Gitiles
Code Review Sign In
gerrit-public.fairphone.software / fp2-dev / platform / external / llvm / d8c5ec6aece5732c766bd7d773ff75976d364168 / . / docs / LangRef.html
blob: e5db8c69cfa666b9baa7b21e9d18c5d1b18cc96a [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">
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]8 <meta name="description"
Reid Spencer3921c742004-08-26 20:44:00 +0000[diff] [blame]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>
Bill Wendling3d10a5a2009-07-20 01:03:30 +0000[diff] [blame]23 <li><a href="#linkage">Linkage Types</a>
24 <ol>
Bill Wendling987e7eb2009-07-20 02:41:50 +0000[diff] [blame]25 <li><a href="#linkage_private">'<tt>private</tt>' Linkage</a></li>
26 <li><a href="#linkage_linker_private">'<tt>linker_private</tt>' Linkage</a></li>
Bill Wendling5e721d72010-07-01 21:55:59 +0000[diff] [blame]27 <li><a href="#linkage_linker_private_weak">'<tt>linker_private_weak</tt>' Linkage</a></li>
Bill Wendling55ae5152010-08-20 22:05:50 +0000[diff] [blame]28 <li><a href="#linkage_linker_private_weak_def_auto">'<tt>linker_private_weak_def_auto</tt>' Linkage</a></li>
Bill Wendling987e7eb2009-07-20 02:41:50 +0000[diff] [blame]29 <li><a href="#linkage_internal">'<tt>internal</tt>' Linkage</a></li>
30 <li><a href="#linkage_available_externally">'<tt>available_externally</tt>' Linkage</a></li>
31 <li><a href="#linkage_linkonce">'<tt>linkonce</tt>' Linkage</a></li>
32 <li><a href="#linkage_common">'<tt>common</tt>' Linkage</a></li>
33 <li><a href="#linkage_weak">'<tt>weak</tt>' Linkage</a></li>
34 <li><a href="#linkage_appending">'<tt>appending</tt>' Linkage</a></li>
35 <li><a href="#linkage_externweak">'<tt>extern_weak</tt>' Linkage</a></li>
Chris Lattner5a2d8752009-10-10 18:26:06 +0000[diff] [blame]36 <li><a href="#linkage_linkonce_odr">'<tt>linkonce_odr</tt>' Linkage</a></li>
Bill Wendling987e7eb2009-07-20 02:41:50 +0000[diff] [blame]37 <li><a href="#linkage_weak">'<tt>weak_odr</tt>' Linkage</a></li>
38 <li><a href="#linkage_external">'<tt>externally visible</tt>' Linkage</a></li>
39 <li><a href="#linkage_dllimport">'<tt>dllimport</tt>' Linkage</a></li>
40 <li><a href="#linkage_dllexport">'<tt>dllexport</tt>' Linkage</a></li>
Bill Wendling3d10a5a2009-07-20 01:03:30 +0000[diff] [blame]41 </ol>
42 </li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]43 <li><a href="#callingconv">Calling Conventions</a></li>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]44 <li><a href="#namedtypes">Named Types</a></li>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]45 <li><a href="#globalvars">Global Variables</a></li>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]46 <li><a href="#functionstructure">Functions</a></li>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]47 <li><a href="#aliasstructure">Aliases</a></li>
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]48 <li><a href="#namedmetadatastructure">Named Metadata</a></li>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]49 <li><a href="#paramattrs">Parameter Attributes</a></li>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]50 <li><a href="#fnattrs">Function Attributes</a></li>
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame]51 <li><a href="#gc">Garbage Collector Names</a></li>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]52 <li><a href="#moduleasm">Module-Level Inline Assembly</a></li>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]53 <li><a href="#datalayout">Data Layout</a></li>
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]54 <li><a href="#pointeraliasing">Pointer Aliasing Rules</a></li>
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]55 <li><a href="#volatile">Volatile Memory Accesses</a></li>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]56 </ol>
57 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]58 <li><a href="#typesystem">Type System</a>
59 <ol>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]60 <li><a href="#t_classifications">Type Classifications</a></li>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]61 <li><a href="#t_primitive">Primitive Types</a>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]62 <ol>
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]63 <li><a href="#t_integer">Integer Type</a></li>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]64 <li><a href="#t_floating">Floating Point Types</a></li>
Dale Johannesen21fe99b2010-10-01 00:48:59 +0000[diff] [blame]65 <li><a href="#t_x86mmx">X86mmx Type</a></li>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]66 <li><a href="#t_void">Void Type</a></li>
67 <li><a href="#t_label">Label Type</a></li>
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]68 <li><a href="#t_metadata">Metadata Type</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]69 </ol>
70 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]71 <li><a href="#t_derived">Derived Types</a>
72 <ol>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]73 <li><a href="#t_aggregate">Aggregate Types</a>
74 <ol>
75 <li><a href="#t_array">Array Type</a></li>
76 <li><a href="#t_struct">Structure Type</a></li>
77 <li><a href="#t_pstruct">Packed Structure Type</a></li>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]78 <li><a href="#t_vector">Vector Type</a></li>
79 </ol>
80 </li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]81 <li><a href="#t_function">Function Type</a></li>
82 <li><a href="#t_pointer">Pointer Type</a></li>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]83 <li><a href="#t_opaque">Opaque Type</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]84 </ol>
85 </li>
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]86 <li><a href="#t_uprefs">Type Up-references</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]87 </ol>
88 </li>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]89 <li><a href="#constants">Constants</a>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]90 <ol>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]91 <li><a href="#simpleconstants">Simple Constants</a></li>
Chris Lattner70882792009-02-28 18:32:25 +0000[diff] [blame]92 <li><a href="#complexconstants">Complex Constants</a></li>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]93 <li><a href="#globalconstants">Global Variable and Function Addresses</a></li>
94 <li><a href="#undefvalues">Undefined Values</a></li>
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]95 <li><a href="#trapvalues">Trap Values</a></li>
Chris Lattnerf9d078e2009-10-27 21:19:13 +0000[diff] [blame]96 <li><a href="#blockaddress">Addresses of Basic Blocks</a></li>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]97 <li><a href="#constantexprs">Constant Expressions</a></li>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]98 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]99 </li>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]100 <li><a href="#othervalues">Other Values</a>
101 <ol>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]102 <li><a href="#inlineasm">Inline Assembler Expressions</a></li>
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]103 <li><a href="#metadata">Metadata Nodes and Metadata Strings</a></li>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]104 </ol>
105 </li>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]106 <li><a href="#intrinsic_globals">Intrinsic Global Variables</a>
107 <ol>
108 <li><a href="#intg_used">The '<tt>llvm.used</tt>' Global Variable</a></li>
Chris Lattner401e10c2009-07-20 06:14:25 +0000[diff] [blame]109 <li><a href="#intg_compiler_used">The '<tt>llvm.compiler.used</tt>'
110 Global Variable</a></li>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]111 <li><a href="#intg_global_ctors">The '<tt>llvm.global_ctors</tt>'
112 Global Variable</a></li>
113 <li><a href="#intg_global_dtors">The '<tt>llvm.global_dtors</tt>'
114 Global Variable</a></li>
115 </ol>
116 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]117 <li><a href="#instref">Instruction Reference</a>
118 <ol>
119 <li><a href="#terminators">Terminator Instructions</a>
120 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]121 <li><a href="#i_ret">'<tt>ret</tt>' Instruction</a></li>
122 <li><a href="#i_br">'<tt>br</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]123 <li><a href="#i_switch">'<tt>switch</tt>' Instruction</a></li>
Chris Lattnerab21db72009-10-28 00:19:10 +0000[diff] [blame]124 <li><a href="#i_indirectbr">'<tt>indirectbr</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]125 <li><a href="#i_invoke">'<tt>invoke</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]126 <li><a href="#i_unwind">'<tt>unwind</tt>' Instruction</a></li>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]127 <li><a href="#i_unreachable">'<tt>unreachable</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]128 </ol>
129 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]130 <li><a href="#binaryops">Binary Operations</a>
131 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]132 <li><a href="#i_add">'<tt>add</tt>' Instruction</a></li>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]133 <li><a href="#i_fadd">'<tt>fadd</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]134 <li><a href="#i_sub">'<tt>sub</tt>' Instruction</a></li>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]135 <li><a href="#i_fsub">'<tt>fsub</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]136 <li><a href="#i_mul">'<tt>mul</tt>' Instruction</a></li>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]137 <li><a href="#i_fmul">'<tt>fmul</tt>' Instruction</a></li>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]138 <li><a href="#i_udiv">'<tt>udiv</tt>' Instruction</a></li>
139 <li><a href="#i_sdiv">'<tt>sdiv</tt>' Instruction</a></li>
140 <li><a href="#i_fdiv">'<tt>fdiv</tt>' Instruction</a></li>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]141 <li><a href="#i_urem">'<tt>urem</tt>' Instruction</a></li>
142 <li><a href="#i_srem">'<tt>srem</tt>' Instruction</a></li>
143 <li><a href="#i_frem">'<tt>frem</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]144 </ol>
145 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]146 <li><a href="#bitwiseops">Bitwise Binary Operations</a>
147 <ol>
Reid Spencer8e11bf82007-02-02 13:57:07 +0000[diff] [blame]148 <li><a href="#i_shl">'<tt>shl</tt>' Instruction</a></li>
149 <li><a href="#i_lshr">'<tt>lshr</tt>' Instruction</a></li>
150 <li><a href="#i_ashr">'<tt>ashr</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]151 <li><a href="#i_and">'<tt>and</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]152 <li><a href="#i_or">'<tt>or</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]153 <li><a href="#i_xor">'<tt>xor</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]154 </ol>
155 </li>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]156 <li><a href="#vectorops">Vector Operations</a>
157 <ol>
158 <li><a href="#i_extractelement">'<tt>extractelement</tt>' Instruction</a></li>
159 <li><a href="#i_insertelement">'<tt>insertelement</tt>' Instruction</a></li>
160 <li><a href="#i_shufflevector">'<tt>shufflevector</tt>' Instruction</a></li>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]161 </ol>
162 </li>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]163 <li><a href="#aggregateops">Aggregate Operations</a>
164 <ol>
165 <li><a href="#i_extractvalue">'<tt>extractvalue</tt>' Instruction</a></li>
166 <li><a href="#i_insertvalue">'<tt>insertvalue</tt>' Instruction</a></li>
167 </ol>
168 </li>
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]169 <li><a href="#memoryops">Memory Access and Addressing Operations</a>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]170 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]171 <li><a href="#i_alloca">'<tt>alloca</tt>' Instruction</a></li>
Robert Bocchino7b81c752006-02-17 21:18:08 +0000[diff] [blame]172 <li><a href="#i_load">'<tt>load</tt>' Instruction</a></li>
173 <li><a href="#i_store">'<tt>store</tt>' Instruction</a></li>
174 <li><a href="#i_getelementptr">'<tt>getelementptr</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]175 </ol>
176 </li>
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]177 <li><a href="#convertops">Conversion Operations</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]178 <ol>
179 <li><a href="#i_trunc">'<tt>trunc .. to</tt>' Instruction</a></li>
180 <li><a href="#i_zext">'<tt>zext .. to</tt>' Instruction</a></li>
181 <li><a href="#i_sext">'<tt>sext .. to</tt>' Instruction</a></li>
182 <li><a href="#i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a></li>
183 <li><a href="#i_fpext">'<tt>fpext .. to</tt>' Instruction</a></li>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]184 <li><a href="#i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a></li>
185 <li><a href="#i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a></li>
186 <li><a href="#i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a></li>
187 <li><a href="#i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a></li>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]188 <li><a href="#i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a></li>
189 <li><a href="#i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a></li>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]190 <li><a href="#i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a></li>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]191 </ol>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]192 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]193 <li><a href="#otherops">Other Operations</a>
194 <ol>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]195 <li><a href="#i_icmp">'<tt>icmp</tt>' Instruction</a></li>
196 <li><a href="#i_fcmp">'<tt>fcmp</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]197 <li><a href="#i_phi">'<tt>phi</tt>' Instruction</a></li>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]198 <li><a href="#i_select">'<tt>select</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]199 <li><a href="#i_call">'<tt>call</tt>' Instruction</a></li>
Chris Lattnerfb6977d2006-01-13 23:26:01 +0000[diff] [blame]200 <li><a href="#i_va_arg">'<tt>va_arg</tt>' Instruction</a></li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]201 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]202 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]203 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]204 </li>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]205 <li><a href="#intrinsics">Intrinsic Functions</a>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]206 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]207 <li><a href="#int_varargs">Variable Argument Handling Intrinsics</a>
208 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]209 <li><a href="#int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a></li>
210 <li><a href="#int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a></li>
211 <li><a href="#int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]212 </ol>
213 </li>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]214 <li><a href="#int_gc">Accurate Garbage Collection Intrinsics</a>
215 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]216 <li><a href="#int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a></li>
217 <li><a href="#int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a></li>
218 <li><a href="#int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a></li>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]219 </ol>
220 </li>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]221 <li><a href="#int_codegen">Code Generator Intrinsics</a>
222 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]223 <li><a href="#int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a></li>
224 <li><a href="#int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a></li>
225 <li><a href="#int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a></li>
226 <li><a href="#int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a></li>
227 <li><a href="#int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a></li>
228 <li><a href="#int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a></li>
Dan Gohman31f1af12010-05-26 21:56:15 +0000[diff] [blame]229 <li><a href="#int_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a></li>
John Criswell7123e272004-04-09 16:43:20 +0000[diff] [blame]230 </ol>
231 </li>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]232 <li><a href="#int_libc">Standard C Library Intrinsics</a>
233 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]234 <li><a href="#int_memcpy">'<tt>llvm.memcpy.*</tt>' Intrinsic</a></li>
235 <li><a href="#int_memmove">'<tt>llvm.memmove.*</tt>' Intrinsic</a></li>
236 <li><a href="#int_memset">'<tt>llvm.memset.*</tt>' Intrinsic</a></li>
237 <li><a href="#int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a></li>
238 <li><a href="#int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a></li>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]239 <li><a href="#int_sin">'<tt>llvm.sin.*</tt>' Intrinsic</a></li>
240 <li><a href="#int_cos">'<tt>llvm.cos.*</tt>' Intrinsic</a></li>
241 <li><a href="#int_pow">'<tt>llvm.pow.*</tt>' Intrinsic</a></li>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]242 </ol>
243 </li>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]244 <li><a href="#int_manip">Bit Manipulation Intrinsics</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]245 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]246 <li><a href="#int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a></li>
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]247 <li><a href="#int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic </a></li>
248 <li><a href="#int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic </a></li>
249 <li><a href="#int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic </a></li>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]250 </ol>
251 </li>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]252 <li><a href="#int_overflow">Arithmetic with Overflow Intrinsics</a>
253 <ol>
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]254 <li><a href="#int_sadd_overflow">'<tt>llvm.sadd.with.overflow.*</tt> Intrinsics</a></li>
255 <li><a href="#int_uadd_overflow">'<tt>llvm.uadd.with.overflow.*</tt> Intrinsics</a></li>
256 <li><a href="#int_ssub_overflow">'<tt>llvm.ssub.with.overflow.*</tt> Intrinsics</a></li>
257 <li><a href="#int_usub_overflow">'<tt>llvm.usub.with.overflow.*</tt> Intrinsics</a></li>
258 <li><a href="#int_smul_overflow">'<tt>llvm.smul.with.overflow.*</tt> Intrinsics</a></li>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]259 <li><a href="#int_umul_overflow">'<tt>llvm.umul.with.overflow.*</tt> Intrinsics</a></li>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]260 </ol>
261 </li>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]262 <li><a href="#int_fp16">Half Precision Floating Point Intrinsics</a>
263 <ol>
Chris Lattner82c3dc62010-03-14 23:03:31 +0000[diff] [blame]264 <li><a href="#int_convert_to_fp16">'<tt>llvm.convert.to.fp16</tt>' Intrinsic</a></li>
265 <li><a href="#int_convert_from_fp16">'<tt>llvm.convert.from.fp16</tt>' Intrinsic</a></li>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]266 </ol>
267 </li>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]268 <li><a href="#int_debugger">Debugger intrinsics</a></li>
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]269 <li><a href="#int_eh">Exception Handling intrinsics</a></li>
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]270 <li><a href="#int_trampoline">Trampoline Intrinsic</a>
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]271 <ol>
272 <li><a href="#int_it">'<tt>llvm.init.trampoline</tt>' Intrinsic</a></li>
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]273 </ol>
274 </li>
Bill Wendling3c44f5b2008-11-18 22:10:53 +0000[diff] [blame]275 <li><a href="#int_atomics">Atomic intrinsics</a>
276 <ol>
277 <li><a href="#int_memory_barrier"><tt>llvm.memory_barrier</tt></a></li>
278 <li><a href="#int_atomic_cmp_swap"><tt>llvm.atomic.cmp.swap</tt></a></li>
279 <li><a href="#int_atomic_swap"><tt>llvm.atomic.swap</tt></a></li>
280 <li><a href="#int_atomic_load_add"><tt>llvm.atomic.load.add</tt></a></li>
281 <li><a href="#int_atomic_load_sub"><tt>llvm.atomic.load.sub</tt></a></li>
282 <li><a href="#int_atomic_load_and"><tt>llvm.atomic.load.and</tt></a></li>
283 <li><a href="#int_atomic_load_nand"><tt>llvm.atomic.load.nand</tt></a></li>
284 <li><a href="#int_atomic_load_or"><tt>llvm.atomic.load.or</tt></a></li>
285 <li><a href="#int_atomic_load_xor"><tt>llvm.atomic.load.xor</tt></a></li>
286 <li><a href="#int_atomic_load_max"><tt>llvm.atomic.load.max</tt></a></li>
287 <li><a href="#int_atomic_load_min"><tt>llvm.atomic.load.min</tt></a></li>
288 <li><a href="#int_atomic_load_umax"><tt>llvm.atomic.load.umax</tt></a></li>
289 <li><a href="#int_atomic_load_umin"><tt>llvm.atomic.load.umin</tt></a></li>
290 </ol>
291 </li>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]292 <li><a href="#int_memorymarkers">Memory Use Markers</a>
293 <ol>
294 <li><a href="#int_lifetime_start"><tt>llvm.lifetime.start</tt></a></li>
295 <li><a href="#int_lifetime_end"><tt>llvm.lifetime.end</tt></a></li>
296 <li><a href="#int_invariant_start"><tt>llvm.invariant.start</tt></a></li>
297 <li><a href="#int_invariant_end"><tt>llvm.invariant.end</tt></a></li>
298 </ol>
299 </li>
Reid Spencer20677642007-07-20 19:59:11 +0000[diff] [blame]300 <li><a href="#int_general">General intrinsics</a>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]301 <ol>
Reid Spencer20677642007-07-20 19:59:11 +0000[diff] [blame]302 <li><a href="#int_var_annotation">
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]303 '<tt>llvm.var.annotation</tt>' Intrinsic</a></li>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]304 <li><a href="#int_annotation">
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]305 '<tt>llvm.annotation.*</tt>' Intrinsic</a></li>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]306 <li><a href="#int_trap">
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]307 '<tt>llvm.trap</tt>' Intrinsic</a></li>
308 <li><a href="#int_stackprotector">
309 '<tt>llvm.stackprotector</tt>' Intrinsic</a></li>
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]310 <li><a href="#int_objectsize">
311 '<tt>llvm.objectsize</tt>' Intrinsic</a></li>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]312 </ol>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]313 </li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]314 </ol>
315 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]316</ol>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]317
318<div class="doc_author">
319 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>
320 and <a href="mailto:vadve@cs.uiuc.edu">Vikram Adve</a></p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]321</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]322
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]323<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]324<div class="doc_section"> <a name="abstract">Abstract </a></div>
325<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]326
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]327<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]328
329<p>This document is a reference manual for the LLVM assembly language. LLVM is
330 a Static Single Assignment (SSA) based representation that provides type
331 safety, low-level operations, flexibility, and the capability of representing
332 'all' high-level languages cleanly. It is the common code representation
333 used throughout all phases of the LLVM compilation strategy.</p>
334
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]335</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]336
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]337<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]338<div class="doc_section"> <a name="introduction">Introduction</a> </div>
339<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]340
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]341<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]342
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]343<p>The LLVM code representation is designed to be used in three different forms:
344 as an in-memory compiler IR, as an on-disk bitcode representation (suitable
345 for fast loading by a Just-In-Time compiler), and as a human readable
346 assembly language representation. This allows LLVM to provide a powerful
347 intermediate representation for efficient compiler transformations and
348 analysis, while providing a natural means to debug and visualize the
349 transformations. The three different forms of LLVM are all equivalent. This
350 document describes the human readable representation and notation.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]351
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]352<p>The LLVM representation aims to be light-weight and low-level while being
353 expressive, typed, and extensible at the same time. It aims to be a
354 "universal IR" of sorts, by being at a low enough level that high-level ideas
355 may be cleanly mapped to it (similar to how microprocessors are "universal
356 IR's", allowing many source languages to be mapped to them). By providing
357 type information, LLVM can be used as the target of optimizations: for
358 example, through pointer analysis, it can be proven that a C automatic
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]359 variable is never accessed outside of the current function, allowing it to
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]360 be promoted to a simple SSA value instead of a memory location.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]361
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]362</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]363
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]364<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]365<div class="doc_subsubsection"> <a name="wellformed">Well-Formedness</a> </div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]366
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]367<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]368
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]369<p>It is important to note that this document describes 'well formed' LLVM
370 assembly language. There is a difference between what the parser accepts and
371 what is considered 'well formed'. For example, the following instruction is
372 syntactically okay, but not well formed:</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]373
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]374<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]375%x = <a href="#i_add">add</a> i32 1, %x
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]376</pre>
377
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]378<p>because the definition of <tt>%x</tt> does not dominate all of its uses. The
379 LLVM infrastructure provides a verification pass that may be used to verify
380 that an LLVM module is well formed. This pass is automatically run by the
381 parser after parsing input assembly and by the optimizer before it outputs
382 bitcode. The violations pointed out by the verifier pass indicate bugs in
383 transformation passes or input to the parser.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]384
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]385</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]386
Chris Lattnercc689392007-10-03 17:34:29 +0000[diff] [blame]387<!-- Describe the typesetting conventions here. -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]388
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]389<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]390<div class="doc_section"> <a name="identifiers">Identifiers</a> </div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]391<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]392
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]393<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]394
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]395<p>LLVM identifiers come in two basic types: global and local. Global
396 identifiers (functions, global variables) begin with the <tt>'@'</tt>
397 character. Local identifiers (register names, types) begin with
398 the <tt>'%'</tt> character. Additionally, there are three different formats
399 for identifiers, for different purposes:</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]400
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]401<ol>
Reid Spencer2c452282007-08-07 14:34:28 +0000[diff] [blame]402 <li>Named values are represented as a string of characters with their prefix.
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]403 For example, <tt>%foo</tt>, <tt>@DivisionByZero</tt>,
404 <tt>%a.really.long.identifier</tt>. The actual regular expression used is
405 '<tt>[%@][a-zA-Z$._][a-zA-Z$._0-9]*</tt>'. Identifiers which require
406 other characters in their names can be surrounded with quotes. Special
407 characters may be escaped using <tt>"\xx"</tt> where <tt>xx</tt> is the
408 ASCII code for the character in hexadecimal. In this way, any character
409 can be used in a name value, even quotes themselves.</li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]410
Reid Spencer2c452282007-08-07 14:34:28 +0000[diff] [blame]411 <li>Unnamed values are represented as an unsigned numeric value with their
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]412 prefix. For example, <tt>%12</tt>, <tt>@2</tt>, <tt>%44</tt>.</li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]413
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]414 <li>Constants, which are described in a <a href="#constants">section about
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]415 constants</a>, below.</li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]416</ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]417
Reid Spencer2c452282007-08-07 14:34:28 +0000[diff] [blame]418<p>LLVM requires that values start with a prefix for two reasons: Compilers
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]419 don't need to worry about name clashes with reserved words, and the set of
420 reserved words may be expanded in the future without penalty. Additionally,
421 unnamed identifiers allow a compiler to quickly come up with a temporary
422 variable without having to avoid symbol table conflicts.</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]423
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]424<p>Reserved words in LLVM are very similar to reserved words in other
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]425 languages. There are keywords for different opcodes
426 ('<tt><a href="#i_add">add</a></tt>',
427 '<tt><a href="#i_bitcast">bitcast</a></tt>',
428 '<tt><a href="#i_ret">ret</a></tt>', etc...), for primitive type names
429 ('<tt><a href="#t_void">void</a></tt>',
430 '<tt><a href="#t_primitive">i32</a></tt>', etc...), and others. These
431 reserved words cannot conflict with variable names, because none of them
432 start with a prefix character (<tt>'%'</tt> or <tt>'@'</tt>).</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]433
434<p>Here is an example of LLVM code to multiply the integer variable
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]435 '<tt>%X</tt>' by 8:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]436
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]437<p>The easy way:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]438
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]439<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]440%result = <a href="#i_mul">mul</a> i32 %X, 8
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]441</pre>
442
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]443<p>After strength reduction:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]444
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]445<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]446%result = <a href="#i_shl">shl</a> i32 %X, i8 3
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]447</pre>
448
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]449<p>And the hard way:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]450
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]451<pre class="doc_code">
Gabor Greifec58f752009-10-28 13:05:07 +0000[diff] [blame]452%0 = <a href="#i_add">add</a> i32 %X, %X <i>; yields {i32}:%0</i>
453%1 = <a href="#i_add">add</a> i32 %0, %0 <i>; yields {i32}:%1</i>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]454%result = <a href="#i_add">add</a> i32 %1, %1
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]455</pre>
456
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]457<p>This last way of multiplying <tt>%X</tt> by 8 illustrates several important
458 lexical features of LLVM:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]459
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]460<ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]461 <li>Comments are delimited with a '<tt>;</tt>' and go until the end of
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]462 line.</li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]463
464 <li>Unnamed temporaries are created when the result of a computation is not
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]465 assigned to a named value.</li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]466
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]467 <li>Unnamed temporaries are numbered sequentially</li>
468</ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]469
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]470<p>It also shows a convention that we follow in this document. When
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]471 demonstrating instructions, we will follow an instruction with a comment that
472 defines the type and name of value produced. Comments are shown in italic
473 text.</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]474
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]475</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]476
477<!-- *********************************************************************** -->
478<div class="doc_section"> <a name="highlevel">High Level Structure</a> </div>
479<!-- *********************************************************************** -->
480
481<!-- ======================================================================= -->
482<div class="doc_subsection"> <a name="modulestructure">Module Structure</a>
483</div>
484
485<div class="doc_text">
486
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]487<p>LLVM programs are composed of "Module"s, each of which is a translation unit
488 of the input programs. Each module consists of functions, global variables,
489 and symbol table entries. Modules may be combined together with the LLVM
490 linker, which merges function (and global variable) definitions, resolves
491 forward declarations, and merges symbol table entries. Here is an example of
492 the "hello world" module:</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]493
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]494<pre class="doc_code">
Chris Lattner63e4ccb2010-08-17 17:13:42 +0000[diff] [blame]495<i>; Declare the string constant as a global constant.</i>&nbsp;
496<a href="#identifiers">@.LC0</a> = <a href="#linkage_internal">internal</a> <a href="#globalvars">constant</a> <a href="#t_array">[13 x i8]</a> c"hello world\0A\00" <i>; [13 x i8]*</i>&nbsp;
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]497
Chris Lattner63e4ccb2010-08-17 17:13:42 +0000[diff] [blame]498<i>; External declaration of the puts function</i>&nbsp;
499<a href="#functionstructure">declare</a> i32 @puts(i8*) <i>; i32 (i8*)* </i>&nbsp;
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]500
501<i>; Definition of main function</i>
Chris Lattner63e4ccb2010-08-17 17:13:42 +0000[diff] [blame]502define i32 @main() { <i>; i32()* </i>&nbsp;
503 <i>; Convert [13 x i8]* to i8 *...</i>&nbsp;
504 %cast210 = <a href="#i_getelementptr">getelementptr</a> [13 x i8]* @.LC0, i64 0, i64 0 <i>; i8*</i>&nbsp;
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]505
Chris Lattner63e4ccb2010-08-17 17:13:42 +0000[diff] [blame]506 <i>; Call puts function to write out the string to stdout.</i>&nbsp;
507 <a href="#i_call">call</a> i32 @puts(i8* %cast210) <i>; i32</i>&nbsp;
508 <a href="#i_ret">ret</a> i32 0&nbsp;
509}
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]510
511<i>; Named metadata</i>
512!1 = metadata !{i32 41}
513!foo = !{!1, null}
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]514</pre>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]515
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]516<p>This example is made up of a <a href="#globalvars">global variable</a> named
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]517 "<tt>.LC0</tt>", an external declaration of the "<tt>puts</tt>" function,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]518 a <a href="#functionstructure">function definition</a> for
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]519 "<tt>main</tt>" and <a href="#namedmetadatastructure">named metadata</a>
520 "<tt>foo"</tt>.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]521
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]522<p>In general, a module is made up of a list of global values, where both
523 functions and global variables are global values. Global values are
524 represented by a pointer to a memory location (in this case, a pointer to an
525 array of char, and a pointer to a function), and have one of the
526 following <a href="#linkage">linkage types</a>.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]527
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]528</div>
529
530<!-- ======================================================================= -->
531<div class="doc_subsection">
532 <a name="linkage">Linkage Types</a>
533</div>
534
535<div class="doc_text">
536
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]537<p>All Global Variables and Functions have one of the following types of
538 linkage:</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]539
540<dl>
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]541 <dt><tt><b><a name="linkage_private">private</a></b></tt></dt>
Bill Wendling5e721d72010-07-01 21:55:59 +0000[diff] [blame]542 <dd>Global values with "<tt>private</tt>" linkage are only directly accessible
543 by objects in the current module. In particular, linking code into a
544 module with an private global value may cause the private to be renamed as
545 necessary to avoid collisions. Because the symbol is private to the
546 module, all references can be updated. This doesn't show up in any symbol
547 table in the object file.</dd>
Rafael Espindolabb46f522009-01-15 20:18:42 +0000[diff] [blame]548
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]549 <dt><tt><b><a name="linkage_linker_private">linker_private</a></b></tt></dt>
Bill Wendling5e721d72010-07-01 21:55:59 +0000[diff] [blame]550 <dd>Similar to <tt>private</tt>, but the symbol is passed through the
551 assembler and evaluated by the linker. Unlike normal strong symbols, they
552 are removed by the linker from the final linked image (executable or
553 dynamic library).</dd>
554
555 <dt><tt><b><a name="linkage_linker_private_weak">linker_private_weak</a></b></tt></dt>
556 <dd>Similar to "<tt>linker_private</tt>", but the symbol is weak. Note that
557 <tt>linker_private_weak</tt> symbols are subject to coalescing by the
558 linker. The symbols are removed by the linker from the final linked image
559 (executable or dynamic library).</dd>
Bill Wendling3d10a5a2009-07-20 01:03:30 +0000[diff] [blame]560
Bill Wendling55ae5152010-08-20 22:05:50 +0000[diff] [blame]561 <dt><tt><b><a name="linkage_linker_private_weak_def_auto">linker_private_weak_def_auto</a></b></tt></dt>
562 <dd>Similar to "<tt>linker_private_weak</tt>", but it's known that the address
563 of the object is not taken. For instance, functions that had an inline
564 definition, but the compiler decided not to inline it. Note,
565 unlike <tt>linker_private</tt> and <tt>linker_private_weak</tt>,
566 <tt>linker_private_weak_def_auto</tt> may have only <tt>default</tt>
567 visibility. The symbols are removed by the linker from the final linked
568 image (executable or dynamic library).</dd>
569
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]570 <dt><tt><b><a name="linkage_internal">internal</a></b></tt></dt>
Bill Wendling07d31772010-06-29 22:34:52 +0000[diff] [blame]571 <dd>Similar to private, but the value shows as a local symbol
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]572 (<tt>STB_LOCAL</tt> in the case of ELF) in the object file. This
573 corresponds to the notion of the '<tt>static</tt>' keyword in C.</dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]574
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]575 <dt><tt><b><a name="linkage_available_externally">available_externally</a></b></tt></dt>
Chris Lattner266c7bb2009-04-13 05:44:34 +0000[diff] [blame]576 <dd>Globals with "<tt>available_externally</tt>" linkage are never emitted
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]577 into the object file corresponding to the LLVM module. They exist to
578 allow inlining and other optimizations to take place given knowledge of
579 the definition of the global, which is known to be somewhere outside the
580 module. Globals with <tt>available_externally</tt> linkage are allowed to
581 be discarded at will, and are otherwise the same as <tt>linkonce_odr</tt>.
582 This linkage type is only allowed on definitions, not declarations.</dd>
Chris Lattner266c7bb2009-04-13 05:44:34 +0000[diff] [blame]583
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]584 <dt><tt><b><a name="linkage_linkonce">linkonce</a></b></tt></dt>
Chris Lattner4887bd82007-01-14 06:51:48 +0000[diff] [blame]585 <dd>Globals with "<tt>linkonce</tt>" linkage are merged with other globals of
Chris Lattner873187c2010-01-09 19:15:14 +0000[diff] [blame]586 the same name when linkage occurs. This can be used to implement
587 some forms of inline functions, templates, or other code which must be
588 generated in each translation unit that uses it, but where the body may
589 be overridden with a more definitive definition later. Unreferenced
590 <tt>linkonce</tt> globals are allowed to be discarded. Note that
591 <tt>linkonce</tt> linkage does not actually allow the optimizer to
592 inline the body of this function into callers because it doesn't know if
593 this definition of the function is the definitive definition within the
594 program or whether it will be overridden by a stronger definition.
595 To enable inlining and other optimizations, use "<tt>linkonce_odr</tt>"
596 linkage.</dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]597
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]598 <dt><tt><b><a name="linkage_weak">weak</a></b></tt></dt>
Chris Lattner26d054d2009-08-05 05:21:07 +0000[diff] [blame]599 <dd>"<tt>weak</tt>" linkage has the same merging semantics as
600 <tt>linkonce</tt> linkage, except that unreferenced globals with
601 <tt>weak</tt> linkage may not be discarded. This is used for globals that
602 are declared "weak" in C source code.</dd>
603
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]604 <dt><tt><b><a name="linkage_common">common</a></b></tt></dt>
Chris Lattner26d054d2009-08-05 05:21:07 +0000[diff] [blame]605 <dd>"<tt>common</tt>" linkage is most similar to "<tt>weak</tt>" linkage, but
606 they are used for tentative definitions in C, such as "<tt>int X;</tt>" at
607 global scope.
608 Symbols with "<tt>common</tt>" linkage are merged in the same way as
609 <tt>weak symbols</tt>, and they may not be deleted if unreferenced.
Chris Lattnercd81f5d2009-08-05 05:41:44 +0000[diff] [blame]610 <tt>common</tt> symbols may not have an explicit section,
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]611 must have a zero initializer, and may not be marked '<a
Chris Lattnercd81f5d2009-08-05 05:41:44 +0000[diff] [blame]612 href="#globalvars"><tt>constant</tt></a>'. Functions and aliases may not
613 have common linkage.</dd>
Chris Lattner26d054d2009-08-05 05:21:07 +0000[diff] [blame]614
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]615
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]616 <dt><tt><b><a name="linkage_appending">appending</a></b></tt></dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]617 <dd>"<tt>appending</tt>" linkage may only be applied to global variables of
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]618 pointer to array type. When two global variables with appending linkage
619 are linked together, the two global arrays are appended together. This is
620 the LLVM, typesafe, equivalent of having the system linker append together
621 "sections" with identical names when .o files are linked.</dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]622
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]623 <dt><tt><b><a name="linkage_externweak">extern_weak</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]624 <dd>The semantics of this linkage follow the ELF object file model: the symbol
625 is weak until linked, if not linked, the symbol becomes null instead of
626 being an undefined reference.</dd>
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame]627
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]628 <dt><tt><b><a name="linkage_linkonce_odr">linkonce_odr</a></b></tt></dt>
629 <dt><tt><b><a name="linkage_weak_odr">weak_odr</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]630 <dd>Some languages allow differing globals to be merged, such as two functions
631 with different semantics. Other languages, such as <tt>C++</tt>, ensure
Bill Wendling5e721d72010-07-01 21:55:59 +0000[diff] [blame]632 that only equivalent globals are ever merged (the "one definition rule"
633 &mdash; "ODR"). Such languages can use the <tt>linkonce_odr</tt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]634 and <tt>weak_odr</tt> linkage types to indicate that the global will only
635 be merged with equivalent globals. These linkage types are otherwise the
636 same as their non-<tt>odr</tt> versions.</dd>
Duncan Sands667d4b82009-03-07 15:45:40 +0000[diff] [blame]637
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]638 <dt><tt><b><a name="linkage_external">externally visible</a></b></tt>:</dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]639 <dd>If none of the above identifiers are used, the global is externally
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]640 visible, meaning that it participates in linkage and can be used to
641 resolve external symbol references.</dd>
Reid Spencerc8910842007-04-11 23:49:50 +0000[diff] [blame]642</dl>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]643
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]644<p>The next two types of linkage are targeted for Microsoft Windows platform
645 only. They are designed to support importing (exporting) symbols from (to)
646 DLLs (Dynamic Link Libraries).</p>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]647
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]648<dl>
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]649 <dt><tt><b><a name="linkage_dllimport">dllimport</a></b></tt></dt>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]650 <dd>"<tt>dllimport</tt>" linkage causes the compiler to reference a function
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]651 or variable via a global pointer to a pointer that is set up by the DLL
652 exporting the symbol. On Microsoft Windows targets, the pointer name is
653 formed by combining <code>__imp_</code> and the function or variable
654 name.</dd>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]655
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]656 <dt><tt><b><a name="linkage_dllexport">dllexport</a></b></tt></dt>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]657 <dd>"<tt>dllexport</tt>" linkage causes the compiler to provide a global
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]658 pointer to a pointer in a DLL, so that it can be referenced with the
659 <tt>dllimport</tt> attribute. On Microsoft Windows targets, the pointer
660 name is formed by combining <code>__imp_</code> and the function or
661 variable name.</dd>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]662</dl>
663
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]664<p>For example, since the "<tt>.LC0</tt>" variable is defined to be internal, if
665 another module defined a "<tt>.LC0</tt>" variable and was linked with this
666 one, one of the two would be renamed, preventing a collision. Since
667 "<tt>main</tt>" and "<tt>puts</tt>" are external (i.e., lacking any linkage
668 declarations), they are accessible outside of the current module.</p>
669
670<p>It is illegal for a function <i>declaration</i> to have any linkage type
671 other than "externally visible", <tt>dllimport</tt>
672 or <tt>extern_weak</tt>.</p>
673
Duncan Sands667d4b82009-03-07 15:45:40 +0000[diff] [blame]674<p>Aliases can have only <tt>external</tt>, <tt>internal</tt>, <tt>weak</tt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]675 or <tt>weak_odr</tt> linkages.</p>
676
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]677</div>
678
679<!-- ======================================================================= -->
680<div class="doc_subsection">
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]681 <a name="callingconv">Calling Conventions</a>
682</div>
683
684<div class="doc_text">
685
686<p>LLVM <a href="#functionstructure">functions</a>, <a href="#i_call">calls</a>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]687 and <a href="#i_invoke">invokes</a> can all have an optional calling
688 convention specified for the call. The calling convention of any pair of
689 dynamic caller/callee must match, or the behavior of the program is
690 undefined. The following calling conventions are supported by LLVM, and more
691 may be added in the future:</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]692
693<dl>
694 <dt><b>"<tt>ccc</tt>" - The C calling convention</b>:</dt>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]695 <dd>This calling convention (the default if no other calling convention is
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]696 specified) matches the target C calling conventions. This calling
697 convention supports varargs function calls and tolerates some mismatch in
698 the declared prototype and implemented declaration of the function (as
699 does normal C).</dd>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]700
701 <dt><b>"<tt>fastcc</tt>" - The fast calling convention</b>:</dt>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]702 <dd>This calling convention attempts to make calls as fast as possible
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]703 (e.g. by passing things in registers). This calling convention allows the
704 target to use whatever tricks it wants to produce fast code for the
705 target, without having to conform to an externally specified ABI
Jeffrey Yasskin95fa80a2010-01-09 19:44:16 +0000[diff] [blame]706 (Application Binary Interface).
707 <a href="CodeGenerator.html#tailcallopt">Tail calls can only be optimized
Chris Lattner29689432010-03-11 00:22:57 +0000[diff] [blame]708 when this or the GHC convention is used.</a> This calling convention
709 does not support varargs and requires the prototype of all callees to
710 exactly match the prototype of the function definition.</dd>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]711
712 <dt><b>"<tt>coldcc</tt>" - The cold calling convention</b>:</dt>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]713 <dd>This calling convention attempts to make code in the caller as efficient
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]714 as possible under the assumption that the call is not commonly executed.
715 As such, these calls often preserve all registers so that the call does
716 not break any live ranges in the caller side. This calling convention
717 does not support varargs and requires the prototype of all callees to
718 exactly match the prototype of the function definition.</dd>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]719
Chris Lattner29689432010-03-11 00:22:57 +0000[diff] [blame]720 <dt><b>"<tt>cc <em>10</em></tt>" - GHC convention</b>:</dt>
721 <dd>This calling convention has been implemented specifically for use by the
722 <a href="http://www.haskell.org/ghc">Glasgow Haskell Compiler (GHC)</a>.
723 It passes everything in registers, going to extremes to achieve this by
724 disabling callee save registers. This calling convention should not be
725 used lightly but only for specific situations such as an alternative to
726 the <em>register pinning</em> performance technique often used when
727 implementing functional programming languages.At the moment only X86
728 supports this convention and it has the following limitations:
729 <ul>
730 <li>On <em>X86-32</em> only supports up to 4 bit type parameters. No
731 floating point types are supported.</li>
732 <li>On <em>X86-64</em> only supports up to 10 bit type parameters and
733 6 floating point parameters.</li>
734 </ul>
735 This calling convention supports
736 <a href="CodeGenerator.html#tailcallopt">tail call optimization</a> but
737 requires both the caller and callee are using it.
738 </dd>
739
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]740 <dt><b>"<tt>cc &lt;<em>n</em>&gt;</tt>" - Numbered convention</b>:</dt>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]741 <dd>Any calling convention may be specified by number, allowing
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]742 target-specific calling conventions to be used. Target specific calling
743 conventions start at 64.</dd>
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]744</dl>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]745
746<p>More calling conventions can be added/defined on an as-needed basis, to
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]747 support Pascal conventions or any other well-known target-independent
748 convention.</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]749
750</div>
751
752<!-- ======================================================================= -->
753<div class="doc_subsection">
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]754 <a name="visibility">Visibility Styles</a>
755</div>
756
757<div class="doc_text">
758
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]759<p>All Global Variables and Functions have one of the following visibility
760 styles:</p>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]761
762<dl>
763 <dt><b>"<tt>default</tt>" - Default style</b>:</dt>
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]764 <dd>On targets that use the ELF object file format, default visibility means
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]765 that the declaration is visible to other modules and, in shared libraries,
766 means that the declared entity may be overridden. On Darwin, default
767 visibility means that the declaration is visible to other modules. Default
768 visibility corresponds to "external linkage" in the language.</dd>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]769
770 <dt><b>"<tt>hidden</tt>" - Hidden style</b>:</dt>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]771 <dd>Two declarations of an object with hidden visibility refer to the same
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]772 object if they are in the same shared object. Usually, hidden visibility
773 indicates that the symbol will not be placed into the dynamic symbol
774 table, so no other module (executable or shared library) can reference it
775 directly.</dd>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]776
Anton Korobeynikov6f9896f2007-04-29 18:35:00 +0000[diff] [blame]777 <dt><b>"<tt>protected</tt>" - Protected style</b>:</dt>
Anton Korobeynikov6f9896f2007-04-29 18:35:00 +0000[diff] [blame]778 <dd>On ELF, protected visibility indicates that the symbol will be placed in
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]779 the dynamic symbol table, but that references within the defining module
780 will bind to the local symbol. That is, the symbol cannot be overridden by
781 another module.</dd>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]782</dl>
783
784</div>
785
786<!-- ======================================================================= -->
787<div class="doc_subsection">
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]788 <a name="namedtypes">Named Types</a>
789</div>
790
791<div class="doc_text">
792
793<p>LLVM IR allows you to specify name aliases for certain types. This can make
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]794 it easier to read the IR and make the IR more condensed (particularly when
795 recursive types are involved). An example of a name specification is:</p>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]796
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]797<pre class="doc_code">
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]798%mytype = type { %mytype*, i32 }
799</pre>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]800
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]801<p>You may give a name to any <a href="#typesystem">type</a> except
Chris Lattnerdc65f222010-08-17 23:26:04 +0000[diff] [blame]802 "<a href="#t_void">void</a>". Type name aliases may be used anywhere a type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]803 is expected with the syntax "%mytype".</p>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]804
805<p>Note that type names are aliases for the structural type that they indicate,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]806 and that you can therefore specify multiple names for the same type. This
807 often leads to confusing behavior when dumping out a .ll file. Since LLVM IR
808 uses structural typing, the name is not part of the type. When printing out
809 LLVM IR, the printer will pick <em>one name</em> to render all types of a
810 particular shape. This means that if you have code where two different
811 source types end up having the same LLVM type, that the dumper will sometimes
812 print the "wrong" or unexpected type. This is an important design point and
813 isn't going to change.</p>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]814
815</div>
816
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]817<!-- ======================================================================= -->
818<div class="doc_subsection">
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]819 <a name="globalvars">Global Variables</a>
820</div>
821
822<div class="doc_text">
823
Chris Lattner3689a342005-02-12 19:30:21 +0000[diff] [blame]824<p>Global variables define regions of memory allocated at compilation time
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]825 instead of run-time. Global variables may optionally be initialized, may
826 have an explicit section to be placed in, and may have an optional explicit
827 alignment specified. A variable may be defined as "thread_local", which
828 means that it will not be shared by threads (each thread will have a
829 separated copy of the variable). A variable may be defined as a global
830 "constant," which indicates that the contents of the variable
831 will <b>never</b> be modified (enabling better optimization, allowing the
832 global data to be placed in the read-only section of an executable, etc).
