Gitiles
Code Review Sign In
gerrit-public.fairphone.software / fp2-dev / platform / external / llvm / dc444e9a236c9e5a9a3a5e6da75f29f7000208d3 / . / docs / LangRef.html
blob: eec06fcc023f886e6f10b762758ab5ec11a2b9ba [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>
27 <li><a href="#linkage_internal">'<tt>internal</tt>' Linkage</a></li>
28 <li><a href="#linkage_available_externally">'<tt>available_externally</tt>' Linkage</a></li>
29 <li><a href="#linkage_linkonce">'<tt>linkonce</tt>' Linkage</a></li>
30 <li><a href="#linkage_common">'<tt>common</tt>' Linkage</a></li>
31 <li><a href="#linkage_weak">'<tt>weak</tt>' Linkage</a></li>
32 <li><a href="#linkage_appending">'<tt>appending</tt>' Linkage</a></li>
33 <li><a href="#linkage_externweak">'<tt>extern_weak</tt>' Linkage</a></li>
Chris Lattner5a2d8752009-10-10 18:26:06 +0000[diff] [blame]34 <li><a href="#linkage_linkonce_odr">'<tt>linkonce_odr</tt>' Linkage</a></li>
Bill Wendling987e7eb2009-07-20 02:41:50 +0000[diff] [blame]35 <li><a href="#linkage_weak">'<tt>weak_odr</tt>' Linkage</a></li>
36 <li><a href="#linkage_external">'<tt>externally visible</tt>' Linkage</a></li>
37 <li><a href="#linkage_dllimport">'<tt>dllimport</tt>' Linkage</a></li>
38 <li><a href="#linkage_dllexport">'<tt>dllexport</tt>' Linkage</a></li>
Bill Wendling3d10a5a2009-07-20 01:03:30 +0000[diff] [blame]39 </ol>
40 </li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]41 <li><a href="#callingconv">Calling Conventions</a></li>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]42 <li><a href="#namedtypes">Named Types</a></li>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]43 <li><a href="#globalvars">Global Variables</a></li>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]44 <li><a href="#functionstructure">Functions</a></li>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]45 <li><a href="#aliasstructure">Aliases</a></li>
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]46 <li><a href="#namedmetadatastructure">Named Metadata</a></li>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]47 <li><a href="#paramattrs">Parameter Attributes</a></li>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]48 <li><a href="#fnattrs">Function Attributes</a></li>
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame]49 <li><a href="#gc">Garbage Collector Names</a></li>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]50 <li><a href="#moduleasm">Module-Level Inline Assembly</a></li>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]51 <li><a href="#datalayout">Data Layout</a></li>
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]52 <li><a href="#pointeraliasing">Pointer Aliasing Rules</a></li>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]53 </ol>
54 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]55 <li><a href="#typesystem">Type System</a>
56 <ol>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]57 <li><a href="#t_classifications">Type Classifications</a></li>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]58 <li><a href="#t_primitive">Primitive Types</a>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]59 <ol>
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]60 <li><a href="#t_integer">Integer Type</a></li>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]61 <li><a href="#t_floating">Floating Point Types</a></li>
62 <li><a href="#t_void">Void Type</a></li>
63 <li><a href="#t_label">Label Type</a></li>
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]64 <li><a href="#t_metadata">Metadata Type</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]65 </ol>
66 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]67 <li><a href="#t_derived">Derived Types</a>
68 <ol>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]69 <li><a href="#t_aggregate">Aggregate Types</a>
70 <ol>
71 <li><a href="#t_array">Array Type</a></li>
72 <li><a href="#t_struct">Structure Type</a></li>
73 <li><a href="#t_pstruct">Packed Structure Type</a></li>
74 <li><a href="#t_union">Union Type</a></li>
75 <li><a href="#t_vector">Vector Type</a></li>
76 </ol>
77 </li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]78 <li><a href="#t_function">Function Type</a></li>
79 <li><a href="#t_pointer">Pointer Type</a></li>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]80 <li><a href="#t_opaque">Opaque Type</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]81 </ol>
82 </li>
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]83 <li><a href="#t_uprefs">Type Up-references</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]84 </ol>
85 </li>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]86 <li><a href="#constants">Constants</a>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]87 <ol>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]88 <li><a href="#simpleconstants">Simple Constants</a></li>
Chris Lattner70882792009-02-28 18:32:25 +0000[diff] [blame]89 <li><a href="#complexconstants">Complex Constants</a></li>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]90 <li><a href="#globalconstants">Global Variable and Function Addresses</a></li>
91 <li><a href="#undefvalues">Undefined Values</a></li>
Chris Lattnerf9d078e2009-10-27 21:19:13 +0000[diff] [blame]92 <li><a href="#blockaddress">Addresses of Basic Blocks</a></li>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]93 <li><a href="#constantexprs">Constant Expressions</a></li>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]94 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]95 </li>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]96 <li><a href="#othervalues">Other Values</a>
97 <ol>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]98 <li><a href="#inlineasm">Inline Assembler Expressions</a></li>
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]99 <li><a href="#metadata">Metadata Nodes and Metadata Strings</a></li>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]100 </ol>
101 </li>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]102 <li><a href="#intrinsic_globals">Intrinsic Global Variables</a>
103 <ol>
104 <li><a href="#intg_used">The '<tt>llvm.used</tt>' Global Variable</a></li>
Chris Lattner401e10c2009-07-20 06:14:25 +0000[diff] [blame]105 <li><a href="#intg_compiler_used">The '<tt>llvm.compiler.used</tt>'
106 Global Variable</a></li>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]107 <li><a href="#intg_global_ctors">The '<tt>llvm.global_ctors</tt>'
108 Global Variable</a></li>
109 <li><a href="#intg_global_dtors">The '<tt>llvm.global_dtors</tt>'
110 Global Variable</a></li>
111 </ol>
112 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]113 <li><a href="#instref">Instruction Reference</a>
114 <ol>
115 <li><a href="#terminators">Terminator Instructions</a>
116 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]117 <li><a href="#i_ret">'<tt>ret</tt>' Instruction</a></li>
118 <li><a href="#i_br">'<tt>br</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]119 <li><a href="#i_switch">'<tt>switch</tt>' Instruction</a></li>
Chris Lattnerab21db72009-10-28 00:19:10 +0000[diff] [blame]120 <li><a href="#i_indirectbr">'<tt>indirectbr</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]121 <li><a href="#i_invoke">'<tt>invoke</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]122 <li><a href="#i_unwind">'<tt>unwind</tt>' Instruction</a></li>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]123 <li><a href="#i_unreachable">'<tt>unreachable</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]124 </ol>
125 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]126 <li><a href="#binaryops">Binary Operations</a>
127 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]128 <li><a href="#i_add">'<tt>add</tt>' Instruction</a></li>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]129 <li><a href="#i_fadd">'<tt>fadd</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]130 <li><a href="#i_sub">'<tt>sub</tt>' Instruction</a></li>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]131 <li><a href="#i_fsub">'<tt>fsub</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]132 <li><a href="#i_mul">'<tt>mul</tt>' Instruction</a></li>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]133 <li><a href="#i_fmul">'<tt>fmul</tt>' Instruction</a></li>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]134 <li><a href="#i_udiv">'<tt>udiv</tt>' Instruction</a></li>
135 <li><a href="#i_sdiv">'<tt>sdiv</tt>' Instruction</a></li>
136 <li><a href="#i_fdiv">'<tt>fdiv</tt>' Instruction</a></li>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]137 <li><a href="#i_urem">'<tt>urem</tt>' Instruction</a></li>
138 <li><a href="#i_srem">'<tt>srem</tt>' Instruction</a></li>
139 <li><a href="#i_frem">'<tt>frem</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]140 </ol>
141 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]142 <li><a href="#bitwiseops">Bitwise Binary Operations</a>
143 <ol>
Reid Spencer8e11bf82007-02-02 13:57:07 +0000[diff] [blame]144 <li><a href="#i_shl">'<tt>shl</tt>' Instruction</a></li>
145 <li><a href="#i_lshr">'<tt>lshr</tt>' Instruction</a></li>
146 <li><a href="#i_ashr">'<tt>ashr</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]147 <li><a href="#i_and">'<tt>and</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]148 <li><a href="#i_or">'<tt>or</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]149 <li><a href="#i_xor">'<tt>xor</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]150 </ol>
151 </li>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]152 <li><a href="#vectorops">Vector Operations</a>
153 <ol>
154 <li><a href="#i_extractelement">'<tt>extractelement</tt>' Instruction</a></li>
155 <li><a href="#i_insertelement">'<tt>insertelement</tt>' Instruction</a></li>
156 <li><a href="#i_shufflevector">'<tt>shufflevector</tt>' Instruction</a></li>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]157 </ol>
158 </li>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]159 <li><a href="#aggregateops">Aggregate Operations</a>
160 <ol>
161 <li><a href="#i_extractvalue">'<tt>extractvalue</tt>' Instruction</a></li>
162 <li><a href="#i_insertvalue">'<tt>insertvalue</tt>' Instruction</a></li>
163 </ol>
164 </li>
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]165 <li><a href="#memoryops">Memory Access and Addressing Operations</a>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]166 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]167 <li><a href="#i_alloca">'<tt>alloca</tt>' Instruction</a></li>
Robert Bocchino7b81c752006-02-17 21:18:08 +0000[diff] [blame]168 <li><a href="#i_load">'<tt>load</tt>' Instruction</a></li>
169 <li><a href="#i_store">'<tt>store</tt>' Instruction</a></li>
170 <li><a href="#i_getelementptr">'<tt>getelementptr</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]171 </ol>
172 </li>
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]173 <li><a href="#convertops">Conversion Operations</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]174 <ol>
175 <li><a href="#i_trunc">'<tt>trunc .. to</tt>' Instruction</a></li>
176 <li><a href="#i_zext">'<tt>zext .. to</tt>' Instruction</a></li>
177 <li><a href="#i_sext">'<tt>sext .. to</tt>' Instruction</a></li>
178 <li><a href="#i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a></li>
179 <li><a href="#i_fpext">'<tt>fpext .. to</tt>' Instruction</a></li>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]180 <li><a href="#i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a></li>
181 <li><a href="#i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a></li>
182 <li><a href="#i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a></li>
183 <li><a href="#i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a></li>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]184 <li><a href="#i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a></li>
185 <li><a href="#i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a></li>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]186 <li><a href="#i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a></li>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]187 </ol>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]188 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]189 <li><a href="#otherops">Other Operations</a>
190 <ol>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]191 <li><a href="#i_icmp">'<tt>icmp</tt>' Instruction</a></li>
192 <li><a href="#i_fcmp">'<tt>fcmp</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]193 <li><a href="#i_phi">'<tt>phi</tt>' Instruction</a></li>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]194 <li><a href="#i_select">'<tt>select</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]195 <li><a href="#i_call">'<tt>call</tt>' Instruction</a></li>
Chris Lattnerfb6977d2006-01-13 23:26:01 +0000[diff] [blame]196 <li><a href="#i_va_arg">'<tt>va_arg</tt>' Instruction</a></li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]197 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]198 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]199 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]200 </li>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]201 <li><a href="#intrinsics">Intrinsic Functions</a>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]202 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]203 <li><a href="#int_varargs">Variable Argument Handling Intrinsics</a>
204 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]205 <li><a href="#int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a></li>
206 <li><a href="#int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a></li>
207 <li><a href="#int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]208 </ol>
209 </li>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]210 <li><a href="#int_gc">Accurate Garbage Collection Intrinsics</a>
211 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]212 <li><a href="#int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a></li>
213 <li><a href="#int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a></li>
214 <li><a href="#int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a></li>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]215 </ol>
216 </li>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]217 <li><a href="#int_codegen">Code Generator Intrinsics</a>
218 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]219 <li><a href="#int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a></li>
220 <li><a href="#int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a></li>
221 <li><a href="#int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a></li>
222 <li><a href="#int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a></li>
223 <li><a href="#int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a></li>
224 <li><a href="#int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a></li>
225 <li><a href="#int_readcyclecounter"><tt>llvm.readcyclecounter</tt>' Intrinsic</a></li>
John Criswell7123e272004-04-09 16:43:20 +0000[diff] [blame]226 </ol>
227 </li>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]228 <li><a href="#int_libc">Standard C Library Intrinsics</a>
229 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]230 <li><a href="#int_memcpy">'<tt>llvm.memcpy.*</tt>' Intrinsic</a></li>
231 <li><a href="#int_memmove">'<tt>llvm.memmove.*</tt>' Intrinsic</a></li>
232 <li><a href="#int_memset">'<tt>llvm.memset.*</tt>' Intrinsic</a></li>
233 <li><a href="#int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a></li>
234 <li><a href="#int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a></li>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]235 <li><a href="#int_sin">'<tt>llvm.sin.*</tt>' Intrinsic</a></li>
236 <li><a href="#int_cos">'<tt>llvm.cos.*</tt>' Intrinsic</a></li>
237 <li><a href="#int_pow">'<tt>llvm.pow.*</tt>' Intrinsic</a></li>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]238 </ol>
239 </li>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]240 <li><a href="#int_manip">Bit Manipulation Intrinsics</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]241 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]242 <li><a href="#int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a></li>
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]243 <li><a href="#int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic </a></li>
244 <li><a href="#int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic </a></li>
245 <li><a href="#int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic </a></li>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]246 </ol>
247 </li>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]248 <li><a href="#int_overflow">Arithmetic with Overflow Intrinsics</a>
249 <ol>
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]250 <li><a href="#int_sadd_overflow">'<tt>llvm.sadd.with.overflow.*</tt> Intrinsics</a></li>
251 <li><a href="#int_uadd_overflow">'<tt>llvm.uadd.with.overflow.*</tt> Intrinsics</a></li>
252 <li><a href="#int_ssub_overflow">'<tt>llvm.ssub.with.overflow.*</tt> Intrinsics</a></li>
253 <li><a href="#int_usub_overflow">'<tt>llvm.usub.with.overflow.*</tt> Intrinsics</a></li>
254 <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]255 <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]256 </ol>
257 </li>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]258 <li><a href="#int_debugger">Debugger intrinsics</a></li>
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]259 <li><a href="#int_eh">Exception Handling intrinsics</a></li>
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]260 <li><a href="#int_trampoline">Trampoline Intrinsic</a>
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]261 <ol>
262 <li><a href="#int_it">'<tt>llvm.init.trampoline</tt>' Intrinsic</a></li>
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]263 </ol>
264 </li>
Bill Wendling3c44f5b2008-11-18 22:10:53 +0000[diff] [blame]265 <li><a href="#int_atomics">Atomic intrinsics</a>
266 <ol>
267 <li><a href="#int_memory_barrier"><tt>llvm.memory_barrier</tt></a></li>
268 <li><a href="#int_atomic_cmp_swap"><tt>llvm.atomic.cmp.swap</tt></a></li>
269 <li><a href="#int_atomic_swap"><tt>llvm.atomic.swap</tt></a></li>
270 <li><a href="#int_atomic_load_add"><tt>llvm.atomic.load.add</tt></a></li>
271 <li><a href="#int_atomic_load_sub"><tt>llvm.atomic.load.sub</tt></a></li>
272 <li><a href="#int_atomic_load_and"><tt>llvm.atomic.load.and</tt></a></li>
273 <li><a href="#int_atomic_load_nand"><tt>llvm.atomic.load.nand</tt></a></li>
274 <li><a href="#int_atomic_load_or"><tt>llvm.atomic.load.or</tt></a></li>
275 <li><a href="#int_atomic_load_xor"><tt>llvm.atomic.load.xor</tt></a></li>
276 <li><a href="#int_atomic_load_max"><tt>llvm.atomic.load.max</tt></a></li>
277 <li><a href="#int_atomic_load_min"><tt>llvm.atomic.load.min</tt></a></li>
278 <li><a href="#int_atomic_load_umax"><tt>llvm.atomic.load.umax</tt></a></li>
279 <li><a href="#int_atomic_load_umin"><tt>llvm.atomic.load.umin</tt></a></li>
280 </ol>
281 </li>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]282 <li><a href="#int_memorymarkers">Memory Use Markers</a>
283 <ol>
284 <li><a href="#int_lifetime_start"><tt>llvm.lifetime.start</tt></a></li>
285 <li><a href="#int_lifetime_end"><tt>llvm.lifetime.end</tt></a></li>
286 <li><a href="#int_invariant_start"><tt>llvm.invariant.start</tt></a></li>
287 <li><a href="#int_invariant_end"><tt>llvm.invariant.end</tt></a></li>
288 </ol>
289 </li>
Reid Spencer20677642007-07-20 19:59:11 +0000[diff] [blame]290 <li><a href="#int_general">General intrinsics</a>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]291 <ol>
Reid Spencer20677642007-07-20 19:59:11 +0000[diff] [blame]292 <li><a href="#int_var_annotation">
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]293 '<tt>llvm.var.annotation</tt>' Intrinsic</a></li>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]294 <li><a href="#int_annotation">
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]295 '<tt>llvm.annotation.*</tt>' Intrinsic</a></li>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]296 <li><a href="#int_trap">
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]297 '<tt>llvm.trap</tt>' Intrinsic</a></li>
298 <li><a href="#int_stackprotector">
299 '<tt>llvm.stackprotector</tt>' Intrinsic</a></li>
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]300 <li><a href="#int_objectsize">
301 '<tt>llvm.objectsize</tt>' Intrinsic</a></li>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]302 </ol>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]303 </li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]304 </ol>
305 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]306</ol>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]307
308<div class="doc_author">
309 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>
310 and <a href="mailto:vadve@cs.uiuc.edu">Vikram Adve</a></p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]311</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]312
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]313<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]314<div class="doc_section"> <a name="abstract">Abstract </a></div>
315<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]316
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]317<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]318
319<p>This document is a reference manual for the LLVM assembly language. LLVM is
320 a Static Single Assignment (SSA) based representation that provides type
321 safety, low-level operations, flexibility, and the capability of representing
322 'all' high-level languages cleanly. It is the common code representation
323 used throughout all phases of the LLVM compilation strategy.</p>
324
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]325</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]326
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]327<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]328<div class="doc_section"> <a name="introduction">Introduction</a> </div>
329<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]330
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]331<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]332
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]333<p>The LLVM code representation is designed to be used in three different forms:
334 as an in-memory compiler IR, as an on-disk bitcode representation (suitable
335 for fast loading by a Just-In-Time compiler), and as a human readable
336 assembly language representation. This allows LLVM to provide a powerful
337 intermediate representation for efficient compiler transformations and
338 analysis, while providing a natural means to debug and visualize the
339 transformations. The three different forms of LLVM are all equivalent. This
340 document describes the human readable representation and notation.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]341
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]342<p>The LLVM representation aims to be light-weight and low-level while being
343 expressive, typed, and extensible at the same time. It aims to be a
344 "universal IR" of sorts, by being at a low enough level that high-level ideas
345 may be cleanly mapped to it (similar to how microprocessors are "universal
346 IR's", allowing many source languages to be mapped to them). By providing
347 type information, LLVM can be used as the target of optimizations: for
348 example, through pointer analysis, it can be proven that a C automatic
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]349 variable is never accessed outside of the current function, allowing it to
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]350 be promoted to a simple SSA value instead of a memory location.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]351
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]352</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]353
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]354<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]355<div class="doc_subsubsection"> <a name="wellformed">Well-Formedness</a> </div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]356
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]357<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]358
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]359<p>It is important to note that this document describes 'well formed' LLVM
360 assembly language. There is a difference between what the parser accepts and
361 what is considered 'well formed'. For example, the following instruction is
362 syntactically okay, but not well formed:</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]363
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]364<div class="doc_code">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]365<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]366%x = <a href="#i_add">add</a> i32 1, %x
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]367</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]368</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]369
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]370<p>because the definition of <tt>%x</tt> does not dominate all of its uses. The
371 LLVM infrastructure provides a verification pass that may be used to verify
372 that an LLVM module is well formed. This pass is automatically run by the
373 parser after parsing input assembly and by the optimizer before it outputs
374 bitcode. The violations pointed out by the verifier pass indicate bugs in
375 transformation passes or input to the parser.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]376
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]377</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]378
Chris Lattnercc689392007-10-03 17:34:29 +0000[diff] [blame]379<!-- Describe the typesetting conventions here. -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]380
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]381<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]382<div class="doc_section"> <a name="identifiers">Identifiers</a> </div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]383<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]384
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]385<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]386
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]387<p>LLVM identifiers come in two basic types: global and local. Global
388 identifiers (functions, global variables) begin with the <tt>'@'</tt>
389 character. Local identifiers (register names, types) begin with
390 the <tt>'%'</tt> character. Additionally, there are three different formats
391 for identifiers, for different purposes:</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]392
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]393<ol>
Reid Spencer2c452282007-08-07 14:34:28 +0000[diff] [blame]394 <li>Named values are represented as a string of characters with their prefix.
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]395 For example, <tt>%foo</tt>, <tt>@DivisionByZero</tt>,
396 <tt>%a.really.long.identifier</tt>. The actual regular expression used is
397 '<tt>[%@][a-zA-Z$._][a-zA-Z$._0-9]*</tt>'. Identifiers which require
398 other characters in their names can be surrounded with quotes. Special
399 characters may be escaped using <tt>"\xx"</tt> where <tt>xx</tt> is the
400 ASCII code for the character in hexadecimal. In this way, any character
401 can be used in a name value, even quotes themselves.</li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]402
Reid Spencer2c452282007-08-07 14:34:28 +0000[diff] [blame]403 <li>Unnamed values are represented as an unsigned numeric value with their
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]404 prefix. For example, <tt>%12</tt>, <tt>@2</tt>, <tt>%44</tt>.</li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]405
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]406 <li>Constants, which are described in a <a href="#constants">section about
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]407 constants</a>, below.</li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]408</ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]409
Reid Spencer2c452282007-08-07 14:34:28 +0000[diff] [blame]410<p>LLVM requires that values start with a prefix for two reasons: Compilers
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]411 don't need to worry about name clashes with reserved words, and the set of
412 reserved words may be expanded in the future without penalty. Additionally,
413 unnamed identifiers allow a compiler to quickly come up with a temporary
414 variable without having to avoid symbol table conflicts.</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]415
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]416<p>Reserved words in LLVM are very similar to reserved words in other
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]417 languages. There are keywords for different opcodes
418 ('<tt><a href="#i_add">add</a></tt>',
419 '<tt><a href="#i_bitcast">bitcast</a></tt>',
420 '<tt><a href="#i_ret">ret</a></tt>', etc...), for primitive type names
421 ('<tt><a href="#t_void">void</a></tt>',
422 '<tt><a href="#t_primitive">i32</a></tt>', etc...), and others. These
423 reserved words cannot conflict with variable names, because none of them
424 start with a prefix character (<tt>'%'</tt> or <tt>'@'</tt>).</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]425
426<p>Here is an example of LLVM code to multiply the integer variable
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]427 '<tt>%X</tt>' by 8:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]428
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]429<p>The easy way:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]430
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]431<div class="doc_code">
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]432<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]433%result = <a href="#i_mul">mul</a> i32 %X, 8
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]434</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]435</div>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]436
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]437<p>After strength reduction:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]438
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]439<div class="doc_code">
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]440<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]441%result = <a href="#i_shl">shl</a> i32 %X, i8 3
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]442</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]443</div>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]444
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]445<p>And the hard way:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]446
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]447<div class="doc_code">
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]448<pre>
Gabor Greifec58f752009-10-28 13:05:07 +0000[diff] [blame]449%0 = <a href="#i_add">add</a> i32 %X, %X <i>; yields {i32}:%0</i>
450%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]451%result = <a href="#i_add">add</a> i32 %1, %1
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]452</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]453</div>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]454
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]455<p>This last way of multiplying <tt>%X</tt> by 8 illustrates several important
456 lexical features of LLVM:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]457
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]458<ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]459 <li>Comments are delimited with a '<tt>;</tt>' and go until the end of
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]460 line.</li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]461
462 <li>Unnamed temporaries are created when the result of a computation is not
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]463 assigned to a named value.</li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]464
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]465 <li>Unnamed temporaries are numbered sequentially</li>
466</ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]467
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]468<p>It also shows a convention that we follow in this document. When
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]469 demonstrating instructions, we will follow an instruction with a comment that
470 defines the type and name of value produced. Comments are shown in italic
471 text.</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]472
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]473</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]474
475<!-- *********************************************************************** -->
476<div class="doc_section"> <a name="highlevel">High Level Structure</a> </div>
477<!-- *********************************************************************** -->
478
479<!-- ======================================================================= -->
480<div class="doc_subsection"> <a name="modulestructure">Module Structure</a>
481</div>
482
483<div class="doc_text">
484
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]485<p>LLVM programs are composed of "Module"s, each of which is a translation unit
486 of the input programs. Each module consists of functions, global variables,
487 and symbol table entries. Modules may be combined together with the LLVM
488 linker, which merges function (and global variable) definitions, resolves
489 forward declarations, and merges symbol table entries. Here is an example of
490 the "hello world" module:</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]491
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]492<div class="doc_code">
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]493<pre>
494<i>; Declare the string constant as a global constant.</i>
495<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>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]496
497<i>; External declaration of the puts function</i>
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]498<a href="#functionstructure">declare</a> i32 @puts(i8 *) <i>; i32(i8 *)* </i>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]499
500<i>; Definition of main function</i>
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]501define i32 @main() { <i>; i32()* </i>
502 <i>; Convert [13 x i8]* to i8 *...</i>
503 %cast210 = <a href="#i_getelementptr">getelementptr</a> [13 x i8]* @.LC0, i64 0, i64 0 <i>; i8 *</i>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]504
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]505 <i>; Call puts function to write out the string to stdout.</i>
506 <a href="#i_call">call</a> i32 @puts(i8 * %cast210) <i>; i32</i>
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]507 <a href="#i_ret">ret</a> i32 0<br>}
508
509<i>; Named metadata</i>
510!1 = metadata !{i32 41}
511!foo = !{!1, null}
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]512</pre>
513</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]514
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]515<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]516 "<tt>.LC0</tt>", an external declaration of the "<tt>puts</tt>" function,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]517 a <a href="#functionstructure">function definition</a> for
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]518 "<tt>main</tt>" and <a href="#namedmetadatastructure">named metadata</a>
519 "<tt>foo"</tt>.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]520
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]521<p>In general, a module is made up of a list of global values, where both
522 functions and global variables are global values. Global values are
523 represented by a pointer to a memory location (in this case, a pointer to an
524 array of char, and a pointer to a function), and have one of the
525 following <a href="#linkage">linkage types</a>.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]526
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]527</div>
528
529<!-- ======================================================================= -->
530<div class="doc_subsection">
531 <a name="linkage">Linkage Types</a>
532</div>
533
534<div class="doc_text">
535
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]536<p>All Global Variables and Functions have one of the following types of
537 linkage:</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]538
539<dl>
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]540 <dt><tt><b><a name="linkage_private">private</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]541 <dd>Global values with private linkage are only directly accessible by objects
542 in the current module. In particular, linking code into a module with an
543 private global value may cause the private to be renamed as necessary to
544 avoid collisions. Because the symbol is private to the module, all
545 references can be updated. This doesn't show up in any symbol table in the
546 object file.</dd>
Rafael Espindolabb46f522009-01-15 20:18:42 +0000[diff] [blame]547
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]548 <dt><tt><b><a name="linkage_linker_private">linker_private</a></b></tt></dt>
Bill Wendling3d10a5a2009-07-20 01:03:30 +0000[diff] [blame]549 <dd>Similar to private, but the symbol is passed through the assembler and
Chris Lattnere1eaf912009-08-24 04:32:16 +0000[diff] [blame]550 removed by the linker after evaluation. Note that (unlike private
551 symbols) linker_private symbols are subject to coalescing by the linker:
552 weak symbols get merged and redefinitions are rejected. However, unlike
553 normal strong symbols, they are removed by the linker from the final
554 linked image (executable or dynamic library).</dd>
Bill Wendling3d10a5a2009-07-20 01:03:30 +0000[diff] [blame]555
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]556 <dt><tt><b><a name="linkage_internal">internal</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]557 <dd>Similar to private, but the value shows as a local symbol
558 (<tt>STB_LOCAL</tt> in the case of ELF) in the object file. This
559 corresponds to the notion of the '<tt>static</tt>' keyword in C.</dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]560
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]561 <dt><tt><b><a name="linkage_available_externally">available_externally</a></b></tt></dt>
Chris Lattner266c7bb2009-04-13 05:44:34 +0000[diff] [blame]562 <dd>Globals with "<tt>available_externally</tt>" linkage are never emitted
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]563 into the object file corresponding to the LLVM module. They exist to
564 allow inlining and other optimizations to take place given knowledge of
565 the definition of the global, which is known to be somewhere outside the
566 module. Globals with <tt>available_externally</tt> linkage are allowed to
567 be discarded at will, and are otherwise the same as <tt>linkonce_odr</tt>.
568 This linkage type is only allowed on definitions, not declarations.</dd>
Chris Lattner266c7bb2009-04-13 05:44:34 +0000[diff] [blame]569
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]570 <dt><tt><b><a name="linkage_linkonce">linkonce</a></b></tt></dt>
Chris Lattner4887bd82007-01-14 06:51:48 +0000[diff] [blame]571 <dd>Globals with "<tt>linkonce</tt>" linkage are merged with other globals of
Chris Lattner873187c2010-01-09 19:15:14 +0000[diff] [blame]572 the same name when linkage occurs. This can be used to implement
573 some forms of inline functions, templates, or other code which must be
574 generated in each translation unit that uses it, but where the body may
575 be overridden with a more definitive definition later. Unreferenced
576 <tt>linkonce</tt> globals are allowed to be discarded. Note that
577 <tt>linkonce</tt> linkage does not actually allow the optimizer to
578 inline the body of this function into callers because it doesn't know if
579 this definition of the function is the definitive definition within the
580 program or whether it will be overridden by a stronger definition.
581 To enable inlining and other optimizations, use "<tt>linkonce_odr</tt>"
582 linkage.</dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]583
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]584 <dt><tt><b><a name="linkage_weak">weak</a></b></tt></dt>
Chris Lattner26d054d2009-08-05 05:21:07 +0000[diff] [blame]585 <dd>"<tt>weak</tt>" linkage has the same merging semantics as
586 <tt>linkonce</tt> linkage, except that unreferenced globals with
587 <tt>weak</tt> linkage may not be discarded. This is used for globals that
588 are declared "weak" in C source code.</dd>
589
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]590 <dt><tt><b><a name="linkage_common">common</a></b></tt></dt>
Chris Lattner26d054d2009-08-05 05:21:07 +0000[diff] [blame]591 <dd>"<tt>common</tt>" linkage is most similar to "<tt>weak</tt>" linkage, but
592 they are used for tentative definitions in C, such as "<tt>int X;</tt>" at
593 global scope.
594 Symbols with "<tt>common</tt>" linkage are merged in the same way as
595 <tt>weak symbols</tt>, and they may not be deleted if unreferenced.
Chris Lattnercd81f5d2009-08-05 05:41:44 +0000[diff] [blame]596 <tt>common</tt> symbols may not have an explicit section,
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]597 must have a zero initializer, and may not be marked '<a
Chris Lattnercd81f5d2009-08-05 05:41:44 +0000[diff] [blame]598 href="#globalvars"><tt>constant</tt></a>'. Functions and aliases may not
599 have common linkage.</dd>
Chris Lattner26d054d2009-08-05 05:21:07 +0000[diff] [blame]600
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]601
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]602 <dt><tt><b><a name="linkage_appending">appending</a></b></tt></dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]603 <dd>"<tt>appending</tt>" linkage may only be applied to global variables of
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]604 pointer to array type. When two global variables with appending linkage
605 are linked together, the two global arrays are appended together. This is
606 the LLVM, typesafe, equivalent of having the system linker append together
607 "sections" with identical names when .o files are linked.</dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]608
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]609 <dt><tt><b><a name="linkage_externweak">extern_weak</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]610 <dd>The semantics of this linkage follow the ELF object file model: the symbol
611 is weak until linked, if not linked, the symbol becomes null instead of
612 being an undefined reference.</dd>
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame]613
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]614 <dt><tt><b><a name="linkage_linkonce_odr">linkonce_odr</a></b></tt></dt>
615 <dt><tt><b><a name="linkage_weak_odr">weak_odr</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]616 <dd>Some languages allow differing globals to be merged, such as two functions
617 with different semantics. Other languages, such as <tt>C++</tt>, ensure
618 that only equivalent globals are ever merged (the "one definition rule" -
619 "ODR"). Such languages can use the <tt>linkonce_odr</tt>
620 and <tt>weak_odr</tt> linkage types to indicate that the global will only
621 be merged with equivalent globals. These linkage types are otherwise the
622 same as their non-<tt>odr</tt> versions.</dd>
Duncan Sands667d4b82009-03-07 15:45:40 +0000[diff] [blame]623
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]624 <dt><tt><b><a name="linkage_external">externally visible</a></b></tt>:</dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]625 <dd>If none of the above identifiers are used, the global is externally
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]626 visible, meaning that it participates in linkage and can be used to
627 resolve external symbol references.</dd>
Reid Spencerc8910842007-04-11 23:49:50 +0000[diff] [blame]628</dl>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]629
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]630<p>The next two types of linkage are targeted for Microsoft Windows platform
631 only. They are designed to support importing (exporting) symbols from (to)
632 DLLs (Dynamic Link Libraries).</p>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]633
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]634<dl>
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]635 <dt><tt><b><a name="linkage_dllimport">dllimport</a></b></tt></dt>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]636 <dd>"<tt>dllimport</tt>" linkage causes the compiler to reference a function
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]637 or variable via a global pointer to a pointer that is set up by the DLL
638 exporting the symbol. On Microsoft Windows targets, the pointer name is
639 formed by combining <code>__imp_</code> and the function or variable
640 name.</dd>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]641
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]642 <dt><tt><b><a name="linkage_dllexport">dllexport</a></b></tt></dt>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]643 <dd>"<tt>dllexport</tt>" linkage causes the compiler to provide a global
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]644 pointer to a pointer in a DLL, so that it can be referenced with the
645 <tt>dllimport</tt> attribute. On Microsoft Windows targets, the pointer
646 name is formed by combining <code>__imp_</code> and the function or
647 variable name.</dd>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]648</dl>
649
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]650<p>For example, since the "<tt>.LC0</tt>" variable is defined to be internal, if
651 another module defined a "<tt>.LC0</tt>" variable and was linked with this
652 one, one of the two would be renamed, preventing a collision. Since
653 "<tt>main</tt>" and "<tt>puts</tt>" are external (i.e., lacking any linkage
654 declarations), they are accessible outside of the current module.</p>
655
656<p>It is illegal for a function <i>declaration</i> to have any linkage type
657 other than "externally visible", <tt>dllimport</tt>
658 or <tt>extern_weak</tt>.</p>
659
Duncan Sands667d4b82009-03-07 15:45:40 +0000[diff] [blame]660<p>Aliases can have only <tt>external</tt>, <tt>internal</tt>, <tt>weak</tt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]661 or <tt>weak_odr</tt> linkages.</p>
662
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]663</div>
664
665<!-- ======================================================================= -->
666<div class="doc_subsection">
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]667 <a name="callingconv">Calling Conventions</a>
668</div>
669
670<div class="doc_text">
671
672<p>LLVM <a href="#functionstructure">functions</a>, <a href="#i_call">calls</a>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]673 and <a href="#i_invoke">invokes</a> can all have an optional calling
674 convention specified for the call. The calling convention of any pair of
675 dynamic caller/callee must match, or the behavior of the program is
676 undefined. The following calling conventions are supported by LLVM, and more
677 may be added in the future:</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]678
679<dl>
680 <dt><b>"<tt>ccc</tt>" - The C calling convention</b>:</dt>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]681 <dd>This calling convention (the default if no other calling convention is
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]682 specified) matches the target C calling conventions. This calling
683 convention supports varargs function calls and tolerates some mismatch in
684 the declared prototype and implemented declaration of the function (as
685 does normal C).</dd>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]686
687 <dt><b>"<tt>fastcc</tt>" - The fast calling convention</b>:</dt>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]688 <dd>This calling convention attempts to make calls as fast as possible
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]689 (e.g. by passing things in registers). This calling convention allows the
690 target to use whatever tricks it wants to produce fast code for the
691 target, without having to conform to an externally specified ABI
Jeffrey Yasskin95fa80a2010-01-09 19:44:16 +0000[diff] [blame]692 (Application Binary Interface).
693 <a href="CodeGenerator.html#tailcallopt">Tail calls can only be optimized
694 when this convention is used.</a> This calling convention does not
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]695 support varargs and requires the prototype of all callees to exactly match
696 the prototype of the function definition.</dd>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]697
698 <dt><b>"<tt>coldcc</tt>" - The cold calling convention</b>:</dt>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]699 <dd>This calling convention attempts to make code in the caller as efficient
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]700 as possible under the assumption that the call is not commonly executed.
701 As such, these calls often preserve all registers so that the call does
702 not break any live ranges in the caller side. This calling convention
703 does not support varargs and requires the prototype of all callees to
704 exactly match the prototype of the function definition.</dd>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]705
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]706 <dt><b>"<tt>cc &lt;<em>n</em>&gt;</tt>" - Numbered convention</b>:</dt>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]707 <dd>Any calling convention may be specified by number, allowing
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]708 target-specific calling conventions to be used. Target specific calling
709 conventions start at 64.</dd>
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]710</dl>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]711
712<p>More calling conventions can be added/defined on an as-needed basis, to
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]713 support Pascal conventions or any other well-known target-independent
714 convention.</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]715
716</div>
717
718<!-- ======================================================================= -->
719<div class="doc_subsection">
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]720 <a name="visibility">Visibility Styles</a>
721</div>
722
723<div class="doc_text">
724
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]725<p>All Global Variables and Functions have one of the following visibility
726 styles:</p>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]727
728<dl>
729 <dt><b>"<tt>default</tt>" - Default style</b>:</dt>
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]730 <dd>On targets that use the ELF object file format, default visibility means
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]731 that the declaration is visible to other modules and, in shared libraries,
732 means that the declared entity may be overridden. On Darwin, default
733 visibility means that the declaration is visible to other modules. Default
734 visibility corresponds to "external linkage" in the language.</dd>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]735
736 <dt><b>"<tt>hidden</tt>" - Hidden style</b>:</dt>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]737 <dd>Two declarations of an object with hidden visibility refer to the same
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]738 object if they are in the same shared object. Usually, hidden visibility
739 indicates that the symbol will not be placed into the dynamic symbol
740 table, so no other module (executable or shared library) can reference it
741 directly.</dd>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]742
Anton Korobeynikov6f9896f2007-04-29 18:35:00 +0000[diff] [blame]743 <dt><b>"<tt>protected</tt>" - Protected style</b>:</dt>
Anton Korobeynikov6f9896f2007-04-29 18:35:00 +0000[diff] [blame]744 <dd>On ELF, protected visibility indicates that the symbol will be placed in
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]745 the dynamic symbol table, but that references within the defining module
746 will bind to the local symbol. That is, the symbol cannot be overridden by
747 another module.</dd>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]748</dl>
749
750</div>
751
752<!-- ======================================================================= -->
753<div class="doc_subsection">
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]754 <a name="namedtypes">Named Types</a>
755</div>
756
757<div class="doc_text">
758
759<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]760 it easier to read the IR and make the IR more condensed (particularly when
761 recursive types are involved). An example of a name specification is:</p>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]762
763<div class="doc_code">
764<pre>
765%mytype = type { %mytype*, i32 }
766</pre>
767</div>
768
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]769<p>You may give a name to any <a href="#typesystem">type</a> except
770 "<a href="t_void">void</a>". Type name aliases may be used anywhere a type
771 is expected with the syntax "%mytype".</p>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]772
773<p>Note that type names are aliases for the structural type that they indicate,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]774 and that you can therefore specify multiple names for the same type. This
775 often leads to confusing behavior when dumping out a .ll file. Since LLVM IR
776 uses structural typing, the name is not part of the type. When printing out
777 LLVM IR, the printer will pick <em>one name</em> to render all types of a
778 particular shape. This means that if you have code where two different
779 source types end up having the same LLVM type, that the dumper will sometimes
780 print the "wrong" or unexpected type. This is an important design point and
781 isn't going to change.</p>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]782
783</div>
784
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]785<!-- ======================================================================= -->
786<div class="doc_subsection">
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]787 <a name="globalvars">Global Variables</a>
788</div>
789
790<div class="doc_text">
791
Chris Lattner3689a342005-02-12 19:30:21 +0000[diff] [blame]792<p>Global variables define regions of memory allocated at compilation time
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]793 instead of run-time. Global variables may optionally be initialized, may
794 have an explicit section to be placed in, and may have an optional explicit
795 alignment specified. A variable may be defined as "thread_local", which
796 means that it will not be shared by threads (each thread will have a
797 separated copy of the variable). A variable may be defined as a global
798 "constant," which indicates that the contents of the variable
799 will <b>never</b> be modified (enabling better optimization, allowing the
800 global data to be placed in the read-only section of an executable, etc).
