Gitiles
Code Review Sign In
gerrit-public.fairphone.software / fp2-dev / platform / external / llvm / dccc03b2423fe65efb5963ae816b99c24fc53374 / . / docs / LangRef.html
blob: ac527c4df554b5924e1e6ee1d0565349e8ec55f5 [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
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]15<h1>LLVM Language Reference Manual</h1>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]16<ol>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]17 <li><a href="#abstract">Abstract</a></li>
18 <li><a href="#introduction">Introduction</a></li>
19 <li><a href="#identifiers">Identifiers</a></li>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]20 <li><a href="#highlevel">High Level Structure</a>
21 <ol>
22 <li><a href="#modulestructure">Module Structure</a></li>
Bill Wendling3d10a5a2009-07-20 01:03:30 +0000[diff] [blame]23 <li><a href="#linkage">Linkage Types</a>
24 <ol>
Bill Wendling987e7eb2009-07-20 02:41:50 +0000[diff] [blame]25 <li><a href="#linkage_private">'<tt>private</tt>' Linkage</a></li>
26 <li><a href="#linkage_linker_private">'<tt>linker_private</tt>' Linkage</a></li>
Bill Wendling5e721d72010-07-01 21:55:59 +0000[diff] [blame]27 <li><a href="#linkage_linker_private_weak">'<tt>linker_private_weak</tt>' Linkage</a></li>
Bill Wendling55ae5152010-08-20 22:05:50 +0000[diff] [blame]28 <li><a href="#linkage_linker_private_weak_def_auto">'<tt>linker_private_weak_def_auto</tt>' Linkage</a></li>
Bill Wendling987e7eb2009-07-20 02:41:50 +0000[diff] [blame]29 <li><a href="#linkage_internal">'<tt>internal</tt>' Linkage</a></li>
30 <li><a href="#linkage_available_externally">'<tt>available_externally</tt>' Linkage</a></li>
31 <li><a href="#linkage_linkonce">'<tt>linkonce</tt>' Linkage</a></li>
32 <li><a href="#linkage_common">'<tt>common</tt>' Linkage</a></li>
33 <li><a href="#linkage_weak">'<tt>weak</tt>' Linkage</a></li>
34 <li><a href="#linkage_appending">'<tt>appending</tt>' Linkage</a></li>
35 <li><a href="#linkage_externweak">'<tt>extern_weak</tt>' Linkage</a></li>
Chris Lattner5a2d8752009-10-10 18:26:06 +0000[diff] [blame]36 <li><a href="#linkage_linkonce_odr">'<tt>linkonce_odr</tt>' Linkage</a></li>
Bill Wendling987e7eb2009-07-20 02:41:50 +0000[diff] [blame]37 <li><a href="#linkage_weak">'<tt>weak_odr</tt>' Linkage</a></li>
38 <li><a href="#linkage_external">'<tt>externally visible</tt>' Linkage</a></li>
39 <li><a href="#linkage_dllimport">'<tt>dllimport</tt>' Linkage</a></li>
40 <li><a href="#linkage_dllexport">'<tt>dllexport</tt>' Linkage</a></li>
Bill Wendling3d10a5a2009-07-20 01:03:30 +0000[diff] [blame]41 </ol>
42 </li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]43 <li><a href="#callingconv">Calling Conventions</a></li>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]44 <li><a href="#namedtypes">Named Types</a></li>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]45 <li><a href="#globalvars">Global Variables</a></li>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]46 <li><a href="#functionstructure">Functions</a></li>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]47 <li><a href="#aliasstructure">Aliases</a></li>
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]48 <li><a href="#namedmetadatastructure">Named Metadata</a></li>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]49 <li><a href="#paramattrs">Parameter Attributes</a></li>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]50 <li><a href="#fnattrs">Function Attributes</a></li>
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame]51 <li><a href="#gc">Garbage Collector Names</a></li>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]52 <li><a href="#moduleasm">Module-Level Inline Assembly</a></li>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]53 <li><a href="#datalayout">Data Layout</a></li>
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]54 <li><a href="#pointeraliasing">Pointer Aliasing Rules</a></li>
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]55 <li><a href="#volatile">Volatile Memory Accesses</a></li>
Eli Friedman5b60e1b2011-07-20 21:35:53 +0000[diff] [blame]56 <li><a href="#memmodel">Memory Model for Concurrent Operations</a></li>
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]57 <li><a href="#ordering">Atomic Memory Ordering Constraints</a></li>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]58 </ol>
59 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]60 <li><a href="#typesystem">Type System</a>
61 <ol>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]62 <li><a href="#t_classifications">Type Classifications</a></li>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]63 <li><a href="#t_primitive">Primitive Types</a>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]64 <ol>
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]65 <li><a href="#t_integer">Integer Type</a></li>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]66 <li><a href="#t_floating">Floating Point Types</a></li>
Dale Johannesen21fe99b2010-10-01 00:48:59 +0000[diff] [blame]67 <li><a href="#t_x86mmx">X86mmx Type</a></li>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]68 <li><a href="#t_void">Void Type</a></li>
69 <li><a href="#t_label">Label Type</a></li>
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]70 <li><a href="#t_metadata">Metadata Type</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]71 </ol>
72 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]73 <li><a href="#t_derived">Derived Types</a>
74 <ol>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]75 <li><a href="#t_aggregate">Aggregate Types</a>
76 <ol>
77 <li><a href="#t_array">Array Type</a></li>
78 <li><a href="#t_struct">Structure Type</a></li>
Chris Lattner628ed392011-07-23 19:59:08 +0000[diff] [blame]79 <li><a href="#t_opaque">Opaque Structure Types</a></li>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]80 <li><a href="#t_vector">Vector Type</a></li>
81 </ol>
82 </li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]83 <li><a href="#t_function">Function Type</a></li>
84 <li><a href="#t_pointer">Pointer Type</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]85 </ol>
86 </li>
87 </ol>
88 </li>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]89 <li><a href="#constants">Constants</a>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]90 <ol>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]91 <li><a href="#simpleconstants">Simple Constants</a></li>
Chris Lattner70882792009-02-28 18:32:25 +0000[diff] [blame]92 <li><a href="#complexconstants">Complex Constants</a></li>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]93 <li><a href="#globalconstants">Global Variable and Function Addresses</a></li>
94 <li><a href="#undefvalues">Undefined Values</a></li>
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]95 <li><a href="#trapvalues">Trap Values</a></li>
Chris Lattnerf9d078e2009-10-27 21:19:13 +0000[diff] [blame]96 <li><a href="#blockaddress">Addresses of Basic Blocks</a></li>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]97 <li><a href="#constantexprs">Constant Expressions</a></li>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]98 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]99 </li>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]100 <li><a href="#othervalues">Other Values</a>
101 <ol>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]102 <li><a href="#inlineasm">Inline Assembler Expressions</a></li>
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]103 <li><a href="#metadata">Metadata Nodes and Metadata Strings</a></li>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]104 </ol>
105 </li>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]106 <li><a href="#intrinsic_globals">Intrinsic Global Variables</a>
107 <ol>
108 <li><a href="#intg_used">The '<tt>llvm.used</tt>' Global Variable</a></li>
Chris Lattner401e10c2009-07-20 06:14:25 +0000[diff] [blame]109 <li><a href="#intg_compiler_used">The '<tt>llvm.compiler.used</tt>'
110 Global Variable</a></li>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]111 <li><a href="#intg_global_ctors">The '<tt>llvm.global_ctors</tt>'
112 Global Variable</a></li>
113 <li><a href="#intg_global_dtors">The '<tt>llvm.global_dtors</tt>'
114 Global Variable</a></li>
115 </ol>
116 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]117 <li><a href="#instref">Instruction Reference</a>
118 <ol>
119 <li><a href="#terminators">Terminator Instructions</a>
120 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]121 <li><a href="#i_ret">'<tt>ret</tt>' Instruction</a></li>
122 <li><a href="#i_br">'<tt>br</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]123 <li><a href="#i_switch">'<tt>switch</tt>' Instruction</a></li>
Chris Lattnerab21db72009-10-28 00:19:10 +0000[diff] [blame]124 <li><a href="#i_indirectbr">'<tt>indirectbr</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]125 <li><a href="#i_invoke">'<tt>invoke</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]126 <li><a href="#i_unwind">'<tt>unwind</tt>' Instruction</a></li>
Bill Wendlingdccc03b2011-07-31 06:30:59 +0000[diff] [blame^]127 <li><a href="#i_resume">'<tt>resume</tt>' Instruction</a></li>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]128 <li><a href="#i_unreachable">'<tt>unreachable</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]129 </ol>
130 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]131 <li><a href="#binaryops">Binary Operations</a>
132 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]133 <li><a href="#i_add">'<tt>add</tt>' Instruction</a></li>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]134 <li><a href="#i_fadd">'<tt>fadd</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]135 <li><a href="#i_sub">'<tt>sub</tt>' Instruction</a></li>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]136 <li><a href="#i_fsub">'<tt>fsub</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]137 <li><a href="#i_mul">'<tt>mul</tt>' Instruction</a></li>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]138 <li><a href="#i_fmul">'<tt>fmul</tt>' Instruction</a></li>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]139 <li><a href="#i_udiv">'<tt>udiv</tt>' Instruction</a></li>
140 <li><a href="#i_sdiv">'<tt>sdiv</tt>' Instruction</a></li>
141 <li><a href="#i_fdiv">'<tt>fdiv</tt>' Instruction</a></li>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]142 <li><a href="#i_urem">'<tt>urem</tt>' Instruction</a></li>
143 <li><a href="#i_srem">'<tt>srem</tt>' Instruction</a></li>
144 <li><a href="#i_frem">'<tt>frem</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]145 </ol>
146 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]147 <li><a href="#bitwiseops">Bitwise Binary Operations</a>
148 <ol>
Reid Spencer8e11bf82007-02-02 13:57:07 +0000[diff] [blame]149 <li><a href="#i_shl">'<tt>shl</tt>' Instruction</a></li>
150 <li><a href="#i_lshr">'<tt>lshr</tt>' Instruction</a></li>
151 <li><a href="#i_ashr">'<tt>ashr</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]152 <li><a href="#i_and">'<tt>and</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]153 <li><a href="#i_or">'<tt>or</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]154 <li><a href="#i_xor">'<tt>xor</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]155 </ol>
156 </li>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]157 <li><a href="#vectorops">Vector Operations</a>
158 <ol>
159 <li><a href="#i_extractelement">'<tt>extractelement</tt>' Instruction</a></li>
160 <li><a href="#i_insertelement">'<tt>insertelement</tt>' Instruction</a></li>
161 <li><a href="#i_shufflevector">'<tt>shufflevector</tt>' Instruction</a></li>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]162 </ol>
163 </li>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]164 <li><a href="#aggregateops">Aggregate Operations</a>
165 <ol>
166 <li><a href="#i_extractvalue">'<tt>extractvalue</tt>' Instruction</a></li>
167 <li><a href="#i_insertvalue">'<tt>insertvalue</tt>' Instruction</a></li>
168 </ol>
169 </li>
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]170 <li><a href="#memoryops">Memory Access and Addressing Operations</a>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]171 <ol>
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]172 <li><a href="#i_alloca">'<tt>alloca</tt>' Instruction</a></li>
173 <li><a href="#i_load">'<tt>load</tt>' Instruction</a></li>
174 <li><a href="#i_store">'<tt>store</tt>' Instruction</a></li>
175 <li><a href="#i_fence">'<tt>fence</tt>' Instruction</a></li>
176 <li><a href="#i_cmpxchg">'<tt>cmpxchg</tt>' Instruction</a></li>
177 <li><a href="#i_atomicrmw">'<tt>atomicrmw</tt>' Instruction</a></li>
Robert Bocchino7b81c752006-02-17 21:18:08 +0000[diff] [blame]178 <li><a href="#i_getelementptr">'<tt>getelementptr</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]179 </ol>
180 </li>
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]181 <li><a href="#convertops">Conversion Operations</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]182 <ol>
183 <li><a href="#i_trunc">'<tt>trunc .. to</tt>' Instruction</a></li>
184 <li><a href="#i_zext">'<tt>zext .. to</tt>' Instruction</a></li>
185 <li><a href="#i_sext">'<tt>sext .. to</tt>' Instruction</a></li>