833 Note that variables that need runtime initialization cannot be marked
834 "constant" as there is a store to the variable.</p>
Chris Lattner3689a342005-02-12 19:30:21 +0000[diff] [blame]835
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]836<p>LLVM explicitly allows <em>declarations</em> of global variables to be marked
837 constant, even if the final definition of the global is not. This capability
838 can be used to enable slightly better optimization of the program, but
839 requires the language definition to guarantee that optimizations based on the
840 'constantness' are valid for the translation units that do not include the
841 definition.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]842
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]843<p>As SSA values, global variables define pointer values that are in scope
844 (i.e. they dominate) all basic blocks in the program. Global variables
845 always define a pointer to their "content" type because they describe a
846 region of memory, and all memory objects in LLVM are accessed through
847 pointers.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]848
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]849<p>A global variable may be declared to reside in a target-specific numbered
850 address space. For targets that support them, address spaces may affect how
851 optimizations are performed and/or what target instructions are used to
852 access the variable. The default address space is zero. The address space
853 qualifier must precede any other attributes.</p>
Christopher Lamb284d9922007-12-11 09:31:00 +0000[diff] [blame]854
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]855<p>LLVM allows an explicit section to be specified for globals. If the target
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]856 supports it, it will emit globals to the section specified.</p>
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]857
Chris Lattnerce99fa92010-04-28 00:13:42 +0000[diff] [blame]858<p>An explicit alignment may be specified for a global, which must be a power
859 of 2. If not present, or if the alignment is set to zero, the alignment of
860 the global is set by the target to whatever it feels convenient. If an
861 explicit alignment is specified, the global is forced to have exactly that
Chris Lattner2d4b8ee2010-04-28 00:31:12 +0000[diff] [blame]862 alignment. Targets and optimizers are not allowed to over-align the global
863 if the global has an assigned section. In this case, the extra alignment
864 could be observable: for example, code could assume that the globals are
865 densely packed in their section and try to iterate over them as an array,
866 alignment padding would break this iteration.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]867
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]868<p>For example, the following defines a global in a numbered address space with
869 an initializer, section, and alignment:</p>
Chris Lattner68027ea2007-01-14 00:27:09 +0000[diff] [blame]870
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]871<pre class="doc_code">
Dan Gohman398873c2009-01-11 00:40:00 +0000[diff] [blame]872@G = addrspace(5) constant float 1.0, section "foo", align 4
Chris Lattner68027ea2007-01-14 00:27:09 +0000[diff] [blame]873</pre>
874
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]875</div>
876
877
878<!-- ======================================================================= -->
879<div class="doc_subsection">
880 <a name="functionstructure">Functions</a>
881</div>
882
883<div class="doc_text">
884
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]885<p>LLVM function definitions consist of the "<tt>define</tt>" keyword, an
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]886 optional <a href="#linkage">linkage type</a>, an optional
887 <a href="#visibility">visibility style</a>, an optional
888 <a href="#callingconv">calling convention</a>, a return type, an optional
889 <a href="#paramattrs">parameter attribute</a> for the return type, a function
890 name, a (possibly empty) argument list (each with optional
891 <a href="#paramattrs">parameter attributes</a>), optional
892 <a href="#fnattrs">function attributes</a>, an optional section, an optional
893 alignment, an optional <a href="#gc">garbage collector name</a>, an opening
894 curly brace, a list of basic blocks, and a closing curly brace.</p>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]895
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]896<p>LLVM function declarations consist of the "<tt>declare</tt>" keyword, an
897 optional <a href="#linkage">linkage type</a>, an optional
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]898 <a href="#visibility">visibility style</a>, an optional
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]899 <a href="#callingconv">calling convention</a>, a return type, an optional
900 <a href="#paramattrs">parameter attribute</a> for the return type, a function
901 name, a possibly empty list of arguments, an optional alignment, and an
902 optional <a href="#gc">garbage collector name</a>.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]903
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]904<p>A function definition contains a list of basic blocks, forming the CFG
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]905 (Control Flow Graph) for the function. Each basic block may optionally start
906 with a label (giving the basic block a symbol table entry), contains a list
907 of instructions, and ends with a <a href="#terminators">terminator</a>
908 instruction (such as a branch or function return).</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]909
Chris Lattner4a3c9012007-06-08 16:52:14 +0000[diff] [blame]910<p>The first basic block in a function is special in two ways: it is immediately
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]911 executed on entrance to the function, and it is not allowed to have
912 predecessor basic blocks (i.e. there can not be any branches to the entry
913 block of a function). Because the block can have no predecessors, it also
914 cannot have any <a href="#i_phi">PHI nodes</a>.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]915
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]916<p>LLVM allows an explicit section to be specified for functions. If the target
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]917 supports it, it will emit functions to the section specified.</p>
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]918
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]919<p>An explicit alignment may be specified for a function. If not present, or if
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]920 the alignment is set to zero, the alignment of the function is set by the
921 target to whatever it feels convenient. If an explicit alignment is
922 specified, the function is forced to have at least that much alignment. All
923 alignments must be a power of 2.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]924
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]925<h5>Syntax:</h5>
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]926<pre class="doc_code">
Chris Lattner50ad45c2008-10-13 16:55:18 +0000[diff] [blame]927define [<a href="#linkage">linkage</a>] [<a href="#visibility">visibility</a>]
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]928 [<a href="#callingconv">cconv</a>] [<a href="#paramattrs">ret attrs</a>]
929 &lt;ResultType&gt; @&lt;FunctionName&gt; ([argument list])
930 [<a href="#fnattrs">fn Attrs</a>] [section "name"] [align N]
931 [<a href="#gc">gc</a>] { ... }
932</pre>
Devang Patel307e8ab2008-10-07 17:48:33 +0000[diff] [blame]933
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]934</div>
935
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]936<!-- ======================================================================= -->
937<div class="doc_subsection">
938 <a name="aliasstructure">Aliases</a>
939</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]940
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]941<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]942
943<p>Aliases act as "second name" for the aliasee value (which can be either
944 function, global variable, another alias or bitcast of global value). Aliases
945 may have an optional <a href="#linkage">linkage type</a>, and an
946 optional <a href="#visibility">visibility style</a>.</p>
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]947
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]948<h5>Syntax:</h5>
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]949<pre class="doc_code">
Duncan Sands0b23ac12008-09-12 20:48:21 +0000[diff] [blame]950@&lt;Name&gt; = alias [Linkage] [Visibility] &lt;AliaseeTy&gt; @&lt;Aliasee&gt;
Bill Wendlingaac388b2007-05-29 09:42:13 +0000[diff] [blame]951</pre>
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]952
953</div>
954
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]955<!-- ======================================================================= -->
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]956<div class="doc_subsection">
957 <a name="namedmetadatastructure">Named Metadata</a>
958</div>
959
960<div class="doc_text">
961
Chris Lattnere6a5ddd2010-01-15 21:50:19 +0000[diff] [blame]962<p>Named metadata is a collection of metadata. <a href="#metadata">Metadata
Dan Gohman872814a2010-07-21 18:54:18 +0000[diff] [blame]963 nodes</a> (but not metadata strings) are the only valid operands for
Chris Lattnere6a5ddd2010-01-15 21:50:19 +0000[diff] [blame]964 a named metadata.</p>
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]965
966<h5>Syntax:</h5>
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]967<pre class="doc_code">
Dan Gohman872814a2010-07-21 18:54:18 +0000[diff] [blame]968; Some unnamed metadata nodes, which are referenced by the named metadata.
969!0 = metadata !{metadata !"zero"}
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]970!1 = metadata !{metadata !"one"}
Dan Gohman872814a2010-07-21 18:54:18 +0000[diff] [blame]971!2 = metadata !{metadata !"two"}
Dan Gohman1005bc52010-07-13 19:48:13 +0000[diff] [blame]972; A named metadata.
Dan Gohman872814a2010-07-21 18:54:18 +0000[diff] [blame]973!name = !{!0, !1, !2}
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]974</pre>
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]975
976</div>
977
978<!-- ======================================================================= -->
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]979<div class="doc_subsection"><a name="paramattrs">Parameter Attributes</a></div>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]980
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]981<div class="doc_text">
982
983<p>The return type and each parameter of a function type may have a set of
984 <i>parameter attributes</i> associated with them. Parameter attributes are
985 used to communicate additional information about the result or parameters of
986 a function. Parameter attributes are considered to be part of the function,
987 not of the function type, so functions with different parameter attributes
988 can have the same function type.</p>
989
990<p>Parameter attributes are simple keywords that follow the type specified. If
991 multiple parameter attributes are needed, they are space separated. For
992 example:</p>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]993
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]994<pre class="doc_code">
Nick Lewyckyb6a7d252009-02-15 23:06:14 +0000[diff] [blame]995declare i32 @printf(i8* noalias nocapture, ...)
Chris Lattner66d922c2008-10-04 18:33:34 +0000[diff] [blame]996declare i32 @atoi(i8 zeroext)
997declare signext i8 @returns_signed_char()
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]998</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]999
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1000<p>Note that any attributes for the function result (<tt>nounwind</tt>,
1001 <tt>readonly</tt>) come immediately after the argument list.</p>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1002
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1003<p>Currently, only the following parameter attributes are defined:</p>
Chris Lattner47507de2008-01-11 06:20:47 +0000[diff] [blame]1004
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1005<dl>
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1006 <dt><tt><b>zeroext</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1007 <dd>This indicates to the code generator that the parameter or return value
1008 should be zero-extended to a 32-bit value by the caller (for a parameter)
1009 or the callee (for a return value).</dd>
Chris Lattner47507de2008-01-11 06:20:47 +0000[diff] [blame]1010
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1011 <dt><tt><b>signext</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1012 <dd>This indicates to the code generator that the parameter or return value
1013 should be sign-extended to a 32-bit value by the caller (for a parameter)
1014 or the callee (for a return value).</dd>
Chris Lattner47507de2008-01-11 06:20:47 +0000[diff] [blame]1015
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1016 <dt><tt><b>inreg</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1017 <dd>This indicates that this parameter or return value should be treated in a
1018 special target-dependent fashion during while emitting code for a function
1019 call or return (usually, by putting it in a register as opposed to memory,
1020 though some targets use it to distinguish between two different kinds of
1021 registers). Use of this attribute is target-specific.</dd>
Chris Lattner47507de2008-01-11 06:20:47 +0000[diff] [blame]1022
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1023 <dt><tt><b><a name="byval">byval</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1024 <dd>This indicates that the pointer parameter should really be passed by value
1025 to the function. The attribute implies that a hidden copy of the pointee
1026 is made between the caller and the callee, so the callee is unable to
1027 modify the value in the callee. This attribute is only valid on LLVM
1028 pointer arguments. It is generally used to pass structs and arrays by
1029 value, but is also valid on pointers to scalars. The copy is considered
1030 to belong to the caller not the callee (for example,
1031 <tt><a href="#readonly">readonly</a></tt> functions should not write to
1032 <tt>byval</tt> parameters). This is not a valid attribute for return
1033 values. The byval attribute also supports specifying an alignment with
1034 the align attribute. This has a target-specific effect on the code
1035 generator that usually indicates a desired alignment for the synthesized
1036 stack slot.</dd>
1037
Dan Gohmanff235352010-07-02 23:18:08 +0000[diff] [blame]1038 <dt><tt><b><a name="sret">sret</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1039 <dd>This indicates that the pointer parameter specifies the address of a
1040 structure that is the return value of the function in the source program.
1041 This pointer must be guaranteed by the caller to be valid: loads and
1042 stores to the structure may be assumed by the callee to not to trap. This
1043 may only be applied to the first parameter. This is not a valid attribute
1044 for return values. </dd>
1045
Dan Gohmanff235352010-07-02 23:18:08 +0000[diff] [blame]1046 <dt><tt><b><a name="noalias">noalias</a></b></tt></dt>
Dan Gohman1e109622010-07-02 18:41:32 +0000[diff] [blame]1047 <dd>This indicates that pointer values
1048 <a href="#pointeraliasing"><i>based</i></a> on the argument or return
Dan Gohmanefca7f92010-07-02 23:46:54 +0000[diff] [blame]1049 value do not alias pointer values which are not <i>based</i> on it,
1050 ignoring certain "irrelevant" dependencies.
1051 For a call to the parent function, dependencies between memory
1052 references from before or after the call and from those during the call
1053 are "irrelevant" to the <tt>noalias</tt> keyword for the arguments and
1054 return value used in that call.
Dan Gohman1e109622010-07-02 18:41:32 +0000[diff] [blame]1055 The caller shares the responsibility with the callee for ensuring that
1056 these requirements are met.
1057 For further details, please see the discussion of the NoAlias response in
Dan Gohmanff70fe42010-07-06 15:26:33 +0000[diff] [blame]1058 <a href="AliasAnalysis.html#MustMayNo">alias analysis</a>.<br>
1059<br>
John McCall191d4ee2010-07-06 21:07:14 +0000[diff] [blame]1060 Note that this definition of <tt>noalias</tt> is intentionally
1061 similar to the definition of <tt>restrict</tt> in C99 for function
Chris Lattner211244a2010-07-06 20:51:35 +0000[diff] [blame]1062 arguments, though it is slightly weaker.
Dan Gohmanff70fe42010-07-06 15:26:33 +0000[diff] [blame]1063<br>
1064 For function return values, C99's <tt>restrict</tt> is not meaningful,
1065 while LLVM's <tt>noalias</tt> is.
1066 </dd>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1067
Dan Gohmanff235352010-07-02 23:18:08 +0000[diff] [blame]1068 <dt><tt><b><a name="nocapture">nocapture</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1069 <dd>This indicates that the callee does not make any copies of the pointer
1070 that outlive the callee itself. This is not a valid attribute for return
1071 values.</dd>
1072
Dan Gohmanff235352010-07-02 23:18:08 +0000[diff] [blame]1073 <dt><tt><b><a name="nest">nest</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1074 <dd>This indicates that the pointer parameter can be excised using the
1075 <a href="#int_trampoline">trampoline intrinsics</a>. This is not a valid
1076 attribute for return values.</dd>
1077</dl>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1078
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1079</div>
1080
1081<!-- ======================================================================= -->
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]1082<div class="doc_subsection">
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame]1083 <a name="gc">Garbage Collector Names</a>
1084</div>
1085
1086<div class="doc_text">
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame]1087
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1088<p>Each function may specify a garbage collector name, which is simply a
1089 string:</p>
1090
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]1091<pre class="doc_code">
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1092define void @f() gc "name" { ... }
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1093</pre>
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame]1094
1095<p>The compiler declares the supported values of <i>name</i>. Specifying a
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1096 collector which will cause the compiler to alter its output in order to
1097 support the named garbage collection algorithm.</p>
1098
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame]1099</div>
1100
1101<!-- ======================================================================= -->
1102<div class="doc_subsection">
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1103 <a name="fnattrs">Function Attributes</a>
Devang Patelf8b94812008-09-04 23:05:13 +0000[diff] [blame]1104</div>
1105
1106<div class="doc_text">
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1107
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1108<p>Function attributes are set to communicate additional information about a
1109 function. Function attributes are considered to be part of the function, not
1110 of the function type, so functions with different parameter attributes can
1111 have the same function type.</p>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1112
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1113<p>Function attributes are simple keywords that follow the type specified. If
1114 multiple attributes are needed, they are space separated. For example:</p>
Devang Patelf8b94812008-09-04 23:05:13 +0000[diff] [blame]1115
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]1116<pre class="doc_code">
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1117define void @f() noinline { ... }
1118define void @f() alwaysinline { ... }
1119define void @f() alwaysinline optsize { ... }
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1120define void @f() optsize { ... }
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1121</pre>
Devang Patelf8b94812008-09-04 23:05:13 +0000[diff] [blame]1122
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1123<dl>
Charles Davis1e063d12010-02-12 00:31:15 +0000[diff] [blame]1124 <dt><tt><b>alignstack(&lt;<em>n</em>&gt;)</b></tt></dt>
1125 <dd>This attribute indicates that, when emitting the prologue and epilogue,
1126 the backend should forcibly align the stack pointer. Specify the
1127 desired alignment, which must be a power of two, in parentheses.
1128
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1129 <dt><tt><b>alwaysinline</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1130 <dd>This attribute indicates that the inliner should attempt to inline this
1131 function into callers whenever possible, ignoring any active inlining size
1132 threshold for this caller.</dd>
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1133
Jakob Stoklund Olesen570a4a52010-02-06 01:16:28 +0000[diff] [blame]1134 <dt><tt><b>inlinehint</b></tt></dt>
1135 <dd>This attribute indicates that the source code contained a hint that inlining
1136 this function is desirable (such as the "inline" keyword in C/C++). It
1137 is just a hint; it imposes no requirements on the inliner.</dd>
1138
Nick Lewycky76ec37a2010-07-06 18:24:09 +0000[diff] [blame]1139 <dt><tt><b>naked</b></tt></dt>
1140 <dd>This attribute disables prologue / epilogue emission for the function.
1141 This can have very system-specific consequences.</dd>
1142
1143 <dt><tt><b>noimplicitfloat</b></tt></dt>
1144 <dd>This attributes disables implicit floating point instructions.</dd>
1145
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1146 <dt><tt><b>noinline</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1147 <dd>This attribute indicates that the inliner should never inline this
1148 function in any situation. This attribute may not be used together with
1149 the <tt>alwaysinline</tt> attribute.</dd>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1150
Nick Lewycky76ec37a2010-07-06 18:24:09 +0000[diff] [blame]1151 <dt><tt><b>noredzone</b></tt></dt>
1152 <dd>This attribute indicates that the code generator should not use a red
1153 zone, even if the target-specific ABI normally permits it.</dd>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1154
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1155 <dt><tt><b>noreturn</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1156 <dd>This function attribute indicates that the function never returns
1157 normally. This produces undefined behavior at runtime if the function
1158 ever does dynamically return.</dd>
Bill Wendling31359ba2008-11-13 01:02:51 +0000[diff] [blame]1159
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1160 <dt><tt><b>nounwind</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1161 <dd>This function attribute indicates that the function never returns with an
1162 unwind or exceptional control flow. If the function does unwind, its
1163 runtime behavior is undefined.</dd>
Bill Wendlingfbaa7ed2008-11-26 19:07:40 +0000[diff] [blame]1164
Nick Lewycky76ec37a2010-07-06 18:24:09 +0000[diff] [blame]1165 <dt><tt><b>optsize</b></tt></dt>
1166 <dd>This attribute suggests that optimization passes and code generator passes
1167 make choices that keep the code size of this function low, and otherwise
1168 do optimizations specifically to reduce code size.</dd>
1169
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1170 <dt><tt><b>readnone</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1171 <dd>This attribute indicates that the function computes its result (or decides
1172 to unwind an exception) based strictly on its arguments, without
1173 dereferencing any pointer arguments or otherwise accessing any mutable
1174 state (e.g. memory, control registers, etc) visible to caller functions.
1175 It does not write through any pointer arguments
1176 (including <tt><a href="#byval">byval</a></tt> arguments) and never
1177 changes any state visible to callers. This means that it cannot unwind
1178 exceptions by calling the <tt>C++</tt> exception throwing methods, but
1179 could use the <tt>unwind</tt> instruction.</dd>
Devang Patel5d96fda2009-06-12 19:45:19 +0000[diff] [blame]1180
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1181 <dt><tt><b><a name="readonly">readonly</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1182 <dd>This attribute indicates that the function does not write through any
1183 pointer arguments (including <tt><a href="#byval">byval</a></tt>
1184 arguments) or otherwise modify any state (e.g. memory, control registers,
1185 etc) visible to caller functions. It may dereference pointer arguments
1186 and read state that may be set in the caller. A readonly function always
1187 returns the same value (or unwinds an exception identically) when called
1188 with the same set of arguments and global state. It cannot unwind an
1189 exception by calling the <tt>C++</tt> exception throwing methods, but may
1190 use the <tt>unwind</tt> instruction.</dd>
Anton Korobeynikovc5ec8a72009-07-17 18:07:26 +0000[diff] [blame]1191
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1192 <dt><tt><b><a name="ssp">ssp</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1193 <dd>This attribute indicates that the function should emit a stack smashing
1194 protector. It is in the form of a "canary"&mdash;a random value placed on
1195 the stack before the local variables that's checked upon return from the
1196 function to see if it has been overwritten. A heuristic is used to
1197 determine if a function needs stack protectors or not.<br>
1198<br>
1199 If a function that has an <tt>ssp</tt> attribute is inlined into a
1200 function that doesn't have an <tt>ssp</tt> attribute, then the resulting
1201 function will have an <tt>ssp</tt> attribute.</dd>
1202
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1203 <dt><tt><b>sspreq</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1204 <dd>This attribute indicates that the function should <em>always</em> emit a
1205 stack smashing protector. This overrides
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1206 the <tt><a href="#ssp">ssp</a></tt> function attribute.<br>
1207<br>
1208 If a function that has an <tt>sspreq</tt> attribute is inlined into a
1209 function that doesn't have an <tt>sspreq</tt> attribute or which has
1210 an <tt>ssp</tt> attribute, then the resulting function will have
1211 an <tt>sspreq</tt> attribute.</dd>
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1212</dl>
1213
Devang Patelf8b94812008-09-04 23:05:13 +0000[diff] [blame]1214</div>
1215
1216<!-- ======================================================================= -->
1217<div class="doc_subsection">
Chris Lattner1eeeb0c2006-04-08 04:40:53 +0000[diff] [blame]1218 <a name="moduleasm">Module-Level Inline Assembly</a>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]1219</div>
1220
1221<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1222
1223<p>Modules may contain "module-level inline asm" blocks, which corresponds to
1224 the GCC "file scope inline asm" blocks. These blocks are internally
1225 concatenated by LLVM and treated as a single unit, but may be separated in
1226 the <tt>.ll</tt> file if desired. The syntax is very simple:</p>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]1227
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]1228<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1229module asm "inline asm code goes here"
1230module asm "more can go here"
1231</pre>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]1232
1233<p>The strings can contain any character by escaping non-printable characters.
1234 The escape sequence used is simply "\xx" where "xx" is the two digit hex code
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1235 for the number.</p>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]1236
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1237<p>The inline asm code is simply printed to the machine code .s file when
1238 assembly code is generated.</p>
1239
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]1240</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]1241
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1242<!-- ======================================================================= -->
1243<div class="doc_subsection">
1244 <a name="datalayout">Data Layout</a>
1245</div>
1246
1247<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1248
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1249<p>A module may specify a target specific data layout string that specifies how
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1250 data is to be laid out in memory. The syntax for the data layout is
1251 simply:</p>
1252
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]1253<pre class="doc_code">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1254target datalayout = "<i>layout specification</i>"
1255</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1256
1257<p>The <i>layout specification</i> consists of a list of specifications
1258 separated by the minus sign character ('-'). Each specification starts with
1259 a letter and may include other information after the letter to define some
1260 aspect of the data layout. The specifications accepted are as follows:</p>
1261
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1262<dl>
1263 <dt><tt>E</tt></dt>
1264 <dd>Specifies that the target lays out data in big-endian form. That is, the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1265 bits with the most significance have the lowest address location.</dd>
1266
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1267 <dt><tt>e</tt></dt>
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]1268 <dd>Specifies that the target lays out data in little-endian form. That is,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1269 the bits with the least significance have the lowest address
1270 location.</dd>
1271
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1272 <dt><tt>p:<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]1273 <dd>This specifies the <i>size</i> of a pointer and its <i>abi</i> and
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1274 <i>preferred</i> alignments. All sizes are in bits. Specifying
1275 the <i>pref</i> alignment is optional. If omitted, the
1276 preceding <tt>:</tt> should be omitted too.</dd>
1277
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1278 <dt><tt>i<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1279 <dd>This specifies the alignment for an integer type of a given bit
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1280 <i>size</i>. The value of <i>size</i> must be in the range [1,2^23).</dd>
1281
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1282 <dt><tt>v<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]1283 <dd>This specifies the alignment for a vector type of a given bit
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1284 <i>size</i>.</dd>
1285
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1286 <dt><tt>f<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]1287 <dd>This specifies the alignment for a floating point type of a given bit
Dale Johannesen9d8d2212010-05-28 18:54:47 +0000[diff] [blame]1288 <i>size</i>. Only values of <i>size</i> that are supported by the target
1289 will work. 32 (float) and 64 (double) are supported on all targets;
1290 80 or 128 (different flavors of long double) are also supported on some
1291 targets.
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1292
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1293 <dt><tt>a<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1294 <dd>This specifies the alignment for an aggregate type of a given bit
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1295 <i>size</i>.</dd>
1296
Daniel Dunbar87bde0b2009-06-08 22:17:53 +0000[diff] [blame]1297 <dt><tt>s<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1298 <dd>This specifies the alignment for a stack object of a given bit
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1299 <i>size</i>.</dd>
Chris Lattnere82bdc42009-11-07 09:35:34 +0000[diff] [blame]1300
1301 <dt><tt>n<i>size1</i>:<i>size2</i>:<i>size3</i>...</tt></dt>
1302 <dd>This specifies a set of native integer widths for the target CPU
1303 in bits. For example, it might contain "n32" for 32-bit PowerPC,
1304 "n32:64" for PowerPC 64, or "n8:16:32:64" for X86-64. Elements of
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]1305 this set are considered to support most general arithmetic
Chris Lattnere82bdc42009-11-07 09:35:34 +0000[diff] [blame]1306 operations efficiently.</dd>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1307</dl>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1308
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1309<p>When constructing the data layout for a given target, LLVM starts with a
Dan Gohman1c70c002010-04-28 00:36:01 +0000[diff] [blame]1310 default set of specifications which are then (possibly) overridden by the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1311 specifications in the <tt>datalayout</tt> keyword. The default specifications
1312 are given in this list:</p>
1313
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1314<ul>
1315 <li><tt>E</tt> - big endian</li>
Dan Gohmanfdf2e8c2010-02-23 02:44:03 +0000[diff] [blame]1316 <li><tt>p:64:64:64</tt> - 64-bit pointers with 64-bit alignment</li>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1317 <li><tt>i1:8:8</tt> - i1 is 8-bit (byte) aligned</li>
1318 <li><tt>i8:8:8</tt> - i8 is 8-bit (byte) aligned</li>
1319 <li><tt>i16:16:16</tt> - i16 is 16-bit aligned</li>
1320 <li><tt>i32:32:32</tt> - i32 is 32-bit aligned</li>
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]1321 <li><tt>i64:32:64</tt> - i64 has ABI alignment of 32-bits but preferred
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1322 alignment of 64-bits</li>
1323 <li><tt>f32:32:32</tt> - float is 32-bit aligned</li>
1324 <li><tt>f64:64:64</tt> - double is 64-bit aligned</li>
1325 <li><tt>v64:64:64</tt> - 64-bit vector is 64-bit aligned</li>
1326 <li><tt>v128:128:128</tt> - 128-bit vector is 128-bit aligned</li>
1327 <li><tt>a0:0:1</tt> - aggregates are 8-bit aligned</li>
Daniel Dunbar87bde0b2009-06-08 22:17:53 +0000[diff] [blame]1328 <li><tt>s0:64:64</tt> - stack objects are 64-bit aligned</li>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1329</ul>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1330
1331<p>When LLVM is determining the alignment for a given type, it uses the
1332 following rules:</p>
1333
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1334<ol>
1335 <li>If the type sought is an exact match for one of the specifications, that
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1336 specification is used.</li>
1337
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1338 <li>If no match is found, and the type sought is an integer type, then the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1339 smallest integer type that is larger than the bitwidth of the sought type
1340 is used. If none of the specifications are larger than the bitwidth then
1341 the the largest integer type is used. For example, given the default
1342 specifications above, the i7 type will use the alignment of i8 (next
1343 largest) while both i65 and i256 will use the alignment of i64 (largest
1344 specified).</li>
1345
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1346 <li>If no match is found, and the type sought is a vector type, then the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1347 largest vector type that is smaller than the sought vector type will be
1348 used as a fall back. This happens because &lt;128 x double&gt; can be
1349 implemented in terms of 64 &lt;2 x double&gt;, for example.</li>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1350</ol>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1351
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1352</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]1353
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1354<!-- ======================================================================= -->
1355<div class="doc_subsection">
1356 <a name="pointeraliasing">Pointer Aliasing Rules</a>
1357</div>
1358
1359<div class="doc_text">
1360
Andreas Bolka55e459a2009-07-29 00:02:05 +0000[diff] [blame]1361<p>Any memory access must be done through a pointer value associated
Andreas Bolka99a82052009-07-27 20:37:10 +0000[diff] [blame]1362with an address range of the memory access, otherwise the behavior
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1363is undefined. Pointer values are associated with address ranges
1364according to the following rules:</p>
1365
1366<ul>
Dan Gohman1e109622010-07-02 18:41:32 +0000[diff] [blame]1367 <li>A pointer value is associated with the addresses associated with
1368 any value it is <i>based</i> on.
Andreas Bolka55e459a2009-07-29 00:02:05 +0000[diff] [blame]1369 <li>An address of a global variable is associated with the address
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1370 range of the variable's storage.</li>
1371 <li>The result value of an allocation instruction is associated with
1372 the address range of the allocated storage.</li>
1373 <li>A null pointer in the default address-space is associated with
Andreas Bolka55e459a2009-07-29 00:02:05 +0000[diff] [blame]1374 no address.</li>
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1375 <li>An integer constant other than zero or a pointer value returned
1376 from a function not defined within LLVM may be associated with address
1377 ranges allocated through mechanisms other than those provided by
Andreas Bolka55e459a2009-07-29 00:02:05 +0000[diff] [blame]1378 LLVM. Such ranges shall not overlap with any ranges of addresses
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1379 allocated by mechanisms provided by LLVM.</li>
Dan Gohman1e109622010-07-02 18:41:32 +0000[diff] [blame]1380</ul>
1381
1382<p>A pointer value is <i>based</i> on another pointer value according
1383 to the following rules:</p>
1384
1385<ul>
1386 <li>A pointer value formed from a
1387 <tt><a href="#i_getelementptr">getelementptr</a></tt> operation
1388 is <i>based</i> on the first operand of the <tt>getelementptr</tt>.</li>
1389 <li>The result value of a
1390 <tt><a href="#i_bitcast">bitcast</a></tt> is <i>based</i> on the operand
1391 of the <tt>bitcast</tt>.</li>
1392 <li>A pointer value formed by an
1393 <tt><a href="#i_inttoptr">inttoptr</a></tt> is <i>based</i> on all
1394 pointer values that contribute (directly or indirectly) to the
1395 computation of the pointer's value.</li>
1396 <li>The "<i>based</i> on" relationship is transitive.</li>
1397</ul>
1398
1399<p>Note that this definition of <i>"based"</i> is intentionally
1400 similar to the definition of <i>"based"</i> in C99, though it is
1401 slightly weaker.</p>
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1402
1403<p>LLVM IR does not associate types with memory. The result type of a
Andreas Bolka55e459a2009-07-29 00:02:05 +0000[diff] [blame]1404<tt><a href="#i_load">load</a></tt> merely indicates the size and
1405alignment of the memory from which to load, as well as the
Dan Gohmanc22c0f32010-06-17 19:23:50 +0000[diff] [blame]1406interpretation of the value. The first operand type of a
Andreas Bolka55e459a2009-07-29 00:02:05 +0000[diff] [blame]1407<tt><a href="#i_store">store</a></tt> similarly only indicates the size
1408and alignment of the store.</p>
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1409
1410<p>Consequently, type-based alias analysis, aka TBAA, aka
1411<tt>-fstrict-aliasing</tt>, is not applicable to general unadorned
1412LLVM IR. <a href="#metadata">Metadata</a> may be used to encode
1413additional information which specialized optimization passes may use
1414to implement type-based alias analysis.</p>
1415
1416</div>
1417
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]1418<!-- ======================================================================= -->
1419<div class="doc_subsection">
1420 <a name="volatile">Volatile Memory Accesses</a>
1421</div>
1422
1423<div class="doc_text">
1424
1425<p>Certain memory accesses, such as <a href="#i_load"><tt>load</tt></a>s, <a
1426href="#i_store"><tt>store</tt></a>s, and <a
1427href="#int_memcpy"><tt>llvm.memcpy</tt></a>s may be marked <tt>volatile</tt>.
1428The optimizers must not change the number of volatile operations or change their
1429order of execution relative to other volatile operations. The optimizers
1430<i>may</i> change the order of volatile operations relative to non-volatile
1431operations. This is not Java's "volatile" and has no cross-thread
1432synchronization behavior.</p>
1433
1434</div>
1435
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1436<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1437<div class="doc_section"> <a name="typesystem">Type System</a> </div>
1438<!-- *********************************************************************** -->
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]1439
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1440<div class="doc_text">
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]1441
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1442<p>The LLVM type system is one of the most important features of the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1443 intermediate representation. Being typed enables a number of optimizations
1444 to be performed on the intermediate representation directly, without having
1445 to do extra analyses on the side before the transformation. A strong type
1446 system makes it easier to read the generated code and enables novel analyses
1447 and transformations that are not feasible to perform on normal three address
1448 code representations.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]1449
1450</div>
1451
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1452<!-- ======================================================================= -->
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1453<div class="doc_subsection"> <a name="t_classifications">Type
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1454Classifications</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1455
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1456<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1457
1458<p>The types fall into a few useful classifications:</p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1459
1460<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1461 <tbody>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1462 <tr><th>Classification</th><th>Types</th></tr>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1463 <tr>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1464 <td><a href="#t_integer">integer</a></td>
Reid Spencer2b916312007-05-16 18:44:01 +0000[diff] [blame]1465 <td><tt>i1, i2, i3, ... i8, ... i16, ... i32, ... i64, ... </tt></td>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1466 </tr>
1467 <tr>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1468 <td><a href="#t_floating">floating point</a></td>
1469 <td><tt>float, double, x86_fp80, fp128, ppc_fp128</tt></td>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1470 </tr>
1471 <tr>
1472 <td><a name="t_firstclass">first class</a></td>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1473 <td><a href="#t_integer">integer</a>,
1474 <a href="#t_floating">floating point</a>,
1475 <a href="#t_pointer">pointer</a>,
Dan Gohman0066db62008-06-18 18:42:13 +0000[diff] [blame]1476 <a href="#t_vector">vector</a>,
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]1477 <a href="#t_struct">structure</a>,
1478 <a href="#t_array">array</a>,
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1479 <a href="#t_label">label</a>,
1480 <a href="#t_metadata">metadata</a>.
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1481 </td>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1482 </tr>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1483 <tr>
1484 <td><a href="#t_primitive">primitive</a></td>
1485 <td><a href="#t_label">label</a>,
1486 <a href="#t_void">void</a>,
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1487 <a href="#t_floating">floating point</a>,
Dale Johannesen21fe99b2010-10-01 00:48:59 +0000[diff] [blame]1488 <a href="#t_x86mmx">x86mmx</a>,
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1489 <a href="#t_metadata">metadata</a>.</td>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1490 </tr>
1491 <tr>
1492 <td><a href="#t_derived">derived</a></td>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]1493 <td><a href="#t_array">array</a>,
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1494 <a href="#t_function">function</a>,
1495 <a href="#t_pointer">pointer</a>,
1496 <a href="#t_struct">structure</a>,
1497 <a href="#t_pstruct">packed structure</a>,
1498 <a href="#t_vector">vector</a>,
1499 <a href="#t_opaque">opaque</a>.
Dan Gohman01ac1012008-10-14 16:32:04 +0000[diff] [blame]1500 </td>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1501 </tr>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1502 </tbody>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1503</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1504
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1505<p>The <a href="#t_firstclass">first class</a> types are perhaps the most
1506 important. Values of these types are the only ones which can be produced by
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]1507 instructions.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1508
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1509</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1510
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1511<!-- ======================================================================= -->
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1512<div class="doc_subsection"> <a name="t_primitive">Primitive Types</a> </div>
Chris Lattner8f8c7b72008-01-04 04:34:14 +0000[diff] [blame]1513
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1514<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1515
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1516<p>The primitive types are the fundamental building blocks of the LLVM
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1517 system.</p>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1518
Chris Lattner8f8c7b72008-01-04 04:34:14 +0000[diff] [blame]1519</div>
1520
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1521<!-- _______________________________________________________________________ -->
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]1522<div class="doc_subsubsection"> <a name="t_integer">Integer Type</a> </div>
1523
1524<div class="doc_text">
1525
1526<h5>Overview:</h5>
1527<p>The integer type is a very simple type that simply specifies an arbitrary
1528 bit width for the integer type desired. Any bit width from 1 bit to
1529 2<sup>23</sup>-1 (about 8 million) can be specified.</p>
1530
1531<h5>Syntax:</h5>
1532<pre>
1533 iN
1534</pre>
1535
1536<p>The number of bits the integer will occupy is specified by the <tt>N</tt>
1537 value.</p>
1538
1539<h5>Examples:</h5>
1540<table class="layout">
1541 <tr class="layout">
1542 <td class="left"><tt>i1</tt></td>
1543 <td class="left">a single-bit integer.</td>
1544 </tr>
1545 <tr class="layout">
1546 <td class="left"><tt>i32</tt></td>
1547 <td class="left">a 32-bit integer.</td>
1548 </tr>
1549 <tr class="layout">
1550 <td class="left"><tt>i1942652</tt></td>
1551 <td class="left">a really big integer of over 1 million bits.</td>
1552 </tr>
1553</table>
1554
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]1555</div>
1556
1557<!-- _______________________________________________________________________ -->
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1558<div class="doc_subsubsection"> <a name="t_floating">Floating Point Types</a> </div>
1559
1560<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1561
1562<table>
1563 <tbody>
1564 <tr><th>Type</th><th>Description</th></tr>
1565 <tr><td><tt>float</tt></td><td>32-bit floating point value</td></tr>
1566 <tr><td><tt>double</tt></td><td>64-bit floating point value</td></tr>
1567 <tr><td><tt>fp128</tt></td><td>128-bit floating point value (112-bit mantissa)</td></tr>
1568 <tr><td><tt>x86_fp80</tt></td><td>80-bit floating point value (X87)</td></tr>
1569 <tr><td><tt>ppc_fp128</tt></td><td>128-bit floating point value (two 64-bits)</td></tr>
1570 </tbody>
1571</table>
1572
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1573</div>
1574
1575<!-- _______________________________________________________________________ -->
Dale Johannesen21fe99b2010-10-01 00:48:59 +0000[diff] [blame]1576<div class="doc_subsubsection"> <a name="t_x86mmx">X86mmx Type</a> </div>
1577
1578<div class="doc_text">
1579
1580<h5>Overview:</h5>
1581<p>The x86mmx type represents a value held in an MMX register on an x86 machine. The operations allowed on it are quite limited: parameters and return values, load and store, and bitcast. User-specified MMX instructions are represented as intrinsic or asm calls with arguments and/or results of this type. There are no arrays, vectors or constants of this type.</p>
1582
1583<h5>Syntax:</h5>
1584<pre>
Dale Johannesen473a8c82010-10-01 01:07:02 +0000[diff] [blame]1585 x86mmx
Dale Johannesen21fe99b2010-10-01 00:48:59 +0000[diff] [blame]1586</pre>
1587
1588</div>
1589
1590<!-- _______________________________________________________________________ -->
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1591<div class="doc_subsubsection"> <a name="t_void">Void Type</a> </div>
1592
1593<div class="doc_text">
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1594
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1595<h5>Overview:</h5>
1596<p>The void type does not represent any value and has no size.</p>
1597
1598<h5>Syntax:</h5>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1599<pre>
1600 void
1601</pre>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1602
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1603</div>
1604
1605<!-- _______________________________________________________________________ -->
1606<div class="doc_subsubsection"> <a name="t_label">Label Type</a> </div>
1607
1608<div class="doc_text">
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1609
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1610<h5>Overview:</h5>
1611<p>The label type represents code labels.</p>
1612
1613<h5>Syntax:</h5>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1614<pre>
1615 label
1616</pre>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1617
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1618</div>
1619
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1620<!-- _______________________________________________________________________ -->
1621<div class="doc_subsubsection"> <a name="t_metadata">Metadata Type</a> </div>
1622
1623<div class="doc_text">
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1624
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1625<h5>Overview:</h5>
Nick Lewyckyc261df92009-09-27 23:27:42 +0000[diff] [blame]1626<p>The metadata type represents embedded metadata. No derived types may be
1627 created from metadata except for <a href="#t_function">function</a>
1628 arguments.
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1629
1630<h5>Syntax:</h5>
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1631<pre>
1632 metadata
1633</pre>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1634
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1635</div>
1636
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1637
1638<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1639<div class="doc_subsection"> <a name="t_derived">Derived Types</a> </div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1640
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1641<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1642
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1643<p>The real power in LLVM comes from the derived types in the system. This is
1644 what allows a programmer to represent arrays, functions, pointers, and other
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]1645 useful types. Each of these types contain one or more element types which
1646 may be a primitive type, or another derived type. For example, it is
1647 possible to have a two dimensional array, using an array as the element type
1648 of another array.</p>
Dan Gohmand8791e52009-01-24 15:58:40 +0000[diff] [blame]1649
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]1650
1651</div>
1652
1653<!-- _______________________________________________________________________ -->
1654<div class="doc_subsubsection"> <a name="t_aggregate">Aggregate Types</a> </div>
1655
1656<div class="doc_text">
1657
1658<p>Aggregate Types are a subset of derived types that can contain multiple
1659 member types. <a href="#t_array">Arrays</a>,
Chris Lattner61c70e92010-08-28 04:09:24 +0000[diff] [blame]1660 <a href="#t_struct">structs</a>, and <a href="#t_vector">vectors</a> are
1661 aggregate types.</p>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]1662
1663</div>
1664
Reid Spencer2b916312007-05-16 18:44:01 +0000[diff] [blame]1665<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1666<div class="doc_subsubsection"> <a name="t_array">Array Type</a> </div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1667
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1668<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1669
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1670<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1671<p>The array type is a very simple derived type that arranges elements
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1672 sequentially in memory. The array type requires a size (number of elements)
1673 and an underlying data type.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1674
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1675<h5>Syntax:</h5>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1676<pre>
1677 [&lt;# elements&gt; x &lt;elementtype&gt;]
1678</pre>
1679
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1680<p>The number of elements is a constant integer value; <tt>elementtype</tt> may
1681 be any type with a size.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1682
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1683<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1684<table class="layout">
1685 <tr class="layout">
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]1686 <td class="left"><tt>[40 x i32]</tt></td>
1687 <td class="left">Array of 40 32-bit integer values.</td>
1688 </tr>
1689 <tr class="layout">
1690 <td class="left"><tt>[41 x i32]</tt></td>
1691 <td class="left">Array of 41 32-bit integer values.</td>
1692 </tr>
1693 <tr class="layout">
1694 <td class="left"><tt>[4 x i8]</tt></td>
1695 <td class="left">Array of 4 8-bit integer values.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1696 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1697</table>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1698<p>Here are some examples of multidimensional arrays:</p>
1699<table class="layout">
1700 <tr class="layout">
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]1701 <td class="left"><tt>[3 x [4 x i32]]</tt></td>
1702 <td class="left">3x4 array of 32-bit integer values.</td>
1703 </tr>
1704 <tr class="layout">
1705 <td class="left"><tt>[12 x [10 x float]]</tt></td>
1706 <td class="left">12x10 array of single precision floating point values.</td>
1707 </tr>
1708 <tr class="layout">
1709 <td class="left"><tt>[2 x [3 x [4 x i16]]]</tt></td>
1710 <td class="left">2x3x4 array of 16-bit integer values.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1711 </tr>
1712</table>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]1713
Dan Gohman7657f6b2009-11-09 19:01:53 +0000[diff] [blame]1714<p>There is no restriction on indexing beyond the end of the array implied by
1715 a static type (though there are restrictions on indexing beyond the bounds
1716 of an allocated object in some cases). This means that single-dimension
1717 'variable sized array' addressing can be implemented in LLVM with a zero
1718 length array type. An implementation of 'pascal style arrays' in LLVM could
1719 use the type "<tt>{ i32, [0 x float]}</tt>", for example.</p>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]1720
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1721</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1722
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1723<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1724<div class="doc_subsubsection"> <a name="t_function">Function Type</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1725
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1726<div class="doc_text">
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]1727
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1728<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1729<p>The function type can be thought of as a function signature. It consists of
1730 a return type and a list of formal parameter types. The return type of a
Chris Lattner61c70e92010-08-28 04:09:24 +0000[diff] [blame]1731 function type is a first class type or a void type.</p>
Devang Patelc3fc6df2008-03-10 20:49:15 +0000[diff] [blame]1732
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1733<h5>Syntax:</h5>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]1734<pre>
Nick Lewycky51386942009-09-27 07:55:32 +0000[diff] [blame]1735 &lt;returntype&gt; (&lt;parameter list&gt;)
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]1736</pre>
1737
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]1738<p>...where '<tt>&lt;parameter list&gt;</tt>' is a comma-separated list of type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1739 specifiers. Optionally, the parameter list may include a type <tt>...</tt>,
1740 which indicates that the function takes a variable number of arguments.