801 Note that variables that need runtime initialization cannot be marked
802 "constant" as there is a store to the variable.</p>
Chris Lattner3689a342005-02-12 19:30:21 +0000[diff] [blame]803
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]804<p>LLVM explicitly allows <em>declarations</em> of global variables to be marked
805 constant, even if the final definition of the global is not. This capability
806 can be used to enable slightly better optimization of the program, but
807 requires the language definition to guarantee that optimizations based on the
808 'constantness' are valid for the translation units that do not include the
809 definition.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]810
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]811<p>As SSA values, global variables define pointer values that are in scope
812 (i.e. they dominate) all basic blocks in the program. Global variables
813 always define a pointer to their "content" type because they describe a
814 region of memory, and all memory objects in LLVM are accessed through
815 pointers.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]816
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]817<p>A global variable may be declared to reside in a target-specific numbered
818 address space. For targets that support them, address spaces may affect how
819 optimizations are performed and/or what target instructions are used to
820 access the variable. The default address space is zero. The address space
821 qualifier must precede any other attributes.</p>
Christopher Lamb284d9922007-12-11 09:31:00 +0000[diff] [blame]822
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]823<p>LLVM allows an explicit section to be specified for globals. If the target
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]824 supports it, it will emit globals to the section specified.</p>
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]825
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]826<p>An explicit alignment may be specified for a global. If not present, or if
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]827 the alignment is set to zero, the alignment of the global is set by the
828 target to whatever it feels convenient. If an explicit alignment is
829 specified, the global is forced to have at least that much alignment. All
830 alignments must be a power of 2.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]831
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]832<p>For example, the following defines a global in a numbered address space with
833 an initializer, section, and alignment:</p>
Chris Lattner68027ea2007-01-14 00:27:09 +0000[diff] [blame]834
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]835<div class="doc_code">
Chris Lattner68027ea2007-01-14 00:27:09 +0000[diff] [blame]836<pre>
Dan Gohman398873c2009-01-11 00:40:00 +0000[diff] [blame]837@G = addrspace(5) constant float 1.0, section "foo", align 4
Chris Lattner68027ea2007-01-14 00:27:09 +0000[diff] [blame]838</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]839</div>
Chris Lattner68027ea2007-01-14 00:27:09 +0000[diff] [blame]840
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]841</div>
842
843
844<!-- ======================================================================= -->
845<div class="doc_subsection">
846 <a name="functionstructure">Functions</a>
847</div>
848
849<div class="doc_text">
850
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]851<p>LLVM function definitions consist of the "<tt>define</tt>" keyword, an
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]852 optional <a href="#linkage">linkage type</a>, an optional
853 <a href="#visibility">visibility style</a>, an optional
854 <a href="#callingconv">calling convention</a>, a return type, an optional
855 <a href="#paramattrs">parameter attribute</a> for the return type, a function
856 name, a (possibly empty) argument list (each with optional
857 <a href="#paramattrs">parameter attributes</a>), optional
858 <a href="#fnattrs">function attributes</a>, an optional section, an optional
859 alignment, an optional <a href="#gc">garbage collector name</a>, an opening
860 curly brace, a list of basic blocks, and a closing curly brace.</p>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]861
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]862<p>LLVM function declarations consist of the "<tt>declare</tt>" keyword, an
863 optional <a href="#linkage">linkage type</a>, an optional
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]864 <a href="#visibility">visibility style</a>, an optional
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]865 <a href="#callingconv">calling convention</a>, a return type, an optional
866 <a href="#paramattrs">parameter attribute</a> for the return type, a function
867 name, a possibly empty list of arguments, an optional alignment, and an
868 optional <a href="#gc">garbage collector name</a>.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]869
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]870<p>A function definition contains a list of basic blocks, forming the CFG
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]871 (Control Flow Graph) for the function. Each basic block may optionally start
872 with a label (giving the basic block a symbol table entry), contains a list
873 of instructions, and ends with a <a href="#terminators">terminator</a>
874 instruction (such as a branch or function return).</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]875
Chris Lattner4a3c9012007-06-08 16:52:14 +0000[diff] [blame]876<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]877 executed on entrance to the function, and it is not allowed to have
878 predecessor basic blocks (i.e. there can not be any branches to the entry
879 block of a function). Because the block can have no predecessors, it also
880 cannot have any <a href="#i_phi">PHI nodes</a>.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]881
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]882<p>LLVM allows an explicit section to be specified for functions. If the target
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]883 supports it, it will emit functions to the section specified.</p>
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]884
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]885<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]886 the alignment is set to zero, the alignment of the function is set by the
887 target to whatever it feels convenient. If an explicit alignment is
888 specified, the function is forced to have at least that much alignment. All
889 alignments must be a power of 2.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]890
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]891<h5>Syntax:</h5>
Devang Patel307e8ab2008-10-07 17:48:33 +0000[diff] [blame]892<div class="doc_code">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]893<pre>
Chris Lattner50ad45c2008-10-13 16:55:18 +0000[diff] [blame]894define [<a href="#linkage">linkage</a>] [<a href="#visibility">visibility</a>]
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]895 [<a href="#callingconv">cconv</a>] [<a href="#paramattrs">ret attrs</a>]
896 &lt;ResultType&gt; @&lt;FunctionName&gt; ([argument list])
897 [<a href="#fnattrs">fn Attrs</a>] [section "name"] [align N]
898 [<a href="#gc">gc</a>] { ... }
899</pre>
Devang Patel307e8ab2008-10-07 17:48:33 +0000[diff] [blame]900</div>
901
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]902</div>
903
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]904<!-- ======================================================================= -->
905<div class="doc_subsection">
906 <a name="aliasstructure">Aliases</a>
907</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]908
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]909<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]910
911<p>Aliases act as "second name" for the aliasee value (which can be either
912 function, global variable, another alias or bitcast of global value). Aliases
913 may have an optional <a href="#linkage">linkage type</a>, and an
914 optional <a href="#visibility">visibility style</a>.</p>
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]915
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]916<h5>Syntax:</h5>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]917<div class="doc_code">
Bill Wendlingaac388b2007-05-29 09:42:13 +0000[diff] [blame]918<pre>
Duncan Sands0b23ac12008-09-12 20:48:21 +0000[diff] [blame]919@&lt;Name&gt; = alias [Linkage] [Visibility] &lt;AliaseeTy&gt; @&lt;Aliasee&gt;
Bill Wendlingaac388b2007-05-29 09:42:13 +0000[diff] [blame]920</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]921</div>
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]922
923</div>
924
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]925<!-- ======================================================================= -->
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]926<div class="doc_subsection">
927 <a name="namedmetadatastructure">Named Metadata</a>
928</div>
929
930<div class="doc_text">
931
Chris Lattnere6a5ddd2010-01-15 21:50:19 +0000[diff] [blame]932<p>Named metadata is a collection of metadata. <a href="#metadata">Metadata
933 nodes</a> (but not metadata strings) and null are the only valid operands for
934 a named metadata.</p>
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]935
936<h5>Syntax:</h5>
937<div class="doc_code">
938<pre>
939!1 = metadata !{metadata !"one"}
940!name = !{null, !1}
941</pre>
942</div>
943
944</div>
945
946<!-- ======================================================================= -->
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]947<div class="doc_subsection"><a name="paramattrs">Parameter Attributes</a></div>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]948
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]949<div class="doc_text">
950
951<p>The return type and each parameter of a function type may have a set of
952 <i>parameter attributes</i> associated with them. Parameter attributes are
953 used to communicate additional information about the result or parameters of
954 a function. Parameter attributes are considered to be part of the function,
955 not of the function type, so functions with different parameter attributes
956 can have the same function type.</p>
957
958<p>Parameter attributes are simple keywords that follow the type specified. If
959 multiple parameter attributes are needed, they are space separated. For
960 example:</p>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]961
962<div class="doc_code">
963<pre>
Nick Lewyckyb6a7d252009-02-15 23:06:14 +0000[diff] [blame]964declare i32 @printf(i8* noalias nocapture, ...)
Chris Lattner66d922c2008-10-04 18:33:34 +0000[diff] [blame]965declare i32 @atoi(i8 zeroext)
966declare signext i8 @returns_signed_char()
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]967</pre>
968</div>
969
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]970<p>Note that any attributes for the function result (<tt>nounwind</tt>,
971 <tt>readonly</tt>) come immediately after the argument list.</p>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]972
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]973<p>Currently, only the following parameter attributes are defined:</p>
Chris Lattner47507de2008-01-11 06:20:47 +0000[diff] [blame]974
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]975<dl>
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]976 <dt><tt><b>zeroext</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]977 <dd>This indicates to the code generator that the parameter or return value
978 should be zero-extended to a 32-bit value by the caller (for a parameter)
979 or the callee (for a return value).</dd>
Chris Lattner47507de2008-01-11 06:20:47 +0000[diff] [blame]980
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]981 <dt><tt><b>signext</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]982 <dd>This indicates to the code generator that the parameter or return value
983 should be sign-extended to a 32-bit value by the caller (for a parameter)
984 or the callee (for a return value).</dd>
Chris Lattner47507de2008-01-11 06:20:47 +0000[diff] [blame]985
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]986 <dt><tt><b>inreg</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]987 <dd>This indicates that this parameter or return value should be treated in a
988 special target-dependent fashion during while emitting code for a function
989 call or return (usually, by putting it in a register as opposed to memory,
990 though some targets use it to distinguish between two different kinds of
991 registers). Use of this attribute is target-specific.</dd>
Chris Lattner47507de2008-01-11 06:20:47 +0000[diff] [blame]992
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]993 <dt><tt><b><a name="byval">byval</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]994 <dd>This indicates that the pointer parameter should really be passed by value
995 to the function. The attribute implies that a hidden copy of the pointee
996 is made between the caller and the callee, so the callee is unable to
997 modify the value in the callee. This attribute is only valid on LLVM
998 pointer arguments. It is generally used to pass structs and arrays by
999 value, but is also valid on pointers to scalars. The copy is considered
1000 to belong to the caller not the callee (for example,
1001 <tt><a href="#readonly">readonly</a></tt> functions should not write to
1002 <tt>byval</tt> parameters). This is not a valid attribute for return
1003 values. The byval attribute also supports specifying an alignment with
1004 the align attribute. This has a target-specific effect on the code
1005 generator that usually indicates a desired alignment for the synthesized
1006 stack slot.</dd>
1007
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1008 <dt><tt><b>sret</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1009 <dd>This indicates that the pointer parameter specifies the address of a
1010 structure that is the return value of the function in the source program.
1011 This pointer must be guaranteed by the caller to be valid: loads and
1012 stores to the structure may be assumed by the callee to not to trap. This
1013 may only be applied to the first parameter. This is not a valid attribute
1014 for return values. </dd>
1015
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1016 <dt><tt><b>noalias</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1017 <dd>This indicates that the pointer does not alias any global or any other
1018 parameter. The caller is responsible for ensuring that this is the
1019 case. On a function return value, <tt>noalias</tt> additionally indicates
1020 that the pointer does not alias any other pointers visible to the
1021 caller. For further details, please see the discussion of the NoAlias
1022 response in
1023 <a href="http://llvm.org/docs/AliasAnalysis.html#MustMayNo">alias
1024 analysis</a>.</dd>
1025
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1026 <dt><tt><b>nocapture</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1027 <dd>This indicates that the callee does not make any copies of the pointer
1028 that outlive the callee itself. This is not a valid attribute for return
1029 values.</dd>
1030
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1031 <dt><tt><b>nest</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1032 <dd>This indicates that the pointer parameter can be excised using the
1033 <a href="#int_trampoline">trampoline intrinsics</a>. This is not a valid
1034 attribute for return values.</dd>
1035</dl>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1036
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1037</div>
1038
1039<!-- ======================================================================= -->
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]1040<div class="doc_subsection">
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame]1041 <a name="gc">Garbage Collector Names</a>
1042</div>
1043
1044<div class="doc_text">
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame]1045
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1046<p>Each function may specify a garbage collector name, which is simply a
1047 string:</p>
1048
1049<div class="doc_code">
1050<pre>
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1051define void @f() gc "name" { ... }
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1052</pre>
1053</div>
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame]1054
1055<p>The compiler declares the supported values of <i>name</i>. Specifying a
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1056 collector which will cause the compiler to alter its output in order to
1057 support the named garbage collection algorithm.</p>
1058
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame]1059</div>
1060
1061<!-- ======================================================================= -->
1062<div class="doc_subsection">
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1063 <a name="fnattrs">Function Attributes</a>
Devang Patelf8b94812008-09-04 23:05:13 +0000[diff] [blame]1064</div>
1065
1066<div class="doc_text">
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1067
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1068<p>Function attributes are set to communicate additional information about a
1069 function. Function attributes are considered to be part of the function, not
1070 of the function type, so functions with different parameter attributes can
1071 have the same function type.</p>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1072
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1073<p>Function attributes are simple keywords that follow the type specified. If
1074 multiple attributes are needed, they are space separated. For example:</p>
Devang Patelf8b94812008-09-04 23:05:13 +0000[diff] [blame]1075
1076<div class="doc_code">
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1077<pre>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1078define void @f() noinline { ... }
1079define void @f() alwaysinline { ... }
1080define void @f() alwaysinline optsize { ... }
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1081define void @f() optsize { ... }
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1082</pre>
Devang Patelf8b94812008-09-04 23:05:13 +0000[diff] [blame]1083</div>
1084
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1085<dl>
Charles Davis1e063d12010-02-12 00:31:15 +0000[diff] [blame]1086 <dt><tt><b>alignstack(&lt;<em>n</em>&gt;)</b></tt></dt>
1087 <dd>This attribute indicates that, when emitting the prologue and epilogue,
1088 the backend should forcibly align the stack pointer. Specify the
1089 desired alignment, which must be a power of two, in parentheses.
1090
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1091 <dt><tt><b>alwaysinline</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1092 <dd>This attribute indicates that the inliner should attempt to inline this
1093 function into callers whenever possible, ignoring any active inlining size
1094 threshold for this caller.</dd>
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1095
Jakob Stoklund Olesen570a4a52010-02-06 01:16:28 +0000[diff] [blame]1096 <dt><tt><b>inlinehint</b></tt></dt>
1097 <dd>This attribute indicates that the source code contained a hint that inlining
1098 this function is desirable (such as the "inline" keyword in C/C++). It
1099 is just a hint; it imposes no requirements on the inliner.</dd>
1100
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1101 <dt><tt><b>noinline</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1102 <dd>This attribute indicates that the inliner should never inline this
1103 function in any situation. This attribute may not be used together with
1104 the <tt>alwaysinline</tt> attribute.</dd>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1105
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1106 <dt><tt><b>optsize</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1107 <dd>This attribute suggests that optimization passes and code generator passes
1108 make choices that keep the code size of this function low, and otherwise
1109 do optimizations specifically to reduce code size.</dd>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1110
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1111 <dt><tt><b>noreturn</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1112 <dd>This function attribute indicates that the function never returns
1113 normally. This produces undefined behavior at runtime if the function
1114 ever does dynamically return.</dd>
Bill Wendling31359ba2008-11-13 01:02:51 +0000[diff] [blame]1115
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1116 <dt><tt><b>nounwind</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1117 <dd>This function attribute indicates that the function never returns with an
1118 unwind or exceptional control flow. If the function does unwind, its
1119 runtime behavior is undefined.</dd>
Bill Wendlingfbaa7ed2008-11-26 19:07:40 +0000[diff] [blame]1120
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1121 <dt><tt><b>readnone</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1122 <dd>This attribute indicates that the function computes its result (or decides
1123 to unwind an exception) based strictly on its arguments, without
1124 dereferencing any pointer arguments or otherwise accessing any mutable
1125 state (e.g. memory, control registers, etc) visible to caller functions.
1126 It does not write through any pointer arguments
1127 (including <tt><a href="#byval">byval</a></tt> arguments) and never
1128 changes any state visible to callers. This means that it cannot unwind
1129 exceptions by calling the <tt>C++</tt> exception throwing methods, but
1130 could use the <tt>unwind</tt> instruction.</dd>
Devang Patel5d96fda2009-06-12 19:45:19 +0000[diff] [blame]1131
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1132 <dt><tt><b><a name="readonly">readonly</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1133 <dd>This attribute indicates that the function does not write through any
1134 pointer arguments (including <tt><a href="#byval">byval</a></tt>
1135 arguments) or otherwise modify any state (e.g. memory, control registers,
1136 etc) visible to caller functions. It may dereference pointer arguments
1137 and read state that may be set in the caller. A readonly function always
1138 returns the same value (or unwinds an exception identically) when called
1139 with the same set of arguments and global state. It cannot unwind an
1140 exception by calling the <tt>C++</tt> exception throwing methods, but may
1141 use the <tt>unwind</tt> instruction.</dd>
Anton Korobeynikovc5ec8a72009-07-17 18:07:26 +0000[diff] [blame]1142
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1143 <dt><tt><b><a name="ssp">ssp</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1144 <dd>This attribute indicates that the function should emit a stack smashing
1145 protector. It is in the form of a "canary"&mdash;a random value placed on
1146 the stack before the local variables that's checked upon return from the
1147 function to see if it has been overwritten. A heuristic is used to
1148 determine if a function needs stack protectors or not.<br>
1149<br>
1150 If a function that has an <tt>ssp</tt> attribute is inlined into a
1151 function that doesn't have an <tt>ssp</tt> attribute, then the resulting
1152 function will have an <tt>ssp</tt> attribute.</dd>
1153
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1154 <dt><tt><b>sspreq</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1155 <dd>This attribute indicates that the function should <em>always</em> emit a
1156 stack smashing protector. This overrides
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1157 the <tt><a href="#ssp">ssp</a></tt> function attribute.<br>
1158<br>
1159 If a function that has an <tt>sspreq</tt> attribute is inlined into a
1160 function that doesn't have an <tt>sspreq</tt> attribute or which has
1161 an <tt>ssp</tt> attribute, then the resulting function will have
1162 an <tt>sspreq</tt> attribute.</dd>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1163
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1164 <dt><tt><b>noredzone</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1165 <dd>This attribute indicates that the code generator should not use a red
1166 zone, even if the target-specific ABI normally permits it.</dd>
1167
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1168 <dt><tt><b>noimplicitfloat</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1169 <dd>This attributes disables implicit floating point instructions.</dd>
1170
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1171 <dt><tt><b>naked</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1172 <dd>This attribute disables prologue / epilogue emission for the function.
1173 This can have very system-specific consequences.</dd>
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1174</dl>
1175
Devang Patelf8b94812008-09-04 23:05:13 +0000[diff] [blame]1176</div>
1177
1178<!-- ======================================================================= -->
1179<div class="doc_subsection">
Chris Lattner1eeeb0c2006-04-08 04:40:53 +0000[diff] [blame]1180 <a name="moduleasm">Module-Level Inline Assembly</a>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]1181</div>
1182
1183<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1184
1185<p>Modules may contain "module-level inline asm" blocks, which corresponds to
1186 the GCC "file scope inline asm" blocks. These blocks are internally
1187 concatenated by LLVM and treated as a single unit, but may be separated in
1188 the <tt>.ll</tt> file if desired. The syntax is very simple:</p>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]1189
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1190<div class="doc_code">
1191<pre>
1192module asm "inline asm code goes here"
1193module asm "more can go here"
1194</pre>
1195</div>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]1196
1197<p>The strings can contain any character by escaping non-printable characters.
1198 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]1199 for the number.</p>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]1200
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1201<p>The inline asm code is simply printed to the machine code .s file when
1202 assembly code is generated.</p>
1203
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]1204</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]1205
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1206<!-- ======================================================================= -->
1207<div class="doc_subsection">
1208 <a name="datalayout">Data Layout</a>
1209</div>
1210
1211<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1212
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1213<p>A module may specify a target specific data layout string that specifies how
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1214 data is to be laid out in memory. The syntax for the data layout is
1215 simply:</p>
1216
1217<div class="doc_code">
1218<pre>
1219target datalayout = "<i>layout specification</i>"
1220</pre>
1221</div>
1222
1223<p>The <i>layout specification</i> consists of a list of specifications
1224 separated by the minus sign character ('-'). Each specification starts with
1225 a letter and may include other information after the letter to define some
1226 aspect of the data layout. The specifications accepted are as follows:</p>
1227
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1228<dl>
1229 <dt><tt>E</tt></dt>
1230 <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]1231 bits with the most significance have the lowest address location.</dd>
1232
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1233 <dt><tt>e</tt></dt>
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]1234 <dd>Specifies that the target lays out data in little-endian form. That is,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1235 the bits with the least significance have the lowest address
1236 location.</dd>
1237
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1238 <dt><tt>p:<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]1239 <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]1240 <i>preferred</i> alignments. All sizes are in bits. Specifying
1241 the <i>pref</i> alignment is optional. If omitted, the
1242 preceding <tt>:</tt> should be omitted too.</dd>
1243
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1244 <dt><tt>i<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1245 <dd>This specifies the alignment for an integer type of a given bit
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1246 <i>size</i>. The value of <i>size</i> must be in the range [1,2^23).</dd>
1247
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1248 <dt><tt>v<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]1249 <dd>This specifies the alignment for a vector type of a given bit
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1250 <i>size</i>.</dd>
1251
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1252 <dt><tt>f<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]1253 <dd>This specifies the alignment for a floating point type of a given bit
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1254 <i>size</i>. The value of <i>size</i> must be either 32 (float) or 64
1255 (double).</dd>
1256
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1257 <dt><tt>a<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1258 <dd>This specifies the alignment for an aggregate type of a given bit
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1259 <i>size</i>.</dd>
1260
Daniel Dunbar87bde0b2009-06-08 22:17:53 +0000[diff] [blame]1261 <dt><tt>s<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1262 <dd>This specifies the alignment for a stack object of a given bit
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1263 <i>size</i>.</dd>
Chris Lattnere82bdc42009-11-07 09:35:34 +0000[diff] [blame]1264
1265 <dt><tt>n<i>size1</i>:<i>size2</i>:<i>size3</i>...</tt></dt>
1266 <dd>This specifies a set of native integer widths for the target CPU
1267 in bits. For example, it might contain "n32" for 32-bit PowerPC,
1268 "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]1269 this set are considered to support most general arithmetic
Chris Lattnere82bdc42009-11-07 09:35:34 +0000[diff] [blame]1270 operations efficiently.</dd>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1271</dl>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1272
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1273<p>When constructing the data layout for a given target, LLVM starts with a
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1274 default set of specifications which are then (possibly) overriden by the
1275 specifications in the <tt>datalayout</tt> keyword. The default specifications
1276 are given in this list:</p>
1277
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1278<ul>
1279 <li><tt>E</tt> - big endian</li>
Dan Gohmanfdf2e8c2010-02-23 02:44:03 +0000[diff] [blame]1280 <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]1281 <li><tt>i1:8:8</tt> - i1 is 8-bit (byte) aligned</li>
1282 <li><tt>i8:8:8</tt> - i8 is 8-bit (byte) aligned</li>
1283 <li><tt>i16:16:16</tt> - i16 is 16-bit aligned</li>
1284 <li><tt>i32:32:32</tt> - i32 is 32-bit aligned</li>
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]1285 <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]1286 alignment of 64-bits</li>
1287 <li><tt>f32:32:32</tt> - float is 32-bit aligned</li>
1288 <li><tt>f64:64:64</tt> - double is 64-bit aligned</li>
1289 <li><tt>v64:64:64</tt> - 64-bit vector is 64-bit aligned</li>
1290 <li><tt>v128:128:128</tt> - 128-bit vector is 128-bit aligned</li>
1291 <li><tt>a0:0:1</tt> - aggregates are 8-bit aligned</li>
Daniel Dunbar87bde0b2009-06-08 22:17:53 +0000[diff] [blame]1292 <li><tt>s0:64:64</tt> - stack objects are 64-bit aligned</li>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1293</ul>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1294
1295<p>When LLVM is determining the alignment for a given type, it uses the
1296 following rules:</p>
1297
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1298<ol>
1299 <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]1300 specification is used.</li>
1301
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1302 <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]1303 smallest integer type that is larger than the bitwidth of the sought type
1304 is used. If none of the specifications are larger than the bitwidth then
1305 the the largest integer type is used. For example, given the default
1306 specifications above, the i7 type will use the alignment of i8 (next
1307 largest) while both i65 and i256 will use the alignment of i64 (largest
1308 specified).</li>
1309
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1310 <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]1311 largest vector type that is smaller than the sought vector type will be
1312 used as a fall back. This happens because &lt;128 x double&gt; can be
1313 implemented in terms of 64 &lt;2 x double&gt;, for example.</li>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1314</ol>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1315
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1316</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]1317
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1318<!-- ======================================================================= -->
1319<div class="doc_subsection">
1320 <a name="pointeraliasing">Pointer Aliasing Rules</a>
1321</div>
1322
1323<div class="doc_text">
1324
Andreas Bolka55e459a2009-07-29 00:02:05 +0000[diff] [blame]1325<p>Any memory access must be done through a pointer value associated
Andreas Bolka99a82052009-07-27 20:37:10 +0000[diff] [blame]1326with an address range of the memory access, otherwise the behavior
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1327is undefined. Pointer values are associated with address ranges
1328according to the following rules:</p>
1329
1330<ul>
Andreas Bolka55e459a2009-07-29 00:02:05 +0000[diff] [blame]1331 <li>A pointer value formed from a
1332 <tt><a href="#i_getelementptr">getelementptr</a></tt> instruction
1333 is associated with the addresses associated with the first operand
1334 of the <tt>getelementptr</tt>.</li>
1335 <li>An address of a global variable is associated with the address
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1336 range of the variable's storage.</li>
1337 <li>The result value of an allocation instruction is associated with
1338 the address range of the allocated storage.</li>
1339 <li>A null pointer in the default address-space is associated with
Andreas Bolka55e459a2009-07-29 00:02:05 +0000[diff] [blame]1340 no address.</li>
1341 <li>A pointer value formed by an
1342 <tt><a href="#i_inttoptr">inttoptr</a></tt> is associated with all
1343 address ranges of all pointer values that contribute (directly or
1344 indirectly) to the computation of the pointer's value.</li>
1345 <li>The result value of a
1346 <tt><a href="#i_bitcast">bitcast</a></tt> is associated with all
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1347 addresses associated with the operand of the <tt>bitcast</tt>.</li>
1348 <li>An integer constant other than zero or a pointer value returned
1349 from a function not defined within LLVM may be associated with address
1350 ranges allocated through mechanisms other than those provided by
Andreas Bolka55e459a2009-07-29 00:02:05 +0000[diff] [blame]1351 LLVM. Such ranges shall not overlap with any ranges of addresses
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1352 allocated by mechanisms provided by LLVM.</li>
1353 </ul>
1354
1355<p>LLVM IR does not associate types with memory. The result type of a
Andreas Bolka55e459a2009-07-29 00:02:05 +0000[diff] [blame]1356<tt><a href="#i_load">load</a></tt> merely indicates the size and
1357alignment of the memory from which to load, as well as the
1358interpretation of the value. The first operand of a
1359<tt><a href="#i_store">store</a></tt> similarly only indicates the size
1360and alignment of the store.</p>
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1361
1362<p>Consequently, type-based alias analysis, aka TBAA, aka
1363<tt>-fstrict-aliasing</tt>, is not applicable to general unadorned
1364LLVM IR. <a href="#metadata">Metadata</a> may be used to encode
1365additional information which specialized optimization passes may use
1366to implement type-based alias analysis.</p>
1367
1368</div>
1369
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1370<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1371<div class="doc_section"> <a name="typesystem">Type System</a> </div>
1372<!-- *********************************************************************** -->
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]1373
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1374<div class="doc_text">
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]1375
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1376<p>The LLVM type system is one of the most important features of the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1377 intermediate representation. Being typed enables a number of optimizations
1378 to be performed on the intermediate representation directly, without having
1379 to do extra analyses on the side before the transformation. A strong type
1380 system makes it easier to read the generated code and enables novel analyses
1381 and transformations that are not feasible to perform on normal three address
1382 code representations.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]1383
1384</div>
1385
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1386<!-- ======================================================================= -->
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1387<div class="doc_subsection"> <a name="t_classifications">Type
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1388Classifications</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1389
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1390<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1391
1392<p>The types fall into a few useful classifications:</p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1393
1394<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1395 <tbody>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1396 <tr><th>Classification</th><th>Types</th></tr>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1397 <tr>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1398 <td><a href="#t_integer">integer</a></td>
Reid Spencer2b916312007-05-16 18:44:01 +0000[diff] [blame]1399 <td><tt>i1, i2, i3, ... i8, ... i16, ... i32, ... i64, ... </tt></td>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1400 </tr>
1401 <tr>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1402 <td><a href="#t_floating">floating point</a></td>
1403 <td><tt>float, double, x86_fp80, fp128, ppc_fp128</tt></td>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1404 </tr>
1405 <tr>
1406 <td><a name="t_firstclass">first class</a></td>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1407 <td><a href="#t_integer">integer</a>,
1408 <a href="#t_floating">floating point</a>,
1409 <a href="#t_pointer">pointer</a>,
Dan Gohman0066db62008-06-18 18:42:13 +0000[diff] [blame]1410 <a href="#t_vector">vector</a>,
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]1411 <a href="#t_struct">structure</a>,
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]1412 <a href="#t_union">union</a>,
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]1413 <a href="#t_array">array</a>,
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1414 <a href="#t_label">label</a>,
1415 <a href="#t_metadata">metadata</a>.
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1416 </td>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1417 </tr>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1418 <tr>
1419 <td><a href="#t_primitive">primitive</a></td>
1420 <td><a href="#t_label">label</a>,
1421 <a href="#t_void">void</a>,
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1422 <a href="#t_floating">floating point</a>,
1423 <a href="#t_metadata">metadata</a>.</td>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1424 </tr>
1425 <tr>
1426 <td><a href="#t_derived">derived</a></td>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]1427 <td><a href="#t_array">array</a>,
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1428 <a href="#t_function">function</a>,
1429 <a href="#t_pointer">pointer</a>,
1430 <a href="#t_struct">structure</a>,
1431 <a href="#t_pstruct">packed structure</a>,
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]1432 <a href="#t_union">union</a>,
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1433 <a href="#t_vector">vector</a>,
1434 <a href="#t_opaque">opaque</a>.
Dan Gohman01ac1012008-10-14 16:32:04 +0000[diff] [blame]1435 </td>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1436 </tr>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1437 </tbody>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1438</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1439
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1440<p>The <a href="#t_firstclass">first class</a> types are perhaps the most
1441 important. Values of these types are the only ones which can be produced by
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]1442 instructions.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1443
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1444</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1445
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1446<!-- ======================================================================= -->
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1447<div class="doc_subsection"> <a name="t_primitive">Primitive Types</a> </div>
Chris Lattner8f8c7b72008-01-04 04:34:14 +0000[diff] [blame]1448
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1449<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1450
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1451<p>The primitive types are the fundamental building blocks of the LLVM
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1452 system.</p>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1453
Chris Lattner8f8c7b72008-01-04 04:34:14 +0000[diff] [blame]1454</div>
1455
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1456<!-- _______________________________________________________________________ -->
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]1457<div class="doc_subsubsection"> <a name="t_integer">Integer Type</a> </div>
1458
1459<div class="doc_text">
1460
1461<h5>Overview:</h5>
1462<p>The integer type is a very simple type that simply specifies an arbitrary
1463 bit width for the integer type desired. Any bit width from 1 bit to
1464 2<sup>23</sup>-1 (about 8 million) can be specified.</p>
1465
1466<h5>Syntax:</h5>
1467<pre>
1468 iN
1469</pre>
1470
1471<p>The number of bits the integer will occupy is specified by the <tt>N</tt>
1472 value.</p>
1473
1474<h5>Examples:</h5>
1475<table class="layout">
1476 <tr class="layout">
1477 <td class="left"><tt>i1</tt></td>
1478 <td class="left">a single-bit integer.</td>
1479 </tr>
1480 <tr class="layout">
1481 <td class="left"><tt>i32</tt></td>
1482 <td class="left">a 32-bit integer.</td>
1483 </tr>
1484 <tr class="layout">
1485 <td class="left"><tt>i1942652</tt></td>
1486 <td class="left">a really big integer of over 1 million bits.</td>
1487 </tr>
1488</table>
1489
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]1490</div>
1491
1492<!-- _______________________________________________________________________ -->
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1493<div class="doc_subsubsection"> <a name="t_floating">Floating Point Types</a> </div>
1494
1495<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1496
1497<table>
1498 <tbody>
1499 <tr><th>Type</th><th>Description</th></tr>
1500 <tr><td><tt>float</tt></td><td>32-bit floating point value</td></tr>
1501 <tr><td><tt>double</tt></td><td>64-bit floating point value</td></tr>
1502 <tr><td><tt>fp128</tt></td><td>128-bit floating point value (112-bit mantissa)</td></tr>
1503 <tr><td><tt>x86_fp80</tt></td><td>80-bit floating point value (X87)</td></tr>
1504 <tr><td><tt>ppc_fp128</tt></td><td>128-bit floating point value (two 64-bits)</td></tr>
1505 </tbody>
1506</table>
1507
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1508</div>
1509
1510<!-- _______________________________________________________________________ -->
1511<div class="doc_subsubsection"> <a name="t_void">Void Type</a> </div>
1512
1513<div class="doc_text">
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1514
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1515<h5>Overview:</h5>
1516<p>The void type does not represent any value and has no size.</p>
1517
1518<h5>Syntax:</h5>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1519<pre>
1520 void
1521</pre>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1522
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1523</div>
1524
1525<!-- _______________________________________________________________________ -->
1526<div class="doc_subsubsection"> <a name="t_label">Label Type</a> </div>
1527
1528<div class="doc_text">
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1529
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1530<h5>Overview:</h5>
1531<p>The label type represents code labels.</p>
1532
1533<h5>Syntax:</h5>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1534<pre>
1535 label
1536</pre>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1537
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1538</div>
1539
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1540<!-- _______________________________________________________________________ -->
1541<div class="doc_subsubsection"> <a name="t_metadata">Metadata Type</a> </div>
1542
1543<div class="doc_text">
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1544
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1545<h5>Overview:</h5>
Nick Lewyckyc261df92009-09-27 23:27:42 +0000[diff] [blame]1546<p>The metadata type represents embedded metadata. No derived types may be
1547 created from metadata except for <a href="#t_function">function</a>
1548 arguments.
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1549
1550<h5>Syntax:</h5>
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1551<pre>
1552 metadata
1553</pre>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1554
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1555</div>
1556
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1557
1558<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1559<div class="doc_subsection"> <a name="t_derived">Derived Types</a> </div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1560
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1561<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1562
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1563<p>The real power in LLVM comes from the derived types in the system. This is
1564 what allows a programmer to represent arrays, functions, pointers, and other
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]1565 useful types. Each of these types contain one or more element types which
1566 may be a primitive type, or another derived type. For example, it is
1567 possible to have a two dimensional array, using an array as the element type
1568 of another array.</p>
Dan Gohmand8791e52009-01-24 15:58:40 +0000[diff] [blame]1569
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]1570
1571</div>
1572
1573<!-- _______________________________________________________________________ -->
1574<div class="doc_subsubsection"> <a name="t_aggregate">Aggregate Types</a> </div>
1575
1576<div class="doc_text">
1577
1578<p>Aggregate Types are a subset of derived types that can contain multiple
1579 member types. <a href="#t_array">Arrays</a>,
1580 <a href="#t_struct">structs</a>, <a href="#t_vector">vectors</a> and
1581 <a href="#t_union">unions</a> are aggregate types.</p>
1582
1583</div>
1584
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1585</div>
Reid Spencer2b916312007-05-16 18:44:01 +0000[diff] [blame]1586
1587<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1588<div class="doc_subsubsection"> <a name="t_array">Array Type</a> </div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1589
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1590<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1591
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1592<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1593<p>The array type is a very simple derived type that arranges elements
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1594 sequentially in memory. The array type requires a size (number of elements)
1595 and an underlying data type.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1596
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1597<h5>Syntax:</h5>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1598<pre>
1599 [&lt;# elements&gt; x &lt;elementtype&gt;]
1600</pre>
1601
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1602<p>The number of elements is a constant integer value; <tt>elementtype</tt> may
1603 be any type with a size.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1604
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1605<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1606<table class="layout">
1607 <tr class="layout">
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]1608 <td class="left"><tt>[40 x i32]</tt></td>
1609 <td class="left">Array of 40 32-bit integer values.</td>
1610 </tr>
1611 <tr class="layout">
1612 <td class="left"><tt>[41 x i32]</tt></td>
1613 <td class="left">Array of 41 32-bit integer values.</td>
1614 </tr>
1615 <tr class="layout">
1616 <td class="left"><tt>[4 x i8]</tt></td>
1617 <td class="left">Array of 4 8-bit integer values.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1618 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1619</table>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1620<p>Here are some examples of multidimensional arrays:</p>
1621<table class="layout">
1622 <tr class="layout">
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]1623 <td class="left"><tt>[3 x [4 x i32]]</tt></td>
1624 <td class="left">3x4 array of 32-bit integer values.</td>
1625 </tr>
1626 <tr class="layout">
1627 <td class="left"><tt>[12 x [10 x float]]</tt></td>
1628 <td class="left">12x10 array of single precision floating point values.</td>
1629 </tr>
1630 <tr class="layout">
1631 <td class="left"><tt>[2 x [3 x [4 x i16]]]</tt></td>
1632 <td class="left">2x3x4 array of 16-bit integer values.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1633 </tr>
1634</table>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]1635
Dan Gohman7657f6b2009-11-09 19:01:53 +0000[diff] [blame]1636<p>There is no restriction on indexing beyond the end of the array implied by
1637 a static type (though there are restrictions on indexing beyond the bounds
1638 of an allocated object in some cases). This means that single-dimension
1639 'variable sized array' addressing can be implemented in LLVM with a zero
1640 length array type. An implementation of 'pascal style arrays' in LLVM could
1641 use the type "<tt>{ i32, [0 x float]}</tt>", for example.</p>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]1642
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1643</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1644
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1645<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1646<div class="doc_subsubsection"> <a name="t_function">Function Type</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1647
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1648<div class="doc_text">
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]1649
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1650<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1651<p>The function type can be thought of as a function signature. It consists of
1652 a return type and a list of formal parameter types. The return type of a
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]1653 function type is a scalar type, a void type, a struct type, or a union
1654 type. If the return type is a struct type then all struct elements must be
1655 of first class types, and the struct must have at least one element.</p>
Devang Patelc3fc6df2008-03-10 20:49:15 +0000[diff] [blame]1656
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1657<h5>Syntax:</h5>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]1658<pre>
Nick Lewycky51386942009-09-27 07:55:32 +0000[diff] [blame]1659 &lt;returntype&gt; (&lt;parameter list&gt;)
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]1660</pre>
1661
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]1662<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]1663 specifiers. Optionally, the parameter list may include a type <tt>...</tt>,
1664 which indicates that the function takes a variable number of arguments.
1665 Variable argument functions can access their arguments with
1666 the <a href="#int_varargs">variable argument handling intrinsic</a>
Chris Lattner0724fbd2010-03-02 06:36:51 +0000[diff] [blame]1667 functions. '<tt>&lt;returntype&gt;</tt>' is any type except
Nick Lewyckyc261df92009-09-27 23:27:42 +0000[diff] [blame]1668 <a href="#t_label">label</a>.</p>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]1669
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1670<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1671<table class="layout">
1672 <tr class="layout">
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]1673 <td class="left"><tt>i32 (i32)</tt></td>
1674 <td class="left">function taking an <tt>i32</tt>, returning an <tt>i32</tt>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1675 </td>
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]1676 </tr><tr class="layout">
Chris Lattner0724fbd2010-03-02 06:36:51 +0000[diff] [blame]1677 <td class="left"><tt>float&nbsp;(i16,&nbsp;i32&nbsp;*)&nbsp;*
Reid Spencerf17a0b72006-12-31 07:20:23 +0000[diff] [blame]1678 </tt></td>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]1679 <td class="left"><a href="#t_pointer">Pointer</a> to a function that takes
Chris Lattner0724fbd2010-03-02 06:36:51 +0000[diff] [blame]1680 an <tt>i16</tt> and a <a href="#t_pointer">pointer</a> to <tt>i32</tt>,
1681 returning <tt>float</tt>.