186 <li><a href="#i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a></li>
187 <li><a href="#i_fpext">'<tt>fpext .. to</tt>' Instruction</a></li>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]188 <li><a href="#i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a></li>
189 <li><a href="#i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a></li>
190 <li><a href="#i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a></li>
191 <li><a href="#i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a></li>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]192 <li><a href="#i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a></li>
193 <li><a href="#i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a></li>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]194 <li><a href="#i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a></li>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]195 </ol>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]196 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]197 <li><a href="#otherops">Other Operations</a>
198 <ol>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]199 <li><a href="#i_icmp">'<tt>icmp</tt>' Instruction</a></li>
200 <li><a href="#i_fcmp">'<tt>fcmp</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]201 <li><a href="#i_phi">'<tt>phi</tt>' Instruction</a></li>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]202 <li><a href="#i_select">'<tt>select</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]203 <li><a href="#i_call">'<tt>call</tt>' Instruction</a></li>
Chris Lattnerfb6977d2006-01-13 23:26:01 +0000[diff] [blame]204 <li><a href="#i_va_arg">'<tt>va_arg</tt>' Instruction</a></li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]205 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]206 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]207 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]208 </li>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]209 <li><a href="#intrinsics">Intrinsic Functions</a>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]210 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]211 <li><a href="#int_varargs">Variable Argument Handling Intrinsics</a>
212 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]213 <li><a href="#int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a></li>
214 <li><a href="#int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a></li>
215 <li><a href="#int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]216 </ol>
217 </li>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]218 <li><a href="#int_gc">Accurate Garbage Collection Intrinsics</a>
219 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]220 <li><a href="#int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a></li>
221 <li><a href="#int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a></li>
222 <li><a href="#int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a></li>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]223 </ol>
224 </li>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]225 <li><a href="#int_codegen">Code Generator Intrinsics</a>
226 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]227 <li><a href="#int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a></li>
228 <li><a href="#int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a></li>
229 <li><a href="#int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a></li>
230 <li><a href="#int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a></li>
231 <li><a href="#int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a></li>
232 <li><a href="#int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a></li>
Dan Gohman31f1af12010-05-26 21:56:15 +0000[diff] [blame]233 <li><a href="#int_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a></li>
John Criswell7123e272004-04-09 16:43:20 +0000[diff] [blame]234 </ol>
235 </li>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]236 <li><a href="#int_libc">Standard C Library Intrinsics</a>
237 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]238 <li><a href="#int_memcpy">'<tt>llvm.memcpy.*</tt>' Intrinsic</a></li>
239 <li><a href="#int_memmove">'<tt>llvm.memmove.*</tt>' Intrinsic</a></li>
240 <li><a href="#int_memset">'<tt>llvm.memset.*</tt>' Intrinsic</a></li>
241 <li><a href="#int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a></li>
242 <li><a href="#int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a></li>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]243 <li><a href="#int_sin">'<tt>llvm.sin.*</tt>' Intrinsic</a></li>
244 <li><a href="#int_cos">'<tt>llvm.cos.*</tt>' Intrinsic</a></li>
245 <li><a href="#int_pow">'<tt>llvm.pow.*</tt>' Intrinsic</a></li>
Dan Gohman08b280b2011-05-27 00:36:31 +0000[diff] [blame]246 <li><a href="#int_exp">'<tt>llvm.exp.*</tt>' Intrinsic</a></li>
247 <li><a href="#int_log">'<tt>llvm.log.*</tt>' Intrinsic</a></li>
Cameron Zwarich33390842011-07-08 21:39:21 +0000[diff] [blame]248 <li><a href="#int_fma">'<tt>llvm.fma.*</tt>' Intrinsic</a></li>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]249 </ol>
250 </li>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]251 <li><a href="#int_manip">Bit Manipulation Intrinsics</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]252 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]253 <li><a href="#int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a></li>
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]254 <li><a href="#int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic </a></li>
255 <li><a href="#int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic </a></li>
256 <li><a href="#int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic </a></li>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]257 </ol>
258 </li>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]259 <li><a href="#int_overflow">Arithmetic with Overflow Intrinsics</a>
260 <ol>
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]261 <li><a href="#int_sadd_overflow">'<tt>llvm.sadd.with.overflow.*</tt> Intrinsics</a></li>
262 <li><a href="#int_uadd_overflow">'<tt>llvm.uadd.with.overflow.*</tt> Intrinsics</a></li>
263 <li><a href="#int_ssub_overflow">'<tt>llvm.ssub.with.overflow.*</tt> Intrinsics</a></li>
264 <li><a href="#int_usub_overflow">'<tt>llvm.usub.with.overflow.*</tt> Intrinsics</a></li>
265 <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]266 <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]267 </ol>
268 </li>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]269 <li><a href="#int_fp16">Half Precision Floating Point Intrinsics</a>
270 <ol>
Chris Lattner82c3dc62010-03-14 23:03:31 +0000[diff] [blame]271 <li><a href="#int_convert_to_fp16">'<tt>llvm.convert.to.fp16</tt>' Intrinsic</a></li>
272 <li><a href="#int_convert_from_fp16">'<tt>llvm.convert.from.fp16</tt>' Intrinsic</a></li>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]273 </ol>
274 </li>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]275 <li><a href="#int_debugger">Debugger intrinsics</a></li>
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]276 <li><a href="#int_eh">Exception Handling intrinsics</a></li>
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]277 <li><a href="#int_trampoline">Trampoline Intrinsic</a>
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]278 <ol>
279 <li><a href="#int_it">'<tt>llvm.init.trampoline</tt>' Intrinsic</a></li>
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]280 </ol>
281 </li>
Bill Wendling3c44f5b2008-11-18 22:10:53 +0000[diff] [blame]282 <li><a href="#int_atomics">Atomic intrinsics</a>
283 <ol>
284 <li><a href="#int_memory_barrier"><tt>llvm.memory_barrier</tt></a></li>
285 <li><a href="#int_atomic_cmp_swap"><tt>llvm.atomic.cmp.swap</tt></a></li>
286 <li><a href="#int_atomic_swap"><tt>llvm.atomic.swap</tt></a></li>
287 <li><a href="#int_atomic_load_add"><tt>llvm.atomic.load.add</tt></a></li>
288 <li><a href="#int_atomic_load_sub"><tt>llvm.atomic.load.sub</tt></a></li>
289 <li><a href="#int_atomic_load_and"><tt>llvm.atomic.load.and</tt></a></li>
290 <li><a href="#int_atomic_load_nand"><tt>llvm.atomic.load.nand</tt></a></li>
291 <li><a href="#int_atomic_load_or"><tt>llvm.atomic.load.or</tt></a></li>
292 <li><a href="#int_atomic_load_xor"><tt>llvm.atomic.load.xor</tt></a></li>
293 <li><a href="#int_atomic_load_max"><tt>llvm.atomic.load.max</tt></a></li>
294 <li><a href="#int_atomic_load_min"><tt>llvm.atomic.load.min</tt></a></li>
295 <li><a href="#int_atomic_load_umax"><tt>llvm.atomic.load.umax</tt></a></li>
296 <li><a href="#int_atomic_load_umin"><tt>llvm.atomic.load.umin</tt></a></li>
297 </ol>
298 </li>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]299 <li><a href="#int_memorymarkers">Memory Use Markers</a>
300 <ol>
301 <li><a href="#int_lifetime_start"><tt>llvm.lifetime.start</tt></a></li>
302 <li><a href="#int_lifetime_end"><tt>llvm.lifetime.end</tt></a></li>
303 <li><a href="#int_invariant_start"><tt>llvm.invariant.start</tt></a></li>
304 <li><a href="#int_invariant_end"><tt>llvm.invariant.end</tt></a></li>
305 </ol>
306 </li>
Reid Spencer20677642007-07-20 19:59:11 +0000[diff] [blame]307 <li><a href="#int_general">General intrinsics</a>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]308 <ol>
Reid Spencer20677642007-07-20 19:59:11 +0000[diff] [blame]309 <li><a href="#int_var_annotation">
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]310 '<tt>llvm.var.annotation</tt>' Intrinsic</a></li>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]311 <li><a href="#int_annotation">
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]312 '<tt>llvm.annotation.*</tt>' Intrinsic</a></li>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]313 <li><a href="#int_trap">
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]314 '<tt>llvm.trap</tt>' Intrinsic</a></li>
315 <li><a href="#int_stackprotector">
316 '<tt>llvm.stackprotector</tt>' Intrinsic</a></li>
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]317 <li><a href="#int_objectsize">
318 '<tt>llvm.objectsize</tt>' Intrinsic</a></li>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]319 </ol>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]320 </li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]321 </ol>
322 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]323</ol>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]324
325<div class="doc_author">
326 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>
327 and <a href="mailto:vadve@cs.uiuc.edu">Vikram Adve</a></p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]328</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]329
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]330<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]331<h2><a name="abstract">Abstract</a></h2>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]332<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]333
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]334<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]335
336<p>This document is a reference manual for the LLVM assembly language. LLVM is
337 a Static Single Assignment (SSA) based representation that provides type
338 safety, low-level operations, flexibility, and the capability of representing
339 'all' high-level languages cleanly. It is the common code representation
340 used throughout all phases of the LLVM compilation strategy.</p>
341
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]342</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]343
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]344<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]345<h2><a name="introduction">Introduction</a></h2>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]346<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]347
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]348<div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]349
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]350<p>The LLVM code representation is designed to be used in three different forms:
351 as an in-memory compiler IR, as an on-disk bitcode representation (suitable
352 for fast loading by a Just-In-Time compiler), and as a human readable
353 assembly language representation. This allows LLVM to provide a powerful
354 intermediate representation for efficient compiler transformations and
355 analysis, while providing a natural means to debug and visualize the
356 transformations. The three different forms of LLVM are all equivalent. This
357 document describes the human readable representation and notation.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]358
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]359<p>The LLVM representation aims to be light-weight and low-level while being
360 expressive, typed, and extensible at the same time. It aims to be a
361 "universal IR" of sorts, by being at a low enough level that high-level ideas
362 may be cleanly mapped to it (similar to how microprocessors are "universal
363 IR's", allowing many source languages to be mapped to them). By providing
364 type information, LLVM can be used as the target of optimizations: for
365 example, through pointer analysis, it can be proven that a C automatic