1741 Variable argument functions can access their arguments with
1742 the <a href="#int_varargs">variable argument handling intrinsic</a>
Chris Lattner0724fbd2010-03-02 06:36:51 +0000[diff] [blame]1743 functions. '<tt>&lt;returntype&gt;</tt>' is any type except
Nick Lewyckyc261df92009-09-27 23:27:42 +0000[diff] [blame]1744 <a href="#t_label">label</a>.</p>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]1745
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1746<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1747<table class="layout">
1748 <tr class="layout">
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]1749 <td class="left"><tt>i32 (i32)</tt></td>
1750 <td class="left">function taking an <tt>i32</tt>, returning an <tt>i32</tt>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1751 </td>
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]1752 </tr><tr class="layout">
Chris Lattner0724fbd2010-03-02 06:36:51 +0000[diff] [blame]1753 <td class="left"><tt>float&nbsp;(i16,&nbsp;i32&nbsp;*)&nbsp;*
Reid Spencerf17a0b72006-12-31 07:20:23 +0000[diff] [blame]1754 </tt></td>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]1755 <td class="left"><a href="#t_pointer">Pointer</a> to a function that takes
Chris Lattner0724fbd2010-03-02 06:36:51 +0000[diff] [blame]1756 an <tt>i16</tt> and a <a href="#t_pointer">pointer</a> to <tt>i32</tt>,
1757 returning <tt>float</tt>.
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]1758 </td>
1759 </tr><tr class="layout">
1760 <td class="left"><tt>i32 (i8*, ...)</tt></td>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]1761 <td class="left">A vararg function that takes at least one
1762 <a href="#t_pointer">pointer</a> to <tt>i8 </tt> (char in C),
1763 which returns an integer. This is the signature for <tt>printf</tt> in
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]1764 LLVM.
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1765 </td>
Devang Patela582f402008-03-24 05:35:41 +0000[diff] [blame]1766 </tr><tr class="layout">
1767 <td class="left"><tt>{i32, i32} (i32)</tt></td>
Nick Lewycky51386942009-09-27 07:55:32 +0000[diff] [blame]1768 <td class="left">A function taking an <tt>i32</tt>, returning a
1769 <a href="#t_struct">structure</a> containing two <tt>i32</tt> values
Devang Patela582f402008-03-24 05:35:41 +0000[diff] [blame]1770 </td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1771 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1772</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1773
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1774</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1775
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1776<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1777<div class="doc_subsubsection"> <a name="t_struct">Structure Type</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1778
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1779<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1780
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1781<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1782<p>The structure type is used to represent a collection of data members together
1783 in memory. The packing of the field types is defined to match the ABI of the
1784 underlying processor. The elements of a structure may be any type that has a
1785 size.</p>
1786
Jeffrey Yasskin7a088cf2010-01-11 19:19:26 +0000[diff] [blame]1787<p>Structures in memory are accessed using '<tt><a href="#i_load">load</a></tt>'
1788 and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a field
1789 with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.
1790 Structures in registers are accessed using the
1791 '<tt><a href="#i_extractvalue">extractvalue</a></tt>' and
1792 '<tt><a href="#i_insertvalue">insertvalue</a></tt>' instructions.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1793<h5>Syntax:</h5>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1794<pre>
1795 { &lt;type list&gt; }
1796</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1797
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1798<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1799<table class="layout">
1800 <tr class="layout">
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]1801 <td class="left"><tt>{ i32, i32, i32 }</tt></td>
1802 <td class="left">A triple of three <tt>i32</tt> values</td>
1803 </tr><tr class="layout">
1804 <td class="left"><tt>{&nbsp;float,&nbsp;i32&nbsp;(i32)&nbsp;*&nbsp;}</tt></td>
1805 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1806 second element is a <a href="#t_pointer">pointer</a> to a
1807 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1808 an <tt>i32</tt>.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1809 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1810</table>
Dan Gohmand8791e52009-01-24 15:58:40 +0000[diff] [blame]1811
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1812</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1813
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1814<!-- _______________________________________________________________________ -->
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]1815<div class="doc_subsubsection"> <a name="t_pstruct">Packed Structure Type</a>
1816</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1817
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]1818<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1819
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]1820<h5>Overview:</h5>
1821<p>The packed structure type is used to represent a collection of data members
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1822 together in memory. There is no padding between fields. Further, the
1823 alignment of a packed structure is 1 byte. The elements of a packed
1824 structure may be any type that has a size.</p>
1825
1826<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt> and
1827 '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a field with
1828 the '<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.</p>
1829
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]1830<h5>Syntax:</h5>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1831<pre>
1832 &lt; { &lt;type list&gt; } &gt;
1833</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1834
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]1835<h5>Examples:</h5>
1836<table class="layout">
1837 <tr class="layout">
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]1838 <td class="left"><tt>&lt; { i32, i32, i32 } &gt;</tt></td>
1839 <td class="left">A triple of three <tt>i32</tt> values</td>
1840 </tr><tr class="layout">
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1841 <td class="left">
1842<tt>&lt;&nbsp;{&nbsp;float,&nbsp;i32&nbsp;(i32)*&nbsp;}&nbsp;&gt;</tt></td>
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]1843 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1844 second element is a <a href="#t_pointer">pointer</a> to a
1845 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1846 an <tt>i32</tt>.</td>
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]1847 </tr>
1848</table>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1849
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]1850</div>
1851
1852<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1853<div class="doc_subsubsection"> <a name="t_pointer">Pointer Type</a> </div>
Chris Lattner0fd4a272009-02-08 19:53:29 +0000[diff] [blame]1854
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1855<div class="doc_text">
1856
1857<h5>Overview:</h5>
Dan Gohmanff3ef322010-02-25 16:50:07 +0000[diff] [blame]1858<p>The pointer type is used to specify memory locations.
1859 Pointers are commonly used to reference objects in memory.</p>
1860
1861<p>Pointer types may have an optional address space attribute defining the
1862 numbered address space where the pointed-to object resides. The default
1863 address space is number zero. The semantics of non-zero address
1864 spaces are target-specific.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1865
1866<p>Note that LLVM does not permit pointers to void (<tt>void*</tt>) nor does it
1867 permit pointers to labels (<tt>label*</tt>). Use <tt>i8*</tt> instead.</p>
Chris Lattner0fd4a272009-02-08 19:53:29 +0000[diff] [blame]1868
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1869<h5>Syntax:</h5>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1870<pre>
1871 &lt;type&gt; *
1872</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1873
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1874<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1875<table class="layout">
1876 <tr class="layout">
Dan Gohman2a08c532009-01-04 23:44:43 +0000[diff] [blame]1877 <td class="left"><tt>[4 x i32]*</tt></td>
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]1878 <td class="left">A <a href="#t_pointer">pointer</a> to <a
1879 href="#t_array">array</a> of four <tt>i32</tt> values.</td>
1880 </tr>
1881 <tr class="layout">
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]1882 <td class="left"><tt>i32 (i32*) *</tt></td>
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]1883 <td class="left"> A <a href="#t_pointer">pointer</a> to a <a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1884 href="#t_function">function</a> that takes an <tt>i32*</tt>, returning an
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]1885 <tt>i32</tt>.</td>
1886 </tr>
1887 <tr class="layout">
1888 <td class="left"><tt>i32 addrspace(5)*</tt></td>
1889 <td class="left">A <a href="#t_pointer">pointer</a> to an <tt>i32</tt> value
1890 that resides in address space #5.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1891 </tr>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1892</table>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1893
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1894</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1895
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1896<!-- _______________________________________________________________________ -->
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1897<div class="doc_subsubsection"> <a name="t_vector">Vector Type</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1898
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1899<div class="doc_text">
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1900
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1901<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1902<p>A vector type is a simple derived type that represents a vector of elements.
1903 Vector types are used when multiple primitive data are operated in parallel
1904 using a single instruction (SIMD). A vector type requires a size (number of
Duncan Sandsd40d14e2009-11-27 13:38:03 +0000[diff] [blame]1905 elements) and an underlying primitive data type. Vector types are considered
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1906 <a href="#t_firstclass">first class</a>.</p>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1907
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1908<h5>Syntax:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1909<pre>
1910 &lt; &lt;# elements&gt; x &lt;elementtype&gt; &gt;
1911</pre>
1912
Chris Lattner7d2e7be2010-10-10 18:20:35 +0000[diff] [blame]1913<p>The number of elements is a constant integer value larger than 0; elementtype
1914 may be any integer or floating point type. Vectors of size zero are not
1915 allowed, and pointers are not allowed as the element type.</p>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1916
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1917<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1918<table class="layout">
1919 <tr class="layout">
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]1920 <td class="left"><tt>&lt;4 x i32&gt;</tt></td>
1921 <td class="left">Vector of 4 32-bit integer values.</td>
1922 </tr>
1923 <tr class="layout">
1924 <td class="left"><tt>&lt;8 x float&gt;</tt></td>
1925 <td class="left">Vector of 8 32-bit floating-point values.</td>
1926 </tr>
1927 <tr class="layout">
1928 <td class="left"><tt>&lt;2 x i64&gt;</tt></td>
1929 <td class="left">Vector of 2 64-bit integer values.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1930 </tr>
1931</table>
Dan Gohmand8791e52009-01-24 15:58:40 +0000[diff] [blame]1932
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1933</div>
1934
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1935<!-- _______________________________________________________________________ -->
1936<div class="doc_subsubsection"> <a name="t_opaque">Opaque Type</a> </div>
1937<div class="doc_text">
1938
1939<h5>Overview:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1940<p>Opaque types are used to represent unknown types in the system. This
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1941 corresponds (for example) to the C notion of a forward declared structure
1942 type. In LLVM, opaque types can eventually be resolved to any type (not just
1943 a structure type).</p>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1944
1945<h5>Syntax:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1946<pre>
1947 opaque
1948</pre>
1949
1950<h5>Examples:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1951<table class="layout">
1952 <tr class="layout">
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]1953 <td class="left"><tt>opaque</tt></td>
1954 <td class="left">An opaque type.</td>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1955 </tr>
1956</table>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1957
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1958</div>
1959
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]1960<!-- ======================================================================= -->
1961<div class="doc_subsection">
1962 <a name="t_uprefs">Type Up-references</a>
1963</div>
1964
1965<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1966
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]1967<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1968<p>An "up reference" allows you to refer to a lexically enclosing type without
1969 requiring it to have a name. For instance, a structure declaration may
1970 contain a pointer to any of the types it is lexically a member of. Example
1971 of up references (with their equivalent as named type declarations)
1972 include:</p>
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]1973
1974<pre>
Chris Lattner3060f5b2009-02-09 10:00:56 +0000[diff] [blame]1975 { \2 * } %x = type { %x* }
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]1976 { \2 }* %y = type { %y }*
1977 \1* %z = type %z*
1978</pre>
1979
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1980<p>An up reference is needed by the asmprinter for printing out cyclic types
1981 when there is no declared name for a type in the cycle. Because the
1982 asmprinter does not want to print out an infinite type string, it needs a
1983 syntax to handle recursive types that have no names (all names are optional
1984 in llvm IR).</p>
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]1985
1986<h5>Syntax:</h5>
1987<pre>
1988 \&lt;level&gt;
1989</pre>
1990
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1991<p>The level is the count of the lexical type that is being referred to.</p>
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]1992
1993<h5>Examples:</h5>
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]1994<table class="layout">
1995 <tr class="layout">
1996 <td class="left"><tt>\1*</tt></td>
1997 <td class="left">Self-referential pointer.</td>
1998 </tr>
1999 <tr class="layout">
2000 <td class="left"><tt>{ { \3*, i8 }, i32 }</tt></td>
2001 <td class="left">Recursive structure where the upref refers to the out-most
2002 structure.</td>
2003 </tr>
2004</table>
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]2005
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2006</div>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]2007
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2008<!-- *********************************************************************** -->
2009<div class="doc_section"> <a name="constants">Constants</a> </div>
2010<!-- *********************************************************************** -->
2011
2012<div class="doc_text">
2013
2014<p>LLVM has several different basic types of constants. This section describes
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2015 them all and their syntax.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2016
2017</div>
2018
2019<!-- ======================================================================= -->
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]2020<div class="doc_subsection"><a name="simpleconstants">Simple Constants</a></div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2021
2022<div class="doc_text">
2023
2024<dl>
2025 <dt><b>Boolean constants</b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2026 <dd>The two strings '<tt>true</tt>' and '<tt>false</tt>' are both valid
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]2027 constants of the <tt><a href="#t_integer">i1</a></tt> type.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2028
2029 <dt><b>Integer constants</b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2030 <dd>Standard integers (such as '4') are constants of
2031 the <a href="#t_integer">integer</a> type. Negative numbers may be used
2032 with integer types.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2033
2034 <dt><b>Floating point constants</b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2035 <dd>Floating point constants use standard decimal notation (e.g. 123.421),
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2036 exponential notation (e.g. 1.23421e+2), or a more precise hexadecimal
2037 notation (see below). The assembler requires the exact decimal value of a
2038 floating-point constant. For example, the assembler accepts 1.25 but
2039 rejects 1.3 because 1.3 is a repeating decimal in binary. Floating point
2040 constants must have a <a href="#t_floating">floating point</a> type. </dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2041
2042 <dt><b>Null pointer constants</b></dt>
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]2043 <dd>The identifier '<tt>null</tt>' is recognized as a null pointer constant
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2044 and must be of <a href="#t_pointer">pointer type</a>.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2045</dl>
2046
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2047<p>The one non-intuitive notation for constants is the hexadecimal form of
2048 floating point constants. For example, the form '<tt>double
2049 0x432ff973cafa8000</tt>' is equivalent to (but harder to read than)
2050 '<tt>double 4.5e+15</tt>'. The only time hexadecimal floating point
2051 constants are required (and the only time that they are generated by the
2052 disassembler) is when a floating point constant must be emitted but it cannot
2053 be represented as a decimal floating point number in a reasonable number of
2054 digits. For example, NaN's, infinities, and other special values are
2055 represented in their IEEE hexadecimal format so that assembly and disassembly
2056 do not cause any bits to change in the constants.</p>
2057
Dale Johannesenbd5e5a82009-02-11 22:14:51 +0000[diff] [blame]2058<p>When using the hexadecimal form, constants of types float and double are
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2059 represented using the 16-digit form shown above (which matches the IEEE754
2060 representation for double); float values must, however, be exactly
2061 representable as IEE754 single precision. Hexadecimal format is always used
2062 for long double, and there are three forms of long double. The 80-bit format
2063 used by x86 is represented as <tt>0xK</tt> followed by 20 hexadecimal digits.
2064 The 128-bit format used by PowerPC (two adjacent doubles) is represented
2065 by <tt>0xM</tt> followed by 32 hexadecimal digits. The IEEE 128-bit format
2066 is represented by <tt>0xL</tt> followed by 32 hexadecimal digits; no
2067 currently supported target uses this format. Long doubles will only work if
2068 they match the long double format on your target. All hexadecimal formats
2069 are big-endian (sign bit at the left).</p>
2070
Dale Johannesen21fe99b2010-10-01 00:48:59 +0000[diff] [blame]2071<p>There are no constants of type x86mmx.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2072</div>
2073
2074<!-- ======================================================================= -->
Chris Lattner70882792009-02-28 18:32:25 +0000[diff] [blame]2075<div class="doc_subsection">
Bill Wendlingd9fe2982009-07-20 02:32:41 +0000[diff] [blame]2076<a name="aggregateconstants"></a> <!-- old anchor -->
2077<a name="complexconstants">Complex Constants</a>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2078</div>
2079
2080<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2081
Chris Lattner70882792009-02-28 18:32:25 +0000[diff] [blame]2082<p>Complex constants are a (potentially recursive) combination of simple
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2083 constants and smaller complex constants.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2084
2085<dl>
2086 <dt><b>Structure constants</b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2087 <dd>Structure constants are represented with notation similar to structure
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2088 type definitions (a comma separated list of elements, surrounded by braces
2089 (<tt>{}</tt>)). For example: "<tt>{ i32 4, float 17.0, i32* @G }</tt>",
2090 where "<tt>@G</tt>" is declared as "<tt>@G = external global i32</tt>".
2091 Structure constants must have <a href="#t_struct">structure type</a>, and
2092 the number and types of elements must match those specified by the
2093 type.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2094
2095 <dt><b>Array constants</b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2096 <dd>Array constants are represented with notation similar to array type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2097 definitions (a comma separated list of elements, surrounded by square
2098 brackets (<tt>[]</tt>)). For example: "<tt>[ i32 42, i32 11, i32 74
2099 ]</tt>". Array constants must have <a href="#t_array">array type</a>, and
2100 the number and types of elements must match those specified by the
2101 type.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2102
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]2103 <dt><b>Vector constants</b></dt>
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]2104 <dd>Vector constants are represented with notation similar to vector type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2105 definitions (a comma separated list of elements, surrounded by
2106 less-than/greater-than's (<tt>&lt;&gt;</tt>)). For example: "<tt>&lt; i32
2107 42, i32 11, i32 74, i32 100 &gt;</tt>". Vector constants must
2108 have <a href="#t_vector">vector type</a>, and the number and types of
2109 elements must match those specified by the type.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2110
2111 <dt><b>Zero initialization</b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2112 <dd>The string '<tt>zeroinitializer</tt>' can be used to zero initialize a
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]2113 value to zero of <em>any</em> type, including scalar and
2114 <a href="#t_aggregate">aggregate</a> types.
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2115 This is often used to avoid having to print large zero initializers
2116 (e.g. for large arrays) and is always exactly equivalent to using explicit
2117 zero initializers.</dd>
Nick Lewycky21cc4462009-04-04 07:22:01 +0000[diff] [blame]2118
2119 <dt><b>Metadata node</b></dt>
Nick Lewycky1e8c7a62009-05-30 16:08:30 +0000[diff] [blame]2120 <dd>A metadata node is a structure-like constant with
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2121 <a href="#t_metadata">metadata type</a>. For example: "<tt>metadata !{
2122 i32 0, metadata !"test" }</tt>". Unlike other constants that are meant to
2123 be interpreted as part of the instruction stream, metadata is a place to
2124 attach additional information such as debug info.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2125</dl>
2126
2127</div>
2128
2129<!-- ======================================================================= -->
2130<div class="doc_subsection">
2131 <a name="globalconstants">Global Variable and Function Addresses</a>
2132</div>
2133
2134<div class="doc_text">
2135
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2136<p>The addresses of <a href="#globalvars">global variables</a>
2137 and <a href="#functionstructure">functions</a> are always implicitly valid
2138 (link-time) constants. These constants are explicitly referenced when
2139 the <a href="#identifiers">identifier for the global</a> is used and always
2140 have <a href="#t_pointer">pointer</a> type. For example, the following is a
2141 legal LLVM file:</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2142
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2143<pre class="doc_code">
Chris Lattnera18a4242007-06-06 18:28:13 +0000[diff] [blame]2144@X = global i32 17
2145@Y = global i32 42
2146@Z = global [2 x i32*] [ i32* @X, i32* @Y ]
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2147</pre>
2148
2149</div>
2150
2151<!-- ======================================================================= -->
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]2152<div class="doc_subsection"><a name="undefvalues">Undefined Values</a></div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2153<div class="doc_text">
2154
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2155<p>The string '<tt>undef</tt>' can be used anywhere a constant is expected, and
Benjamin Kramer8040cd32009-10-12 14:46:08 +0000[diff] [blame]2156 indicates that the user of the value may receive an unspecified bit-pattern.
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2157 Undefined values may be of any type (other than label or void) and be used
2158 anywhere a constant is permitted.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2159
Chris Lattnerc608cb12009-09-11 01:49:31 +0000[diff] [blame]2160<p>Undefined values are useful because they indicate to the compiler that the
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2161 program is well defined no matter what value is used. This gives the
2162 compiler more freedom to optimize. Here are some examples of (potentially
2163 surprising) transformations that are valid (in pseudo IR):</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2164
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2165
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2166<pre class="doc_code">
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2167 %A = add %X, undef
2168 %B = sub %X, undef
2169 %C = xor %X, undef
2170Safe:
2171 %A = undef
2172 %B = undef
2173 %C = undef
2174</pre>
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2175
2176<p>This is safe because all of the output bits are affected by the undef bits.
2177Any output bit can have a zero or one depending on the input bits.</p>
2178
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2179<pre class="doc_code">
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2180 %A = or %X, undef
2181 %B = and %X, undef
2182Safe:
2183 %A = -1
2184 %B = 0
2185Unsafe:
2186 %A = undef
2187 %B = undef
2188</pre>
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2189
2190<p>These logical operations have bits that are not always affected by the input.
2191For example, if "%X" has a zero bit, then the output of the 'and' operation will
2192always be a zero, no matter what the corresponding bit from the undef is. As
Chris Lattnerc608cb12009-09-11 01:49:31 +0000[diff] [blame]2193such, it is unsafe to optimize or assume that the result of the and is undef.
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]2194However, it is safe to assume that all bits of the undef could be 0, and
2195optimize the and to 0. Likewise, it is safe to assume that all the bits of
2196the undef operand to the or could be set, allowing the or to be folded to
Chris Lattnerc608cb12009-09-11 01:49:31 +0000[diff] [blame]2197-1.</p>
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2198
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2199<pre class="doc_code">
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2200 %A = select undef, %X, %Y
2201 %B = select undef, 42, %Y
2202 %C = select %X, %Y, undef
2203Safe:
2204 %A = %X (or %Y)
2205 %B = 42 (or %Y)
2206 %C = %Y
2207Unsafe:
2208 %A = undef
2209 %B = undef
2210 %C = undef
2211</pre>
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2212
2213<p>This set of examples show that undefined select (and conditional branch)
2214conditions can go "either way" but they have to come from one of the two
2215operands. In the %A example, if %X and %Y were both known to have a clear low
2216bit, then %A would have to have a cleared low bit. However, in the %C example,
2217the optimizer is allowed to assume that the undef operand could be the same as
2218%Y, allowing the whole select to be eliminated.</p>
2219
2220
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2221<pre class="doc_code">
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2222 %A = xor undef, undef
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]2223
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2224 %B = undef
2225 %C = xor %B, %B
2226
2227 %D = undef
2228 %E = icmp lt %D, 4
2229 %F = icmp gte %D, 4
2230
2231Safe:
2232 %A = undef
2233 %B = undef
2234 %C = undef
2235 %D = undef
2236 %E = undef
2237 %F = undef
2238</pre>
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2239
2240<p>This example points out that two undef operands are not necessarily the same.
2241This can be surprising to people (and also matches C semantics) where they
2242assume that "X^X" is always zero, even if X is undef. This isn't true for a
2243number of reasons, but the short answer is that an undef "variable" can
2244arbitrarily change its value over its "live range". This is true because the
2245"variable" doesn't actually <em>have a live range</em>. Instead, the value is
2246logically read from arbitrary registers that happen to be around when needed,
Benjamin Kramer8040cd32009-10-12 14:46:08 +0000[diff] [blame]2247so the value is not necessarily consistent over time. In fact, %A and %C need
Chris Lattner349eb412009-09-08 15:13:16 +0000[diff] [blame]2248to have the same semantics or the core LLVM "replace all uses with" concept
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2249would not hold.</p>
Chris Lattner6e9057b2009-09-07 23:33:52 +0000[diff] [blame]2250
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2251<pre class="doc_code">
Chris Lattner6e9057b2009-09-07 23:33:52 +0000[diff] [blame]2252 %A = fdiv undef, %X
2253 %B = fdiv %X, undef
2254Safe:
2255 %A = undef
2256b: unreachable
2257</pre>
Chris Lattner6e9057b2009-09-07 23:33:52 +0000[diff] [blame]2258
2259<p>These examples show the crucial difference between an <em>undefined
2260value</em> and <em>undefined behavior</em>. An undefined value (like undef) is
2261allowed to have an arbitrary bit-pattern. This means that the %A operation
2262can be constant folded to undef because the undef could be an SNaN, and fdiv is
2263not (currently) defined on SNaN's. However, in the second example, we can make
2264a more aggressive assumption: because the undef is allowed to be an arbitrary
2265value, we are allowed to assume that it could be zero. Since a divide by zero
Chris Lattner8e371aa2009-09-08 19:45:34 +0000[diff] [blame]2266has <em>undefined behavior</em>, we are allowed to assume that the operation
Chris Lattner6e9057b2009-09-07 23:33:52 +0000[diff] [blame]2267does not execute at all. This allows us to delete the divide and all code after
2268it: since the undefined operation "can't happen", the optimizer can assume that
2269it occurs in dead code.
2270</p>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]2271
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2272<pre class="doc_code">
Chris Lattner6e9057b2009-09-07 23:33:52 +0000[diff] [blame]2273a: store undef -> %X
2274b: store %X -> undef
2275Safe:
2276a: &lt;deleted&gt;
2277b: unreachable
2278</pre>
Chris Lattner6e9057b2009-09-07 23:33:52 +0000[diff] [blame]2279
2280<p>These examples reiterate the fdiv example: a store "of" an undefined value
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]2281can be assumed to not have any effect: we can assume that the value is
Chris Lattner6e9057b2009-09-07 23:33:52 +0000[diff] [blame]2282overwritten with bits that happen to match what was already there. However, a
2283store "to" an undefined location could clobber arbitrary memory, therefore, it
2284has undefined behavior.</p>
2285
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2286</div>
2287
2288<!-- ======================================================================= -->
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]2289<div class="doc_subsection"><a name="trapvalues">Trap Values</a></div>
2290<div class="doc_text">
2291
Dan Gohmanc68ce062010-04-26 20:21:21 +0000[diff] [blame]2292<p>Trap values are similar to <a href="#undefvalues">undef values</a>, however
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]2293 instead of representing an unspecified bit pattern, they represent the
2294 fact that an instruction or constant expression which cannot evoke side
2295 effects has nevertheless detected a condition which results in undefined
Dan Gohmanc68ce062010-04-26 20:21:21 +0000[diff] [blame]2296 behavior.</p>
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]2297
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2298<p>There is currently no way of representing a trap value in the IR; they
Dan Gohman855abed2010-05-03 14:51:43 +0000[diff] [blame]2299 only exist when produced by operations such as
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2300 <a href="#i_add"><tt>add</tt></a> with the <tt>nsw</tt> flag.</p>
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2301
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2302<p>Trap value behavior is defined in terms of value <i>dependence</i>:</p>
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2303
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2304<ul>
2305<li>Values other than <a href="#i_phi"><tt>phi</tt></a> nodes depend on
2306 their operands.</li>
2307
2308<li><a href="#i_phi"><tt>Phi</tt></a> nodes depend on the operand corresponding
2309 to their dynamic predecessor basic block.</li>
2310
2311<li>Function arguments depend on the corresponding actual argument values in
2312 the dynamic callers of their functions.</li>
2313
2314<li><a href="#i_call"><tt>Call</tt></a> instructions depend on the
2315 <a href="#i_ret"><tt>ret</tt></a> instructions that dynamically transfer
2316 control back to them.</li>
2317
Dan Gohmanb5328162010-05-03 14:55:22 +0000[diff] [blame]2318<li><a href="#i_invoke"><tt>Invoke</tt></a> instructions depend on the
2319 <a href="#i_ret"><tt>ret</tt></a>, <a href="#i_unwind"><tt>unwind</tt></a>,
2320 or exception-throwing call instructions that dynamically transfer control
2321 back to them.</li>
2322
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2323<li>Non-volatile loads and stores depend on the most recent stores to all of the
2324 referenced memory addresses, following the order in the IR
2325 (including loads and stores implied by intrinsics such as
2326 <a href="#int_memcpy"><tt>@llvm.memcpy</tt></a>.)</li>
2327
Dan Gohman7c24ff12010-05-03 14:59:34 +0000[diff] [blame]2328<!-- TODO: In the case of multiple threads, this only applies if the store
2329 "happens-before" the load or store. -->
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2330
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2331<!-- TODO: floating-point exception state -->
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2332
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2333<li>An instruction with externally visible side effects depends on the most
2334 recent preceding instruction with externally visible side effects, following
Dan Gohmanff70fe42010-07-06 15:26:33 +0000[diff] [blame]2335 the order in the IR. (This includes
2336 <a href="#volatile">volatile operations</a>.)</li>
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]2337
Dan Gohmanb5328162010-05-03 14:55:22 +0000[diff] [blame]2338<li>An instruction <i>control-depends</i> on a
2339 <a href="#terminators">terminator instruction</a>
2340 if the terminator instruction has multiple successors and the instruction
2341 is always executed when control transfers to one of the successors, and
2342 may not be executed when control is transfered to another.</li>
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2343
2344<li>Dependence is transitive.</li>
2345
2346</ul>
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2347
2348<p>Whenever a trap value is generated, all values which depend on it evaluate
2349 to trap. If they have side effects, the evoke their side effects as if each
2350 operand with a trap value were undef. If they have externally-visible side
2351 effects, the behavior is undefined.</p>
2352
2353<p>Here are some examples:</p>
Dan Gohmanc30f6e12010-04-26 20:54:53 +0000[diff] [blame]2354
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2355<pre class="doc_code">
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2356entry:
2357 %trap = sub nuw i32 0, 1 ; Results in a trap value.
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2358 %still_trap = and i32 %trap, 0 ; Whereas (and i32 undef, 0) would return 0.
2359 %trap_yet_again = getelementptr i32* @h, i32 %still_trap
2360 store i32 0, i32* %trap_yet_again ; undefined behavior
2361
2362 store i32 %trap, i32* @g ; Trap value conceptually stored to memory.
2363 %trap2 = load i32* @g ; Returns a trap value, not just undef.
2364
2365 volatile store i32 %trap, i32* @g ; External observation; undefined behavior.
2366
2367 %narrowaddr = bitcast i32* @g to i16*
2368 %wideaddr = bitcast i32* @g to i64*
2369 %trap3 = load 16* %narrowaddr ; Returns a trap value.
2370 %trap4 = load i64* %widaddr ; Returns a trap value.
2371
2372 %cmp = icmp i32 slt %trap, 0 ; Returns a trap value.
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2373 %br i1 %cmp, %true, %end ; Branch to either destination.
2374
2375true:
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2376 volatile store i32 0, i32* @g ; This is control-dependent on %cmp, so
2377 ; it has undefined behavior.
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2378 br label %end
2379
2380end:
2381 %p = phi i32 [ 0, %entry ], [ 1, %true ]
2382 ; Both edges into this PHI are
2383 ; control-dependent on %cmp, so this
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2384 ; always results in a trap value.
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2385
2386 volatile store i32 0, i32* @g ; %end is control-equivalent to %entry
2387 ; so this is defined (ignoring earlier
2388 ; undefined behavior in this example).
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2389</pre>
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]2390
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]2391</div>
2392
2393<!-- ======================================================================= -->
Chris Lattnerf9d078e2009-10-27 21:19:13 +0000[diff] [blame]2394<div class="doc_subsection"><a name="blockaddress">Addresses of Basic
2395 Blocks</a></div>
Chris Lattnerc6f44362009-10-27 21:01:34 +0000[diff] [blame]2396<div class="doc_text">
2397
Chris Lattnercdfc9402009-11-01 01:27:45 +0000[diff] [blame]2398<p><b><tt>blockaddress(@function, %block)</tt></b></p>
Chris Lattnerc6f44362009-10-27 21:01:34 +0000[diff] [blame]2399
2400<p>The '<tt>blockaddress</tt>' constant computes the address of the specified
Chris Lattner2dfdf2a2009-10-27 21:49:40 +0000[diff] [blame]2401 basic block in the specified function, and always has an i8* type. Taking
Chris Lattnercdfc9402009-11-01 01:27:45 +0000[diff] [blame]2402 the address of the entry block is illegal.</p>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]2403
Chris Lattnerc6f44362009-10-27 21:01:34 +0000[diff] [blame]2404<p>This value only has defined behavior when used as an operand to the
Chris Lattnerab21db72009-10-28 00:19:10 +0000[diff] [blame]2405 '<a href="#i_indirectbr"><tt>indirectbr</tt></a>' instruction or for comparisons
Chris Lattnerc6f44362009-10-27 21:01:34 +0000[diff] [blame]2406 against null. Pointer equality tests between labels addresses is undefined
2407 behavior - though, again, comparison against null is ok, and no label is
Chris Lattnerf9d078e2009-10-27 21:19:13 +0000[diff] [blame]2408 equal to the null pointer. This may also be passed around as an opaque
2409 pointer sized value as long as the bits are not inspected. This allows
Chris Lattner3fd77ce2009-10-27 21:44:20 +0000[diff] [blame]2410 <tt>ptrtoint</tt> and arithmetic to be performed on these values so long as
Chris Lattnerab21db72009-10-28 00:19:10 +0000[diff] [blame]2411 the original value is reconstituted before the <tt>indirectbr</tt>.</p>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]2412
Chris Lattnerf9d078e2009-10-27 21:19:13 +0000[diff] [blame]2413<p>Finally, some targets may provide defined semantics when
Chris Lattnerc6f44362009-10-27 21:01:34 +0000[diff] [blame]2414 using the value as the operand to an inline assembly, but that is target
2415 specific.
2416 </p>
2417
2418</div>
2419
2420
2421<!-- ======================================================================= -->
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2422<div class="doc_subsection"><a name="constantexprs">Constant Expressions</a>
2423</div>
2424
2425<div class="doc_text">
2426
2427<p>Constant expressions are used to allow expressions involving other constants
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2428 to be used as constants. Constant expressions may be of
2429 any <a href="#t_firstclass">first class</a> type and may involve any LLVM
2430 operation that does not have side effects (e.g. load and call are not
2431 supported). The following is the syntax for constant expressions:</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2432
2433<dl>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2434 <dt><b><tt>trunc (CST to TYPE)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2435 <dd>Truncate a constant to another type. The bit size of CST must be larger
2436 than the bit size of TYPE. Both types must be integers.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2437
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2438 <dt><b><tt>zext (CST to TYPE)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2439 <dd>Zero extend a constant to another type. The bit size of CST must be
Duncan Sands28afd432010-07-13 12:06:14 +0000[diff] [blame]2440 smaller than the bit size of TYPE. Both types must be integers.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2441
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2442 <dt><b><tt>sext (CST to TYPE)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2443 <dd>Sign extend a constant to another type. The bit size of CST must be
Duncan Sands28afd432010-07-13 12:06:14 +0000[diff] [blame]2444 smaller than the bit size of TYPE. Both types must be integers.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2445
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2446 <dt><b><tt>fptrunc (CST to TYPE)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2447 <dd>Truncate a floating point constant to another floating point type. The
2448 size of CST must be larger than the size of TYPE. Both types must be
2449 floating point.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2450
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2451 <dt><b><tt>fpext (CST to TYPE)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2452 <dd>Floating point extend a constant to another type. The size of CST must be
2453 smaller or equal to the size of TYPE. Both types must be floating
2454 point.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2455
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2456 <dt><b><tt>fptoui (CST to TYPE)</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2457 <dd>Convert a floating point constant to the corresponding unsigned integer
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2458 constant. TYPE must be a scalar or vector integer type. CST must be of
2459 scalar or vector floating point type. Both CST and TYPE must be scalars,
2460 or vectors of the same number of elements. If the value won't fit in the
2461 integer type, the results are undefined.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2462
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2463 <dt><b><tt>fptosi (CST to TYPE)</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2464 <dd>Convert a floating point constant to the corresponding signed integer
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2465 constant. TYPE must be a scalar or vector integer type. CST must be of
2466 scalar or vector floating point type. Both CST and TYPE must be scalars,
2467 or vectors of the same number of elements. If the value won't fit in the
2468 integer type, the results are undefined.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2469
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2470 <dt><b><tt>uitofp (CST to TYPE)</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2471 <dd>Convert an unsigned integer constant to the corresponding floating point
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2472 constant. TYPE must be a scalar or vector floating point type. CST must be
2473 of scalar or vector integer type. Both CST and TYPE must be scalars, or
2474 vectors of the same number of elements. If the value won't fit in the
2475 floating point type, the results are undefined.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2476
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2477 <dt><b><tt>sitofp (CST to TYPE)</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2478 <dd>Convert a signed integer constant to the corresponding floating point
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2479 constant. TYPE must be a scalar or vector floating point type. CST must be
2480 of scalar or vector integer type. Both CST and TYPE must be scalars, or
2481 vectors of the same number of elements. If the value won't fit in the
2482 floating point type, the results are undefined.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2483
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2484 <dt><b><tt>ptrtoint (CST to TYPE)</tt></b></dt>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]2485 <dd>Convert a pointer typed constant to the corresponding integer constant
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2486 <tt>TYPE</tt> must be an integer type. <tt>CST</tt> must be of pointer
2487 type. The <tt>CST</tt> value is zero extended, truncated, or unchanged to
2488 make it fit in <tt>TYPE</tt>.</dd>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]2489
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2490 <dt><b><tt>inttoptr (CST to TYPE)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2491 <dd>Convert a integer constant to a pointer constant. TYPE must be a pointer
2492 type. CST must be of integer type. The CST value is zero extended,
2493 truncated, or unchanged to make it fit in a pointer size. This one is
2494 <i>really</i> dangerous!</dd>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]2495
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2496 <dt><b><tt>bitcast (CST to TYPE)</tt></b></dt>
Chris Lattner03bbad62009-02-28 18:27:03 +0000[diff] [blame]2497 <dd>Convert a constant, CST, to another TYPE. The constraints of the operands
2498 are the same as those for the <a href="#i_bitcast">bitcast
2499 instruction</a>.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2500
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2501 <dt><b><tt>getelementptr (CSTPTR, IDX0, IDX1, ...)</tt></b></dt>
2502 <dt><b><tt>getelementptr inbounds (CSTPTR, IDX0, IDX1, ...)</tt></b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2503 <dd>Perform the <a href="#i_getelementptr">getelementptr operation</a> on
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2504 constants. As with the <a href="#i_getelementptr">getelementptr</a>
2505 instruction, the index list may have zero or more indexes, which are
2506 required to make sense for the type of "CSTPTR".</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2507
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2508 <dt><b><tt>select (COND, VAL1, VAL2)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2509 <dd>Perform the <a href="#i_select">select operation</a> on constants.</dd>
Reid Spencer01c42592006-12-04 19:23:19 +0000[diff] [blame]2510
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2511 <dt><b><tt>icmp COND (VAL1, VAL2)</tt></b></dt>
Reid Spencer01c42592006-12-04 19:23:19 +0000[diff] [blame]2512 <dd>Performs the <a href="#i_icmp">icmp operation</a> on constants.</dd>
2513
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2514 <dt><b><tt>fcmp COND (VAL1, VAL2)</tt></b></dt>
Reid Spencer01c42592006-12-04 19:23:19 +0000[diff] [blame]2515 <dd>Performs the <a href="#i_fcmp">fcmp operation</a> on constants.</dd>
Robert Bocchino9fbe1452006-01-10 19:31:34 +0000[diff] [blame]2516
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2517 <dt><b><tt>extractelement (VAL, IDX)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2518 <dd>Perform the <a href="#i_extractelement">extractelement operation</a> on
2519 constants.</dd>
Robert Bocchino9fbe1452006-01-10 19:31:34 +0000[diff] [blame]2520
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2521 <dt><b><tt>insertelement (VAL, ELT, IDX)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2522 <dd>Perform the <a href="#i_insertelement">insertelement operation</a> on
2523 constants.</dd>
Chris Lattnerc1989542006-04-08 00:13:41 +0000[diff] [blame]2524
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2525 <dt><b><tt>shufflevector (VEC1, VEC2, IDXMASK)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2526 <dd>Perform the <a href="#i_shufflevector">shufflevector operation</a> on
2527 constants.</dd>
Chris Lattnerc1989542006-04-08 00:13:41 +0000[diff] [blame]2528
Nick Lewycky9e130ce2010-05-29 06:44:15 +0000[diff] [blame]2529 <dt><b><tt>extractvalue (VAL, IDX0, IDX1, ...)</tt></b></dt>
2530 <dd>Perform the <a href="#i_extractvalue">extractvalue operation</a> on
2531 constants. The index list is interpreted in a similar manner as indices in
2532 a '<a href="#i_getelementptr">getelementptr</a>' operation. At least one
2533 index value must be specified.</dd>
2534
2535 <dt><b><tt>insertvalue (VAL, ELT, IDX0, IDX1, ...)</tt></b></dt>
2536 <dd>Perform the <a href="#i_insertvalue">insertvalue operation</a> on
2537 constants. The index list is interpreted in a similar manner as indices in
2538 a '<a href="#i_getelementptr">getelementptr</a>' operation. At least one
2539 index value must be specified.</dd>
2540
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2541 <dt><b><tt>OPCODE (LHS, RHS)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2542 <dd>Perform the specified operation of the LHS and RHS constants. OPCODE may
2543 be any of the <a href="#binaryops">binary</a>
2544 or <a href="#bitwiseops">bitwise binary</a> operations. The constraints
2545 on operands are the same as those for the corresponding instruction
2546 (e.g. no bitwise operations on floating point values are allowed).</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2547</dl>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2548
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2549</div>
Chris Lattner9ee5d222004-03-08 16:49:10 +0000[diff] [blame]2550
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2551<!-- *********************************************************************** -->
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2552<div class="doc_section"> <a name="othervalues">Other Values</a> </div>
2553<!-- *********************************************************************** -->
2554
2555<!-- ======================================================================= -->
2556<div class="doc_subsection">
2557<a name="inlineasm">Inline Assembler Expressions</a>
2558</div>
2559
2560<div class="doc_text">
2561
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2562<p>LLVM supports inline assembler expressions (as opposed
2563 to <a href="#moduleasm"> Module-Level Inline Assembly</a>) through the use of
2564 a special value. This value represents the inline assembler as a string
2565 (containing the instructions to emit), a list of operand constraints (stored
Dale Johannesen09fed252009-10-13 21:56:55 +0000[diff] [blame]2566 as a string), a flag that indicates whether or not the inline asm
Dale Johannesen8ba2d5b2009-10-21 23:28:00 +0000[diff] [blame]2567 expression has side effects, and a flag indicating whether the function
2568 containing the asm needs to align its stack conservatively. An example
2569 inline assembler expression is:</p>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2570
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2571<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2572i32 (i32) asm "bswap $0", "=r,r"
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2573</pre>
2574
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2575<p>Inline assembler expressions may <b>only</b> be used as the callee operand of
2576 a <a href="#i_call"><tt>call</tt> instruction</a>. Thus, typically we
2577 have:</p>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2578
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2579<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2580%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]2581</pre>
2582
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2583<p>Inline asms with side effects not visible in the constraint list must be
2584 marked as having side effects. This is done through the use of the
2585 '<tt>sideeffect</tt>' keyword, like so:</p>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2586
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2587<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2588call void asm sideeffect "eieio", ""()
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2589</pre>
2590
Dale Johannesen8ba2d5b2009-10-21 23:28:00 +0000[diff] [blame]2591<p>In some cases inline asms will contain code that will not work unless the
2592 stack is aligned in some way, such as calls or SSE instructions on x86,
2593 yet will not contain code that does that alignment within the asm.
2594 The compiler should make conservative assumptions about what the asm might
2595 contain and should generate its usual stack alignment code in the prologue
2596 if the '<tt>alignstack</tt>' keyword is present:</p>
Dale Johannesen09fed252009-10-13 21:56:55 +0000[diff] [blame]2597
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2598<pre class="doc_code">
Dale Johannesen8ba2d5b2009-10-21 23:28:00 +0000[diff] [blame]2599call void asm alignstack "eieio", ""()
Dale Johannesen09fed252009-10-13 21:56:55 +0000[diff] [blame]2600</pre>
Dale Johannesen09fed252009-10-13 21:56:55 +0000[diff] [blame]2601
2602<p>If both keywords appear the '<tt>sideeffect</tt>' keyword must come
2603 first.</p>
2604
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2605<p>TODO: The format of the asm and constraints string still need to be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2606 documented here. Constraints on what can be done (e.g. duplication, moving,
2607 etc need to be documented). This is probably best done by reference to
2608 another document that covers inline asm from a holistic perspective.</p>
Chris Lattnercf9a4152010-04-07 05:38:05 +0000[diff] [blame]2609</div>
2610
2611<div class="doc_subsubsection">
2612<a name="inlineasm_md">Inline Asm Metadata</a>
2613</div>
2614
2615<div class="doc_text">
2616
2617<p>The call instructions that wrap inline asm nodes may have a "!srcloc" MDNode
2618 attached to it that contains a constant integer. If present, the code
2619 generator will use the integer as the location cookie value when report
2620 errors through the LLVMContext error reporting mechanisms. This allows a
Dan Gohman1c70c002010-04-28 00:36:01 +0000[diff] [blame]2621 front-end to correlate backend errors that occur with inline asm back to the
Chris Lattnercf9a4152010-04-07 05:38:05 +0000[diff] [blame]2622 source code that produced it. For example:</p>
2623
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2624<pre class="doc_code">
Chris Lattnercf9a4152010-04-07 05:38:05 +0000[diff] [blame]2625call void asm sideeffect "something bad", ""()<b>, !srcloc !42</b>
2626...