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]1682 </td>
1683 </tr><tr class="layout">
1684 <td class="left"><tt>i32 (i8*, ...)</tt></td>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]1685 <td class="left">A vararg function that takes at least one
1686 <a href="#t_pointer">pointer</a> to <tt>i8 </tt> (char in C),
1687 which returns an integer. This is the signature for <tt>printf</tt> in
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]1688 LLVM.
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1689 </td>
Devang Patela582f402008-03-24 05:35:41 +0000[diff] [blame]1690 </tr><tr class="layout">
1691 <td class="left"><tt>{i32, i32} (i32)</tt></td>
Nick Lewycky51386942009-09-27 07:55:32 +0000[diff] [blame]1692 <td class="left">A function taking an <tt>i32</tt>, returning a
1693 <a href="#t_struct">structure</a> containing two <tt>i32</tt> values
Devang Patela582f402008-03-24 05:35:41 +0000[diff] [blame]1694 </td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1695 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1696</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1697
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1698</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1699
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1700<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1701<div class="doc_subsubsection"> <a name="t_struct">Structure Type</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1702
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1703<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1704
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1705<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1706<p>The structure type is used to represent a collection of data members together
1707 in memory. The packing of the field types is defined to match the ABI of the
1708 underlying processor. The elements of a structure may be any type that has a
1709 size.</p>
1710
Jeffrey Yasskin7a088cf2010-01-11 19:19:26 +0000[diff] [blame]1711<p>Structures in memory are accessed using '<tt><a href="#i_load">load</a></tt>'
1712 and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a field
1713 with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.
1714 Structures in registers are accessed using the
1715 '<tt><a href="#i_extractvalue">extractvalue</a></tt>' and
1716 '<tt><a href="#i_insertvalue">insertvalue</a></tt>' instructions.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1717<h5>Syntax:</h5>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1718<pre>
1719 { &lt;type list&gt; }
1720</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1721
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1722<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1723<table class="layout">
1724 <tr class="layout">
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]1725 <td class="left"><tt>{ i32, i32, i32 }</tt></td>
1726 <td class="left">A triple of three <tt>i32</tt> values</td>
1727 </tr><tr class="layout">
1728 <td class="left"><tt>{&nbsp;float,&nbsp;i32&nbsp;(i32)&nbsp;*&nbsp;}</tt></td>
1729 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1730 second element is a <a href="#t_pointer">pointer</a> to a
1731 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1732 an <tt>i32</tt>.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1733 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1734</table>
Dan Gohmand8791e52009-01-24 15:58:40 +0000[diff] [blame]1735
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1736</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1737
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1738<!-- _______________________________________________________________________ -->
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]1739<div class="doc_subsubsection"> <a name="t_pstruct">Packed Structure Type</a>
1740</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1741
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]1742<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1743
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]1744<h5>Overview:</h5>
1745<p>The packed structure type is used to represent a collection of data members
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1746 together in memory. There is no padding between fields. Further, the
1747 alignment of a packed structure is 1 byte. The elements of a packed
1748 structure may be any type that has a size.</p>
1749
1750<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt> and
1751 '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a field with
1752 the '<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.</p>
1753
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]1754<h5>Syntax:</h5>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1755<pre>
1756 &lt; { &lt;type list&gt; } &gt;
1757</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1758
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]1759<h5>Examples:</h5>
1760<table class="layout">
1761 <tr class="layout">
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]1762 <td class="left"><tt>&lt; { i32, i32, i32 } &gt;</tt></td>
1763 <td class="left">A triple of three <tt>i32</tt> values</td>
1764 </tr><tr class="layout">
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1765 <td class="left">
1766<tt>&lt;&nbsp;{&nbsp;float,&nbsp;i32&nbsp;(i32)*&nbsp;}&nbsp;&gt;</tt></td>
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]1767 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1768 second element is a <a href="#t_pointer">pointer</a> to a
1769 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1770 an <tt>i32</tt>.</td>
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]1771 </tr>
1772</table>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1773
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]1774</div>
1775
1776<!-- _______________________________________________________________________ -->
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]1777<div class="doc_subsubsection"> <a name="t_union">Union Type</a> </div>
1778
1779<div class="doc_text">
1780
1781<h5>Overview:</h5>
1782<p>A union type describes an object with size and alignment suitable for
1783 an object of any one of a given set of types (also known as an "untagged"
1784 union). It is similar in concept and usage to a
1785 <a href="#t_struct">struct</a>, except that all members of the union
1786 have an offset of zero. The elements of a union may be any type that has a
1787 size. Unions must have at least one member - empty unions are not allowed.
1788 </p>
1789
1790<p>The size of the union as a whole will be the size of its largest member,
1791 and the alignment requirements of the union as a whole will be the largest
1792 alignment requirement of any member.</p>
1793
Dan Gohman2eddfef2010-02-25 16:51:31 +0000[diff] [blame]1794<p>Union members are accessed using '<tt><a href="#i_load">load</a></tt> and
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]1795 '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a field with
1796 the '<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.
1797 Since all members are at offset zero, the getelementptr instruction does
1798 not affect the address, only the type of the resulting pointer.</p>
1799
1800<h5>Syntax:</h5>
1801<pre>
1802 union { &lt;type list&gt; }
1803</pre>
1804
1805<h5>Examples:</h5>
1806<table class="layout">
1807 <tr class="layout">
1808 <td class="left"><tt>union { i32, i32*, float }</tt></td>
1809 <td class="left">A union of three types: an <tt>i32</tt>, a pointer to
1810 an <tt>i32</tt>, and a <tt>float</tt>.</td>
1811 </tr><tr class="layout">
1812 <td class="left">
1813 <tt>union {&nbsp;float,&nbsp;i32&nbsp;(i32)&nbsp;*&nbsp;}</tt></td>
1814 <td class="left">A union, where the first element is a <tt>float</tt> and the
1815 second element is a <a href="#t_pointer">pointer</a> to a
1816 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1817 an <tt>i32</tt>.</td>
1818 </tr>
1819</table>
1820
1821</div>
1822
1823<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1824<div class="doc_subsubsection"> <a name="t_pointer">Pointer Type</a> </div>
Chris Lattner0fd4a272009-02-08 19:53:29 +0000[diff] [blame]1825
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1826<div class="doc_text">
1827
1828<h5>Overview:</h5>
Dan Gohmanff3ef322010-02-25 16:50:07 +0000[diff] [blame]1829<p>The pointer type is used to specify memory locations.
1830 Pointers are commonly used to reference objects in memory.</p>
1831
1832<p>Pointer types may have an optional address space attribute defining the
1833 numbered address space where the pointed-to object resides. The default
1834 address space is number zero. The semantics of non-zero address
1835 spaces are target-specific.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1836
1837<p>Note that LLVM does not permit pointers to void (<tt>void*</tt>) nor does it
1838 permit pointers to labels (<tt>label*</tt>). Use <tt>i8*</tt> instead.</p>
Chris Lattner0fd4a272009-02-08 19:53:29 +0000[diff] [blame]1839
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1840<h5>Syntax:</h5>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1841<pre>
1842 &lt;type&gt; *
1843</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1844
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1845<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1846<table class="layout">
1847 <tr class="layout">
Dan Gohman2a08c532009-01-04 23:44:43 +0000[diff] [blame]1848 <td class="left"><tt>[4 x i32]*</tt></td>
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]1849 <td class="left">A <a href="#t_pointer">pointer</a> to <a
1850 href="#t_array">array</a> of four <tt>i32</tt> values.</td>
1851 </tr>
1852 <tr class="layout">
1853 <td class="left"><tt>i32 (i32 *) *</tt></td>
1854 <td class="left"> A <a href="#t_pointer">pointer</a> to a <a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1855 href="#t_function">function</a> that takes an <tt>i32*</tt>, returning an
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]1856 <tt>i32</tt>.</td>
1857 </tr>
1858 <tr class="layout">
1859 <td class="left"><tt>i32 addrspace(5)*</tt></td>
1860 <td class="left">A <a href="#t_pointer">pointer</a> to an <tt>i32</tt> value
1861 that resides in address space #5.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1862 </tr>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1863</table>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1864
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1865</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1866
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1867<!-- _______________________________________________________________________ -->
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1868<div class="doc_subsubsection"> <a name="t_vector">Vector Type</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1869
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1870<div class="doc_text">
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1871
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1872<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1873<p>A vector type is a simple derived type that represents a vector of elements.
1874 Vector types are used when multiple primitive data are operated in parallel
1875 using a single instruction (SIMD). A vector type requires a size (number of
Duncan Sandsd40d14e2009-11-27 13:38:03 +0000[diff] [blame]1876 elements) and an underlying primitive data type. Vector types are considered
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1877 <a href="#t_firstclass">first class</a>.</p>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1878
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1879<h5>Syntax:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1880<pre>
1881 &lt; &lt;# elements&gt; x &lt;elementtype&gt; &gt;
1882</pre>
1883
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1884<p>The number of elements is a constant integer value; elementtype may be any
1885 integer or floating point type.</p>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1886
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1887<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1888<table class="layout">
1889 <tr class="layout">
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]1890 <td class="left"><tt>&lt;4 x i32&gt;</tt></td>
1891 <td class="left">Vector of 4 32-bit integer values.</td>
1892 </tr>
1893 <tr class="layout">
1894 <td class="left"><tt>&lt;8 x float&gt;</tt></td>
1895 <td class="left">Vector of 8 32-bit floating-point values.</td>
1896 </tr>
1897 <tr class="layout">
1898 <td class="left"><tt>&lt;2 x i64&gt;</tt></td>
1899 <td class="left">Vector of 2 64-bit integer values.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1900 </tr>
1901</table>
Dan Gohmand8791e52009-01-24 15:58:40 +0000[diff] [blame]1902
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1903</div>
1904
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1905<!-- _______________________________________________________________________ -->
1906<div class="doc_subsubsection"> <a name="t_opaque">Opaque Type</a> </div>
1907<div class="doc_text">
1908
1909<h5>Overview:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1910<p>Opaque types are used to represent unknown types in the system. This
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1911 corresponds (for example) to the C notion of a forward declared structure
1912 type. In LLVM, opaque types can eventually be resolved to any type (not just
1913 a structure type).</p>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1914
1915<h5>Syntax:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1916<pre>
1917 opaque
1918</pre>
1919
1920<h5>Examples:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1921<table class="layout">
1922 <tr class="layout">
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]1923 <td class="left"><tt>opaque</tt></td>
1924 <td class="left">An opaque type.</td>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1925 </tr>
1926</table>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1927
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1928</div>
1929
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]1930<!-- ======================================================================= -->
1931<div class="doc_subsection">
1932 <a name="t_uprefs">Type Up-references</a>
1933</div>
1934
1935<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1936
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]1937<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1938<p>An "up reference" allows you to refer to a lexically enclosing type without
1939 requiring it to have a name. For instance, a structure declaration may
1940 contain a pointer to any of the types it is lexically a member of. Example
1941 of up references (with their equivalent as named type declarations)
1942 include:</p>
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]1943
1944<pre>
Chris Lattner3060f5b2009-02-09 10:00:56 +0000[diff] [blame]1945 { \2 * } %x = type { %x* }
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]1946 { \2 }* %y = type { %y }*
1947 \1* %z = type %z*
1948</pre>
1949
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1950<p>An up reference is needed by the asmprinter for printing out cyclic types
1951 when there is no declared name for a type in the cycle. Because the
1952 asmprinter does not want to print out an infinite type string, it needs a
1953 syntax to handle recursive types that have no names (all names are optional
1954 in llvm IR).</p>
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]1955
1956<h5>Syntax:</h5>
1957<pre>
1958 \&lt;level&gt;
1959</pre>
1960
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1961<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]1962
1963<h5>Examples:</h5>
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]1964<table class="layout">
1965 <tr class="layout">
1966 <td class="left"><tt>\1*</tt></td>
1967 <td class="left">Self-referential pointer.</td>
1968 </tr>
1969 <tr class="layout">
1970 <td class="left"><tt>{ { \3*, i8 }, i32 }</tt></td>
1971 <td class="left">Recursive structure where the upref refers to the out-most
1972 structure.</td>
1973 </tr>
1974</table>
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]1975
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1976</div>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1977
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1978<!-- *********************************************************************** -->
1979<div class="doc_section"> <a name="constants">Constants</a> </div>
1980<!-- *********************************************************************** -->
1981
1982<div class="doc_text">
1983
1984<p>LLVM has several different basic types of constants. This section describes
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1985 them all and their syntax.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1986
1987</div>
1988
1989<!-- ======================================================================= -->
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]1990<div class="doc_subsection"><a name="simpleconstants">Simple Constants</a></div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1991
1992<div class="doc_text">
1993
1994<dl>
1995 <dt><b>Boolean constants</b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1996 <dd>The two strings '<tt>true</tt>' and '<tt>false</tt>' are both valid
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]1997 constants of the <tt><a href="#t_integer">i1</a></tt> type.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1998
1999 <dt><b>Integer constants</b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2000 <dd>Standard integers (such as '4') are constants of
2001 the <a href="#t_integer">integer</a> type. Negative numbers may be used
2002 with integer types.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2003
2004 <dt><b>Floating point constants</b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2005 <dd>Floating point constants use standard decimal notation (e.g. 123.421),
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2006 exponential notation (e.g. 1.23421e+2), or a more precise hexadecimal
2007 notation (see below). The assembler requires the exact decimal value of a
2008 floating-point constant. For example, the assembler accepts 1.25 but
2009 rejects 1.3 because 1.3 is a repeating decimal in binary. Floating point
2010 constants must have a <a href="#t_floating">floating point</a> type. </dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2011
2012 <dt><b>Null pointer constants</b></dt>
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]2013 <dd>The identifier '<tt>null</tt>' is recognized as a null pointer constant
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2014 and must be of <a href="#t_pointer">pointer type</a>.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2015</dl>
2016
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2017<p>The one non-intuitive notation for constants is the hexadecimal form of
2018 floating point constants. For example, the form '<tt>double
2019 0x432ff973cafa8000</tt>' is equivalent to (but harder to read than)
2020 '<tt>double 4.5e+15</tt>'. The only time hexadecimal floating point
2021 constants are required (and the only time that they are generated by the
2022 disassembler) is when a floating point constant must be emitted but it cannot
2023 be represented as a decimal floating point number in a reasonable number of
2024 digits. For example, NaN's, infinities, and other special values are
2025 represented in their IEEE hexadecimal format so that assembly and disassembly
2026 do not cause any bits to change in the constants.</p>
2027
Dale Johannesenbd5e5a82009-02-11 22:14:51 +0000[diff] [blame]2028<p>When using the hexadecimal form, constants of types float and double are
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2029 represented using the 16-digit form shown above (which matches the IEEE754
2030 representation for double); float values must, however, be exactly
2031 representable as IEE754 single precision. Hexadecimal format is always used
2032 for long double, and there are three forms of long double. The 80-bit format
2033 used by x86 is represented as <tt>0xK</tt> followed by 20 hexadecimal digits.
2034 The 128-bit format used by PowerPC (two adjacent doubles) is represented
2035 by <tt>0xM</tt> followed by 32 hexadecimal digits. The IEEE 128-bit format
2036 is represented by <tt>0xL</tt> followed by 32 hexadecimal digits; no
2037 currently supported target uses this format. Long doubles will only work if
2038 they match the long double format on your target. All hexadecimal formats
2039 are big-endian (sign bit at the left).</p>
2040
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2041</div>
2042
2043<!-- ======================================================================= -->
Chris Lattner70882792009-02-28 18:32:25 +0000[diff] [blame]2044<div class="doc_subsection">
Bill Wendlingd9fe2982009-07-20 02:32:41 +0000[diff] [blame]2045<a name="aggregateconstants"></a> <!-- old anchor -->
2046<a name="complexconstants">Complex Constants</a>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2047</div>
2048
2049<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2050
Chris Lattner70882792009-02-28 18:32:25 +0000[diff] [blame]2051<p>Complex constants are a (potentially recursive) combination of simple
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2052 constants and smaller complex constants.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2053
2054<dl>
2055 <dt><b>Structure constants</b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2056 <dd>Structure constants are represented with notation similar to structure
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2057 type definitions (a comma separated list of elements, surrounded by braces
2058 (<tt>{}</tt>)). For example: "<tt>{ i32 4, float 17.0, i32* @G }</tt>",
2059 where "<tt>@G</tt>" is declared as "<tt>@G = external global i32</tt>".
2060 Structure constants must have <a href="#t_struct">structure type</a>, and
2061 the number and types of elements must match those specified by the
2062 type.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2063
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]2064 <dt><b>Union constants</b></dt>
2065 <dd>Union constants are represented with notation similar to a structure with
2066 a single element - that is, a single typed element surrounded
2067 by braces (<tt>{}</tt>)). For example: "<tt>{ i32 4 }</tt>". The
2068 <a href="#t_union">union type</a> can be initialized with a single-element
2069 struct as long as the type of the struct element matches the type of
2070 one of the union members.</dd>
2071
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2072 <dt><b>Array constants</b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2073 <dd>Array constants are represented with notation similar to array type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2074 definitions (a comma separated list of elements, surrounded by square
2075 brackets (<tt>[]</tt>)). For example: "<tt>[ i32 42, i32 11, i32 74
2076 ]</tt>". Array constants must have <a href="#t_array">array type</a>, and
2077 the number and types of elements must match those specified by the
2078 type.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2079
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]2080 <dt><b>Vector constants</b></dt>
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]2081 <dd>Vector constants are represented with notation similar to vector type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2082 definitions (a comma separated list of elements, surrounded by
2083 less-than/greater-than's (<tt>&lt;&gt;</tt>)). For example: "<tt>&lt; i32
2084 42, i32 11, i32 74, i32 100 &gt;</tt>". Vector constants must
2085 have <a href="#t_vector">vector type</a>, and the number and types of
2086 elements must match those specified by the type.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2087
2088 <dt><b>Zero initialization</b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2089 <dd>The string '<tt>zeroinitializer</tt>' can be used to zero initialize a
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]2090 value to zero of <em>any</em> type, including scalar and
2091 <a href="#t_aggregate">aggregate</a> types.
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2092 This is often used to avoid having to print large zero initializers
2093 (e.g. for large arrays) and is always exactly equivalent to using explicit
2094 zero initializers.</dd>
Nick Lewycky21cc4462009-04-04 07:22:01 +0000[diff] [blame]2095
2096 <dt><b>Metadata node</b></dt>
Nick Lewycky1e8c7a62009-05-30 16:08:30 +0000[diff] [blame]2097 <dd>A metadata node is a structure-like constant with
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2098 <a href="#t_metadata">metadata type</a>. For example: "<tt>metadata !{
2099 i32 0, metadata !"test" }</tt>". Unlike other constants that are meant to
2100 be interpreted as part of the instruction stream, metadata is a place to
2101 attach additional information such as debug info.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2102</dl>
2103
2104</div>
2105
2106<!-- ======================================================================= -->
2107<div class="doc_subsection">
2108 <a name="globalconstants">Global Variable and Function Addresses</a>
2109</div>
2110
2111<div class="doc_text">
2112
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2113<p>The addresses of <a href="#globalvars">global variables</a>
2114 and <a href="#functionstructure">functions</a> are always implicitly valid
2115 (link-time) constants. These constants are explicitly referenced when
2116 the <a href="#identifiers">identifier for the global</a> is used and always
2117 have <a href="#t_pointer">pointer</a> type. For example, the following is a
2118 legal LLVM file:</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2119
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2120<div class="doc_code">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2121<pre>
Chris Lattnera18a4242007-06-06 18:28:13 +0000[diff] [blame]2122@X = global i32 17
2123@Y = global i32 42
2124@Z = global [2 x i32*] [ i32* @X, i32* @Y ]
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2125</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2126</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2127
2128</div>
2129
2130<!-- ======================================================================= -->
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]2131<div class="doc_subsection"><a name="undefvalues">Undefined Values</a></div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2132<div class="doc_text">
2133
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2134<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]2135 indicates that the user of the value may receive an unspecified bit-pattern.
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2136 Undefined values may be of any type (other than label or void) and be used
2137 anywhere a constant is permitted.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2138
Chris Lattnerc608cb12009-09-11 01:49:31 +0000[diff] [blame]2139<p>Undefined values are useful because they indicate to the compiler that the
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2140 program is well defined no matter what value is used. This gives the
2141 compiler more freedom to optimize. Here are some examples of (potentially
2142 surprising) transformations that are valid (in pseudo IR):</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2143
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2144
2145<div class="doc_code">
2146<pre>
2147 %A = add %X, undef
2148 %B = sub %X, undef
2149 %C = xor %X, undef
2150Safe:
2151 %A = undef
2152 %B = undef
2153 %C = undef
2154</pre>
2155</div>
2156
2157<p>This is safe because all of the output bits are affected by the undef bits.
2158Any output bit can have a zero or one depending on the input bits.</p>
2159
2160<div class="doc_code">
2161<pre>
2162 %A = or %X, undef
2163 %B = and %X, undef
2164Safe:
2165 %A = -1
2166 %B = 0
2167Unsafe:
2168 %A = undef
2169 %B = undef
2170</pre>
2171</div>
2172
2173<p>These logical operations have bits that are not always affected by the input.
2174For example, if "%X" has a zero bit, then the output of the 'and' operation will
2175always be a zero, no matter what the corresponding bit from the undef is. As
Chris Lattnerc608cb12009-09-11 01:49:31 +0000[diff] [blame]2176such, 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]2177However, it is safe to assume that all bits of the undef could be 0, and
2178optimize the and to 0. Likewise, it is safe to assume that all the bits of
2179the 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]2180-1.</p>
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2181
2182<div class="doc_code">
2183<pre>
2184 %A = select undef, %X, %Y
2185 %B = select undef, 42, %Y
2186 %C = select %X, %Y, undef
2187Safe:
2188 %A = %X (or %Y)
2189 %B = 42 (or %Y)
2190 %C = %Y
2191Unsafe:
2192 %A = undef
2193 %B = undef
2194 %C = undef
2195</pre>
2196</div>
2197
2198<p>This set of examples show that undefined select (and conditional branch)
2199conditions can go "either way" but they have to come from one of the two
2200operands. In the %A example, if %X and %Y were both known to have a clear low
2201bit, then %A would have to have a cleared low bit. However, in the %C example,
2202the optimizer is allowed to assume that the undef operand could be the same as
2203%Y, allowing the whole select to be eliminated.</p>
2204
2205
2206<div class="doc_code">
2207<pre>
2208 %A = xor undef, undef
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]2209
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2210 %B = undef
2211 %C = xor %B, %B
2212
2213 %D = undef
2214 %E = icmp lt %D, 4
2215 %F = icmp gte %D, 4
2216
2217Safe:
2218 %A = undef
2219 %B = undef
2220 %C = undef
2221 %D = undef
2222 %E = undef
2223 %F = undef
2224</pre>
2225</div>
2226
2227<p>This example points out that two undef operands are not necessarily the same.
2228This can be surprising to people (and also matches C semantics) where they
2229assume that "X^X" is always zero, even if X is undef. This isn't true for a
2230number of reasons, but the short answer is that an undef "variable" can
2231arbitrarily change its value over its "live range". This is true because the
2232"variable" doesn't actually <em>have a live range</em>. Instead, the value is
2233logically read from arbitrary registers that happen to be around when needed,
Benjamin Kramer8040cd32009-10-12 14:46:08 +0000[diff] [blame]2234so the value is not necessarily consistent over time. In fact, %A and %C need
Chris Lattner349eb412009-09-08 15:13:16 +0000[diff] [blame]2235to have the same semantics or the core LLVM "replace all uses with" concept
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2236would not hold.</p>
Chris Lattner6e9057b2009-09-07 23:33:52 +0000[diff] [blame]2237
2238<div class="doc_code">
2239<pre>
2240 %A = fdiv undef, %X
2241 %B = fdiv %X, undef
2242Safe:
2243 %A = undef
2244b: unreachable
2245</pre>
2246</div>
2247
2248<p>These examples show the crucial difference between an <em>undefined
2249value</em> and <em>undefined behavior</em>. An undefined value (like undef) is
2250allowed to have an arbitrary bit-pattern. This means that the %A operation
2251can be constant folded to undef because the undef could be an SNaN, and fdiv is
2252not (currently) defined on SNaN's. However, in the second example, we can make
2253a more aggressive assumption: because the undef is allowed to be an arbitrary
2254value, 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]2255has <em>undefined behavior</em>, we are allowed to assume that the operation
Chris Lattner6e9057b2009-09-07 23:33:52 +0000[diff] [blame]2256does not execute at all. This allows us to delete the divide and all code after
2257it: since the undefined operation "can't happen", the optimizer can assume that
2258it occurs in dead code.
2259</p>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]2260
Chris Lattner6e9057b2009-09-07 23:33:52 +0000[diff] [blame]2261<div class="doc_code">
2262<pre>
2263a: store undef -> %X
2264b: store %X -> undef
2265Safe:
2266a: &lt;deleted&gt;
2267b: unreachable
2268</pre>
2269</div>
2270
2271<p>These examples reiterate the fdiv example: a store "of" an undefined value
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]2272can be assumed to not have any effect: we can assume that the value is
Chris Lattner6e9057b2009-09-07 23:33:52 +0000[diff] [blame]2273overwritten with bits that happen to match what was already there. However, a
2274store "to" an undefined location could clobber arbitrary memory, therefore, it
2275has undefined behavior.</p>
2276
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2277</div>
2278
2279<!-- ======================================================================= -->
Chris Lattnerf9d078e2009-10-27 21:19:13 +0000[diff] [blame]2280<div class="doc_subsection"><a name="blockaddress">Addresses of Basic
2281 Blocks</a></div>
Chris Lattnerc6f44362009-10-27 21:01:34 +0000[diff] [blame]2282<div class="doc_text">
2283
Chris Lattnercdfc9402009-11-01 01:27:45 +0000[diff] [blame]2284<p><b><tt>blockaddress(@function, %block)</tt></b></p>
Chris Lattnerc6f44362009-10-27 21:01:34 +0000[diff] [blame]2285
2286<p>The '<tt>blockaddress</tt>' constant computes the address of the specified
Chris Lattner2dfdf2a2009-10-27 21:49:40 +0000[diff] [blame]2287 basic block in the specified function, and always has an i8* type. Taking
Chris Lattnercdfc9402009-11-01 01:27:45 +0000[diff] [blame]2288 the address of the entry block is illegal.</p>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]2289
Chris Lattnerc6f44362009-10-27 21:01:34 +0000[diff] [blame]2290<p>This value only has defined behavior when used as an operand to the
Chris Lattnerab21db72009-10-28 00:19:10 +0000[diff] [blame]2291 '<a href="#i_indirectbr"><tt>indirectbr</tt></a>' instruction or for comparisons
Chris Lattnerc6f44362009-10-27 21:01:34 +0000[diff] [blame]2292 against null. Pointer equality tests between labels addresses is undefined
2293 behavior - though, again, comparison against null is ok, and no label is
Chris Lattnerf9d078e2009-10-27 21:19:13 +0000[diff] [blame]2294 equal to the null pointer. This may also be passed around as an opaque
2295 pointer sized value as long as the bits are not inspected. This allows
Chris Lattner3fd77ce2009-10-27 21:44:20 +0000[diff] [blame]2296 <tt>ptrtoint</tt> and arithmetic to be performed on these values so long as
Chris Lattnerab21db72009-10-28 00:19:10 +0000[diff] [blame]2297 the original value is reconstituted before the <tt>indirectbr</tt>.</p>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]2298
Chris Lattnerf9d078e2009-10-27 21:19:13 +0000[diff] [blame]2299<p>Finally, some targets may provide defined semantics when
Chris Lattnerc6f44362009-10-27 21:01:34 +0000[diff] [blame]2300 using the value as the operand to an inline assembly, but that is target
2301 specific.
2302 </p>
2303
2304</div>
2305
2306
2307<!-- ======================================================================= -->
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2308<div class="doc_subsection"><a name="constantexprs">Constant Expressions</a>
2309</div>
2310
2311<div class="doc_text">
2312
2313<p>Constant expressions are used to allow expressions involving other constants
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2314 to be used as constants. Constant expressions may be of
2315 any <a href="#t_firstclass">first class</a> type and may involve any LLVM
2316 operation that does not have side effects (e.g. load and call are not
2317 supported). The following is the syntax for constant expressions:</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2318
2319<dl>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2320 <dt><b><tt>trunc ( CST to TYPE )</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2321 <dd>Truncate a constant to another type. The bit size of CST must be larger
2322 than the bit size of TYPE. Both types must be integers.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2323
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2324 <dt><b><tt>zext ( CST to TYPE )</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2325 <dd>Zero extend a constant to another type. The bit size of CST must be
2326 smaller or equal to the bit size of TYPE. Both types must be
2327 integers.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2328
2329 <dt><b><tt>sext ( CST to TYPE )</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2330 <dd>Sign extend a constant to another type. The bit size of CST must be
2331 smaller or equal to the bit size of TYPE. Both types must be
2332 integers.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2333
2334 <dt><b><tt>fptrunc ( CST to TYPE )</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2335 <dd>Truncate a floating point constant to another floating point type. The
2336 size of CST must be larger than the size of TYPE. Both types must be
2337 floating point.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2338
2339 <dt><b><tt>fpext ( CST to TYPE )</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2340 <dd>Floating point extend a constant to another type. The size of CST must be
2341 smaller or equal to the size of TYPE. Both types must be floating
2342 point.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2343
Reid Spencer1539a1c2007-07-31 14:40:14 +0000[diff] [blame]2344 <dt><b><tt>fptoui ( CST to TYPE )</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2345 <dd>Convert a floating point constant to the corresponding unsigned integer
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2346 constant. TYPE must be a scalar or vector integer type. CST must be of
2347 scalar or vector floating point type. Both CST and TYPE must be scalars,
2348 or vectors of the same number of elements. If the value won't fit in the
2349 integer type, the results are undefined.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2350
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2351 <dt><b><tt>fptosi ( CST to TYPE )</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2352 <dd>Convert a floating point constant to the corresponding signed integer
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2353 constant. TYPE must be a scalar or vector integer type. CST must be of
2354 scalar or vector floating point type. Both CST and TYPE must be scalars,
2355 or vectors of the same number of elements. If the value won't fit in the
2356 integer type, the results are undefined.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2357
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2358 <dt><b><tt>uitofp ( CST to TYPE )</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2359 <dd>Convert an unsigned integer constant to the corresponding floating point
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2360 constant. TYPE must be a scalar or vector floating point type. CST must be
2361 of scalar or vector integer type. Both CST and TYPE must be scalars, or
2362 vectors of the same number of elements. If the value won't fit in the
2363 floating point type, the results are undefined.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2364
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2365 <dt><b><tt>sitofp ( CST to TYPE )</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2366 <dd>Convert a signed integer constant to the corresponding floating point
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2367 constant. TYPE must be a scalar or vector floating point type. CST must be
2368 of scalar or vector integer type. Both CST and TYPE must be scalars, or
2369 vectors of the same number of elements. If the value won't fit in the
2370 floating point type, the results are undefined.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2371
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]2372 <dt><b><tt>ptrtoint ( CST to TYPE )</tt></b></dt>
2373 <dd>Convert a pointer typed constant to the corresponding integer constant
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2374 <tt>TYPE</tt> must be an integer type. <tt>CST</tt> must be of pointer
2375 type. The <tt>CST</tt> value is zero extended, truncated, or unchanged to
2376 make it fit in <tt>TYPE</tt>.</dd>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]2377
2378 <dt><b><tt>inttoptr ( CST to TYPE )</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2379 <dd>Convert a integer constant to a pointer constant. TYPE must be a pointer
2380 type. CST must be of integer type. The CST value is zero extended,
2381 truncated, or unchanged to make it fit in a pointer size. This one is
2382 <i>really</i> dangerous!</dd>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]2383
2384 <dt><b><tt>bitcast ( CST to TYPE )</tt></b></dt>
Chris Lattner03bbad62009-02-28 18:27:03 +0000[diff] [blame]2385 <dd>Convert a constant, CST, to another TYPE. The constraints of the operands
2386 are the same as those for the <a href="#i_bitcast">bitcast
2387 instruction</a>.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2388
2389 <dt><b><tt>getelementptr ( CSTPTR, IDX0, IDX1, ... )</tt></b></dt>
Dan Gohmandd8004d2009-07-27 21:53:46 +0000[diff] [blame]2390 <dt><b><tt>getelementptr inbounds ( CSTPTR, IDX0, IDX1, ... )</tt></b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2391 <dd>Perform the <a href="#i_getelementptr">getelementptr operation</a> on
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2392 constants. As with the <a href="#i_getelementptr">getelementptr</a>
2393 instruction, the index list may have zero or more indexes, which are
2394 required to make sense for the type of "CSTPTR".</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2395
Robert Bocchino9fbe1452006-01-10 19:31:34 +0000[diff] [blame]2396 <dt><b><tt>select ( COND, VAL1, VAL2 )</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2397 <dd>Perform the <a href="#i_select">select operation</a> on constants.</dd>
Reid Spencer01c42592006-12-04 19:23:19 +0000[diff] [blame]2398
2399 <dt><b><tt>icmp COND ( VAL1, VAL2 )</tt></b></dt>
2400 <dd>Performs the <a href="#i_icmp">icmp operation</a> on constants.</dd>
2401
2402 <dt><b><tt>fcmp COND ( VAL1, VAL2 )</tt></b></dt>
2403 <dd>Performs the <a href="#i_fcmp">fcmp operation</a> on constants.</dd>
Robert Bocchino9fbe1452006-01-10 19:31:34 +0000[diff] [blame]2404
2405 <dt><b><tt>extractelement ( VAL, IDX )</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2406 <dd>Perform the <a href="#i_extractelement">extractelement operation</a> on
2407 constants.</dd>
Robert Bocchino9fbe1452006-01-10 19:31:34 +0000[diff] [blame]2408
Robert Bocchino05ccd702006-01-15 20:48:27 +0000[diff] [blame]2409 <dt><b><tt>insertelement ( VAL, ELT, IDX )</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2410 <dd>Perform the <a href="#i_insertelement">insertelement operation</a> on
2411 constants.</dd>
Chris Lattnerc1989542006-04-08 00:13:41 +0000[diff] [blame]2412
2413 <dt><b><tt>shufflevector ( VEC1, VEC2, IDXMASK )</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2414 <dd>Perform the <a href="#i_shufflevector">shufflevector operation</a> on
2415 constants.</dd>
Chris Lattnerc1989542006-04-08 00:13:41 +0000[diff] [blame]2416
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2417 <dt><b><tt>OPCODE ( LHS, RHS )</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2418 <dd>Perform the specified operation of the LHS and RHS constants. OPCODE may
2419 be any of the <a href="#binaryops">binary</a>
2420 or <a href="#bitwiseops">bitwise binary</a> operations. The constraints
2421 on operands are the same as those for the corresponding instruction
2422 (e.g. no bitwise operations on floating point values are allowed).</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2423</dl>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2424
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2425</div>
Chris Lattner9ee5d222004-03-08 16:49:10 +0000[diff] [blame]2426
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2427<!-- *********************************************************************** -->
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2428<div class="doc_section"> <a name="othervalues">Other Values</a> </div>
2429<!-- *********************************************************************** -->
2430
2431<!-- ======================================================================= -->
2432<div class="doc_subsection">
2433<a name="inlineasm">Inline Assembler Expressions</a>
2434</div>
2435
2436<div class="doc_text">
2437
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2438<p>LLVM supports inline assembler expressions (as opposed
2439 to <a href="#moduleasm"> Module-Level Inline Assembly</a>) through the use of
2440 a special value. This value represents the inline assembler as a string
2441 (containing the instructions to emit), a list of operand constraints (stored
Dale Johannesen09fed252009-10-13 21:56:55 +0000[diff] [blame]2442 as a string), a flag that indicates whether or not the inline asm
Dale Johannesen8ba2d5b2009-10-21 23:28:00 +0000[diff] [blame]2443 expression has side effects, and a flag indicating whether the function
2444 containing the asm needs to align its stack conservatively. An example
2445 inline assembler expression is:</p>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2446
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2447<div class="doc_code">
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2448<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2449i32 (i32) asm "bswap $0", "=r,r"
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2450</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2451</div>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2452
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2453<p>Inline assembler expressions may <b>only</b> be used as the callee operand of
2454 a <a href="#i_call"><tt>call</tt> instruction</a>. Thus, typically we
2455 have:</p>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2456
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2457<div class="doc_code">
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2458<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2459%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]2460</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2461</div>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2462
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2463<p>Inline asms with side effects not visible in the constraint list must be
2464 marked as having side effects. This is done through the use of the
2465 '<tt>sideeffect</tt>' keyword, like so:</p>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2466
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2467<div class="doc_code">
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2468<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2469call void asm sideeffect "eieio", ""()
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2470</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2471</div>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2472
Dale Johannesen8ba2d5b2009-10-21 23:28:00 +0000[diff] [blame]2473<p>In some cases inline asms will contain code that will not work unless the
2474 stack is aligned in some way, such as calls or SSE instructions on x86,
2475 yet will not contain code that does that alignment within the asm.
2476 The compiler should make conservative assumptions about what the asm might
2477 contain and should generate its usual stack alignment code in the prologue
2478 if the '<tt>alignstack</tt>' keyword is present:</p>
Dale Johannesen09fed252009-10-13 21:56:55 +0000[diff] [blame]2479
2480<div class="doc_code">
2481<pre>
Dale Johannesen8ba2d5b2009-10-21 23:28:00 +0000[diff] [blame]2482call void asm alignstack "eieio", ""()
Dale Johannesen09fed252009-10-13 21:56:55 +0000[diff] [blame]2483</pre>
2484</div>
2485
2486<p>If both keywords appear the '<tt>sideeffect</tt>' keyword must come
2487 first.</p>
2488
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2489<p>TODO: The format of the asm and constraints string still need to be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2490 documented here. Constraints on what can be done (e.g. duplication, moving,
2491 etc need to be documented). This is probably best done by reference to
2492 another document that covers inline asm from a holistic perspective.</p>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2493
2494</div>
2495
Chris Lattnere6a5ddd2010-01-15 21:50:19 +0000[diff] [blame]2496<!-- ======================================================================= -->
2497<div class="doc_subsection"><a name="metadata">Metadata Nodes and Metadata
2498 Strings</a>
2499</div>
2500
2501<div class="doc_text">
2502
2503<p>LLVM IR allows metadata to be attached to instructions in the program that
2504 can convey extra information about the code to the optimizers and code
2505 generator. One example application of metadata is source-level debug
2506 information. There are two metadata primitives: strings and nodes. All
2507 metadata has the <tt>metadata</tt> type and is identified in syntax by a
2508 preceding exclamation point ('<tt>!</tt>').</p>
2509
2510<p>A metadata string is a string surrounded by double quotes. It can contain
2511 any character by escaping non-printable characters with "\xx" where "xx" is
2512 the two digit hex code. For example: "<tt>!"test\00"</tt>".</p>
2513
2514<p>Metadata nodes are represented with notation similar to structure constants
2515 (a comma separated list of elements, surrounded by braces and preceded by an
2516 exclamation point). For example: "<tt>!{ metadata !"test\00", i32
2517 10}</tt>". Metadata nodes can have any values as their operand.</p>
2518
2519<p>A <a href="#namedmetadatastructure">named metadata</a> is a collection of
2520 metadata nodes, which can be looked up in the module symbol table. For
2521 example: "<tt>!foo = metadata !{!4, !3}</tt>".
2522
Devang Patele1d50cd2010-03-04 23:44:48 +0000[diff] [blame]2523<p>Metadata can be used as function arguments. Here <tt>llvm.dbg.value</tt>
2524 function is using two metadata arguments.
2525
2526 <div class="doc_code">
2527 <pre>
2528 call void @llvm.dbg.value(metadata !24, i64 0, metadata !25)
2529 </pre>
2530 </div></p>
2531
2532<p>Metadata can be attached with an instruction. Here metadata <tt>!21</tt> is
2533 attached with <tt>add</tt> instruction using <tt>!dbg</tt> identifier.