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]366 variable is never accessed outside of the current function, allowing it to
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]367 be promoted to a simple SSA value instead of a memory location.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]368
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]369<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]370<h4>
371 <a name="wellformed">Well-Formedness</a>
372</h4>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]373
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]374<div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]375
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]376<p>It is important to note that this document describes 'well formed' LLVM
377 assembly language. There is a difference between what the parser accepts and
378 what is considered 'well formed'. For example, the following instruction is
379 syntactically okay, but not well formed:</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]380
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]381<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]382%x = <a href="#i_add">add</a> i32 1, %x
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]383</pre>
384
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]385<p>because the definition of <tt>%x</tt> does not dominate all of its uses. The
386 LLVM infrastructure provides a verification pass that may be used to verify
387 that an LLVM module is well formed. This pass is automatically run by the
388 parser after parsing input assembly and by the optimizer before it outputs
389 bitcode. The violations pointed out by the verifier pass indicate bugs in
390 transformation passes or input to the parser.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]391
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]392</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]393
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]394</div>
395
Chris Lattnercc689392007-10-03 17:34:29 +0000[diff] [blame]396<!-- Describe the typesetting conventions here. -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]397
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]398<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]399<h2><a name="identifiers">Identifiers</a></h2>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]400<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]401
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]402<div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]403
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]404<p>LLVM identifiers come in two basic types: global and local. Global
405 identifiers (functions, global variables) begin with the <tt>'@'</tt>
406 character. Local identifiers (register names, types) begin with
407 the <tt>'%'</tt> character. Additionally, there are three different formats
408 for identifiers, for different purposes:</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]409
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]410<ol>
Reid Spencer2c452282007-08-07 14:34:28 +0000[diff] [blame]411 <li>Named values are represented as a string of characters with their prefix.
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]412 For example, <tt>%foo</tt>, <tt>@DivisionByZero</tt>,
413 <tt>%a.really.long.identifier</tt>. The actual regular expression used is
414 '<tt>[%@][a-zA-Z$._][a-zA-Z$._0-9]*</tt>'. Identifiers which require
415 other characters in their names can be surrounded with quotes. Special
416 characters may be escaped using <tt>"\xx"</tt> where <tt>xx</tt> is the
417 ASCII code for the character in hexadecimal. In this way, any character
418 can be used in a name value, even quotes themselves.</li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]419
Reid Spencer2c452282007-08-07 14:34:28 +0000[diff] [blame]420 <li>Unnamed values are represented as an unsigned numeric value with their
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]421 prefix. For example, <tt>%12</tt>, <tt>@2</tt>, <tt>%44</tt>.</li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]422
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]423 <li>Constants, which are described in a <a href="#constants">section about
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]424 constants</a>, below.</li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]425</ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]426
Reid Spencer2c452282007-08-07 14:34:28 +0000[diff] [blame]427<p>LLVM requires that values start with a prefix for two reasons: Compilers
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]428 don't need to worry about name clashes with reserved words, and the set of
429 reserved words may be expanded in the future without penalty. Additionally,
430 unnamed identifiers allow a compiler to quickly come up with a temporary
431 variable without having to avoid symbol table conflicts.</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]432
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]433<p>Reserved words in LLVM are very similar to reserved words in other
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]434 languages. There are keywords for different opcodes
435 ('<tt><a href="#i_add">add</a></tt>',
436 '<tt><a href="#i_bitcast">bitcast</a></tt>',
437 '<tt><a href="#i_ret">ret</a></tt>', etc...), for primitive type names
438 ('<tt><a href="#t_void">void</a></tt>',
439 '<tt><a href="#t_primitive">i32</a></tt>', etc...), and others. These
440 reserved words cannot conflict with variable names, because none of them
441 start with a prefix character (<tt>'%'</tt> or <tt>'@'</tt>).</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]442
443<p>Here is an example of LLVM code to multiply the integer variable
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]444 '<tt>%X</tt>' by 8:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]445
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]446<p>The easy way:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]447
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]448<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]449%result = <a href="#i_mul">mul</a> i32 %X, 8
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]450</pre>
451
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]452<p>After strength reduction:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]453
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]454<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]455%result = <a href="#i_shl">shl</a> i32 %X, i8 3
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]456</pre>
457
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]458<p>And the hard way:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]459
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]460<pre class="doc_code">
Gabor Greifec58f752009-10-28 13:05:07 +0000[diff] [blame]461%0 = <a href="#i_add">add</a> i32 %X, %X <i>; yields {i32}:%0</i>
462%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]463%result = <a href="#i_add">add</a> i32 %1, %1
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]464</pre>
465
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]466<p>This last way of multiplying <tt>%X</tt> by 8 illustrates several important
467 lexical features of LLVM:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]468
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]469<ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]470 <li>Comments are delimited with a '<tt>;</tt>' and go until the end of
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]471 line.</li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]472
473 <li>Unnamed temporaries are created when the result of a computation is not
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]474 assigned to a named value.</li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]475
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]476 <li>Unnamed temporaries are numbered sequentially</li>
477</ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]478
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]479<p>It also shows a convention that we follow in this document. When
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]480 demonstrating instructions, we will follow an instruction with a comment that
481 defines the type and name of value produced. Comments are shown in italic
482 text.</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]483
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]484</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]485
486<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]487<h2><a name="highlevel">High Level Structure</a></h2>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]488<!-- *********************************************************************** -->
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]489<div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]490<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]491<h3>
492 <a name="modulestructure">Module Structure</a>
493</h3>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]494
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]495<div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]496
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]497<p>LLVM programs are composed of "Module"s, each of which is a translation unit
498 of the input programs. Each module consists of functions, global variables,
499 and symbol table entries. Modules may be combined together with the LLVM
500 linker, which merges function (and global variable) definitions, resolves
501 forward declarations, and merges symbol table entries. Here is an example of
502 the "hello world" module:</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]503
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]504<pre class="doc_code">
Chris Lattner63e4ccb2010-08-17 17:13:42 +0000[diff] [blame]505<i>; Declare the string constant as a global constant.</i>&nbsp;
Nick Lewyckydb9cd762011-01-29 01:09:53 +0000[diff] [blame]506<a href="#identifiers">@.LC0</a> = <a href="#linkage_internal">internal</a>&nbsp;<a href="#globalvars">constant</a>&nbsp;<a href="#t_array">[13 x i8]</a> c"hello world\0A\00" <i>; [13 x i8]*</i>&nbsp;
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]507
Chris Lattner63e4ccb2010-08-17 17:13:42 +0000[diff] [blame]508<i>; External declaration of the puts function</i>&nbsp;
509<a href="#functionstructure">declare</a> i32 @puts(i8*) <i>; i32 (i8*)* </i>&nbsp;
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]510
511<i>; Definition of main function</i>
Chris Lattner63e4ccb2010-08-17 17:13:42 +0000[diff] [blame]512define i32 @main() { <i>; i32()* </i>&nbsp;
513 <i>; Convert [13 x i8]* to i8 *...</i>&nbsp;
514 %cast210 = <a href="#i_getelementptr">getelementptr</a> [13 x i8]* @.LC0, i64 0, i64 0 <i>; i8*</i>&nbsp;
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]515
Chris Lattner63e4ccb2010-08-17 17:13:42 +0000[diff] [blame]516 <i>; Call puts function to write out the string to stdout.</i>&nbsp;
517 <a href="#i_call">call</a> i32 @puts(i8* %cast210) <i>; i32</i>&nbsp;
518 <a href="#i_ret">ret</a> i32 0&nbsp;
519}
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]520
521<i>; Named metadata</i>
522!1 = metadata !{i32 41}
523!foo = !{!1, null}
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]524</pre>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]525
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]526<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]527 "<tt>.LC0</tt>", an external declaration of the "<tt>puts</tt>" function,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]528 a <a href="#functionstructure">function definition</a> for
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]529 "<tt>main</tt>" and <a href="#namedmetadatastructure">named metadata</a>
530 "<tt>foo"</tt>.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]531
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]532<p>In general, a module is made up of a list of global values, where both
533 functions and global variables are global values. Global values are
534 represented by a pointer to a memory location (in this case, a pointer to an
535 array of char, and a pointer to a function), and have one of the
536 following <a href="#linkage">linkage types</a>.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]537
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]538</div>
539
540<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]541<h3>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]542 <a name="linkage">Linkage Types</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]543</h3>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]544
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]545<div>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]546
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]547<p>All Global Variables and Functions have one of the following types of
548 linkage:</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]549
550<dl>