2627!42 = !{ i32 1234567 }
2628</pre>
Chris Lattnercf9a4152010-04-07 05:38:05 +0000[diff] [blame]2629
2630<p>It is up to the front-end to make sense of the magic numbers it places in the
2631 IR.</p>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2632
2633</div>
2634
Chris Lattnere6a5ddd2010-01-15 21:50:19 +0000[diff] [blame]2635<!-- ======================================================================= -->
2636<div class="doc_subsection"><a name="metadata">Metadata Nodes and Metadata
2637 Strings</a>
2638</div>
2639
2640<div class="doc_text">
2641
2642<p>LLVM IR allows metadata to be attached to instructions in the program that
2643 can convey extra information about the code to the optimizers and code
2644 generator. One example application of metadata is source-level debug
2645 information. There are two metadata primitives: strings and nodes. All
2646 metadata has the <tt>metadata</tt> type and is identified in syntax by a
2647 preceding exclamation point ('<tt>!</tt>').</p>
2648
2649<p>A metadata string is a string surrounded by double quotes. It can contain
2650 any character by escaping non-printable characters with "\xx" where "xx" is
2651 the two digit hex code. For example: "<tt>!"test\00"</tt>".</p>
2652
2653<p>Metadata nodes are represented with notation similar to structure constants
2654 (a comma separated list of elements, surrounded by braces and preceded by an
2655 exclamation point). For example: "<tt>!{ metadata !"test\00", i32
2656 10}</tt>". Metadata nodes can have any values as their operand.</p>
2657
2658<p>A <a href="#namedmetadatastructure">named metadata</a> is a collection of
2659 metadata nodes, which can be looked up in the module symbol table. For
2660 example: "<tt>!foo = metadata !{!4, !3}</tt>".
2661
Devang Patele1d50cd2010-03-04 23:44:48 +0000[diff] [blame]2662<p>Metadata can be used as function arguments. Here <tt>llvm.dbg.value</tt>
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2663 function is using two metadata arguments.</p>
Devang Patele1d50cd2010-03-04 23:44:48 +0000[diff] [blame]2664
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2665 <pre class="doc_code">
Devang Patele1d50cd2010-03-04 23:44:48 +0000[diff] [blame]2666 call void @llvm.dbg.value(metadata !24, i64 0, metadata !25)
2667 </pre>
Devang Patele1d50cd2010-03-04 23:44:48 +0000[diff] [blame]2668
2669<p>Metadata can be attached with an instruction. Here metadata <tt>!21</tt> is
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2670 attached with <tt>add</tt> instruction using <tt>!dbg</tt> identifier.</p>
Devang Patele1d50cd2010-03-04 23:44:48 +0000[diff] [blame]2671
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2672 <pre class="doc_code">
Devang Patele1d50cd2010-03-04 23:44:48 +0000[diff] [blame]2673 %indvar.next = add i64 %indvar, 1, !dbg !21
2674 </pre>
Chris Lattnere6a5ddd2010-01-15 21:50:19 +0000[diff] [blame]2675</div>
2676
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2677
2678<!-- *********************************************************************** -->
2679<div class="doc_section">
2680 <a name="intrinsic_globals">Intrinsic Global Variables</a>
2681</div>
2682<!-- *********************************************************************** -->
2683
2684<p>LLVM has a number of "magic" global variables that contain data that affect
2685code generation or other IR semantics. These are documented here. All globals
Chris Lattner401e10c2009-07-20 06:14:25 +0000[diff] [blame]2686of this sort should have a section specified as "<tt>llvm.metadata</tt>". This
2687section and all globals that start with "<tt>llvm.</tt>" are reserved for use
2688by LLVM.</p>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2689
2690<!-- ======================================================================= -->
2691<div class="doc_subsection">
2692<a name="intg_used">The '<tt>llvm.used</tt>' Global Variable</a>
2693</div>
2694
2695<div class="doc_text">
2696
2697<p>The <tt>@llvm.used</tt> global is an array with i8* element type which has <a
2698href="#linkage_appending">appending linkage</a>. This array contains a list of
2699pointers to global variables and functions which may optionally have a pointer
2700cast formed of bitcast or getelementptr. For example, a legal use of it is:</p>
2701
2702<pre>
2703 @X = global i8 4
2704 @Y = global i32 123
2705
2706 @llvm.used = appending global [2 x i8*] [
2707 i8* @X,
2708 i8* bitcast (i32* @Y to i8*)
2709 ], section "llvm.metadata"
2710</pre>
2711
2712<p>If a global variable appears in the <tt>@llvm.used</tt> list, then the
2713compiler, assembler, and linker are required to treat the symbol as if there is
2714a reference to the global that it cannot see. For example, if a variable has
2715internal linkage and no references other than that from the <tt>@llvm.used</tt>
2716list, it cannot be deleted. This is commonly used to represent references from
2717inline asms and other things the compiler cannot "see", and corresponds to
2718"attribute((used))" in GNU C.</p>
2719
2720<p>On some targets, the code generator must emit a directive to the assembler or
2721object file to prevent the assembler and linker from molesting the symbol.</p>
2722
2723</div>
2724
2725<!-- ======================================================================= -->
2726<div class="doc_subsection">
Chris Lattner401e10c2009-07-20 06:14:25 +0000[diff] [blame]2727<a name="intg_compiler_used">The '<tt>llvm.compiler.used</tt>' Global Variable</a>
2728</div>
2729
2730<div class="doc_text">
2731
2732<p>The <tt>@llvm.compiler.used</tt> directive is the same as the
2733<tt>@llvm.used</tt> directive, except that it only prevents the compiler from
2734touching the symbol. On targets that support it, this allows an intelligent
2735linker to optimize references to the symbol without being impeded as it would be
2736by <tt>@llvm.used</tt>.</p>
2737
2738<p>This is a rare construct that should only be used in rare circumstances, and
2739should not be exposed to source languages.</p>
2740
2741</div>
2742
2743<!-- ======================================================================= -->
2744<div class="doc_subsection">
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2745<a name="intg_global_ctors">The '<tt>llvm.global_ctors</tt>' Global Variable</a>
2746</div>
2747
2748<div class="doc_text">
David Chisnalle31e9962010-04-30 19:23:49 +0000[diff] [blame]2749<pre>
2750%0 = type { i32, void ()* }
David Chisnall27195a52010-04-30 19:27:35 +0000[diff] [blame]2751@llvm.global_ctors = appending global [1 x %0] [%0 { i32 65535, void ()* @ctor }]
David Chisnalle31e9962010-04-30 19:23:49 +0000[diff] [blame]2752</pre>
2753<p>The <tt>@llvm.global_ctors</tt> array contains a list of constructor functions and associated priorities. The functions referenced by this array will be called in ascending order of priority (i.e. lowest first) when the module is loaded. The order of functions with the same priority is not defined.
2754</p>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2755
2756</div>
2757
2758<!-- ======================================================================= -->
2759<div class="doc_subsection">
2760<a name="intg_global_dtors">The '<tt>llvm.global_dtors</tt>' Global Variable</a>
2761</div>
2762
2763<div class="doc_text">
David Chisnalle31e9962010-04-30 19:23:49 +0000[diff] [blame]2764<pre>
2765%0 = type { i32, void ()* }
David Chisnall27195a52010-04-30 19:27:35 +0000[diff] [blame]2766@llvm.global_dtors = appending global [1 x %0] [%0 { i32 65535, void ()* @dtor }]
David Chisnalle31e9962010-04-30 19:23:49 +0000[diff] [blame]2767</pre>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2768
David Chisnalle31e9962010-04-30 19:23:49 +0000[diff] [blame]2769<p>The <tt>@llvm.global_dtors</tt> array contains a list of destructor functions and associated priorities. The functions referenced by this array will be called in descending order of priority (i.e. highest first) when the module is loaded. The order of functions with the same priority is not defined.
2770</p>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2771
2772</div>
2773
2774
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2775<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2776<div class="doc_section"> <a name="instref">Instruction Reference</a> </div>
2777<!-- *********************************************************************** -->
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2778
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2779<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2780
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2781<p>The LLVM instruction set consists of several different classifications of
2782 instructions: <a href="#terminators">terminator
2783 instructions</a>, <a href="#binaryops">binary instructions</a>,
2784 <a href="#bitwiseops">bitwise binary instructions</a>,
2785 <a href="#memoryops">memory instructions</a>, and
2786 <a href="#otherops">other instructions</a>.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2787
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2788</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2789
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2790<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2791<div class="doc_subsection"> <a name="terminators">Terminator
2792Instructions</a> </div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2793
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2794<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2795
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2796<p>As mentioned <a href="#functionstructure">previously</a>, every basic block
2797 in a program ends with a "Terminator" instruction, which indicates which
2798 block should be executed after the current block is finished. These
2799 terminator instructions typically yield a '<tt>void</tt>' value: they produce
2800 control flow, not values (the one exception being the
2801 '<a href="#i_invoke"><tt>invoke</tt></a>' instruction).</p>
2802
Duncan Sands83821c82010-04-15 20:35:54 +0000[diff] [blame]2803<p>There are seven different terminator instructions: the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2804 '<a href="#i_ret"><tt>ret</tt></a>' instruction, the
2805 '<a href="#i_br"><tt>br</tt></a>' instruction, the
2806 '<a href="#i_switch"><tt>switch</tt></a>' instruction, the
Bill Wendling21c346e2009-11-02 00:25:26 +0000[diff] [blame]2807 '<a href="#i_indirectbr">'<tt>indirectbr</tt></a>' Instruction, the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2808 '<a href="#i_invoke"><tt>invoke</tt></a>' instruction, the
2809 '<a href="#i_unwind"><tt>unwind</tt></a>' instruction, and the
2810 '<a href="#i_unreachable"><tt>unreachable</tt></a>' instruction.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2811
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2812</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2813
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2814<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2815<div class="doc_subsubsection"> <a name="i_ret">'<tt>ret</tt>'
2816Instruction</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2817
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2818<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2819
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2820<h5>Syntax:</h5>
Dan Gohmanb1e6b962008-10-04 19:00:07 +0000[diff] [blame]2821<pre>
2822 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]2823 ret void <i>; Return from void function</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2824</pre>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]2825
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2826<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2827<p>The '<tt>ret</tt>' instruction is used to return control flow (and optionally
2828 a value) from a function back to the caller.</p>
2829
2830<p>There are two forms of the '<tt>ret</tt>' instruction: one that returns a
2831 value and then causes control flow, and one that just causes control flow to
2832 occur.</p>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]2833
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2834<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2835<p>The '<tt>ret</tt>' instruction optionally accepts a single argument, the
2836 return value. The type of the return value must be a
2837 '<a href="#t_firstclass">first class</a>' type.</p>
Dan Gohmanb1e6b962008-10-04 19:00:07 +0000[diff] [blame]2838
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2839<p>A function is not <a href="#wellformed">well formed</a> if it it has a
2840 non-void return type and contains a '<tt>ret</tt>' instruction with no return
2841 value or a return value with a type that does not match its type, or if it
2842 has a void return type and contains a '<tt>ret</tt>' instruction with a
2843 return value.</p>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]2844
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2845<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2846<p>When the '<tt>ret</tt>' instruction is executed, control flow returns back to
2847 the calling function's context. If the caller is a
2848 "<a href="#i_call"><tt>call</tt></a>" instruction, execution continues at the
2849 instruction after the call. If the caller was an
2850 "<a href="#i_invoke"><tt>invoke</tt></a>" instruction, execution continues at
2851 the beginning of the "normal" destination block. If the instruction returns
2852 a value, that value shall set the call or invoke instruction's return
2853 value.</p>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]2854
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2855<h5>Example:</h5>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]2856<pre>
2857 ret i32 5 <i>; Return an integer value of 5</i>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2858 ret void <i>; Return from a void function</i>
Bill Wendling0a4bbbf2009-02-28 22:12:54 +0000[diff] [blame]2859 ret { i32, i8 } { i32 4, i8 2 } <i>; Return a struct of values 4 and 2</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2860</pre>
Dan Gohmanf3e60bd2009-01-12 23:12:39 +0000[diff] [blame]2861
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2862</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2863<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2864<div class="doc_subsubsection"> <a name="i_br">'<tt>br</tt>' Instruction</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2865
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2866<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2867
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2868<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2869<pre>
2870 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]2871</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2872
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2873<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2874<p>The '<tt>br</tt>' instruction is used to cause control flow to transfer to a
2875 different basic block in the current function. There are two forms of this
2876 instruction, corresponding to a conditional branch and an unconditional
2877 branch.</p>
2878
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2879<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2880<p>The conditional branch form of the '<tt>br</tt>' instruction takes a single
2881 '<tt>i1</tt>' value and two '<tt>label</tt>' values. The unconditional form
2882 of the '<tt>br</tt>' instruction takes a single '<tt>label</tt>' value as a
2883 target.</p>
2884
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2885<h5>Semantics:</h5>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]2886<p>Upon execution of a conditional '<tt>br</tt>' instruction, the '<tt>i1</tt>'
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2887 argument is evaluated. If the value is <tt>true</tt>, control flows to the
2888 '<tt>iftrue</tt>' <tt>label</tt> argument. If "cond" is <tt>false</tt>,
2889 control flows to the '<tt>iffalse</tt>' <tt>label</tt> argument.</p>
2890
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2891<h5>Example:</h5>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]2892<pre>
2893Test:
2894 %cond = <a href="#i_icmp">icmp</a> eq i32 %a, %b
2895 br i1 %cond, label %IfEqual, label %IfUnequal
2896IfEqual:
2897 <a href="#i_ret">ret</a> i32 1
2898IfUnequal:
2899 <a href="#i_ret">ret</a> i32 0
2900</pre>
2901
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2902</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2903
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2904<!-- _______________________________________________________________________ -->
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2905<div class="doc_subsubsection">
2906 <a name="i_switch">'<tt>switch</tt>' Instruction</a>
2907</div>
2908
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2909<div class="doc_text">
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2910
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2911<h5>Syntax:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2912<pre>
2913 switch &lt;intty&gt; &lt;value&gt;, label &lt;defaultdest&gt; [ &lt;intty&gt; &lt;val&gt;, label &lt;dest&gt; ... ]
2914</pre>
2915
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2916<h5>Overview:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2917<p>The '<tt>switch</tt>' instruction is used to transfer control flow to one of
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2918 several different places. It is a generalization of the '<tt>br</tt>'
2919 instruction, allowing a branch to occur to one of many possible
2920 destinations.</p>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2921
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2922<h5>Arguments:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2923<p>The '<tt>switch</tt>' instruction uses three parameters: an integer
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2924 comparison value '<tt>value</tt>', a default '<tt>label</tt>' destination,
2925 and an array of pairs of comparison value constants and '<tt>label</tt>'s.
2926 The table is not allowed to contain duplicate constant entries.</p>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2927
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2928<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2929<p>The <tt>switch</tt> instruction specifies a table of values and
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2930 destinations. When the '<tt>switch</tt>' instruction is executed, this table
2931 is searched for the given value. If the value is found, control flow is
Benjamin Kramer8040cd32009-10-12 14:46:08 +0000[diff] [blame]2932 transferred to the corresponding destination; otherwise, control flow is
2933 transferred to the default destination.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2934
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2935<h5>Implementation:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2936<p>Depending on properties of the target machine and the particular
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2937 <tt>switch</tt> instruction, this instruction may be code generated in
2938 different ways. For example, it could be generated as a series of chained
2939 conditional branches or with a lookup table.</p>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2940
2941<h5>Example:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2942<pre>
2943 <i>; Emulate a conditional br instruction</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]2944 %Val = <a href="#i_zext">zext</a> i1 %value to i32
Dan Gohman2a08c532009-01-04 23:44:43 +0000[diff] [blame]2945 switch i32 %Val, label %truedest [ i32 0, label %falsedest ]
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2946
2947 <i>; Emulate an unconditional br instruction</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2948 switch i32 0, label %dest [ ]
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2949
2950 <i>; Implement a jump table:</i>
Dan Gohman2a08c532009-01-04 23:44:43 +0000[diff] [blame]2951 switch i32 %val, label %otherwise [ i32 0, label %onzero
2952 i32 1, label %onone
2953 i32 2, label %ontwo ]
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2954</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2955
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2956</div>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2957
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]2958
2959<!-- _______________________________________________________________________ -->
2960<div class="doc_subsubsection">
Chris Lattnerab21db72009-10-28 00:19:10 +0000[diff] [blame]2961 <a name="i_indirectbr">'<tt>indirectbr</tt>' Instruction</a>
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]2962</div>
2963
2964<div class="doc_text">
2965
2966<h5>Syntax:</h5>
2967<pre>
Chris Lattnerab21db72009-10-28 00:19:10 +0000[diff] [blame]2968 indirectbr &lt;somety&gt;* &lt;address&gt;, [ label &lt;dest1&gt;, label &lt;dest2&gt;, ... ]
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]2969</pre>
2970
2971<h5>Overview:</h5>
2972
Chris Lattnerab21db72009-10-28 00:19:10 +0000[diff] [blame]2973<p>The '<tt>indirectbr</tt>' instruction implements an indirect branch to a label
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]2974 within the current function, whose address is specified by
Chris Lattnerc6f44362009-10-27 21:01:34 +0000[diff] [blame]2975 "<tt>address</tt>". Address must be derived from a <a
2976 href="#blockaddress">blockaddress</a> constant.</p>
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]2977
2978<h5>Arguments:</h5>
2979
2980<p>The '<tt>address</tt>' argument is the address of the label to jump to. The
2981 rest of the arguments indicate the full set of possible destinations that the
2982 address may point to. Blocks are allowed to occur multiple times in the
2983 destination list, though this isn't particularly useful.</p>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]2984
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]2985<p>This destination list is required so that dataflow analysis has an accurate
2986 understanding of the CFG.</p>
2987
2988<h5>Semantics:</h5>
2989
2990<p>Control transfers to the block specified in the address argument. All
2991 possible destination blocks must be listed in the label list, otherwise this
2992 instruction has undefined behavior. This implies that jumps to labels
2993 defined in other functions have undefined behavior as well.</p>
2994
2995<h5>Implementation:</h5>
2996
2997<p>This is typically implemented with a jump through a register.</p>
2998
2999<h5>Example:</h5>
3000<pre>
Chris Lattnerab21db72009-10-28 00:19:10 +0000[diff] [blame]3001 indirectbr i8* %Addr, [ label %bb1, label %bb2, label %bb3 ]
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]3002</pre>
3003
3004</div>
3005
3006
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3007<!-- _______________________________________________________________________ -->
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3008<div class="doc_subsubsection">
3009 <a name="i_invoke">'<tt>invoke</tt>' Instruction</a>
3010</div>
3011
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3012<div class="doc_text">
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3013
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3014<h5>Syntax:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3015<pre>
Devang Patel307e8ab2008-10-07 17:48:33 +0000[diff] [blame]3016 &lt;result&gt; = invoke [<a href="#callingconv">cconv</a>] [<a href="#paramattrs">ret attrs</a>] &lt;ptr to function ty&gt; &lt;function ptr val&gt;(&lt;function args&gt;) [<a href="#fnattrs">fn attrs</a>]
Chris Lattner76b8a332006-05-14 18:23:06 +0000[diff] [blame]3017 to label &lt;normal label&gt; unwind label &lt;exception label&gt;
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3018</pre>
3019
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]3020<h5>Overview:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3021<p>The '<tt>invoke</tt>' instruction causes control to transfer to a specified
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3022 function, with the possibility of control flow transfer to either the
3023 '<tt>normal</tt>' label or the '<tt>exception</tt>' label. If the callee
3024 function returns with the "<tt><a href="#i_ret">ret</a></tt>" instruction,
3025 control flow will return to the "normal" label. If the callee (or any
3026 indirect callees) returns with the "<a href="#i_unwind"><tt>unwind</tt></a>"
3027 instruction, control is interrupted and continued at the dynamically nearest
3028 "exception" label.</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3029
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3030<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3031<p>This instruction requires several arguments:</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3032
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3033<ol>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3034 <li>The optional "cconv" marker indicates which <a href="#callingconv">calling
3035 convention</a> the call should use. If none is specified, the call
3036 defaults to using C calling conventions.</li>
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]3037
3038 <li>The optional <a href="#paramattrs">Parameter Attributes</a> list for
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3039 return values. Only '<tt>zeroext</tt>', '<tt>signext</tt>', and
3040 '<tt>inreg</tt>' attributes are valid here.</li>
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]3041
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3042 <li>'<tt>ptr to function ty</tt>': shall be the signature of the pointer to
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3043 function value being invoked. In most cases, this is a direct function
3044 invocation, but indirect <tt>invoke</tt>s are just as possible, branching
3045 off an arbitrary pointer to function value.</li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3046
3047 <li>'<tt>function ptr val</tt>': An LLVM value containing a pointer to a
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3048 function to be invoked. </li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3049
3050 <li>'<tt>function args</tt>': argument list whose types match the function
Chris Lattner0724fbd2010-03-02 06:36:51 +0000[diff] [blame]3051 signature argument types and parameter attributes. All arguments must be
3052 of <a href="#t_firstclass">first class</a> type. If the function
3053 signature indicates the function accepts a variable number of arguments,
3054 the extra arguments can be specified.</li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3055
3056 <li>'<tt>normal label</tt>': the label reached when the called function
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3057 executes a '<tt><a href="#i_ret">ret</a></tt>' instruction. </li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3058
3059 <li>'<tt>exception label</tt>': the label reached when a callee returns with
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3060 the <a href="#i_unwind"><tt>unwind</tt></a> instruction. </li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3061
Devang Patel307e8ab2008-10-07 17:48:33 +0000[diff] [blame]3062 <li>The optional <a href="#fnattrs">function attributes</a> list. Only
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3063 '<tt>noreturn</tt>', '<tt>nounwind</tt>', '<tt>readonly</tt>' and
3064 '<tt>readnone</tt>' attributes are valid here.</li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3065</ol>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3066
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3067<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3068<p>This instruction is designed to operate as a standard
3069 '<tt><a href="#i_call">call</a></tt>' instruction in most regards. The
3070 primary difference is that it establishes an association with a label, which
3071 is used by the runtime library to unwind the stack.</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3072
3073<p>This instruction is used in languages with destructors to ensure that proper
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3074 cleanup is performed in the case of either a <tt>longjmp</tt> or a thrown
3075 exception. Additionally, this is important for implementation of
3076 '<tt>catch</tt>' clauses in high-level languages that support them.</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3077
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3078<p>For the purposes of the SSA form, the definition of the value returned by the
3079 '<tt>invoke</tt>' instruction is deemed to occur on the edge from the current
3080 block to the "normal" label. If the callee unwinds then no return value is
3081 available.</p>
Dan Gohmanf96a4992009-05-22 21:47:08 +0000[diff] [blame]3082
Chris Lattnerdf7a6802010-01-15 18:08:37 +0000[diff] [blame]3083<p>Note that the code generator does not yet completely support unwind, and
3084that the invoke/unwind semantics are likely to change in future versions.</p>
3085
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3086<h5>Example:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3087<pre>
Nick Lewyckyd703f652008-03-16 07:18:12 +0000[diff] [blame]3088 %retval = invoke i32 @Test(i32 15) to label %Continue
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]3089 unwind label %TestCleanup <i>; {i32}:retval set</i>
Nick Lewyckyd703f652008-03-16 07:18:12 +0000[diff] [blame]3090 %retval = invoke <a href="#callingconv">coldcc</a> i32 %Testfnptr(i32 15) to label %Continue
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]3091 unwind label %TestCleanup <i>; {i32}:retval set</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3092</pre>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3093
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3094</div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3095
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]3096<!-- _______________________________________________________________________ -->
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3097
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3098<div class="doc_subsubsection"> <a name="i_unwind">'<tt>unwind</tt>'
3099Instruction</a> </div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3100
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3101<div class="doc_text">
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3102
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]3103<h5>Syntax:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3104<pre>
3105 unwind
3106</pre>
3107
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]3108<h5>Overview:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3109<p>The '<tt>unwind</tt>' instruction unwinds the stack, continuing control flow
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3110 at the first callee in the dynamic call stack which used
3111 an <a href="#i_invoke"><tt>invoke</tt></a> instruction to perform the call.
3112 This is primarily used to implement exception handling.</p>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3113
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]3114<h5>Semantics:</h5>
Chris Lattner72ed2002008-04-19 21:01:16 +0000[diff] [blame]3115<p>The '<tt>unwind</tt>' instruction causes execution of the current function to
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3116 immediately halt. The dynamic call stack is then searched for the
3117 first <a href="#i_invoke"><tt>invoke</tt></a> instruction on the call stack.
3118 Once found, execution continues at the "exceptional" destination block
3119 specified by the <tt>invoke</tt> instruction. If there is no <tt>invoke</tt>
3120 instruction in the dynamic call chain, undefined behavior results.</p>
3121
Chris Lattnerdf7a6802010-01-15 18:08:37 +0000[diff] [blame]3122<p>Note that the code generator does not yet completely support unwind, and
3123that the invoke/unwind semantics are likely to change in future versions.</p>
3124
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3125</div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3126
3127<!-- _______________________________________________________________________ -->
3128
3129<div class="doc_subsubsection"> <a name="i_unreachable">'<tt>unreachable</tt>'
3130Instruction</a> </div>
3131
3132<div class="doc_text">
3133
3134<h5>Syntax:</h5>
3135<pre>
3136 unreachable
3137</pre>
3138
3139<h5>Overview:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3140<p>The '<tt>unreachable</tt>' instruction has no defined semantics. This
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3141 instruction is used to inform the optimizer that a particular portion of the
3142 code is not reachable. This can be used to indicate that the code after a
3143 no-return function cannot be reached, and other facts.</p>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3144
3145<h5>Semantics:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3146<p>The '<tt>unreachable</tt>' instruction has no defined semantics.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3147
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3148</div>
3149
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3150<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3151<div class="doc_subsection"> <a name="binaryops">Binary Operations</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3152
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3153<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3154
3155<p>Binary operators are used to do most of the computation in a program. They
3156 require two operands of the same type, execute an operation on them, and
3157 produce a single value. The operands might represent multiple data, as is
3158 the case with the <a href="#t_vector">vector</a> data type. The result value
3159 has the same type as its operands.</p>
3160
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3161<p>There are several different binary operators:</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3162
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3163</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3164
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3165<!-- _______________________________________________________________________ -->
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3166<div class="doc_subsubsection">
3167 <a name="i_add">'<tt>add</tt>' Instruction</a>
3168</div>
3169
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3170<div class="doc_text">
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3171
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3172<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3173<pre>
Dan Gohmancbb38f22009-07-20 22:41:19 +0000[diff] [blame]3174 &lt;result&gt; = add &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanfdfca792009-09-02 17:31:42 +0000[diff] [blame]3175 &lt;result&gt; = add nuw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3176 &lt;result&gt; = add nsw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3177 &lt;result&gt; = add nuw nsw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3178</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3179
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3180<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3181<p>The '<tt>add</tt>' instruction returns the sum of its two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3182
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3183<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3184<p>The two arguments to the '<tt>add</tt>' instruction must
3185 be <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of
3186 integer values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3187
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3188<h5>Semantics:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3189<p>The value produced is the integer sum of the two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3190
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3191<p>If the sum has unsigned overflow, the result returned is the mathematical
3192 result modulo 2<sup>n</sup>, where n is the bit width of the result.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3193
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3194<p>Because LLVM integers use a two's complement representation, this instruction
3195 is appropriate for both signed and unsigned integers.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3196
Dan Gohman08d012e2009-07-22 22:44:56 +0000[diff] [blame]3197<p><tt>nuw</tt> and <tt>nsw</tt> stand for &quot;No Unsigned Wrap&quot;
3198 and &quot;No Signed Wrap&quot;, respectively. If the <tt>nuw</tt> and/or
3199 <tt>nsw</tt> keywords are present, the result value of the <tt>add</tt>
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]3200 is a <a href="#trapvalues">trap value</a> if unsigned and/or signed overflow,
3201 respectively, occurs.</p>
Dan Gohmancbb38f22009-07-20 22:41:19 +0000[diff] [blame]3202
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3203<h5>Example:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3204<pre>
3205 &lt;result&gt; = add i32 4, %var <i>; yields {i32}:result = 4 + %var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3206</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3207
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3208</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3209
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3210<!-- _______________________________________________________________________ -->
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3211<div class="doc_subsubsection">
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3212 <a name="i_fadd">'<tt>fadd</tt>' Instruction</a>
3213</div>
3214
3215<div class="doc_text">
3216
3217<h5>Syntax:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3218<pre>
3219 &lt;result&gt; = fadd &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3220</pre>
3221
3222<h5>Overview:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3223<p>The '<tt>fadd</tt>' instruction returns the sum of its two operands.</p>
3224
3225<h5>Arguments:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3226<p>The two arguments to the '<tt>fadd</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3227 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
3228 floating point values. Both arguments must have identical types.</p>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3229
3230<h5>Semantics:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3231<p>The value produced is the floating point sum of the two operands.</p>
3232
3233<h5>Example:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3234<pre>
3235 &lt;result&gt; = fadd float 4.0, %var <i>; yields {float}:result = 4.0 + %var</i>
3236</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3237
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3238</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3239
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3240<!-- _______________________________________________________________________ -->
3241<div class="doc_subsubsection">
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3242 <a name="i_sub">'<tt>sub</tt>' Instruction</a>
3243</div>
3244
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3245<div class="doc_text">
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3246
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3247<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3248<pre>
Dan Gohman08d012e2009-07-22 22:44:56 +0000[diff] [blame]3249 &lt;result&gt; = sub &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanfdfca792009-09-02 17:31:42 +0000[diff] [blame]3250 &lt;result&gt; = sub nuw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3251 &lt;result&gt; = sub nsw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3252 &lt;result&gt; = sub nuw nsw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3253</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3254
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3255<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3256<p>The '<tt>sub</tt>' instruction returns the difference of its two
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3257 operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3258
3259<p>Note that the '<tt>sub</tt>' instruction is used to represent the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3260 '<tt>neg</tt>' instruction present in most other intermediate
3261 representations.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3262
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3263<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3264<p>The two arguments to the '<tt>sub</tt>' instruction must
3265 be <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of
3266 integer values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3267
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3268<h5>Semantics:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3269<p>The value produced is the integer difference of the two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3270
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3271<p>If the difference has unsigned overflow, the result returned is the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3272 mathematical result modulo 2<sup>n</sup>, where n is the bit width of the
3273 result.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3274
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3275<p>Because LLVM integers use a two's complement representation, this instruction
3276 is appropriate for both signed and unsigned integers.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3277
Dan Gohman08d012e2009-07-22 22:44:56 +0000[diff] [blame]3278<p><tt>nuw</tt> and <tt>nsw</tt> stand for &quot;No Unsigned Wrap&quot;
3279 and &quot;No Signed Wrap&quot;, respectively. If the <tt>nuw</tt> and/or
3280 <tt>nsw</tt> keywords are present, the result value of the <tt>sub</tt>
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]3281 is a <a href="#trapvalues">trap value</a> if unsigned and/or signed overflow,
3282 respectively, occurs.</p>
Dan Gohmancbb38f22009-07-20 22:41:19 +0000[diff] [blame]3283
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3284<h5>Example:</h5>
Bill Wendlingaac388b2007-05-29 09:42:13 +0000[diff] [blame]3285<pre>
3286 &lt;result&gt; = sub i32 4, %var <i>; yields {i32}:result = 4 - %var</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3287 &lt;result&gt; = sub i32 0, %val <i>; yields {i32}:result = -%var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3288</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3289
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3290</div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3291
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3292<!-- _______________________________________________________________________ -->
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3293<div class="doc_subsubsection">
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3294 <a name="i_fsub">'<tt>fsub</tt>' Instruction</a>
3295</div>
3296
3297<div class="doc_text">
3298
3299<h5>Syntax:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3300<pre>
3301 &lt;result&gt; = fsub &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3302</pre>
3303
3304<h5>Overview:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3305<p>The '<tt>fsub</tt>' instruction returns the difference of its two
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3306 operands.</p>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3307
3308<p>Note that the '<tt>fsub</tt>' instruction is used to represent the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3309 '<tt>fneg</tt>' instruction present in most other intermediate
3310 representations.</p>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3311
3312<h5>Arguments:</h5>
Bill Wendlingd9fe2982009-07-20 02:32:41 +0000[diff] [blame]3313<p>The two arguments to the '<tt>fsub</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3314 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
3315 floating point values. Both arguments must have identical types.</p>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3316
3317<h5>Semantics:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3318<p>The value produced is the floating point difference of the two operands.</p>
3319
3320<h5>Example:</h5>
3321<pre>
3322 &lt;result&gt; = fsub float 4.0, %var <i>; yields {float}:result = 4.0 - %var</i>
3323 &lt;result&gt; = fsub float -0.0, %val <i>; yields {float}:result = -%var</i>
3324</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3325
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3326</div>
3327
3328<!-- _______________________________________________________________________ -->
3329<div class="doc_subsubsection">
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3330 <a name="i_mul">'<tt>mul</tt>' Instruction</a>
3331</div>
3332
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3333<div class="doc_text">
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3334
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3335<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3336<pre>
Dan Gohman08d012e2009-07-22 22:44:56 +0000[diff] [blame]3337 &lt;result&gt; = mul &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanfdfca792009-09-02 17:31:42 +0000[diff] [blame]3338 &lt;result&gt; = mul nuw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3339 &lt;result&gt; = mul nsw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3340 &lt;result&gt; = mul nuw nsw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3341</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3342
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3343<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3344<p>The '<tt>mul</tt>' instruction returns the product of its two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3345
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3346<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3347<p>The two arguments to the '<tt>mul</tt>' instruction must
3348 be <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of
3349 integer values. Both arguments must have identical types.</p>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3350
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3351<h5>Semantics:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3352<p>The value produced is the integer product of the two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3353
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3354<p>If the result of the multiplication has unsigned overflow, the result
3355 returned is the mathematical result modulo 2<sup>n</sup>, where n is the bit
3356 width of the result.</p>
3357
3358<p>Because LLVM integers use a two's complement representation, and the result
3359 is the same width as the operands, this instruction returns the correct
3360 result for both signed and unsigned integers. If a full product
3361 (e.g. <tt>i32</tt>x<tt>i32</tt>-><tt>i64</tt>) is needed, the operands should
3362 be sign-extended or zero-extended as appropriate to the width of the full
3363 product.</p>
3364
Dan Gohman08d012e2009-07-22 22:44:56 +0000[diff] [blame]3365<p><tt>nuw</tt> and <tt>nsw</tt> stand for &quot;No Unsigned Wrap&quot;
3366 and &quot;No Signed Wrap&quot;, respectively. If the <tt>nuw</tt> and/or
3367 <tt>nsw</tt> keywords are present, the result value of the <tt>mul</tt>
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]3368 is a <a href="#trapvalues">trap value</a> if unsigned and/or signed overflow,
3369 respectively, occurs.</p>
Dan Gohmancbb38f22009-07-20 22:41:19 +0000[diff] [blame]3370
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3371<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3372<pre>
3373 &lt;result&gt; = mul i32 4, %var <i>; yields {i32}:result = 4 * %var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3374</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3375
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3376</div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3377
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3378<!-- _______________________________________________________________________ -->
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3379<div class="doc_subsubsection">
3380 <a name="i_fmul">'<tt>fmul</tt>' Instruction</a>
3381</div>
3382
3383<div class="doc_text">
3384
3385<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3386<pre>
3387 &lt;result&gt; = fmul &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3388</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3389
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3390<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3391<p>The '<tt>fmul</tt>' instruction returns the product of its two operands.</p>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3392
3393<h5>Arguments:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3394<p>The two arguments to the '<tt>fmul</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3395 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
3396 floating point values. Both arguments must have identical types.</p>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3397
3398<h5>Semantics:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3399<p>The value produced is the floating point product of the two operands.</p>
3400
3401<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3402<pre>
3403 &lt;result&gt; = fmul float 4.0, %var <i>; yields {float}:result = 4.0 * %var</i>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3404</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3405
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3406</div>
3407
3408<!-- _______________________________________________________________________ -->
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3409<div class="doc_subsubsection"> <a name="i_udiv">'<tt>udiv</tt>' Instruction
3410</a></div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3411
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3412<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3413
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3414<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3415<pre>
3416 &lt;result&gt; = udiv &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3417</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3418
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3419<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3420<p>The '<tt>udiv</tt>' instruction returns the quotient of its two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3421
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3422<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3423<p>The two arguments to the '<tt>udiv</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3424 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3425 values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3426
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3427<h5>Semantics:</h5>
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3428<p>The value produced is the unsigned integer quotient of the two operands.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3429
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3430<p>Note that unsigned integer division and signed integer division are distinct
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3431 operations; for signed integer division, use '<tt>sdiv</tt>'.</p>
3432
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3433<p>Division by zero leads to undefined behavior.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3434
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3435<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3436<pre>
3437 &lt;result&gt; = udiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3438</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3439
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3440</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3441
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3442<!-- _______________________________________________________________________ -->
3443<div class="doc_subsubsection"> <a name="i_sdiv">'<tt>sdiv</tt>' Instruction
3444</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3445
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3446<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3447
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3448<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3449<pre>
Dan Gohmancbb38f22009-07-20 22:41:19 +0000[diff] [blame]3450 &lt;result&gt; = sdiv &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanfdfca792009-09-02 17:31:42 +0000[diff] [blame]3451 &lt;result&gt; = sdiv exact &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3452</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3453
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3454<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3455<p>The '<tt>sdiv</tt>' instruction returns the quotient of its two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3456
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3457<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3458<p>The two arguments to the '<tt>sdiv</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3459 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3460 values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3461
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3462<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3463<p>The value produced is the signed integer quotient of the two operands rounded
3464 towards zero.</p>
3465
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3466<p>Note that signed integer division and unsigned integer division are distinct
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3467 operations; for unsigned integer division, use '<tt>udiv</tt>'.</p>
3468
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3469<p>Division by zero leads to undefined behavior. Overflow also leads to
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3470 undefined behavior; this is a rare case, but can occur, for example, by doing
3471 a 32-bit division of -2147483648 by -1.</p>
3472
Dan Gohman9c5beed2009-07-22 00:04:19 +0000[diff] [blame]3473<p>If the <tt>exact</tt> keyword is present, the result value of the
Dan Gohman27ef9972010-04-23 15:23:32 +0000[diff] [blame]3474 <tt>sdiv</tt> is a <a href="#trapvalues">trap value</a> if the result would
Dan Gohman38da9272010-07-11 00:08:34 +0000[diff] [blame]3475 be rounded.</p>
Dan Gohmancbb38f22009-07-20 22:41:19 +0000[diff] [blame]3476
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3477<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3478<pre>
3479 &lt;result&gt; = sdiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3480</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3481
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3482</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3483
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3484<!-- _______________________________________________________________________ -->
3485<div class="doc_subsubsection"> <a name="i_fdiv">'<tt>fdiv</tt>'
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3486Instruction</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3487
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3488<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3489
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3490<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3491<pre>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]3492 &lt;result&gt; = fdiv &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3493</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3494
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3495<h5>Overview:</h5>
3496<p>The '<tt>fdiv</tt>' instruction returns the quotient of its two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3497
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3498<h5>Arguments:</h5>
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]3499<p>The two arguments to the '<tt>fdiv</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3500 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
3501 floating point values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3502
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3503<h5>Semantics:</h5>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3504<p>The value produced is the floating point quotient of the two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3505
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3506<h5>Example:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3507<pre>
3508 &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]3509</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3510
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3511</div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3512
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3513<!-- _______________________________________________________________________ -->
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3514<div class="doc_subsubsection"> <a name="i_urem">'<tt>urem</tt>' Instruction</a>
3515</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3516
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3517<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3518
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3519<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3520<pre>
3521 &lt;result&gt; = urem &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3522</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3523
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3524<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3525<p>The '<tt>urem</tt>' instruction returns the remainder from the unsigned
3526 division of its two arguments.</p>
3527
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3528<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3529<p>The two arguments to the '<tt>urem</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3530 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3531 values. Both arguments must have identical types.</p>
3532
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3533<h5>Semantics:</h5>
3534<p>This instruction returns the unsigned integer <i>remainder</i> of a division.
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3535 This instruction always performs an unsigned division to get the
3536 remainder.</p>
3537
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3538<p>Note that unsigned integer remainder and signed integer remainder are
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3539 distinct operations; for signed integer remainder, use '<tt>srem</tt>'.</p>
3540
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3541<p>Taking the remainder of a division by zero leads to undefined behavior.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3542
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3543<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3544<pre>
3545 &lt;result&gt; = urem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3546</pre>
3547
3548</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3549
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3550<!-- _______________________________________________________________________ -->
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3551<div class="doc_subsubsection">
3552 <a name="i_srem">'<tt>srem</tt>' Instruction</a>
3553</div>
3554
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3555<div class="doc_text">
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3556
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3557<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3558<pre>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]3559 &lt;result&gt; = srem &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3560</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3561
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3562<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3563<p>The '<tt>srem</tt>' instruction returns the remainder from the signed
3564 division of its two operands. This instruction can also take
3565 <a href="#t_vector">vector</a> versions of the values in which case the
3566 elements must be integers.</p>
Chris Lattnerc7d3ab32008-01-04 04:33:49 +0000[diff] [blame]3567
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3568<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3569<p>The two arguments to the '<tt>srem</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3570 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3571 values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3572
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3573<h5>Semantics:</h5>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3574<p>This instruction returns the <i>remainder</i> of a division (where the result
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3575 has the same sign as the dividend, <tt>op1</tt>), not the <i>modulo</i>
3576 operator (where the result has the same sign as the divisor, <tt>op2</tt>) of
3577 a value. For more information about the difference,
3578 see <a href="http://mathforum.org/dr.math/problems/anne.4.28.99.html">The
3579 Math Forum</a>. For a table of how this is implemented in various languages,
3580 please see <a href="http://en.wikipedia.org/wiki/Modulo_operation">
3581 Wikipedia: modulo operation</a>.</p>
3582
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3583<p>Note that signed integer remainder and unsigned integer remainder are
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3584 distinct operations; for unsigned integer remainder, use '<tt>urem</tt>'.</p>
3585
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3586<p>Taking the remainder of a division by zero leads to undefined behavior.
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3587 Overflow also leads to undefined behavior; this is a rare case, but can
3588 occur, for example, by taking the remainder of a 32-bit division of
3589 -2147483648 by -1. (The remainder doesn't actually overflow, but this rule
3590 lets srem be implemented using instructions that return both the result of
3591 the division and the remainder.)</p>
3592
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3593<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3594<pre>
3595 &lt;result&gt; = srem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3596</pre>
3597
3598</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3599
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3600<!-- _______________________________________________________________________ -->
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3601<div class="doc_subsubsection">
3602 <a name="i_frem">'<tt>frem</tt>' Instruction</a> </div>
3603
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3604<div class="doc_text">
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3605
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3606<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3607<pre>
3608 &lt;result&gt; = frem &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3609</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3610
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3611<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3612<p>The '<tt>frem</tt>' instruction returns the remainder from the division of
3613 its two operands.</p>
3614
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3615<h5>Arguments:</h5>
3616<p>The two arguments to the '<tt>frem</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3617 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
3618 floating point values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3619
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3620<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3621<p>This instruction returns the <i>remainder</i> of a division. The remainder
3622 has the same sign as the dividend.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3623
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3624<h5>Example:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3625<pre>
3626 &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]3627</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3628
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3629</div>
Robert Bocchino7b81c752006-02-17 21:18:08 +0000[diff] [blame]3630
Reid Spencer8e11bf82007-02-02 13:57:07 +0000[diff] [blame]3631<!-- ======================================================================= -->
3632<div class="doc_subsection"> <a name="bitwiseops">Bitwise Binary
3633Operations</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3634
Reid Spencer8e11bf82007-02-02 13:57:07 +0000[diff] [blame]3635<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3636
3637<p>Bitwise binary operators are used to do various forms of bit-twiddling in a
3638 program. They are generally very efficient instructions and can commonly be
3639 strength reduced from other instructions. They require two operands of the
3640 same type, execute an operation on them, and produce a single value. The
3641 resulting value is the same type as its operands.</p>
3642
Reid Spencer8e11bf82007-02-02 13:57:07 +0000[diff] [blame]3643</div>
3644
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3645<!-- _______________________________________________________________________ -->
3646<div class="doc_subsubsection"> <a name="i_shl">'<tt>shl</tt>'
3647Instruction</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3648
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3649<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3650
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3651<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3652<pre>
3653 &lt;result&gt; = shl &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3654</pre>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]3655
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3656<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3657<p>The '<tt>shl</tt>' instruction returns the first operand shifted to the left
3658 a specified number of bits.</p>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]3659
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3660<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3661<p>Both arguments to the '<tt>shl</tt>' instruction must be the
3662 same <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of
3663 integer type. '<tt>op2</tt>' is treated as an unsigned value.</p>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3664
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3665<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3666<p>The value produced is <tt>op1</tt> * 2<sup><tt>op2</tt></sup> mod
3667 2<sup>n</sup>, where <tt>n</tt> is the width of the result. If <tt>op2</tt>
3668 is (statically or dynamically) negative or equal to or larger than the number
3669 of bits in <tt>op1</tt>, the result is undefined. If the arguments are
3670 vectors, each vector element of <tt>op1</tt> is shifted by the corresponding
3671 shift amount in <tt>op2</tt>.</p>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]3672
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3673<h5>Example:</h5>
3674<pre>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3675 &lt;result&gt; = shl i32 4, %var <i>; yields {i32}: 4 &lt;&lt; %var</i>
3676 &lt;result&gt; = shl i32 4, 2 <i>; yields {i32}: 16</i>
3677 &lt;result&gt; = shl i32 1, 10 <i>; yields {i32}: 1024</i>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]3678 &lt;result&gt; = shl i32 1, 32 <i>; undefined</i>
Mon P Wange9f10152008-12-09 05:46:39 +0000[diff] [blame]3679 &lt;result&gt; = shl &lt;2 x i32&gt; &lt; i32 1, i32 1&gt;, &lt; i32 1, i32 2&gt; <i>; yields: result=&lt;2 x i32&gt; &lt; i32 2, i32 4&gt;</i>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3680</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3681
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3682</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3683
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3684<!-- _______________________________________________________________________ -->
3685<div class="doc_subsubsection"> <a name="i_lshr">'<tt>lshr</tt>'
3686Instruction</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3687
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3688<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3689
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3690<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3691<pre>
3692 &lt;result&gt; = lshr &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3693</pre>
3694
3695<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3696<p>The '<tt>lshr</tt>' instruction (logical shift right) returns the first
3697 operand shifted to the right a specified number of bits with zero fill.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3698
3699<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3700<p>Both arguments to the '<tt>lshr</tt>' instruction must be the same
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3701 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3702 type. '<tt>op2</tt>' is treated as an unsigned value.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3703
3704<h5>Semantics:</h5>
3705<p>This instruction always performs a logical shift right operation. The most
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3706 significant bits of the result will be filled with zero bits after the shift.