2534
2535 <div class="doc_code">
2536 <pre>
2537 %indvar.next = add i64 %indvar, 1, !dbg !21
2538 </pre>
2539 </div></p>
Chris Lattnere6a5ddd2010-01-15 21:50:19 +0000[diff] [blame]2540</div>
2541
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2542
2543<!-- *********************************************************************** -->
2544<div class="doc_section">
2545 <a name="intrinsic_globals">Intrinsic Global Variables</a>
2546</div>
2547<!-- *********************************************************************** -->
2548
2549<p>LLVM has a number of "magic" global variables that contain data that affect
2550code generation or other IR semantics. These are documented here. All globals
Chris Lattner401e10c2009-07-20 06:14:25 +0000[diff] [blame]2551of this sort should have a section specified as "<tt>llvm.metadata</tt>". This
2552section and all globals that start with "<tt>llvm.</tt>" are reserved for use
2553by LLVM.</p>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2554
2555<!-- ======================================================================= -->
2556<div class="doc_subsection">
2557<a name="intg_used">The '<tt>llvm.used</tt>' Global Variable</a>
2558</div>
2559
2560<div class="doc_text">
2561
2562<p>The <tt>@llvm.used</tt> global is an array with i8* element type which has <a
2563href="#linkage_appending">appending linkage</a>. This array contains a list of
2564pointers to global variables and functions which may optionally have a pointer
2565cast formed of bitcast or getelementptr. For example, a legal use of it is:</p>
2566
2567<pre>
2568 @X = global i8 4
2569 @Y = global i32 123
2570
2571 @llvm.used = appending global [2 x i8*] [
2572 i8* @X,
2573 i8* bitcast (i32* @Y to i8*)
2574 ], section "llvm.metadata"
2575</pre>
2576
2577<p>If a global variable appears in the <tt>@llvm.used</tt> list, then the
2578compiler, assembler, and linker are required to treat the symbol as if there is
2579a reference to the global that it cannot see. For example, if a variable has
2580internal linkage and no references other than that from the <tt>@llvm.used</tt>
2581list, it cannot be deleted. This is commonly used to represent references from
2582inline asms and other things the compiler cannot "see", and corresponds to
2583"attribute((used))" in GNU C.</p>
2584
2585<p>On some targets, the code generator must emit a directive to the assembler or
2586object file to prevent the assembler and linker from molesting the symbol.</p>
2587
2588</div>
2589
2590<!-- ======================================================================= -->
2591<div class="doc_subsection">
Chris Lattner401e10c2009-07-20 06:14:25 +0000[diff] [blame]2592<a name="intg_compiler_used">The '<tt>llvm.compiler.used</tt>' Global Variable</a>
2593</div>
2594
2595<div class="doc_text">
2596
2597<p>The <tt>@llvm.compiler.used</tt> directive is the same as the
2598<tt>@llvm.used</tt> directive, except that it only prevents the compiler from
2599touching the symbol. On targets that support it, this allows an intelligent
2600linker to optimize references to the symbol without being impeded as it would be
2601by <tt>@llvm.used</tt>.</p>
2602
2603<p>This is a rare construct that should only be used in rare circumstances, and
2604should not be exposed to source languages.</p>
2605
2606</div>
2607
2608<!-- ======================================================================= -->
2609<div class="doc_subsection">
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2610<a name="intg_global_ctors">The '<tt>llvm.global_ctors</tt>' Global Variable</a>
2611</div>
2612
2613<div class="doc_text">
2614
2615<p>TODO: Describe this.</p>
2616
2617</div>
2618
2619<!-- ======================================================================= -->
2620<div class="doc_subsection">
2621<a name="intg_global_dtors">The '<tt>llvm.global_dtors</tt>' Global Variable</a>
2622</div>
2623
2624<div class="doc_text">
2625
2626<p>TODO: Describe this.</p>
2627
2628</div>
2629
2630
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2631<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2632<div class="doc_section"> <a name="instref">Instruction Reference</a> </div>
2633<!-- *********************************************************************** -->
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2634
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2635<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2636
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2637<p>The LLVM instruction set consists of several different classifications of
2638 instructions: <a href="#terminators">terminator
2639 instructions</a>, <a href="#binaryops">binary instructions</a>,
2640 <a href="#bitwiseops">bitwise binary instructions</a>,
2641 <a href="#memoryops">memory instructions</a>, and
2642 <a href="#otherops">other instructions</a>.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2643
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2644</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2645
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2646<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2647<div class="doc_subsection"> <a name="terminators">Terminator
2648Instructions</a> </div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2649
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2650<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2651
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2652<p>As mentioned <a href="#functionstructure">previously</a>, every basic block
2653 in a program ends with a "Terminator" instruction, which indicates which
2654 block should be executed after the current block is finished. These
2655 terminator instructions typically yield a '<tt>void</tt>' value: they produce
2656 control flow, not values (the one exception being the
2657 '<a href="#i_invoke"><tt>invoke</tt></a>' instruction).</p>
2658
2659<p>There are six different terminator instructions: the
2660 '<a href="#i_ret"><tt>ret</tt></a>' instruction, the
2661 '<a href="#i_br"><tt>br</tt></a>' instruction, the
2662 '<a href="#i_switch"><tt>switch</tt></a>' instruction, the
Bill Wendling21c346e2009-11-02 00:25:26 +0000[diff] [blame]2663 '<a href="#i_indirectbr">'<tt>indirectbr</tt></a>' Instruction, the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2664 '<a href="#i_invoke"><tt>invoke</tt></a>' instruction, the
2665 '<a href="#i_unwind"><tt>unwind</tt></a>' instruction, and the
2666 '<a href="#i_unreachable"><tt>unreachable</tt></a>' instruction.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2667
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2668</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2669
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2670<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2671<div class="doc_subsubsection"> <a name="i_ret">'<tt>ret</tt>'
2672Instruction</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2673
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2674<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2675
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2676<h5>Syntax:</h5>
Dan Gohmanb1e6b962008-10-04 19:00:07 +0000[diff] [blame]2677<pre>
2678 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]2679 ret void <i>; Return from void function</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2680</pre>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]2681
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2682<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2683<p>The '<tt>ret</tt>' instruction is used to return control flow (and optionally
2684 a value) from a function back to the caller.</p>
2685
2686<p>There are two forms of the '<tt>ret</tt>' instruction: one that returns a
2687 value and then causes control flow, and one that just causes control flow to
2688 occur.</p>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]2689
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2690<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2691<p>The '<tt>ret</tt>' instruction optionally accepts a single argument, the
2692 return value. The type of the return value must be a
2693 '<a href="#t_firstclass">first class</a>' type.</p>
Dan Gohmanb1e6b962008-10-04 19:00:07 +0000[diff] [blame]2694
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2695<p>A function is not <a href="#wellformed">well formed</a> if it it has a
2696 non-void return type and contains a '<tt>ret</tt>' instruction with no return
2697 value or a return value with a type that does not match its type, or if it
2698 has a void return type and contains a '<tt>ret</tt>' instruction with a
2699 return value.</p>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]2700
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2701<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2702<p>When the '<tt>ret</tt>' instruction is executed, control flow returns back to
2703 the calling function's context. If the caller is a
2704 "<a href="#i_call"><tt>call</tt></a>" instruction, execution continues at the
2705 instruction after the call. If the caller was an
2706 "<a href="#i_invoke"><tt>invoke</tt></a>" instruction, execution continues at
2707 the beginning of the "normal" destination block. If the instruction returns
2708 a value, that value shall set the call or invoke instruction's return
2709 value.</p>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]2710
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2711<h5>Example:</h5>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]2712<pre>
2713 ret i32 5 <i>; Return an integer value of 5</i>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2714 ret void <i>; Return from a void function</i>
Bill Wendling0a4bbbf2009-02-28 22:12:54 +0000[diff] [blame]2715 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]2716</pre>
Dan Gohmanf3e60bd2009-01-12 23:12:39 +0000[diff] [blame]2717
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2718</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2719<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2720<div class="doc_subsubsection"> <a name="i_br">'<tt>br</tt>' Instruction</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2721
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2722<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2723
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2724<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2725<pre>
2726 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]2727</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2728
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2729<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2730<p>The '<tt>br</tt>' instruction is used to cause control flow to transfer to a
2731 different basic block in the current function. There are two forms of this
2732 instruction, corresponding to a conditional branch and an unconditional
2733 branch.</p>
2734
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2735<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2736<p>The conditional branch form of the '<tt>br</tt>' instruction takes a single
2737 '<tt>i1</tt>' value and two '<tt>label</tt>' values. The unconditional form
2738 of the '<tt>br</tt>' instruction takes a single '<tt>label</tt>' value as a
2739 target.</p>
2740
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2741<h5>Semantics:</h5>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]2742<p>Upon execution of a conditional '<tt>br</tt>' instruction, the '<tt>i1</tt>'
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2743 argument is evaluated. If the value is <tt>true</tt>, control flows to the
2744 '<tt>iftrue</tt>' <tt>label</tt> argument. If "cond" is <tt>false</tt>,
2745 control flows to the '<tt>iffalse</tt>' <tt>label</tt> argument.</p>
2746
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2747<h5>Example:</h5>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]2748<pre>
2749Test:
2750 %cond = <a href="#i_icmp">icmp</a> eq i32 %a, %b
2751 br i1 %cond, label %IfEqual, label %IfUnequal
2752IfEqual:
2753 <a href="#i_ret">ret</a> i32 1
2754IfUnequal:
2755 <a href="#i_ret">ret</a> i32 0
2756</pre>
2757
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2758</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2759
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2760<!-- _______________________________________________________________________ -->
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2761<div class="doc_subsubsection">
2762 <a name="i_switch">'<tt>switch</tt>' Instruction</a>
2763</div>
2764
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2765<div class="doc_text">
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2766
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2767<h5>Syntax:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2768<pre>
2769 switch &lt;intty&gt; &lt;value&gt;, label &lt;defaultdest&gt; [ &lt;intty&gt; &lt;val&gt;, label &lt;dest&gt; ... ]
2770</pre>
2771
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2772<h5>Overview:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2773<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]2774 several different places. It is a generalization of the '<tt>br</tt>'
2775 instruction, allowing a branch to occur to one of many possible
2776 destinations.</p>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2777
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2778<h5>Arguments:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2779<p>The '<tt>switch</tt>' instruction uses three parameters: an integer
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2780 comparison value '<tt>value</tt>', a default '<tt>label</tt>' destination,
2781 and an array of pairs of comparison value constants and '<tt>label</tt>'s.
2782 The table is not allowed to contain duplicate constant entries.</p>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2783
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2784<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2785<p>The <tt>switch</tt> instruction specifies a table of values and
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2786 destinations. When the '<tt>switch</tt>' instruction is executed, this table
2787 is searched for the given value. If the value is found, control flow is
Benjamin Kramer8040cd32009-10-12 14:46:08 +0000[diff] [blame]2788 transferred to the corresponding destination; otherwise, control flow is
2789 transferred to the default destination.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2790
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2791<h5>Implementation:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2792<p>Depending on properties of the target machine and the particular
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2793 <tt>switch</tt> instruction, this instruction may be code generated in
2794 different ways. For example, it could be generated as a series of chained
2795 conditional branches or with a lookup table.</p>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2796
2797<h5>Example:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2798<pre>
2799 <i>; Emulate a conditional br instruction</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]2800 %Val = <a href="#i_zext">zext</a> i1 %value to i32
Dan Gohman2a08c532009-01-04 23:44:43 +0000[diff] [blame]2801 switch i32 %Val, label %truedest [ i32 0, label %falsedest ]
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2802
2803 <i>; Emulate an unconditional br instruction</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2804 switch i32 0, label %dest [ ]
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2805
2806 <i>; Implement a jump table:</i>
Dan Gohman2a08c532009-01-04 23:44:43 +0000[diff] [blame]2807 switch i32 %val, label %otherwise [ i32 0, label %onzero
2808 i32 1, label %onone
2809 i32 2, label %ontwo ]
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2810</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2811
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2812</div>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2813
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]2814
2815<!-- _______________________________________________________________________ -->
2816<div class="doc_subsubsection">
Chris Lattnerab21db72009-10-28 00:19:10 +0000[diff] [blame]2817 <a name="i_indirectbr">'<tt>indirectbr</tt>' Instruction</a>
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]2818</div>
2819
2820<div class="doc_text">
2821
2822<h5>Syntax:</h5>
2823<pre>
Chris Lattnerab21db72009-10-28 00:19:10 +0000[diff] [blame]2824 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]2825</pre>
2826
2827<h5>Overview:</h5>
2828
Chris Lattnerab21db72009-10-28 00:19:10 +0000[diff] [blame]2829<p>The '<tt>indirectbr</tt>' instruction implements an indirect branch to a label
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]2830 within the current function, whose address is specified by
Chris Lattnerc6f44362009-10-27 21:01:34 +0000[diff] [blame]2831 "<tt>address</tt>". Address must be derived from a <a
2832 href="#blockaddress">blockaddress</a> constant.</p>
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]2833
2834<h5>Arguments:</h5>
2835
2836<p>The '<tt>address</tt>' argument is the address of the label to jump to. The
2837 rest of the arguments indicate the full set of possible destinations that the
2838 address may point to. Blocks are allowed to occur multiple times in the
2839 destination list, though this isn't particularly useful.</p>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]2840
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]2841<p>This destination list is required so that dataflow analysis has an accurate
2842 understanding of the CFG.</p>
2843
2844<h5>Semantics:</h5>
2845
2846<p>Control transfers to the block specified in the address argument. All
2847 possible destination blocks must be listed in the label list, otherwise this
2848 instruction has undefined behavior. This implies that jumps to labels
2849 defined in other functions have undefined behavior as well.</p>
2850
2851<h5>Implementation:</h5>
2852
2853<p>This is typically implemented with a jump through a register.</p>
2854
2855<h5>Example:</h5>
2856<pre>
Chris Lattnerab21db72009-10-28 00:19:10 +0000[diff] [blame]2857 indirectbr i8* %Addr, [ label %bb1, label %bb2, label %bb3 ]
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]2858</pre>
2859
2860</div>
2861
2862
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2863<!-- _______________________________________________________________________ -->
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2864<div class="doc_subsubsection">
2865 <a name="i_invoke">'<tt>invoke</tt>' Instruction</a>
2866</div>
2867
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2868<div class="doc_text">
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2869
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2870<h5>Syntax:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2871<pre>
Devang Patel307e8ab2008-10-07 17:48:33 +0000[diff] [blame]2872 &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]2873 to label &lt;normal label&gt; unwind label &lt;exception label&gt;
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2874</pre>
2875
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]2876<h5>Overview:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2877<p>The '<tt>invoke</tt>' instruction causes control to transfer to a specified
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2878 function, with the possibility of control flow transfer to either the
2879 '<tt>normal</tt>' label or the '<tt>exception</tt>' label. If the callee
2880 function returns with the "<tt><a href="#i_ret">ret</a></tt>" instruction,
2881 control flow will return to the "normal" label. If the callee (or any
2882 indirect callees) returns with the "<a href="#i_unwind"><tt>unwind</tt></a>"
2883 instruction, control is interrupted and continued at the dynamically nearest
2884 "exception" label.</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2885
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2886<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2887<p>This instruction requires several arguments:</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2888
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2889<ol>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2890 <li>The optional "cconv" marker indicates which <a href="#callingconv">calling
2891 convention</a> the call should use. If none is specified, the call
2892 defaults to using C calling conventions.</li>
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]2893
2894 <li>The optional <a href="#paramattrs">Parameter Attributes</a> list for
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2895 return values. Only '<tt>zeroext</tt>', '<tt>signext</tt>', and
2896 '<tt>inreg</tt>' attributes are valid here.</li>
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]2897
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2898 <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]2899 function value being invoked. In most cases, this is a direct function
2900 invocation, but indirect <tt>invoke</tt>s are just as possible, branching
2901 off an arbitrary pointer to function value.</li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2902
2903 <li>'<tt>function ptr val</tt>': An LLVM value containing a pointer to a
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2904 function to be invoked. </li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2905
2906 <li>'<tt>function args</tt>': argument list whose types match the function
Chris Lattner0724fbd2010-03-02 06:36:51 +0000[diff] [blame]2907 signature argument types and parameter attributes. All arguments must be
2908 of <a href="#t_firstclass">first class</a> type. If the function
2909 signature indicates the function accepts a variable number of arguments,
2910 the extra arguments can be specified.</li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2911
2912 <li>'<tt>normal label</tt>': the label reached when the called function
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2913 executes a '<tt><a href="#i_ret">ret</a></tt>' instruction. </li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2914
2915 <li>'<tt>exception label</tt>': the label reached when a callee returns with
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2916 the <a href="#i_unwind"><tt>unwind</tt></a> instruction. </li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2917
Devang Patel307e8ab2008-10-07 17:48:33 +0000[diff] [blame]2918 <li>The optional <a href="#fnattrs">function attributes</a> list. Only
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2919 '<tt>noreturn</tt>', '<tt>nounwind</tt>', '<tt>readonly</tt>' and
2920 '<tt>readnone</tt>' attributes are valid here.</li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2921</ol>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2922
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2923<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2924<p>This instruction is designed to operate as a standard
2925 '<tt><a href="#i_call">call</a></tt>' instruction in most regards. The
2926 primary difference is that it establishes an association with a label, which
2927 is used by the runtime library to unwind the stack.</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2928
2929<p>This instruction is used in languages with destructors to ensure that proper
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2930 cleanup is performed in the case of either a <tt>longjmp</tt> or a thrown
2931 exception. Additionally, this is important for implementation of
2932 '<tt>catch</tt>' clauses in high-level languages that support them.</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2933
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2934<p>For the purposes of the SSA form, the definition of the value returned by the
2935 '<tt>invoke</tt>' instruction is deemed to occur on the edge from the current
2936 block to the "normal" label. If the callee unwinds then no return value is
2937 available.</p>
Dan Gohmanf96a4992009-05-22 21:47:08 +0000[diff] [blame]2938
Chris Lattnerdf7a6802010-01-15 18:08:37 +0000[diff] [blame]2939<p>Note that the code generator does not yet completely support unwind, and
2940that the invoke/unwind semantics are likely to change in future versions.</p>
2941
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2942<h5>Example:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2943<pre>
Nick Lewyckyd703f652008-03-16 07:18:12 +0000[diff] [blame]2944 %retval = invoke i32 @Test(i32 15) to label %Continue
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]2945 unwind label %TestCleanup <i>; {i32}:retval set</i>
Nick Lewyckyd703f652008-03-16 07:18:12 +0000[diff] [blame]2946 %retval = invoke <a href="#callingconv">coldcc</a> i32 %Testfnptr(i32 15) to label %Continue
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]2947 unwind label %TestCleanup <i>; {i32}:retval set</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2948</pre>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]2949
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2950</div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]2951
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]2952<!-- _______________________________________________________________________ -->
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]2953
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2954<div class="doc_subsubsection"> <a name="i_unwind">'<tt>unwind</tt>'
2955Instruction</a> </div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]2956
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2957<div class="doc_text">
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]2958
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]2959<h5>Syntax:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]2960<pre>
2961 unwind
2962</pre>
2963
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]2964<h5>Overview:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]2965<p>The '<tt>unwind</tt>' instruction unwinds the stack, continuing control flow
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2966 at the first callee in the dynamic call stack which used
2967 an <a href="#i_invoke"><tt>invoke</tt></a> instruction to perform the call.
2968 This is primarily used to implement exception handling.</p>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]2969
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]2970<h5>Semantics:</h5>
Chris Lattner72ed2002008-04-19 21:01:16 +0000[diff] [blame]2971<p>The '<tt>unwind</tt>' instruction causes execution of the current function to
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2972 immediately halt. The dynamic call stack is then searched for the
2973 first <a href="#i_invoke"><tt>invoke</tt></a> instruction on the call stack.
2974 Once found, execution continues at the "exceptional" destination block
2975 specified by the <tt>invoke</tt> instruction. If there is no <tt>invoke</tt>
2976 instruction in the dynamic call chain, undefined behavior results.</p>
2977
Chris Lattnerdf7a6802010-01-15 18:08:37 +0000[diff] [blame]2978<p>Note that the code generator does not yet completely support unwind, and
2979that the invoke/unwind semantics are likely to change in future versions.</p>
2980
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2981</div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]2982
2983<!-- _______________________________________________________________________ -->
2984
2985<div class="doc_subsubsection"> <a name="i_unreachable">'<tt>unreachable</tt>'
2986Instruction</a> </div>
2987
2988<div class="doc_text">
2989
2990<h5>Syntax:</h5>
2991<pre>
2992 unreachable
2993</pre>
2994
2995<h5>Overview:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]2996<p>The '<tt>unreachable</tt>' instruction has no defined semantics. This
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2997 instruction is used to inform the optimizer that a particular portion of the
2998 code is not reachable. This can be used to indicate that the code after a
2999 no-return function cannot be reached, and other facts.</p>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3000
3001<h5>Semantics:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3002<p>The '<tt>unreachable</tt>' instruction has no defined semantics.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3003
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3004</div>
3005
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3006<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3007<div class="doc_subsection"> <a name="binaryops">Binary Operations</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3008
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3009<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3010
3011<p>Binary operators are used to do most of the computation in a program. They
3012 require two operands of the same type, execute an operation on them, and
3013 produce a single value. The operands might represent multiple data, as is
3014 the case with the <a href="#t_vector">vector</a> data type. The result value
3015 has the same type as its operands.</p>
3016
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3017<p>There are several different binary operators:</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3018
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3019</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3020
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3021<!-- _______________________________________________________________________ -->
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3022<div class="doc_subsubsection">
3023 <a name="i_add">'<tt>add</tt>' Instruction</a>
3024</div>
3025
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3026<div class="doc_text">
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3027
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3028<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3029<pre>
Dan Gohmancbb38f22009-07-20 22:41:19 +0000[diff] [blame]3030 &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]3031 &lt;result&gt; = add nuw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3032 &lt;result&gt; = add nsw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3033 &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]3034</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3035
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3036<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3037<p>The '<tt>add</tt>' instruction returns the sum of its two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3038
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3039<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3040<p>The two arguments to the '<tt>add</tt>' instruction must
3041 be <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of
3042 integer values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3043
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3044<h5>Semantics:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3045<p>The value produced is the integer sum of the two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3046
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3047<p>If the sum has unsigned overflow, the result returned is the mathematical
3048 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]3049
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3050<p>Because LLVM integers use a two's complement representation, this instruction
3051 is appropriate for both signed and unsigned integers.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3052
Dan Gohman08d012e2009-07-22 22:44:56 +0000[diff] [blame]3053<p><tt>nuw</tt> and <tt>nsw</tt> stand for &quot;No Unsigned Wrap&quot;
3054 and &quot;No Signed Wrap&quot;, respectively. If the <tt>nuw</tt> and/or
3055 <tt>nsw</tt> keywords are present, the result value of the <tt>add</tt>
3056 is undefined if unsigned and/or signed overflow, respectively, occurs.</p>
Dan Gohmancbb38f22009-07-20 22:41:19 +0000[diff] [blame]3057
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3058<h5>Example:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3059<pre>
3060 &lt;result&gt; = add i32 4, %var <i>; yields {i32}:result = 4 + %var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3061</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3062
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3063</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3064
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3065<!-- _______________________________________________________________________ -->
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3066<div class="doc_subsubsection">
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3067 <a name="i_fadd">'<tt>fadd</tt>' Instruction</a>
3068</div>
3069
3070<div class="doc_text">
3071
3072<h5>Syntax:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3073<pre>
3074 &lt;result&gt; = fadd &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3075</pre>
3076
3077<h5>Overview:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3078<p>The '<tt>fadd</tt>' instruction returns the sum of its two operands.</p>
3079
3080<h5>Arguments:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3081<p>The two arguments to the '<tt>fadd</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3082 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
3083 floating point values. Both arguments must have identical types.</p>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3084
3085<h5>Semantics:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3086<p>The value produced is the floating point sum of the two operands.</p>
3087
3088<h5>Example:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3089<pre>
3090 &lt;result&gt; = fadd float 4.0, %var <i>; yields {float}:result = 4.0 + %var</i>
3091</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3092
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3093</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3094
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3095<!-- _______________________________________________________________________ -->
3096<div class="doc_subsubsection">
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3097 <a name="i_sub">'<tt>sub</tt>' Instruction</a>
3098</div>
3099
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3100<div class="doc_text">
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3101
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3102<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3103<pre>
Dan Gohman08d012e2009-07-22 22:44:56 +0000[diff] [blame]3104 &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]3105 &lt;result&gt; = sub nuw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3106 &lt;result&gt; = sub nsw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3107 &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]3108</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3109
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3110<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3111<p>The '<tt>sub</tt>' instruction returns the difference of its two
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3112 operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3113
3114<p>Note that the '<tt>sub</tt>' instruction is used to represent the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3115 '<tt>neg</tt>' instruction present in most other intermediate
3116 representations.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3117
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3118<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3119<p>The two arguments to the '<tt>sub</tt>' instruction must
3120 be <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of
3121 integer values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3122
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3123<h5>Semantics:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3124<p>The value produced is the integer difference of the two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3125
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3126<p>If the difference has unsigned overflow, the result returned is the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3127 mathematical result modulo 2<sup>n</sup>, where n is the bit width of the
3128 result.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3129
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3130<p>Because LLVM integers use a two's complement representation, this instruction
3131 is appropriate for both signed and unsigned integers.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3132
Dan Gohman08d012e2009-07-22 22:44:56 +0000[diff] [blame]3133<p><tt>nuw</tt> and <tt>nsw</tt> stand for &quot;No Unsigned Wrap&quot;
3134 and &quot;No Signed Wrap&quot;, respectively. If the <tt>nuw</tt> and/or
3135 <tt>nsw</tt> keywords are present, the result value of the <tt>sub</tt>
3136 is undefined if unsigned and/or signed overflow, respectively, occurs.</p>
Dan Gohmancbb38f22009-07-20 22:41:19 +0000[diff] [blame]3137
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3138<h5>Example:</h5>
Bill Wendlingaac388b2007-05-29 09:42:13 +0000[diff] [blame]3139<pre>
3140 &lt;result&gt; = sub i32 4, %var <i>; yields {i32}:result = 4 - %var</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3141 &lt;result&gt; = sub i32 0, %val <i>; yields {i32}:result = -%var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3142</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3143
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3144</div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3145
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3146<!-- _______________________________________________________________________ -->
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3147<div class="doc_subsubsection">
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3148 <a name="i_fsub">'<tt>fsub</tt>' Instruction</a>
3149</div>
3150
3151<div class="doc_text">
3152
3153<h5>Syntax:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3154<pre>
3155 &lt;result&gt; = fsub &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3156</pre>
3157
3158<h5>Overview:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3159<p>The '<tt>fsub</tt>' instruction returns the difference of its two
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3160 operands.</p>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3161
3162<p>Note that the '<tt>fsub</tt>' instruction is used to represent the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3163 '<tt>fneg</tt>' instruction present in most other intermediate
3164 representations.</p>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3165
3166<h5>Arguments:</h5>
Bill Wendlingd9fe2982009-07-20 02:32:41 +0000[diff] [blame]3167<p>The two arguments to the '<tt>fsub</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3168 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
3169 floating point values. Both arguments must have identical types.</p>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3170
3171<h5>Semantics:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3172<p>The value produced is the floating point difference of the two operands.</p>
3173
3174<h5>Example:</h5>
3175<pre>
3176 &lt;result&gt; = fsub float 4.0, %var <i>; yields {float}:result = 4.0 - %var</i>
3177 &lt;result&gt; = fsub float -0.0, %val <i>; yields {float}:result = -%var</i>
3178</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3179
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3180</div>
3181
3182<!-- _______________________________________________________________________ -->
3183<div class="doc_subsubsection">
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3184 <a name="i_mul">'<tt>mul</tt>' Instruction</a>
3185</div>
3186
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3187<div class="doc_text">
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3188
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3189<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3190<pre>
Dan Gohman08d012e2009-07-22 22:44:56 +0000[diff] [blame]3191 &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]3192 &lt;result&gt; = mul nuw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3193 &lt;result&gt; = mul nsw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3194 &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]3195</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3196
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3197<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3198<p>The '<tt>mul</tt>' instruction returns the product of its two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3199
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3200<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3201<p>The two arguments to the '<tt>mul</tt>' instruction must
3202 be <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of
3203 integer values. Both arguments must have identical types.</p>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3204
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3205<h5>Semantics:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3206<p>The value produced is the integer product of the two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3207
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3208<p>If the result of the multiplication has unsigned overflow, the result
3209 returned is the mathematical result modulo 2<sup>n</sup>, where n is the bit
3210 width of the result.</p>
3211
3212<p>Because LLVM integers use a two's complement representation, and the result
3213 is the same width as the operands, this instruction returns the correct
3214 result for both signed and unsigned integers. If a full product
3215 (e.g. <tt>i32</tt>x<tt>i32</tt>-><tt>i64</tt>) is needed, the operands should
3216 be sign-extended or zero-extended as appropriate to the width of the full
3217 product.</p>
3218
Dan Gohman08d012e2009-07-22 22:44:56 +0000[diff] [blame]3219<p><tt>nuw</tt> and <tt>nsw</tt> stand for &quot;No Unsigned Wrap&quot;
3220 and &quot;No Signed Wrap&quot;, respectively. If the <tt>nuw</tt> and/or
3221 <tt>nsw</tt> keywords are present, the result value of the <tt>mul</tt>
3222 is undefined if unsigned and/or signed overflow, respectively, occurs.</p>
Dan Gohmancbb38f22009-07-20 22:41:19 +0000[diff] [blame]3223
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3224<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3225<pre>
3226 &lt;result&gt; = mul i32 4, %var <i>; yields {i32}:result = 4 * %var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3227</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3228
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3229</div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3230
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3231<!-- _______________________________________________________________________ -->
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3232<div class="doc_subsubsection">
3233 <a name="i_fmul">'<tt>fmul</tt>' Instruction</a>
3234</div>
3235
3236<div class="doc_text">
3237
3238<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3239<pre>
3240 &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]3241</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3242
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3243<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3244<p>The '<tt>fmul</tt>' instruction returns the product of its two operands.</p>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3245
3246<h5>Arguments:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3247<p>The two arguments to the '<tt>fmul</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3248 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
3249 floating point values. Both arguments must have identical types.</p>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3250
3251<h5>Semantics:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3252<p>The value produced is the floating point product of the two operands.</p>
3253
3254<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3255<pre>
3256 &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]3257</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3258
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3259</div>
3260
3261<!-- _______________________________________________________________________ -->
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3262<div class="doc_subsubsection"> <a name="i_udiv">'<tt>udiv</tt>' Instruction
3263</a></div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3264
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3265<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3266
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3267<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3268<pre>
3269 &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]3270</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3271
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3272<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3273<p>The '<tt>udiv</tt>' instruction returns the quotient of its two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3274
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3275<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3276<p>The two arguments to the '<tt>udiv</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3277 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3278 values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3279
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3280<h5>Semantics:</h5>
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3281<p>The value produced is the unsigned integer quotient of the two operands.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3282
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3283<p>Note that unsigned integer division and signed integer division are distinct
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3284 operations; for signed integer division, use '<tt>sdiv</tt>'.</p>
3285
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3286<p>Division by zero leads to undefined behavior.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3287
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3288<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3289<pre>
3290 &lt;result&gt; = udiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3291</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3292
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3293</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3294
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3295<!-- _______________________________________________________________________ -->
3296<div class="doc_subsubsection"> <a name="i_sdiv">'<tt>sdiv</tt>' Instruction
3297</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3298
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3299<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3300
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3301<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3302<pre>
Dan Gohmancbb38f22009-07-20 22:41:19 +0000[diff] [blame]3303 &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]3304 &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]3305</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3306
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3307<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3308<p>The '<tt>sdiv</tt>' instruction returns the quotient of its two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3309
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3310<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3311<p>The two arguments to the '<tt>sdiv</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3312 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3313 values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3314
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3315<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3316<p>The value produced is the signed integer quotient of the two operands rounded
3317 towards zero.</p>
3318
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3319<p>Note that signed integer division and unsigned integer division are distinct
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3320 operations; for unsigned integer division, use '<tt>udiv</tt>'.</p>
3321
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3322<p>Division by zero leads to undefined behavior. Overflow also leads to
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3323 undefined behavior; this is a rare case, but can occur, for example, by doing
3324 a 32-bit division of -2147483648 by -1.</p>
3325
Dan Gohman9c5beed2009-07-22 00:04:19 +0000[diff] [blame]3326<p>If the <tt>exact</tt> keyword is present, the result value of the
3327 <tt>sdiv</tt> is undefined if the result would be rounded or if overflow
3328 would occur.</p>
Dan Gohmancbb38f22009-07-20 22:41:19 +0000[diff] [blame]3329
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3330<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3331<pre>
3332 &lt;result&gt; = sdiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3333</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3334
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3335</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3336
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3337<!-- _______________________________________________________________________ -->
3338<div class="doc_subsubsection"> <a name="i_fdiv">'<tt>fdiv</tt>'
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3339Instruction</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3340
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3341<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3342
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3343<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3344<pre>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]3345 &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]3346</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3347
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3348<h5>Overview:</h5>
3349<p>The '<tt>fdiv</tt>' instruction returns the quotient of its two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3350
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3351<h5>Arguments:</h5>
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]3352<p>The two arguments to the '<tt>fdiv</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3353 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
3354 floating point values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3355
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3356<h5>Semantics:</h5>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3357<p>The value produced is the floating point quotient of the two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3358
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3359<h5>Example:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3360<pre>
3361 &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]3362</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3363
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3364</div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3365
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3366<!-- _______________________________________________________________________ -->
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3367<div class="doc_subsubsection"> <a name="i_urem">'<tt>urem</tt>' Instruction</a>
3368</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3369
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3370<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3371
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3372<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3373<pre>
3374 &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]3375</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3376
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3377<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3378<p>The '<tt>urem</tt>' instruction returns the remainder from the unsigned
3379 division of its two arguments.</p>
3380
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3381<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3382<p>The two arguments to the '<tt>urem</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3383 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3384 values. Both arguments must have identical types.</p>
3385
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3386<h5>Semantics:</h5>
3387<p>This instruction returns the unsigned integer <i>remainder</i> of a division.
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3388 This instruction always performs an unsigned division to get the
3389 remainder.</p>
3390
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3391<p>Note that unsigned integer remainder and signed integer remainder are
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3392 distinct operations; for signed integer remainder, use '<tt>srem</tt>'.</p>
3393
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3394<p>Taking the remainder of a division by zero leads to undefined behavior.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3395
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3396<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3397<pre>
3398 &lt;result&gt; = urem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3399</pre>
3400
3401</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3402
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3403<!-- _______________________________________________________________________ -->
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3404<div class="doc_subsubsection">
3405 <a name="i_srem">'<tt>srem</tt>' Instruction</a>
3406</div>
3407
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3408<div class="doc_text">
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3409
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3410<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3411<pre>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]3412 &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]3413</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3414
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3415<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3416<p>The '<tt>srem</tt>' instruction returns the remainder from the signed
3417 division of its two operands. This instruction can also take
3418 <a href="#t_vector">vector</a> versions of the values in which case the
3419 elements must be integers.</p>
Chris Lattnerc7d3ab32008-01-04 04:33:49 +0000[diff] [blame]3420
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3421<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3422<p>The two arguments to the '<tt>srem</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3423 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3424 values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3425
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3426<h5>Semantics:</h5>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3427<p>This instruction returns the <i>remainder</i> of a division (where the result
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3428 has the same sign as the dividend, <tt>op1</tt>), not the <i>modulo</i>
3429 operator (where the result has the same sign as the divisor, <tt>op2</tt>) of
3430 a value. For more information about the difference,
3431 see <a href="http://mathforum.org/dr.math/problems/anne.4.28.99.html">The
3432 Math Forum</a>. For a table of how this is implemented in various languages,
3433 please see <a href="http://en.wikipedia.org/wiki/Modulo_operation">
3434 Wikipedia: modulo operation</a>.</p>
3435
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3436<p>Note that signed integer remainder and unsigned integer remainder are
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3437 distinct operations; for unsigned integer remainder, use '<tt>urem</tt>'.</p>
3438
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3439<p>Taking the remainder of a division by zero leads to undefined behavior.
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3440 Overflow also leads to undefined behavior; this is a rare case, but can
3441 occur, for example, by taking the remainder of a 32-bit division of
3442 -2147483648 by -1. (The remainder doesn't actually overflow, but this rule
3443 lets srem be implemented using instructions that return both the result of
3444 the division and the remainder.)</p>
3445
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3446<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3447<pre>
3448 &lt;result&gt; = srem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3449</pre>
3450
3451</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3452
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3453<!-- _______________________________________________________________________ -->
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3454<div class="doc_subsubsection">
3455 <a name="i_frem">'<tt>frem</tt>' Instruction</a> </div>
3456
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3457<div class="doc_text">
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3458
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3459<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3460<pre>
3461 &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]3462</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3463
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3464<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3465<p>The '<tt>frem</tt>' instruction returns the remainder from the division of
3466 its two operands.</p>
3467
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3468<h5>Arguments:</h5>
3469<p>The two arguments to the '<tt>frem</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3470 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
3471 floating point values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3472
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3473<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3474<p>This instruction returns the <i>remainder</i> of a division. The remainder
3475 has the same sign as the dividend.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3476
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3477<h5>Example:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3478<pre>
3479 &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]3480</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3481
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3482</div>
Robert Bocchino7b81c752006-02-17 21:18:08 +0000[diff] [blame]3483
Reid Spencer8e11bf82007-02-02 13:57:07 +0000[diff] [blame]3484<!-- ======================================================================= -->
3485<div class="doc_subsection"> <a name="bitwiseops">Bitwise Binary
3486Operations</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3487
Reid Spencer8e11bf82007-02-02 13:57:07 +0000[diff] [blame]3488<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3489
3490<p>Bitwise binary operators are used to do various forms of bit-twiddling in a
3491 program. They are generally very efficient instructions and can commonly be
3492 strength reduced from other instructions. They require two operands of the
3493 same type, execute an operation on them, and produce a single value. The
3494 resulting value is the same type as its operands.</p>
3495
Reid Spencer8e11bf82007-02-02 13:57:07 +0000[diff] [blame]3496</div>
3497
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3498<!-- _______________________________________________________________________ -->
3499<div class="doc_subsubsection"> <a name="i_shl">'<tt>shl</tt>'
3500Instruction</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3501
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3502<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3503
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3504<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3505<pre>
3506 &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]3507</pre>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]3508
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3509<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3510<p>The '<tt>shl</tt>' instruction returns the first operand shifted to the left
3511 a specified number of bits.</p>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]3512
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3513<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3514<p>Both arguments to the '<tt>shl</tt>' instruction must be the
3515 same <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of
3516 integer type. '<tt>op2</tt>' is treated as an unsigned value.</p>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3517
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3518<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3519<p>The value produced is <tt>op1</tt> * 2<sup><tt>op2</tt></sup> mod
3520 2<sup>n</sup>, where <tt>n</tt> is the width of the result. If <tt>op2</tt>
3521 is (statically or dynamically) negative or equal to or larger than the number
3522 of bits in <tt>op1</tt>, the result is undefined. If the arguments are
3523 vectors, each vector element of <tt>op1</tt> is shifted by the corresponding
3524 shift amount in <tt>op2</tt>.</p>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]3525
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3526<h5>Example:</h5>
3527<pre>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3528 &lt;result&gt; = shl i32 4, %var <i>; yields {i32}: 4 &lt;&lt; %var</i>
3529 &lt;result&gt; = shl i32 4, 2 <i>; yields {i32}: 16</i>
3530 &lt;result&gt; = shl i32 1, 10 <i>; yields {i32}: 1024</i>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]3531 &lt;result&gt; = shl i32 1, 32 <i>; undefined</i>
Mon P Wange9f10152008-12-09 05:46:39 +0000[diff] [blame]3532 &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]3533</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3534
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3535</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3536
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3537<!-- _______________________________________________________________________ -->
3538<div class="doc_subsubsection"> <a name="i_lshr">'<tt>lshr</tt>'
3539Instruction</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3540
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3541<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3542
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3543<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3544<pre>
3545 &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]3546</pre>
3547
3548<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3549<p>The '<tt>lshr</tt>' instruction (logical shift right) returns the first
3550 operand shifted to the right a specified number of bits with zero fill.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3551
3552<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3553<p>Both arguments to the '<tt>lshr</tt>' instruction must be the same
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3554 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3555 type. '<tt>op2</tt>' is treated as an unsigned value.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3556
3557<h5>Semantics:</h5>
3558<p>This instruction always performs a logical shift right operation. The most
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3559 significant bits of the result will be filled with zero bits after the shift.
