <a name="int_debugger">Debugger Intrinsics</a>
5014</div>
5015
5016<div class="doc_text">
5017<p>
5018The LLVM debugger intrinsics (which all start with <tt>llvm.dbg.</tt> prefix),
5019are described in the <a
5020href="SourceLevelDebugging.html#format_common_intrinsics">LLVM Source Level
5021Debugging</a> document.
5022</p>
5023</div>
5024
5025
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]5026<!-- ======================================================================= -->
5027<div class="doc_subsection">
5028 <a name="int_eh">Exception Handling Intrinsics</a>
5029</div>
5030
5031<div class="doc_text">
5032<p> The LLVM exception handling intrinsics (which all start with
5033<tt>llvm.eh.</tt> prefix), are described in the <a
5034href="ExceptionHandling.html#format_common_intrinsics">LLVM Exception
5035Handling</a> document. </p>
5036</div>
5037
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]5038<!-- ======================================================================= -->
5039<div class="doc_subsection">
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]5040 <a name="int_trampoline">Trampoline Intrinsic</a>
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]5041</div>
5042
5043<div class="doc_text">
5044<p>
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]5045 This intrinsic makes it possible to excise one parameter, marked with
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]5046 the <tt>nest</tt> attribute, from a function. The result is a callable
5047 function pointer lacking the nest parameter - the caller does not need
5048 to provide a value for it. Instead, the value to use is stored in
5049 advance in a "trampoline", a block of memory usually allocated
5050 on the stack, which also contains code to splice the nest value into the
5051 argument list. This is used to implement the GCC nested function address
5052 extension.
5053</p>
5054<p>
5055 For example, if the function is
5056 <tt>i32 f(i8* nest %c, i32 %x, i32 %y)</tt> then the resulting function
Bill Wendling03295ca2007-09-22 09:23:55 +0000[diff] [blame]5057 pointer has signature <tt>i32 (i32, i32)*</tt>. It can be created as follows:</p>
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]5058<pre>
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]5059 %tramp = alloca [10 x i8], align 4 ; size and alignment only correct for X86
5060 %tramp1 = getelementptr [10 x i8]* %tramp, i32 0, i32 0
5061 %p = call i8* @llvm.init.trampoline( i8* %tramp1, i8* bitcast (i32 (i8* nest , i32, i32)* @f to i8*), i8* %nval )
5062 %fp = bitcast i8* %p to i32 (i32, i32)*
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]5063</pre>
Bill Wendling03295ca2007-09-22 09:23:55 +0000[diff] [blame]5064 <p>The call <tt>%val = call i32 %fp( i32 %x, i32 %y )</tt> is then equivalent
5065 to <tt>%val = call i32 %f( i8* %nval, i32 %x, i32 %y )</tt>.</p>
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]5066</div>
5067
5068<!-- _______________________________________________________________________ -->
5069<div class="doc_subsubsection">
5070 <a name="int_it">'<tt>llvm.init.trampoline</tt>' Intrinsic</a>
5071</div>
5072<div class="doc_text">
5073<h5>Syntax:</h5>
5074<pre>
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]5075declare i8* @llvm.init.trampoline(i8* &lt;tramp&gt;, i8* &lt;func&gt;, i8* &lt;nval&gt;)
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]5076</pre>
5077<h5>Overview:</h5>
5078<p>
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]5079 This fills the memory pointed to by <tt>tramp</tt> with code
5080 and returns a function pointer suitable for executing it.
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]5081</p>
5082<h5>Arguments:</h5>
5083<p>
5084 The <tt>llvm.init.trampoline</tt> intrinsic takes three arguments, all
5085 pointers. The <tt>tramp</tt> argument must point to a sufficiently large
5086 and sufficiently aligned block of memory; this memory is written to by the
Duncan Sandsc00c2ba2007-08-22 23:39:54 +0000[diff] [blame]5087 intrinsic. Note that the size and the alignment are target-specific - LLVM
5088 currently provides no portable way of determining them, so a front-end that
5089 generates this intrinsic needs to have some target-specific knowledge.
5090 The <tt>func</tt> argument must hold a function bitcast to an <tt>i8*</tt>.
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]5091</p>
5092<h5>Semantics:</h5>
5093<p>
5094 The block of memory pointed to by <tt>tramp</tt> is filled with target
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]5095 dependent code, turning it into a function. A pointer to this function is
5096 returned, but needs to be bitcast to an
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]5097 <a href="#int_trampoline">appropriate function pointer type</a>
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]5098 before being called. The new function's signature is the same as that of
5099 <tt>func</tt> with any arguments marked with the <tt>nest</tt> attribute
5100 removed. At most one such <tt>nest</tt> argument is allowed, and it must be
5101 of pointer type. Calling the new function is equivalent to calling
5102 <tt>func</tt> with the same argument list, but with <tt>nval</tt> used for the
5103 missing <tt>nest</tt> argument. If, after calling
5104 <tt>llvm.init.trampoline</tt>, the memory pointed to by <tt>tramp</tt> is
5105 modified, then the effect of any later call to the returned function pointer is
5106 undefined.