3707 If <tt>op2</tt> is (statically or dynamically) equal to or larger than the
3708 number of bits in <tt>op1</tt>, the result is undefined. If the arguments are
3709 vectors, each vector element of <tt>op1</tt> is shifted by the corresponding
3710 shift amount in <tt>op2</tt>.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3711
3712<h5>Example:</h5>
3713<pre>
3714 &lt;result&gt; = lshr i32 4, 1 <i>; yields {i32}:result = 2</i>
3715 &lt;result&gt; = lshr i32 4, 2 <i>; yields {i32}:result = 1</i>
3716 &lt;result&gt; = lshr i8 4, 3 <i>; yields {i8}:result = 0</i>
3717 &lt;result&gt; = lshr i8 -2, 1 <i>; yields {i8}:result = 0x7FFFFFFF </i>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]3718 &lt;result&gt; = lshr i32 1, 32 <i>; undefined</i>
Mon P Wange9f10152008-12-09 05:46:39 +0000[diff] [blame]3719 &lt;result&gt; = lshr &lt;2 x i32&gt; &lt; i32 -2, i32 4&gt;, &lt; i32 1, i32 2&gt; <i>; yields: result=&lt;2 x i32&gt; &lt; i32 0x7FFFFFFF, i32 1&gt;</i>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3720</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3721
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3722</div>
3723
Reid Spencer8e11bf82007-02-02 13:57:07 +0000[diff] [blame]3724<!-- _______________________________________________________________________ -->
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3725<div class="doc_subsubsection"> <a name="i_ashr">'<tt>ashr</tt>'
3726Instruction</a> </div>
3727<div class="doc_text">
3728
3729<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3730<pre>
3731 &lt;result&gt; = ashr &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3732</pre>
3733
3734<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3735<p>The '<tt>ashr</tt>' instruction (arithmetic shift right) returns the first
3736 operand shifted to the right a specified number of bits with sign
3737 extension.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3738
3739<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3740<p>Both arguments to the '<tt>ashr</tt>' instruction must be the same
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3741 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3742 type. '<tt>op2</tt>' is treated as an unsigned value.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3743
3744<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3745<p>This instruction always performs an arithmetic shift right operation, The
3746 most significant bits of the result will be filled with the sign bit
3747 of <tt>op1</tt>. If <tt>op2</tt> is (statically or dynamically) equal to or
3748 larger than the number of bits in <tt>op1</tt>, the result is undefined. If
3749 the arguments are vectors, each vector element of <tt>op1</tt> is shifted by
3750 the corresponding shift amount in <tt>op2</tt>.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3751
3752<h5>Example:</h5>
3753<pre>
3754 &lt;result&gt; = ashr i32 4, 1 <i>; yields {i32}:result = 2</i>
3755 &lt;result&gt; = ashr i32 4, 2 <i>; yields {i32}:result = 1</i>
3756 &lt;result&gt; = ashr i8 4, 3 <i>; yields {i8}:result = 0</i>
3757 &lt;result&gt; = ashr i8 -2, 1 <i>; yields {i8}:result = -1</i>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]3758 &lt;result&gt; = ashr i32 1, 32 <i>; undefined</i>
Mon P Wange9f10152008-12-09 05:46:39 +0000[diff] [blame]3759 &lt;result&gt; = ashr &lt;2 x i32&gt; &lt; i32 -2, i32 4&gt;, &lt; i32 1, i32 3&gt; <i>; yields: result=&lt;2 x i32&gt; &lt; i32 -1, i32 0&gt;</i>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3760</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3761
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3762</div>
3763
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3764<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3765<div class="doc_subsubsection"> <a name="i_and">'<tt>and</tt>'
3766Instruction</a> </div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3767
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3768<div class="doc_text">
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3769
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3770<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3771<pre>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]3772 &lt;result&gt; = and &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3773</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3774
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3775<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3776<p>The '<tt>and</tt>' instruction returns the bitwise logical and of its two
3777 operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3778
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3779<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3780<p>The two arguments to the '<tt>and</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3781 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3782 values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3783
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3784<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3785<p>The truth table used for the '<tt>and</tt>' instruction is:</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3786
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3787<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3788 <tbody>
3789 <tr>
3790 <td>In0</td>
3791 <td>In1</td>
3792 <td>Out</td>
3793 </tr>
3794 <tr>
3795 <td>0</td>
3796 <td>0</td>
3797 <td>0</td>
3798 </tr>
3799 <tr>
3800 <td>0</td>
3801 <td>1</td>
3802 <td>0</td>
3803 </tr>
3804 <tr>
3805 <td>1</td>
3806 <td>0</td>
3807 <td>0</td>
3808 </tr>
3809 <tr>
3810 <td>1</td>
3811 <td>1</td>
3812 <td>1</td>
3813 </tr>
3814 </tbody>
3815</table>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3816
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3817<h5>Example:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3818<pre>
3819 &lt;result&gt; = and i32 4, %var <i>; yields {i32}:result = 4 &amp; %var</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3820 &lt;result&gt; = and i32 15, 40 <i>; yields {i32}:result = 8</i>
3821 &lt;result&gt; = and i32 4, 8 <i>; yields {i32}:result = 0</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3822</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3823</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3824<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3825<div class="doc_subsubsection"> <a name="i_or">'<tt>or</tt>' Instruction</a> </div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3826
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3827<div class="doc_text">
3828
3829<h5>Syntax:</h5>
3830<pre>
3831 &lt;result&gt; = or &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3832</pre>
3833
3834<h5>Overview:</h5>
3835<p>The '<tt>or</tt>' instruction returns the bitwise logical inclusive or of its
3836 two operands.</p>
3837
3838<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3839<p>The two arguments to the '<tt>or</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3840 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3841 values. Both arguments must have identical types.</p>
3842
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3843<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3844<p>The truth table used for the '<tt>or</tt>' instruction is:</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3845
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3846<table border="1" cellspacing="0" cellpadding="4">
3847 <tbody>
3848 <tr>
3849 <td>In0</td>
3850 <td>In1</td>
3851 <td>Out</td>
3852 </tr>
3853 <tr>
3854 <td>0</td>
3855 <td>0</td>
3856 <td>0</td>
3857 </tr>
3858 <tr>
3859 <td>0</td>
3860 <td>1</td>
3861 <td>1</td>
3862 </tr>
3863 <tr>
3864 <td>1</td>
3865 <td>0</td>
3866 <td>1</td>
3867 </tr>
3868 <tr>
3869 <td>1</td>
3870 <td>1</td>
3871 <td>1</td>
3872 </tr>
3873 </tbody>
3874</table>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3875
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3876<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3877<pre>
3878 &lt;result&gt; = or i32 4, %var <i>; yields {i32}:result = 4 | %var</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3879 &lt;result&gt; = or i32 15, 40 <i>; yields {i32}:result = 47</i>
3880 &lt;result&gt; = or i32 4, 8 <i>; yields {i32}:result = 12</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3881</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3882
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3883</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3884
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3885<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3886<div class="doc_subsubsection"> <a name="i_xor">'<tt>xor</tt>'
3887Instruction</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3888
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3889<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3890
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3891<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3892<pre>
3893 &lt;result&gt; = xor &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3894</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3895
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3896<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3897<p>The '<tt>xor</tt>' instruction returns the bitwise logical exclusive or of
3898 its two operands. The <tt>xor</tt> is used to implement the "one's
3899 complement" operation, which is the "~" operator in C.</p>
3900
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3901<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3902<p>The two arguments to the '<tt>xor</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3903 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3904 values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3905
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3906<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3907<p>The truth table used for the '<tt>xor</tt>' instruction is:</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3908
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3909<table border="1" cellspacing="0" cellpadding="4">
3910 <tbody>
3911 <tr>
3912 <td>In0</td>
3913 <td>In1</td>
3914 <td>Out</td>
3915 </tr>
3916 <tr>
3917 <td>0</td>
3918 <td>0</td>
3919 <td>0</td>
3920 </tr>
3921 <tr>
3922 <td>0</td>
3923 <td>1</td>
3924 <td>1</td>
3925 </tr>
3926 <tr>
3927 <td>1</td>
3928 <td>0</td>
3929 <td>1</td>
3930 </tr>
3931 <tr>
3932 <td>1</td>
3933 <td>1</td>
3934 <td>0</td>
3935 </tr>
3936 </tbody>
3937</table>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3938
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3939<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3940<pre>
3941 &lt;result&gt; = xor i32 4, %var <i>; yields {i32}:result = 4 ^ %var</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3942 &lt;result&gt; = xor i32 15, 40 <i>; yields {i32}:result = 39</i>
3943 &lt;result&gt; = xor i32 4, 8 <i>; yields {i32}:result = 12</i>
3944 &lt;result&gt; = xor i32 %V, -1 <i>; yields {i32}:result = ~%V</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3945</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3946
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3947</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3948
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3949<!-- ======================================================================= -->
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3950<div class="doc_subsection">
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3951 <a name="vectorops">Vector Operations</a>
3952</div>
3953
3954<div class="doc_text">
3955
3956<p>LLVM supports several instructions to represent vector operations in a
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3957 target-independent manner. These instructions cover the element-access and
3958 vector-specific operations needed to process vectors effectively. While LLVM
3959 does directly support these vector operations, many sophisticated algorithms
3960 will want to use target-specific intrinsics to take full advantage of a
3961 specific target.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3962
3963</div>
3964
3965<!-- _______________________________________________________________________ -->
3966<div class="doc_subsubsection">
3967 <a name="i_extractelement">'<tt>extractelement</tt>' Instruction</a>
3968</div>
3969
3970<div class="doc_text">
3971
3972<h5>Syntax:</h5>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3973<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3974 &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]3975</pre>
3976
3977<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3978<p>The '<tt>extractelement</tt>' instruction extracts a single scalar element
3979 from a vector at a specified index.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3980
3981
3982<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3983<p>The first operand of an '<tt>extractelement</tt>' instruction is a value
3984 of <a href="#t_vector">vector</a> type. The second operand is an index
3985 indicating the position from which to extract the element. The index may be
3986 a variable.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3987
3988<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3989<p>The result is a scalar of the same type as the element type of
3990 <tt>val</tt>. Its value is the value at position <tt>idx</tt> of
3991 <tt>val</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
3992 results are undefined.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3993
3994<h5>Example:</h5>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3995<pre>
Gabor Greifa5b6f452009-10-28 13:14:50 +0000[diff] [blame]3996 &lt;result&gt; = extractelement &lt;4 x i32&gt; %vec, i32 0 <i>; yields i32</i>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3997</pre>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3998
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3999</div>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4000
4001<!-- _______________________________________________________________________ -->
4002<div class="doc_subsubsection">
4003 <a name="i_insertelement">'<tt>insertelement</tt>' Instruction</a>
4004</div>
4005
4006<div class="doc_text">
4007
4008<h5>Syntax:</h5>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4009<pre>
Dan Gohmanf3480b92008-05-12 23:38:42 +0000[diff] [blame]4010 &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]4011</pre>
4012
4013<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4014<p>The '<tt>insertelement</tt>' instruction inserts a scalar element into a
4015 vector at a specified index.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4016
4017<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4018<p>The first operand of an '<tt>insertelement</tt>' instruction is a value
4019 of <a href="#t_vector">vector</a> type. The second operand is a scalar value
4020 whose type must equal the element type of the first operand. The third
4021 operand is an index indicating the position at which to insert the value.
4022 The index may be a variable.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4023
4024<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4025<p>The result is a vector of the same type as <tt>val</tt>. Its element values
4026 are those of <tt>val</tt> except at position <tt>idx</tt>, where it gets the
4027 value <tt>elt</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
4028 results are undefined.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4029
4030<h5>Example:</h5>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4031<pre>
Gabor Greifa5b6f452009-10-28 13:14:50 +0000[diff] [blame]4032 &lt;result&gt; = 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]4033</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4034
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4035</div>
4036
4037<!-- _______________________________________________________________________ -->
4038<div class="doc_subsubsection">
4039 <a name="i_shufflevector">'<tt>shufflevector</tt>' Instruction</a>
4040</div>
4041
4042<div class="doc_text">
4043
4044<h5>Syntax:</h5>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4045<pre>
Mon P Wangaeb06d22008-11-10 04:46:22 +0000[diff] [blame]4046 &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;m x i32&gt; &lt;mask&gt; <i>; yields &lt;m x &lt;ty&gt;&gt;</i>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4047</pre>
4048
4049<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4050<p>The '<tt>shufflevector</tt>' instruction constructs a permutation of elements
4051 from two input vectors, returning a vector with the same element type as the
4052 input and length that is the same as the shuffle mask.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4053
4054<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4055<p>The first two operands of a '<tt>shufflevector</tt>' instruction are vectors
4056 with types that match each other. The third argument is a shuffle mask whose
4057 element type is always 'i32'. The result of the instruction is a vector
4058 whose length is the same as the shuffle mask and whose element type is the
4059 same as the element type of the first two operands.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4060
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4061<p>The shuffle mask operand is required to be a constant vector with either
4062 constant integer or undef values.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4063
4064<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4065<p>The elements of the two input vectors are numbered from left to right across
4066 both of the vectors. The shuffle mask operand specifies, for each element of
4067 the result vector, which element of the two input vectors the result element
4068 gets. The element selector may be undef (meaning "don't care") and the
4069 second operand may be undef if performing a shuffle from only one vector.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4070
4071<h5>Example:</h5>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4072<pre>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4073 &lt;result&gt; = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; %v2,
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]4074 &lt;4 x i32&gt; &lt;i32 0, i32 4, i32 1, i32 5&gt; <i>; yields &lt;4 x i32&gt;</i>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4075 &lt;result&gt; = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; undef,
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4076 &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.
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4077 &lt;result&gt; = shufflevector &lt;8 x i32&gt; %v1, &lt;8 x i32&gt; undef,
Mon P Wangaeb06d22008-11-10 04:46:22 +0000[diff] [blame]4078 &lt;4 x i32&gt; &lt;i32 0, i32 1, i32 2, i32 3&gt; <i>; yields &lt;4 x i32&gt;</i>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4079 &lt;result&gt; = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; %v2,
Mon P Wangaeb06d22008-11-10 04:46:22 +0000[diff] [blame]4080 &lt;8 x i32&gt; &lt;i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7 &gt; <i>; yields &lt;8 x i32&gt;</i>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4081</pre>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4082
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4083</div>
Tanya Lattner09474292006-04-14 19:24:33 +0000[diff] [blame]4084
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4085<!-- ======================================================================= -->
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4086<div class="doc_subsection">
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4087 <a name="aggregateops">Aggregate Operations</a>
4088</div>
4089
4090<div class="doc_text">
4091
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4092<p>LLVM supports several instructions for working with
4093 <a href="#t_aggregate">aggregate</a> values.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4094
4095</div>
4096
4097<!-- _______________________________________________________________________ -->
4098<div class="doc_subsubsection">
4099 <a name="i_extractvalue">'<tt>extractvalue</tt>' Instruction</a>
4100</div>
4101
4102<div class="doc_text">
4103
4104<h5>Syntax:</h5>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4105<pre>
4106 &lt;result&gt; = extractvalue &lt;aggregate type&gt; &lt;val&gt;, &lt;idx&gt;{, &lt;idx&gt;}*
4107</pre>
4108
4109<h5>Overview:</h5>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4110<p>The '<tt>extractvalue</tt>' instruction extracts the value of a member field
4111 from an <a href="#t_aggregate">aggregate</a> value.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4112
4113<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4114<p>The first operand of an '<tt>extractvalue</tt>' instruction is a value
Chris Lattner61c70e92010-08-28 04:09:24 +0000[diff] [blame]4115 of <a href="#t_struct">struct</a> or
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4116 <a href="#t_array">array</a> type. The operands are constant indices to
4117 specify which value to extract in a similar manner as indices in a
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4118 '<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4119
4120<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4121<p>The result is the value at the position in the aggregate specified by the
4122 index operands.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4123
4124<h5>Example:</h5>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4125<pre>
Gabor Greifa5b6f452009-10-28 13:14:50 +0000[diff] [blame]4126 &lt;result&gt; = extractvalue {i32, float} %agg, 0 <i>; yields i32</i>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4127</pre>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4128
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4129</div>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4130
4131<!-- _______________________________________________________________________ -->
4132<div class="doc_subsubsection">
4133 <a name="i_insertvalue">'<tt>insertvalue</tt>' Instruction</a>
4134</div>
4135
4136<div class="doc_text">
4137
4138<h5>Syntax:</h5>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4139<pre>
Jeffrey Yasskin7a088cf2010-01-11 19:19:26 +0000[diff] [blame]4140 &lt;result&gt; = insertvalue &lt;aggregate type&gt; &lt;val&gt;, &lt;ty&gt; &lt;elt&gt;, &lt;idx&gt; <i>; yields &lt;aggregate type&gt;</i>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4141</pre>
4142
4143<h5>Overview:</h5>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4144<p>The '<tt>insertvalue</tt>' instruction inserts a value into a member field
4145 in an <a href="#t_aggregate">aggregate</a> value.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4146
4147<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4148<p>The first operand of an '<tt>insertvalue</tt>' instruction is a value
Chris Lattner61c70e92010-08-28 04:09:24 +0000[diff] [blame]4149 of <a href="#t_struct">struct</a> or
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4150 <a href="#t_array">array</a> type. The second operand is a first-class
4151 value to insert. The following operands are constant indices indicating
4152 the position at which to insert the value in a similar manner as indices in a
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4153 '<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction. The
4154 value to insert must have the same type as the value identified by the
4155 indices.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4156
4157<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4158<p>The result is an aggregate of the same type as <tt>val</tt>. Its value is
4159 that of <tt>val</tt> except that the value at the position specified by the
4160 indices is that of <tt>elt</tt>.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4161
4162<h5>Example:</h5>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4163<pre>
Jeffrey Yasskin7a088cf2010-01-11 19:19:26 +0000[diff] [blame]4164 %agg1 = insertvalue {i32, float} undef, i32 1, 0 <i>; yields {i32 1, float undef}</i>
4165 %agg2 = insertvalue {i32, float} %agg1, float %val, 1 <i>; yields {i32 1, float %val}</i>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4166</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4167
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4168</div>
4169
4170
4171<!-- ======================================================================= -->
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4172<div class="doc_subsection">
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]4173 <a name="memoryops">Memory Access and Addressing Operations</a>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4174</div>
4175
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4176<div class="doc_text">
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4177
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4178<p>A key design point of an SSA-based representation is how it represents
4179 memory. In LLVM, no memory locations are in SSA form, which makes things
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]4180 very simple. This section describes how to read, write, and allocate
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4181 memory in LLVM.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4182
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4183</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4184
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4185<!-- _______________________________________________________________________ -->
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4186<div class="doc_subsubsection">
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4187 <a name="i_alloca">'<tt>alloca</tt>' Instruction</a>
4188</div>
4189
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4190<div class="doc_text">
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4191
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4192<h5>Syntax:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4193<pre>
Dan Gohmanf75a7d32010-05-28 01:14:11 +0000[diff] [blame]4194 &lt;result&gt; = alloca &lt;type&gt;[, &lt;ty&gt; &lt;NumElements&gt;][, align &lt;alignment&gt;] <i>; yields {type*}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4195</pre>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4196
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4197<h5>Overview:</h5>
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]4198<p>The '<tt>alloca</tt>' instruction allocates memory on the stack frame of the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4199 currently executing function, to be automatically released when this function
4200 returns to its caller. The object is always allocated in the generic address
4201 space (address space zero).</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4202
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4203<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4204<p>The '<tt>alloca</tt>' instruction
4205 allocates <tt>sizeof(&lt;type&gt;)*NumElements</tt> bytes of memory on the
4206 runtime stack, returning a pointer of the appropriate type to the program.
4207 If "NumElements" is specified, it is the number of elements allocated,
4208 otherwise "NumElements" is defaulted to be one. If a constant alignment is
4209 specified, the value result of the allocation is guaranteed to be aligned to
4210 at least that boundary. If not specified, or if zero, the target can choose
4211 to align the allocation on any convenient boundary compatible with the
4212 type.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4213
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4214<p>'<tt>type</tt>' may be any sized type.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4215
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4216<h5>Semantics:</h5>
Bill Wendling871eb0a2009-05-08 20:49:29 +0000[diff] [blame]4217<p>Memory is allocated; a pointer is returned. The operation is undefined if
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4218 there is insufficient stack space for the allocation. '<tt>alloca</tt>'d
4219 memory is automatically released when the function returns. The
4220 '<tt>alloca</tt>' instruction is commonly used to represent automatic
4221 variables that must have an address available. When the function returns
4222 (either with the <tt><a href="#i_ret">ret</a></tt>
4223 or <tt><a href="#i_unwind">unwind</a></tt> instructions), the memory is
4224 reclaimed. Allocating zero bytes is legal, but the result is undefined.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4225
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4226<h5>Example:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4227<pre>
Dan Gohman81e21672009-01-04 23:49:44 +0000[diff] [blame]4228 %ptr = alloca i32 <i>; yields {i32*}:ptr</i>
4229 %ptr = alloca i32, i32 4 <i>; yields {i32*}:ptr</i>
4230 %ptr = alloca i32, i32 4, align 1024 <i>; yields {i32*}:ptr</i>
4231 %ptr = alloca i32, align 1024 <i>; yields {i32*}:ptr</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4232</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4233
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4234</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4235
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4236<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4237<div class="doc_subsubsection"> <a name="i_load">'<tt>load</tt>'
4238Instruction</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4239
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4240<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4241
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4242<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4243<pre>
Bill Wendling7c78dbb2010-02-25 21:23:24 +0000[diff] [blame]4244 &lt;result&gt; = load &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;][, !nontemporal !&lt;index&gt;]
4245 &lt;result&gt; = volatile load &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;][, !nontemporal !&lt;index&gt;]
4246 !&lt;index&gt; = !{ i32 1 }
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4247</pre>
4248
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4249<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4250<p>The '<tt>load</tt>' instruction is used to read from memory.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4251
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4252<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4253<p>The argument to the '<tt>load</tt>' instruction specifies the memory address
4254 from which to load. The pointer must point to
4255 a <a href="#t_firstclass">first class</a> type. If the <tt>load</tt> is
4256 marked as <tt>volatile</tt>, then the optimizer is not allowed to modify the
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]4257 number or order of execution of this <tt>load</tt> with other <a
4258 href="#volatile">volatile operations</a>.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4259
Bill Wendling7c78dbb2010-02-25 21:23:24 +0000[diff] [blame]4260<p>The optional constant <tt>align</tt> argument specifies the alignment of the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4261 operation (that is, the alignment of the memory address). A value of 0 or an
Bill Wendling7c78dbb2010-02-25 21:23:24 +0000[diff] [blame]4262 omitted <tt>align</tt> argument means that the operation has the preferential
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4263 alignment for the target. It is the responsibility of the code emitter to
4264 ensure that the alignment information is correct. Overestimating the
Bill Wendling7c78dbb2010-02-25 21:23:24 +0000[diff] [blame]4265 alignment results in undefined behavior. Underestimating the alignment may
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4266 produce less efficient code. An alignment of 1 is always safe.</p>
4267
Bill Wendling7c78dbb2010-02-25 21:23:24 +0000[diff] [blame]4268<p>The optional <tt>!nontemporal</tt> metadata must reference a single
4269 metatadata name &lt;index&gt; corresponding to a metadata node with
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]4270 one <tt>i32</tt> entry of value 1. The existence of
Bill Wendling7c78dbb2010-02-25 21:23:24 +0000[diff] [blame]4271 the <tt>!nontemporal</tt> metatadata on the instruction tells the optimizer
4272 and code generator that this load is not expected to be reused in the cache.
4273 The code generator may select special instructions to save cache bandwidth,
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]4274 such as the <tt>MOVNT</tt> instruction on x86.</p>
David Greene8939b0d2010-02-16 20:50:18 +0000[diff] [blame]4275
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4276<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4277<p>The location of memory pointed to is loaded. If the value being loaded is of
4278 scalar type then the number of bytes read does not exceed the minimum number
4279 of bytes needed to hold all bits of the type. For example, loading an
4280 <tt>i24</tt> reads at most three bytes. When loading a value of a type like
4281 <tt>i20</tt> with a size that is not an integral number of bytes, the result
4282 is undefined if the value was not originally written using a store of the
4283 same type.</p>
4284
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4285<h5>Examples:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4286<pre>
4287 %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
4288 <a href="#i_store">store</a> i32 3, i32* %ptr <i>; yields {void}</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4289 %val = load i32* %ptr <i>; yields {i32}:val = i32 3</i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4290</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4291
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4292</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4293
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4294<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4295<div class="doc_subsubsection"> <a name="i_store">'<tt>store</tt>'
4296Instruction</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4297
Reid Spencer035ab572006-11-09 21:18:01 +0000[diff] [blame]4298<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4299
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4300<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4301<pre>
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]4302 store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;][, !nontemporal !&lt;index&gt;] <i>; yields {void}</i>
4303 volatile store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;][, !nontemporal !&lt;index&gt;] <i>; yields {void}</i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4304</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4305
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4306<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4307<p>The '<tt>store</tt>' instruction is used to write to memory.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4308
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4309<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4310<p>There are two arguments to the '<tt>store</tt>' instruction: a value to store
4311 and an address at which to store it. The type of the
4312 '<tt>&lt;pointer&gt;</tt>' operand must be a pointer to
4313 the <a href="#t_firstclass">first class</a> type of the
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]4314 '<tt>&lt;value&gt;</tt>' operand. If the <tt>store</tt> is marked as
4315 <tt>volatile</tt>, then the optimizer is not allowed to modify the number or
4316 order of execution of this <tt>store</tt> with other <a
4317 href="#volatile">volatile operations</a>.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4318
4319<p>The optional constant "align" argument specifies the alignment of the
4320 operation (that is, the alignment of the memory address). A value of 0 or an
4321 omitted "align" argument means that the operation has the preferential
4322 alignment for the target. It is the responsibility of the code emitter to
4323 ensure that the alignment information is correct. Overestimating the
4324 alignment results in an undefined behavior. Underestimating the alignment may
4325 produce less efficient code. An alignment of 1 is always safe.</p>
4326
David Greene8939b0d2010-02-16 20:50:18 +0000[diff] [blame]4327<p>The optional !nontemporal metadata must reference a single metatadata
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]4328 name &lt;index&gt; corresponding to a metadata node with one i32 entry of
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]4329 value 1. The existence of the !nontemporal metatadata on the
David Greene8939b0d2010-02-16 20:50:18 +0000[diff] [blame]4330 instruction tells the optimizer and code generator that this load is
4331 not expected to be reused in the cache. The code generator may
4332 select special instructions to save cache bandwidth, such as the
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]4333 MOVNT instruction on x86.</p>
David Greene8939b0d2010-02-16 20:50:18 +0000[diff] [blame]4334
4335
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4336<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4337<p>The contents of memory are updated to contain '<tt>&lt;value&gt;</tt>' at the
4338 location specified by the '<tt>&lt;pointer&gt;</tt>' operand. If
4339 '<tt>&lt;value&gt;</tt>' is of scalar type then the number of bytes written
4340 does not exceed the minimum number of bytes needed to hold all bits of the
4341 type. For example, storing an <tt>i24</tt> writes at most three bytes. When
4342 writing a value of a type like <tt>i20</tt> with a size that is not an
4343 integral number of bytes, it is unspecified what happens to the extra bits
4344 that do not belong to the type, but they will typically be overwritten.</p>
4345
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4346<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4347<pre>
4348 %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
Bill Wendling8c6c72d2007-10-22 05:10:05 +0000[diff] [blame]4349 store i32 3, i32* %ptr <i>; yields {void}</i>
4350 %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]4351</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4352
Reid Spencer47ce1792006-11-09 21:15:49 +0000[diff] [blame]4353</div>
4354
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4355<!-- _______________________________________________________________________ -->
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4356<div class="doc_subsubsection">
4357 <a name="i_getelementptr">'<tt>getelementptr</tt>' Instruction</a>
4358</div>
4359
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4360<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4361
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]4362<h5>Syntax:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4363<pre>
Matthijs Kooijmane49d0bc2008-10-13 13:44:15 +0000[diff] [blame]4364 &lt;result&gt; = getelementptr &lt;pty&gt;* &lt;ptrval&gt;{, &lt;ty&gt; &lt;idx&gt;}*
Dan Gohmandd8004d2009-07-27 21:53:46 +0000[diff] [blame]4365 &lt;result&gt; = getelementptr inbounds &lt;pty&gt;* &lt;ptrval&gt;{, &lt;ty&gt; &lt;idx&gt;}*
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4366</pre>
4367
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]4368<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4369<p>The '<tt>getelementptr</tt>' instruction is used to get the address of a
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4370 subelement of an <a href="#t_aggregate">aggregate</a> data structure.
4371 It performs address calculation only and does not access memory.</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4372
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]4373<h5>Arguments:</h5>
Matthijs Kooijmane49d0bc2008-10-13 13:44:15 +0000[diff] [blame]4374<p>The first argument is always a pointer, and forms the basis of the
Chris Lattnerc8eef442009-07-29 06:44:13 +0000[diff] [blame]4375 calculation. The remaining arguments are indices that indicate which of the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4376 elements of the aggregate object are indexed. The interpretation of each
4377 index is dependent on the type being indexed into. The first index always
4378 indexes the pointer value given as the first argument, the second index
4379 indexes a value of the type pointed to (not necessarily the value directly
4380 pointed to, since the first index can be non-zero), etc. The first type
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4381 indexed into must be a pointer value, subsequent types can be arrays,
Chris Lattner61c70e92010-08-28 04:09:24 +0000[diff] [blame]4382 vectors, and structs. Note that subsequent types being indexed into
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4383 can never be pointers, since that would require loading the pointer before
4384 continuing calculation.</p>
Matthijs Kooijmane49d0bc2008-10-13 13:44:15 +0000[diff] [blame]4385
4386<p>The type of each index argument depends on the type it is indexing into.
Chris Lattner61c70e92010-08-28 04:09:24 +0000[diff] [blame]4387 When indexing into a (optionally packed) structure, only <tt>i32</tt>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4388 integer <b>constants</b> are allowed. When indexing into an array, pointer
4389 or vector, integers of any width are allowed, and they are not required to be
Chris Lattnerc8eef442009-07-29 06:44:13 +0000[diff] [blame]4390 constant.</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4391
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4392<p>For example, let's consider a C code fragment and how it gets compiled to
4393 LLVM:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4394
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]4395<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]4396struct RT {
4397 char A;
Chris Lattnercabc8462007-05-29 15:43:56 +0000[diff] [blame]4398 int B[10][20];
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]4399 char C;
4400};
4401struct ST {
Chris Lattnercabc8462007-05-29 15:43:56 +0000[diff] [blame]4402 int X;
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]4403 double Y;
4404 struct RT Z;
4405};
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4406
Chris Lattnercabc8462007-05-29 15:43:56 +0000[diff] [blame]4407int *foo(struct ST *s) {
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]4408 return &amp;s[1].Z.B[5][13];
4409}
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4410</pre>
4411
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4412<p>The LLVM code generated by the GCC frontend is:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4413
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]4414<pre class="doc_code">
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]4415%RT = <a href="#namedtypes">type</a> { i8 , [10 x [20 x i32]], i8 }
4416%ST = <a href="#namedtypes">type</a> { i32, double, %RT }
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4417
Dan Gohman4df605b2009-07-25 02:23:48 +0000[diff] [blame]4418define i32* @foo(%ST* %s) {
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]4419entry:
4420 %reg = getelementptr %ST* %s, i32 1, i32 2, i32 1, i32 5, i32 13
4421 ret i32* %reg
4422}
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4423</pre>
4424
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]4425<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4426<p>In the example above, the first index is indexing into the '<tt>%ST*</tt>'
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4427 type, which is a pointer, yielding a '<tt>%ST</tt>' = '<tt>{ i32, double, %RT
4428 }</tt>' type, a structure. The second index indexes into the third element
4429 of the structure, yielding a '<tt>%RT</tt>' = '<tt>{ i8 , [10 x [20 x i32]],
4430 i8 }</tt>' type, another structure. The third index indexes into the second
4431 element of the structure, yielding a '<tt>[10 x [20 x i32]]</tt>' type, an
4432 array. The two dimensions of the array are subscripted into, yielding an
4433 '<tt>i32</tt>' type. The '<tt>getelementptr</tt>' instruction returns a
4434 pointer to this element, thus computing a value of '<tt>i32*</tt>' type.</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4435
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4436<p>Note that it is perfectly legal to index partially through a structure,
4437 returning a pointer to an inner element. Because of this, the LLVM code for
4438 the given testcase is equivalent to:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4439
4440<pre>
Dan Gohman4df605b2009-07-25 02:23:48 +0000[diff] [blame]4441 define i32* @foo(%ST* %s) {
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4442 %t1 = getelementptr %ST* %s, i32 1 <i>; yields %ST*:%t1</i>
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]4443 %t2 = getelementptr %ST* %t1, i32 0, i32 2 <i>; yields %RT*:%t2</i>
4444 %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]4445 %t4 = getelementptr [10 x [20 x i32]]* %t3, i32 0, i32 5 <i>; yields [20 x i32]*:%t4</i>
4446 %t5 = getelementptr [20 x i32]* %t4, i32 0, i32 13 <i>; yields i32*:%t5</i>
4447 ret i32* %t5
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4448 }
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]4449</pre>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]4450
Dan Gohmandd8004d2009-07-27 21:53:46 +0000[diff] [blame]4451<p>If the <tt>inbounds</tt> keyword is present, the result value of the
Dan Gohman27ef9972010-04-23 15:23:32 +0000[diff] [blame]4452 <tt>getelementptr</tt> is a <a href="#trapvalues">trap value</a> if the
4453 base pointer is not an <i>in bounds</i> address of an allocated object,
4454 or if any of the addresses that would be formed by successive addition of
4455 the offsets implied by the indices to the base address with infinitely
4456 precise arithmetic are not an <i>in bounds</i> address of that allocated
4457 object. The <i>in bounds</i> addresses for an allocated object are all
4458 the addresses that point into the object, plus the address one byte past
4459 the end.</p>
Dan Gohmandd8004d2009-07-27 21:53:46 +0000[diff] [blame]4460
4461<p>If the <tt>inbounds</tt> keyword is not present, the offsets are added to
4462 the base address with silently-wrapping two's complement arithmetic, and
4463 the result value of the <tt>getelementptr</tt> may be outside the object
4464 pointed to by the base pointer. The result value may not necessarily be
4465 used to access memory though, even if it happens to point into allocated
4466 storage. See the <a href="#pointeraliasing">Pointer Aliasing Rules</a>
4467 section for more information.</p>
4468
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4469<p>The getelementptr instruction is often confusing. For some more insight into
4470 how it works, see <a href="GetElementPtr.html">the getelementptr FAQ</a>.</p>
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]4471
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]4472<h5>Example:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4473<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4474 <i>; yields [12 x i8]*:aptr</i>
Matthijs Kooijmane49d0bc2008-10-13 13:44:15 +0000[diff] [blame]4475 %aptr = getelementptr {i32, [12 x i8]}* %saptr, i64 0, i32 1
4476 <i>; yields i8*:vptr</i>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]4477 %vptr = getelementptr {i32, &lt;2 x i8&gt;}* %svptr, i64 0, i32 1, i32 1
Matthijs Kooijmane49d0bc2008-10-13 13:44:15 +0000[diff] [blame]4478 <i>; yields i8*:eptr</i>
4479 %eptr = getelementptr [12 x i8]* %aptr, i64 0, i32 1
Sanjiv Gupta9f805c22009-04-25 07:27:44 +0000[diff] [blame]4480 <i>; yields i32*:iptr</i>
Sanjiv Gupta16ffa802009-04-24 16:38:13 +0000[diff] [blame]4481 %iptr = getelementptr [10 x i32]* @arr, i16 0, i16 0
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4482</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4483
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4484</div>
Reid Spencer47ce1792006-11-09 21:15:49 +0000[diff] [blame]4485
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4486<!-- ======================================================================= -->
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]4487<div class="doc_subsection"> <a name="convertops">Conversion Operations</a>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4488</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4489
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4490<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4491
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]4492<p>The instructions in this category are the conversion instructions (casting)
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4493 which all take a single operand and a type. They perform various bit
4494 conversions on the operand.</p>
4495
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4496</div>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]4497
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]4498<!-- _______________________________________________________________________ -->
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]4499<div class="doc_subsubsection">
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4500 <a name="i_trunc">'<tt>trunc .. to</tt>' Instruction</a>
4501</div>
4502<div class="doc_text">
4503
4504<h5>Syntax:</h5>
4505<pre>
4506 &lt;result&gt; = trunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4507</pre>
4508
4509<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4510<p>The '<tt>trunc</tt>' instruction truncates its operand to the
4511 type <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4512
4513<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4514<p>The '<tt>trunc</tt>' instruction takes a <tt>value</tt> to trunc, which must
4515 be an <a href="#t_integer">integer</a> type, and a type that specifies the
4516 size and type of the result, which must be
4517 an <a href="#t_integer">integer</a> type. The bit size of <tt>value</tt> must
4518 be larger than the bit size of <tt>ty2</tt>. Equal sized types are not
4519 allowed.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4520
4521<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4522<p>The '<tt>trunc</tt>' instruction truncates the high order bits
4523 in <tt>value</tt> and converts the remaining bits to <tt>ty2</tt>. Since the
4524 source size must be larger than the destination size, <tt>trunc</tt> cannot
4525 be a <i>no-op cast</i>. It will always truncate bits.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4526
4527<h5>Example:</h5>
4528<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4529 %X = trunc i32 257 to i8 <i>; yields i8:1</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]4530 %Y = trunc i32 123 to i1 <i>; yields i1:true</i>
Gabor Greif6a292012009-10-28 09:21:30 +0000[diff] [blame]4531 %Z = trunc i32 122 to i1 <i>; yields i1:false</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4532</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4533
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4534</div>
4535
4536<!-- _______________________________________________________________________ -->
4537<div class="doc_subsubsection">
4538 <a name="i_zext">'<tt>zext .. to</tt>' Instruction</a>
4539</div>
4540<div class="doc_text">
4541
4542<h5>Syntax:</h5>
4543<pre>
4544 &lt;result&gt; = zext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4545</pre>
4546
4547<h5>Overview:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4548<p>The '<tt>zext</tt>' instruction zero extends its operand to type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4549 <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4550
4551
4552<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4553<p>The '<tt>zext</tt>' instruction takes a value to cast, which must be of
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4554 <a href="#t_integer">integer</a> type, and a type to cast it to, which must
4555 also be of <a href="#t_integer">integer</a> type. The bit size of the
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4556 <tt>value</tt> must be smaller than the bit size of the destination type,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4557 <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4558
4559<h5>Semantics:</h5>
4560<p>The <tt>zext</tt> fills the high order bits of the <tt>value</tt> with zero
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4561 bits until it reaches the size of the destination type, <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4562
Reid Spencerb5929522007-01-12 15:46:11 +0000[diff] [blame]4563<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]4564
4565<h5>Example:</h5>
4566<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4567 %X = zext i32 257 to i64 <i>; yields i64:257</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]4568 %Y = zext i1 true to i32 <i>; yields i32:1</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4569</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4570
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4571</div>
4572
4573<!-- _______________________________________________________________________ -->
4574<div class="doc_subsubsection">
4575 <a name="i_sext">'<tt>sext .. to</tt>' Instruction</a>
4576</div>
4577<div class="doc_text">
4578
4579<h5>Syntax:</h5>
4580<pre>
4581 &lt;result&gt; = sext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4582</pre>
4583
4584<h5>Overview:</h5>
4585<p>The '<tt>sext</tt>' sign extends <tt>value</tt> to the type <tt>ty2</tt>.</p>
4586
4587<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4588<p>The '<tt>sext</tt>' instruction takes a value to cast, which must be of
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4589 <a href="#t_integer">integer</a> type, and a type to cast it to, which must
4590 also be of <a href="#t_integer">integer</a> type. The bit size of the
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4591 <tt>value</tt> must be smaller than the bit size of the destination type,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4592 <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4593
4594<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4595<p>The '<tt>sext</tt>' instruction performs a sign extension by copying the sign
4596 bit (highest order bit) of the <tt>value</tt> until it reaches the bit size
4597 of the type <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4598
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]4599<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]4600
4601<h5>Example:</h5>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4602<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4603 %X = sext i8 -1 to i16 <i>; yields i16 :65535</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]4604 %Y = sext i1 true to i32 <i>; yields i32:-1</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4605</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4606
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4607</div>
4608
4609<!-- _______________________________________________________________________ -->
4610<div class="doc_subsubsection">
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]4611 <a name="i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a>
4612</div>
4613
4614<div class="doc_text">
4615
4616<h5>Syntax:</h5>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]4617<pre>
4618 &lt;result&gt; = fptrunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4619</pre>
4620
4621<h5>Overview:</h5>
4622<p>The '<tt>fptrunc</tt>' instruction truncates <tt>value</tt> to type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4623 <tt>ty2</tt>.</p>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]4624
4625<h5>Arguments:</h5>
4626<p>The '<tt>fptrunc</tt>' instruction takes a <a href="#t_floating">floating
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4627 point</a> value to cast and a <a href="#t_floating">floating point</a> type
4628 to cast it to. The size of <tt>value</tt> must be larger than the size of
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4629 <tt>ty2</tt>. This implies that <tt>fptrunc</tt> cannot be used to make a
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4630 <i>no-op cast</i>.</p>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]4631
4632<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4633<p>The '<tt>fptrunc</tt>' instruction truncates a <tt>value</tt> from a larger
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4634 <a href="#t_floating">floating point</a> type to a smaller
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4635 <a href="#t_floating">floating point</a> type. If the value cannot fit
4636 within the destination type, <tt>ty2</tt>, then the results are
4637 undefined.</p>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]4638
4639<h5>Example:</h5>
4640<pre>
4641 %X = fptrunc double 123.0 to float <i>; yields float:123.0</i>
4642 %Y = fptrunc double 1.0E+300 to float <i>; yields undefined</i>
4643</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4644
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]4645</div>
4646
4647<!-- _______________________________________________________________________ -->
4648<div class="doc_subsubsection">
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4649 <a name="i_fpext">'<tt>fpext .. to</tt>' Instruction</a>
4650</div>
4651<div class="doc_text">
4652
4653<h5>Syntax:</h5>
4654<pre>
4655 &lt;result&gt; = fpext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4656</pre>
4657
4658<h5>Overview:</h5>
4659<p>The '<tt>fpext</tt>' extends a floating point <tt>value</tt> to a larger
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4660 floating point value.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4661
4662<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4663<p>The '<tt>fpext</tt>' instruction takes a
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4664 <a href="#t_floating">floating point</a> <tt>value</tt> to cast, and
4665 a <a href="#t_floating">floating point</a> type to cast it to. The source
4666 type must be smaller than the destination type.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4667
4668<h5>Semantics:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]4669<p>The '<tt>fpext</tt>' instruction extends the <tt>value</tt> from a smaller
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4670 <a href="#t_floating">floating point</a> type to a larger
4671 <a href="#t_floating">floating point</a> type. The <tt>fpext</tt> cannot be
4672 used to make a <i>no-op cast</i> because it always changes bits. Use
4673 <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]4674
4675<h5>Example:</h5>
4676<pre>
4677 %X = fpext float 3.1415 to double <i>; yields double:3.1415</i>
4678 %Y = fpext float 1.0 to float <i>; yields float:1.0 (no-op)</i>
4679</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4680
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4681</div>
4682
4683<!-- _______________________________________________________________________ -->
4684<div class="doc_subsubsection">
Reid Spencer24d6da52007-01-21 00:29:26 +0000[diff] [blame]4685 <a name="i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4686</div>
4687<div class="doc_text">
4688
4689<h5>Syntax:</h5>
4690<pre>
Reid Spencer1539a1c2007-07-31 14:40:14 +0000[diff] [blame]4691 &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]4692</pre>
4693
4694<h5>Overview:</h5>
Reid Spencer1539a1c2007-07-31 14:40:14 +0000[diff] [blame]4695<p>The '<tt>fptoui</tt>' converts a floating point <tt>value</tt> to its
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4696 unsigned integer equivalent of type <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4697
4698<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4699<p>The '<tt>fptoui</tt>' instruction takes a value to cast, which must be a
4700 scalar or vector <a href="#t_floating">floating point</a> value, and a type
4701 to cast it to <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a>
4702 type. If <tt>ty</tt> is a vector floating point type, <tt>ty2</tt> must be a
4703 vector integer type with the same number of elements as <tt>ty</tt></p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4704
4705<h5>Semantics:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4706<p>The '<tt>fptoui</tt>' instruction converts its
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4707 <a href="#t_floating">floating point</a> operand into the nearest (rounding
4708 towards zero) unsigned integer value. If the value cannot fit
4709 in <tt>ty2</tt>, the results are undefined.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4710
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4711<h5>Example:</h5>
4712<pre>
Reid Spencer1539a1c2007-07-31 14:40:14 +0000[diff] [blame]4713 %X = fptoui double 123.0 to i32 <i>; yields i32:123</i>
Chris Lattner88519042007-09-22 03:17:52 +0000[diff] [blame]4714 %Y = fptoui float 1.0E+300 to i1 <i>; yields undefined:1</i>
Gabor Greif6a292012009-10-28 09:21:30 +0000[diff] [blame]4715 %Z = fptoui float 1.04E+17 to i8 <i>; yields undefined:1</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4716</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4717
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4718</div>
4719
4720<!-- _______________________________________________________________________ -->
4721<div class="doc_subsubsection">
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]4722 <a name="i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4723</div>
4724<div class="doc_text">
4725
4726<h5>Syntax:</h5>
4727<pre>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]4728 &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]4729</pre>
4730
4731<h5>Overview:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4732<p>The '<tt>fptosi</tt>' instruction converts
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4733 <a href="#t_floating">floating point</a> <tt>value</tt> to
4734 type <tt>ty2</tt>.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]4735
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]4736<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4737<p>The '<tt>fptosi</tt>' instruction takes a value to cast, which must be a
4738 scalar or vector <a href="#t_floating">floating point</a> value, and a type
4739 to cast it to <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a>
4740 type. If <tt>ty</tt> is a vector floating point type, <tt>ty2</tt> must be a
4741 vector integer type with the same number of elements as <tt>ty</tt></p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]4742
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]4743<h5>Semantics:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4744<p>The '<tt>fptosi</tt>' instruction converts its
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4745 <a href="#t_floating">floating point</a> operand into the nearest (rounding
4746 towards zero) signed integer value. If the value cannot fit in <tt>ty2</tt>,
4747 the results are undefined.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]4748
Chris Lattner33ba0d92001-07-09 00:26:23 +0000[diff] [blame]4749<h5>Example:</h5>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]4750<pre>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]4751 %X = fptosi double -123.0 to i32 <i>; yields i32:-123</i>
Chris Lattner88519042007-09-22 03:17:52 +0000[diff] [blame]4752 %Y = fptosi float 1.0E-247 to i1 <i>; yields undefined:1</i>
Gabor Greif6a292012009-10-28 09:21:30 +0000[diff] [blame]4753 %Z = fptosi float 1.04E+17 to i8 <i>; yields undefined:1</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4754</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4755
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4756</div>
4757
4758<!-- _______________________________________________________________________ -->
4759<div class="doc_subsubsection">
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]4760 <a name="i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4761</div>
4762<div class="doc_text">
4763
4764<h5>Syntax:</h5>
4765<pre>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]4766 &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]4767</pre>
4768
4769<h5>Overview:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]4770<p>The '<tt>uitofp</tt>' instruction regards <tt>value</tt> as an unsigned
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4771 integer and converts that value to the <tt>ty2</tt> type.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4772
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4773<h5>Arguments:</h5>