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]551 <dt><tt><b><a name="linkage_private">private</a></b></tt></dt>
Bill Wendling5e721d72010-07-01 21:55:59 +0000[diff] [blame]552 <dd>Global values with "<tt>private</tt>" linkage are only directly accessible
553 by objects in the current module. In particular, linking code into a
554 module with an private global value may cause the private to be renamed as
555 necessary to avoid collisions. Because the symbol is private to the
556 module, all references can be updated. This doesn't show up in any symbol
557 table in the object file.</dd>
Rafael Espindolabb46f522009-01-15 20:18:42 +0000[diff] [blame]558
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]559 <dt><tt><b><a name="linkage_linker_private">linker_private</a></b></tt></dt>
Bill Wendling5e721d72010-07-01 21:55:59 +0000[diff] [blame]560 <dd>Similar to <tt>private</tt>, but the symbol is passed through the
561 assembler and evaluated by the linker. Unlike normal strong symbols, they
562 are removed by the linker from the final linked image (executable or
563 dynamic library).</dd>
564
565 <dt><tt><b><a name="linkage_linker_private_weak">linker_private_weak</a></b></tt></dt>
566 <dd>Similar to "<tt>linker_private</tt>", but the symbol is weak. Note that
567 <tt>linker_private_weak</tt> symbols are subject to coalescing by the
568 linker. The symbols are removed by the linker from the final linked image
569 (executable or dynamic library).</dd>
Bill Wendling3d10a5a2009-07-20 01:03:30 +0000[diff] [blame]570
Bill Wendling55ae5152010-08-20 22:05:50 +0000[diff] [blame]571 <dt><tt><b><a name="linkage_linker_private_weak_def_auto">linker_private_weak_def_auto</a></b></tt></dt>
572 <dd>Similar to "<tt>linker_private_weak</tt>", but it's known that the address
573 of the object is not taken. For instance, functions that had an inline
574 definition, but the compiler decided not to inline it. Note,
575 unlike <tt>linker_private</tt> and <tt>linker_private_weak</tt>,
576 <tt>linker_private_weak_def_auto</tt> may have only <tt>default</tt>
577 visibility. The symbols are removed by the linker from the final linked
578 image (executable or dynamic library).</dd>
579
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]580 <dt><tt><b><a name="linkage_internal">internal</a></b></tt></dt>
Bill Wendling07d31772010-06-29 22:34:52 +0000[diff] [blame]581 <dd>Similar to private, but the value shows as a local symbol
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]582 (<tt>STB_LOCAL</tt> in the case of ELF) in the object file. This
583 corresponds to the notion of the '<tt>static</tt>' keyword in C.</dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]584
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]585 <dt><tt><b><a name="linkage_available_externally">available_externally</a></b></tt></dt>
Chris Lattner266c7bb2009-04-13 05:44:34 +0000[diff] [blame]586 <dd>Globals with "<tt>available_externally</tt>" linkage are never emitted
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]587 into the object file corresponding to the LLVM module. They exist to
588 allow inlining and other optimizations to take place given knowledge of
589 the definition of the global, which is known to be somewhere outside the
590 module. Globals with <tt>available_externally</tt> linkage are allowed to
591 be discarded at will, and are otherwise the same as <tt>linkonce_odr</tt>.
592 This linkage type is only allowed on definitions, not declarations.</dd>
Chris Lattner266c7bb2009-04-13 05:44:34 +0000[diff] [blame]593
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]594 <dt><tt><b><a name="linkage_linkonce">linkonce</a></b></tt></dt>
Chris Lattner4887bd82007-01-14 06:51:48 +0000[diff] [blame]595 <dd>Globals with "<tt>linkonce</tt>" linkage are merged with other globals of
Chris Lattner873187c2010-01-09 19:15:14 +0000[diff] [blame]596 the same name when linkage occurs. This can be used to implement
597 some forms of inline functions, templates, or other code which must be
598 generated in each translation unit that uses it, but where the body may
599 be overridden with a more definitive definition later. Unreferenced
600 <tt>linkonce</tt> globals are allowed to be discarded. Note that
601 <tt>linkonce</tt> linkage does not actually allow the optimizer to
602 inline the body of this function into callers because it doesn't know if
603 this definition of the function is the definitive definition within the
604 program or whether it will be overridden by a stronger definition.
605 To enable inlining and other optimizations, use "<tt>linkonce_odr</tt>"
606 linkage.</dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]607
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]608 <dt><tt><b><a name="linkage_weak">weak</a></b></tt></dt>
Chris Lattner26d054d2009-08-05 05:21:07 +0000[diff] [blame]609 <dd>"<tt>weak</tt>" linkage has the same merging semantics as
610 <tt>linkonce</tt> linkage, except that unreferenced globals with
611 <tt>weak</tt> linkage may not be discarded. This is used for globals that
612 are declared "weak" in C source code.</dd>
613
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]614 <dt><tt><b><a name="linkage_common">common</a></b></tt></dt>
Chris Lattner26d054d2009-08-05 05:21:07 +0000[diff] [blame]615 <dd>"<tt>common</tt>" linkage is most similar to "<tt>weak</tt>" linkage, but
616 they are used for tentative definitions in C, such as "<tt>int X;</tt>" at
617 global scope.
618 Symbols with "<tt>common</tt>" linkage are merged in the same way as
619 <tt>weak symbols</tt>, and they may not be deleted if unreferenced.
Chris Lattnercd81f5d2009-08-05 05:41:44 +0000[diff] [blame]620 <tt>common</tt> symbols may not have an explicit section,
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]621 must have a zero initializer, and may not be marked '<a
Chris Lattnercd81f5d2009-08-05 05:41:44 +0000[diff] [blame]622 href="#globalvars"><tt>constant</tt></a>'. Functions and aliases may not
623 have common linkage.</dd>
Chris Lattner26d054d2009-08-05 05:21:07 +0000[diff] [blame]624
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]625
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]626 <dt><tt><b><a name="linkage_appending">appending</a></b></tt></dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]627 <dd>"<tt>appending</tt>" linkage may only be applied to global variables of
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]628 pointer to array type. When two global variables with appending linkage
629 are linked together, the two global arrays are appended together. This is
630 the LLVM, typesafe, equivalent of having the system linker append together
631 "sections" with identical names when .o files are linked.</dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]632
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]633 <dt><tt><b><a name="linkage_externweak">extern_weak</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]634 <dd>The semantics of this linkage follow the ELF object file model: the symbol
635 is weak until linked, if not linked, the symbol becomes null instead of
636 being an undefined reference.</dd>
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame]637
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]638 <dt><tt><b><a name="linkage_linkonce_odr">linkonce_odr</a></b></tt></dt>
639 <dt><tt><b><a name="linkage_weak_odr">weak_odr</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]640 <dd>Some languages allow differing globals to be merged, such as two functions
641 with different semantics. Other languages, such as <tt>C++</tt>, ensure
Bill Wendling5e721d72010-07-01 21:55:59 +0000[diff] [blame]642 that only equivalent globals are ever merged (the "one definition rule"
643 &mdash; "ODR"). Such languages can use the <tt>linkonce_odr</tt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]644 and <tt>weak_odr</tt> linkage types to indicate that the global will only
645 be merged with equivalent globals. These linkage types are otherwise the
646 same as their non-<tt>odr</tt> versions.</dd>
Duncan Sands667d4b82009-03-07 15:45:40 +0000[diff] [blame]647
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]648 <dt><tt><b><a name="linkage_external">externally visible</a></b></tt>:</dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]649 <dd>If none of the above identifiers are used, the global is externally
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]650 visible, meaning that it participates in linkage and can be used to
651 resolve external symbol references.</dd>
Reid Spencerc8910842007-04-11 23:49:50 +0000[diff] [blame]652</dl>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]653
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]654<p>The next two types of linkage are targeted for Microsoft Windows platform
655 only. They are designed to support importing (exporting) symbols from (to)
656 DLLs (Dynamic Link Libraries).</p>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]657
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]658<dl>
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]659 <dt><tt><b><a name="linkage_dllimport">dllimport</a></b></tt></dt>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]660 <dd>"<tt>dllimport</tt>" linkage causes the compiler to reference a function
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]661 or variable via a global pointer to a pointer that is set up by the DLL
662 exporting the symbol. On Microsoft Windows targets, the pointer name is
663 formed by combining <code>__imp_</code> and the function or variable
664 name.</dd>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]665
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]666 <dt><tt><b><a name="linkage_dllexport">dllexport</a></b></tt></dt>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]667 <dd>"<tt>dllexport</tt>" linkage causes the compiler to provide a global
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]668 pointer to a pointer in a DLL, so that it can be referenced with the
669 <tt>dllimport</tt> attribute. On Microsoft Windows targets, the pointer
670 name is formed by combining <code>__imp_</code> and the function or
671 variable name.</dd>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]672</dl>
673
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]674<p>For example, since the "<tt>.LC0</tt>" variable is defined to be internal, if
675 another module defined a "<tt>.LC0</tt>" variable and was linked with this
676 one, one of the two would be renamed, preventing a collision. Since
677 "<tt>main</tt>" and "<tt>puts</tt>" are external (i.e., lacking any linkage
678 declarations), they are accessible outside of the current module.</p>
679
680<p>It is illegal for a function <i>declaration</i> to have any linkage type
681 other than "externally visible", <tt>dllimport</tt>
682 or <tt>extern_weak</tt>.</p>
683
Duncan Sands667d4b82009-03-07 15:45:40 +0000[diff] [blame]684<p>Aliases can have only <tt>external</tt>, <tt>internal</tt>, <tt>weak</tt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]685 or <tt>weak_odr</tt> linkages.</p>
686
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]687</div>
688
689<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]690<h3>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]691 <a name="callingconv">Calling Conventions</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]692</h3>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]693
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]694<div>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]695
696<p>LLVM <a href="#functionstructure">functions</a>, <a href="#i_call">calls</a>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]697 and <a href="#i_invoke">invokes</a> can all have an optional calling
698 convention specified for the call. The calling convention of any pair of
699 dynamic caller/callee must match, or the behavior of the program is
700 undefined. The following calling conventions are supported by LLVM, and more
701 may be added in the future:</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]702
703<dl>
704 <dt><b>"<tt>ccc</tt>" - The C calling convention</b>:</dt>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]705 <dd>This calling convention (the default if no other calling convention is
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]706 specified) matches the target C calling conventions. This calling
707 convention supports varargs function calls and tolerates some mismatch in
708 the declared prototype and implemented declaration of the function (as
709 does normal C).</dd>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]710
711 <dt><b>"<tt>fastcc</tt>" - The fast calling convention</b>:</dt>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]712 <dd>This calling convention attempts to make calls as fast as possible
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]713 (e.g. by passing things in registers). This calling convention allows the
714 target to use whatever tricks it wants to produce fast code for the
715 target, without having to conform to an externally specified ABI
Jeffrey Yasskin95fa80a2010-01-09 19:44:16 +0000[diff] [blame]716 (Application Binary Interface).