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]5107</p>
5108</div>
5109
5110<!-- ======================================================================= -->
5111<div class="doc_subsection">
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]5112 <a name="int_general">General Intrinsics</a>
5113</div>
5114
5115<div class="doc_text">
5116<p> This class of intrinsics is designed to be generic and has
5117no specific purpose. </p>
5118</div>
5119
5120<!-- _______________________________________________________________________ -->
5121<div class="doc_subsubsection">
5122 <a name="int_var_annotation">'<tt>llvm.var.annotation</tt>' Intrinsic</a>
5123</div>
5124
5125<div class="doc_text">
5126
5127<h5>Syntax:</h5>
5128<pre>
Tanya Lattnerd2e84422007-06-18 23:42:37 +0000[diff] [blame]5129 declare void @llvm.var.annotation(i8* &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]5130</pre>
5131
5132<h5>Overview:</h5>
5133
5134<p>
5135The '<tt>llvm.var.annotation</tt>' intrinsic
5136</p>
5137
5138<h5>Arguments:</h5>
5139
5140<p>
Tanya Lattnerd2e84422007-06-18 23:42:37 +0000[diff] [blame]5141The first argument is a pointer to a value, the second is a pointer to a
5142global string, the third is a pointer to a global string which is the source
5143file name, and the last argument is the line number.
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]5144</p>
5145
5146<h5>Semantics:</h5>
5147
5148<p>
5149This intrinsic allows annotation of local variables with arbitrary strings.
5150This can be useful for special purpose optimizations that want to look for these
5151 annotations. These have no other defined use, they are ignored by code
5152 generation and optimization.
5153</div>
5154
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]5155<!-- _______________________________________________________________________ -->
5156<div class="doc_subsubsection">
Tanya Lattnere1a8da02007-09-21 23:57:59 +0000[diff] [blame]5157 <a name="int_annotation">'<tt>llvm.annotation.*</tt>' Intrinsic</a>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]5158</div>
5159
5160<div class="doc_text">
5161
5162<h5>Syntax:</h5>
Tanya Lattner39cfba62007-09-21 23:56:27 +0000[diff] [blame]5163<p>This is an overloaded intrinsic. You can use '<tt>llvm.annotation</tt>' on
5164any integer bit width.
5165</p>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]5166<pre>
Tanya Lattnerd3989a82007-09-22 00:03:01 +0000[diff] [blame]5167 declare i8 @llvm.annotation.i8(i8 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
5168 declare i16 @llvm.annotation.i16(i16 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
5169 declare i32 @llvm.annotation.i32(i32 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
5170 declare i64 @llvm.annotation.i64(i64 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
5171 declare i256 @llvm.annotation.i256(i256 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]5172</pre>
5173
5174<h5>Overview:</h5>
Tanya Lattner39cfba62007-09-21 23:56:27 +0000[diff] [blame]5175
5176<p>
5177The '<tt>llvm.annotation</tt>' intrinsic.
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]5178</p>
5179
5180<h5>Arguments:</h5>
5181
5182<p>
5183The first argument is an integer value (result of some expression),
5184the second is a pointer to a global string, the third is a pointer to a global
5185string which is the source file name, and the last argument is the line number.
Tanya Lattner39cfba62007-09-21 23:56:27 +0000[diff] [blame]5186It returns the value of the first argument.
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]5187</p>
5188
5189<h5>Semantics:</h5>
5190
5191<p>
5192This intrinsic allows annotations to be put on arbitrary expressions
5193with arbitrary strings. This can be useful for special purpose optimizations
5194that want to look for these annotations. These have no other defined use, they
5195are ignored by code generation and optimization.
5196</div>
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]5197
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]5198<!-- *********************************************************************** -->
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]5199<hr>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]5200<address>
5201 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
5202 src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
5203 <a href="http://validator.w3.org/check/referer"><img
5204 src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!" /></a>
5205
5206 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
Reid Spencer05fe4b02006-03-14 05:39:39 +0000[diff] [blame]5207 <a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]5208 Last modified: $Date$
5209</address>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5210</body>
5211</html>