Nate Begemanb348d182007-11-17 03:58:34 +0000[diff] [blame]4774<p>The '<tt>uitofp</tt>' instruction takes a value to cast, which must be a
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4775 scalar or vector <a href="#t_integer">integer</a> value, and a type to cast
4776 it to <tt>ty2</tt>, which must be an <a href="#t_floating">floating point</a>
4777 type. If <tt>ty</tt> is a vector integer type, <tt>ty2</tt> must be a vector
4778 floating point type with the same number of elements as <tt>ty</tt></p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4779
4780<h5>Semantics:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]4781<p>The '<tt>uitofp</tt>' instruction interprets its operand as an unsigned
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4782 integer quantity and converts it to the corresponding floating point
4783 value. If the value cannot fit in the floating point value, the results are
4784 undefined.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4785
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4786<h5>Example:</h5>
4787<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4788 %X = uitofp i32 257 to float <i>; yields float:257.0</i>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]4789 %Y = uitofp i8 -1 to double <i>; yields double:255.0</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4790</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4791
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4792</div>
4793
4794<!-- _______________________________________________________________________ -->
4795<div class="doc_subsubsection">
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]4796 <a name="i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4797</div>
4798<div class="doc_text">
4799
4800<h5>Syntax:</h5>
4801<pre>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]4802 &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]4803</pre>
4804
4805<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4806<p>The '<tt>sitofp</tt>' instruction regards <tt>value</tt> as a signed integer
4807 and converts that value to the <tt>ty2</tt> type.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4808
4809<h5>Arguments:</h5>
Nate Begemanb348d182007-11-17 03:58:34 +0000[diff] [blame]4810<p>The '<tt>sitofp</tt>' instruction takes a value to cast, which must be a
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4811 scalar or vector <a href="#t_integer">integer</a> value, and a type to cast
4812 it to <tt>ty2</tt>, which must be an <a href="#t_floating">floating point</a>
4813 type. If <tt>ty</tt> is a vector integer type, <tt>ty2</tt> must be a vector
4814 floating point type with the same number of elements as <tt>ty</tt></p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4815
4816<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4817<p>The '<tt>sitofp</tt>' instruction interprets its operand as a signed integer
4818 quantity and converts it to the corresponding floating point value. If the
4819 value cannot fit in the floating point value, the results are undefined.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4820
4821<h5>Example:</h5>
4822<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4823 %X = sitofp i32 257 to float <i>; yields float:257.0</i>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]4824 %Y = sitofp i8 -1 to double <i>; yields double:-1.0</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4825</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4826
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4827</div>
4828
4829<!-- _______________________________________________________________________ -->
4830<div class="doc_subsubsection">
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]4831 <a name="i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a>
4832</div>
4833<div class="doc_text">
4834
4835<h5>Syntax:</h5>
4836<pre>
4837 &lt;result&gt; = ptrtoint &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4838</pre>
4839
4840<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4841<p>The '<tt>ptrtoint</tt>' instruction converts the pointer <tt>value</tt> to
4842 the integer type <tt>ty2</tt>.</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]4843
4844<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4845<p>The '<tt>ptrtoint</tt>' instruction takes a <tt>value</tt> to cast, which
4846 must be a <a href="#t_pointer">pointer</a> value, and a type to cast it to
4847 <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a> type.</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]4848
4849<h5>Semantics:</h5>
4850<p>The '<tt>ptrtoint</tt>' instruction converts <tt>value</tt> to integer type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4851 <tt>ty2</tt> by interpreting the pointer value as an integer and either
4852 truncating or zero extending that value to the size of the integer type. If
4853 <tt>value</tt> is smaller than <tt>ty2</tt> then a zero extension is done. If
4854 <tt>value</tt> is larger than <tt>ty2</tt> then a truncation is done. If they
4855 are the same size, then nothing is done (<i>no-op cast</i>) other than a type
4856 change.</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]4857
4858<h5>Example:</h5>
4859<pre>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4860 %X = ptrtoint i32* %X to i8 <i>; yields truncation on 32-bit architecture</i>
4861 %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]4862</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4863
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]4864</div>
4865
4866<!-- _______________________________________________________________________ -->
4867<div class="doc_subsubsection">
4868 <a name="i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a>
4869</div>
4870<div class="doc_text">
4871
4872<h5>Syntax:</h5>
4873<pre>
4874 &lt;result&gt; = inttoptr &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4875</pre>
4876
4877<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4878<p>The '<tt>inttoptr</tt>' instruction converts an integer <tt>value</tt> to a
4879 pointer type, <tt>ty2</tt>.</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]4880
4881<h5>Arguments:</h5>
Duncan Sands8036ca42007-03-30 12:22:09 +0000[diff] [blame]4882<p>The '<tt>inttoptr</tt>' instruction takes an <a href="#t_integer">integer</a>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4883 value to cast, and a type to cast it to, which must be a
4884 <a href="#t_pointer">pointer</a> type.</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]4885
4886<h5>Semantics:</h5>
4887<p>The '<tt>inttoptr</tt>' instruction converts <tt>value</tt> to type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4888 <tt>ty2</tt> by applying either a zero extension or a truncation depending on
4889 the size of the integer <tt>value</tt>. If <tt>value</tt> is larger than the
4890 size of a pointer then a truncation is done. If <tt>value</tt> is smaller
4891 than the size of a pointer then a zero extension is done. If they are the
4892 same size, nothing is done (<i>no-op cast</i>).</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]4893
4894<h5>Example:</h5>
4895<pre>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4896 %X = inttoptr i32 255 to i32* <i>; yields zero extension on 64-bit architecture</i>
Gabor Greif6a292012009-10-28 09:21:30 +0000[diff] [blame]4897 %Y = inttoptr i32 255 to i32* <i>; yields no-op on 32-bit architecture</i>
4898 %Z = inttoptr i64 0 to i32* <i>; yields truncation on 32-bit architecture</i>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]4899</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4900
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]4901</div>
4902
4903<!-- _______________________________________________________________________ -->
4904<div class="doc_subsubsection">
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]4905 <a name="i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4906</div>
4907<div class="doc_text">
4908
4909<h5>Syntax:</h5>
4910<pre>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]4911 &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]4912</pre>
4913
4914<h5>Overview:</h5>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]4915<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4916 <tt>ty2</tt> without changing any bits.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4917
4918<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4919<p>The '<tt>bitcast</tt>' instruction takes a value to cast, which must be a
4920 non-aggregate first class value, and a type to cast it to, which must also be
4921 a non-aggregate <a href="#t_firstclass">first class</a> type. The bit sizes
4922 of <tt>value</tt> and the destination type, <tt>ty2</tt>, must be
4923 identical. If the source type is a pointer, the destination type must also be
4924 a pointer. This instruction supports bitwise conversion of vectors to
4925 integers and to vectors of other types (as long as they have the same
4926 size).</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4927
4928<h5>Semantics:</h5>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]4929<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4930 <tt>ty2</tt>. It is always a <i>no-op cast</i> because no bits change with
4931 this conversion. The conversion is done as if the <tt>value</tt> had been
4932 stored to memory and read back as type <tt>ty2</tt>. Pointer types may only
4933 be converted to other pointer types with this instruction. To convert
4934 pointers to other types, use the <a href="#i_inttoptr">inttoptr</a> or
4935 <a href="#i_ptrtoint">ptrtoint</a> instructions first.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4936
4937<h5>Example:</h5>
4938<pre>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4939 %X = bitcast i8 255 to i8 <i>; yields i8 :-1</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4940 %Y = bitcast i32* %x to sint* <i>; yields sint*:%x</i>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4941 %Z = bitcast &lt;2 x int&gt; %V to i64; <i>; yields i64: %V</i>
Chris Lattner33ba0d92001-07-09 00:26:23 +0000[diff] [blame]4942</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4943
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4944</div>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]4945
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]4946<!-- ======================================================================= -->
4947<div class="doc_subsection"> <a name="otherops">Other Operations</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4948
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]4949<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4950
4951<p>The instructions in this category are the "miscellaneous" instructions, which
4952 defy better classification.</p>
4953
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]4954</div>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4955
4956<!-- _______________________________________________________________________ -->
4957<div class="doc_subsubsection"><a name="i_icmp">'<tt>icmp</tt>' Instruction</a>
4958</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4959
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4960<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4961
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4962<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4963<pre>
4964 &lt;result&gt; = icmp &lt;cond&gt; &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {i1} or {&lt;N x i1&gt;}:result</i>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4965</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4966
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4967<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4968<p>The '<tt>icmp</tt>' instruction returns a boolean value or a vector of
4969 boolean values based on comparison of its two integer, integer vector, or
4970 pointer operands.</p>
4971
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4972<h5>Arguments:</h5>
4973<p>The '<tt>icmp</tt>' instruction takes three operands. The first operand is
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4974 the condition code indicating the kind of comparison to perform. It is not a
4975 value, just a keyword. The possible condition code are:</p>
4976
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4977<ol>
4978 <li><tt>eq</tt>: equal</li>
4979 <li><tt>ne</tt>: not equal </li>
4980 <li><tt>ugt</tt>: unsigned greater than</li>
4981 <li><tt>uge</tt>: unsigned greater or equal</li>
4982 <li><tt>ult</tt>: unsigned less than</li>
4983 <li><tt>ule</tt>: unsigned less or equal</li>
4984 <li><tt>sgt</tt>: signed greater than</li>
4985 <li><tt>sge</tt>: signed greater or equal</li>
4986 <li><tt>slt</tt>: signed less than</li>
4987 <li><tt>sle</tt>: signed less or equal</li>
4988</ol>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4989
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]4990<p>The remaining two arguments must be <a href="#t_integer">integer</a> or
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4991 <a href="#t_pointer">pointer</a> or integer <a href="#t_vector">vector</a>
4992 typed. They must also be identical types.</p>
4993
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4994<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4995<p>The '<tt>icmp</tt>' compares <tt>op1</tt> and <tt>op2</tt> according to the
4996 condition code given as <tt>cond</tt>. The comparison performed always yields
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]4997 either an <a href="#t_integer"><tt>i1</tt></a> or vector of <tt>i1</tt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4998 result, as follows:</p>
4999
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5000<ol>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5001 <li><tt>eq</tt>: yields <tt>true</tt> if the operands are equal,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5002 <tt>false</tt> otherwise. No sign interpretation is necessary or
5003 performed.</li>
5004
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5005 <li><tt>ne</tt>: yields <tt>true</tt> if the operands are unequal,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5006 <tt>false</tt> otherwise. No sign interpretation is necessary or
5007 performed.</li>
5008
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5009 <li><tt>ugt</tt>: interprets the operands as unsigned values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5010 <tt>true</tt> if <tt>op1</tt> is greater than <tt>op2</tt>.</li>
5011
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5012 <li><tt>uge</tt>: interprets the operands as unsigned values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5013 <tt>true</tt> if <tt>op1</tt> is greater than or equal
5014 to <tt>op2</tt>.</li>
5015
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5016 <li><tt>ult</tt>: interprets the operands as unsigned values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5017 <tt>true</tt> if <tt>op1</tt> is less than <tt>op2</tt>.</li>
5018
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5019 <li><tt>ule</tt>: interprets the operands as unsigned values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5020 <tt>true</tt> if <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
5021
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5022 <li><tt>sgt</tt>: interprets the operands as signed values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5023 <tt>true</tt> if <tt>op1</tt> is greater than <tt>op2</tt>.</li>
5024
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5025 <li><tt>sge</tt>: interprets the operands as signed values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5026 <tt>true</tt> if <tt>op1</tt> is greater than or equal
5027 to <tt>op2</tt>.</li>
5028
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5029 <li><tt>slt</tt>: interprets the operands as signed values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5030 <tt>true</tt> if <tt>op1</tt> is less than <tt>op2</tt>.</li>
5031
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5032 <li><tt>sle</tt>: interprets the operands as signed values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5033 <tt>true</tt> if <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5034</ol>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5035
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5036<p>If the operands are <a href="#t_pointer">pointer</a> typed, the pointer
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5037 values are compared as if they were integers.</p>
5038
5039<p>If the operands are integer vectors, then they are compared element by
5040 element. The result is an <tt>i1</tt> vector with the same number of elements
5041 as the values being compared. Otherwise, the result is an <tt>i1</tt>.</p>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5042
5043<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5044<pre>
5045 &lt;result&gt; = icmp eq i32 4, 5 <i>; yields: result=false</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5046 &lt;result&gt; = icmp ne float* %X, %X <i>; yields: result=false</i>
5047 &lt;result&gt; = icmp ult i16 4, 5 <i>; yields: result=true</i>
5048 &lt;result&gt; = icmp sgt i16 4, 5 <i>; yields: result=false</i>
5049 &lt;result&gt; = icmp ule i16 -4, 5 <i>; yields: result=false</i>
5050 &lt;result&gt; = icmp sge i16 4, 5 <i>; yields: result=false</i>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5051</pre>
Dan Gohmanc766f722009-01-22 01:39:38 +0000[diff] [blame]5052
5053<p>Note that the code generator does not yet support vector types with
5054 the <tt>icmp</tt> instruction.</p>
5055
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5056</div>
5057
5058<!-- _______________________________________________________________________ -->
5059<div class="doc_subsubsection"><a name="i_fcmp">'<tt>fcmp</tt>' Instruction</a>
5060</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5061
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5062<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5063
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5064<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5065<pre>
5066 &lt;result&gt; = fcmp &lt;cond&gt; &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {i1} or {&lt;N x i1&gt;}:result</i>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5067</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5068
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5069<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5070<p>The '<tt>fcmp</tt>' instruction returns a boolean value or vector of boolean
5071 values based on comparison of its operands.</p>
5072
5073<p>If the operands are floating point scalars, then the result type is a boolean
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]5074(<a href="#t_integer"><tt>i1</tt></a>).</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5075
5076<p>If the operands are floating point vectors, then the result type is a vector
5077 of boolean with the same number of elements as the operands being
5078 compared.</p>
5079
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5080<h5>Arguments:</h5>
5081<p>The '<tt>fcmp</tt>' instruction takes three operands. The first operand is
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5082 the condition code indicating the kind of comparison to perform. It is not a
5083 value, just a keyword. The possible condition code are:</p>
5084
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5085<ol>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]5086 <li><tt>false</tt>: no comparison, always returns false</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5087 <li><tt>oeq</tt>: ordered and equal</li>
5088 <li><tt>ogt</tt>: ordered and greater than </li>
5089 <li><tt>oge</tt>: ordered and greater than or equal</li>
5090 <li><tt>olt</tt>: ordered and less than </li>
5091 <li><tt>ole</tt>: ordered and less than or equal</li>
5092 <li><tt>one</tt>: ordered and not equal</li>
5093 <li><tt>ord</tt>: ordered (no nans)</li>
5094 <li><tt>ueq</tt>: unordered or equal</li>
5095 <li><tt>ugt</tt>: unordered or greater than </li>
5096 <li><tt>uge</tt>: unordered or greater than or equal</li>
5097 <li><tt>ult</tt>: unordered or less than </li>
5098 <li><tt>ule</tt>: unordered or less than or equal</li>
5099 <li><tt>une</tt>: unordered or not equal</li>
5100 <li><tt>uno</tt>: unordered (either nans)</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]5101 <li><tt>true</tt>: no comparison, always returns true</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5102</ol>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5103
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5104<p><i>Ordered</i> means that neither operand is a QNAN while
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5105 <i>unordered</i> means that either operand may be a QNAN.</p>
5106
5107<p>Each of <tt>val1</tt> and <tt>val2</tt> arguments must be either
5108 a <a href="#t_floating">floating point</a> type or
5109 a <a href="#t_vector">vector</a> of floating point type. They must have
5110 identical types.</p>
5111
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5112<h5>Semantics:</h5>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]5113<p>The '<tt>fcmp</tt>' instruction compares <tt>op1</tt> and <tt>op2</tt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5114 according to the condition code given as <tt>cond</tt>. If the operands are
5115 vectors, then the vectors are compared element by element. Each comparison
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]5116 performed always yields an <a href="#t_integer">i1</a> result, as
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5117 follows:</p>
5118
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5119<ol>
5120 <li><tt>false</tt>: always yields <tt>false</tt>, regardless of operands.</li>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5121
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5122 <li><tt>oeq</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5123 <tt>op1</tt> is equal to <tt>op2</tt>.</li>
5124
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]5125 <li><tt>ogt</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]5126 <tt>op1</tt> is greater than <tt>op2</tt>.</li>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5127
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5128 <li><tt>oge</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5129 <tt>op1</tt> is greater than or equal to <tt>op2</tt>.</li>
5130
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5131 <li><tt>olt</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5132 <tt>op1</tt> is less than <tt>op2</tt>.</li>
5133
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5134 <li><tt>ole</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5135 <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
5136
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5137 <li><tt>one</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5138 <tt>op1</tt> is not equal to <tt>op2</tt>.</li>
5139
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]5140 <li><tt>ord</tt>: yields <tt>true</tt> if both operands are not a QNAN.</li>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5141
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5142 <li><tt>ueq</tt>: yields <tt>true</tt> if either operand is a QNAN or
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5143 <tt>op1</tt> is equal to <tt>op2</tt>.</li>
5144
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5145 <li><tt>ugt</tt>: yields <tt>true</tt> if either operand is a QNAN or
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5146 <tt>op1</tt> is greater than <tt>op2</tt>.</li>
5147
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5148 <li><tt>uge</tt>: yields <tt>true</tt> if either operand is a QNAN or
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5149 <tt>op1</tt> is greater than or equal to <tt>op2</tt>.</li>
5150
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5151 <li><tt>ult</tt>: yields <tt>true</tt> if either operand is a QNAN or
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5152 <tt>op1</tt> is less than <tt>op2</tt>.</li>
5153
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5154 <li><tt>ule</tt>: yields <tt>true</tt> if either operand is a QNAN or
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5155 <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
5156
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5157 <li><tt>une</tt>: yields <tt>true</tt> if either operand is a QNAN or
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5158 <tt>op1</tt> is not equal to <tt>op2</tt>.</li>
5159
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]5160 <li><tt>uno</tt>: yields <tt>true</tt> if either operand is a QNAN.</li>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5161
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5162 <li><tt>true</tt>: always yields <tt>true</tt>, regardless of operands.</li>
5163</ol>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5164
5165<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5166<pre>
5167 &lt;result&gt; = fcmp oeq float 4.0, 5.0 <i>; yields: result=false</i>
Dan Gohmanf72fb672008-09-09 01:02:47 +0000[diff] [blame]5168 &lt;result&gt; = fcmp one float 4.0, 5.0 <i>; yields: result=true</i>
5169 &lt;result&gt; = fcmp olt float 4.0, 5.0 <i>; yields: result=true</i>
5170 &lt;result&gt; = fcmp ueq double 1.0, 2.0 <i>; yields: result=false</i>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5171</pre>
Dan Gohmanc766f722009-01-22 01:39:38 +0000[diff] [blame]5172
5173<p>Note that the code generator does not yet support vector types with
5174 the <tt>fcmp</tt> instruction.</p>
5175
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5176</div>
5177
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5178<!-- _______________________________________________________________________ -->
Nate Begemanac80ade2008-05-12 19:01:56 +0000[diff] [blame]5179<div class="doc_subsubsection">
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]5180 <a name="i_phi">'<tt>phi</tt>' Instruction</a>
5181</div>
5182
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5183<div class="doc_text">
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]5184
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5185<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5186<pre>
5187 &lt;result&gt; = phi &lt;ty&gt; [ &lt;val0&gt;, &lt;label0&gt;], ...
5188</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]5189
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5190<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5191<p>The '<tt>phi</tt>' instruction is used to implement the &#966; node in the
5192 SSA graph representing the function.</p>
5193
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5194<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5195<p>The type of the incoming values is specified with the first type field. After
5196 this, the '<tt>phi</tt>' instruction takes a list of pairs as arguments, with
5197 one pair for each predecessor basic block of the current block. Only values
5198 of <a href="#t_firstclass">first class</a> type may be used as the value
5199 arguments to the PHI node. Only labels may be used as the label
5200 arguments.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]5201
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5202<p>There must be no non-phi instructions between the start of a basic block and
5203 the PHI instructions: i.e. PHI instructions must be first in a basic
5204 block.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]5205
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5206<p>For the purposes of the SSA form, the use of each incoming value is deemed to
5207 occur on the edge from the corresponding predecessor block to the current
5208 block (but after any definition of an '<tt>invoke</tt>' instruction's return
5209 value on the same edge).</p>
Jay Foadd2449092009-06-03 10:20:10 +0000[diff] [blame]5210
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5211<h5>Semantics:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5212<p>At runtime, the '<tt>phi</tt>' instruction logically takes on the value
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5213 specified by the pair corresponding to the predecessor basic block that
5214 executed just prior to the current block.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]5215
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5216<h5>Example:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]5217<pre>
5218Loop: ; Infinite loop that counts from 0 on up...
5219 %indvar = phi i32 [ 0, %LoopHeader ], [ %nextindvar, %Loop ]
5220 %nextindvar = add i32 %indvar, 1
5221 br label %Loop
5222</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5223
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5224</div>
5225
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5226<!-- _______________________________________________________________________ -->
5227<div class="doc_subsubsection">
5228 <a name="i_select">'<tt>select</tt>' Instruction</a>
5229</div>
5230
5231<div class="doc_text">
5232
5233<h5>Syntax:</h5>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5234<pre>
Dan Gohmanf72fb672008-09-09 01:02:47 +0000[diff] [blame]5235 &lt;result&gt; = select <i>selty</i> &lt;cond&gt;, &lt;ty&gt; &lt;val1&gt;, &lt;ty&gt; &lt;val2&gt; <i>; yields ty</i>
5236
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5237 <i>selty</i> is either i1 or {&lt;N x i1&gt;}
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5238</pre>
5239
5240<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5241<p>The '<tt>select</tt>' instruction is used to choose one value based on a
5242 condition, without branching.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5243
5244
5245<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5246<p>The '<tt>select</tt>' instruction requires an 'i1' value or a vector of 'i1'
5247 values indicating the condition, and two values of the
5248 same <a href="#t_firstclass">first class</a> type. If the val1/val2 are
5249 vectors and the condition is a scalar, then entire vectors are selected, not
5250 individual elements.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5251
5252<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5253<p>If the condition is an i1 and it evaluates to 1, the instruction returns the
5254 first value argument; otherwise, it returns the second value argument.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5255
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5256<p>If the condition is a vector of i1, then the value arguments must be vectors
5257 of the same size, and the selection is done element by element.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5258
5259<h5>Example:</h5>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5260<pre>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]5261 %X = select i1 true, i8 17, i8 42 <i>; yields i8:17</i>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5262</pre>
Dan Gohmanc766f722009-01-22 01:39:38 +0000[diff] [blame]5263
5264<p>Note that the code generator does not yet support conditions
5265 with vector type.</p>
5266
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5267</div>
5268
Robert Bocchino05ccd702006-01-15 20:48:27 +0000[diff] [blame]5269<!-- _______________________________________________________________________ -->
5270<div class="doc_subsubsection">
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5271 <a name="i_call">'<tt>call</tt>' Instruction</a>
5272</div>
5273
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5274<div class="doc_text">
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5275
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]5276<h5>Syntax:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5277<pre>
Devang Patel307e8ab2008-10-07 17:48:33 +0000[diff] [blame]5278 &lt;result&gt; = [tail] call [<a href="#callingconv">cconv</a>] [<a href="#paramattrs">ret attrs</a>] &lt;ty&gt; [&lt;fnty&gt;*] &lt;fnptrval&gt;(&lt;function args&gt;) [<a href="#fnattrs">fn attrs</a>]
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5279</pre>
5280
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]5281<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5282<p>The '<tt>call</tt>' instruction represents a simple function call.</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5283
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]5284<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5285<p>This instruction requires several arguments:</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5286
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]5287<ol>
Jeffrey Yasskin95fa80a2010-01-09 19:44:16 +0000[diff] [blame]5288 <li>The optional "tail" marker indicates that the callee function does not
5289 access any allocas or varargs in the caller. Note that calls may be
5290 marked "tail" even if they do not occur before
5291 a <a href="#i_ret"><tt>ret</tt></a> instruction. If the "tail" marker is
5292 present, the function call is eligible for tail call optimization,
5293 but <a href="CodeGenerator.html#tailcallopt">might not in fact be
Evan Chengdc444e92010-03-08 21:05:02 +0000[diff] [blame]5294 optimized into a jump</a>. The code generator may optimize calls marked
5295 "tail" with either 1) automatic <a href="CodeGenerator.html#sibcallopt">
5296 sibling call optimization</a> when the caller and callee have
5297 matching signatures, or 2) forced tail call optimization when the
5298 following extra requirements are met:
Jeffrey Yasskin95fa80a2010-01-09 19:44:16 +0000[diff] [blame]5299 <ul>
5300 <li>Caller and callee both have the calling
5301 convention <tt>fastcc</tt>.</li>
5302 <li>The call is in tail position (ret immediately follows call and ret
5303 uses value of call or is void).</li>
5304 <li>Option <tt>-tailcallopt</tt> is enabled,
Dan Gohmanfbbee8d2010-03-02 01:08:11 +0000[diff] [blame]5305 or <code>llvm::GuaranteedTailCallOpt</code> is <code>true</code>.</li>
Jeffrey Yasskin95fa80a2010-01-09 19:44:16 +0000[diff] [blame]5306 <li><a href="CodeGenerator.html#tailcallopt">Platform specific
5307 constraints are met.</a></li>
5308 </ul>
5309 </li>
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]5310
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5311 <li>The optional "cconv" marker indicates which <a href="#callingconv">calling
5312 convention</a> the call should use. If none is specified, the call
Jeffrey Yasskin95fa80a2010-01-09 19:44:16 +0000[diff] [blame]5313 defaults to using C calling conventions. The calling convention of the
5314 call must match the calling convention of the target function, or else the
5315 behavior is undefined.</li>
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]5316
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5317 <li>The optional <a href="#paramattrs">Parameter Attributes</a> list for
5318 return values. Only '<tt>zeroext</tt>', '<tt>signext</tt>', and
5319 '<tt>inreg</tt>' attributes are valid here.</li>
5320
5321 <li>'<tt>ty</tt>': the type of the call instruction itself which is also the
5322 type of the return value. Functions that return no value are marked
5323 <tt><a href="#t_void">void</a></tt>.</li>
5324
5325 <li>'<tt>fnty</tt>': shall be the signature of the pointer to function value
5326 being invoked. The argument types must match the types implied by this
5327 signature. This type can be omitted if the function is not varargs and if
5328 the function type does not return a pointer to a function.</li>
5329
5330 <li>'<tt>fnptrval</tt>': An LLVM value containing a pointer to a function to
5331 be invoked. In most cases, this is a direct function invocation, but
5332 indirect <tt>call</tt>s are just as possible, calling an arbitrary pointer
5333 to function value.</li>
5334
5335 <li>'<tt>function args</tt>': argument list whose types match the function
Chris Lattner0724fbd2010-03-02 06:36:51 +0000[diff] [blame]5336 signature argument types and parameter attributes. All arguments must be
5337 of <a href="#t_firstclass">first class</a> type. If the function
5338 signature indicates the function accepts a variable number of arguments,
5339 the extra arguments can be specified.</li>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5340
5341 <li>The optional <a href="#fnattrs">function attributes</a> list. Only
5342 '<tt>noreturn</tt>', '<tt>nounwind</tt>', '<tt>readonly</tt>' and
5343 '<tt>readnone</tt>' attributes are valid here.</li>
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]5344</ol>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5345
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]5346<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5347<p>The '<tt>call</tt>' instruction is used to cause control flow to transfer to
5348 a specified function, with its incoming arguments bound to the specified
5349 values. Upon a '<tt><a href="#i_ret">ret</a></tt>' instruction in the called
5350 function, control flow continues with the instruction after the function
5351 call, and the return value of the function is bound to the result
5352 argument.</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5353
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]5354<h5>Example:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5355<pre>
Nick Lewyckydb7e3c92007-09-08 13:57:50 +0000[diff] [blame]5356 %retval = call i32 @test(i32 %argc)
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]5357 call i32 (i8*, ...)* @printf(i8* %msg, i32 12, i8 42) <i>; yields i32</i>
Chris Lattner772fccf2008-03-21 17:24:17 +0000[diff] [blame]5358 %X = tail call i32 @foo() <i>; yields i32</i>
5359 %Y = tail call <a href="#callingconv">fastcc</a> i32 @foo() <i>; yields i32</i>
5360 call void %foo(i8 97 signext)
Devang Patelc3fc6df2008-03-10 20:49:15 +0000[diff] [blame]5361
5362 %struct.A = type { i32, i8 }
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]5363 %r = call %struct.A @foo() <i>; yields { 32, i8 }</i>
Dan Gohmanb1e6b962008-10-04 19:00:07 +0000[diff] [blame]5364 %gr = extractvalue %struct.A %r, 0 <i>; yields i32</i>
5365 %gr1 = extractvalue %struct.A %r, 1 <i>; yields i8</i>
Chris Lattner85a350f2008-10-08 06:26:11 +0000[diff] [blame]5366 %Z = call void @foo() noreturn <i>; indicates that %foo never returns normally</i>
Matthijs Kooijmancb73d192008-10-07 10:03:45 +0000[diff] [blame]5367 %ZZ = call zeroext i32 @bar() <i>; Return value is %zero extended</i>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5368</pre>
5369
Dale Johannesen07de8d12009-09-24 18:38:21 +0000[diff] [blame]5370<p>llvm treats calls to some functions with names and arguments that match the
Dale Johannesen9f8380b2009-09-25 17:04:42 +0000[diff] [blame]5371standard C99 library as being the C99 library functions, and may perform
5372optimizations or generate code for them under that assumption. This is
5373something we'd like to change in the future to provide better support for
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]5374freestanding environments and non-C-based languages.</p>
Dale Johannesen07de8d12009-09-24 18:38:21 +0000[diff] [blame]5375
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5376</div>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5377
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5378<!-- _______________________________________________________________________ -->
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5379<div class="doc_subsubsection">
Chris Lattnerfb6977d2006-01-13 23:26:01 +0000[diff] [blame]5380 <a name="i_va_arg">'<tt>va_arg</tt>' Instruction</a>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5381</div>
5382
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5383<div class="doc_text">
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5384
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]5385<h5>Syntax:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5386<pre>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]5387 &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]5388</pre>
5389
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]5390<h5>Overview:</h5>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]5391<p>The '<tt>va_arg</tt>' instruction is used to access arguments passed through
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5392 the "variable argument" area of a function call. It is used to implement the
5393 <tt>va_arg</tt> macro in C.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5394
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]5395<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5396<p>This instruction takes a <tt>va_list*</tt> value and the type of the
5397 argument. It returns a value of the specified argument type and increments
5398 the <tt>va_list</tt> to point to the next argument. The actual type
5399 of <tt>va_list</tt> is target specific.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5400
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]5401<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5402<p>The '<tt>va_arg</tt>' instruction loads an argument of the specified type
5403 from the specified <tt>va_list</tt> and causes the <tt>va_list</tt> to point
5404 to the next argument. For more information, see the variable argument
5405 handling <a href="#int_varargs">Intrinsic Functions</a>.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5406
5407<p>It is legal for this instruction to be called in a function which does not
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5408 take a variable number of arguments, for example, the <tt>vfprintf</tt>
5409 function.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5410
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5411<p><tt>va_arg</tt> is an LLVM instruction instead of
5412 an <a href="#intrinsics">intrinsic function</a> because it takes a type as an
5413 argument.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5414
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]5415<h5>Example:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5416<p>See the <a href="#int_varargs">variable argument processing</a> section.</p>
5417
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5418<p>Note that the code generator does not yet fully support va_arg on many
5419 targets. Also, it does not currently support va_arg with aggregate types on
5420 any target.</p>
Dan Gohmanf3e60bd2009-01-12 23:12:39 +0000[diff] [blame]5421
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5422</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5423
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5424<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]5425<div class="doc_section"> <a name="intrinsics">Intrinsic Functions</a> </div>
5426<!-- *********************************************************************** -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5427
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5428<div class="doc_text">
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5429
5430<p>LLVM supports the notion of an "intrinsic function". These functions have
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5431 well known names and semantics and are required to follow certain
5432 restrictions. Overall, these intrinsics represent an extension mechanism for
5433 the LLVM language that does not require changing all of the transformations
5434 in LLVM when adding to the language (or the bitcode reader/writer, the
5435 parser, etc...).</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5436
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]5437<p>Intrinsic function names must all start with an "<tt>llvm.</tt>" prefix. This
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5438 prefix is reserved in LLVM for intrinsic names; thus, function names may not
5439 begin with this prefix. Intrinsic functions must always be external
5440 functions: you cannot define the body of intrinsic functions. Intrinsic
5441 functions may only be used in call or invoke instructions: it is illegal to
5442 take the address of an intrinsic function. Additionally, because intrinsic
5443 functions are part of the LLVM language, it is required if any are added that
5444 they be documented here.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5445
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5446<p>Some intrinsic functions can be overloaded, i.e., the intrinsic represents a
5447 family of functions that perform the same operation but on different data
5448 types. Because LLVM can represent over 8 million different integer types,
5449 overloading is used commonly to allow an intrinsic function to operate on any
5450 integer type. One or more of the argument types or the result type can be
5451 overloaded to accept any integer type. Argument types may also be defined as
5452 exactly matching a previous argument's type or the result type. This allows
5453 an intrinsic function which accepts multiple arguments, but needs all of them
5454 to be of the same type, to only be overloaded with respect to a single
5455 argument or the result.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5456
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5457<p>Overloaded intrinsics will have the names of its overloaded argument types
5458 encoded into its function name, each preceded by a period. Only those types
5459 which are overloaded result in a name suffix. Arguments whose type is matched
5460 against another type do not. For example, the <tt>llvm.ctpop</tt> function
5461 can take an integer of any width and returns an integer of exactly the same
5462 integer width. This leads to a family of functions such as
5463 <tt>i8 @llvm.ctpop.i8(i8 %val)</tt> and <tt>i29 @llvm.ctpop.i29(i29
5464 %val)</tt>. Only one type, the return type, is overloaded, and only one type
5465 suffix is required. Because the argument's type is matched against the return
5466 type, it does not require its own name suffix.</p>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]5467
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5468<p>To learn how to add an intrinsic function, please see the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5469 <a href="ExtendingLLVM.html">Extending LLVM Guide</a>.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5470
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5471</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5472
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5473<!-- ======================================================================= -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5474<div class="doc_subsection">
5475 <a name="int_varargs">Variable Argument Handling Intrinsics</a>
5476</div>
5477
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5478<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5479
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5480<p>Variable argument support is defined in LLVM with
5481 the <a href="#i_va_arg"><tt>va_arg</tt></a> instruction and these three
5482 intrinsic functions. These functions are related to the similarly named
5483 macros defined in the <tt>&lt;stdarg.h&gt;</tt> header file.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5484
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5485<p>All of these functions operate on arguments that use a target-specific value
5486 type "<tt>va_list</tt>". The LLVM assembly language reference manual does
5487 not define what this type is, so all transformations should be prepared to
5488 handle these functions regardless of the type used.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5489
Chris Lattner374ab302006-05-15 17:26:46 +0000[diff] [blame]5490<p>This example shows how the <a href="#i_va_arg"><tt>va_arg</tt></a>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5491 instruction and the variable argument handling intrinsic functions are
5492 used.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5493
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]5494<pre class="doc_code">
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]5495define i32 @test(i32 %X, ...) {
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5496 ; Initialize variable argument processing
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5497 %ap = alloca i8*
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]5498 %ap2 = bitcast i8** %ap to i8*
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]5499 call void @llvm.va_start(i8* %ap2)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5500
5501 ; Read a single integer argument
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5502 %tmp = va_arg i8** %ap, i32
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5503
5504 ; Demonstrate usage of llvm.va_copy and llvm.va_end
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5505 %aq = alloca i8*
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]5506 %aq2 = bitcast i8** %aq to i8*
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5507 call void @llvm.va_copy(i8* %aq2, i8* %ap2)
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]5508 call void @llvm.va_end(i8* %aq2)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5509
5510 ; Stop processing of arguments.
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]5511 call void @llvm.va_end(i8* %ap2)
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5512 ret i32 %tmp
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5513}
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]5514
5515declare void @llvm.va_start(i8*)
5516declare void @llvm.va_copy(i8*, i8*)
5517declare void @llvm.va_end(i8*)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5518</pre>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5519
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]5520</div>
5521
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5522<!-- _______________________________________________________________________ -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5523<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5524 <a name="int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5525</div>
5526
5527
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5528<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5529
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5530<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5531<pre>
5532 declare void %llvm.va_start(i8* &lt;arglist&gt;)
5533</pre>
5534
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5535<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5536<p>The '<tt>llvm.va_start</tt>' intrinsic initializes <tt>*&lt;arglist&gt;</tt>
5537 for subsequent use by <tt><a href="#i_va_arg">va_arg</a></tt>.</p>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]5538
5539<h5>Arguments:</h5>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5540<p>The argument is a pointer to a <tt>va_list</tt> element to initialize.</p>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]5541
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5542<h5>Semantics:</h5>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5543<p>The '<tt>llvm.va_start</tt>' intrinsic works just like the <tt>va_start</tt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5544 macro available in C. In a target-dependent way, it initializes
5545 the <tt>va_list</tt> element to which the argument points, so that the next
5546 call to <tt>va_arg</tt> will produce the first variable argument passed to
5547 the function. Unlike the C <tt>va_start</tt> macro, this intrinsic does not
5548 need to know the last argument of the function as the compiler can figure
5549 that out.</p>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]5550
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5551</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5552
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5553<!-- _______________________________________________________________________ -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5554<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5555 <a name="int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5556</div>
5557
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5558<div class="doc_text">
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]5559
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5560<h5>Syntax:</h5>
5561<pre>
5562 declare void @llvm.va_end(i8* &lt;arglist&gt;)
5563</pre>
5564
5565<h5>Overview:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5566<p>The '<tt>llvm.va_end</tt>' intrinsic destroys <tt>*&lt;arglist&gt;</tt>,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5567 which has been initialized previously
5568 with <tt><a href="#int_va_start">llvm.va_start</a></tt>
5569 or <tt><a href="#i_va_copy">llvm.va_copy</a></tt>.</p>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]5570
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5571<h5>Arguments:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5572<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]5573
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5574<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5575<p>The '<tt>llvm.va_end</tt>' intrinsic works just like the <tt>va_end</tt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5576 macro available in C. In a target-dependent way, it destroys
5577 the <tt>va_list</tt> element to which the argument points. Calls
5578 to <a href="#int_va_start"><tt>llvm.va_start</tt></a>
5579 and <a href="#int_va_copy"> <tt>llvm.va_copy</tt></a> must be matched exactly
5580 with calls to <tt>llvm.va_end</tt>.</p>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]5581
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5582</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5583
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5584<!-- _______________________________________________________________________ -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5585<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5586 <a name="int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5587</div>
5588
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5589<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5590
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5591<h5>Syntax:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5592<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]5593 declare void @llvm.va_copy(i8* &lt;destarglist&gt;, i8* &lt;srcarglist&gt;)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5594</pre>
5595
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5596<h5>Overview:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5597<p>The '<tt>llvm.va_copy</tt>' intrinsic copies the current argument position
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5598 from the source argument list to the destination argument list.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5599
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5600<h5>Arguments:</h5>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]5601<p>The first argument is a pointer to a <tt>va_list</tt> element to initialize.
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5602 The second argument is a pointer to a <tt>va_list</tt> element to copy
5603 from.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5604
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5605<h5>Semantics:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5606<p>The '<tt>llvm.va_copy</tt>' intrinsic works just like the <tt>va_copy</tt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5607 macro available in C. In a target-dependent way, it copies the
5608 source <tt>va_list</tt> element into the destination <tt>va_list</tt>
5609 element. This intrinsic is necessary because
5610 the <tt><a href="#int_va_start"> llvm.va_start</a></tt> intrinsic may be
5611 arbitrarily complex and require, for example, memory allocation.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5612
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5613</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5614
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5615<!-- ======================================================================= -->
5616<div class="doc_subsection">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5617 <a name="int_gc">Accurate Garbage Collection Intrinsics</a>
5618</div>
5619
5620<div class="doc_text">
5621
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5622<p>LLVM support for <a href="GarbageCollection.html">Accurate Garbage
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]5623Collection</a> (GC) requires the implementation and generation of these
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5624intrinsics. These intrinsics allow identification of <a href="#int_gcroot">GC
5625roots on the stack</a>, as well as garbage collector implementations that
5626require <a href="#int_gcread">read</a> and <a href="#int_gcwrite">write</a>
5627barriers. Front-ends for type-safe garbage collected languages should generate
5628these intrinsics to make use of the LLVM garbage collectors. For more details,
5629see <a href="GarbageCollection.html">Accurate Garbage Collection with
5630LLVM</a>.</p>
Christopher Lamb303dae92007-12-17 01:00:21 +0000[diff] [blame]5631
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5632<p>The garbage collection intrinsics only operate on objects in the generic
5633 address space (address space zero).</p>
Christopher Lamb303dae92007-12-17 01:00:21 +0000[diff] [blame]5634
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5635</div>
5636
5637<!-- _______________________________________________________________________ -->
5638<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5639 <a name="int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5640</div>
5641
5642<div class="doc_text">
5643
5644<h5>Syntax:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5645<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]5646 declare void @llvm.gcroot(i8** %ptrloc, i8* %metadata)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5647</pre>
5648
5649<h5>Overview:</h5>
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]5650<p>The '<tt>llvm.gcroot</tt>' intrinsic declares the existence of a GC root to
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5651 the code generator, and allows some metadata to be associated with it.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5652
5653<h5>Arguments:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5654<p>The first argument specifies the address of a stack object that contains the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5655 root pointer. The second pointer (which must be either a constant or a
5656 global value address) contains the meta-data to be associated with the
5657 root.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5658
5659<h5>Semantics:</h5>
Chris Lattner05d67092008-04-24 05:59:56 +0000[diff] [blame]5660<p>At runtime, a call to this intrinsic stores a null pointer into the "ptrloc"
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5661 location. At compile-time, the code generator generates information to allow
5662 the runtime to find the pointer at GC safe points. The '<tt>llvm.gcroot</tt>'
5663 intrinsic may only be used in a function which <a href="#gc">specifies a GC
5664 algorithm</a>.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5665
5666</div>
5667
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5668<!-- _______________________________________________________________________ -->
5669<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5670 <a name="int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5671</div>
5672
5673<div class="doc_text">
5674
5675<h5>Syntax:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5676<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]5677 declare i8* @llvm.gcread(i8* %ObjPtr, i8** %Ptr)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5678</pre>
5679
5680<h5>Overview:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5681<p>The '<tt>llvm.gcread</tt>' intrinsic identifies reads of references from heap
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5682 locations, allowing garbage collector implementations that require read
5683 barriers.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5684
5685<h5>Arguments:</h5>
Chris Lattner80626e92006-03-14 20:02:51 +0000[diff] [blame]5686<p>The second argument is the address to read from, which should be an address
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5687 allocated from the garbage collector. The first object is a pointer to the
5688 start of the referenced object, if needed by the language runtime (otherwise
5689 null).</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5690
5691<h5>Semantics:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5692<p>The '<tt>llvm.gcread</tt>' intrinsic has the same semantics as a load
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5693 instruction, but may be replaced with substantially more complex code by the
5694 garbage collector runtime, as needed. The '<tt>llvm.gcread</tt>' intrinsic
5695 may only be used in a function which <a href="#gc">specifies a GC
5696 algorithm</a>.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5697
5698</div>
5699
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5700<!-- _______________________________________________________________________ -->
5701<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5702 <a name="int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5703</div>
5704
5705<div class="doc_text">
5706
5707<h5>Syntax:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5708<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]5709 declare void @llvm.gcwrite(i8* %P1, i8* %Obj, i8** %P2)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5710</pre>
5711
5712<h5>Overview:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5713<p>The '<tt>llvm.gcwrite</tt>' intrinsic identifies writes of references to heap
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5714 locations, allowing garbage collector implementations that require write
5715 barriers (such as generational or reference counting collectors).</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5716
5717<h5>Arguments:</h5>
Chris Lattner80626e92006-03-14 20:02:51 +0000[diff] [blame]5718<p>The first argument is the reference to store, the second is the start of the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5719 object to store it to, and the third is the address of the field of Obj to
5720 store to. If the runtime does not require a pointer to the object, Obj may
5721 be null.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5722
5723<h5>Semantics:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5724<p>The '<tt>llvm.gcwrite</tt>' intrinsic has the same semantics as a store
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5725 instruction, but may be replaced with substantially more complex code by the
5726 garbage collector runtime, as needed. The '<tt>llvm.gcwrite</tt>' intrinsic
5727 may only be used in a function which <a href="#gc">specifies a GC
5728 algorithm</a>.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5729
5730</div>
5731
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5732<!-- ======================================================================= -->
5733<div class="doc_subsection">
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5734 <a name="int_codegen">Code Generator Intrinsics</a>
5735</div>
5736
5737<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5738
5739<p>These intrinsics are provided by LLVM to expose special features that may
5740 only be implemented with code generator support.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5741
5742</div>
5743
5744<!-- _______________________________________________________________________ -->
5745<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5746 <a name="int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5747</div>
5748
5749<div class="doc_text">
5750
5751<h5>Syntax:</h5>
5752<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]5753 declare i8 *@llvm.returnaddress(i32 &lt;level&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5754</pre>
5755
5756<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5757<p>The '<tt>llvm.returnaddress</tt>' intrinsic attempts to compute a
5758 target-specific value indicating the return address of the current function
5759 or one of its callers.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5760
5761<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5762<p>The argument to this intrinsic indicates which function to return the address
5763 for. Zero indicates the calling function, one indicates its caller, etc.