717 <a href="CodeGenerator.html#tailcallopt">Tail calls can only be optimized
Chris Lattner29689432010-03-11 00:22:57 +0000[diff] [blame]718 when this or the GHC convention is used.</a> This calling convention
719 does not support varargs and requires the prototype of all callees to
720 exactly match the prototype of the function definition.</dd>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]721
722 <dt><b>"<tt>coldcc</tt>" - The cold calling convention</b>:</dt>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]723 <dd>This calling convention attempts to make code in the caller as efficient
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]724 as possible under the assumption that the call is not commonly executed.
725 As such, these calls often preserve all registers so that the call does
726 not break any live ranges in the caller side. This calling convention
727 does not support varargs and requires the prototype of all callees to
728 exactly match the prototype of the function definition.</dd>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]729
Chris Lattner29689432010-03-11 00:22:57 +0000[diff] [blame]730 <dt><b>"<tt>cc <em>10</em></tt>" - GHC convention</b>:</dt>
731 <dd>This calling convention has been implemented specifically for use by the
732 <a href="http://www.haskell.org/ghc">Glasgow Haskell Compiler (GHC)</a>.
733 It passes everything in registers, going to extremes to achieve this by
734 disabling callee save registers. This calling convention should not be
735 used lightly but only for specific situations such as an alternative to
736 the <em>register pinning</em> performance technique often used when
737 implementing functional programming languages.At the moment only X86
738 supports this convention and it has the following limitations:
739 <ul>
740 <li>On <em>X86-32</em> only supports up to 4 bit type parameters. No
741 floating point types are supported.</li>
742 <li>On <em>X86-64</em> only supports up to 10 bit type parameters and
743 6 floating point parameters.</li>
744 </ul>
745 This calling convention supports
746 <a href="CodeGenerator.html#tailcallopt">tail call optimization</a> but
747 requires both the caller and callee are using it.
748 </dd>
749
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]750 <dt><b>"<tt>cc &lt;<em>n</em>&gt;</tt>" - Numbered convention</b>:</dt>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]751 <dd>Any calling convention may be specified by number, allowing
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]752 target-specific calling conventions to be used. Target specific calling
753 conventions start at 64.</dd>
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]754</dl>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]755
756<p>More calling conventions can be added/defined on an as-needed basis, to
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]757 support Pascal conventions or any other well-known target-independent
758 convention.</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]759
760</div>
761
762<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]763<h3>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]764 <a name="visibility">Visibility Styles</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]765</h3>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]766
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]767<div>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]768
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]769<p>All Global Variables and Functions have one of the following visibility
770 styles:</p>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]771
772<dl>
773 <dt><b>"<tt>default</tt>" - Default style</b>:</dt>
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]774 <dd>On targets that use the ELF object file format, default visibility means
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]775 that the declaration is visible to other modules and, in shared libraries,
776 means that the declared entity may be overridden. On Darwin, default
777 visibility means that the declaration is visible to other modules. Default
778 visibility corresponds to "external linkage" in the language.</dd>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]779
780 <dt><b>"<tt>hidden</tt>" - Hidden style</b>:</dt>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]781 <dd>Two declarations of an object with hidden visibility refer to the same
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]782 object if they are in the same shared object. Usually, hidden visibility
783 indicates that the symbol will not be placed into the dynamic symbol
784 table, so no other module (executable or shared library) can reference it
785 directly.</dd>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]786
Anton Korobeynikov6f9896f2007-04-29 18:35:00 +0000[diff] [blame]787 <dt><b>"<tt>protected</tt>" - Protected style</b>:</dt>
Anton Korobeynikov6f9896f2007-04-29 18:35:00 +0000[diff] [blame]788 <dd>On ELF, protected visibility indicates that the symbol will be placed in
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]789 the dynamic symbol table, but that references within the defining module
790 will bind to the local symbol. That is, the symbol cannot be overridden by
791 another module.</dd>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]792</dl>
793
794</div>
795
796<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]797<h3>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]798 <a name="namedtypes">Named Types</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]799</h3>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]800
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]801<div>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]802
803<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]804 it easier to read the IR and make the IR more condensed (particularly when
805 recursive types are involved). An example of a name specification is:</p>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]806
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]807<pre class="doc_code">
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]808%mytype = type { %mytype*, i32 }
809</pre>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]810
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]811<p>You may give a name to any <a href="#typesystem">type</a> except
Chris Lattnerdc65f222010-08-17 23:26:04 +0000[diff] [blame]812 "<a href="#t_void">void</a>". Type name aliases may be used anywhere a type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]813 is expected with the syntax "%mytype".</p>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]814
815<p>Note that type names are aliases for the structural type that they indicate,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]816 and that you can therefore specify multiple names for the same type. This
817 often leads to confusing behavior when dumping out a .ll file. Since LLVM IR
818 uses structural typing, the name is not part of the type. When printing out
819 LLVM IR, the printer will pick <em>one name</em> to render all types of a
820 particular shape. This means that if you have code where two different
821 source types end up having the same LLVM type, that the dumper will sometimes
822 print the "wrong" or unexpected type. This is an important design point and
823 isn't going to change.</p>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]824
825</div>
826
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]827<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]828<h3>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]829 <a name="globalvars">Global Variables</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]830</h3>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]831
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]832<div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]833
Chris Lattner3689a342005-02-12 19:30:21 +0000[diff] [blame]834<p>Global variables define regions of memory allocated at compilation time
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]835 instead of run-time. Global variables may optionally be initialized, may
836 have an explicit section to be placed in, and may have an optional explicit
837 alignment specified. A variable may be defined as "thread_local", which
838 means that it will not be shared by threads (each thread will have a
839 separated copy of the variable). A variable may be defined as a global
840 "constant," which indicates that the contents of the variable
841 will <b>never</b> be modified (enabling better optimization, allowing the
842 global data to be placed in the read-only section of an executable, etc).