3560 If <tt>op2</tt> is (statically or dynamically) equal to or larger than the
3561 number of bits in <tt>op1</tt>, the result is undefined. If the arguments are
3562 vectors, each vector element of <tt>op1</tt> is shifted by the corresponding
3563 shift amount in <tt>op2</tt>.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3564
3565<h5>Example:</h5>
3566<pre>
3567 &lt;result&gt; = lshr i32 4, 1 <i>; yields {i32}:result = 2</i>
3568 &lt;result&gt; = lshr i32 4, 2 <i>; yields {i32}:result = 1</i>
3569 &lt;result&gt; = lshr i8 4, 3 <i>; yields {i8}:result = 0</i>
3570 &lt;result&gt; = lshr i8 -2, 1 <i>; yields {i8}:result = 0x7FFFFFFF </i>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]3571 &lt;result&gt; = lshr i32 1, 32 <i>; undefined</i>
Mon P Wange9f10152008-12-09 05:46:39 +0000[diff] [blame]3572 &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]3573</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3574
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3575</div>
3576
Reid Spencer8e11bf82007-02-02 13:57:07 +0000[diff] [blame]3577<!-- _______________________________________________________________________ -->
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3578<div class="doc_subsubsection"> <a name="i_ashr">'<tt>ashr</tt>'
3579Instruction</a> </div>
3580<div class="doc_text">
3581
3582<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3583<pre>
3584 &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]3585</pre>
3586
3587<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3588<p>The '<tt>ashr</tt>' instruction (arithmetic shift right) returns the first
3589 operand shifted to the right a specified number of bits with sign
3590 extension.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3591
3592<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3593<p>Both arguments to the '<tt>ashr</tt>' instruction must be the same
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3594 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3595 type. '<tt>op2</tt>' is treated as an unsigned value.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3596
3597<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3598<p>This instruction always performs an arithmetic shift right operation, The
3599 most significant bits of the result will be filled with the sign bit
3600 of <tt>op1</tt>. If <tt>op2</tt> is (statically or dynamically) equal to or
3601 larger than the number of bits in <tt>op1</tt>, the result is undefined. If
3602 the arguments are vectors, each vector element of <tt>op1</tt> is shifted by
3603 the corresponding shift amount in <tt>op2</tt>.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3604
3605<h5>Example:</h5>
3606<pre>
3607 &lt;result&gt; = ashr i32 4, 1 <i>; yields {i32}:result = 2</i>
3608 &lt;result&gt; = ashr i32 4, 2 <i>; yields {i32}:result = 1</i>
3609 &lt;result&gt; = ashr i8 4, 3 <i>; yields {i8}:result = 0</i>
3610 &lt;result&gt; = ashr i8 -2, 1 <i>; yields {i8}:result = -1</i>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]3611 &lt;result&gt; = ashr i32 1, 32 <i>; undefined</i>
Mon P Wange9f10152008-12-09 05:46:39 +0000[diff] [blame]3612 &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]3613</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3614
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3615</div>
3616
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3617<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3618<div class="doc_subsubsection"> <a name="i_and">'<tt>and</tt>'
3619Instruction</a> </div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3620
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3621<div class="doc_text">
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3622
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3623<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3624<pre>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]3625 &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]3626</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3627
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3628<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3629<p>The '<tt>and</tt>' instruction returns the bitwise logical and of its two
3630 operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3631
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3632<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3633<p>The two arguments to the '<tt>and</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3634 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3635 values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3636
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3637<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3638<p>The truth table used for the '<tt>and</tt>' instruction is:</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3639
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3640<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3641 <tbody>
3642 <tr>
3643 <td>In0</td>
3644 <td>In1</td>
3645 <td>Out</td>
3646 </tr>
3647 <tr>
3648 <td>0</td>
3649 <td>0</td>
3650 <td>0</td>
3651 </tr>
3652 <tr>
3653 <td>0</td>
3654 <td>1</td>
3655 <td>0</td>
3656 </tr>
3657 <tr>
3658 <td>1</td>
3659 <td>0</td>
3660 <td>0</td>
3661 </tr>
3662 <tr>
3663 <td>1</td>
3664 <td>1</td>
3665 <td>1</td>
3666 </tr>
3667 </tbody>
3668</table>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3669
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3670<h5>Example:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3671<pre>
3672 &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]3673 &lt;result&gt; = and i32 15, 40 <i>; yields {i32}:result = 8</i>
3674 &lt;result&gt; = and i32 4, 8 <i>; yields {i32}:result = 0</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3675</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3676</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3677<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3678<div class="doc_subsubsection"> <a name="i_or">'<tt>or</tt>' Instruction</a> </div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3679
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3680<div class="doc_text">
3681
3682<h5>Syntax:</h5>
3683<pre>
3684 &lt;result&gt; = or &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3685</pre>
3686
3687<h5>Overview:</h5>
3688<p>The '<tt>or</tt>' instruction returns the bitwise logical inclusive or of its
3689 two operands.</p>
3690
3691<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3692<p>The two arguments to the '<tt>or</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3693 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3694 values. Both arguments must have identical types.</p>
3695
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3696<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3697<p>The truth table used for the '<tt>or</tt>' instruction is:</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3698
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3699<table border="1" cellspacing="0" cellpadding="4">
3700 <tbody>
3701 <tr>
3702 <td>In0</td>
3703 <td>In1</td>
3704 <td>Out</td>
3705 </tr>
3706 <tr>
3707 <td>0</td>
3708 <td>0</td>
3709 <td>0</td>
3710 </tr>
3711 <tr>
3712 <td>0</td>
3713 <td>1</td>
3714 <td>1</td>
3715 </tr>
3716 <tr>
3717 <td>1</td>
3718 <td>0</td>
3719 <td>1</td>
3720 </tr>
3721 <tr>
3722 <td>1</td>
3723 <td>1</td>
3724 <td>1</td>
3725 </tr>
3726 </tbody>
3727</table>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3728
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3729<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3730<pre>
3731 &lt;result&gt; = or i32 4, %var <i>; yields {i32}:result = 4 | %var</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3732 &lt;result&gt; = or i32 15, 40 <i>; yields {i32}:result = 47</i>
3733 &lt;result&gt; = or i32 4, 8 <i>; yields {i32}:result = 12</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3734</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3735
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3736</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3737
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3738<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3739<div class="doc_subsubsection"> <a name="i_xor">'<tt>xor</tt>'
3740Instruction</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3741
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3742<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3743
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3744<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3745<pre>
3746 &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]3747</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3748
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3749<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3750<p>The '<tt>xor</tt>' instruction returns the bitwise logical exclusive or of
3751 its two operands. The <tt>xor</tt> is used to implement the "one's
3752 complement" operation, which is the "~" operator in C.</p>
3753
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3754<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3755<p>The two arguments to the '<tt>xor</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3756 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3757 values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3758
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3759<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3760<p>The truth table used for the '<tt>xor</tt>' instruction is:</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3761
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3762<table border="1" cellspacing="0" cellpadding="4">
3763 <tbody>
3764 <tr>
3765 <td>In0</td>
3766 <td>In1</td>
3767 <td>Out</td>
3768 </tr>
3769 <tr>
3770 <td>0</td>
3771 <td>0</td>
3772 <td>0</td>
3773 </tr>
3774 <tr>
3775 <td>0</td>
3776 <td>1</td>
3777 <td>1</td>
3778 </tr>
3779 <tr>
3780 <td>1</td>
3781 <td>0</td>
3782 <td>1</td>
3783 </tr>
3784 <tr>
3785 <td>1</td>
3786 <td>1</td>
3787 <td>0</td>
3788 </tr>
3789 </tbody>
3790</table>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3791
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3792<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3793<pre>
3794 &lt;result&gt; = xor i32 4, %var <i>; yields {i32}:result = 4 ^ %var</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3795 &lt;result&gt; = xor i32 15, 40 <i>; yields {i32}:result = 39</i>
3796 &lt;result&gt; = xor i32 4, 8 <i>; yields {i32}:result = 12</i>
3797 &lt;result&gt; = xor i32 %V, -1 <i>; yields {i32}:result = ~%V</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3798</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3799
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3800</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3801
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3802<!-- ======================================================================= -->
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3803<div class="doc_subsection">
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3804 <a name="vectorops">Vector Operations</a>
3805</div>
3806
3807<div class="doc_text">
3808
3809<p>LLVM supports several instructions to represent vector operations in a
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3810 target-independent manner. These instructions cover the element-access and
3811 vector-specific operations needed to process vectors effectively. While LLVM
3812 does directly support these vector operations, many sophisticated algorithms
3813 will want to use target-specific intrinsics to take full advantage of a
3814 specific target.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3815
3816</div>
3817
3818<!-- _______________________________________________________________________ -->
3819<div class="doc_subsubsection">
3820 <a name="i_extractelement">'<tt>extractelement</tt>' Instruction</a>
3821</div>
3822
3823<div class="doc_text">
3824
3825<h5>Syntax:</h5>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3826<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3827 &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]3828</pre>
3829
3830<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3831<p>The '<tt>extractelement</tt>' instruction extracts a single scalar element
3832 from a vector at a specified index.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3833
3834
3835<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3836<p>The first operand of an '<tt>extractelement</tt>' instruction is a value
3837 of <a href="#t_vector">vector</a> type. The second operand is an index
3838 indicating the position from which to extract the element. The index may be
3839 a variable.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3840
3841<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3842<p>The result is a scalar of the same type as the element type of
3843 <tt>val</tt>. Its value is the value at position <tt>idx</tt> of
3844 <tt>val</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
3845 results are undefined.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3846
3847<h5>Example:</h5>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3848<pre>
Gabor Greifa5b6f452009-10-28 13:14:50 +0000[diff] [blame]3849 &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]3850</pre>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3851
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3852</div>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3853
3854<!-- _______________________________________________________________________ -->
3855<div class="doc_subsubsection">
3856 <a name="i_insertelement">'<tt>insertelement</tt>' Instruction</a>
3857</div>
3858
3859<div class="doc_text">
3860
3861<h5>Syntax:</h5>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3862<pre>
Dan Gohmanf3480b92008-05-12 23:38:42 +0000[diff] [blame]3863 &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]3864</pre>
3865
3866<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3867<p>The '<tt>insertelement</tt>' instruction inserts a scalar element into a
3868 vector at a specified index.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3869
3870<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3871<p>The first operand of an '<tt>insertelement</tt>' instruction is a value
3872 of <a href="#t_vector">vector</a> type. The second operand is a scalar value
3873 whose type must equal the element type of the first operand. The third
3874 operand is an index indicating the position at which to insert the value.
3875 The index may be a variable.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3876
3877<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3878<p>The result is a vector of the same type as <tt>val</tt>. Its element values
3879 are those of <tt>val</tt> except at position <tt>idx</tt>, where it gets the
3880 value <tt>elt</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
3881 results are undefined.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3882
3883<h5>Example:</h5>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3884<pre>
Gabor Greifa5b6f452009-10-28 13:14:50 +0000[diff] [blame]3885 &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]3886</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3887
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3888</div>
3889
3890<!-- _______________________________________________________________________ -->
3891<div class="doc_subsubsection">
3892 <a name="i_shufflevector">'<tt>shufflevector</tt>' Instruction</a>
3893</div>
3894
3895<div class="doc_text">
3896
3897<h5>Syntax:</h5>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3898<pre>
Mon P Wangaeb06d22008-11-10 04:46:22 +0000[diff] [blame]3899 &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]3900</pre>
3901
3902<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3903<p>The '<tt>shufflevector</tt>' instruction constructs a permutation of elements
3904 from two input vectors, returning a vector with the same element type as the
3905 input and length that is the same as the shuffle mask.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3906
3907<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3908<p>The first two operands of a '<tt>shufflevector</tt>' instruction are vectors
3909 with types that match each other. The third argument is a shuffle mask whose
3910 element type is always 'i32'. The result of the instruction is a vector
3911 whose length is the same as the shuffle mask and whose element type is the
3912 same as the element type of the first two operands.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3913
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3914<p>The shuffle mask operand is required to be a constant vector with either
3915 constant integer or undef values.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3916
3917<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3918<p>The elements of the two input vectors are numbered from left to right across
3919 both of the vectors. The shuffle mask operand specifies, for each element of
3920 the result vector, which element of the two input vectors the result element
3921 gets. The element selector may be undef (meaning "don't care") and the
3922 second operand may be undef if performing a shuffle from only one vector.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3923
3924<h5>Example:</h5>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3925<pre>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3926 &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]3927 &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]3928 &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]3929 &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]3930 &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]3931 &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]3932 &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]3933 &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]3934</pre>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3935
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3936</div>
Tanya Lattner09474292006-04-14 19:24:33 +0000[diff] [blame]3937
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3938<!-- ======================================================================= -->
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3939<div class="doc_subsection">
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]3940 <a name="aggregateops">Aggregate Operations</a>
3941</div>
3942
3943<div class="doc_text">
3944
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]3945<p>LLVM supports several instructions for working with
3946 <a href="#t_aggregate">aggregate</a> values.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]3947
3948</div>
3949
3950<!-- _______________________________________________________________________ -->
3951<div class="doc_subsubsection">
3952 <a name="i_extractvalue">'<tt>extractvalue</tt>' Instruction</a>
3953</div>
3954
3955<div class="doc_text">
3956
3957<h5>Syntax:</h5>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]3958<pre>
3959 &lt;result&gt; = extractvalue &lt;aggregate type&gt; &lt;val&gt;, &lt;idx&gt;{, &lt;idx&gt;}*
3960</pre>
3961
3962<h5>Overview:</h5>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]3963<p>The '<tt>extractvalue</tt>' instruction extracts the value of a member field
3964 from an <a href="#t_aggregate">aggregate</a> value.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]3965
3966<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3967<p>The first operand of an '<tt>extractvalue</tt>' instruction is a value
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]3968 of <a href="#t_struct">struct</a>, <a href="#t_union">union</a> or
3969 <a href="#t_array">array</a> type. The operands are constant indices to
3970 specify which value to extract in a similar manner as indices in a
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3971 '<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]3972
3973<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3974<p>The result is the value at the position in the aggregate specified by the
3975 index operands.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]3976
3977<h5>Example:</h5>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]3978<pre>
Gabor Greifa5b6f452009-10-28 13:14:50 +0000[diff] [blame]3979 &lt;result&gt; = extractvalue {i32, float} %agg, 0 <i>; yields i32</i>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]3980</pre>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]3981
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3982</div>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]3983
3984<!-- _______________________________________________________________________ -->
3985<div class="doc_subsubsection">
3986 <a name="i_insertvalue">'<tt>insertvalue</tt>' Instruction</a>
3987</div>
3988
3989<div class="doc_text">
3990
3991<h5>Syntax:</h5>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]3992<pre>
Jeffrey Yasskin7a088cf2010-01-11 19:19:26 +0000[diff] [blame]3993 &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]3994</pre>
3995
3996<h5>Overview:</h5>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]3997<p>The '<tt>insertvalue</tt>' instruction inserts a value into a member field
3998 in an <a href="#t_aggregate">aggregate</a> value.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]3999
4000<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4001<p>The first operand of an '<tt>insertvalue</tt>' instruction is a value
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4002 of <a href="#t_struct">struct</a>, <a href="#t_union">union</a> or
4003 <a href="#t_array">array</a> type. The second operand is a first-class
4004 value to insert. The following operands are constant indices indicating
4005 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]4006 '<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction. The
4007 value to insert must have the same type as the value identified by the
4008 indices.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4009
4010<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4011<p>The result is an aggregate of the same type as <tt>val</tt>. Its value is
4012 that of <tt>val</tt> except that the value at the position specified by the
4013 indices is that of <tt>elt</tt>.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4014
4015<h5>Example:</h5>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4016<pre>
Jeffrey Yasskin7a088cf2010-01-11 19:19:26 +0000[diff] [blame]4017 %agg1 = insertvalue {i32, float} undef, i32 1, 0 <i>; yields {i32 1, float undef}</i>
4018 %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]4019</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4020
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4021</div>
4022
4023
4024<!-- ======================================================================= -->
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4025<div class="doc_subsection">
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]4026 <a name="memoryops">Memory Access and Addressing Operations</a>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4027</div>
4028
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4029<div class="doc_text">
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4030
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4031<p>A key design point of an SSA-based representation is how it represents
4032 memory. In LLVM, no memory locations are in SSA form, which makes things
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]4033 very simple. This section describes how to read, write, and allocate
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4034 memory in LLVM.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4035
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4036</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4037
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4038<!-- _______________________________________________________________________ -->
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4039<div class="doc_subsubsection">
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4040 <a name="i_alloca">'<tt>alloca</tt>' Instruction</a>
4041</div>
4042
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4043<div class="doc_text">
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4044
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4045<h5>Syntax:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4046<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4047 &lt;result&gt; = alloca &lt;type&gt;[, i32 &lt;NumElements&gt;][, align &lt;alignment&gt;] <i>; yields {type*}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4048</pre>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4049
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4050<h5>Overview:</h5>
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]4051<p>The '<tt>alloca</tt>' instruction allocates memory on the stack frame of the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4052 currently executing function, to be automatically released when this function
4053 returns to its caller. The object is always allocated in the generic address
4054 space (address space zero).</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4055
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4056<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4057<p>The '<tt>alloca</tt>' instruction
4058 allocates <tt>sizeof(&lt;type&gt;)*NumElements</tt> bytes of memory on the
4059 runtime stack, returning a pointer of the appropriate type to the program.
4060 If "NumElements" is specified, it is the number of elements allocated,
4061 otherwise "NumElements" is defaulted to be one. If a constant alignment is
4062 specified, the value result of the allocation is guaranteed to be aligned to
4063 at least that boundary. If not specified, or if zero, the target can choose
4064 to align the allocation on any convenient boundary compatible with the
4065 type.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4066
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4067<p>'<tt>type</tt>' may be any sized type.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4068
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4069<h5>Semantics:</h5>
Bill Wendling871eb0a2009-05-08 20:49:29 +0000[diff] [blame]4070<p>Memory is allocated; a pointer is returned. The operation is undefined if
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4071 there is insufficient stack space for the allocation. '<tt>alloca</tt>'d
4072 memory is automatically released when the function returns. The
4073 '<tt>alloca</tt>' instruction is commonly used to represent automatic
4074 variables that must have an address available. When the function returns
4075 (either with the <tt><a href="#i_ret">ret</a></tt>
4076 or <tt><a href="#i_unwind">unwind</a></tt> instructions), the memory is
4077 reclaimed. Allocating zero bytes is legal, but the result is undefined.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4078
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4079<h5>Example:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4080<pre>
Dan Gohman81e21672009-01-04 23:49:44 +0000[diff] [blame]4081 %ptr = alloca i32 <i>; yields {i32*}:ptr</i>
4082 %ptr = alloca i32, i32 4 <i>; yields {i32*}:ptr</i>
4083 %ptr = alloca i32, i32 4, align 1024 <i>; yields {i32*}:ptr</i>
4084 %ptr = alloca i32, align 1024 <i>; yields {i32*}:ptr</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4085</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4086
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4087</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4088
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4089<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4090<div class="doc_subsubsection"> <a name="i_load">'<tt>load</tt>'
4091Instruction</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4092
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4093<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4094
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4095<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4096<pre>
Bill Wendling7c78dbb2010-02-25 21:23:24 +0000[diff] [blame]4097 &lt;result&gt; = load &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;][, !nontemporal !&lt;index&gt;]
4098 &lt;result&gt; = volatile load &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;][, !nontemporal !&lt;index&gt;]
4099 !&lt;index&gt; = !{ i32 1 }
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4100</pre>
4101
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4102<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4103<p>The '<tt>load</tt>' instruction is used to read from memory.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4104
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4105<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4106<p>The argument to the '<tt>load</tt>' instruction specifies the memory address
4107 from which to load. The pointer must point to
4108 a <a href="#t_firstclass">first class</a> type. If the <tt>load</tt> is
4109 marked as <tt>volatile</tt>, then the optimizer is not allowed to modify the
4110 number or order of execution of this <tt>load</tt> with other
4111 volatile <tt>load</tt> and <tt><a href="#i_store">store</a></tt>
David Greene8939b0d2010-02-16 20:50:18 +0000[diff] [blame]4112 instructions.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4113
Bill Wendling7c78dbb2010-02-25 21:23:24 +0000[diff] [blame]4114<p>The optional constant <tt>align</tt> argument specifies the alignment of the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4115 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]4116 omitted <tt>align</tt> argument means that the operation has the preferential
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4117 alignment for the target. It is the responsibility of the code emitter to
4118 ensure that the alignment information is correct. Overestimating the
Bill Wendling7c78dbb2010-02-25 21:23:24 +0000[diff] [blame]4119 alignment results in undefined behavior. Underestimating the alignment may
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4120 produce less efficient code. An alignment of 1 is always safe.</p>
4121
Bill Wendling7c78dbb2010-02-25 21:23:24 +0000[diff] [blame]4122<p>The optional <tt>!nontemporal</tt> metadata must reference a single
4123 metatadata name &lt;index&gt; corresponding to a metadata node with
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]4124 one <tt>i32</tt> entry of value 1. The existence of
Bill Wendling7c78dbb2010-02-25 21:23:24 +0000[diff] [blame]4125 the <tt>!nontemporal</tt> metatadata on the instruction tells the optimizer
4126 and code generator that this load is not expected to be reused in the cache.
4127 The code generator may select special instructions to save cache bandwidth,
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]4128 such as the <tt>MOVNT</tt> instruction on x86.</p>
David Greene8939b0d2010-02-16 20:50:18 +0000[diff] [blame]4129
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4130<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4131<p>The location of memory pointed to is loaded. If the value being loaded is of
4132 scalar type then the number of bytes read does not exceed the minimum number
4133 of bytes needed to hold all bits of the type. For example, loading an
4134 <tt>i24</tt> reads at most three bytes. When loading a value of a type like
4135 <tt>i20</tt> with a size that is not an integral number of bytes, the result
4136 is undefined if the value was not originally written using a store of the
4137 same type.</p>
4138
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4139<h5>Examples:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4140<pre>
4141 %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
4142 <a href="#i_store">store</a> i32 3, i32* %ptr <i>; yields {void}</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4143 %val = load i32* %ptr <i>; yields {i32}:val = i32 3</i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4144</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4145
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4146</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4147
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4148<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4149<div class="doc_subsubsection"> <a name="i_store">'<tt>store</tt>'
4150Instruction</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4151
Reid Spencer035ab572006-11-09 21:18:01 +0000[diff] [blame]4152<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4153
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4154<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4155<pre>
David Greene8939b0d2010-02-16 20:50:18 +0000[diff] [blame]4156 store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;][, !nontemporal !<index>] <i>; yields {void}</i>
4157 volatile store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;][, !nontemporal !<index>] <i>; yields {void}</i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4158</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4159
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4160<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4161<p>The '<tt>store</tt>' instruction is used to write to memory.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4162
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4163<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4164<p>There are two arguments to the '<tt>store</tt>' instruction: a value to store
4165 and an address at which to store it. The type of the
4166 '<tt>&lt;pointer&gt;</tt>' operand must be a pointer to
4167 the <a href="#t_firstclass">first class</a> type of the
4168 '<tt>&lt;value&gt;</tt>' operand. If the <tt>store</tt> is marked
4169 as <tt>volatile</tt>, then the optimizer is not allowed to modify the number
4170 or order of execution of this <tt>store</tt> with other
4171 volatile <tt>load</tt> and <tt><a href="#i_store">store</a></tt>
4172 instructions.</p>
4173
4174<p>The optional constant "align" argument specifies the alignment of the
4175 operation (that is, the alignment of the memory address). A value of 0 or an
4176 omitted "align" argument means that the operation has the preferential
4177 alignment for the target. It is the responsibility of the code emitter to
4178 ensure that the alignment information is correct. Overestimating the
4179 alignment results in an undefined behavior. Underestimating the alignment may
4180 produce less efficient code. An alignment of 1 is always safe.</p>
4181
David Greene8939b0d2010-02-16 20:50:18 +0000[diff] [blame]4182<p>The optional !nontemporal metadata must reference a single metatadata
4183 name <index> corresponding to a metadata node with one i32 entry of
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]4184 value 1. The existence of the !nontemporal metatadata on the
David Greene8939b0d2010-02-16 20:50:18 +0000[diff] [blame]4185 instruction tells the optimizer and code generator that this load is
4186 not expected to be reused in the cache. The code generator may
4187 select special instructions to save cache bandwidth, such as the
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]4188 MOVNT instruction on x86.</p>
David Greene8939b0d2010-02-16 20:50:18 +0000[diff] [blame]4189
4190
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4191<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4192<p>The contents of memory are updated to contain '<tt>&lt;value&gt;</tt>' at the
4193 location specified by the '<tt>&lt;pointer&gt;</tt>' operand. If
4194 '<tt>&lt;value&gt;</tt>' is of scalar type then the number of bytes written
4195 does not exceed the minimum number of bytes needed to hold all bits of the
4196 type. For example, storing an <tt>i24</tt> writes at most three bytes. When
4197 writing a value of a type like <tt>i20</tt> with a size that is not an
4198 integral number of bytes, it is unspecified what happens to the extra bits
4199 that do not belong to the type, but they will typically be overwritten.</p>
4200
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4201<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4202<pre>
4203 %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
Bill Wendling8c6c72d2007-10-22 05:10:05 +0000[diff] [blame]4204 store i32 3, i32* %ptr <i>; yields {void}</i>
4205 %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]4206</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4207
Reid Spencer47ce1792006-11-09 21:15:49 +0000[diff] [blame]4208</div>
4209
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4210<!-- _______________________________________________________________________ -->
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4211<div class="doc_subsubsection">
4212 <a name="i_getelementptr">'<tt>getelementptr</tt>' Instruction</a>
4213</div>
4214
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4215<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4216
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]4217<h5>Syntax:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4218<pre>
Matthijs Kooijmane49d0bc2008-10-13 13:44:15 +0000[diff] [blame]4219 &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]4220 &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]4221</pre>
4222
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]4223<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4224<p>The '<tt>getelementptr</tt>' instruction is used to get the address of a
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4225 subelement of an <a href="#t_aggregate">aggregate</a> data structure.
4226 It performs address calculation only and does not access memory.</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4227
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]4228<h5>Arguments:</h5>
Matthijs Kooijmane49d0bc2008-10-13 13:44:15 +0000[diff] [blame]4229<p>The first argument is always a pointer, and forms the basis of the
Chris Lattnerc8eef442009-07-29 06:44:13 +0000[diff] [blame]4230 calculation. The remaining arguments are indices that indicate which of the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4231 elements of the aggregate object are indexed. The interpretation of each
4232 index is dependent on the type being indexed into. The first index always
4233 indexes the pointer value given as the first argument, the second index
4234 indexes a value of the type pointed to (not necessarily the value directly
4235 pointed to, since the first index can be non-zero), etc. The first type
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4236 indexed into must be a pointer value, subsequent types can be arrays,
4237 vectors, structs and unions. Note that subsequent types being indexed into
4238 can never be pointers, since that would require loading the pointer before
4239 continuing calculation.</p>
Matthijs Kooijmane49d0bc2008-10-13 13:44:15 +0000[diff] [blame]4240
4241<p>The type of each index argument depends on the type it is indexing into.
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4242 When indexing into a (optionally packed) structure or union, only <tt>i32</tt>
4243 integer <b>constants</b> are allowed. When indexing into an array, pointer
4244 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]4245 constant.</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4246
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4247<p>For example, let's consider a C code fragment and how it gets compiled to
4248 LLVM:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4249
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]4250<div class="doc_code">
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4251<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]4252struct RT {
4253 char A;
Chris Lattnercabc8462007-05-29 15:43:56 +0000[diff] [blame]4254 int B[10][20];
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]4255 char C;
4256};
4257struct ST {
Chris Lattnercabc8462007-05-29 15:43:56 +0000[diff] [blame]4258 int X;
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]4259 double Y;
4260 struct RT Z;
4261};
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4262
Chris Lattnercabc8462007-05-29 15:43:56 +0000[diff] [blame]4263int *foo(struct ST *s) {
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]4264 return &amp;s[1].Z.B[5][13];
4265}
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4266</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]4267</div>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4268
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4269<p>The LLVM code generated by the GCC frontend is:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4270
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]4271<div class="doc_code">
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4272<pre>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]4273%RT = <a href="#namedtypes">type</a> { i8 , [10 x [20 x i32]], i8 }
4274%ST = <a href="#namedtypes">type</a> { i32, double, %RT }
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4275
Dan Gohman4df605b2009-07-25 02:23:48 +0000[diff] [blame]4276define i32* @foo(%ST* %s) {
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]4277entry:
4278 %reg = getelementptr %ST* %s, i32 1, i32 2, i32 1, i32 5, i32 13
4279 ret i32* %reg
4280}
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4281</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]4282</div>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4283
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]4284<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4285<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]4286 type, which is a pointer, yielding a '<tt>%ST</tt>' = '<tt>{ i32, double, %RT
4287 }</tt>' type, a structure. The second index indexes into the third element
4288 of the structure, yielding a '<tt>%RT</tt>' = '<tt>{ i8 , [10 x [20 x i32]],
4289 i8 }</tt>' type, another structure. The third index indexes into the second
4290 element of the structure, yielding a '<tt>[10 x [20 x i32]]</tt>' type, an
4291 array. The two dimensions of the array are subscripted into, yielding an
4292 '<tt>i32</tt>' type. The '<tt>getelementptr</tt>' instruction returns a
4293 pointer to this element, thus computing a value of '<tt>i32*</tt>' type.</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4294
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4295<p>Note that it is perfectly legal to index partially through a structure,
4296 returning a pointer to an inner element. Because of this, the LLVM code for
4297 the given testcase is equivalent to:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4298
4299<pre>
Dan Gohman4df605b2009-07-25 02:23:48 +0000[diff] [blame]4300 define i32* @foo(%ST* %s) {
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4301 %t1 = getelementptr %ST* %s, i32 1 <i>; yields %ST*:%t1</i>
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]4302 %t2 = getelementptr %ST* %t1, i32 0, i32 2 <i>; yields %RT*:%t2</i>
4303 %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]4304 %t4 = getelementptr [10 x [20 x i32]]* %t3, i32 0, i32 5 <i>; yields [20 x i32]*:%t4</i>
4305 %t5 = getelementptr [20 x i32]* %t4, i32 0, i32 13 <i>; yields i32*:%t5</i>
4306 ret i32* %t5
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4307 }
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]4308</pre>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]4309
Dan Gohmandd8004d2009-07-27 21:53:46 +0000[diff] [blame]4310<p>If the <tt>inbounds</tt> keyword is present, the result value of the
Dan Gohman0a28d182009-07-29 16:00:30 +0000[diff] [blame]4311 <tt>getelementptr</tt> is undefined if the base pointer is not an
4312 <i>in bounds</i> address of an allocated object, or if any of the addresses
Dan Gohmanb255b882009-08-20 17:08:17 +0000[diff] [blame]4313 that would be formed by successive addition of the offsets implied by the
4314 indices to the base address with infinitely precise arithmetic are not an
4315 <i>in bounds</i> address of that allocated object.