5764 The argument is <b>required</b> to be a constant integer value.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5765
5766<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5767<p>The '<tt>llvm.returnaddress</tt>' intrinsic either returns a pointer
5768 indicating the return address of the specified call frame, or zero if it
5769 cannot be identified. The value returned by this intrinsic is likely to be
5770 incorrect or 0 for arguments other than zero, so it should only be used for
5771 debugging purposes.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5772
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5773<p>Note that calling this intrinsic does not prevent function inlining or other
5774 aggressive transformations, so the value returned may not be that of the
5775 obvious source-language caller.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5776
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5777</div>
5778
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5779<!-- _______________________________________________________________________ -->
5780<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5781 <a name="int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5782</div>
5783
5784<div class="doc_text">
5785
5786<h5>Syntax:</h5>
5787<pre>
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]5788 declare i8* @llvm.frameaddress(i32 &lt;level&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5789</pre>
5790
5791<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5792<p>The '<tt>llvm.frameaddress</tt>' intrinsic attempts to return the
5793 target-specific frame pointer value for the specified stack frame.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5794
5795<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5796<p>The argument to this intrinsic indicates which function to return the frame
5797 pointer for. Zero indicates the calling function, one indicates its caller,
5798 etc. The argument is <b>required</b> to be a constant integer value.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5799
5800<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5801<p>The '<tt>llvm.frameaddress</tt>' intrinsic either returns a pointer
5802 indicating the frame address of the specified call frame, or zero if it
5803 cannot be identified. The value returned by this intrinsic is likely to be
5804 incorrect or 0 for arguments other than zero, so it should only be used for
5805 debugging purposes.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5806
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5807<p>Note that calling this intrinsic does not prevent function inlining or other
5808 aggressive transformations, so the value returned may not be that of the
5809 obvious source-language caller.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5810
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5811</div>
5812
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]5813<!-- _______________________________________________________________________ -->
5814<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5815 <a name="int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]5816</div>
5817
5818<div class="doc_text">
5819
5820<h5>Syntax:</h5>
5821<pre>
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]5822 declare i8* @llvm.stacksave()
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]5823</pre>
5824
5825<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5826<p>The '<tt>llvm.stacksave</tt>' intrinsic is used to remember the current state
5827 of the function stack, for use
5828 with <a href="#int_stackrestore"> <tt>llvm.stackrestore</tt></a>. This is
5829 useful for implementing language features like scoped automatic variable
5830 sized arrays in C99.</p>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]5831
5832<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5833<p>This intrinsic returns a opaque pointer value that can be passed
5834 to <a href="#int_stackrestore"><tt>llvm.stackrestore</tt></a>. When
5835 an <tt>llvm.stackrestore</tt> intrinsic is executed with a value saved
5836 from <tt>llvm.stacksave</tt>, it effectively restores the state of the stack
5837 to the state it was in when the <tt>llvm.stacksave</tt> intrinsic executed.
5838 In practice, this pops any <a href="#i_alloca">alloca</a> blocks from the
5839 stack that were allocated after the <tt>llvm.stacksave</tt> was executed.</p>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]5840
5841</div>
5842
5843<!-- _______________________________________________________________________ -->
5844<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5845 <a name="int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]5846</div>
5847
5848<div class="doc_text">
5849
5850<h5>Syntax:</h5>
5851<pre>
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]5852 declare void @llvm.stackrestore(i8* %ptr)
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]5853</pre>
5854
5855<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5856<p>The '<tt>llvm.stackrestore</tt>' intrinsic is used to restore the state of
5857 the function stack to the state it was in when the
5858 corresponding <a href="#int_stacksave"><tt>llvm.stacksave</tt></a> intrinsic
5859 executed. This is useful for implementing language features like scoped
5860 automatic variable sized arrays in C99.</p>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]5861
5862<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5863<p>See the description
5864 for <a href="#int_stacksave"><tt>llvm.stacksave</tt></a>.</p>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]5865
5866</div>
5867
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]5868<!-- _______________________________________________________________________ -->
5869<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5870 <a name="int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a>
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]5871</div>
5872
5873<div class="doc_text">
5874
5875<h5>Syntax:</h5>
5876<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]5877 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]5878</pre>
5879
5880<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5881<p>The '<tt>llvm.prefetch</tt>' intrinsic is a hint to the code generator to
5882 insert a prefetch instruction if supported; otherwise, it is a noop.
5883 Prefetches have no effect on the behavior of the program but can change its
5884 performance characteristics.</p>
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]5885
5886<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5887<p><tt>address</tt> is the address to be prefetched, <tt>rw</tt> is the
5888 specifier determining if the fetch should be for a read (0) or write (1),
5889 and <tt>locality</tt> is a temporal locality specifier ranging from (0) - no
5890 locality, to (3) - extremely local keep in cache. The <tt>rw</tt>
5891 and <tt>locality</tt> arguments must be constant integers.</p>
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]5892
5893<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5894<p>This intrinsic does not modify the behavior of the program. In particular,
5895 prefetches cannot trap and do not produce a value. On targets that support
5896 this intrinsic, the prefetch can provide hints to the processor cache for
5897 better performance.</p>
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]5898
5899</div>
5900
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]5901<!-- _______________________________________________________________________ -->
5902<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5903 <a name="int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a>
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]5904</div>
5905
5906<div class="doc_text">
5907
5908<h5>Syntax:</h5>
5909<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]5910 declare void @llvm.pcmarker(i32 &lt;id&gt;)
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]5911</pre>
5912
5913<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5914<p>The '<tt>llvm.pcmarker</tt>' intrinsic is a method to export a Program
5915 Counter (PC) in a region of code to simulators and other tools. The method
5916 is target specific, but it is expected that the marker will use exported
5917 symbols to transmit the PC of the marker. The marker makes no guarantees
5918 that it will remain with any specific instruction after optimizations. It is
5919 possible that the presence of a marker will inhibit optimizations. The
5920 intended use is to be inserted after optimizations to allow correlations of
5921 simulation runs.</p>
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]5922
5923<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5924<p><tt>id</tt> is a numerical id identifying the marker.</p>
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]5925
5926<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5927<p>This intrinsic does not modify the behavior of the program. Backends that do
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]5928 not support this intrinsic may ignore it.</p>
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]5929
5930</div>
5931
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]5932<!-- _______________________________________________________________________ -->
5933<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5934 <a name="int_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a>
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]5935</div>
5936
5937<div class="doc_text">
5938
5939<h5>Syntax:</h5>
5940<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]5941 declare i64 @llvm.readcyclecounter()
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]5942</pre>
5943
5944<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5945<p>The '<tt>llvm.readcyclecounter</tt>' intrinsic provides access to the cycle
5946 counter register (or similar low latency, high accuracy clocks) on those
5947 targets that support it. On X86, it should map to RDTSC. On Alpha, it
5948 should map to RPCC. As the backing counters overflow quickly (on the order
5949 of 9 seconds on alpha), this should only be used for small timings.</p>
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]5950
5951<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5952<p>When directly supported, reading the cycle counter should not modify any
5953 memory. Implementations are allowed to either return a application specific
5954 value or a system wide value. On backends without support, this is lowered
5955 to a constant 0.</p>
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]5956
5957</div>
5958
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5959<!-- ======================================================================= -->
5960<div class="doc_subsection">
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5961 <a name="int_libc">Standard C Library Intrinsics</a>
5962</div>
5963
5964<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5965
5966<p>LLVM provides intrinsics for a few important standard C library functions.
5967 These intrinsics allow source-language front-ends to pass information about
5968 the alignment of the pointer arguments to the code generator, providing
5969 opportunity for more efficient code generation.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5970
5971</div>
5972
5973<!-- _______________________________________________________________________ -->
5974<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5975 <a name="int_memcpy">'<tt>llvm.memcpy</tt>' Intrinsic</a>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5976</div>
5977
5978<div class="doc_text">
5979
5980<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5981<p>This is an overloaded intrinsic. You can use <tt>llvm.memcpy</tt> on any
Mon P Wange88909b2010-04-07 06:35:53 +0000[diff] [blame]5982 integer bit width and for different address spaces. Not all targets support
5983 all bit widths however.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5984
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5985<pre>
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]5986 declare void @llvm.memcpy.p0i8.p0i8.i32(i8* &lt;dest&gt;, i8* &lt;src&gt;,
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]5987 i32 &lt;len&gt;, i32 &lt;align&gt;, i1 &lt;isvolatile&gt;)
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]5988 declare void @llvm.memcpy.p0i8.p0i8.i64(i8* &lt;dest&gt;, i8* &lt;src&gt;,
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]5989 i64 &lt;len&gt;, i32 &lt;align&gt;, i1 &lt;isvolatile&gt;)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5990</pre>
5991
5992<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5993<p>The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the
5994 source location to the destination location.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5995
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5996<p>Note that, unlike the standard libc function, the <tt>llvm.memcpy.*</tt>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]5997 intrinsics do not return a value, takes extra alignment/isvolatile arguments
5998 and the pointers can be in specified address spaces.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5999
6000<h5>Arguments:</h5>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6001
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6002<p>The first argument is a pointer to the destination, the second is a pointer
6003 to the source. The third argument is an integer argument specifying the
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6004 number of bytes to copy, the fourth argument is the alignment of the
6005 source and destination locations, and the fifth is a boolean indicating a
6006 volatile access.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6007
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]6008<p>If the call to this intrinsic has an alignment value that is not 0 or 1,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6009 then the caller guarantees that both the source and destination pointers are
6010 aligned to that boundary.</p>
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]6011
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]6012<p>If the <tt>isvolatile</tt> parameter is <tt>true</tt>, the
6013 <tt>llvm.memcpy</tt> call is a <a href="#volatile">volatile operation</a>.
6014 The detailed access behavior is not very cleanly specified and it is unwise
6015 to depend on it.</p>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6016
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6017<h5>Semantics:</h5>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6018
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6019<p>The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the
6020 source location to the destination location, which are not allowed to
6021 overlap. It copies "len" bytes of memory over. If the argument is known to
6022 be aligned to some boundary, this can be specified as the fourth argument,
6023 otherwise it should be set to 0 or 1.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6024
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6025</div>
6026
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6027<!-- _______________________________________________________________________ -->
6028<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6029 <a name="int_memmove">'<tt>llvm.memmove</tt>' Intrinsic</a>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6030</div>
6031
6032<div class="doc_text">
6033
6034<h5>Syntax:</h5>
Chris Lattner824b9582008-11-21 16:42:48 +0000[diff] [blame]6035<p>This is an overloaded intrinsic. You can use llvm.memmove on any integer bit
Mon P Wange88909b2010-04-07 06:35:53 +0000[diff] [blame]6036 width and for different address space. Not all targets support all bit
6037 widths however.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6038
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6039<pre>
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]6040 declare void @llvm.memmove.p0i8.p0i8.i32(i8* &lt;dest&gt;, i8* &lt;src&gt;,
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6041 i32 &lt;len&gt;, i32 &lt;align&gt;, i1 &lt;isvolatile&gt;)
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]6042 declare void @llvm.memmove.p0i8.p0i8.i64(i8* &lt;dest&gt;, i8* &lt;src&gt;,
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6043 i64 &lt;len&gt;, i32 &lt;align&gt;, i1 &lt;isvolatile&gt;)
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6044</pre>
6045
6046<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6047<p>The '<tt>llvm.memmove.*</tt>' intrinsics move a block of memory from the
6048 source location to the destination location. It is similar to the
6049 '<tt>llvm.memcpy</tt>' intrinsic but allows the two memory locations to
6050 overlap.</p>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6051
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6052<p>Note that, unlike the standard libc function, the <tt>llvm.memmove.*</tt>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6053 intrinsics do not return a value, takes extra alignment/isvolatile arguments
6054 and the pointers can be in specified address spaces.</p>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6055
6056<h5>Arguments:</h5>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6057
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6058<p>The first argument is a pointer to the destination, the second is a pointer
6059 to the source. The third argument is an integer argument specifying the
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6060 number of bytes to copy, the fourth argument is the alignment of the
6061 source and destination locations, and the fifth is a boolean indicating a
6062 volatile access.</p>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6063
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]6064<p>If the call to this intrinsic has an alignment value that is not 0 or 1,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6065 then the caller guarantees that the source and destination pointers are
6066 aligned to that boundary.</p>
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]6067
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]6068<p>If the <tt>isvolatile</tt> parameter is <tt>true</tt>, the
6069 <tt>llvm.memmove</tt> call is a <a href="#volatile">volatile operation</a>.
6070 The detailed access behavior is not very cleanly specified and it is unwise
6071 to depend on it.</p>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6072
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6073<h5>Semantics:</h5>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6074
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6075<p>The '<tt>llvm.memmove.*</tt>' intrinsics copy a block of memory from the
6076 source location to the destination location, which may overlap. It copies
6077 "len" bytes of memory over. If the argument is known to be aligned to some
6078 boundary, this can be specified as the fourth argument, otherwise it should
6079 be set to 0 or 1.</p>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6080
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6081</div>
6082
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6083<!-- _______________________________________________________________________ -->
6084<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6085 <a name="int_memset">'<tt>llvm.memset.*</tt>' Intrinsics</a>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6086</div>
6087
6088<div class="doc_text">
6089
6090<h5>Syntax:</h5>
Chris Lattner824b9582008-11-21 16:42:48 +0000[diff] [blame]6091<p>This is an overloaded intrinsic. You can use llvm.memset on any integer bit
John Criswellcdcbbfc2010-07-30 16:30:28 +0000[diff] [blame]6092 width and for different address spaces. However, not all targets support all
6093 bit widths.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6094
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6095<pre>
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]6096 declare void @llvm.memset.p0i8.i32(i8* &lt;dest&gt;, i8 &lt;val&gt;,
Chris Lattnerff35c3f2010-04-08 00:54:34 +0000[diff] [blame]6097 i32 &lt;len&gt;, i32 &lt;align&gt;, i1 &lt;isvolatile&gt;)
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]6098 declare void @llvm.memset.p0i8.i64(i8* &lt;dest&gt;, i8 &lt;val&gt;,
Chris Lattnerff35c3f2010-04-08 00:54:34 +0000[diff] [blame]6099 i64 &lt;len&gt;, i32 &lt;align&gt;, i1 &lt;isvolatile&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6100</pre>
6101
6102<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6103<p>The '<tt>llvm.memset.*</tt>' intrinsics fill a block of memory with a
6104 particular byte value.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6105
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6106<p>Note that, unlike the standard libc function, the <tt>llvm.memset</tt>
John Criswellcdcbbfc2010-07-30 16:30:28 +0000[diff] [blame]6107 intrinsic does not return a value and takes extra alignment/volatile
6108 arguments. Also, the destination can be in an arbitrary address space.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6109
6110<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6111<p>The first argument is a pointer to the destination to fill, the second is the
John Criswellcdcbbfc2010-07-30 16:30:28 +0000[diff] [blame]6112 byte value with which to fill it, the third argument is an integer argument
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6113 specifying the number of bytes to fill, and the fourth argument is the known
John Criswellcdcbbfc2010-07-30 16:30:28 +0000[diff] [blame]6114 alignment of the destination location.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6115
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]6116<p>If the call to this intrinsic has an alignment value that is not 0 or 1,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6117 then the caller guarantees that the destination pointer is aligned to that
6118 boundary.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6119
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]6120<p>If the <tt>isvolatile</tt> parameter is <tt>true</tt>, the
6121 <tt>llvm.memset</tt> call is a <a href="#volatile">volatile operation</a>.
6122 The detailed access behavior is not very cleanly specified and it is unwise
6123 to depend on it.</p>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6124
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6125<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6126<p>The '<tt>llvm.memset.*</tt>' intrinsics fill "len" bytes of memory starting
6127 at the destination location. If the argument is known to be aligned to some
6128 boundary, this can be specified as the fourth argument, otherwise it should
6129 be set to 0 or 1.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6130
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6131</div>
6132
Chris Lattner32006282004-06-11 02:28:03 +0000[diff] [blame]6133<!-- _______________________________________________________________________ -->
6134<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6135 <a name="int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a>
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]6136</div>
6137
6138<div class="doc_text">
6139
6140<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6141<p>This is an overloaded intrinsic. You can use <tt>llvm.sqrt</tt> on any
6142 floating point or vector of floating point type. Not all targets support all
6143 types however.</p>
6144
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]6145<pre>
Dale Johannesen408f9c12007-10-02 17:47:38 +0000[diff] [blame]6146 declare float @llvm.sqrt.f32(float %Val)
6147 declare double @llvm.sqrt.f64(double %Val)
6148 declare x86_fp80 @llvm.sqrt.f80(x86_fp80 %Val)
6149 declare fp128 @llvm.sqrt.f128(fp128 %Val)
6150 declare ppc_fp128 @llvm.sqrt.ppcf128(ppc_fp128 %Val)
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]6151</pre>
6152
6153<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6154<p>The '<tt>llvm.sqrt</tt>' intrinsics return the sqrt of the specified operand,
6155 returning the same value as the libm '<tt>sqrt</tt>' functions would.
6156 Unlike <tt>sqrt</tt> in libm, however, <tt>llvm.sqrt</tt> has undefined
6157 behavior for negative numbers other than -0.0 (which allows for better
6158 optimization, because there is no need to worry about errno being
6159 set). <tt>llvm.sqrt(-0.0)</tt> is defined to return -0.0 like IEEE sqrt.</p>
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]6160
6161<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6162<p>The argument and return value are floating point numbers of the same
6163 type.</p>
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]6164
6165<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6166<p>This function returns the sqrt of the specified operand if it is a
6167 nonnegative floating point number.</p>
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]6168
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]6169</div>
6170
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6171<!-- _______________________________________________________________________ -->
6172<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6173 <a name="int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6174</div>
6175
6176<div class="doc_text">
6177
6178<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6179<p>This is an overloaded intrinsic. You can use <tt>llvm.powi</tt> on any
6180 floating point or vector of floating point type. Not all targets support all
6181 types however.</p>
6182
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6183<pre>
Dale Johannesen408f9c12007-10-02 17:47:38 +0000[diff] [blame]6184 declare float @llvm.powi.f32(float %Val, i32 %power)
6185 declare double @llvm.powi.f64(double %Val, i32 %power)
6186 declare x86_fp80 @llvm.powi.f80(x86_fp80 %Val, i32 %power)
6187 declare fp128 @llvm.powi.f128(fp128 %Val, i32 %power)
6188 declare ppc_fp128 @llvm.powi.ppcf128(ppc_fp128 %Val, i32 %power)
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6189</pre>
6190
6191<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6192<p>The '<tt>llvm.powi.*</tt>' intrinsics return the first operand raised to the
6193 specified (positive or negative) power. The order of evaluation of
6194 multiplications is not defined. When a vector of floating point type is
6195 used, the second argument remains a scalar integer value.</p>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6196
6197<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6198<p>The second argument is an integer power, and the first is a value to raise to
6199 that power.</p>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6200
6201<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6202<p>This function returns the first value raised to the second power with an
6203 unspecified sequence of rounding operations.</p>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6204
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6205</div>
6206
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6207<!-- _______________________________________________________________________ -->
6208<div class="doc_subsubsection">
6209 <a name="int_sin">'<tt>llvm.sin.*</tt>' Intrinsic</a>
6210</div>
6211
6212<div class="doc_text">
6213
6214<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6215<p>This is an overloaded intrinsic. You can use <tt>llvm.sin</tt> on any
6216 floating point or vector of floating point type. Not all targets support all
6217 types however.</p>
6218
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6219<pre>
6220 declare float @llvm.sin.f32(float %Val)
6221 declare double @llvm.sin.f64(double %Val)
6222 declare x86_fp80 @llvm.sin.f80(x86_fp80 %Val)
6223 declare fp128 @llvm.sin.f128(fp128 %Val)
6224 declare ppc_fp128 @llvm.sin.ppcf128(ppc_fp128 %Val)
6225</pre>
6226
6227<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6228<p>The '<tt>llvm.sin.*</tt>' intrinsics return the sine of the operand.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6229
6230<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6231<p>The argument and return value are floating point numbers of the same
6232 type.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6233
6234<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6235<p>This function returns the sine of the specified operand, returning the same
6236 values as the libm <tt>sin</tt> functions would, and handles error conditions
6237 in the same way.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6238
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6239</div>
6240
6241<!-- _______________________________________________________________________ -->
6242<div class="doc_subsubsection">
6243 <a name="int_cos">'<tt>llvm.cos.*</tt>' Intrinsic</a>
6244</div>
6245
6246<div class="doc_text">
6247
6248<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6249<p>This is an overloaded intrinsic. You can use <tt>llvm.cos</tt> on any
6250 floating point or vector of floating point type. Not all targets support all
6251 types however.</p>
6252
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6253<pre>
6254 declare float @llvm.cos.f32(float %Val)
6255 declare double @llvm.cos.f64(double %Val)
6256 declare x86_fp80 @llvm.cos.f80(x86_fp80 %Val)
6257 declare fp128 @llvm.cos.f128(fp128 %Val)
6258 declare ppc_fp128 @llvm.cos.ppcf128(ppc_fp128 %Val)
6259</pre>
6260
6261<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6262<p>The '<tt>llvm.cos.*</tt>' intrinsics return the cosine of the operand.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6263
6264<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6265<p>The argument and return value are floating point numbers of the same
6266 type.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6267
6268<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6269<p>This function returns the cosine of the specified operand, returning the same
6270 values as the libm <tt>cos</tt> functions would, and handles error conditions
6271 in the same way.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6272
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6273</div>
6274
6275<!-- _______________________________________________________________________ -->
6276<div class="doc_subsubsection">
6277 <a name="int_pow">'<tt>llvm.pow.*</tt>' Intrinsic</a>
6278</div>
6279
6280<div class="doc_text">
6281
6282<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6283<p>This is an overloaded intrinsic. You can use <tt>llvm.pow</tt> on any
6284 floating point or vector of floating point type. Not all targets support all
6285 types however.</p>
6286
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6287<pre>
6288 declare float @llvm.pow.f32(float %Val, float %Power)
6289 declare double @llvm.pow.f64(double %Val, double %Power)
6290 declare x86_fp80 @llvm.pow.f80(x86_fp80 %Val, x86_fp80 %Power)
6291 declare fp128 @llvm.pow.f128(fp128 %Val, fp128 %Power)
6292 declare ppc_fp128 @llvm.pow.ppcf128(ppc_fp128 %Val, ppc_fp128 Power)
6293</pre>
6294
6295<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6296<p>The '<tt>llvm.pow.*</tt>' intrinsics return the first operand raised to the
6297 specified (positive or negative) power.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6298
6299<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6300<p>The second argument is a floating point power, and the first is a value to
6301 raise to that power.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6302
6303<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6304<p>This function returns the first value raised to the second power, returning
6305 the same values as the libm <tt>pow</tt> functions would, and handles error
6306 conditions in the same way.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6307
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6308</div>
6309
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6310<!-- ======================================================================= -->
6311<div class="doc_subsection">
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]6312 <a name="int_manip">Bit Manipulation Intrinsics</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6313</div>
6314
6315<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6316
6317<p>LLVM provides intrinsics for a few important bit manipulation operations.
6318 These allow efficient code generation for some algorithms.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6319
6320</div>
6321
6322<!-- _______________________________________________________________________ -->
6323<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6324 <a name="int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]6325</div>
6326
6327<div class="doc_text">
6328
6329<h5>Syntax:</h5>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]6330<p>This is an overloaded intrinsic function. You can use bswap on any integer
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6331 type that is an even number of bytes (i.e. BitWidth % 16 == 0).</p>
6332
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]6333<pre>
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]6334 declare i16 @llvm.bswap.i16(i16 &lt;id&gt;)
6335 declare i32 @llvm.bswap.i32(i32 &lt;id&gt;)
6336 declare i64 @llvm.bswap.i64(i64 &lt;id&gt;)
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]6337</pre>
6338
6339<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6340<p>The '<tt>llvm.bswap</tt>' family of intrinsics is used to byte swap integer
6341 values with an even number of bytes (positive multiple of 16 bits). These
6342 are useful for performing operations on data that is not in the target's
6343 native byte order.</p>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]6344
6345<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6346<p>The <tt>llvm.bswap.i16</tt> intrinsic returns an i16 value that has the high
6347 and low byte of the input i16 swapped. Similarly,
6348 the <tt>llvm.bswap.i32</tt> intrinsic returns an i32 value that has the four
6349 bytes of the input i32 swapped, so that if the input bytes are numbered 0, 1,
6350 2, 3 then the returned i32 will have its bytes in 3, 2, 1, 0 order.
6351 The <tt>llvm.bswap.i48</tt>, <tt>llvm.bswap.i64</tt> and other intrinsics
6352 extend this concept to additional even-byte lengths (6 bytes, 8 bytes and
6353 more, respectively).</p>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]6354
6355</div>
6356
6357<!-- _______________________________________________________________________ -->
6358<div class="doc_subsubsection">
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]6359 <a name="int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6360</div>
6361
6362<div class="doc_text">
6363
6364<h5>Syntax:</h5>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]6365<p>This is an overloaded intrinsic. You can use llvm.ctpop on any integer bit
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6366 width. Not all targets support all bit widths however.</p>
6367
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6368<pre>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6369 declare i8 @llvm.ctpop.i8(i8 &lt;src&gt;)
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]6370 declare i16 @llvm.ctpop.i16(i16 &lt;src&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]6371 declare i32 @llvm.ctpop.i32(i32 &lt;src&gt;)
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]6372 declare i64 @llvm.ctpop.i64(i64 &lt;src&gt;)
6373 declare i256 @llvm.ctpop.i256(i256 &lt;src&gt;)
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6374</pre>
6375
6376<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6377<p>The '<tt>llvm.ctpop</tt>' family of intrinsics counts the number of bits set
6378 in a value.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6379
6380<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6381<p>The only argument is the value to be counted. The argument may be of any
6382 integer type. The return type must match the argument type.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6383
6384<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6385<p>The '<tt>llvm.ctpop</tt>' intrinsic counts the 1's in a variable.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6386
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6387</div>
6388
6389<!-- _______________________________________________________________________ -->
6390<div class="doc_subsubsection">
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]6391 <a name="int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6392</div>
6393
6394<div class="doc_text">
6395
6396<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6397<p>This is an overloaded intrinsic. You can use <tt>llvm.ctlz</tt> on any
6398 integer bit width. Not all targets support all bit widths however.</p>
6399
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6400<pre>
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]6401 declare i8 @llvm.ctlz.i8 (i8 &lt;src&gt;)
6402 declare i16 @llvm.ctlz.i16(i16 &lt;src&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]6403 declare i32 @llvm.ctlz.i32(i32 &lt;src&gt;)
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]6404 declare i64 @llvm.ctlz.i64(i64 &lt;src&gt;)
6405 declare i256 @llvm.ctlz.i256(i256 &lt;src&gt;)
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6406</pre>
6407
6408<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6409<p>The '<tt>llvm.ctlz</tt>' family of intrinsic functions counts the number of
6410 leading zeros in a variable.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6411
6412<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6413<p>The only argument is the value to be counted. The argument may be of any
6414 integer type. The return type must match the argument type.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6415
6416<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6417<p>The '<tt>llvm.ctlz</tt>' intrinsic counts the leading (most significant)
6418 zeros in a variable. If the src == 0 then the result is the size in bits of
6419 the type of src. For example, <tt>llvm.ctlz(i32 2) = 30</tt>.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6420
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6421</div>
Chris Lattner32006282004-06-11 02:28:03 +0000[diff] [blame]6422
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]6423<!-- _______________________________________________________________________ -->
6424<div class="doc_subsubsection">
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]6425 <a name="int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic</a>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]6426</div>
6427
6428<div class="doc_text">
6429
6430<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6431<p>This is an overloaded intrinsic. You can use <tt>llvm.cttz</tt> on any
6432 integer bit width. Not all targets support all bit widths however.</p>
6433
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]6434<pre>
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]6435 declare i8 @llvm.cttz.i8 (i8 &lt;src&gt;)
6436 declare i16 @llvm.cttz.i16(i16 &lt;src&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]6437 declare i32 @llvm.cttz.i32(i32 &lt;src&gt;)
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]6438 declare i64 @llvm.cttz.i64(i64 &lt;src&gt;)
6439 declare i256 @llvm.cttz.i256(i256 &lt;src&gt;)
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]6440</pre>
6441
6442<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6443<p>The '<tt>llvm.cttz</tt>' family of intrinsic functions counts the number of
6444 trailing zeros.</p>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]6445
6446<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6447<p>The only argument is the value to be counted. The argument may be of any
6448 integer type. The return type must match the argument type.</p>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]6449
6450<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6451<p>The '<tt>llvm.cttz</tt>' intrinsic counts the trailing (least significant)
6452 zeros in a variable. If the src == 0 then the result is the size in bits of
6453 the type of src. For example, <tt>llvm.cttz(2) = 1</tt>.</p>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]6454
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]6455</div>
6456
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]6457<!-- ======================================================================= -->
6458<div class="doc_subsection">
6459 <a name="int_overflow">Arithmetic with Overflow Intrinsics</a>
6460</div>
6461
6462<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6463
6464<p>LLVM provides intrinsics for some arithmetic with overflow operations.</p>
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]6465
6466</div>
6467
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6468<!-- _______________________________________________________________________ -->
6469<div class="doc_subsubsection">
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]6470 <a name="int_sadd_overflow">'<tt>llvm.sadd.with.overflow.*</tt>' Intrinsics</a>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6471</div>
6472
6473<div class="doc_text">
6474
6475<h5>Syntax:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6476<p>This is an overloaded intrinsic. You can use <tt>llvm.sadd.with.overflow</tt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6477 on any integer bit width.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6478
6479<pre>
6480 declare {i16, i1} @llvm.sadd.with.overflow.i16(i16 %a, i16 %b)
6481 declare {i32, i1} @llvm.sadd.with.overflow.i32(i32 %a, i32 %b)
6482 declare {i64, i1} @llvm.sadd.with.overflow.i64(i64 %a, i64 %b)
6483</pre>
6484
6485<h5>Overview:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6486<p>The '<tt>llvm.sadd.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6487 a signed addition of the two arguments, and indicate whether an overflow
6488 occurred during the signed summation.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6489
6490<h5>Arguments:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6491<p>The arguments (%a and %b) and the first element of the result structure may
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6492 be of integer types of any bit width, but they must have the same bit
6493 width. The second element of the result structure must be of
6494 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
6495 undergo signed addition.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6496
6497<h5>Semantics:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6498<p>The '<tt>llvm.sadd.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6499 a signed addition of the two variables. They return a structure &mdash; the
6500 first element of which is the signed summation, and the second element of
6501 which is a bit specifying if the signed summation resulted in an
6502 overflow.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6503
6504<h5>Examples:</h5>
6505<pre>
6506 %res = call {i32, i1} @llvm.sadd.with.overflow.i32(i32 %a, i32 %b)
6507 %sum = extractvalue {i32, i1} %res, 0
6508 %obit = extractvalue {i32, i1} %res, 1
6509 br i1 %obit, label %overflow, label %normal
6510</pre>
6511
6512</div>
6513
6514<!-- _______________________________________________________________________ -->
6515<div class="doc_subsubsection">
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]6516 <a name="int_uadd_overflow">'<tt>llvm.uadd.with.overflow.*</tt>' Intrinsics</a>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6517</div>
6518
6519<div class="doc_text">
6520
6521<h5>Syntax:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6522<p>This is an overloaded intrinsic. You can use <tt>llvm.uadd.with.overflow</tt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6523 on any integer bit width.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6524
6525<pre>
6526 declare {i16, i1} @llvm.uadd.with.overflow.i16(i16 %a, i16 %b)
6527 declare {i32, i1} @llvm.uadd.with.overflow.i32(i32 %a, i32 %b)
6528 declare {i64, i1} @llvm.uadd.with.overflow.i64(i64 %a, i64 %b)
6529</pre>
6530
6531<h5>Overview:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6532<p>The '<tt>llvm.uadd.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6533 an unsigned addition of the two arguments, and indicate whether a carry
6534 occurred during the unsigned summation.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6535
6536<h5>Arguments:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6537<p>The arguments (%a and %b) and the first element of the result structure may
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6538 be of integer types of any bit width, but they must have the same bit
6539 width. The second element of the result structure must be of
6540 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
6541 undergo unsigned addition.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6542
6543<h5>Semantics:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6544<p>The '<tt>llvm.uadd.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6545 an unsigned addition of the two arguments. They return a structure &mdash;
6546 the first element of which is the sum, and the second element of which is a
6547 bit specifying if the unsigned summation resulted in a carry.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6548
6549<h5>Examples:</h5>
6550<pre>
6551 %res = call {i32, i1} @llvm.uadd.with.overflow.i32(i32 %a, i32 %b)
6552 %sum = extractvalue {i32, i1} %res, 0
6553 %obit = extractvalue {i32, i1} %res, 1
6554 br i1 %obit, label %carry, label %normal
6555</pre>
6556
6557</div>
6558
6559<!-- _______________________________________________________________________ -->
6560<div class="doc_subsubsection">
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]6561 <a name="int_ssub_overflow">'<tt>llvm.ssub.with.overflow.*</tt>' Intrinsics</a>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6562</div>
6563
6564<div class="doc_text">
6565
6566<h5>Syntax:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6567<p>This is an overloaded intrinsic. You can use <tt>llvm.ssub.with.overflow</tt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6568 on any integer bit width.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6569
6570<pre>
6571 declare {i16, i1} @llvm.ssub.with.overflow.i16(i16 %a, i16 %b)
6572 declare {i32, i1} @llvm.ssub.with.overflow.i32(i32 %a, i32 %b)
6573 declare {i64, i1} @llvm.ssub.with.overflow.i64(i64 %a, i64 %b)
6574</pre>
6575
6576<h5>Overview:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6577<p>The '<tt>llvm.ssub.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6578 a signed subtraction of the two arguments, and indicate whether an overflow
6579 occurred during the signed subtraction.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6580
6581<h5>Arguments:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6582<p>The arguments (%a and %b) and the first element of the result structure may
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6583 be of integer types of any bit width, but they must have the same bit
6584 width. The second element of the result structure must be of
6585 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
6586 undergo signed subtraction.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6587
6588<h5>Semantics:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6589<p>The '<tt>llvm.ssub.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6590 a signed subtraction of the two arguments. They return a structure &mdash;
6591 the first element of which is the subtraction, and the second element of
6592 which is a bit specifying if the signed subtraction resulted in an
6593 overflow.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6594
6595<h5>Examples:</h5>
6596<pre>
6597 %res = call {i32, i1} @llvm.ssub.with.overflow.i32(i32 %a, i32 %b)
6598 %sum = extractvalue {i32, i1} %res, 0
6599 %obit = extractvalue {i32, i1} %res, 1
6600 br i1 %obit, label %overflow, label %normal
6601</pre>
6602
6603</div>
6604
6605<!-- _______________________________________________________________________ -->
6606<div class="doc_subsubsection">
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]6607 <a name="int_usub_overflow">'<tt>llvm.usub.with.overflow.*</tt>' Intrinsics</a>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6608</div>
6609
6610<div class="doc_text">
6611
6612<h5>Syntax:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6613<p>This is an overloaded intrinsic. You can use <tt>llvm.usub.with.overflow</tt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6614 on any integer bit width.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6615
6616<pre>
6617 declare {i16, i1} @llvm.usub.with.overflow.i16(i16 %a, i16 %b)
6618 declare {i32, i1} @llvm.usub.with.overflow.i32(i32 %a, i32 %b)
6619 declare {i64, i1} @llvm.usub.with.overflow.i64(i64 %a, i64 %b)
6620</pre>
6621
6622<h5>Overview:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6623<p>The '<tt>llvm.usub.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6624 an unsigned subtraction of the two arguments, and indicate whether an
6625 overflow occurred during the unsigned subtraction.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6626
6627<h5>Arguments:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6628<p>The arguments (%a and %b) and the first element of the result structure may
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6629 be of integer types of any bit width, but they must have the same bit
6630 width. The second element of the result structure must be of
6631 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
6632 undergo unsigned subtraction.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6633
6634<h5>Semantics:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6635<p>The '<tt>llvm.usub.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6636 an unsigned subtraction of the two arguments. They return a structure &mdash;
6637 the first element of which is the subtraction, and the second element of
6638 which is a bit specifying if the unsigned subtraction resulted in an
6639 overflow.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6640
6641<h5>Examples:</h5>
6642<pre>
6643 %res = call {i32, i1} @llvm.usub.with.overflow.i32(i32 %a, i32 %b)
6644 %sum = extractvalue {i32, i1} %res, 0
6645 %obit = extractvalue {i32, i1} %res, 1
6646 br i1 %obit, label %overflow, label %normal
6647</pre>
6648
6649</div>
6650
6651<!-- _______________________________________________________________________ -->
6652<div class="doc_subsubsection">
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]6653 <a name="int_smul_overflow">'<tt>llvm.smul.with.overflow.*</tt>' Intrinsics</a>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6654</div>
6655
6656<div class="doc_text">
6657
6658<h5>Syntax:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6659<p>This is an overloaded intrinsic. You can use <tt>llvm.smul.with.overflow</tt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6660 on any integer bit width.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6661
6662<pre>
6663 declare {i16, i1} @llvm.smul.with.overflow.i16(i16 %a, i16 %b)
6664 declare {i32, i1} @llvm.smul.with.overflow.i32(i32 %a, i32 %b)
6665 declare {i64, i1} @llvm.smul.with.overflow.i64(i64 %a, i64 %b)
6666</pre>
6667
6668<h5>Overview:</h5>
6669
6670<p>The '<tt>llvm.smul.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6671 a signed multiplication of the two arguments, and indicate whether an
6672 overflow occurred during the signed multiplication.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6673
6674<h5>Arguments:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6675<p>The arguments (%a and %b) and the first element of the result structure may
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6676 be of integer types of any bit width, but they must have the same bit
6677 width. The second element of the result structure must be of
6678 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
6679 undergo signed multiplication.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6680
6681<h5>Semantics:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6682<p>The '<tt>llvm.smul.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6683 a signed multiplication of the two arguments. They return a structure &mdash;
6684 the first element of which is the multiplication, and the second element of
6685 which is a bit specifying if the signed multiplication resulted in an
6686 overflow.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6687
6688<h5>Examples:</h5>
6689<pre>
6690 %res = call {i32, i1} @llvm.smul.with.overflow.i32(i32 %a, i32 %b)
6691 %sum = extractvalue {i32, i1} %res, 0
6692 %obit = extractvalue {i32, i1} %res, 1
6693 br i1 %obit, label %overflow, label %normal
6694</pre>
6695
Reid Spencerf86037f2007-04-11 23:23:49 +0000[diff] [blame]6696</div>
6697
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]6698<!-- _______________________________________________________________________ -->
6699<div class="doc_subsubsection">
6700 <a name="int_umul_overflow">'<tt>llvm.umul.with.overflow.*</tt>' Intrinsics</a>
6701</div>
6702
6703<div class="doc_text">
6704
6705<h5>Syntax:</h5>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]6706<p>This is an overloaded intrinsic. You can use <tt>llvm.umul.with.overflow</tt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6707 on any integer bit width.</p>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]6708
6709<pre>
6710 declare {i16, i1} @llvm.umul.with.overflow.i16(i16 %a, i16 %b)
6711 declare {i32, i1} @llvm.umul.with.overflow.i32(i32 %a, i32 %b)
6712 declare {i64, i1} @llvm.umul.with.overflow.i64(i64 %a, i64 %b)
6713</pre>
6714
6715<h5>Overview:</h5>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]6716<p>The '<tt>llvm.umul.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6717 a unsigned multiplication of the two arguments, and indicate whether an
6718 overflow occurred during the unsigned multiplication.</p>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]6719
6720<h5>Arguments:</h5>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]6721<p>The arguments (%a and %b) and the first element of the result structure may
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6722 be of integer types of any bit width, but they must have the same bit
6723 width. The second element of the result structure must be of
6724 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
6725 undergo unsigned multiplication.</p>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]6726
6727<h5>Semantics:</h5>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]6728<p>The '<tt>llvm.umul.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6729 an unsigned multiplication of the two arguments. They return a structure
6730 &mdash; the first element of which is the multiplication, and the second
6731 element of which is a bit specifying if the unsigned multiplication resulted
6732 in an overflow.</p>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]6733
6734<h5>Examples:</h5>
6735<pre>
6736 %res = call {i32, i1} @llvm.umul.with.overflow.i32(i32 %a, i32 %b)
6737 %sum = extractvalue {i32, i1} %res, 0
6738 %obit = extractvalue {i32, i1} %res, 1
6739 br i1 %obit, label %overflow, label %normal
6740</pre>
6741
6742</div>
6743
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]6744<!-- ======================================================================= -->
6745<div class="doc_subsection">
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]6746 <a name="int_fp16">Half Precision Floating Point Intrinsics</a>
6747</div>
6748
6749<div class="doc_text">
6750
Chris Lattner0cec9c82010-03-15 04:12:21 +0000[diff] [blame]6751<p>Half precision floating point is a storage-only format. This means that it is
6752 a dense encoding (in memory) but does not support computation in the
6753 format.</p>
Chris Lattner82c3dc62010-03-14 23:03:31 +0000[diff] [blame]6754
Chris Lattner0cec9c82010-03-15 04:12:21 +0000[diff] [blame]6755<p>This means that code must first load the half-precision floating point
Chris Lattner82c3dc62010-03-14 23:03:31 +0000[diff] [blame]6756 value as an i16, then convert it to float with <a
6757 href="#int_convert_from_fp16"><tt>llvm.convert.from.fp16</tt></a>.