843 Note that variables that need runtime initialization cannot be marked
844 "constant" as there is a store to the variable.</p>
Chris Lattner3689a342005-02-12 19:30:21 +0000[diff] [blame]845
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]846<p>LLVM explicitly allows <em>declarations</em> of global variables to be marked
847 constant, even if the final definition of the global is not. This capability
848 can be used to enable slightly better optimization of the program, but
849 requires the language definition to guarantee that optimizations based on the
850 'constantness' are valid for the translation units that do not include the
851 definition.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]852
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]853<p>As SSA values, global variables define pointer values that are in scope
854 (i.e. they dominate) all basic blocks in the program. Global variables
855 always define a pointer to their "content" type because they describe a
856 region of memory, and all memory objects in LLVM are accessed through
857 pointers.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]858
Rafael Espindolabea46262011-01-08 16:42:36 +0000[diff] [blame]859<p>Global variables can be marked with <tt>unnamed_addr</tt> which indicates
860 that the address is not significant, only the content. Constants marked
Rafael Espindolaa5eaa862011-01-15 08:20:57 +0000[diff] [blame]861 like this can be merged with other constants if they have the same
862 initializer. Note that a constant with significant address <em>can</em>
863 be merged with a <tt>unnamed_addr</tt> constant, the result being a
864 constant whose address is significant.</p>
Rafael Espindolabea46262011-01-08 16:42:36 +0000[diff] [blame]865
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]866<p>A global variable may be declared to reside in a target-specific numbered
867 address space. For targets that support them, address spaces may affect how
868 optimizations are performed and/or what target instructions are used to
869 access the variable. The default address space is zero. The address space
870 qualifier must precede any other attributes.</p>
Christopher Lamb284d9922007-12-11 09:31:00 +0000[diff] [blame]871
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]872<p>LLVM allows an explicit section to be specified for globals. If the target
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]873 supports it, it will emit globals to the section specified.</p>
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]874
Chris Lattnerce99fa92010-04-28 00:13:42 +0000[diff] [blame]875<p>An explicit alignment may be specified for a global, which must be a power
876 of 2. If not present, or if the alignment is set to zero, the alignment of
877 the global is set by the target to whatever it feels convenient. If an
878 explicit alignment is specified, the global is forced to have exactly that
Chris Lattner2d4b8ee2010-04-28 00:31:12 +0000[diff] [blame]879 alignment. Targets and optimizers are not allowed to over-align the global
880 if the global has an assigned section. In this case, the extra alignment
881 could be observable: for example, code could assume that the globals are
882 densely packed in their section and try to iterate over them as an array,
883 alignment padding would break this iteration.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]884
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]885<p>For example, the following defines a global in a numbered address space with
886 an initializer, section, and alignment:</p>
Chris Lattner68027ea2007-01-14 00:27:09 +0000[diff] [blame]887
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]888<pre class="doc_code">
Dan Gohman398873c2009-01-11 00:40:00 +0000[diff] [blame]889@G = addrspace(5) constant float 1.0, section "foo", align 4
Chris Lattner68027ea2007-01-14 00:27:09 +0000[diff] [blame]890</pre>
891
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]892</div>
893
894
895<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]896<h3>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]897 <a name="functionstructure">Functions</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]898</h3>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]899
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]900<div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]901
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]902<p>LLVM function definitions consist of the "<tt>define</tt>" keyword, an
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]903 optional <a href="#linkage">linkage type</a>, an optional
904 <a href="#visibility">visibility style</a>, an optional
Rafael Espindolabea46262011-01-08 16:42:36 +0000[diff] [blame]905 <a href="#callingconv">calling convention</a>,
906 an optional <tt>unnamed_addr</tt> attribute, a return type, an optional
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]907 <a href="#paramattrs">parameter attribute</a> for the return type, a function
908 name, a (possibly empty) argument list (each with optional
909 <a href="#paramattrs">parameter attributes</a>), optional
910 <a href="#fnattrs">function attributes</a>, an optional section, an optional
911 alignment, an optional <a href="#gc">garbage collector name</a>, an opening
912 curly brace, a list of basic blocks, and a closing curly brace.</p>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]913
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]914<p>LLVM function declarations consist of the "<tt>declare</tt>" keyword, an
915 optional <a href="#linkage">linkage type</a>, an optional
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]916 <a href="#visibility">visibility style</a>, an optional
Rafael Espindolabea46262011-01-08 16:42:36 +0000[diff] [blame]917 <a href="#callingconv">calling convention</a>,
918 an optional <tt>unnamed_addr</tt> attribute, a return type, an optional
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]919 <a href="#paramattrs">parameter attribute</a> for the return type, a function
920 name, a possibly empty list of arguments, an optional alignment, and an
921 optional <a href="#gc">garbage collector name</a>.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]922
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]923<p>A function definition contains a list of basic blocks, forming the CFG
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]924 (Control Flow Graph) for the function. Each basic block may optionally start
925 with a label (giving the basic block a symbol table entry), contains a list
926 of instructions, and ends with a <a href="#terminators">terminator</a>
927 instruction (such as a branch or function return).</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]928
Chris Lattner4a3c9012007-06-08 16:52:14 +0000[diff] [blame]929<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]930 executed on entrance to the function, and it is not allowed to have
931 predecessor basic blocks (i.e. there can not be any branches to the entry
932 block of a function). Because the block can have no predecessors, it also
933 cannot have any <a href="#i_phi">PHI nodes</a>.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]934
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]935<p>LLVM allows an explicit section to be specified for functions. If the target
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]936 supports it, it will emit functions to the section specified.</p>
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]937
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]938<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]939 the alignment is set to zero, the alignment of the function is set by the
940 target to whatever it feels convenient. If an explicit alignment is
941 specified, the function is forced to have at least that much alignment. All
942 alignments must be a power of 2.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]943
Rafael Espindolabea46262011-01-08 16:42:36 +0000[diff] [blame]944<p>If the <tt>unnamed_addr</tt> attribute is given, the address is know to not
945 be significant and two identical functions can be merged</p>.
946
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]947<h5>Syntax:</h5>
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]948<pre class="doc_code">
Chris Lattner50ad45c2008-10-13 16:55:18 +0000[diff] [blame]949define [<a href="#linkage">linkage</a>] [<a href="#visibility">visibility</a>]
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]950 [<a href="#callingconv">cconv</a>] [<a href="#paramattrs">ret attrs</a>]
951 &lt;ResultType&gt; @&lt;FunctionName&gt; ([argument list])
952 [<a href="#fnattrs">fn Attrs</a>] [section "name"] [align N]
953 [<a href="#gc">gc</a>] { ... }
954</pre>
Devang Patel307e8ab2008-10-07 17:48:33 +0000[diff] [blame]955
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]956</div>
957
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]958<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]959<h3>
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]960 <a name="aliasstructure">Aliases</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]961</h3>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]962
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]963<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]964
965<p>Aliases act as "second name" for the aliasee value (which can be either
966 function, global variable, another alias or bitcast of global value). Aliases
967 may have an optional <a href="#linkage">linkage type</a>, and an
968 optional <a href="#visibility">visibility style</a>.</p>
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]969
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]970<h5>Syntax:</h5>
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]971<pre class="doc_code">
Duncan Sands0b23ac12008-09-12 20:48:21 +0000[diff] [blame]972@&lt;Name&gt; = alias [Linkage] [Visibility] &lt;AliaseeTy&gt; @&lt;Aliasee&gt;
Bill Wendlingaac388b2007-05-29 09:42:13 +0000[diff] [blame]973</pre>
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]974
975</div>
976
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]977<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]978<h3>
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]979 <a name="namedmetadatastructure">Named Metadata</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]980</h3>
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]981
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]982<div>
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]983
Chris Lattnere6a5ddd2010-01-15 21:50:19 +0000[diff] [blame]984<p>Named metadata is a collection of metadata. <a href="#metadata">Metadata
Dan Gohman872814a2010-07-21 18:54:18 +0000[diff] [blame]985 nodes</a> (but not metadata strings) are the only valid operands for
Chris Lattnere6a5ddd2010-01-15 21:50:19 +0000[diff] [blame]986 a named metadata.</p>
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]987
988<h5>Syntax:</h5>
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]989<pre class="doc_code">
Dan Gohman872814a2010-07-21 18:54:18 +0000[diff] [blame]990; Some unnamed metadata nodes, which are referenced by the named metadata.
991!0 = metadata !{metadata !"zero"}
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]992!1 = metadata !{metadata !"one"}
Dan Gohman872814a2010-07-21 18:54:18 +0000[diff] [blame]993!2 = metadata !{metadata !"two"}
Dan Gohman1005bc52010-07-13 19:48:13 +0000[diff] [blame]994; A named metadata.
Dan Gohman872814a2010-07-21 18:54:18 +0000[diff] [blame]995!name = !{!0, !1, !2}
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]996</pre>
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]997
998</div>
999
1000<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1001<h3>
1002 <a name="paramattrs">Parameter Attributes</a>
1003</h3>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1004
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1005<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1006
1007<p>The return type and each parameter of a function type may have a set of
1008 <i>parameter attributes</i> associated with them. Parameter attributes are
1009 used to communicate additional information about the result or parameters of
1010 a function. Parameter attributes are considered to be part of the function,
1011 not of the function type, so functions with different parameter attributes
1012 can have the same function type.</p>
1013
1014<p>Parameter attributes are simple keywords that follow the type specified. If
1015 multiple parameter attributes are needed, they are space separated. For
1016 example:</p>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1017
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]1018<pre class="doc_code">
Nick Lewyckyb6a7d252009-02-15 23:06:14 +0000[diff] [blame]1019declare i32 @printf(i8* noalias nocapture, ...)
Chris Lattner66d922c2008-10-04 18:33:34 +0000[diff] [blame]1020declare i32 @atoi(i8 zeroext)
1021declare signext i8 @returns_signed_char()
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1022</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1023
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1024<p>Note that any attributes for the function result (<tt>nounwind</tt>,
1025 <tt>readonly</tt>) come immediately after the argument list.</p>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1026
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1027<p>Currently, only the following parameter attributes are defined:</p>
Chris Lattner47507de2008-01-11 06:20:47 +0000[diff] [blame]1028
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1029<dl>
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1030 <dt><tt><b>zeroext</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1031 <dd>This indicates to the code generator that the parameter or return value
Cameron Zwarichebe81732011-03-16 22:20:18 +0000[diff] [blame]1032 should be zero-extended to the extent required by the target's ABI (which
1033 is usually 32-bits, but is 8-bits for a i1 on x86-64) by the caller (for a
1034 parameter) or the callee (for a return value).</dd>
Chris Lattner47507de2008-01-11 06:20:47 +0000[diff] [blame]1035
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1036 <dt><tt><b>signext</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1037 <dd>This indicates to the code generator that the parameter or return value
Cameron Zwarich9e69ff92011-03-17 14:21:58 +0000[diff] [blame]1038 should be sign-extended to the extent required by the target's ABI (which
1039 is usually 32-bits) by the caller (for a parameter) or the callee (for a
1040 return value).</dd>
Chris Lattner47507de2008-01-11 06:20:47 +0000[diff] [blame]1041
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1042 <dt><tt><b>inreg</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1043 <dd>This indicates that this parameter or return value should be treated in a
1044 special target-dependent fashion during while emitting code for a function
1045 call or return (usually, by putting it in a register as opposed to memory,
1046 though some targets use it to distinguish between two different kinds of
1047 registers). Use of this attribute is target-specific.</dd>
Chris Lattner47507de2008-01-11 06:20:47 +0000[diff] [blame]1048
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1049 <dt><tt><b><a name="byval">byval</a></b></tt></dt>
Chris Lattnera6fd81d2010-11-20 23:49:06 +0000[diff] [blame]1050 <dd><p>This indicates that the pointer parameter should really be passed by
1051 value to the function. The attribute implies that a hidden copy of the
1052 pointee
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1053 is made between the caller and the callee, so the callee is unable to
1054 modify the value in the callee. This attribute is only valid on LLVM
1055 pointer arguments. It is generally used to pass structs and arrays by
1056 value, but is also valid on pointers to scalars. The copy is considered
1057 to belong to the caller not the callee (for example,
1058 <tt><a href="#readonly">readonly</a></tt> functions should not write to
1059 <tt>byval</tt> parameters). This is not a valid attribute for return
Chris Lattnera6fd81d2010-11-20 23:49:06 +0000[diff] [blame]1060 values.</p>
1061
1062 <p>The byval attribute also supports specifying an alignment with
1063 the align attribute. It indicates the alignment of the stack slot to
1064 form and the known alignment of the pointer specified to the call site. If
1065 the alignment is not specified, then the code generator makes a
1066 target-specific assumption.</p></dd>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1067
Dan Gohmanff235352010-07-02 23:18:08 +0000[diff] [blame]1068 <dt><tt><b><a name="sret">sret</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1069 <dd>This indicates that the pointer parameter specifies the address of a
1070 structure that is the return value of the function in the source program.