Dan Gohman0a28d182009-07-29 16:00:30 +0000[diff] [blame]4316 The <i>in bounds</i> addresses for an allocated object are all the addresses
Dan Gohmanb255b882009-08-20 17:08:17 +0000[diff] [blame]4317 that point into the object, plus the address one byte past the end.</p>
Dan Gohmandd8004d2009-07-27 21:53:46 +0000[diff] [blame]4318
4319<p>If the <tt>inbounds</tt> keyword is not present, the offsets are added to
4320 the base address with silently-wrapping two's complement arithmetic, and
4321 the result value of the <tt>getelementptr</tt> may be outside the object
4322 pointed to by the base pointer. The result value may not necessarily be
4323 used to access memory though, even if it happens to point into allocated
4324 storage. See the <a href="#pointeraliasing">Pointer Aliasing Rules</a>
4325 section for more information.</p>
4326
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4327<p>The getelementptr instruction is often confusing. For some more insight into
4328 how it works, see <a href="GetElementPtr.html">the getelementptr FAQ</a>.</p>
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]4329
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]4330<h5>Example:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4331<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4332 <i>; yields [12 x i8]*:aptr</i>
Matthijs Kooijmane49d0bc2008-10-13 13:44:15 +0000[diff] [blame]4333 %aptr = getelementptr {i32, [12 x i8]}* %saptr, i64 0, i32 1
4334 <i>; yields i8*:vptr</i>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]4335 %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]4336 <i>; yields i8*:eptr</i>
4337 %eptr = getelementptr [12 x i8]* %aptr, i64 0, i32 1
Sanjiv Gupta9f805c22009-04-25 07:27:44 +0000[diff] [blame]4338 <i>; yields i32*:iptr</i>
Sanjiv Gupta16ffa802009-04-24 16:38:13 +0000[diff] [blame]4339 %iptr = getelementptr [10 x i32]* @arr, i16 0, i16 0
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4340</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4341
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4342</div>
Reid Spencer47ce1792006-11-09 21:15:49 +0000[diff] [blame]4343
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4344<!-- ======================================================================= -->
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]4345<div class="doc_subsection"> <a name="convertops">Conversion Operations</a>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4346</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4347
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4348<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4349
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]4350<p>The instructions in this category are the conversion instructions (casting)
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4351 which all take a single operand and a type. They perform various bit
4352 conversions on the operand.</p>
4353
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4354</div>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]4355
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]4356<!-- _______________________________________________________________________ -->
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]4357<div class="doc_subsubsection">
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4358 <a name="i_trunc">'<tt>trunc .. to</tt>' Instruction</a>
4359</div>
4360<div class="doc_text">
4361
4362<h5>Syntax:</h5>
4363<pre>
4364 &lt;result&gt; = trunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4365</pre>
4366
4367<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4368<p>The '<tt>trunc</tt>' instruction truncates its operand to the
4369 type <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4370
4371<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4372<p>The '<tt>trunc</tt>' instruction takes a <tt>value</tt> to trunc, which must
4373 be an <a href="#t_integer">integer</a> type, and a type that specifies the
4374 size and type of the result, which must be
4375 an <a href="#t_integer">integer</a> type. The bit size of <tt>value</tt> must
4376 be larger than the bit size of <tt>ty2</tt>. Equal sized types are not
4377 allowed.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4378
4379<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4380<p>The '<tt>trunc</tt>' instruction truncates the high order bits
4381 in <tt>value</tt> and converts the remaining bits to <tt>ty2</tt>. Since the
4382 source size must be larger than the destination size, <tt>trunc</tt> cannot
4383 be a <i>no-op cast</i>. It will always truncate bits.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4384
4385<h5>Example:</h5>
4386<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4387 %X = trunc i32 257 to i8 <i>; yields i8:1</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]4388 %Y = trunc i32 123 to i1 <i>; yields i1:true</i>
Gabor Greif6a292012009-10-28 09:21:30 +0000[diff] [blame]4389 %Z = trunc i32 122 to i1 <i>; yields i1:false</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4390</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4391
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4392</div>
4393
4394<!-- _______________________________________________________________________ -->
4395<div class="doc_subsubsection">
4396 <a name="i_zext">'<tt>zext .. to</tt>' Instruction</a>
4397</div>
4398<div class="doc_text">
4399
4400<h5>Syntax:</h5>
4401<pre>
4402 &lt;result&gt; = zext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4403</pre>
4404
4405<h5>Overview:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4406<p>The '<tt>zext</tt>' instruction zero extends its operand to type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4407 <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4408
4409
4410<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4411<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]4412 <a href="#t_integer">integer</a> type, and a type to cast it to, which must
4413 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]4414 <tt>value</tt> must be smaller than the bit size of the destination type,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4415 <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4416
4417<h5>Semantics:</h5>
4418<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]4419 bits until it reaches the size of the destination type, <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4420
Reid Spencerb5929522007-01-12 15:46:11 +0000[diff] [blame]4421<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]4422
4423<h5>Example:</h5>
4424<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4425 %X = zext i32 257 to i64 <i>; yields i64:257</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]4426 %Y = zext i1 true to i32 <i>; yields i32:1</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4427</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4428
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4429</div>
4430
4431<!-- _______________________________________________________________________ -->
4432<div class="doc_subsubsection">
4433 <a name="i_sext">'<tt>sext .. to</tt>' Instruction</a>
4434</div>
4435<div class="doc_text">
4436
4437<h5>Syntax:</h5>
4438<pre>
4439 &lt;result&gt; = sext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4440</pre>
4441
4442<h5>Overview:</h5>
4443<p>The '<tt>sext</tt>' sign extends <tt>value</tt> to the type <tt>ty2</tt>.</p>
4444
4445<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4446<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]4447 <a href="#t_integer">integer</a> type, and a type to cast it to, which must
4448 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]4449 <tt>value</tt> must be smaller than the bit size of the destination type,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4450 <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4451
4452<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4453<p>The '<tt>sext</tt>' instruction performs a sign extension by copying the sign
4454 bit (highest order bit) of the <tt>value</tt> until it reaches the bit size
4455 of the type <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4456
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]4457<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]4458
4459<h5>Example:</h5>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4460<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4461 %X = sext i8 -1 to i16 <i>; yields i16 :65535</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]4462 %Y = sext i1 true to i32 <i>; yields i32:-1</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4463</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4464
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4465</div>
4466
4467<!-- _______________________________________________________________________ -->
4468<div class="doc_subsubsection">
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]4469 <a name="i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a>
4470</div>
4471
4472<div class="doc_text">
4473
4474<h5>Syntax:</h5>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]4475<pre>
4476 &lt;result&gt; = fptrunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4477</pre>
4478
4479<h5>Overview:</h5>
4480<p>The '<tt>fptrunc</tt>' instruction truncates <tt>value</tt> to type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4481 <tt>ty2</tt>.</p>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]4482
4483<h5>Arguments:</h5>
4484<p>The '<tt>fptrunc</tt>' instruction takes a <a href="#t_floating">floating
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4485 point</a> value to cast and a <a href="#t_floating">floating point</a> type
4486 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]4487 <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]4488 <i>no-op cast</i>.</p>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]4489
4490<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4491<p>The '<tt>fptrunc</tt>' instruction truncates a <tt>value</tt> from a larger
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4492 <a href="#t_floating">floating point</a> type to a smaller
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4493 <a href="#t_floating">floating point</a> type. If the value cannot fit
4494 within the destination type, <tt>ty2</tt>, then the results are
4495 undefined.</p>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]4496
4497<h5>Example:</h5>
4498<pre>
4499 %X = fptrunc double 123.0 to float <i>; yields float:123.0</i>
4500 %Y = fptrunc double 1.0E+300 to float <i>; yields undefined</i>
4501</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4502
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]4503</div>
4504
4505<!-- _______________________________________________________________________ -->
4506<div class="doc_subsubsection">
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4507 <a name="i_fpext">'<tt>fpext .. to</tt>' Instruction</a>
4508</div>
4509<div class="doc_text">
4510
4511<h5>Syntax:</h5>
4512<pre>
4513 &lt;result&gt; = fpext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4514</pre>
4515
4516<h5>Overview:</h5>
4517<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]4518 floating point value.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4519
4520<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4521<p>The '<tt>fpext</tt>' instruction takes a
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4522 <a href="#t_floating">floating point</a> <tt>value</tt> to cast, and
4523 a <a href="#t_floating">floating point</a> type to cast it to. The source
4524 type must be smaller than the destination type.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4525
4526<h5>Semantics:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]4527<p>The '<tt>fpext</tt>' instruction extends the <tt>value</tt> from a smaller
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4528 <a href="#t_floating">floating point</a> type to a larger
4529 <a href="#t_floating">floating point</a> type. The <tt>fpext</tt> cannot be
4530 used to make a <i>no-op cast</i> because it always changes bits. Use
4531 <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]4532
4533<h5>Example:</h5>
4534<pre>
4535 %X = fpext float 3.1415 to double <i>; yields double:3.1415</i>
4536 %Y = fpext float 1.0 to float <i>; yields float:1.0 (no-op)</i>
4537</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4538
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4539</div>
4540
4541<!-- _______________________________________________________________________ -->
4542<div class="doc_subsubsection">
Reid Spencer24d6da52007-01-21 00:29:26 +0000[diff] [blame]4543 <a name="i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4544</div>
4545<div class="doc_text">
4546
4547<h5>Syntax:</h5>
4548<pre>
Reid Spencer1539a1c2007-07-31 14:40:14 +0000[diff] [blame]4549 &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]4550</pre>
4551
4552<h5>Overview:</h5>
Reid Spencer1539a1c2007-07-31 14:40:14 +0000[diff] [blame]4553<p>The '<tt>fptoui</tt>' converts a floating point <tt>value</tt> to its
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4554 unsigned integer equivalent of type <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4555
4556<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4557<p>The '<tt>fptoui</tt>' instruction takes a value to cast, which must be a
4558 scalar or vector <a href="#t_floating">floating point</a> value, and a type
4559 to cast it to <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a>
4560 type. If <tt>ty</tt> is a vector floating point type, <tt>ty2</tt> must be a
4561 vector integer type with the same number of elements as <tt>ty</tt></p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4562
4563<h5>Semantics:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4564<p>The '<tt>fptoui</tt>' instruction converts its
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4565 <a href="#t_floating">floating point</a> operand into the nearest (rounding
4566 towards zero) unsigned integer value. If the value cannot fit
4567 in <tt>ty2</tt>, the results are undefined.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4568
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4569<h5>Example:</h5>
4570<pre>
Reid Spencer1539a1c2007-07-31 14:40:14 +0000[diff] [blame]4571 %X = fptoui double 123.0 to i32 <i>; yields i32:123</i>
Chris Lattner88519042007-09-22 03:17:52 +0000[diff] [blame]4572 %Y = fptoui float 1.0E+300 to i1 <i>; yields undefined:1</i>
Gabor Greif6a292012009-10-28 09:21:30 +0000[diff] [blame]4573 %Z = fptoui float 1.04E+17 to i8 <i>; yields undefined:1</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4574</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4575
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4576</div>
4577
4578<!-- _______________________________________________________________________ -->
4579<div class="doc_subsubsection">
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]4580 <a name="i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4581</div>
4582<div class="doc_text">
4583
4584<h5>Syntax:</h5>
4585<pre>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]4586 &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]4587</pre>
4588
4589<h5>Overview:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4590<p>The '<tt>fptosi</tt>' instruction converts
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4591 <a href="#t_floating">floating point</a> <tt>value</tt> to
4592 type <tt>ty2</tt>.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]4593
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]4594<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4595<p>The '<tt>fptosi</tt>' instruction takes a value to cast, which must be a
4596 scalar or vector <a href="#t_floating">floating point</a> value, and a type
4597 to cast it to <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a>
4598 type. If <tt>ty</tt> is a vector floating point type, <tt>ty2</tt> must be a
4599 vector integer type with the same number of elements as <tt>ty</tt></p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]4600
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]4601<h5>Semantics:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4602<p>The '<tt>fptosi</tt>' instruction converts its
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4603 <a href="#t_floating">floating point</a> operand into the nearest (rounding
4604 towards zero) signed integer value. If the value cannot fit in <tt>ty2</tt>,
4605 the results are undefined.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]4606
Chris Lattner33ba0d92001-07-09 00:26:23 +0000[diff] [blame]4607<h5>Example:</h5>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]4608<pre>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]4609 %X = fptosi double -123.0 to i32 <i>; yields i32:-123</i>
Chris Lattner88519042007-09-22 03:17:52 +0000[diff] [blame]4610 %Y = fptosi float 1.0E-247 to i1 <i>; yields undefined:1</i>
Gabor Greif6a292012009-10-28 09:21:30 +0000[diff] [blame]4611 %Z = fptosi float 1.04E+17 to i8 <i>; yields undefined:1</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4612</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4613
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4614</div>
4615
4616<!-- _______________________________________________________________________ -->
4617<div class="doc_subsubsection">
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]4618 <a name="i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4619</div>
4620<div class="doc_text">
4621
4622<h5>Syntax:</h5>
4623<pre>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]4624 &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]4625</pre>
4626
4627<h5>Overview:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]4628<p>The '<tt>uitofp</tt>' instruction regards <tt>value</tt> as an unsigned
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4629 integer and converts that value to the <tt>ty2</tt> type.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4630
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4631<h5>Arguments:</h5>
Nate Begemanb348d182007-11-17 03:58:34 +0000[diff] [blame]4632<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]4633 scalar or vector <a href="#t_integer">integer</a> value, and a type to cast
4634 it to <tt>ty2</tt>, which must be an <a href="#t_floating">floating point</a>
4635 type. If <tt>ty</tt> is a vector integer type, <tt>ty2</tt> must be a vector
4636 floating point type with the same number of elements as <tt>ty</tt></p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4637
4638<h5>Semantics:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]4639<p>The '<tt>uitofp</tt>' instruction interprets its operand as an unsigned
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4640 integer quantity and converts it to the corresponding floating point
4641 value. If the value cannot fit in the floating point value, the results are
4642 undefined.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4643
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4644<h5>Example:</h5>
4645<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4646 %X = uitofp i32 257 to float <i>; yields float:257.0</i>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]4647 %Y = uitofp i8 -1 to double <i>; yields double:255.0</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4648</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4649
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4650</div>
4651
4652<!-- _______________________________________________________________________ -->
4653<div class="doc_subsubsection">
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]4654 <a name="i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4655</div>
4656<div class="doc_text">
4657
4658<h5>Syntax:</h5>
4659<pre>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]4660 &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]4661</pre>
4662
4663<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4664<p>The '<tt>sitofp</tt>' instruction regards <tt>value</tt> as a signed integer
4665 and converts that value to the <tt>ty2</tt> type.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4666
4667<h5>Arguments:</h5>
Nate Begemanb348d182007-11-17 03:58:34 +0000[diff] [blame]4668<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]4669 scalar or vector <a href="#t_integer">integer</a> value, and a type to cast
4670 it to <tt>ty2</tt>, which must be an <a href="#t_floating">floating point</a>
4671 type. If <tt>ty</tt> is a vector integer type, <tt>ty2</tt> must be a vector
4672 floating point type with the same number of elements as <tt>ty</tt></p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4673
4674<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4675<p>The '<tt>sitofp</tt>' instruction interprets its operand as a signed integer
4676 quantity and converts it to the corresponding floating point value. If the
4677 value cannot fit in the floating point value, the results are undefined.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4678
4679<h5>Example:</h5>
4680<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4681 %X = sitofp i32 257 to float <i>; yields float:257.0</i>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]4682 %Y = sitofp i8 -1 to double <i>; yields double:-1.0</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4683</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4684
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4685</div>
4686
4687<!-- _______________________________________________________________________ -->
4688<div class="doc_subsubsection">
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]4689 <a name="i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a>
4690</div>
4691<div class="doc_text">
4692
4693<h5>Syntax:</h5>
4694<pre>
4695 &lt;result&gt; = ptrtoint &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4696</pre>
4697
4698<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4699<p>The '<tt>ptrtoint</tt>' instruction converts the pointer <tt>value</tt> to
4700 the integer type <tt>ty2</tt>.</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]4701
4702<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4703<p>The '<tt>ptrtoint</tt>' instruction takes a <tt>value</tt> to cast, which
4704 must be a <a href="#t_pointer">pointer</a> value, and a type to cast it to
4705 <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]4706
4707<h5>Semantics:</h5>
4708<p>The '<tt>ptrtoint</tt>' instruction converts <tt>value</tt> to integer type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4709 <tt>ty2</tt> by interpreting the pointer value as an integer and either
4710 truncating or zero extending that value to the size of the integer type. If
4711 <tt>value</tt> is smaller than <tt>ty2</tt> then a zero extension is done. If
4712 <tt>value</tt> is larger than <tt>ty2</tt> then a truncation is done. If they
4713 are the same size, then nothing is done (<i>no-op cast</i>) other than a type
4714 change.</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]4715
4716<h5>Example:</h5>
4717<pre>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4718 %X = ptrtoint i32* %X to i8 <i>; yields truncation on 32-bit architecture</i>
4719 %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]4720</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4721
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]4722</div>
4723
4724<!-- _______________________________________________________________________ -->
4725<div class="doc_subsubsection">
4726 <a name="i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a>
4727</div>
4728<div class="doc_text">
4729
4730<h5>Syntax:</h5>
4731<pre>
4732 &lt;result&gt; = inttoptr &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4733</pre>
4734
4735<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4736<p>The '<tt>inttoptr</tt>' instruction converts an integer <tt>value</tt> to a
4737 pointer type, <tt>ty2</tt>.</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]4738
4739<h5>Arguments:</h5>
Duncan Sands8036ca42007-03-30 12:22:09 +0000[diff] [blame]4740<p>The '<tt>inttoptr</tt>' instruction takes an <a href="#t_integer">integer</a>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4741 value to cast, and a type to cast it to, which must be a
4742 <a href="#t_pointer">pointer</a> type.</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]4743
4744<h5>Semantics:</h5>
4745<p>The '<tt>inttoptr</tt>' instruction converts <tt>value</tt> to type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4746 <tt>ty2</tt> by applying either a zero extension or a truncation depending on
4747 the size of the integer <tt>value</tt>. If <tt>value</tt> is larger than the
4748 size of a pointer then a truncation is done. If <tt>value</tt> is smaller
4749 than the size of a pointer then a zero extension is done. If they are the
4750 same size, nothing is done (<i>no-op cast</i>).</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]4751
4752<h5>Example:</h5>
4753<pre>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4754 %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]4755 %Y = inttoptr i32 255 to i32* <i>; yields no-op on 32-bit architecture</i>
4756 %Z = inttoptr i64 0 to i32* <i>; yields truncation on 32-bit architecture</i>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]4757</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4758
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]4759</div>
4760
4761<!-- _______________________________________________________________________ -->
4762<div class="doc_subsubsection">
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]4763 <a name="i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4764</div>
4765<div class="doc_text">
4766
4767<h5>Syntax:</h5>
4768<pre>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]4769 &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]4770</pre>
4771
4772<h5>Overview:</h5>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]4773<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4774 <tt>ty2</tt> without changing any bits.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4775
4776<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4777<p>The '<tt>bitcast</tt>' instruction takes a value to cast, which must be a
4778 non-aggregate first class value, and a type to cast it to, which must also be
4779 a non-aggregate <a href="#t_firstclass">first class</a> type. The bit sizes
4780 of <tt>value</tt> and the destination type, <tt>ty2</tt>, must be
4781 identical. If the source type is a pointer, the destination type must also be
4782 a pointer. This instruction supports bitwise conversion of vectors to
4783 integers and to vectors of other types (as long as they have the same
4784 size).</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4785
4786<h5>Semantics:</h5>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]4787<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4788 <tt>ty2</tt>. It is always a <i>no-op cast</i> because no bits change with
4789 this conversion. The conversion is done as if the <tt>value</tt> had been
4790 stored to memory and read back as type <tt>ty2</tt>. Pointer types may only
4791 be converted to other pointer types with this instruction. To convert
4792 pointers to other types, use the <a href="#i_inttoptr">inttoptr</a> or
4793 <a href="#i_ptrtoint">ptrtoint</a> instructions first.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4794
4795<h5>Example:</h5>
4796<pre>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4797 %X = bitcast i8 255 to i8 <i>; yields i8 :-1</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4798 %Y = bitcast i32* %x to sint* <i>; yields sint*:%x</i>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4799 %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]4800</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4801
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4802</div>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]4803
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]4804<!-- ======================================================================= -->
4805<div class="doc_subsection"> <a name="otherops">Other Operations</a> </div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4806
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]4807<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4808
4809<p>The instructions in this category are the "miscellaneous" instructions, which
4810 defy better classification.</p>
4811
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]4812</div>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4813
4814<!-- _______________________________________________________________________ -->
4815<div class="doc_subsubsection"><a name="i_icmp">'<tt>icmp</tt>' Instruction</a>
4816</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4817
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4818<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4819
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4820<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4821<pre>
4822 &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]4823</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4824
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4825<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4826<p>The '<tt>icmp</tt>' instruction returns a boolean value or a vector of
4827 boolean values based on comparison of its two integer, integer vector, or
4828 pointer operands.</p>
4829
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4830<h5>Arguments:</h5>
4831<p>The '<tt>icmp</tt>' instruction takes three operands. The first operand is
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4832 the condition code indicating the kind of comparison to perform. It is not a
4833 value, just a keyword. The possible condition code are:</p>
4834
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4835<ol>
4836 <li><tt>eq</tt>: equal</li>
4837 <li><tt>ne</tt>: not equal </li>
4838 <li><tt>ugt</tt>: unsigned greater than</li>
4839 <li><tt>uge</tt>: unsigned greater or equal</li>
4840 <li><tt>ult</tt>: unsigned less than</li>
4841 <li><tt>ule</tt>: unsigned less or equal</li>
4842 <li><tt>sgt</tt>: signed greater than</li>
4843 <li><tt>sge</tt>: signed greater or equal</li>
4844 <li><tt>slt</tt>: signed less than</li>
4845 <li><tt>sle</tt>: signed less or equal</li>
4846</ol>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4847
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]4848<p>The remaining two arguments must be <a href="#t_integer">integer</a> or
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4849 <a href="#t_pointer">pointer</a> or integer <a href="#t_vector">vector</a>
4850 typed. They must also be identical types.</p>
4851
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4852<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4853<p>The '<tt>icmp</tt>' compares <tt>op1</tt> and <tt>op2</tt> according to the
4854 condition code given as <tt>cond</tt>. The comparison performed always yields
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]4855 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]4856 result, as follows:</p>
4857
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4858<ol>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4859 <li><tt>eq</tt>: yields <tt>true</tt> if the operands are equal,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4860 <tt>false</tt> otherwise. No sign interpretation is necessary or
4861 performed.</li>
4862
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4863 <li><tt>ne</tt>: yields <tt>true</tt> if the operands are unequal,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4864 <tt>false</tt> otherwise. No sign interpretation is necessary or
4865 performed.</li>
4866
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4867 <li><tt>ugt</tt>: interprets the operands as unsigned values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4868 <tt>true</tt> if <tt>op1</tt> is greater than <tt>op2</tt>.</li>
4869
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4870 <li><tt>uge</tt>: interprets the operands as unsigned values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4871 <tt>true</tt> if <tt>op1</tt> is greater than or equal
4872 to <tt>op2</tt>.</li>
4873
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4874 <li><tt>ult</tt>: interprets the operands as unsigned values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4875 <tt>true</tt> if <tt>op1</tt> is less than <tt>op2</tt>.</li>
4876
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4877 <li><tt>ule</tt>: interprets the operands as unsigned values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4878 <tt>true</tt> if <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
4879
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4880 <li><tt>sgt</tt>: interprets the operands as signed values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4881 <tt>true</tt> if <tt>op1</tt> is greater than <tt>op2</tt>.</li>
4882
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4883 <li><tt>sge</tt>: interprets the operands as signed values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4884 <tt>true</tt> if <tt>op1</tt> is greater than or equal
4885 to <tt>op2</tt>.</li>
4886
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4887 <li><tt>slt</tt>: interprets the operands as signed values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4888 <tt>true</tt> if <tt>op1</tt> is less than <tt>op2</tt>.</li>
4889
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4890 <li><tt>sle</tt>: interprets the operands as signed values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4891 <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]4892</ol>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4893
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4894<p>If the operands are <a href="#t_pointer">pointer</a> typed, the pointer
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4895 values are compared as if they were integers.</p>
4896
4897<p>If the operands are integer vectors, then they are compared element by
4898 element. The result is an <tt>i1</tt> vector with the same number of elements
4899 as the values being compared. Otherwise, the result is an <tt>i1</tt>.</p>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4900
4901<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4902<pre>
4903 &lt;result&gt; = icmp eq i32 4, 5 <i>; yields: result=false</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4904 &lt;result&gt; = icmp ne float* %X, %X <i>; yields: result=false</i>
4905 &lt;result&gt; = icmp ult i16 4, 5 <i>; yields: result=true</i>
4906 &lt;result&gt; = icmp sgt i16 4, 5 <i>; yields: result=false</i>
4907 &lt;result&gt; = icmp ule i16 -4, 5 <i>; yields: result=false</i>
4908 &lt;result&gt; = icmp sge i16 4, 5 <i>; yields: result=false</i>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4909</pre>
Dan Gohmanc766f722009-01-22 01:39:38 +0000[diff] [blame]4910
4911<p>Note that the code generator does not yet support vector types with
4912 the <tt>icmp</tt> instruction.</p>
4913
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4914</div>
4915
4916<!-- _______________________________________________________________________ -->
4917<div class="doc_subsubsection"><a name="i_fcmp">'<tt>fcmp</tt>' Instruction</a>
4918</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4919
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4920<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4921
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4922<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4923<pre>
4924 &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]4925</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4926
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4927<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4928<p>The '<tt>fcmp</tt>' instruction returns a boolean value or vector of boolean
4929 values based on comparison of its operands.</p>
4930
4931<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]4932(<a href="#t_integer"><tt>i1</tt></a>).</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4933
4934<p>If the operands are floating point vectors, then the result type is a vector
4935 of boolean with the same number of elements as the operands being
4936 compared.</p>
4937
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4938<h5>Arguments:</h5>
4939<p>The '<tt>fcmp</tt>' instruction takes three operands. The first operand is
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4940 the condition code indicating the kind of comparison to perform. It is not a
4941 value, just a keyword. The possible condition code are:</p>
4942
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4943<ol>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]4944 <li><tt>false</tt>: no comparison, always returns false</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4945 <li><tt>oeq</tt>: ordered and equal</li>
4946 <li><tt>ogt</tt>: ordered and greater than </li>
4947 <li><tt>oge</tt>: ordered and greater than or equal</li>
4948 <li><tt>olt</tt>: ordered and less than </li>
4949 <li><tt>ole</tt>: ordered and less than or equal</li>
4950 <li><tt>one</tt>: ordered and not equal</li>
4951 <li><tt>ord</tt>: ordered (no nans)</li>
4952 <li><tt>ueq</tt>: unordered or equal</li>
4953 <li><tt>ugt</tt>: unordered or greater than </li>
4954 <li><tt>uge</tt>: unordered or greater than or equal</li>
4955 <li><tt>ult</tt>: unordered or less than </li>
4956 <li><tt>ule</tt>: unordered or less than or equal</li>
4957 <li><tt>une</tt>: unordered or not equal</li>
4958 <li><tt>uno</tt>: unordered (either nans)</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]4959 <li><tt>true</tt>: no comparison, always returns true</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4960</ol>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4961
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4962<p><i>Ordered</i> means that neither operand is a QNAN while
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4963 <i>unordered</i> means that either operand may be a QNAN.</p>
4964
4965<p>Each of <tt>val1</tt> and <tt>val2</tt> arguments must be either
4966 a <a href="#t_floating">floating point</a> type or
4967 a <a href="#t_vector">vector</a> of floating point type. They must have
4968 identical types.</p>
4969
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4970<h5>Semantics:</h5>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4971<p>The '<tt>fcmp</tt>' instruction compares <tt>op1</tt> and <tt>op2</tt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4972 according to the condition code given as <tt>cond</tt>. If the operands are
4973 vectors, then the vectors are compared element by element. Each comparison
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]4974 performed always yields an <a href="#t_integer">i1</a> result, as
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4975 follows:</p>
4976
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4977<ol>
4978 <li><tt>false</tt>: always yields <tt>false</tt>, regardless of operands.</li>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4979
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4980 <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]4981 <tt>op1</tt> is equal to <tt>op2</tt>.</li>
4982
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]4983 <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]4984 <tt>op1</tt> is greater than <tt>op2</tt>.</li>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4985
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4986 <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]4987 <tt>op1</tt> is greater than or equal to <tt>op2</tt>.</li>
4988
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4989 <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]4990 <tt>op1</tt> is less than <tt>op2</tt>.</li>
4991
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4992 <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]4993 <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
4994
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4995 <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]4996 <tt>op1</tt> is not equal to <tt>op2</tt>.</li>
4997
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]4998 <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]4999
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5000 <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]5001 <tt>op1</tt> is equal to <tt>op2</tt>.</li>
5002
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5003 <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]5004 <tt>op1</tt> is greater than <tt>op2</tt>.</li>
5005
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5006 <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]5007 <tt>op1</tt> is greater than or equal to <tt>op2</tt>.</li>
5008
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5009 <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]5010 <tt>op1</tt> is less than <tt>op2</tt>.</li>
5011
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5012 <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]5013 <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
5014
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5015 <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]5016 <tt>op1</tt> is not equal to <tt>op2</tt>.</li>
5017
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]5018 <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]5019
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5020 <li><tt>true</tt>: always yields <tt>true</tt>, regardless of operands.</li>
5021</ol>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5022
5023<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5024<pre>
5025 &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]5026 &lt;result&gt; = fcmp one float 4.0, 5.0 <i>; yields: result=true</i>
5027 &lt;result&gt; = fcmp olt float 4.0, 5.0 <i>; yields: result=true</i>
5028 &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]5029</pre>
Dan Gohmanc766f722009-01-22 01:39:38 +0000[diff] [blame]5030
5031<p>Note that the code generator does not yet support vector types with
5032 the <tt>fcmp</tt> instruction.</p>
5033
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5034</div>
5035
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5036<!-- _______________________________________________________________________ -->
Nate Begemanac80ade2008-05-12 19:01:56 +0000[diff] [blame]5037<div class="doc_subsubsection">
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]5038 <a name="i_phi">'<tt>phi</tt>' Instruction</a>
5039</div>
5040
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5041<div class="doc_text">
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]5042
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5043<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5044<pre>
5045 &lt;result&gt; = phi &lt;ty&gt; [ &lt;val0&gt;, &lt;label0&gt;], ...
5046</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]5047
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5048<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5049<p>The '<tt>phi</tt>' instruction is used to implement the &#966; node in the
5050 SSA graph representing the function.</p>
5051
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5052<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5053<p>The type of the incoming values is specified with the first type field. After
5054 this, the '<tt>phi</tt>' instruction takes a list of pairs as arguments, with
5055 one pair for each predecessor basic block of the current block. Only values
5056 of <a href="#t_firstclass">first class</a> type may be used as the value
5057 arguments to the PHI node. Only labels may be used as the label
5058 arguments.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]5059
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5060<p>There must be no non-phi instructions between the start of a basic block and
5061 the PHI instructions: i.e. PHI instructions must be first in a basic
5062 block.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]5063
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5064<p>For the purposes of the SSA form, the use of each incoming value is deemed to
5065 occur on the edge from the corresponding predecessor block to the current
5066 block (but after any definition of an '<tt>invoke</tt>' instruction's return
5067 value on the same edge).</p>
Jay Foadd2449092009-06-03 10:20:10 +0000[diff] [blame]5068
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5069<h5>Semantics:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5070<p>At runtime, the '<tt>phi</tt>' instruction logically takes on the value
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5071 specified by the pair corresponding to the predecessor basic block that
5072 executed just prior to the current block.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]5073
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5074<h5>Example:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]5075<pre>
5076Loop: ; Infinite loop that counts from 0 on up...
5077 %indvar = phi i32 [ 0, %LoopHeader ], [ %nextindvar, %Loop ]
5078 %nextindvar = add i32 %indvar, 1
5079 br label %Loop
5080</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5081
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5082</div>
5083
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5084<!-- _______________________________________________________________________ -->
5085<div class="doc_subsubsection">
5086 <a name="i_select">'<tt>select</tt>' Instruction</a>
5087</div>
5088
5089<div class="doc_text">
5090
5091<h5>Syntax:</h5>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5092<pre>
Dan Gohmanf72fb672008-09-09 01:02:47 +0000[diff] [blame]5093 &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>
5094
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5095 <i>selty</i> is either i1 or {&lt;N x i1&gt;}
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5096</pre>
5097
5098<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5099<p>The '<tt>select</tt>' instruction is used to choose one value based on a
5100 condition, without branching.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5101
5102
5103<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5104<p>The '<tt>select</tt>' instruction requires an 'i1' value or a vector of 'i1'
5105 values indicating the condition, and two values of the
5106 same <a href="#t_firstclass">first class</a> type. If the val1/val2 are
5107 vectors and the condition is a scalar, then entire vectors are selected, not
5108 individual elements.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5109
5110<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5111<p>If the condition is an i1 and it evaluates to 1, the instruction returns the
5112 first value argument; otherwise, it returns the second value argument.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5113
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5114<p>If the condition is a vector of i1, then the value arguments must be vectors
5115 of the same size, and the selection is done element by element.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5116
5117<h5>Example:</h5>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5118<pre>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]5119 %X = select i1 true, i8 17, i8 42 <i>; yields i8:17</i>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5120</pre>
Dan Gohmanc766f722009-01-22 01:39:38 +0000[diff] [blame]5121
5122<p>Note that the code generator does not yet support conditions
5123 with vector type.</p>
5124
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5125</div>
5126
Robert Bocchino05ccd702006-01-15 20:48:27 +0000[diff] [blame]5127<!-- _______________________________________________________________________ -->
5128<div class="doc_subsubsection">
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5129 <a name="i_call">'<tt>call</tt>' Instruction</a>
5130</div>
5131
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5132<div class="doc_text">
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5133
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]5134<h5>Syntax:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5135<pre>
Devang Patel307e8ab2008-10-07 17:48:33 +0000[diff] [blame]5136 &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]5137</pre>
5138
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]5139<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5140<p>The '<tt>call</tt>' instruction represents a simple function call.</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5141
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]5142<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5143<p>This instruction requires several arguments:</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5144
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]5145<ol>
Jeffrey Yasskin95fa80a2010-01-09 19:44:16 +0000[diff] [blame]5146 <li>The optional "tail" marker indicates that the callee function does not
5147 access any allocas or varargs in the caller. Note that calls may be
5148 marked "tail" even if they do not occur before
5149 a <a href="#i_ret"><tt>ret</tt></a> instruction. If the "tail" marker is
5150 present, the function call is eligible for tail call optimization,
5151 but <a href="CodeGenerator.html#tailcallopt">might not in fact be
Evan Chengdc444e92010-03-08 21:05:02 +0000[diff] [blame^]5152 optimized into a jump</a>. The code generator may optimize calls marked
5153 "tail" with either 1) automatic <a href="CodeGenerator.html#sibcallopt">
5154 sibling call optimization</a> when the caller and callee have
5155 matching signatures, or 2) forced tail call optimization when the
5156 following extra requirements are met:
Jeffrey Yasskin95fa80a2010-01-09 19:44:16 +0000[diff] [blame]5157 <ul>
5158 <li>Caller and callee both have the calling
5159 convention <tt>fastcc</tt>.</li>
5160 <li>The call is in tail position (ret immediately follows call and ret
5161 uses value of call or is void).</li>
5162 <li>Option <tt>-tailcallopt</tt> is enabled,
Dan Gohmanfbbee8d2010-03-02 01:08:11 +0000[diff] [blame]5163 or <code>llvm::GuaranteedTailCallOpt</code> is <code>true</code>.</li>
Jeffrey Yasskin95fa80a2010-01-09 19:44:16 +0000[diff] [blame]5164 <li><a href="CodeGenerator.html#tailcallopt">Platform specific
5165 constraints are met.</a></li>
5166 </ul>
5167 </li>
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]5168
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5169 <li>The optional "cconv" marker indicates which <a href="#callingconv">calling
5170 convention</a> the call should use. If none is specified, the call
Jeffrey Yasskin95fa80a2010-01-09 19:44:16 +0000[diff] [blame]5171 defaults to using C calling conventions. The calling convention of the
5172 call must match the calling convention of the target function, or else the
5173 behavior is undefined.</li>
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]5174
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5175 <li>The optional <a href="#paramattrs">Parameter Attributes</a> list for
5176 return values. Only '<tt>zeroext</tt>', '<tt>signext</tt>', and
5177 '<tt>inreg</tt>' attributes are valid here.</li>
5178
5179 <li>'<tt>ty</tt>': the type of the call instruction itself which is also the
5180 type of the return value. Functions that return no value are marked
5181 <tt><a href="#t_void">void</a></tt>.</li>
5182
5183 <li>'<tt>fnty</tt>': shall be the signature of the pointer to function value
5184 being invoked. The argument types must match the types implied by this
5185 signature. This type can be omitted if the function is not varargs and if
5186 the function type does not return a pointer to a function.</li>
5187
5188 <li>'<tt>fnptrval</tt>': An LLVM value containing a pointer to a function to
5189 be invoked. In most cases, this is a direct function invocation, but
5190 indirect <tt>call</tt>s are just as possible, calling an arbitrary pointer
5191 to function value.</li>
5192
5193 <li>'<tt>function args</tt>': argument list whose types match the function
Chris Lattner0724fbd2010-03-02 06:36:51 +0000[diff] [blame]5194 signature argument types and parameter attributes. All arguments must be
5195 of <a href="#t_firstclass">first class</a> type. If the function
5196 signature indicates the function accepts a variable number of arguments,
5197 the extra arguments can be specified.</li>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5198
5199 <li>The optional <a href="#fnattrs">function attributes</a> list. Only
5200 '<tt>noreturn</tt>', '<tt>nounwind</tt>', '<tt>readonly</tt>' and
5201 '<tt>readnone</tt>' attributes are valid here.</li>
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]5202</ol>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5203
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]5204<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5205<p>The '<tt>call</tt>' instruction is used to cause control flow to transfer to
5206 a specified function, with its incoming arguments bound to the specified
5207 values. Upon a '<tt><a href="#i_ret">ret</a></tt>' instruction in the called
5208 function, control flow continues with the instruction after the function
5209 call, and the return value of the function is bound to the result
5210 argument.</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5211
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]5212<h5>Example:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5213<pre>
Nick Lewyckydb7e3c92007-09-08 13:57:50 +0000[diff] [blame]5214 %retval = call i32 @test(i32 %argc)
Chris Lattner772fccf2008-03-21 17:24:17 +0000[diff] [blame]5215 call i32 (i8 *, ...)* @printf(i8 * %msg, i32 12, i8 42) <i>; yields i32</i>
5216 %X = tail call i32 @foo() <i>; yields i32</i>
5217 %Y = tail call <a href="#callingconv">fastcc</a> i32 @foo() <i>; yields i32</i>
5218 call void %foo(i8 97 signext)
Devang Patelc3fc6df2008-03-10 20:49:15 +0000[diff] [blame]5219
5220 %struct.A = type { i32, i8 }
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]5221 %r = call %struct.A @foo() <i>; yields { 32, i8 }</i>
Dan Gohmanb1e6b962008-10-04 19:00:07 +0000[diff] [blame]5222 %gr = extractvalue %struct.A %r, 0 <i>; yields i32</i>
5223 %gr1 = extractvalue %struct.A %r, 1 <i>; yields i8</i>
Chris Lattner85a350f2008-10-08 06:26:11 +0000[diff] [blame]5224 %Z = call void @foo() noreturn <i>; indicates that %foo never returns normally</i>
Matthijs Kooijmancb73d192008-10-07 10:03:45 +0000[diff] [blame]5225 %ZZ = call zeroext i32 @bar() <i>; Return value is %zero extended</i>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5226</pre>
5227
Dale Johannesen07de8d12009-09-24 18:38:21 +0000[diff] [blame]5228<p>llvm treats calls to some functions with names and arguments that match the
Dale Johannesen9f8380b2009-09-25 17:04:42 +0000[diff] [blame]5229standard C99 library as being the C99 library functions, and may perform
5230optimizations or generate code for them under that assumption. This is
5231something we'd like to change in the future to provide better support for
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]5232freestanding environments and non-C-based languages.</p>
Dale Johannesen07de8d12009-09-24 18:38:21 +0000[diff] [blame]5233
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5234</div>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5235
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5236<!-- _______________________________________________________________________ -->
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5237<div class="doc_subsubsection">
Chris Lattnerfb6977d2006-01-13 23:26:01 +0000[diff] [blame]5238 <a name="i_va_arg">'<tt>va_arg</tt>' Instruction</a>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5239</div>
5240
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5241<div class="doc_text">
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5242
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]5243<h5>Syntax:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5244<pre>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]5245 &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]5246</pre>
5247
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]5248<h5>Overview:</h5>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]5249<p>The '<tt>va_arg</tt>' instruction is used to access arguments passed through
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5250 the "variable argument" area of a function call. It is used to implement the
5251 <tt>va_arg</tt> macro in C.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5252
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]5253<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5254<p>This instruction takes a <tt>va_list*</tt> value and the type of the
5255 argument. It returns a value of the specified argument type and increments
5256 the <tt>va_list</tt> to point to the next argument. The actual type
5257 of <tt>va_list</tt> is target specific.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5258
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]5259<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5260<p>The '<tt>va_arg</tt>' instruction loads an argument of the specified type
5261 from the specified <tt>va_list</tt> and causes the <tt>va_list</tt> to point
5262 to the next argument. For more information, see the variable argument
5263 handling <a href="#int_varargs">Intrinsic Functions</a>.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5264
5265<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]5266 take a variable number of arguments, for example, the <tt>vfprintf</tt>
5267 function.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5268
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5269<p><tt>va_arg</tt> is an LLVM instruction instead of
5270 an <a href="#intrinsics">intrinsic function</a> because it takes a type as an
5271 argument.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5272
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]5273<h5>Example:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5274<p>See the <a href="#int_varargs">variable argument processing</a> section.</p>
5275
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5276<p>Note that the code generator does not yet fully support va_arg on many
5277 targets. Also, it does not currently support va_arg with aggregate types on
5278 any target.</p>
Dan Gohmanf3e60bd2009-01-12 23:12:39 +0000[diff] [blame]5279
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5280</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5281
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5282<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]5283<div class="doc_section"> <a name="intrinsics">Intrinsic Functions</a> </div>
5284<!-- *********************************************************************** -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5285
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5286<div class="doc_text">
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5287
5288<p>LLVM supports the notion of an "intrinsic function". These functions have
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5289 well known names and semantics and are required to follow certain
5290 restrictions. Overall, these intrinsics represent an extension mechanism for
5291 the LLVM language that does not require changing all of the transformations
5292 in LLVM when adding to the language (or the bitcode reader/writer, the
5293 parser, etc...).</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5294
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]5295<p>Intrinsic function names must all start with an "<tt>llvm.</tt>" prefix. This
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5296 prefix is reserved in LLVM for intrinsic names; thus, function names may not
5297 begin with this prefix. Intrinsic functions must always be external
5298 functions: you cannot define the body of intrinsic functions. Intrinsic
5299 functions may only be used in call or invoke instructions: it is illegal to
5300 take the address of an intrinsic function. Additionally, because intrinsic
5301 functions are part of the LLVM language, it is required if any are added that
5302 they be documented here.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5303
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5304<p>Some intrinsic functions can be overloaded, i.e., the intrinsic represents a
5305 family of functions that perform the same operation but on different data
5306 types. Because LLVM can represent over 8 million different integer types,
5307 overloading is used commonly to allow an intrinsic function to operate on any
5308 integer type. One or more of the argument types or the result type can be
5309 overloaded to accept any integer type. Argument types may also be defined as
5310 exactly matching a previous argument's type or the result type. This allows
5311 an intrinsic function which accepts multiple arguments, but needs all of them
5312 to be of the same type, to only be overloaded with respect to a single
5313 argument or the result.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5314
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5315<p>Overloaded intrinsics will have the names of its overloaded argument types
5316 encoded into its function name, each preceded by a period. Only those types
5317 which are overloaded result in a name suffix. Arguments whose type is matched
5318 against another type do not. For example, the <tt>llvm.ctpop</tt> function
5319 can take an integer of any width and returns an integer of exactly the same
5320 integer width. This leads to a family of functions such as
5321 <tt>i8 @llvm.ctpop.i8(i8 %val)</tt> and <tt>i29 @llvm.ctpop.i29(i29
5322 %val)</tt>. Only one type, the return type, is overloaded, and only one type
5323 suffix is required. Because the argument's type is matched against the return
5324 type, it does not require its own name suffix.</p>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]5325
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5326<p>To learn how to add an intrinsic function, please see the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5327 <a href="ExtendingLLVM.html">Extending LLVM Guide</a>.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5328
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5329</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5330
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5331<!-- ======================================================================= -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5332<div class="doc_subsection">
5333 <a name="int_varargs">Variable Argument Handling Intrinsics</a>
5334</div>
5335
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5336<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5337
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5338<p>Variable argument support is defined in LLVM with
5339 the <a href="#i_va_arg"><tt>va_arg</tt></a> instruction and these three
5340 intrinsic functions. These functions are related to the similarly named
5341 macros defined in the <tt>&lt;stdarg.h&gt;</tt> header file.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5342
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5343<p>All of these functions operate on arguments that use a target-specific value
5344 type "<tt>va_list</tt>". The LLVM assembly language reference manual does
5345 not define what this type is, so all transformations should be prepared to
5346 handle these functions regardless of the type used.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5347
Chris Lattner374ab302006-05-15 17:26:46 +0000[diff] [blame]5348<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]5349 instruction and the variable argument handling intrinsic functions are
5350 used.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5351
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]5352<div class="doc_code">
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5353<pre>
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]5354define i32 @test(i32 %X, ...) {
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5355 ; Initialize variable argument processing
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5356 %ap = alloca i8*
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]5357 %ap2 = bitcast i8** %ap to i8*
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]5358 call void @llvm.va_start(i8* %ap2)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5359
5360 ; Read a single integer argument
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5361 %tmp = va_arg i8** %ap, i32
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5362
5363 ; Demonstrate usage of llvm.va_copy and llvm.va_end
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5364 %aq = alloca i8*
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]5365 %aq2 = bitcast i8** %aq to i8*
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5366 call void @llvm.va_copy(i8* %aq2, i8* %ap2)
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]5367 call void @llvm.va_end(i8* %aq2)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5368
5369 ; Stop processing of arguments.