6758 Computation can then be performed on the float value (including extending to
Chris Lattner0cec9c82010-03-15 04:12:21 +0000[diff] [blame]6759 double etc). To store the value back to memory, it is first converted to
6760 float if needed, then converted to i16 with
Chris Lattner82c3dc62010-03-14 23:03:31 +0000[diff] [blame]6761 <a href="#int_convert_to_fp16"><tt>llvm.convert.to.fp16</tt></a>, then
6762 storing as an i16 value.</p>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]6763</div>
6764
6765<!-- _______________________________________________________________________ -->
6766<div class="doc_subsubsection">
Chris Lattner82c3dc62010-03-14 23:03:31 +0000[diff] [blame]6767 <a name="int_convert_to_fp16">'<tt>llvm.convert.to.fp16</tt>' Intrinsic</a>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]6768</div>
6769
6770<div class="doc_text">
6771
6772<h5>Syntax:</h5>
6773<pre>
6774 declare i16 @llvm.convert.to.fp16(f32 %a)
6775</pre>
6776
6777<h5>Overview:</h5>
6778<p>The '<tt>llvm.convert.to.fp16</tt>' intrinsic function performs
6779 a conversion from single precision floating point format to half precision
6780 floating point format.</p>
6781
6782<h5>Arguments:</h5>
6783<p>The intrinsic function contains single argument - the value to be
6784 converted.</p>
6785
6786<h5>Semantics:</h5>
6787<p>The '<tt>llvm.convert.to.fp16</tt>' intrinsic function performs
6788 a conversion from single precision floating point format to half precision
Chris Lattner0cec9c82010-03-15 04:12:21 +0000[diff] [blame]6789 floating point format. The return value is an <tt>i16</tt> which
Chris Lattner82c3dc62010-03-14 23:03:31 +0000[diff] [blame]6790 contains the converted number.</p>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]6791
6792<h5>Examples:</h5>
6793<pre>
6794 %res = call i16 @llvm.convert.to.fp16(f32 %a)
6795 store i16 %res, i16* @x, align 2
6796</pre>
6797
6798</div>
6799
6800<!-- _______________________________________________________________________ -->
6801<div class="doc_subsubsection">
Chris Lattner82c3dc62010-03-14 23:03:31 +0000[diff] [blame]6802 <a name="int_convert_from_fp16">'<tt>llvm.convert.from.fp16</tt>' Intrinsic</a>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]6803</div>
6804
6805<div class="doc_text">
6806
6807<h5>Syntax:</h5>
6808<pre>
6809 declare f32 @llvm.convert.from.fp16(i16 %a)
6810</pre>
6811
6812<h5>Overview:</h5>
6813<p>The '<tt>llvm.convert.from.fp16</tt>' intrinsic function performs
6814 a conversion from half precision floating point format to single precision
6815 floating point format.</p>
6816
6817<h5>Arguments:</h5>
6818<p>The intrinsic function contains single argument - the value to be
6819 converted.</p>
6820
6821<h5>Semantics:</h5>
6822<p>The '<tt>llvm.convert.from.fp16</tt>' intrinsic function performs a
Chris Lattner0cec9c82010-03-15 04:12:21 +0000[diff] [blame]6823 conversion from half single precision floating point format to single
Chris Lattner82c3dc62010-03-14 23:03:31 +0000[diff] [blame]6824 precision floating point format. The input half-float value is represented by
6825 an <tt>i16</tt> value.</p>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]6826
6827<h5>Examples:</h5>
6828<pre>
6829 %a = load i16* @x, align 2
6830 %res = call f32 @llvm.convert.from.fp16(i16 %a)
6831</pre>
6832
6833</div>
6834
6835<!-- ======================================================================= -->
6836<div class="doc_subsection">
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]6837 <a name="int_debugger">Debugger Intrinsics</a>
6838</div>
6839
6840<div class="doc_text">
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]6841
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6842<p>The LLVM debugger intrinsics (which all start with <tt>llvm.dbg.</tt>
6843 prefix), are described in
6844 the <a href="SourceLevelDebugging.html#format_common_intrinsics">LLVM Source
6845 Level Debugging</a> document.</p>
6846
6847</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]6848
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]6849<!-- ======================================================================= -->
6850<div class="doc_subsection">
6851 <a name="int_eh">Exception Handling Intrinsics</a>
6852</div>
6853
6854<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6855
6856<p>The LLVM exception handling intrinsics (which all start with
6857 <tt>llvm.eh.</tt> prefix), are described in
6858 the <a href="ExceptionHandling.html#format_common_intrinsics">LLVM Exception
6859 Handling</a> document.</p>
6860
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]6861</div>
6862
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]6863<!-- ======================================================================= -->
6864<div class="doc_subsection">
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]6865 <a name="int_trampoline">Trampoline Intrinsic</a>
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]6866</div>
6867
6868<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6869
6870<p>This intrinsic makes it possible to excise one parameter, marked with
Dan Gohmanff235352010-07-02 23:18:08 +0000[diff] [blame]6871 the <a href="#nest"><tt>nest</tt></a> attribute, from a function.
6872 The result is a callable
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6873 function pointer lacking the nest parameter - the caller does not need to
6874 provide a value for it. Instead, the value to use is stored in advance in a
6875 "trampoline", a block of memory usually allocated on the stack, which also
6876 contains code to splice the nest value into the argument list. This is used
6877 to implement the GCC nested function address extension.</p>
6878
6879<p>For example, if the function is
6880 <tt>i32 f(i8* nest %c, i32 %x, i32 %y)</tt> then the resulting function
6881 pointer has signature <tt>i32 (i32, i32)*</tt>. It can be created as
6882 follows:</p>
6883
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]6884<pre class="doc_code">
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]6885 %tramp = alloca [10 x i8], align 4 ; size and alignment only correct for X86
6886 %tramp1 = getelementptr [10 x i8]* %tramp, i32 0, i32 0
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]6887 %p = call i8* @llvm.init.trampoline(i8* %tramp1, i8* bitcast (i32 (i8* nest , i32, i32)* @f to i8*), i8* %nval)
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]6888 %fp = bitcast i8* %p to i32 (i32, i32)*
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]6889</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6890
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]6891<p>The call <tt>%val = call i32 %fp(i32 %x, i32 %y)</tt> is then equivalent
6892 to <tt>%val = call i32 %f(i8* %nval, i32 %x, i32 %y)</tt>.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6893
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]6894</div>
6895
6896<!-- _______________________________________________________________________ -->
6897<div class="doc_subsubsection">
6898 <a name="int_it">'<tt>llvm.init.trampoline</tt>' Intrinsic</a>
6899</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6900
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]6901<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6902
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]6903<h5>Syntax:</h5>
6904<pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6905 declare 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]6906</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6907
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]6908<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6909<p>This fills the memory pointed to by <tt>tramp</tt> with code and returns a
6910 function pointer suitable for executing it.</p>
6911
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]6912<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6913<p>The <tt>llvm.init.trampoline</tt> intrinsic takes three arguments, all
6914 pointers. The <tt>tramp</tt> argument must point to a sufficiently large and
6915 sufficiently aligned block of memory; this memory is written to by the
6916 intrinsic. Note that the size and the alignment are target-specific - LLVM
6917 currently provides no portable way of determining them, so a front-end that
6918 generates this intrinsic needs to have some target-specific knowledge.
6919 The <tt>func</tt> argument must hold a function bitcast to
6920 an <tt>i8*</tt>.</p>
6921
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]6922<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6923<p>The block of memory pointed to by <tt>tramp</tt> is filled with target
6924 dependent code, turning it into a function. A pointer to this function is
6925 returned, but needs to be bitcast to an <a href="#int_trampoline">appropriate
6926 function pointer type</a> before being called. The new function's signature
6927 is the same as that of <tt>func</tt> with any arguments marked with
6928 the <tt>nest</tt> attribute removed. At most one such <tt>nest</tt> argument
6929 is allowed, and it must be of pointer type. Calling the new function is
6930 equivalent to calling <tt>func</tt> with the same argument list, but
6931 with <tt>nval</tt> used for the missing <tt>nest</tt> argument. If, after
6932 calling <tt>llvm.init.trampoline</tt>, the memory pointed to
6933 by <tt>tramp</tt> is modified, then the effect of any later call to the
6934 returned function pointer is undefined.</p>
6935
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]6936</div>
6937
6938<!-- ======================================================================= -->
6939<div class="doc_subsection">
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]6940 <a name="int_atomics">Atomic Operations and Synchronization Intrinsics</a>
6941</div>
6942
6943<div class="doc_text">
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]6944
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6945<p>These intrinsic functions expand the "universal IR" of LLVM to represent
6946 hardware constructs for atomic operations and memory synchronization. This
6947 provides an interface to the hardware, not an interface to the programmer. It
6948 is aimed at a low enough level to allow any programming models or APIs
6949 (Application Programming Interfaces) which need atomic behaviors to map
6950 cleanly onto it. It is also modeled primarily on hardware behavior. Just as
6951 hardware provides a "universal IR" for source languages, it also provides a
6952 starting point for developing a "universal" atomic operation and
6953 synchronization IR.</p>
6954
6955<p>These do <em>not</em> form an API such as high-level threading libraries,
6956 software transaction memory systems, atomic primitives, and intrinsic
6957 functions as found in BSD, GNU libc, atomic_ops, APR, and other system and
6958 application libraries. The hardware interface provided by LLVM should allow
6959 a clean implementation of all of these APIs and parallel programming models.
6960 No one model or paradigm should be selected above others unless the hardware
6961 itself ubiquitously does so.</p>
6962
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]6963</div>
6964
6965<!-- _______________________________________________________________________ -->
6966<div class="doc_subsubsection">
6967 <a name="int_memory_barrier">'<tt>llvm.memory.barrier</tt>' Intrinsic</a>
6968</div>
6969<div class="doc_text">
6970<h5>Syntax:</h5>
6971<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]6972 declare void @llvm.memory.barrier(i1 &lt;ll&gt;, i1 &lt;ls&gt;, i1 &lt;sl&gt;, i1 &lt;ss&gt;, i1 &lt;device&gt;)
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]6973</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6974
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]6975<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6976<p>The <tt>llvm.memory.barrier</tt> intrinsic guarantees ordering between
6977 specific pairs of memory access types.</p>
6978
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]6979<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6980<p>The <tt>llvm.memory.barrier</tt> intrinsic requires five boolean arguments.
6981 The first four arguments enables a specific barrier as listed below. The
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]6982 fifth argument specifies that the barrier applies to io or device or uncached
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6983 memory.</p>
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]6984
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6985<ul>
6986 <li><tt>ll</tt>: load-load barrier</li>
6987 <li><tt>ls</tt>: load-store barrier</li>
6988 <li><tt>sl</tt>: store-load barrier</li>
6989 <li><tt>ss</tt>: store-store barrier</li>
6990 <li><tt>device</tt>: barrier applies to device and uncached memory also.</li>
6991</ul>
6992
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]6993<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6994<p>This intrinsic causes the system to enforce some ordering constraints upon
6995 the loads and stores of the program. This barrier does not
6996 indicate <em>when</em> any events will occur, it only enforces
6997 an <em>order</em> in which they occur. For any of the specified pairs of load
6998 and store operations (f.ex. load-load, or store-load), all of the first
6999 operations preceding the barrier will complete before any of the second
7000 operations succeeding the barrier begin. Specifically the semantics for each
7001 pairing is as follows:</p>
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7002
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7003<ul>
7004 <li><tt>ll</tt>: All loads before the barrier must complete before any load
7005 after the barrier begins.</li>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]7006 <li><tt>ls</tt>: All loads before the barrier must complete before any
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7007 store after the barrier begins.</li>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]7008 <li><tt>ss</tt>: All stores before the barrier must complete before any
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7009 store after the barrier begins.</li>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]7010 <li><tt>sl</tt>: All stores before the barrier must complete before any
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7011 load after the barrier begins.</li>
7012</ul>
7013
7014<p>These semantics are applied with a logical "and" behavior when more than one
7015 is enabled in a single memory barrier intrinsic.</p>
7016
7017<p>Backends may implement stronger barriers than those requested when they do
7018 not support as fine grained a barrier as requested. Some architectures do
7019 not need all types of barriers and on such architectures, these become
7020 noops.</p>
7021
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7022<h5>Example:</h5>
7023<pre>
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]7024%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
7025%ptr = bitcast i8* %mallocP to i32*
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7026 store i32 4, %ptr
7027
7028%result1 = load i32* %ptr <i>; yields {i32}:result1 = 4</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7029 call void @llvm.memory.barrier(i1 false, i1 true, i1 false, i1 false)
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7030 <i>; guarantee the above finishes</i>
7031 store i32 8, %ptr <i>; before this begins</i>
7032</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7033
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7034</div>
7035
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7036<!-- _______________________________________________________________________ -->
7037<div class="doc_subsubsection">
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7038 <a name="int_atomic_cmp_swap">'<tt>llvm.atomic.cmp.swap.*</tt>' Intrinsic</a>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7039</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7040
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7041<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7042
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7043<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7044<p>This is an overloaded intrinsic. You can use <tt>llvm.atomic.cmp.swap</tt> on
7045 any integer bit width and for different address spaces. Not all targets
7046 support all bit widths however.</p>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7047
7048<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7049 declare i8 @llvm.atomic.cmp.swap.i8.p0i8(i8* &lt;ptr&gt;, i8 &lt;cmp&gt;, i8 &lt;val&gt;)
7050 declare i16 @llvm.atomic.cmp.swap.i16.p0i16(i16* &lt;ptr&gt;, i16 &lt;cmp&gt;, i16 &lt;val&gt;)
7051 declare i32 @llvm.atomic.cmp.swap.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;cmp&gt;, i32 &lt;val&gt;)
7052 declare i64 @llvm.atomic.cmp.swap.i64.p0i64(i64* &lt;ptr&gt;, i64 &lt;cmp&gt;, i64 &lt;val&gt;)
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7053</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7054
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7055<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7056<p>This loads a value in memory and compares it to a given value. If they are
7057 equal, it stores a new value into the memory.</p>
7058
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7059<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7060<p>The <tt>llvm.atomic.cmp.swap</tt> intrinsic takes three arguments. The result
7061 as well as both <tt>cmp</tt> and <tt>val</tt> must be integer values with the
7062 same bit width. The <tt>ptr</tt> argument must be a pointer to a value of
7063 this integer type. While any bit width integer may be used, targets may only
7064 lower representations they support in hardware.</p>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7065
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7066<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7067<p>This entire intrinsic must be executed atomically. It first loads the value
7068 in memory pointed to by <tt>ptr</tt> and compares it with the
7069 value <tt>cmp</tt>. If they are equal, <tt>val</tt> is stored into the
7070 memory. The loaded value is yielded in all cases. This provides the
7071 equivalent of an atomic compare-and-swap operation within the SSA
7072 framework.</p>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7073
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7074<h5>Examples:</h5>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7075<pre>
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]7076%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
7077%ptr = bitcast i8* %mallocP to i32*
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7078 store i32 4, %ptr
7079
7080%val1 = add i32 4, 4
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7081%result1 = call i32 @llvm.atomic.cmp.swap.i32.p0i32(i32* %ptr, i32 4, %val1)
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7082 <i>; yields {i32}:result1 = 4</i>
7083%stored1 = icmp eq i32 %result1, 4 <i>; yields {i1}:stored1 = true</i>
7084%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 8</i>
7085
7086%val2 = add i32 1, 1
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7087%result2 = call i32 @llvm.atomic.cmp.swap.i32.p0i32(i32* %ptr, i32 5, %val2)
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7088 <i>; yields {i32}:result2 = 8</i>
7089%stored2 = icmp eq i32 %result2, 5 <i>; yields {i1}:stored2 = false</i>
7090
7091%memval2 = load i32* %ptr <i>; yields {i32}:memval2 = 8</i>
7092</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7093
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7094</div>
7095
7096<!-- _______________________________________________________________________ -->
7097<div class="doc_subsubsection">
7098 <a name="int_atomic_swap">'<tt>llvm.atomic.swap.*</tt>' Intrinsic</a>
7099</div>
7100<div class="doc_text">
7101<h5>Syntax:</h5>
7102
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7103<p>This is an overloaded intrinsic. You can use <tt>llvm.atomic.swap</tt> on any
7104 integer bit width. Not all targets support all bit widths however.</p>
7105
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7106<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7107 declare i8 @llvm.atomic.swap.i8.p0i8(i8* &lt;ptr&gt;, i8 &lt;val&gt;)
7108 declare i16 @llvm.atomic.swap.i16.p0i16(i16* &lt;ptr&gt;, i16 &lt;val&gt;)
7109 declare i32 @llvm.atomic.swap.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;val&gt;)
7110 declare i64 @llvm.atomic.swap.i64.p0i64(i64* &lt;ptr&gt;, i64 &lt;val&gt;)
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7111</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7112
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7113<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7114<p>This intrinsic loads the value stored in memory at <tt>ptr</tt> and yields
7115 the value from memory. It then stores the value in <tt>val</tt> in the memory
7116 at <tt>ptr</tt>.</p>
7117
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7118<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7119<p>The <tt>llvm.atomic.swap</tt> intrinsic takes two arguments. Both
7120 the <tt>val</tt> argument and the result must be integers of the same bit
7121 width. The first argument, <tt>ptr</tt>, must be a pointer to a value of this
7122 integer type. The targets may only lower integer representations they
7123 support.</p>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7124
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7125<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7126<p>This intrinsic loads the value pointed to by <tt>ptr</tt>, yields it, and
7127 stores <tt>val</tt> back into <tt>ptr</tt> atomically. This provides the
7128 equivalent of an atomic swap operation within the SSA framework.</p>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7129
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7130<h5>Examples:</h5>
7131<pre>
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]7132%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
7133%ptr = bitcast i8* %mallocP to i32*
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7134 store i32 4, %ptr
7135
7136%val1 = add i32 4, 4
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7137%result1 = call i32 @llvm.atomic.swap.i32.p0i32(i32* %ptr, i32 %val1)
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7138 <i>; yields {i32}:result1 = 4</i>
7139%stored1 = icmp eq i32 %result1, 4 <i>; yields {i1}:stored1 = true</i>
7140%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 8</i>
7141
7142%val2 = add i32 1, 1
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7143%result2 = call i32 @llvm.atomic.swap.i32.p0i32(i32* %ptr, i32 %val2)
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7144 <i>; yields {i32}:result2 = 8</i>
7145
7146%stored2 = icmp eq i32 %result2, 8 <i>; yields {i1}:stored2 = true</i>
7147%memval2 = load i32* %ptr <i>; yields {i32}:memval2 = 2</i>
7148</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7149
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7150</div>
7151
7152<!-- _______________________________________________________________________ -->
7153<div class="doc_subsubsection">
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7154 <a name="int_atomic_load_add">'<tt>llvm.atomic.load.add.*</tt>' Intrinsic</a>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7155
7156</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7157
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7158<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7159
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7160<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7161<p>This is an overloaded intrinsic. You can use <tt>llvm.atomic.load.add</tt> on
7162 any integer bit width. Not all targets support all bit widths however.</p>
7163
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7164<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7165 declare i8 @llvm.atomic.load.add.i8.p0i8(i8* &lt;ptr&gt;, i8 &lt;delta&gt;)
7166 declare i16 @llvm.atomic.load.add.i16.p0i16(i16* &lt;ptr&gt;, i16 &lt;delta&gt;)
7167 declare i32 @llvm.atomic.load.add.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;delta&gt;)
7168 declare i64 @llvm.atomic.load.add.i64.p0i64(i64* &lt;ptr&gt;, i64 &lt;delta&gt;)
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7169</pre>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7170
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7171<h5>Overview:</h5>
7172<p>This intrinsic adds <tt>delta</tt> to the value stored in memory
7173 at <tt>ptr</tt>. It yields the original value at <tt>ptr</tt>.</p>
7174
7175<h5>Arguments:</h5>
7176<p>The intrinsic takes two arguments, the first a pointer to an integer value
7177 and the second an integer value. The result is also an integer value. These
7178 integer types can have any bit width, but they must all have the same bit
7179 width. The targets may only lower integer representations they support.</p>
7180
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7181<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7182<p>This intrinsic does a series of operations atomically. It first loads the
7183 value stored at <tt>ptr</tt>. It then adds <tt>delta</tt>, stores the result
7184 to <tt>ptr</tt>. It yields the original value stored at <tt>ptr</tt>.</p>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7185
7186<h5>Examples:</h5>
7187<pre>
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]7188%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
7189%ptr = bitcast i8* %mallocP to i32*
7190 store i32 4, %ptr
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7191%result1 = call i32 @llvm.atomic.load.add.i32.p0i32(i32* %ptr, i32 4)
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7192 <i>; yields {i32}:result1 = 4</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7193%result2 = call i32 @llvm.atomic.load.add.i32.p0i32(i32* %ptr, i32 2)
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7194 <i>; yields {i32}:result2 = 8</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7195%result3 = call i32 @llvm.atomic.load.add.i32.p0i32(i32* %ptr, i32 5)
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7196 <i>; yields {i32}:result3 = 10</i>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7197%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 15</i>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7198</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7199
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7200</div>
7201
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7202<!-- _______________________________________________________________________ -->
7203<div class="doc_subsubsection">
7204 <a name="int_atomic_load_sub">'<tt>llvm.atomic.load.sub.*</tt>' Intrinsic</a>
7205
7206</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7207
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7208<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7209
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7210<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7211<p>This is an overloaded intrinsic. You can use <tt>llvm.atomic.load.sub</tt> on
7212 any integer bit width and for different address spaces. Not all targets
7213 support all bit widths however.</p>
7214
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7215<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7216 declare i8 @llvm.atomic.load.sub.i8.p0i32(i8* &lt;ptr&gt;, i8 &lt;delta&gt;)
7217 declare i16 @llvm.atomic.load.sub.i16.p0i32(i16* &lt;ptr&gt;, i16 &lt;delta&gt;)
7218 declare i32 @llvm.atomic.load.sub.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;delta&gt;)
7219 declare i64 @llvm.atomic.load.sub.i64.p0i32(i64* &lt;ptr&gt;, i64 &lt;delta&gt;)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7220</pre>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7221
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7222<h5>Overview:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]7223<p>This intrinsic subtracts <tt>delta</tt> to the value stored in memory at
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7224 <tt>ptr</tt>. It yields the original value at <tt>ptr</tt>.</p>
7225
7226<h5>Arguments:</h5>
7227<p>The intrinsic takes two arguments, the first a pointer to an integer value
7228 and the second an integer value. The result is also an integer value. These
7229 integer types can have any bit width, but they must all have the same bit
7230 width. The targets may only lower integer representations they support.</p>
7231
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7232<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7233<p>This intrinsic does a series of operations atomically. It first loads the
7234 value stored at <tt>ptr</tt>. It then subtracts <tt>delta</tt>, stores the
7235 result to <tt>ptr</tt>. It yields the original value stored
7236 at <tt>ptr</tt>.</p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7237
7238<h5>Examples:</h5>
7239<pre>
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]7240%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
7241%ptr = bitcast i8* %mallocP to i32*
7242 store i32 8, %ptr
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7243%result1 = call i32 @llvm.atomic.load.sub.i32.p0i32(i32* %ptr, i32 4)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7244 <i>; yields {i32}:result1 = 8</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7245%result2 = call i32 @llvm.atomic.load.sub.i32.p0i32(i32* %ptr, i32 2)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7246 <i>; yields {i32}:result2 = 4</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7247%result3 = call i32 @llvm.atomic.load.sub.i32.p0i32(i32* %ptr, i32 5)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7248 <i>; yields {i32}:result3 = 2</i>
7249%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = -3</i>
7250</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7251
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7252</div>
7253
7254<!-- _______________________________________________________________________ -->
7255<div class="doc_subsubsection">
7256 <a name="int_atomic_load_and">'<tt>llvm.atomic.load.and.*</tt>' Intrinsic</a><br>
7257 <a name="int_atomic_load_nand">'<tt>llvm.atomic.load.nand.*</tt>' Intrinsic</a><br>
7258 <a name="int_atomic_load_or">'<tt>llvm.atomic.load.or.*</tt>' Intrinsic</a><br>
7259 <a name="int_atomic_load_xor">'<tt>llvm.atomic.load.xor.*</tt>' Intrinsic</a><br>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7260</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7261
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7262<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7263
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7264<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7265<p>These are overloaded intrinsics. You can
7266 use <tt>llvm.atomic.load_and</tt>, <tt>llvm.atomic.load_nand</tt>,
7267 <tt>llvm.atomic.load_or</tt>, and <tt>llvm.atomic.load_xor</tt> on any integer
7268 bit width and for different address spaces. Not all targets support all bit
7269 widths however.</p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7270
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7271<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7272 declare i8 @llvm.atomic.load.and.i8.p0i8(i8* &lt;ptr&gt;, i8 &lt;delta&gt;)
7273 declare i16 @llvm.atomic.load.and.i16.p0i16(i16* &lt;ptr&gt;, i16 &lt;delta&gt;)
7274 declare i32 @llvm.atomic.load.and.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;delta&gt;)
7275 declare i64 @llvm.atomic.load.and.i64.p0i64(i64* &lt;ptr&gt;, i64 &lt;delta&gt;)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7276</pre>
7277
7278<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7279 declare i8 @llvm.atomic.load.or.i8.p0i8(i8* &lt;ptr&gt;, i8 &lt;delta&gt;)
7280 declare i16 @llvm.atomic.load.or.i16.p0i16(i16* &lt;ptr&gt;, i16 &lt;delta&gt;)
7281 declare i32 @llvm.atomic.load.or.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;delta&gt;)
7282 declare i64 @llvm.atomic.load.or.i64.p0i64(i64* &lt;ptr&gt;, i64 &lt;delta&gt;)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7283</pre>
7284
7285<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7286 declare i8 @llvm.atomic.load.nand.i8.p0i32(i8* &lt;ptr&gt;, i8 &lt;delta&gt;)
7287 declare i16 @llvm.atomic.load.nand.i16.p0i32(i16* &lt;ptr&gt;, i16 &lt;delta&gt;)
7288 declare i32 @llvm.atomic.load.nand.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;delta&gt;)
7289 declare i64 @llvm.atomic.load.nand.i64.p0i32(i64* &lt;ptr&gt;, i64 &lt;delta&gt;)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7290</pre>
7291
7292<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7293 declare i8 @llvm.atomic.load.xor.i8.p0i32(i8* &lt;ptr&gt;, i8 &lt;delta&gt;)
7294 declare i16 @llvm.atomic.load.xor.i16.p0i32(i16* &lt;ptr&gt;, i16 &lt;delta&gt;)
7295 declare i32 @llvm.atomic.load.xor.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;delta&gt;)
7296 declare i64 @llvm.atomic.load.xor.i64.p0i32(i64* &lt;ptr&gt;, i64 &lt;delta&gt;)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7297</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7298
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7299<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7300<p>These intrinsics bitwise the operation (and, nand, or, xor) <tt>delta</tt> to
7301 the value stored in memory at <tt>ptr</tt>. It yields the original value
7302 at <tt>ptr</tt>.</p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7303
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7304<h5>Arguments:</h5>
7305<p>These intrinsics take two arguments, the first a pointer to an integer value
7306 and the second an integer value. The result is also an integer value. These
7307 integer types can have any bit width, but they must all have the same bit
7308 width. The targets may only lower integer representations they support.</p>
7309
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7310<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7311<p>These intrinsics does a series of operations atomically. They first load the
7312 value stored at <tt>ptr</tt>. They then do the bitwise
7313 operation <tt>delta</tt>, store the result to <tt>ptr</tt>. They yield the
7314 original value stored at <tt>ptr</tt>.</p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7315
7316<h5>Examples:</h5>
7317<pre>
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]7318%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
7319%ptr = bitcast i8* %mallocP to i32*
7320 store i32 0x0F0F, %ptr
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7321%result0 = call i32 @llvm.atomic.load.nand.i32.p0i32(i32* %ptr, i32 0xFF)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7322 <i>; yields {i32}:result0 = 0x0F0F</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7323%result1 = call i32 @llvm.atomic.load.and.i32.p0i32(i32* %ptr, i32 0xFF)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7324 <i>; yields {i32}:result1 = 0xFFFFFFF0</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7325%result2 = call i32 @llvm.atomic.load.or.i32.p0i32(i32* %ptr, i32 0F)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7326 <i>; yields {i32}:result2 = 0xF0</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7327%result3 = call i32 @llvm.atomic.load.xor.i32.p0i32(i32* %ptr, i32 0F)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7328 <i>; yields {i32}:result3 = FF</i>
7329%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = F0</i>
7330</pre>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7331
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7332</div>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7333
7334<!-- _______________________________________________________________________ -->
7335<div class="doc_subsubsection">
7336 <a name="int_atomic_load_max">'<tt>llvm.atomic.load.max.*</tt>' Intrinsic</a><br>
7337 <a name="int_atomic_load_min">'<tt>llvm.atomic.load.min.*</tt>' Intrinsic</a><br>
7338 <a name="int_atomic_load_umax">'<tt>llvm.atomic.load.umax.*</tt>' Intrinsic</a><br>
7339 <a name="int_atomic_load_umin">'<tt>llvm.atomic.load.umin.*</tt>' Intrinsic</a><br>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7340</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7341
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7342<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7343
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7344<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7345<p>These are overloaded intrinsics. You can use <tt>llvm.atomic.load_max</tt>,
7346 <tt>llvm.atomic.load_min</tt>, <tt>llvm.atomic.load_umax</tt>, and
7347 <tt>llvm.atomic.load_umin</tt> on any integer bit width and for different
7348 address spaces. Not all targets support all bit widths however.</p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7349
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7350<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7351 declare i8 @llvm.atomic.load.max.i8.p0i8(i8* &lt;ptr&gt;, i8 &lt;delta&gt;)
7352 declare i16 @llvm.atomic.load.max.i16.p0i16(i16* &lt;ptr&gt;, i16 &lt;delta&gt;)
7353 declare i32 @llvm.atomic.load.max.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;delta&gt;)
7354 declare i64 @llvm.atomic.load.max.i64.p0i64(i64* &lt;ptr&gt;, i64 &lt;delta&gt;)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7355</pre>
7356
7357<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7358 declare i8 @llvm.atomic.load.min.i8.p0i8(i8* &lt;ptr&gt;, i8 &lt;delta&gt;)
7359 declare i16 @llvm.atomic.load.min.i16.p0i16(i16* &lt;ptr&gt;, i16 &lt;delta&gt;)
7360 declare i32 @llvm.atomic.load.min.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;delta&gt;)
7361 declare i64 @llvm.atomic.load.min.i64.p0i64(i64* &lt;ptr&gt;, i64 &lt;delta&gt;)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7362</pre>
7363
7364<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7365 declare i8 @llvm.atomic.load.umax.i8.p0i8(i8* &lt;ptr&gt;, i8 &lt;delta&gt;)
7366 declare i16 @llvm.atomic.load.umax.i16.p0i16(i16* &lt;ptr&gt;, i16 &lt;delta&gt;)
7367 declare i32 @llvm.atomic.load.umax.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;delta&gt;)
7368 declare i64 @llvm.atomic.load.umax.i64.p0i64(i64* &lt;ptr&gt;, i64 &lt;delta&gt;)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7369</pre>
7370
7371<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7372 declare i8 @llvm.atomic.load.umin.i8.p0i8(i8* &lt;ptr&gt;, i8 &lt;delta&gt;)
7373 declare i16 @llvm.atomic.load.umin.i16.p0i16(i16* &lt;ptr&gt;, i16 &lt;delta&gt;)
7374 declare i32 @llvm.atomic.load.umin.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;delta&gt;)
7375 declare i64 @llvm.atomic.load.umin.i64.p0i64(i64* &lt;ptr&gt;, i64 &lt;delta&gt;)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7376</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7377
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7378<h5>Overview:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]7379<p>These intrinsics takes the signed or unsigned minimum or maximum of
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7380 <tt>delta</tt> and the value stored in memory at <tt>ptr</tt>. It yields the
7381 original value at <tt>ptr</tt>.</p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7382
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7383<h5>Arguments:</h5>
7384<p>These intrinsics take two arguments, the first a pointer to an integer value
7385 and the second an integer value. The result is also an integer value. These
7386 integer types can have any bit width, but they must all have the same bit
7387 width. The targets may only lower integer representations they support.</p>
7388
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7389<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7390<p>These intrinsics does a series of operations atomically. They first load the
7391 value stored at <tt>ptr</tt>. They then do the signed or unsigned min or
7392 max <tt>delta</tt> and the value, store the result to <tt>ptr</tt>. They
7393 yield the original value stored at <tt>ptr</tt>.</p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7394
7395<h5>Examples:</h5>
7396<pre>
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]7397%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
7398%ptr = bitcast i8* %mallocP to i32*
7399 store i32 7, %ptr
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7400%result0 = call i32 @llvm.atomic.load.min.i32.p0i32(i32* %ptr, i32 -2)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7401 <i>; yields {i32}:result0 = 7</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7402%result1 = call i32 @llvm.atomic.load.max.i32.p0i32(i32* %ptr, i32 8)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7403 <i>; yields {i32}:result1 = -2</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7404%result2 = call i32 @llvm.atomic.load.umin.i32.p0i32(i32* %ptr, i32 10)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7405 <i>; yields {i32}:result2 = 8</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7406%result3 = call i32 @llvm.atomic.load.umax.i32.p0i32(i32* %ptr, i32 30)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7407 <i>; yields {i32}:result3 = 8</i>
7408%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 30</i>
7409</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7410
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7411</div>
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7412
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7413
7414<!-- ======================================================================= -->
7415<div class="doc_subsection">
7416 <a name="int_memorymarkers">Memory Use Markers</a>
7417</div>
7418
7419<div class="doc_text">
7420
7421<p>This class of intrinsics exists to information about the lifetime of memory
7422 objects and ranges where variables are immutable.</p>
7423
7424</div>
7425
7426<!-- _______________________________________________________________________ -->
7427<div class="doc_subsubsection">
7428 <a name="int_lifetime_start">'<tt>llvm.lifetime.start</tt>' Intrinsic</a>
7429</div>
7430
7431<div class="doc_text">
7432
7433<h5>Syntax:</h5>
7434<pre>
7435 declare void @llvm.lifetime.start(i64 &lt;size&gt;, i8* nocapture &lt;ptr&gt;)
7436</pre>
7437
7438<h5>Overview:</h5>
7439<p>The '<tt>llvm.lifetime.start</tt>' intrinsic specifies the start of a memory
7440 object's lifetime.</p>
7441
7442<h5>Arguments:</h5>
Nick Lewycky321333e2009-10-13 07:57:33 +0000[diff] [blame]7443<p>The first argument is a constant integer representing the size of the
7444 object, or -1 if it is variable sized. The second argument is a pointer to
7445 the object.</p>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7446
7447<h5>Semantics:</h5>
7448<p>This intrinsic indicates that before this point in the code, the value of the
7449 memory pointed to by <tt>ptr</tt> is dead. This means that it is known to
Nick Lewycky8d336592009-10-27 16:56:58 +0000[diff] [blame]7450 never be used and has an undefined value. A load from the pointer that
7451 precedes this intrinsic can be replaced with
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7452 <tt>'<a href="#undefvalues">undef</a>'</tt>.</p>
7453
7454</div>
7455
7456<!-- _______________________________________________________________________ -->
7457<div class="doc_subsubsection">
7458 <a name="int_lifetime_end">'<tt>llvm.lifetime.end</tt>' Intrinsic</a>
7459</div>
7460
7461<div class="doc_text">
7462
7463<h5>Syntax:</h5>
7464<pre>
7465 declare void @llvm.lifetime.end(i64 &lt;size&gt;, i8* nocapture &lt;ptr&gt;)
7466</pre>
7467
7468<h5>Overview:</h5>
7469<p>The '<tt>llvm.lifetime.end</tt>' intrinsic specifies the end of a memory
7470 object's lifetime.</p>
7471
7472<h5>Arguments:</h5>
Nick Lewycky321333e2009-10-13 07:57:33 +0000[diff] [blame]7473<p>The first argument is a constant integer representing the size of the
7474 object, or -1 if it is variable sized. The second argument is a pointer to
7475 the object.</p>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7476
7477<h5>Semantics:</h5>
7478<p>This intrinsic indicates that after this point in the code, the value of the
7479 memory pointed to by <tt>ptr</tt> is dead. This means that it is known to
7480 never be used and has an undefined value. Any stores into the memory object
7481 following this intrinsic may be removed as dead.
7482
7483</div>
7484
7485<!-- _______________________________________________________________________ -->
7486<div class="doc_subsubsection">
7487 <a name="int_invariant_start">'<tt>llvm.invariant.start</tt>' Intrinsic</a>
7488</div>
7489
7490<div class="doc_text">
7491
7492<h5>Syntax:</h5>
7493<pre>
7494 declare {}* @llvm.invariant.start(i64 &lt;size&gt;, i8* nocapture &lt;ptr&gt;) readonly
7495</pre>
7496
7497<h5>Overview:</h5>
7498<p>The '<tt>llvm.invariant.start</tt>' intrinsic specifies that the contents of
7499 a memory object will not change.</p>
7500
7501<h5>Arguments:</h5>
Nick Lewycky321333e2009-10-13 07:57:33 +0000[diff] [blame]7502<p>The first argument is a constant integer representing the size of the
7503 object, or -1 if it is variable sized. The second argument is a pointer to
7504 the object.</p>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7505
7506<h5>Semantics:</h5>
7507<p>This intrinsic indicates that until an <tt>llvm.invariant.end</tt> that uses
7508 the return value, the referenced memory location is constant and
7509 unchanging.</p>
7510
7511</div>
7512
7513<!-- _______________________________________________________________________ -->
7514<div class="doc_subsubsection">
7515 <a name="int_invariant_end">'<tt>llvm.invariant.end</tt>' Intrinsic</a>
7516</div>
7517
7518<div class="doc_text">
7519
7520<h5>Syntax:</h5>
7521<pre>
7522 declare void @llvm.invariant.end({}* &lt;start&gt;, i64 &lt;size&gt;, i8* nocapture &lt;ptr&gt;)
7523</pre>
7524
7525<h5>Overview:</h5>
7526<p>The '<tt>llvm.invariant.end</tt>' intrinsic specifies that the contents of
7527 a memory object are mutable.</p>
7528
7529<h5>Arguments:</h5>
7530<p>The first argument is the matching <tt>llvm.invariant.start</tt> intrinsic.
Nick Lewycky321333e2009-10-13 07:57:33 +0000[diff] [blame]7531 The second argument is a constant integer representing the size of the
7532 object, or -1 if it is variable sized and the third argument is a pointer
7533 to the object.</p>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7534
7535<h5>Semantics:</h5>
7536<p>This intrinsic indicates that the memory is mutable again.</p>
7537
7538</div>
7539
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7540<!-- ======================================================================= -->
7541<div class="doc_subsection">
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]7542 <a name="int_general">General Intrinsics</a>
7543</div>
7544
7545<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7546
7547<p>This class of intrinsics is designed to be generic and has no specific
7548 purpose.</p>
7549
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]7550</div>
7551
7552<!-- _______________________________________________________________________ -->
7553<div class="doc_subsubsection">
7554 <a name="int_var_annotation">'<tt>llvm.var.annotation</tt>' Intrinsic</a>
7555</div>
7556
7557<div class="doc_text">
7558
7559<h5>Syntax:</h5>
7560<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7561 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]7562</pre>
7563
7564<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7565<p>The '<tt>llvm.var.annotation</tt>' intrinsic.</p>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]7566
7567<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7568<p>The first argument is a pointer to a value, the second is a pointer to a
7569 global string, the third is a pointer to a global string which is the source
7570 file name, and the last argument is the line number.</p>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]7571
7572<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7573<p>This intrinsic allows annotation of local variables with arbitrary strings.
7574 This can be useful for special purpose optimizations that want to look for
7575 these annotations. These have no other defined use, they are ignored by code
7576 generation and optimization.</p>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]7577
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]7578</div>
7579
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]7580<!-- _______________________________________________________________________ -->
7581<div class="doc_subsubsection">
Tanya Lattnere1a8da02007-09-21 23:57:59 +0000[diff] [blame]7582 <a name="int_annotation">'<tt>llvm.annotation.*</tt>' Intrinsic</a>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]7583</div>
7584
7585<div class="doc_text">
7586
7587<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7588<p>This is an overloaded intrinsic. You can use '<tt>llvm.annotation</tt>' on
7589 any integer bit width.</p>
7590
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]7591<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7592 declare i8 @llvm.annotation.i8(i8 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt;)
7593 declare i16 @llvm.annotation.i16(i16 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt;)
7594 declare i32 @llvm.annotation.i32(i32 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt;)
7595 declare i64 @llvm.annotation.i64(i64 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt;)
7596 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]7597</pre>
7598
7599<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7600<p>The '<tt>llvm.annotation</tt>' intrinsic.</p>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]7601
7602<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7603<p>The first argument is an integer value (result of some expression), the
7604 second is a pointer to a global string, the third is a pointer to a global
7605 string which is the source file name, and the last argument is the line
7606 number. It returns the value of the first argument.</p>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]7607
7608<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7609<p>This intrinsic allows annotations to be put on arbitrary expressions with
7610 arbitrary strings. This can be useful for special purpose optimizations that
7611 want to look for these annotations. These have no other defined use, they
7612 are ignored by code generation and optimization.</p>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]7613
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]7614</div>
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]7615
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]7616<!-- _______________________________________________________________________ -->
7617<div class="doc_subsubsection">
7618 <a name="int_trap">'<tt>llvm.trap</tt>' Intrinsic</a>
7619</div>
7620
7621<div class="doc_text">
7622
7623<h5>Syntax:</h5>
7624<pre>
7625 declare void @llvm.trap()
7626</pre>
7627
7628<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7629<p>The '<tt>llvm.trap</tt>' intrinsic.</p>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]7630
7631<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7632<p>None.</p>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]7633
7634<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7635<p>This intrinsics is lowered to the target dependent trap instruction. If the
7636 target does not have a trap instruction, this intrinsic will be lowered to
7637 the call of the <tt>abort()</tt> function.</p>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]7638
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]7639</div>
7640
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]7641<!-- _______________________________________________________________________ -->
7642<div class="doc_subsubsection">
Misha Brukmandccb0252008-11-22 23:55:29 +0000[diff] [blame]7643 <a name="int_stackprotector">'<tt>llvm.stackprotector</tt>' Intrinsic</a>
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]7644</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7645
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]7646<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7647
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]7648<h5>Syntax:</h5>
7649<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7650 declare void @llvm.stackprotector(i8* &lt;guard&gt;, i8** &lt;slot&gt;)
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]7651</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7652
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]7653<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7654<p>The <tt>llvm.stackprotector</tt> intrinsic takes the <tt>guard</tt> and
7655 stores it onto the stack at <tt>slot</tt>. The stack slot is adjusted to
7656 ensure that it is placed on the stack before local variables.</p>
7657
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]7658<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7659<p>The <tt>llvm.stackprotector</tt> intrinsic requires two pointer
7660 arguments. The first argument is the value loaded from the stack
7661 guard <tt>@__stack_chk_guard</tt>. The second variable is an <tt>alloca</tt>
7662 that has enough space to hold the value of the guard.</p>
7663
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]7664<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7665<p>This intrinsic causes the prologue/epilogue inserter to force the position of
7666 the <tt>AllocaInst</tt> stack slot to be before local variables on the
7667 stack. This is to ensure that if a local variable on the stack is
7668 overwritten, it will destroy the value of the guard. When the function exits,
7669 the guard on the stack is checked against the original guard. If they're
7670 different, then the program aborts by calling the <tt>__stack_chk_fail()</tt>
7671 function.</p>
7672
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]7673</div>
7674
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]7675<!-- _______________________________________________________________________ -->
7676<div class="doc_subsubsection">
7677 <a name="int_objectsize">'<tt>llvm.objectsize</tt>' Intrinsic</a>
7678</div>
7679
7680<div class="doc_text">
7681
7682<h5>Syntax:</h5>
7683<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7684 declare i32 @llvm.objectsize.i32(i8* &lt;object&gt;, i1 &lt;type&gt;)
7685 declare i64 @llvm.objectsize.i64(i8* &lt;object&gt;, i1 &lt;type&gt;)
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]7686</pre>
7687
7688<h5>Overview:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]7689<p>The <tt>llvm.objectsize</tt> intrinsic is designed to provide information
Eric Christopherd003c5b2010-01-08 21:42:39 +0000[diff] [blame]7690 to the optimizers to discover at compile time either a) when an
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]7691 operation like memcpy will either overflow a buffer that corresponds to
7692 an object, or b) to determine that a runtime check for overflow isn't
7693 necessary. An object in this context means an allocation of a
Eric Christopher8295a0a2009-12-23 00:29:49 +0000[diff] [blame]7694 specific class, structure, array, or other object.</p>
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]7695
7696<h5>Arguments:</h5>
7697<p>The <tt>llvm.objectsize</tt> intrinsic takes two arguments. The first
Eric Christopher8295a0a2009-12-23 00:29:49 +0000[diff] [blame]7698 argument is a pointer to or into the <tt>object</tt>. The second argument
7699 is a boolean 0 or 1. This argument determines whether you want the
7700 maximum (0) or minimum (1) bytes remaining. This needs to be a literal 0 or
7701 1, variables are not allowed.</p>
7702
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]7703<h5>Semantics:</h5>
7704<p>The <tt>llvm.objectsize</tt> intrinsic is lowered to either a constant
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]7705 representing the size of the object concerned or <tt>i32/i64 -1 or 0</tt>
7706 (depending on the <tt>type</tt> argument if the size cannot be determined
7707 at compile time.</p>
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]7708
7709</div>
7710
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]7711<!-- *********************************************************************** -->
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]7712<hr>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]7713<address>
7714 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
Misha Brukman44408702008-12-11 17:34:48 +0000[diff] [blame]7715 src="http://jigsaw.w3.org/css-validator/images/vcss-blue" alt="Valid CSS"></a>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]7716 <a href="http://validator.w3.org/check/referer"><img
Misha Brukman44408702008-12-11 17:34:48 +0000[diff] [blame]7717 src="http://www.w3.org/Icons/valid-html401-blue" alt="Valid HTML 4.01"></a>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]7718
7719 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
Reid Spencer05fe4b02006-03-14 05:39:39 +0000[diff] [blame]7720 <a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]7721 Last modified: $Date$
7722</address>
Chris Lattnerc7d3ab32008-01-04 04:33:49 +0000[diff] [blame]7723
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]7724</body>
7725</html>