1071 This pointer must be guaranteed by the caller to be valid: loads and
1072 stores to the structure may be assumed by the callee to not to trap. This
1073 may only be applied to the first parameter. This is not a valid attribute
1074 for return values. </dd>
1075
Dan Gohmanff235352010-07-02 23:18:08 +0000[diff] [blame]1076 <dt><tt><b><a name="noalias">noalias</a></b></tt></dt>
Dan Gohman1e109622010-07-02 18:41:32 +0000[diff] [blame]1077 <dd>This indicates that pointer values
1078 <a href="#pointeraliasing"><i>based</i></a> on the argument or return
Dan Gohmanefca7f92010-07-02 23:46:54 +0000[diff] [blame]1079 value do not alias pointer values which are not <i>based</i> on it,
1080 ignoring certain "irrelevant" dependencies.
1081 For a call to the parent function, dependencies between memory
1082 references from before or after the call and from those during the call
1083 are "irrelevant" to the <tt>noalias</tt> keyword for the arguments and
1084 return value used in that call.
Dan Gohman1e109622010-07-02 18:41:32 +0000[diff] [blame]1085 The caller shares the responsibility with the callee for ensuring that
1086 these requirements are met.
1087 For further details, please see the discussion of the NoAlias response in
Dan Gohmanff70fe42010-07-06 15:26:33 +0000[diff] [blame]1088 <a href="AliasAnalysis.html#MustMayNo">alias analysis</a>.<br>
1089<br>
John McCall191d4ee2010-07-06 21:07:14 +0000[diff] [blame]1090 Note that this definition of <tt>noalias</tt> is intentionally
1091 similar to the definition of <tt>restrict</tt> in C99 for function
Chris Lattner211244a2010-07-06 20:51:35 +0000[diff] [blame]1092 arguments, though it is slightly weaker.
Dan Gohmanff70fe42010-07-06 15:26:33 +0000[diff] [blame]1093<br>
1094 For function return values, C99's <tt>restrict</tt> is not meaningful,
1095 while LLVM's <tt>noalias</tt> is.
1096 </dd>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1097
Dan Gohmanff235352010-07-02 23:18:08 +0000[diff] [blame]1098 <dt><tt><b><a name="nocapture">nocapture</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1099 <dd>This indicates that the callee does not make any copies of the pointer
1100 that outlive the callee itself. This is not a valid attribute for return
1101 values.</dd>
1102
Dan Gohmanff235352010-07-02 23:18:08 +0000[diff] [blame]1103 <dt><tt><b><a name="nest">nest</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1104 <dd>This indicates that the pointer parameter can be excised using the
1105 <a href="#int_trampoline">trampoline intrinsics</a>. This is not a valid
1106 attribute for return values.</dd>
1107</dl>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1108
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1109</div>
1110
1111<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1112<h3>
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame]1113 <a name="gc">Garbage Collector Names</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1114</h3>
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame]1115
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1116<div>
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame]1117
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1118<p>Each function may specify a garbage collector name, which is simply a
1119 string:</p>
1120
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]1121<pre class="doc_code">
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1122define void @f() gc "name" { ... }
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1123</pre>
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame]1124
1125<p>The compiler declares the supported values of <i>name</i>. Specifying a
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1126 collector which will cause the compiler to alter its output in order to
1127 support the named garbage collection algorithm.</p>
1128
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame]1129</div>
1130
1131<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1132<h3>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1133 <a name="fnattrs">Function Attributes</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1134</h3>
Devang Patelf8b94812008-09-04 23:05:13 +0000[diff] [blame]1135
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1136<div>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1137
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1138<p>Function attributes are set to communicate additional information about a
1139 function. Function attributes are considered to be part of the function, not
1140 of the function type, so functions with different parameter attributes can
1141 have the same function type.</p>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1142
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1143<p>Function attributes are simple keywords that follow the type specified. If
1144 multiple attributes are needed, they are space separated. For example:</p>
Devang Patelf8b94812008-09-04 23:05:13 +0000[diff] [blame]1145
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]1146<pre class="doc_code">
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1147define void @f() noinline { ... }
1148define void @f() alwaysinline { ... }
1149define void @f() alwaysinline optsize { ... }
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1150define void @f() optsize { ... }
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1151</pre>
Devang Patelf8b94812008-09-04 23:05:13 +0000[diff] [blame]1152
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1153<dl>
Charles Davis1e063d12010-02-12 00:31:15 +0000[diff] [blame]1154 <dt><tt><b>alignstack(&lt;<em>n</em>&gt;)</b></tt></dt>
1155 <dd>This attribute indicates that, when emitting the prologue and epilogue,
1156 the backend should forcibly align the stack pointer. Specify the
1157 desired alignment, which must be a power of two, in parentheses.
1158
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1159 <dt><tt><b>alwaysinline</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1160 <dd>This attribute indicates that the inliner should attempt to inline this
1161 function into callers whenever possible, ignoring any active inlining size
1162 threshold for this caller.</dd>
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1163
Charles Davis970bfcc2010-10-25 15:37:09 +0000[diff] [blame]1164 <dt><tt><b>hotpatch</b></tt></dt>
Charles Davis6f12e292010-10-25 16:29:03 +0000[diff] [blame]1165 <dd>This attribute indicates that the function should be 'hotpatchable',
Charles Davis0076d202010-10-25 19:07:39 +0000[diff] [blame]1166 meaning the function can be patched and/or hooked even while it is
1167 loaded into memory. On x86, the function prologue will be preceded
1168 by six bytes of padding and will begin with a two-byte instruction.
1169 Most of the functions in the Windows system DLLs in Windows XP SP2 or
1170 higher were compiled in this fashion.</dd>
Charles Davis970bfcc2010-10-25 15:37:09 +0000[diff] [blame]1171
Dan Gohman129bd562011-06-16 16:03:13 +0000[diff] [blame]1172 <dt><tt><b>nonlazybind</b></tt></dt>
1173 <dd>This attribute suppresses lazy symbol binding for the function. This
1174 may make calls to the function faster, at the cost of extra program
1175 startup time if the function is not called during program startup.</dd>
1176
Jakob Stoklund Olesen570a4a52010-02-06 01:16:28 +0000[diff] [blame]1177 <dt><tt><b>inlinehint</b></tt></dt>
1178 <dd>This attribute indicates that the source code contained a hint that inlining
1179 this function is desirable (such as the "inline" keyword in C/C++). It
1180 is just a hint; it imposes no requirements on the inliner.</dd>
1181
Nick Lewycky76ec37a2010-07-06 18:24:09 +0000[diff] [blame]1182 <dt><tt><b>naked</b></tt></dt>
1183 <dd>This attribute disables prologue / epilogue emission for the function.
1184 This can have very system-specific consequences.</dd>
1185
1186 <dt><tt><b>noimplicitfloat</b></tt></dt>
1187 <dd>This attributes disables implicit floating point instructions.</dd>
1188
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1189 <dt><tt><b>noinline</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1190 <dd>This attribute indicates that the inliner should never inline this
1191 function in any situation. This attribute may not be used together with
1192 the <tt>alwaysinline</tt> attribute.</dd>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1193
Nick Lewycky76ec37a2010-07-06 18:24:09 +0000[diff] [blame]1194 <dt><tt><b>noredzone</b></tt></dt>
1195 <dd>This attribute indicates that the code generator should not use a red
1196 zone, even if the target-specific ABI normally permits it.</dd>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1197
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1198 <dt><tt><b>noreturn</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1199 <dd>This function attribute indicates that the function never returns
1200 normally. This produces undefined behavior at runtime if the function
1201 ever does dynamically return.</dd>
Bill Wendling31359ba2008-11-13 01:02:51 +0000[diff] [blame]1202
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1203 <dt><tt><b>nounwind</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1204 <dd>This function attribute indicates that the function never returns with an
1205 unwind or exceptional control flow. If the function does unwind, its
1206 runtime behavior is undefined.</dd>
Bill Wendlingfbaa7ed2008-11-26 19:07:40 +0000[diff] [blame]1207
Nick Lewycky76ec37a2010-07-06 18:24:09 +0000[diff] [blame]1208 <dt><tt><b>optsize</b></tt></dt>
1209 <dd>This attribute suggests that optimization passes and code generator passes
1210 make choices that keep the code size of this function low, and otherwise
1211 do optimizations specifically to reduce code size.</dd>
1212
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1213 <dt><tt><b>readnone</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1214 <dd>This attribute indicates that the function computes its result (or decides
1215 to unwind an exception) based strictly on its arguments, without
1216 dereferencing any pointer arguments or otherwise accessing any mutable
1217 state (e.g. memory, control registers, etc) visible to caller functions.