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]5370 call void @llvm.va_end(i8* %ap2)
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5371 ret i32 %tmp
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5372}
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]5373
5374declare void @llvm.va_start(i8*)
5375declare void @llvm.va_copy(i8*, i8*)
5376declare void @llvm.va_end(i8*)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5377</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5378</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5379
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]5380</div>
5381
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5382<!-- _______________________________________________________________________ -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5383<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5384 <a name="int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5385</div>
5386
5387
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5388<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5389
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5390<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5391<pre>
5392 declare void %llvm.va_start(i8* &lt;arglist&gt;)
5393</pre>
5394
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5395<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5396<p>The '<tt>llvm.va_start</tt>' intrinsic initializes <tt>*&lt;arglist&gt;</tt>
5397 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]5398
5399<h5>Arguments:</h5>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5400<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]5401
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5402<h5>Semantics:</h5>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5403<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]5404 macro available in C. In a target-dependent way, it initializes
5405 the <tt>va_list</tt> element to which the argument points, so that the next
5406 call to <tt>va_arg</tt> will produce the first variable argument passed to
5407 the function. Unlike the C <tt>va_start</tt> macro, this intrinsic does not
5408 need to know the last argument of the function as the compiler can figure
5409 that out.</p>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]5410
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5411</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5412
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5413<!-- _______________________________________________________________________ -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5414<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5415 <a name="int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5416</div>
5417
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5418<div class="doc_text">
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]5419
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5420<h5>Syntax:</h5>
5421<pre>
5422 declare void @llvm.va_end(i8* &lt;arglist&gt;)
5423</pre>
5424
5425<h5>Overview:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5426<p>The '<tt>llvm.va_end</tt>' intrinsic destroys <tt>*&lt;arglist&gt;</tt>,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5427 which has been initialized previously
5428 with <tt><a href="#int_va_start">llvm.va_start</a></tt>
5429 or <tt><a href="#i_va_copy">llvm.va_copy</a></tt>.</p>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]5430
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5431<h5>Arguments:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5432<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]5433
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5434<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5435<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]5436 macro available in C. In a target-dependent way, it destroys
5437 the <tt>va_list</tt> element to which the argument points. Calls
5438 to <a href="#int_va_start"><tt>llvm.va_start</tt></a>
5439 and <a href="#int_va_copy"> <tt>llvm.va_copy</tt></a> must be matched exactly
5440 with calls to <tt>llvm.va_end</tt>.</p>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]5441
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5442</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5443
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5444<!-- _______________________________________________________________________ -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5445<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5446 <a name="int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5447</div>
5448
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5449<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5450
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5451<h5>Syntax:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5452<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]5453 declare void @llvm.va_copy(i8* &lt;destarglist&gt;, i8* &lt;srcarglist&gt;)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5454</pre>
5455
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5456<h5>Overview:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5457<p>The '<tt>llvm.va_copy</tt>' intrinsic copies the current argument position
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5458 from the source argument list to the destination argument list.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5459
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5460<h5>Arguments:</h5>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]5461<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]5462 The second argument is a pointer to a <tt>va_list</tt> element to copy
5463 from.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5464
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5465<h5>Semantics:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5466<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]5467 macro available in C. In a target-dependent way, it copies the
5468 source <tt>va_list</tt> element into the destination <tt>va_list</tt>
5469 element. This intrinsic is necessary because
5470 the <tt><a href="#int_va_start"> llvm.va_start</a></tt> intrinsic may be
5471 arbitrarily complex and require, for example, memory allocation.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5472
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5473</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5474
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5475<!-- ======================================================================= -->
5476<div class="doc_subsection">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5477 <a name="int_gc">Accurate Garbage Collection Intrinsics</a>
5478</div>
5479
5480<div class="doc_text">
5481
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5482<p>LLVM support for <a href="GarbageCollection.html">Accurate Garbage
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]5483Collection</a> (GC) requires the implementation and generation of these
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5484intrinsics. These intrinsics allow identification of <a href="#int_gcroot">GC
5485roots on the stack</a>, as well as garbage collector implementations that
5486require <a href="#int_gcread">read</a> and <a href="#int_gcwrite">write</a>
5487barriers. Front-ends for type-safe garbage collected languages should generate
5488these intrinsics to make use of the LLVM garbage collectors. For more details,
5489see <a href="GarbageCollection.html">Accurate Garbage Collection with
5490LLVM</a>.</p>
Christopher Lamb303dae92007-12-17 01:00:21 +0000[diff] [blame]5491
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5492<p>The garbage collection intrinsics only operate on objects in the generic
5493 address space (address space zero).</p>
Christopher Lamb303dae92007-12-17 01:00:21 +0000[diff] [blame]5494
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5495</div>
5496
5497<!-- _______________________________________________________________________ -->
5498<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5499 <a name="int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5500</div>
5501
5502<div class="doc_text">
5503
5504<h5>Syntax:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5505<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]5506 declare void @llvm.gcroot(i8** %ptrloc, i8* %metadata)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5507</pre>
5508
5509<h5>Overview:</h5>
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]5510<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]5511 the code generator, and allows some metadata to be associated with it.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5512
5513<h5>Arguments:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5514<p>The first argument specifies the address of a stack object that contains the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5515 root pointer. The second pointer (which must be either a constant or a
5516 global value address) contains the meta-data to be associated with the
5517 root.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5518
5519<h5>Semantics:</h5>
Chris Lattner05d67092008-04-24 05:59:56 +0000[diff] [blame]5520<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]5521 location. At compile-time, the code generator generates information to allow
5522 the runtime to find the pointer at GC safe points. The '<tt>llvm.gcroot</tt>'
5523 intrinsic may only be used in a function which <a href="#gc">specifies a GC
5524 algorithm</a>.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5525
5526</div>
5527
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5528<!-- _______________________________________________________________________ -->
5529<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5530 <a name="int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5531</div>
5532
5533<div class="doc_text">
5534
5535<h5>Syntax:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5536<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]5537 declare i8* @llvm.gcread(i8* %ObjPtr, i8** %Ptr)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5538</pre>
5539
5540<h5>Overview:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5541<p>The '<tt>llvm.gcread</tt>' intrinsic identifies reads of references from heap
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5542 locations, allowing garbage collector implementations that require read
5543 barriers.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5544
5545<h5>Arguments:</h5>
Chris Lattner80626e92006-03-14 20:02:51 +0000[diff] [blame]5546<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]5547 allocated from the garbage collector. The first object is a pointer to the
5548 start of the referenced object, if needed by the language runtime (otherwise
5549 null).</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5550
5551<h5>Semantics:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5552<p>The '<tt>llvm.gcread</tt>' intrinsic has the same semantics as a load
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5553 instruction, but may be replaced with substantially more complex code by the
5554 garbage collector runtime, as needed. The '<tt>llvm.gcread</tt>' intrinsic
5555 may only be used in a function which <a href="#gc">specifies a GC
5556 algorithm</a>.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5557
5558</div>
5559
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5560<!-- _______________________________________________________________________ -->
5561<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5562 <a name="int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5563</div>
5564
5565<div class="doc_text">
5566
5567<h5>Syntax:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5568<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]5569 declare void @llvm.gcwrite(i8* %P1, i8* %Obj, i8** %P2)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5570</pre>
5571
5572<h5>Overview:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5573<p>The '<tt>llvm.gcwrite</tt>' intrinsic identifies writes of references to heap
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5574 locations, allowing garbage collector implementations that require write
5575 barriers (such as generational or reference counting collectors).</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5576
5577<h5>Arguments:</h5>
Chris Lattner80626e92006-03-14 20:02:51 +0000[diff] [blame]5578<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]5579 object to store it to, and the third is the address of the field of Obj to
5580 store to. If the runtime does not require a pointer to the object, Obj may
5581 be null.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5582
5583<h5>Semantics:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5584<p>The '<tt>llvm.gcwrite</tt>' intrinsic has the same semantics as a store
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5585 instruction, but may be replaced with substantially more complex code by the
5586 garbage collector runtime, as needed. The '<tt>llvm.gcwrite</tt>' intrinsic
5587 may only be used in a function which <a href="#gc">specifies a GC
5588 algorithm</a>.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5589
5590</div>
5591
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5592<!-- ======================================================================= -->
5593<div class="doc_subsection">
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5594 <a name="int_codegen">Code Generator Intrinsics</a>
5595</div>
5596
5597<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5598
5599<p>These intrinsics are provided by LLVM to expose special features that may
5600 only be implemented with code generator support.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5601
5602</div>
5603
5604<!-- _______________________________________________________________________ -->
5605<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5606 <a name="int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5607</div>
5608
5609<div class="doc_text">
5610
5611<h5>Syntax:</h5>
5612<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]5613 declare i8 *@llvm.returnaddress(i32 &lt;level&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5614</pre>
5615
5616<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5617<p>The '<tt>llvm.returnaddress</tt>' intrinsic attempts to compute a
5618 target-specific value indicating the return address of the current function
5619 or one of its callers.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5620
5621<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5622<p>The argument to this intrinsic indicates which function to return the address
5623 for. Zero indicates the calling function, one indicates its caller, etc.
5624 The argument is <b>required</b> to be a constant integer value.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5625
5626<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5627<p>The '<tt>llvm.returnaddress</tt>' intrinsic either returns a pointer
5628 indicating the return address of the specified call frame, or zero if it
5629 cannot be identified. The value returned by this intrinsic is likely to be
5630 incorrect or 0 for arguments other than zero, so it should only be used for
5631 debugging purposes.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5632
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5633<p>Note that calling this intrinsic does not prevent function inlining or other
5634 aggressive transformations, so the value returned may not be that of the
5635 obvious source-language caller.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5636
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5637</div>
5638
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5639<!-- _______________________________________________________________________ -->
5640<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5641 <a name="int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5642</div>
5643
5644<div class="doc_text">
5645
5646<h5>Syntax:</h5>
5647<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]5648 declare i8 *@llvm.frameaddress(i32 &lt;level&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5649</pre>
5650
5651<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5652<p>The '<tt>llvm.frameaddress</tt>' intrinsic attempts to return the
5653 target-specific frame pointer value for the specified stack frame.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5654
5655<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5656<p>The argument to this intrinsic indicates which function to return the frame
5657 pointer for. Zero indicates the calling function, one indicates its caller,
5658 etc. The argument is <b>required</b> to be a constant integer value.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5659
5660<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5661<p>The '<tt>llvm.frameaddress</tt>' intrinsic either returns a pointer
5662 indicating the frame address of the specified call frame, or zero if it
5663 cannot be identified. The value returned by this intrinsic is likely to be
5664 incorrect or 0 for arguments other than zero, so it should only be used for
5665 debugging purposes.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5666
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5667<p>Note that calling this intrinsic does not prevent function inlining or other
5668 aggressive transformations, so the value returned may not be that of the
5669 obvious source-language caller.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5670
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5671</div>
5672
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]5673<!-- _______________________________________________________________________ -->
5674<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5675 <a name="int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]5676</div>
5677
5678<div class="doc_text">
5679
5680<h5>Syntax:</h5>
5681<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]5682 declare i8 *@llvm.stacksave()
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]5683</pre>
5684
5685<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5686<p>The '<tt>llvm.stacksave</tt>' intrinsic is used to remember the current state
5687 of the function stack, for use
5688 with <a href="#int_stackrestore"> <tt>llvm.stackrestore</tt></a>. This is
5689 useful for implementing language features like scoped automatic variable
5690 sized arrays in C99.</p>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]5691
5692<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5693<p>This intrinsic returns a opaque pointer value that can be passed
5694 to <a href="#int_stackrestore"><tt>llvm.stackrestore</tt></a>. When
5695 an <tt>llvm.stackrestore</tt> intrinsic is executed with a value saved
5696 from <tt>llvm.stacksave</tt>, it effectively restores the state of the stack
5697 to the state it was in when the <tt>llvm.stacksave</tt> intrinsic executed.
5698 In practice, this pops any <a href="#i_alloca">alloca</a> blocks from the
5699 stack that were allocated after the <tt>llvm.stacksave</tt> was executed.</p>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]5700
5701</div>
5702
5703<!-- _______________________________________________________________________ -->
5704<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5705 <a name="int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]5706</div>
5707
5708<div class="doc_text">
5709
5710<h5>Syntax:</h5>
5711<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]5712 declare void @llvm.stackrestore(i8 * %ptr)
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]5713</pre>
5714
5715<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5716<p>The '<tt>llvm.stackrestore</tt>' intrinsic is used to restore the state of
5717 the function stack to the state it was in when the
5718 corresponding <a href="#int_stacksave"><tt>llvm.stacksave</tt></a> intrinsic
5719 executed. This is useful for implementing language features like scoped
5720 automatic variable sized arrays in C99.</p>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]5721
5722<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5723<p>See the description
5724 for <a href="#int_stacksave"><tt>llvm.stacksave</tt></a>.</p>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]5725
5726</div>
5727
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]5728<!-- _______________________________________________________________________ -->
5729<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5730 <a name="int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a>
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]5731</div>
5732
5733<div class="doc_text">
5734
5735<h5>Syntax:</h5>
5736<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]5737 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]5738</pre>
5739
5740<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5741<p>The '<tt>llvm.prefetch</tt>' intrinsic is a hint to the code generator to
5742 insert a prefetch instruction if supported; otherwise, it is a noop.
5743 Prefetches have no effect on the behavior of the program but can change its
5744 performance characteristics.</p>
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]5745
5746<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5747<p><tt>address</tt> is the address to be prefetched, <tt>rw</tt> is the
5748 specifier determining if the fetch should be for a read (0) or write (1),
5749 and <tt>locality</tt> is a temporal locality specifier ranging from (0) - no
5750 locality, to (3) - extremely local keep in cache. The <tt>rw</tt>
5751 and <tt>locality</tt> arguments must be constant integers.</p>
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]5752
5753<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5754<p>This intrinsic does not modify the behavior of the program. In particular,
5755 prefetches cannot trap and do not produce a value. On targets that support
5756 this intrinsic, the prefetch can provide hints to the processor cache for
5757 better performance.</p>
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]5758
5759</div>
5760
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]5761<!-- _______________________________________________________________________ -->
5762<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5763 <a name="int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a>
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]5764</div>
5765
5766<div class="doc_text">
5767
5768<h5>Syntax:</h5>
5769<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]5770 declare void @llvm.pcmarker(i32 &lt;id&gt;)
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]5771</pre>
5772
5773<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5774<p>The '<tt>llvm.pcmarker</tt>' intrinsic is a method to export a Program
5775 Counter (PC) in a region of code to simulators and other tools. The method
5776 is target specific, but it is expected that the marker will use exported
5777 symbols to transmit the PC of the marker. The marker makes no guarantees
5778 that it will remain with any specific instruction after optimizations. It is
5779 possible that the presence of a marker will inhibit optimizations. The
5780 intended use is to be inserted after optimizations to allow correlations of
5781 simulation runs.</p>
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]5782
5783<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5784<p><tt>id</tt> is a numerical id identifying the marker.</p>
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]5785
5786<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5787<p>This intrinsic does not modify the behavior of the program. Backends that do
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]5788 not support this intrinsic may ignore it.</p>
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]5789
5790</div>
5791
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]5792<!-- _______________________________________________________________________ -->
5793<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5794 <a name="int_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a>
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]5795</div>
5796
5797<div class="doc_text">
5798
5799<h5>Syntax:</h5>
5800<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]5801 declare i64 @llvm.readcyclecounter( )
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]5802</pre>
5803
5804<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5805<p>The '<tt>llvm.readcyclecounter</tt>' intrinsic provides access to the cycle
5806 counter register (or similar low latency, high accuracy clocks) on those
5807 targets that support it. On X86, it should map to RDTSC. On Alpha, it
5808 should map to RPCC. As the backing counters overflow quickly (on the order
5809 of 9 seconds on alpha), this should only be used for small timings.</p>
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]5810
5811<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5812<p>When directly supported, reading the cycle counter should not modify any
5813 memory. Implementations are allowed to either return a application specific
5814 value or a system wide value. On backends without support, this is lowered
5815 to a constant 0.</p>
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]5816
5817</div>
5818
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5819<!-- ======================================================================= -->
5820<div class="doc_subsection">
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5821 <a name="int_libc">Standard C Library Intrinsics</a>
5822</div>
5823
5824<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5825
5826<p>LLVM provides intrinsics for a few important standard C library functions.
5827 These intrinsics allow source-language front-ends to pass information about
5828 the alignment of the pointer arguments to the code generator, providing
5829 opportunity for more efficient code generation.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5830
5831</div>
5832
5833<!-- _______________________________________________________________________ -->
5834<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5835 <a name="int_memcpy">'<tt>llvm.memcpy</tt>' Intrinsic</a>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5836</div>
5837
5838<div class="doc_text">
5839
5840<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5841<p>This is an overloaded intrinsic. You can use <tt>llvm.memcpy</tt> on any
5842 integer bit width. Not all targets support all bit widths however.</p>
5843
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5844<pre>
Chris Lattner824b9582008-11-21 16:42:48 +0000[diff] [blame]5845 declare void @llvm.memcpy.i8(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5846 i8 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattner824b9582008-11-21 16:42:48 +0000[diff] [blame]5847 declare void @llvm.memcpy.i16(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
5848 i16 &lt;len&gt;, i32 &lt;align&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]5849 declare void @llvm.memcpy.i32(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5850 i32 &lt;len&gt;, i32 &lt;align&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]5851 declare void @llvm.memcpy.i64(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5852 i64 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattner33aec9e2004-02-12 17:01:32 +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.memcpy.*</tt>' intrinsics copy a block of memory from the
5857 source location to the destination location.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5858
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5859<p>Note that, unlike the standard libc function, the <tt>llvm.memcpy.*</tt>
5860 intrinsics do not return a value, and takes an extra alignment argument.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5861
5862<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5863<p>The first argument is a pointer to the destination, the second is a pointer
5864 to the source. The third argument is an integer argument specifying the
5865 number of bytes to copy, and the fourth argument is the alignment of the
5866 source and destination locations.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5867
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]5868<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]5869 then the caller guarantees that both the source and destination pointers are
5870 aligned to that boundary.</p>
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]5871
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5872<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5873<p>The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the
5874 source location to the destination location, which are not allowed to
5875 overlap. It copies "len" bytes of memory over. If the argument is known to
5876 be aligned to some boundary, this can be specified as the fourth argument,
5877 otherwise it should be set to 0 or 1.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5878
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5879</div>
5880
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]5881<!-- _______________________________________________________________________ -->
5882<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5883 <a name="int_memmove">'<tt>llvm.memmove</tt>' Intrinsic</a>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]5884</div>
5885
5886<div class="doc_text">
5887
5888<h5>Syntax:</h5>
Chris Lattner824b9582008-11-21 16:42:48 +0000[diff] [blame]5889<p>This is an overloaded intrinsic. You can use llvm.memmove on any integer bit
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5890 width. Not all targets support all bit widths however.</p>
5891
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]5892<pre>
Chris Lattner824b9582008-11-21 16:42:48 +0000[diff] [blame]5893 declare void @llvm.memmove.i8(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5894 i8 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattner824b9582008-11-21 16:42:48 +0000[diff] [blame]5895 declare void @llvm.memmove.i16(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
5896 i16 &lt;len&gt;, i32 &lt;align&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]5897 declare void @llvm.memmove.i32(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5898 i32 &lt;len&gt;, i32 &lt;align&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]5899 declare void @llvm.memmove.i64(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5900 i64 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]5901</pre>
5902
5903<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5904<p>The '<tt>llvm.memmove.*</tt>' intrinsics move a block of memory from the
5905 source location to the destination location. It is similar to the
5906 '<tt>llvm.memcpy</tt>' intrinsic but allows the two memory locations to
5907 overlap.</p>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]5908
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5909<p>Note that, unlike the standard libc function, the <tt>llvm.memmove.*</tt>
5910 intrinsics do not return a value, and takes an extra alignment argument.</p>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]5911
5912<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5913<p>The first argument is a pointer to the destination, the second is a pointer
5914 to the source. The third argument is an integer argument specifying the
5915 number of bytes to copy, and the fourth argument is the alignment of the
5916 source and destination locations.</p>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]5917
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]5918<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]5919 then the caller guarantees that the source and destination pointers are
5920 aligned to that boundary.</p>
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]5921
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]5922<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5923<p>The '<tt>llvm.memmove.*</tt>' intrinsics copy a block of memory from the
5924 source location to the destination location, which may overlap. It copies
5925 "len" bytes of memory over. If the argument is known to be aligned to some
5926 boundary, this can be specified as the fourth argument, otherwise it should
5927 be set to 0 or 1.</p>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]5928
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]5929</div>
5930
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5931<!-- _______________________________________________________________________ -->
5932<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5933 <a name="int_memset">'<tt>llvm.memset.*</tt>' Intrinsics</a>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5934</div>
5935
5936<div class="doc_text">
5937
5938<h5>Syntax:</h5>
Chris Lattner824b9582008-11-21 16:42:48 +0000[diff] [blame]5939<p>This is an overloaded intrinsic. You can use llvm.memset on any integer bit
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5940 width. Not all targets support all bit widths however.</p>
5941
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5942<pre>
Chris Lattner824b9582008-11-21 16:42:48 +0000[diff] [blame]5943 declare void @llvm.memset.i8(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5944 i8 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattner824b9582008-11-21 16:42:48 +0000[diff] [blame]5945 declare void @llvm.memset.i16(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
5946 i16 &lt;len&gt;, i32 &lt;align&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]5947 declare void @llvm.memset.i32(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5948 i32 &lt;len&gt;, i32 &lt;align&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]5949 declare void @llvm.memset.i64(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5950 i64 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5951</pre>
5952
5953<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5954<p>The '<tt>llvm.memset.*</tt>' intrinsics fill a block of memory with a
5955 particular byte value.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5956
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5957<p>Note that, unlike the standard libc function, the <tt>llvm.memset</tt>
5958 intrinsic does not return a value, and takes an extra alignment argument.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5959
5960<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5961<p>The first argument is a pointer to the destination to fill, the second is the
5962 byte value to fill it with, the third argument is an integer argument
5963 specifying the number of bytes to fill, and the fourth argument is the known
5964 alignment of destination location.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5965
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]5966<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]5967 then the caller guarantees that the destination pointer is aligned to that
5968 boundary.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5969
5970<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5971<p>The '<tt>llvm.memset.*</tt>' intrinsics fill "len" bytes of memory starting
5972 at the destination location. If the argument is known to be aligned to some
5973 boundary, this can be specified as the fourth argument, otherwise it should
5974 be set to 0 or 1.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5975
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5976</div>
5977
Chris Lattner32006282004-06-11 02:28:03 +0000[diff] [blame]5978<!-- _______________________________________________________________________ -->
5979<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5980 <a name="int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a>
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]5981</div>
5982
5983<div class="doc_text">
5984
5985<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5986<p>This is an overloaded intrinsic. You can use <tt>llvm.sqrt</tt> on any
5987 floating point or vector of floating point type. Not all targets support all
5988 types however.</p>
5989
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]5990<pre>
Dale Johannesen408f9c12007-10-02 17:47:38 +0000[diff] [blame]5991 declare float @llvm.sqrt.f32(float %Val)
5992 declare double @llvm.sqrt.f64(double %Val)
5993 declare x86_fp80 @llvm.sqrt.f80(x86_fp80 %Val)
5994 declare fp128 @llvm.sqrt.f128(fp128 %Val)
5995 declare ppc_fp128 @llvm.sqrt.ppcf128(ppc_fp128 %Val)
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]5996</pre>
5997
5998<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5999<p>The '<tt>llvm.sqrt</tt>' intrinsics return the sqrt of the specified operand,
6000 returning the same value as the libm '<tt>sqrt</tt>' functions would.
6001 Unlike <tt>sqrt</tt> in libm, however, <tt>llvm.sqrt</tt> has undefined
6002 behavior for negative numbers other than -0.0 (which allows for better
6003 optimization, because there is no need to worry about errno being
6004 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]6005
6006<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6007<p>The argument and return value are floating point numbers of the same
6008 type.</p>
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]6009
6010<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6011<p>This function returns the sqrt of the specified operand if it is a
6012 nonnegative floating point number.</p>
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]6013
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]6014</div>
6015
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6016<!-- _______________________________________________________________________ -->
6017<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6018 <a name="int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6019</div>
6020
6021<div class="doc_text">
6022
6023<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6024<p>This is an overloaded intrinsic. You can use <tt>llvm.powi</tt> on any
6025 floating point or vector of floating point type. Not all targets support all
6026 types however.</p>
6027
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6028<pre>
Dale Johannesen408f9c12007-10-02 17:47:38 +0000[diff] [blame]6029 declare float @llvm.powi.f32(float %Val, i32 %power)
6030 declare double @llvm.powi.f64(double %Val, i32 %power)
6031 declare x86_fp80 @llvm.powi.f80(x86_fp80 %Val, i32 %power)
6032 declare fp128 @llvm.powi.f128(fp128 %Val, i32 %power)
6033 declare ppc_fp128 @llvm.powi.ppcf128(ppc_fp128 %Val, i32 %power)
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6034</pre>
6035
6036<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6037<p>The '<tt>llvm.powi.*</tt>' intrinsics return the first operand raised to the
6038 specified (positive or negative) power. The order of evaluation of
6039 multiplications is not defined. When a vector of floating point type is
6040 used, the second argument remains a scalar integer value.</p>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6041
6042<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6043<p>The second argument is an integer power, and the first is a value to raise to
6044 that power.</p>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6045
6046<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6047<p>This function returns the first value raised to the second power with an
6048 unspecified sequence of rounding operations.</p>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6049
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6050</div>
6051
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6052<!-- _______________________________________________________________________ -->
6053<div class="doc_subsubsection">
6054 <a name="int_sin">'<tt>llvm.sin.*</tt>' Intrinsic</a>
6055</div>
6056
6057<div class="doc_text">
6058
6059<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6060<p>This is an overloaded intrinsic. You can use <tt>llvm.sin</tt> on any
6061 floating point or vector of floating point type. Not all targets support all
6062 types however.</p>
6063
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6064<pre>
6065 declare float @llvm.sin.f32(float %Val)
6066 declare double @llvm.sin.f64(double %Val)
6067 declare x86_fp80 @llvm.sin.f80(x86_fp80 %Val)
6068 declare fp128 @llvm.sin.f128(fp128 %Val)
6069 declare ppc_fp128 @llvm.sin.ppcf128(ppc_fp128 %Val)
6070</pre>
6071
6072<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6073<p>The '<tt>llvm.sin.*</tt>' intrinsics return the sine of the operand.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6074
6075<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6076<p>The argument and return value are floating point numbers of the same
6077 type.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6078
6079<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6080<p>This function returns the sine of the specified operand, returning the same
6081 values as the libm <tt>sin</tt> functions would, and handles error conditions
6082 in the same way.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6083
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6084</div>
6085
6086<!-- _______________________________________________________________________ -->
6087<div class="doc_subsubsection">
6088 <a name="int_cos">'<tt>llvm.cos.*</tt>' Intrinsic</a>
6089</div>
6090
6091<div class="doc_text">
6092
6093<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6094<p>This is an overloaded intrinsic. You can use <tt>llvm.cos</tt> on any
6095 floating point or vector of floating point type. Not all targets support all
6096 types however.</p>
6097
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6098<pre>
6099 declare float @llvm.cos.f32(float %Val)
6100 declare double @llvm.cos.f64(double %Val)
6101 declare x86_fp80 @llvm.cos.f80(x86_fp80 %Val)
6102 declare fp128 @llvm.cos.f128(fp128 %Val)
6103 declare ppc_fp128 @llvm.cos.ppcf128(ppc_fp128 %Val)
6104</pre>
6105
6106<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6107<p>The '<tt>llvm.cos.*</tt>' intrinsics return the cosine of the operand.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6108
6109<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6110<p>The argument and return value are floating point numbers of the same
6111 type.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6112
6113<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6114<p>This function returns the cosine of the specified operand, returning the same
6115 values as the libm <tt>cos</tt> functions would, and handles error conditions
6116 in the same way.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6117
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6118</div>
6119
6120<!-- _______________________________________________________________________ -->
6121<div class="doc_subsubsection">
6122 <a name="int_pow">'<tt>llvm.pow.*</tt>' Intrinsic</a>
6123</div>
6124
6125<div class="doc_text">
6126
6127<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6128<p>This is an overloaded intrinsic. You can use <tt>llvm.pow</tt> on any
6129 floating point or vector of floating point type. Not all targets support all
6130 types however.</p>
6131
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6132<pre>
6133 declare float @llvm.pow.f32(float %Val, float %Power)
6134 declare double @llvm.pow.f64(double %Val, double %Power)
6135 declare x86_fp80 @llvm.pow.f80(x86_fp80 %Val, x86_fp80 %Power)
6136 declare fp128 @llvm.pow.f128(fp128 %Val, fp128 %Power)
6137 declare ppc_fp128 @llvm.pow.ppcf128(ppc_fp128 %Val, ppc_fp128 Power)
6138</pre>
6139
6140<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6141<p>The '<tt>llvm.pow.*</tt>' intrinsics return the first operand raised to the
6142 specified (positive or negative) power.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6143
6144<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6145<p>The second argument is a floating point power, and the first is a value to
6146 raise to that power.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6147
6148<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6149<p>This function returns the first value raised to the second power, returning
6150 the same values as the libm <tt>pow</tt> functions would, and handles error
6151 conditions in the same way.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6152
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6153</div>
6154
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6155<!-- ======================================================================= -->
6156<div class="doc_subsection">
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]6157 <a name="int_manip">Bit Manipulation Intrinsics</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6158</div>
6159
6160<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6161
6162<p>LLVM provides intrinsics for a few important bit manipulation operations.
6163 These allow efficient code generation for some algorithms.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6164
6165</div>
6166
6167<!-- _______________________________________________________________________ -->
6168<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6169 <a name="int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]6170</div>
6171
6172<div class="doc_text">
6173
6174<h5>Syntax:</h5>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]6175<p>This is an overloaded intrinsic function. You can use bswap on any integer
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6176 type that is an even number of bytes (i.e. BitWidth % 16 == 0).</p>
6177
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]6178<pre>
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]6179 declare i16 @llvm.bswap.i16(i16 &lt;id&gt;)
6180 declare i32 @llvm.bswap.i32(i32 &lt;id&gt;)
6181 declare i64 @llvm.bswap.i64(i64 &lt;id&gt;)
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]6182</pre>
6183
6184<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6185<p>The '<tt>llvm.bswap</tt>' family of intrinsics is used to byte swap integer
6186 values with an even number of bytes (positive multiple of 16 bits). These
6187 are useful for performing operations on data that is not in the target's
6188 native byte order.</p>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]6189
6190<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6191<p>The <tt>llvm.bswap.i16</tt> intrinsic returns an i16 value that has the high
6192 and low byte of the input i16 swapped. Similarly,
6193 the <tt>llvm.bswap.i32</tt> intrinsic returns an i32 value that has the four
6194 bytes of the input i32 swapped, so that if the input bytes are numbered 0, 1,
6195 2, 3 then the returned i32 will have its bytes in 3, 2, 1, 0 order.
6196 The <tt>llvm.bswap.i48</tt>, <tt>llvm.bswap.i64</tt> and other intrinsics
6197 extend this concept to additional even-byte lengths (6 bytes, 8 bytes and
6198 more, respectively).</p>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]6199
6200</div>
6201
6202<!-- _______________________________________________________________________ -->
6203<div class="doc_subsubsection">
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]6204 <a name="int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6205</div>
6206
6207<div class="doc_text">
6208
6209<h5>Syntax:</h5>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]6210<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]6211 width. Not all targets support all bit widths however.</p>
6212
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6213<pre>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6214 declare i8 @llvm.ctpop.i8(i8 &lt;src&gt;)
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]6215 declare i16 @llvm.ctpop.i16(i16 &lt;src&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]6216 declare i32 @llvm.ctpop.i32(i32 &lt;src&gt;)
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]6217 declare i64 @llvm.ctpop.i64(i64 &lt;src&gt;)
6218 declare i256 @llvm.ctpop.i256(i256 &lt;src&gt;)
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6219</pre>
6220
6221<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6222<p>The '<tt>llvm.ctpop</tt>' family of intrinsics counts the number of bits set
6223 in a value.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6224
6225<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6226<p>The only argument is the value to be counted. The argument may be of any
6227 integer type. The return type must match the argument type.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6228
6229<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6230<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]6231
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6232</div>
6233
6234<!-- _______________________________________________________________________ -->
6235<div class="doc_subsubsection">
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]6236 <a name="int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6237</div>
6238
6239<div class="doc_text">
6240
6241<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6242<p>This is an overloaded intrinsic. You can use <tt>llvm.ctlz</tt> on any
6243 integer bit width. Not all targets support all bit widths however.</p>
6244
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6245<pre>
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]6246 declare i8 @llvm.ctlz.i8 (i8 &lt;src&gt;)
6247 declare i16 @llvm.ctlz.i16(i16 &lt;src&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]6248 declare i32 @llvm.ctlz.i32(i32 &lt;src&gt;)
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]6249 declare i64 @llvm.ctlz.i64(i64 &lt;src&gt;)
6250 declare i256 @llvm.ctlz.i256(i256 &lt;src&gt;)
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6251</pre>
6252
6253<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6254<p>The '<tt>llvm.ctlz</tt>' family of intrinsic functions counts the number of
6255 leading zeros in a variable.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6256
6257<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6258<p>The only argument is the value to be counted. The argument may be of any
6259 integer type. The return type must match the argument type.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6260
6261<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6262<p>The '<tt>llvm.ctlz</tt>' intrinsic counts the leading (most significant)
6263 zeros in a variable. If the src == 0 then the result is the size in bits of
6264 the type of src. For example, <tt>llvm.ctlz(i32 2) = 30</tt>.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6265
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6266</div>
Chris Lattner32006282004-06-11 02:28:03 +0000[diff] [blame]6267
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]6268<!-- _______________________________________________________________________ -->
6269<div class="doc_subsubsection">
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]6270 <a name="int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic</a>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]6271</div>
6272
6273<div class="doc_text">
6274
6275<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6276<p>This is an overloaded intrinsic. You can use <tt>llvm.cttz</tt> on any
6277 integer bit width. Not all targets support all bit widths however.</p>
6278
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]6279<pre>
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]6280 declare i8 @llvm.cttz.i8 (i8 &lt;src&gt;)
6281 declare i16 @llvm.cttz.i16(i16 &lt;src&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]6282 declare i32 @llvm.cttz.i32(i32 &lt;src&gt;)
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]6283 declare i64 @llvm.cttz.i64(i64 &lt;src&gt;)
6284 declare i256 @llvm.cttz.i256(i256 &lt;src&gt;)
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]6285</pre>
6286
6287<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6288<p>The '<tt>llvm.cttz</tt>' family of intrinsic functions counts the number of
6289 trailing zeros.</p>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]6290
6291<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6292<p>The only argument is the value to be counted. The argument may be of any
6293 integer type. The return type must match the argument type.</p>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]6294
6295<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6296<p>The '<tt>llvm.cttz</tt>' intrinsic counts the trailing (least significant)
6297 zeros in a variable. If the src == 0 then the result is the size in bits of
6298 the type of src. For example, <tt>llvm.cttz(2) = 1</tt>.</p>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]6299
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]6300</div>
6301
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]6302<!-- ======================================================================= -->
6303<div class="doc_subsection">
6304 <a name="int_overflow">Arithmetic with Overflow Intrinsics</a>
6305</div>
6306
6307<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6308
6309<p>LLVM provides intrinsics for some arithmetic with overflow operations.</p>
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]6310
6311</div>
6312
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6313<!-- _______________________________________________________________________ -->
6314<div class="doc_subsubsection">
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]6315 <a name="int_sadd_overflow">'<tt>llvm.sadd.with.overflow.*</tt>' Intrinsics</a>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6316</div>
6317
6318<div class="doc_text">
6319
6320<h5>Syntax:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6321<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]6322 on any integer bit width.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6323
6324<pre>
6325 declare {i16, i1} @llvm.sadd.with.overflow.i16(i16 %a, i16 %b)
6326 declare {i32, i1} @llvm.sadd.with.overflow.i32(i32 %a, i32 %b)
6327 declare {i64, i1} @llvm.sadd.with.overflow.i64(i64 %a, i64 %b)
6328</pre>
6329
6330<h5>Overview:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6331<p>The '<tt>llvm.sadd.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6332 a signed addition of the two arguments, and indicate whether an overflow
6333 occurred during the signed summation.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6334
6335<h5>Arguments:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6336<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]6337 be of integer types of any bit width, but they must have the same bit
6338 width. The second element of the result structure must be of
6339 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
6340 undergo signed addition.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6341
6342<h5>Semantics:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6343<p>The '<tt>llvm.sadd.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6344 a signed addition of the two variables. They return a structure &mdash; the
6345 first element of which is the signed summation, and the second element of
6346 which is a bit specifying if the signed summation resulted in an
6347 overflow.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6348
6349<h5>Examples:</h5>
6350<pre>
6351 %res = call {i32, i1} @llvm.sadd.with.overflow.i32(i32 %a, i32 %b)
6352 %sum = extractvalue {i32, i1} %res, 0
6353 %obit = extractvalue {i32, i1} %res, 1
6354 br i1 %obit, label %overflow, label %normal
6355</pre>
6356
6357</div>
6358
6359<!-- _______________________________________________________________________ -->
6360<div class="doc_subsubsection">
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]6361 <a name="int_uadd_overflow">'<tt>llvm.uadd.with.overflow.*</tt>' Intrinsics</a>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6362</div>
6363
6364<div class="doc_text">
6365
6366<h5>Syntax:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6367<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]6368 on any integer bit width.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6369
6370<pre>
6371 declare {i16, i1} @llvm.uadd.with.overflow.i16(i16 %a, i16 %b)
6372 declare {i32, i1} @llvm.uadd.with.overflow.i32(i32 %a, i32 %b)
6373 declare {i64, i1} @llvm.uadd.with.overflow.i64(i64 %a, i64 %b)
6374</pre>
6375
6376<h5>Overview:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6377<p>The '<tt>llvm.uadd.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6378 an unsigned addition of the two arguments, and indicate whether a carry
6379 occurred during the unsigned summation.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6380
6381<h5>Arguments:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6382<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]6383 be of integer types of any bit width, but they must have the same bit
6384 width. The second element of the result structure must be of
6385 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
6386 undergo unsigned addition.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6387
6388<h5>Semantics:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6389<p>The '<tt>llvm.uadd.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6390 an unsigned addition of the two arguments. They return a structure &mdash;
6391 the first element of which is the sum, and the second element of which is a
6392 bit specifying if the unsigned summation resulted in a carry.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6393
6394<h5>Examples:</h5>
6395<pre>
6396 %res = call {i32, i1} @llvm.uadd.with.overflow.i32(i32 %a, i32 %b)
6397 %sum = extractvalue {i32, i1} %res, 0
6398 %obit = extractvalue {i32, i1} %res, 1
6399 br i1 %obit, label %carry, label %normal
6400</pre>
6401
6402</div>
6403
6404<!-- _______________________________________________________________________ -->
6405<div class="doc_subsubsection">
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]6406 <a name="int_ssub_overflow">'<tt>llvm.ssub.with.overflow.*</tt>' Intrinsics</a>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6407</div>
6408
6409<div class="doc_text">
6410
6411<h5>Syntax:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6412<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]6413 on any integer bit width.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6414
6415<pre>
6416 declare {i16, i1} @llvm.ssub.with.overflow.i16(i16 %a, i16 %b)
6417 declare {i32, i1} @llvm.ssub.with.overflow.i32(i32 %a, i32 %b)
6418 declare {i64, i1} @llvm.ssub.with.overflow.i64(i64 %a, i64 %b)
6419</pre>
6420
6421<h5>Overview:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6422<p>The '<tt>llvm.ssub.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6423 a signed subtraction of the two arguments, and indicate whether an overflow
6424 occurred during the signed subtraction.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6425
6426<h5>Arguments:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6427<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]6428 be of integer types of any bit width, but they must have the same bit
6429 width. The second element of the result structure must be of
6430 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
6431 undergo signed subtraction.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6432
6433<h5>Semantics:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6434<p>The '<tt>llvm.ssub.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6435 a signed subtraction of the two arguments. They return a structure &mdash;
6436 the first element of which is the subtraction, and the second element of
6437 which is a bit specifying if the signed subtraction resulted in an
6438 overflow.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6439
6440<h5>Examples:</h5>
6441<pre>
6442 %res = call {i32, i1} @llvm.ssub.with.overflow.i32(i32 %a, i32 %b)
6443 %sum = extractvalue {i32, i1} %res, 0
6444 %obit = extractvalue {i32, i1} %res, 1
6445 br i1 %obit, label %overflow, label %normal
6446</pre>
6447
6448</div>
6449
6450<!-- _______________________________________________________________________ -->
6451<div class="doc_subsubsection">
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]6452 <a name="int_usub_overflow">'<tt>llvm.usub.with.overflow.*</tt>' Intrinsics</a>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6453</div>
6454
6455<div class="doc_text">
6456
6457<h5>Syntax:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6458<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]6459 on any integer bit width.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6460
6461<pre>
6462 declare {i16, i1} @llvm.usub.with.overflow.i16(i16 %a, i16 %b)
6463 declare {i32, i1} @llvm.usub.with.overflow.i32(i32 %a, i32 %b)
6464 declare {i64, i1} @llvm.usub.with.overflow.i64(i64 %a, i64 %b)
6465</pre>
6466
6467<h5>Overview:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6468<p>The '<tt>llvm.usub.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6469 an unsigned subtraction of the two arguments, and indicate whether an
6470 overflow occurred during the unsigned subtraction.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6471
6472<h5>Arguments:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6473<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]6474 be of integer types of any bit width, but they must have the same bit
6475 width. The second element of the result structure must be of
6476 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
6477 undergo unsigned subtraction.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6478
6479<h5>Semantics:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6480<p>The '<tt>llvm.usub.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6481 an unsigned subtraction of the two arguments. They return a structure &mdash;
6482 the first element of which is the subtraction, and the second element of
6483 which is a bit specifying if the unsigned subtraction resulted in an
6484 overflow.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6485
6486<h5>Examples:</h5>
6487<pre>
6488 %res = call {i32, i1} @llvm.usub.with.overflow.i32(i32 %a, i32 %b)
6489 %sum = extractvalue {i32, i1} %res, 0
6490 %obit = extractvalue {i32, i1} %res, 1
6491 br i1 %obit, label %overflow, label %normal
6492</pre>
6493
6494</div>
6495
6496<!-- _______________________________________________________________________ -->
6497<div class="doc_subsubsection">
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]6498 <a name="int_smul_overflow">'<tt>llvm.smul.with.overflow.*</tt>' Intrinsics</a>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6499</div>
6500
6501<div class="doc_text">
6502
6503<h5>Syntax:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6504<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]6505 on any integer bit width.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6506
6507<pre>
6508 declare {i16, i1} @llvm.smul.with.overflow.i16(i16 %a, i16 %b)
6509 declare {i32, i1} @llvm.smul.with.overflow.i32(i32 %a, i32 %b)
6510 declare {i64, i1} @llvm.smul.with.overflow.i64(i64 %a, i64 %b)
6511</pre>
6512
6513<h5>Overview:</h5>
6514
6515<p>The '<tt>llvm.smul.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6516 a signed multiplication of the two arguments, and indicate whether an
6517 overflow occurred during the signed multiplication.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6518
6519<h5>Arguments:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6520<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]6521 be of integer types of any bit width, but they must have the same bit
6522 width. The second element of the result structure must be of
6523 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
6524 undergo signed multiplication.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6525
6526<h5>Semantics:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6527<p>The '<tt>llvm.smul.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6528 a signed multiplication of the two arguments. They return a structure &mdash;
6529 the first element of which is the multiplication, and the second element of
6530 which is a bit specifying if the signed multiplication resulted in an
6531 overflow.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6532
6533<h5>Examples:</h5>
6534<pre>
6535 %res = call {i32, i1} @llvm.smul.with.overflow.i32(i32 %a, i32 %b)
6536 %sum = extractvalue {i32, i1} %res, 0
6537 %obit = extractvalue {i32, i1} %res, 1
6538 br i1 %obit, label %overflow, label %normal
6539</pre>
6540
Reid Spencerf86037f2007-04-11 23:23:49 +0000[diff] [blame]6541</div>
6542
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]6543<!-- _______________________________________________________________________ -->
6544<div class="doc_subsubsection">
6545 <a name="int_umul_overflow">'<tt>llvm.umul.with.overflow.*</tt>' Intrinsics</a>
6546</div>
6547
6548<div class="doc_text">
6549
6550<h5>Syntax:</h5>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]6551<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]6552 on any integer bit width.</p>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]6553
6554<pre>
6555 declare {i16, i1} @llvm.umul.with.overflow.i16(i16 %a, i16 %b)
6556 declare {i32, i1} @llvm.umul.with.overflow.i32(i32 %a, i32 %b)
6557 declare {i64, i1} @llvm.umul.with.overflow.i64(i64 %a, i64 %b)
6558</pre>
6559
6560<h5>Overview:</h5>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]6561<p>The '<tt>llvm.umul.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6562 a unsigned multiplication of the two arguments, and indicate whether an
6563 overflow occurred during the unsigned multiplication.</p>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]6564
6565<h5>Arguments:</h5>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]6566<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]6567 be of integer types of any bit width, but they must have the same bit
6568 width. The second element of the result structure must be of
6569 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
6570 undergo unsigned multiplication.</p>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]6571
6572<h5>Semantics:</h5>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]6573<p>The '<tt>llvm.umul.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6574 an unsigned multiplication of the two arguments. They return a structure
6575 &mdash; the first element of which is the multiplication, and the second
6576 element of which is a bit specifying if the unsigned multiplication resulted
6577 in an overflow.</p>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]6578
6579<h5>Examples:</h5>
6580<pre>
6581 %res = call {i32, i1} @llvm.umul.with.overflow.i32(i32 %a, i32 %b)
6582 %sum = extractvalue {i32, i1} %res, 0
6583 %obit = extractvalue {i32, i1} %res, 1
6584 br i1 %obit, label %overflow, label %normal
6585</pre>
6586
6587</div>
6588
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]6589<!-- ======================================================================= -->
6590<div class="doc_subsection">
6591 <a name="int_debugger">Debugger Intrinsics</a>
6592</div>
6593
6594<div class="doc_text">
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]6595
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6596<p>The LLVM debugger intrinsics (which all start with <tt>llvm.dbg.</tt>
6597 prefix), are described in
6598 the <a href="SourceLevelDebugging.html#format_common_intrinsics">LLVM Source
6599 Level Debugging</a> document.</p>
6600
6601</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]6602
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]6603<!-- ======================================================================= -->
6604<div class="doc_subsection">
6605 <a name="int_eh">Exception Handling Intrinsics</a>
6606</div>
6607
6608<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6609
6610<p>The LLVM exception handling intrinsics (which all start with
6611 <tt>llvm.eh.</tt> prefix), are described in
6612 the <a href="ExceptionHandling.html#format_common_intrinsics">LLVM Exception
6613 Handling</a> document.</p>
6614
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]6615</div>
6616
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]6617<!-- ======================================================================= -->
6618<div class="doc_subsection">
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]6619 <a name="int_trampoline">Trampoline Intrinsic</a>
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]6620</div>
6621
6622<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6623
6624<p>This intrinsic makes it possible to excise one parameter, marked with
6625 the <tt>nest</tt> attribute, from a function. The result is a callable
6626 function pointer lacking the nest parameter - the caller does not need to
6627 provide a value for it. Instead, the value to use is stored in advance in a
6628 "trampoline", a block of memory usually allocated on the stack, which also
6629 contains code to splice the nest value into the argument list. This is used
6630 to implement the GCC nested function address extension.</p>
6631
6632<p>For example, if the function is
6633 <tt>i32 f(i8* nest %c, i32 %x, i32 %y)</tt> then the resulting function
6634 pointer has signature <tt>i32 (i32, i32)*</tt>. It can be created as
6635 follows:</p>
6636
6637<div class="doc_code">
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]6638<pre>
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]6639 %tramp = alloca [10 x i8], align 4 ; size and alignment only correct for X86
6640 %tramp1 = getelementptr [10 x i8]* %tramp, i32 0, i32 0
6641 %p = call i8* @llvm.init.trampoline( i8* %tramp1, i8* bitcast (i32 (i8* nest , i32, i32)* @f to i8*), i8* %nval )
6642 %fp = bitcast i8* %p to i32 (i32, i32)*
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]6643</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6644</div>
6645
6646<p>The call <tt>%val = call i32 %fp( i32 %x, i32 %y )</tt> is then equivalent
6647 to <tt>%val = call i32 %f( i8* %nval, i32 %x, i32 %y )</tt>.</p>
6648
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]6649</div>
6650
6651<!-- _______________________________________________________________________ -->
6652<div class="doc_subsubsection">
6653 <a name="int_it">'<tt>llvm.init.trampoline</tt>' Intrinsic</a>
6654</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6655
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]6656<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6657
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]6658<h5>Syntax:</h5>
6659<pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6660 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]6661</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6662
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]6663<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6664<p>This fills the memory pointed to by <tt>tramp</tt> with code and returns a
6665 function pointer suitable for executing it.</p>
6666
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]6667<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6668<p>The <tt>llvm.init.trampoline</tt> intrinsic takes three arguments, all
6669 pointers. The <tt>tramp</tt> argument must point to a sufficiently large and
6670 sufficiently aligned block of memory; this memory is written to by the
6671 intrinsic. Note that the size and the alignment are target-specific - LLVM
6672 currently provides no portable way of determining them, so a front-end that
6673 generates this intrinsic needs to have some target-specific knowledge.
6674 The <tt>func</tt> argument must hold a function bitcast to
6675 an <tt>i8*</tt>.</p>
6676
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]6677<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6678<p>The block of memory pointed to by <tt>tramp</tt> is filled with target
6679 dependent code, turning it into a function. A pointer to this function is
6680 returned, but needs to be bitcast to an <a href="#int_trampoline">appropriate
6681 function pointer type</a> before being called. The new function's signature
6682 is the same as that of <tt>func</tt> with any arguments marked with
6683 the <tt>nest</tt> attribute removed. At most one such <tt>nest</tt> argument
6684 is allowed, and it must be of pointer type. Calling the new function is
6685 equivalent to calling <tt>func</tt> with the same argument list, but
6686 with <tt>nval</tt> used for the missing <tt>nest</tt> argument. If, after
6687 calling <tt>llvm.init.trampoline</tt>, the memory pointed to
6688 by <tt>tramp</tt> is modified, then the effect of any later call to the
6689 returned function pointer is undefined.</p>
6690
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]6691</div>
6692
6693<!-- ======================================================================= -->
6694<div class="doc_subsection">
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]6695 <a name="int_atomics">Atomic Operations and Synchronization Intrinsics</a>
6696</div>
6697
6698<div class="doc_text">
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]6699
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6700<p>These intrinsic functions expand the "universal IR" of LLVM to represent
6701 hardware constructs for atomic operations and memory synchronization. This
6702 provides an interface to the hardware, not an interface to the programmer. It
6703 is aimed at a low enough level to allow any programming models or APIs
6704 (Application Programming Interfaces) which need atomic behaviors to map
6705 cleanly onto it. It is also modeled primarily on hardware behavior. Just as
6706 hardware provides a "universal IR" for source languages, it also provides a
6707 starting point for developing a "universal" atomic operation and
6708 synchronization IR.</p>
6709
6710<p>These do <em>not</em> form an API such as high-level threading libraries,
6711 software transaction memory systems, atomic primitives, and intrinsic
6712 functions as found in BSD, GNU libc, atomic_ops, APR, and other system and
6713 application libraries. The hardware interface provided by LLVM should allow
6714 a clean implementation of all of these APIs and parallel programming models.
6715 No one model or paradigm should be selected above others unless the hardware
6716 itself ubiquitously does so.</p>
6717
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]6718</div>
6719
6720<!-- _______________________________________________________________________ -->
6721<div class="doc_subsubsection">
6722 <a name="int_memory_barrier">'<tt>llvm.memory.barrier</tt>' Intrinsic</a>
6723</div>
6724<div class="doc_text">
6725<h5>Syntax:</h5>
6726<pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6727 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]6728</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6729
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]6730<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6731<p>The <tt>llvm.memory.barrier</tt> intrinsic guarantees ordering between
6732 specific pairs of memory access types.</p>
6733
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]6734<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6735<p>The <tt>llvm.memory.barrier</tt> intrinsic requires five boolean arguments.
6736 The first four arguments enables a specific barrier as listed below. The
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]6737 fifth argument specifies that the barrier applies to io or device or uncached
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6738 memory.</p>
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]6739
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6740<ul>
6741 <li><tt>ll</tt>: load-load barrier</li>
6742 <li><tt>ls</tt>: load-store barrier</li>
6743 <li><tt>sl</tt>: store-load barrier</li>
6744 <li><tt>ss</tt>: store-store barrier</li>
6745 <li><tt>device</tt>: barrier applies to device and uncached memory also.</li>
6746</ul>
6747
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]6748<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6749<p>This intrinsic causes the system to enforce some ordering constraints upon
6750 the loads and stores of the program. This barrier does not
6751 indicate <em>when</em> any events will occur, it only enforces
6752 an <em>order</em> in which they occur. For any of the specified pairs of load
6753 and store operations (f.ex. load-load, or store-load), all of the first
6754 operations preceding the barrier will complete before any of the second
6755 operations succeeding the barrier begin. Specifically the semantics for each
6756 pairing is as follows:</p>
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]6757
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6758<ul>
6759 <li><tt>ll</tt>: All loads before the barrier must complete before any load
6760 after the barrier begins.</li>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]6761 <li><tt>ls</tt>: All loads before the barrier must complete before any
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6762 store after the barrier begins.</li>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]6763 <li><tt>ss</tt>: All stores before the barrier must complete before any
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6764 store after the barrier begins.</li>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]6765 <li><tt>sl</tt>: All stores before the barrier must complete before any
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6766 load after the barrier begins.</li>
6767</ul>
6768
6769<p>These semantics are applied with a logical "and" behavior when more than one
6770 is enabled in a single memory barrier intrinsic.</p>
6771
6772<p>Backends may implement stronger barriers than those requested when they do
6773 not support as fine grained a barrier as requested. Some architectures do
6774 not need all types of barriers and on such architectures, these become
6775 noops.</p>
6776
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]6777<h5>Example:</h5>
6778<pre>
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]6779%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
6780%ptr = bitcast i8* %mallocP to i32*
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]6781 store i32 4, %ptr
6782
6783%result1 = load i32* %ptr <i>; yields {i32}:result1 = 4</i>
6784 call void @llvm.memory.barrier( i1 false, i1 true, i1 false, i1 false )
6785 <i>; guarantee the above finishes</i>
6786 store i32 8, %ptr <i>; before this begins</i>
6787</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6788
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]6789</div>
6790
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6791<!-- _______________________________________________________________________ -->
6792<div class="doc_subsubsection">
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6793 <a name="int_atomic_cmp_swap">'<tt>llvm.atomic.cmp.swap.*</tt>' Intrinsic</a>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6794</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6795
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6796<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6797
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6798<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6799<p>This is an overloaded intrinsic. You can use <tt>llvm.atomic.cmp.swap</tt> on
6800 any integer bit width and for different address spaces. Not all targets
6801 support all bit widths however.</p>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6802
6803<pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6804 declare i8 @llvm.atomic.cmp.swap.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;cmp&gt;, i8 &lt;val&gt; )
6805 declare i16 @llvm.atomic.cmp.swap.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;cmp&gt;, i16 &lt;val&gt; )
6806 declare i32 @llvm.atomic.cmp.swap.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;cmp&gt;, i32 &lt;val&gt; )
6807 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]6808</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6809
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6810<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6811<p>This loads a value in memory and compares it to a given value. If they are
6812 equal, it stores a new value into the memory.</p>
6813
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6814<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6815<p>The <tt>llvm.atomic.cmp.swap</tt> intrinsic takes three arguments. The result
6816 as well as both <tt>cmp</tt> and <tt>val</tt> must be integer values with the
6817 same bit width. The <tt>ptr</tt> argument must be a pointer to a value of
6818 this integer type. While any bit width integer may be used, targets may only
6819 lower representations they support in hardware.</p>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6820
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6821<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6822<p>This entire intrinsic must be executed atomically. It first loads the value
6823 in memory pointed to by <tt>ptr</tt> and compares it with the
6824 value <tt>cmp</tt>. If they are equal, <tt>val</tt> is stored into the
6825 memory. The loaded value is yielded in all cases. This provides the
6826 equivalent of an atomic compare-and-swap operation within the SSA
6827 framework.</p>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6828
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6829<h5>Examples:</h5>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6830<pre>
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]6831%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
6832%ptr = bitcast i8* %mallocP to i32*
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6833 store i32 4, %ptr
6834
6835%val1 = add i32 4, 4
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6836%result1 = call i32 @llvm.atomic.cmp.swap.i32.p0i32( i32* %ptr, i32 4, %val1 )
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6837 <i>; yields {i32}:result1 = 4</i>
6838%stored1 = icmp eq i32 %result1, 4 <i>; yields {i1}:stored1 = true</i>
6839%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 8</i>
6840
6841%val2 = add i32 1, 1
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6842%result2 = call i32 @llvm.atomic.cmp.swap.i32.p0i32( i32* %ptr, i32 5, %val2 )
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6843 <i>; yields {i32}:result2 = 8</i>
6844%stored2 = icmp eq i32 %result2, 5 <i>; yields {i1}:stored2 = false</i>
6845
6846%memval2 = load i32* %ptr <i>; yields {i32}:memval2 = 8</i>
6847</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6848
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6849</div>
6850
6851<!-- _______________________________________________________________________ -->
6852<div class="doc_subsubsection">
6853 <a name="int_atomic_swap">'<tt>llvm.atomic.swap.*</tt>' Intrinsic</a>
6854</div>
6855<div class="doc_text">
6856<h5>Syntax:</h5>
6857
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6858<p>This is an overloaded intrinsic. You can use <tt>llvm.atomic.swap</tt> on any
6859 integer bit width. Not all targets support all bit widths however.</p>
6860
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6861<pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6862 declare i8 @llvm.atomic.swap.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;val&gt; )
6863 declare i16 @llvm.atomic.swap.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;val&gt; )
6864 declare i32 @llvm.atomic.swap.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;val&gt; )
6865 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]6866</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6867
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6868<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6869<p>This intrinsic loads the value stored in memory at <tt>ptr</tt> and yields
6870 the value from memory. It then stores the value in <tt>val</tt> in the memory
6871 at <tt>ptr</tt>.</p>
6872
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6873<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6874<p>The <tt>llvm.atomic.swap</tt> intrinsic takes two arguments. Both
6875 the <tt>val</tt> argument and the result must be integers of the same bit
6876 width. The first argument, <tt>ptr</tt>, must be a pointer to a value of this
6877 integer type. The targets may only lower integer representations they
6878 support.</p>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6879
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6880<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6881<p>This intrinsic loads the value pointed to by <tt>ptr</tt>, yields it, and
6882 stores <tt>val</tt> back into <tt>ptr</tt> atomically. This provides the
6883 equivalent of an atomic swap operation within the SSA framework.</p>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6884
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6885<h5>Examples:</h5>
6886<pre>
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]6887%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
6888%ptr = bitcast i8* %mallocP to i32*
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6889 store i32 4, %ptr
6890
6891%val1 = add i32 4, 4
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6892%result1 = call i32 @llvm.atomic.swap.i32.p0i32( i32* %ptr, i32 %val1 )
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6893 <i>; yields {i32}:result1 = 4</i>
6894%stored1 = icmp eq i32 %result1, 4 <i>; yields {i1}:stored1 = true</i>
6895%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 8</i>
6896
6897%val2 = add i32 1, 1
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6898%result2 = call i32 @llvm.atomic.swap.i32.p0i32( i32* %ptr, i32 %val2 )
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6899 <i>; yields {i32}:result2 = 8</i>
6900
6901%stored2 = icmp eq i32 %result2, 8 <i>; yields {i1}:stored2 = true</i>
6902%memval2 = load i32* %ptr <i>; yields {i32}:memval2 = 2</i>
6903</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6904
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6905</div>
6906
6907<!-- _______________________________________________________________________ -->
6908<div class="doc_subsubsection">
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6909 <a name="int_atomic_load_add">'<tt>llvm.atomic.load.add.*</tt>' Intrinsic</a>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6910
6911</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6912
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6913<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6914
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6915<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6916<p>This is an overloaded intrinsic. You can use <tt>llvm.atomic.load.add</tt> on
6917 any integer bit width. Not all targets support all bit widths however.</p>
6918
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6919<pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6920 declare i8 @llvm.atomic.load.add.i8..p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6921 declare i16 @llvm.atomic.load.add.i16..p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6922 declare i32 @llvm.atomic.load.add.i32..p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6923 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]6924</pre>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6925
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6926<h5>Overview:</h5>
6927<p>This intrinsic adds <tt>delta</tt> to the value stored in memory
6928 at <tt>ptr</tt>. It yields the original value at <tt>ptr</tt>.</p>
6929
6930<h5>Arguments:</h5>
6931<p>The intrinsic takes two arguments, the first a pointer to an integer value
6932 and the second an integer value. The result is also an integer value. These
6933 integer types can have any bit width, but they must all have the same bit
6934 width. The targets may only lower integer representations they support.</p>
6935
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6936<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6937<p>This intrinsic does a series of operations atomically. It first loads the
6938 value stored at <tt>ptr</tt>. It then adds <tt>delta</tt>, stores the result
6939 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]6940
6941<h5>Examples:</h5>
6942<pre>
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]6943%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
6944%ptr = bitcast i8* %mallocP to i32*
6945 store i32 4, %ptr
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6946%result1 = call i32 @llvm.atomic.load.add.i32.p0i32( i32* %ptr, i32 4 )
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6947 <i>; yields {i32}:result1 = 4</i>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6948%result2 = call i32 @llvm.atomic.load.add.i32.p0i32( i32* %ptr, i32 2 )
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6949 <i>; yields {i32}:result2 = 8</i>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6950%result3 = call i32 @llvm.atomic.load.add.i32.p0i32( i32* %ptr, i32 5 )
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6951 <i>; yields {i32}:result3 = 10</i>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6952%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 15</i>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6953</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6954
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6955</div>
6956
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6957<!-- _______________________________________________________________________ -->
6958<div class="doc_subsubsection">
6959 <a name="int_atomic_load_sub">'<tt>llvm.atomic.load.sub.*</tt>' Intrinsic</a>
6960
6961</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6962
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6963<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6964
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6965<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6966<p>This is an overloaded intrinsic. You can use <tt>llvm.atomic.load.sub</tt> on
6967 any integer bit width and for different address spaces. Not all targets
6968 support all bit widths however.</p>
6969
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6970<pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6971 declare i8 @llvm.atomic.load.sub.i8.p0i32( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6972 declare i16 @llvm.atomic.load.sub.i16.p0i32( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6973 declare i32 @llvm.atomic.load.sub.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6974 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]6975</pre>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6976
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6977<h5>Overview:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]6978<p>This intrinsic subtracts <tt>delta</tt> to the value stored in memory at
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6979 <tt>ptr</tt>. It yields the original value at <tt>ptr</tt>.</p>
6980
6981<h5>Arguments:</h5>
6982<p>The intrinsic takes two arguments, the first a pointer to an integer value
6983 and the second an integer value. The result is also an integer value. These
6984 integer types can have any bit width, but they must all have the same bit
6985 width. The targets may only lower integer representations they support.</p>
6986
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6987<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6988<p>This intrinsic does a series of operations atomically. It first loads the
6989 value stored at <tt>ptr</tt>. It then subtracts <tt>delta</tt>, stores the
6990 result to <tt>ptr</tt>. It yields the original value stored
6991 at <tt>ptr</tt>.</p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6992
6993<h5>Examples:</h5>
6994<pre>
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]6995%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
6996%ptr = bitcast i8* %mallocP to i32*
6997 store i32 8, %ptr
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6998%result1 = call i32 @llvm.atomic.load.sub.i32.p0i32( i32* %ptr, i32 4 )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6999 <i>; yields {i32}:result1 = 8</i>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]7000%result2 = call i32 @llvm.atomic.load.sub.i32.p0i32( i32* %ptr, i32 2 )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7001 <i>; yields {i32}:result2 = 4</i>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]7002%result3 = call i32 @llvm.atomic.load.sub.i32.p0i32( i32* %ptr, i32 5 )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7003 <i>; yields {i32}:result3 = 2</i>
7004%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = -3</i>
7005</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7006
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7007</div>
7008
7009<!-- _______________________________________________________________________ -->
7010<div class="doc_subsubsection">
7011 <a name="int_atomic_load_and">'<tt>llvm.atomic.load.and.*</tt>' Intrinsic</a><br>
7012 <a name="int_atomic_load_nand">'<tt>llvm.atomic.load.nand.*</tt>' Intrinsic</a><br>
7013 <a name="int_atomic_load_or">'<tt>llvm.atomic.load.or.*</tt>' Intrinsic</a><br>
7014 <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]7015</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7016
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7017<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7018
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7019<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7020<p>These are overloaded intrinsics. You can
7021 use <tt>llvm.atomic.load_and</tt>, <tt>llvm.atomic.load_nand</tt>,
7022 <tt>llvm.atomic.load_or</tt>, and <tt>llvm.atomic.load_xor</tt> on any integer
7023 bit width and for different address spaces. Not all targets support all bit
7024 widths however.</p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7025
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7026<pre>
7027 declare i8 @llvm.atomic.load.and.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
7028 declare i16 @llvm.atomic.load.and.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
7029 declare i32 @llvm.atomic.load.and.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
7030 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]7031</pre>
7032
7033<pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7034 declare i8 @llvm.atomic.load.or.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
7035 declare i16 @llvm.atomic.load.or.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
7036 declare i32 @llvm.atomic.load.or.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
7037 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]7038</pre>
7039
7040<pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7041 declare i8 @llvm.atomic.load.nand.i8.p0i32( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
7042 declare i16 @llvm.atomic.load.nand.i16.p0i32( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
7043 declare i32 @llvm.atomic.load.nand.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
7044 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]7045</pre>
7046
7047<pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7048 declare i8 @llvm.atomic.load.xor.i8.p0i32( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
7049 declare i16 @llvm.atomic.load.xor.i16.p0i32( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
7050 declare i32 @llvm.atomic.load.xor.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
7051 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]7052</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7053
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7054<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7055<p>These intrinsics bitwise the operation (and, nand, or, xor) <tt>delta</tt> to
7056 the value stored in memory at <tt>ptr</tt>. It yields the original value
7057 at <tt>ptr</tt>.</p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7058
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7059<h5>Arguments:</h5>
7060<p>These intrinsics take two arguments, the first a pointer to an integer value
7061 and the second an integer value. The result is also an integer value. These
7062 integer types can have any bit width, but they must all have the same bit
7063 width. The targets may only lower integer representations they support.</p>
7064
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7065<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7066<p>These intrinsics does a series of operations atomically. They first load the
7067 value stored at <tt>ptr</tt>. They then do the bitwise
7068 operation <tt>delta</tt>, store the result to <tt>ptr</tt>. They yield the
7069 original value stored at <tt>ptr</tt>.</p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7070
7071<h5>Examples:</h5>
7072<pre>
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]7073%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
7074%ptr = bitcast i8* %mallocP to i32*
7075 store i32 0x0F0F, %ptr
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]7076%result0 = call i32 @llvm.atomic.load.nand.i32.p0i32( i32* %ptr, i32 0xFF )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7077 <i>; yields {i32}:result0 = 0x0F0F</i>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]7078%result1 = call i32 @llvm.atomic.load.and.i32.p0i32( i32* %ptr, i32 0xFF )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7079 <i>; yields {i32}:result1 = 0xFFFFFFF0</i>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]7080%result2 = call i32 @llvm.atomic.load.or.i32.p0i32( i32* %ptr, i32 0F )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7081 <i>; yields {i32}:result2 = 0xF0</i>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]7082%result3 = call i32 @llvm.atomic.load.xor.i32.p0i32( i32* %ptr, i32 0F )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7083 <i>; yields {i32}:result3 = FF</i>
7084%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = F0</i>
7085</pre>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7086
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7087</div>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7088
7089<!-- _______________________________________________________________________ -->
7090<div class="doc_subsubsection">
7091 <a name="int_atomic_load_max">'<tt>llvm.atomic.load.max.*</tt>' Intrinsic</a><br>
7092 <a name="int_atomic_load_min">'<tt>llvm.atomic.load.min.*</tt>' Intrinsic</a><br>
7093 <a name="int_atomic_load_umax">'<tt>llvm.atomic.load.umax.*</tt>' Intrinsic</a><br>
7094 <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]7095</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7096
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7097<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7098
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7099<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7100<p>These are overloaded intrinsics. You can use <tt>llvm.atomic.load_max</tt>,
7101 <tt>llvm.atomic.load_min</tt>, <tt>llvm.atomic.load_umax</tt>, and
7102 <tt>llvm.atomic.load_umin</tt> on any integer bit width and for different
7103 address spaces. Not all targets support all bit widths however.</p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7104
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7105<pre>
7106 declare i8 @llvm.atomic.load.max.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
7107 declare i16 @llvm.atomic.load.max.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
7108 declare i32 @llvm.atomic.load.max.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
7109 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]7110</pre>
7111
7112<pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7113 declare i8 @llvm.atomic.load.min.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
7114 declare i16 @llvm.atomic.load.min.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
7115 declare i32 @llvm.atomic.load.min.i32..p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
7116 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]7117</pre>
7118
7119<pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7120 declare i8 @llvm.atomic.load.umax.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
7121 declare i16 @llvm.atomic.load.umax.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
7122 declare i32 @llvm.atomic.load.umax.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
7123 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]7124</pre>
7125
7126<pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7127 declare i8 @llvm.atomic.load.umin.i8..p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
7128 declare i16 @llvm.atomic.load.umin.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
7129 declare i32 @llvm.atomic.load.umin.i32..p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
7130 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]7131</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7132
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7133<h5>Overview:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]7134<p>These intrinsics takes the signed or unsigned minimum or maximum of
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7135 <tt>delta</tt> and the value stored in memory at <tt>ptr</tt>. It yields the
7136 original value at <tt>ptr</tt>.</p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7137
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7138<h5>Arguments:</h5>
7139<p>These intrinsics take two arguments, the first a pointer to an integer value
7140 and the second an integer value. The result is also an integer value. These
7141 integer types can have any bit width, but they must all have the same bit
7142 width. The targets may only lower integer representations they support.</p>
7143
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7144<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7145<p>These intrinsics does a series of operations atomically. They first load the
7146 value stored at <tt>ptr</tt>. They then do the signed or unsigned min or
7147 max <tt>delta</tt> and the value, store the result to <tt>ptr</tt>. They
7148 yield the original value stored at <tt>ptr</tt>.</p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7149
7150<h5>Examples:</h5>
7151<pre>
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]7152%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
7153%ptr = bitcast i8* %mallocP to i32*
7154 store i32 7, %ptr
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]7155%result0 = call i32 @llvm.atomic.load.min.i32.p0i32( i32* %ptr, i32 -2 )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7156 <i>; yields {i32}:result0 = 7</i>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]7157%result1 = call i32 @llvm.atomic.load.max.i32.p0i32( i32* %ptr, i32 8 )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7158 <i>; yields {i32}:result1 = -2</i>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]7159%result2 = call i32 @llvm.atomic.load.umin.i32.p0i32( i32* %ptr, i32 10 )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7160 <i>; yields {i32}:result2 = 8</i>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]7161%result3 = call i32 @llvm.atomic.load.umax.i32.p0i32( i32* %ptr, i32 30 )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7162 <i>; yields {i32}:result3 = 8</i>
7163%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 30</i>
7164</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7165
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7166</div>
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7167
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7168
7169<!-- ======================================================================= -->
7170<div class="doc_subsection">
7171 <a name="int_memorymarkers">Memory Use Markers</a>
7172</div>
7173
7174<div class="doc_text">
7175
7176<p>This class of intrinsics exists to information about the lifetime of memory
7177 objects and ranges where variables are immutable.</p>
7178
7179</div>
7180
7181<!-- _______________________________________________________________________ -->
7182<div class="doc_subsubsection">
7183 <a name="int_lifetime_start">'<tt>llvm.lifetime.start</tt>' Intrinsic</a>
7184</div>
7185
7186<div class="doc_text">
7187
7188<h5>Syntax:</h5>
7189<pre>
7190 declare void @llvm.lifetime.start(i64 &lt;size&gt;, i8* nocapture &lt;ptr&gt;)
7191</pre>
7192
7193<h5>Overview:</h5>
7194<p>The '<tt>llvm.lifetime.start</tt>' intrinsic specifies the start of a memory
7195 object's lifetime.</p>
7196
7197<h5>Arguments:</h5>
Nick Lewycky321333e2009-10-13 07:57:33 +0000[diff] [blame]7198<p>The first argument is a constant integer representing the size of the
7199 object, or -1 if it is variable sized. The second argument is a pointer to
7200 the object.</p>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7201
7202<h5>Semantics:</h5>
7203<p>This intrinsic indicates that before this point in the code, the value of the
7204 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]7205 never be used and has an undefined value. A load from the pointer that
7206 precedes this intrinsic can be replaced with
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7207 <tt>'<a href="#undefvalues">undef</a>'</tt>.</p>
7208
7209</div>
7210
7211<!-- _______________________________________________________________________ -->
7212<div class="doc_subsubsection">
7213 <a name="int_lifetime_end">'<tt>llvm.lifetime.end</tt>' Intrinsic</a>
7214</div>
7215
7216<div class="doc_text">
7217
7218<h5>Syntax:</h5>
7219<pre>
7220 declare void @llvm.lifetime.end(i64 &lt;size&gt;, i8* nocapture &lt;ptr&gt;)
7221</pre>
7222
7223<h5>Overview:</h5>
7224<p>The '<tt>llvm.lifetime.end</tt>' intrinsic specifies the end of a memory
7225 object's lifetime.</p>
7226
7227<h5>Arguments:</h5>
Nick Lewycky321333e2009-10-13 07:57:33 +0000[diff] [blame]7228<p>The first argument is a constant integer representing the size of the
7229 object, or -1 if it is variable sized. The second argument is a pointer to
7230 the object.</p>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7231
7232<h5>Semantics:</h5>
7233<p>This intrinsic indicates that after this point in the code, the value of the
7234 memory pointed to by <tt>ptr</tt> is dead. This means that it is known to
7235 never be used and has an undefined value. Any stores into the memory object
7236 following this intrinsic may be removed as dead.
7237
7238</div>
7239
7240<!-- _______________________________________________________________________ -->
7241<div class="doc_subsubsection">
7242 <a name="int_invariant_start">'<tt>llvm.invariant.start</tt>' Intrinsic</a>
7243</div>
7244
7245<div class="doc_text">
7246
7247<h5>Syntax:</h5>
7248<pre>
7249 declare {}* @llvm.invariant.start(i64 &lt;size&gt;, i8* nocapture &lt;ptr&gt;) readonly
7250</pre>
7251
7252<h5>Overview:</h5>
7253<p>The '<tt>llvm.invariant.start</tt>' intrinsic specifies that the contents of
7254 a memory object will not change.</p>
7255
7256<h5>Arguments:</h5>
Nick Lewycky321333e2009-10-13 07:57:33 +0000[diff] [blame]7257<p>The first argument is a constant integer representing the size of the
7258 object, or -1 if it is variable sized. The second argument is a pointer to
7259 the object.</p>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7260
7261<h5>Semantics:</h5>
7262<p>This intrinsic indicates that until an <tt>llvm.invariant.end</tt> that uses
7263 the return value, the referenced memory location is constant and
7264 unchanging.</p>
7265
7266</div>
7267
7268<!-- _______________________________________________________________________ -->
7269<div class="doc_subsubsection">
7270 <a name="int_invariant_end">'<tt>llvm.invariant.end</tt>' Intrinsic</a>
7271</div>
7272
7273<div class="doc_text">
7274
7275<h5>Syntax:</h5>
7276<pre>
7277 declare void @llvm.invariant.end({}* &lt;start&gt;, i64 &lt;size&gt;, i8* nocapture &lt;ptr&gt;)
7278</pre>
7279
7280<h5>Overview:</h5>
7281<p>The '<tt>llvm.invariant.end</tt>' intrinsic specifies that the contents of
7282 a memory object are mutable.</p>
7283
7284<h5>Arguments:</h5>
7285<p>The first argument is the matching <tt>llvm.invariant.start</tt> intrinsic.
Nick Lewycky321333e2009-10-13 07:57:33 +0000[diff] [blame]7286 The second argument is a constant integer representing the size of the
7287 object, or -1 if it is variable sized and the third argument is a pointer
7288 to the object.</p>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7289
7290<h5>Semantics:</h5>
7291<p>This intrinsic indicates that the memory is mutable again.</p>
7292
7293</div>
7294
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7295<!-- ======================================================================= -->
7296<div class="doc_subsection">
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]7297 <a name="int_general">General Intrinsics</a>
7298</div>
7299
7300<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7301
7302<p>This class of intrinsics is designed to be generic and has no specific
7303 purpose.</p>
7304
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]7305</div>
7306
7307<!-- _______________________________________________________________________ -->
7308<div class="doc_subsubsection">
7309 <a name="int_var_annotation">'<tt>llvm.var.annotation</tt>' Intrinsic</a>
7310</div>
7311
7312<div class="doc_text">
7313
7314<h5>Syntax:</h5>
7315<pre>
Tanya Lattnerd2e84422007-06-18 23:42:37 +0000[diff] [blame]7316 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]7317</pre>
7318
7319<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7320<p>The '<tt>llvm.var.annotation</tt>' intrinsic.</p>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]7321
7322<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7323<p>The first argument is a pointer to a value, the second is a pointer to a
7324 global string, the third is a pointer to a global string which is the source
7325 file name, and the last argument is the line number.</p>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]7326
7327<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7328<p>This intrinsic allows annotation of local variables with arbitrary strings.
7329 This can be useful for special purpose optimizations that want to look for
7330 these annotations. These have no other defined use, they are ignored by code
7331 generation and optimization.</p>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]7332
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]7333</div>
7334
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]7335<!-- _______________________________________________________________________ -->
7336<div class="doc_subsubsection">
Tanya Lattnere1a8da02007-09-21 23:57:59 +0000[diff] [blame]7337 <a name="int_annotation">'<tt>llvm.annotation.*</tt>' Intrinsic</a>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]7338</div>
7339
7340<div class="doc_text">
7341
7342<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7343<p>This is an overloaded intrinsic. You can use '<tt>llvm.annotation</tt>' on
7344 any integer bit width.</p>
7345
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]7346<pre>
Tanya Lattnerd3989a82007-09-22 00:03:01 +0000[diff] [blame]7347 declare i8 @llvm.annotation.i8(i8 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
7348 declare i16 @llvm.annotation.i16(i16 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
7349 declare i32 @llvm.annotation.i32(i32 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
7350 declare i64 @llvm.annotation.i64(i64 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
7351 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]7352</pre>
7353
7354<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7355<p>The '<tt>llvm.annotation</tt>' intrinsic.</p>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]7356
7357<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7358<p>The first argument is an integer value (result of some expression), the
7359 second is a pointer to a global string, the third is a pointer to a global
7360 string which is the source file name, and the last argument is the line
7361 number. It returns the value of the first argument.</p>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]7362
7363<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7364<p>This intrinsic allows annotations to be put on arbitrary expressions with
7365 arbitrary strings. This can be useful for special purpose optimizations that
7366 want to look for these annotations. These have no other defined use, they
7367 are ignored by code generation and optimization.</p>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]7368
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]7369</div>
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]7370
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]7371<!-- _______________________________________________________________________ -->
7372<div class="doc_subsubsection">
7373 <a name="int_trap">'<tt>llvm.trap</tt>' Intrinsic</a>
7374</div>
7375
7376<div class="doc_text">
7377
7378<h5>Syntax:</h5>
7379<pre>
7380 declare void @llvm.trap()
7381</pre>
7382
7383<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7384<p>The '<tt>llvm.trap</tt>' intrinsic.</p>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]7385
7386<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7387<p>None.</p>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]7388
7389<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7390<p>This intrinsics is lowered to the target dependent trap instruction. If the
7391 target does not have a trap instruction, this intrinsic will be lowered to
7392 the call of the <tt>abort()</tt> function.</p>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]7393
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]7394</div>
7395
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]7396<!-- _______________________________________________________________________ -->
7397<div class="doc_subsubsection">
Misha Brukmandccb0252008-11-22 23:55:29 +0000[diff] [blame]7398 <a name="int_stackprotector">'<tt>llvm.stackprotector</tt>' Intrinsic</a>
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]7399</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7400
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]7401<div class="doc_text">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7402
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]7403<h5>Syntax:</h5>
7404<pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7405 declare void @llvm.stackprotector( i8* &lt;guard&gt;, i8** &lt;slot&gt; )
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]7406</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7407
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]7408<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7409<p>The <tt>llvm.stackprotector</tt> intrinsic takes the <tt>guard</tt> and
7410 stores it onto the stack at <tt>slot</tt>. The stack slot is adjusted to
7411 ensure that it is placed on the stack before local variables.</p>
7412
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]7413<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7414<p>The <tt>llvm.stackprotector</tt> intrinsic requires two pointer
7415 arguments. The first argument is the value loaded from the stack
7416 guard <tt>@__stack_chk_guard</tt>. The second variable is an <tt>alloca</tt>
7417 that has enough space to hold the value of the guard.</p>
7418
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]7419<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7420<p>This intrinsic causes the prologue/epilogue inserter to force the position of
7421 the <tt>AllocaInst</tt> stack slot to be before local variables on the
7422 stack. This is to ensure that if a local variable on the stack is
7423 overwritten, it will destroy the value of the guard. When the function exits,
7424 the guard on the stack is checked against the original guard. If they're
7425 different, then the program aborts by calling the <tt>__stack_chk_fail()</tt>
7426 function.</p>
7427
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]7428</div>
7429
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]7430<!-- _______________________________________________________________________ -->
7431<div class="doc_subsubsection">
7432 <a name="int_objectsize">'<tt>llvm.objectsize</tt>' Intrinsic</a>
7433</div>
7434
7435<div class="doc_text">
7436
7437<h5>Syntax:</h5>
7438<pre>
Eric Christopher8295a0a2009-12-23 00:29:49 +0000[diff] [blame]7439 declare i32 @llvm.objectsize.i32( i8* &lt;object&gt;, i1 &lt;type&gt; )
7440 declare i64 @llvm.objectsize.i64( i8* &lt;object&gt;, i1 &lt;type&gt; )
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]7441</pre>
7442
7443<h5>Overview:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]7444<p>The <tt>llvm.objectsize</tt> intrinsic is designed to provide information
Eric Christopherd003c5b2010-01-08 21:42:39 +0000[diff] [blame]7445 to the optimizers to discover at compile time either a) when an
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]7446 operation like memcpy will either overflow a buffer that corresponds to
7447 an object, or b) to determine that a runtime check for overflow isn't
7448 necessary. An object in this context means an allocation of a
Eric Christopher8295a0a2009-12-23 00:29:49 +0000[diff] [blame]7449 specific class, structure, array, or other object.</p>
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]7450
7451<h5>Arguments:</h5>
7452<p>The <tt>llvm.objectsize</tt> intrinsic takes two arguments. The first
Eric Christopher8295a0a2009-12-23 00:29:49 +0000[diff] [blame]7453 argument is a pointer to or into the <tt>object</tt>. The second argument
7454 is a boolean 0 or 1. This argument determines whether you want the
7455 maximum (0) or minimum (1) bytes remaining. This needs to be a literal 0 or
7456 1, variables are not allowed.</p>
7457
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]7458<h5>Semantics:</h5>
7459<p>The <tt>llvm.objectsize</tt> intrinsic is lowered to either a constant
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]7460 representing the size of the object concerned or <tt>i32/i64 -1 or 0</tt>
7461 (depending on the <tt>type</tt> argument if the size cannot be determined
7462 at compile time.</p>
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]7463
7464</div>
7465
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]7466<!-- *********************************************************************** -->
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]7467<hr>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]7468<address>
7469 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
Misha Brukman44408702008-12-11 17:34:48 +0000[diff] [blame]7470 src="http://jigsaw.w3.org/css-validator/images/vcss-blue" alt="Valid CSS"></a>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]7471 <a href="http://validator.w3.org/check/referer"><img
Misha Brukman44408702008-12-11 17:34:48 +0000[diff] [blame]7472 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]7473
7474 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
Reid Spencer05fe4b02006-03-14 05:39:39 +0000[diff] [blame]7475 <a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]7476 Last modified: $Date$
7477</address>
Chris Lattnerc7d3ab32008-01-04 04:33:49 +0000[diff] [blame]7478
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]7479</body>
7480</html>