1218 It does not write through any pointer arguments
1219 (including <tt><a href="#byval">byval</a></tt> arguments) and never
1220 changes any state visible to callers. This means that it cannot unwind
1221 exceptions by calling the <tt>C++</tt> exception throwing methods, but
1222 could use the <tt>unwind</tt> instruction.</dd>
Devang Patel5d96fda2009-06-12 19:45:19 +0000[diff] [blame]1223
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1224 <dt><tt><b><a name="readonly">readonly</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1225 <dd>This attribute indicates that the function does not write through any
1226 pointer arguments (including <tt><a href="#byval">byval</a></tt>
1227 arguments) or otherwise modify any state (e.g. memory, control registers,
1228 etc) visible to caller functions. It may dereference pointer arguments
1229 and read state that may be set in the caller. A readonly function always
1230 returns the same value (or unwinds an exception identically) when called
1231 with the same set of arguments and global state. It cannot unwind an
1232 exception by calling the <tt>C++</tt> exception throwing methods, but may
1233 use the <tt>unwind</tt> instruction.</dd>
Anton Korobeynikovc5ec8a72009-07-17 18:07:26 +0000[diff] [blame]1234
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1235 <dt><tt><b><a name="ssp">ssp</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1236 <dd>This attribute indicates that the function should emit a stack smashing
1237 protector. It is in the form of a "canary"&mdash;a random value placed on
1238 the stack before the local variables that's checked upon return from the
1239 function to see if it has been overwritten. A heuristic is used to
1240 determine if a function needs stack protectors or not.<br>
1241<br>
1242 If a function that has an <tt>ssp</tt> attribute is inlined into a
1243 function that doesn't have an <tt>ssp</tt> attribute, then the resulting
1244 function will have an <tt>ssp</tt> attribute.</dd>
1245
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1246 <dt><tt><b>sspreq</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1247 <dd>This attribute indicates that the function should <em>always</em> emit a
1248 stack smashing protector. This overrides
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1249 the <tt><a href="#ssp">ssp</a></tt> function attribute.<br>
1250<br>
1251 If a function that has an <tt>sspreq</tt> attribute is inlined into a
1252 function that doesn't have an <tt>sspreq</tt> attribute or which has
1253 an <tt>ssp</tt> attribute, then the resulting function will have
1254 an <tt>sspreq</tt> attribute.</dd>
Rafael Espindolafbff0ec2011-07-25 15:27:59 +0000[diff] [blame]1255
1256 <dt><tt><b><a name="uwtable">uwtable</a></b></tt></dt>
1257 <dd>This attribute indicates that the ABI being targeted requires that
1258 an unwind table entry be produce for this function even if we can
1259 show that no exceptions passes by it. This is normally the case for
1260 the ELF x86-64 abi, but it can be disabled for some compilation
1261 units.</dd>
1262
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1263</dl>
1264
Devang Patelf8b94812008-09-04 23:05:13 +0000[diff] [blame]1265</div>
1266
1267<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1268<h3>
Chris Lattner1eeeb0c2006-04-08 04:40:53 +0000[diff] [blame]1269 <a name="moduleasm">Module-Level Inline Assembly</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1270</h3>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]1271
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1272<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1273
1274<p>Modules may contain "module-level inline asm" blocks, which corresponds to
1275 the GCC "file scope inline asm" blocks. These blocks are internally
1276 concatenated by LLVM and treated as a single unit, but may be separated in
1277 the <tt>.ll</tt> file if desired. The syntax is very simple:</p>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]1278
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]1279<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1280module asm "inline asm code goes here"
1281module asm "more can go here"
1282</pre>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]1283
1284<p>The strings can contain any character by escaping non-printable characters.
1285 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]1286 for the number.</p>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]1287
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1288<p>The inline asm code is simply printed to the machine code .s file when
1289 assembly code is generated.</p>
1290
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]1291</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]1292
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1293<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1294<h3>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1295 <a name="datalayout">Data Layout</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1296</h3>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1297
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1298<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1299
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1300<p>A module may specify a target specific data layout string that specifies how
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1301 data is to be laid out in memory. The syntax for the data layout is
1302 simply:</p>
1303
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]1304<pre class="doc_code">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1305target datalayout = "<i>layout specification</i>"
1306</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1307
1308<p>The <i>layout specification</i> consists of a list of specifications
1309 separated by the minus sign character ('-'). Each specification starts with
1310 a letter and may include other information after the letter to define some
1311 aspect of the data layout. The specifications accepted are as follows:</p>
1312
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1313<dl>
1314 <dt><tt>E</tt></dt>
1315 <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]1316 bits with the most significance have the lowest address location.</dd>
1317
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1318 <dt><tt>e</tt></dt>
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]1319 <dd>Specifies that the target lays out data in little-endian form. That is,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1320 the bits with the least significance have the lowest address
1321 location.</dd>
1322
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1323 <dt><tt>p:<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]1324 <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]1325 <i>preferred</i> alignments. All sizes are in bits. Specifying
1326 the <i>pref</i> alignment is optional. If omitted, the
1327 preceding <tt>:</tt> should be omitted too.</dd>
1328
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1329 <dt><tt>i<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1330 <dd>This specifies the alignment for an integer type of a given bit
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1331 <i>size</i>. The value of <i>size</i> must be in the range [1,2^23).</dd>
1332
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1333 <dt><tt>v<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]1334 <dd>This specifies the alignment for a vector type of a given bit
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1335 <i>size</i>.</dd>
1336
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1337 <dt><tt>f<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]1338 <dd>This specifies the alignment for a floating point type of a given bit
Dale Johannesen9d8d2212010-05-28 18:54:47 +0000[diff] [blame]1339 <i>size</i>. Only values of <i>size</i> that are supported by the target
1340 will work. 32 (float) and 64 (double) are supported on all targets;
1341 80 or 128 (different flavors of long double) are also supported on some
1342 targets.
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1343
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1344 <dt><tt>a<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1345 <dd>This specifies the alignment for an aggregate type of a given bit
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1346 <i>size</i>.</dd>
1347
Daniel Dunbar87bde0b2009-06-08 22:17:53 +0000[diff] [blame]1348 <dt><tt>s<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1349 <dd>This specifies the alignment for a stack object of a given bit
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1350 <i>size</i>.</dd>
Chris Lattnere82bdc42009-11-07 09:35:34 +0000[diff] [blame]1351
1352 <dt><tt>n<i>size1</i>:<i>size2</i>:<i>size3</i>...</tt></dt>
1353 <dd>This specifies a set of native integer widths for the target CPU
1354 in bits. For example, it might contain "n32" for 32-bit PowerPC,
1355 "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]1356 this set are considered to support most general arithmetic
Chris Lattnere82bdc42009-11-07 09:35:34 +0000[diff] [blame]1357 operations efficiently.</dd>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1358</dl>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1359
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1360<p>When constructing the data layout for a given target, LLVM starts with a
Dan Gohman1c70c002010-04-28 00:36:01 +0000[diff] [blame]1361 default set of specifications which are then (possibly) overridden by the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1362 specifications in the <tt>datalayout</tt> keyword. The default specifications
1363 are given in this list:</p>
1364
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1365<ul>
1366 <li><tt>E</tt> - big endian</li>
Dan Gohmanfdf2e8c2010-02-23 02:44:03 +0000[diff] [blame]1367 <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]1368 <li><tt>i1:8:8</tt> - i1 is 8-bit (byte) aligned</li>
1369 <li><tt>i8:8:8</tt> - i8 is 8-bit (byte) aligned</li>
1370 <li><tt>i16:16:16</tt> - i16 is 16-bit aligned</li>
1371 <li><tt>i32:32:32</tt> - i32 is 32-bit aligned</li>
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]1372 <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]1373 alignment of 64-bits</li>
1374 <li><tt>f32:32:32</tt> - float is 32-bit aligned</li>
1375 <li><tt>f64:64:64</tt> - double is 64-bit aligned</li>
1376 <li><tt>v64:64:64</tt> - 64-bit vector is 64-bit aligned</li>
1377 <li><tt>v128:128:128</tt> - 128-bit vector is 128-bit aligned</li>
1378 <li><tt>a0:0:1</tt> - aggregates are 8-bit aligned</li>
Daniel Dunbar87bde0b2009-06-08 22:17:53 +0000[diff] [blame]1379 <li><tt>s0:64:64</tt> - stack objects are 64-bit aligned</li>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1380</ul>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1381
1382<p>When LLVM is determining the alignment for a given type, it uses the
1383 following rules:</p>
1384
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1385<ol>
1386 <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]1387 specification is used.</li>
1388
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1389 <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]1390 smallest integer type that is larger than the bitwidth of the sought type
1391 is used. If none of the specifications are larger than the bitwidth then
1392 the the largest integer type is used. For example, given the default
1393 specifications above, the i7 type will use the alignment of i8 (next
1394 largest) while both i65 and i256 will use the alignment of i64 (largest
1395 specified).</li>
1396
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1397 <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]1398 largest vector type that is smaller than the sought vector type will be
1399 used as a fall back. This happens because &lt;128 x double&gt; can be
1400 implemented in terms of 64 &lt;2 x double&gt;, for example.</li>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1401</ol>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1402
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1403</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]1404
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1405<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1406<h3>
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1407 <a name="pointeraliasing">Pointer Aliasing Rules</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1408</h3>
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1409
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1410<div>
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1411
Andreas Bolka55e459a2009-07-29 00:02:05 +0000[diff] [blame]1412<p>Any memory access must be done through a pointer value associated
Andreas Bolka99a82052009-07-27 20:37:10 +0000[diff] [blame]1413with an address range of the memory access, otherwise the behavior
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1414is undefined. Pointer values are associated with address ranges
1415according to the following rules:</p>
1416
1417<ul>
Dan Gohman1e109622010-07-02 18:41:32 +0000[diff] [blame]1418 <li>A pointer value is associated with the addresses associated with
1419 any value it is <i>based</i> on.
Andreas Bolka55e459a2009-07-29 00:02:05 +0000[diff] [blame]1420 <li>An address of a global variable is associated with the address
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1421 range of the variable's storage.</li>
1422 <li>The result value of an allocation instruction is associated with
1423 the address range of the allocated storage.</li>
1424 <li>A null pointer in the default address-space is associated with
Andreas Bolka55e459a2009-07-29 00:02:05 +0000[diff] [blame]1425 no address.</li>
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1426 <li>An integer constant other than zero or a pointer value returned
1427 from a function not defined within LLVM may be associated with address
1428 ranges allocated through mechanisms other than those provided by
Andreas Bolka55e459a2009-07-29 00:02:05 +0000[diff] [blame]1429 LLVM. Such ranges shall not overlap with any ranges of addresses
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1430 allocated by mechanisms provided by LLVM.</li>
Dan Gohman1e109622010-07-02 18:41:32 +0000[diff] [blame]1431</ul>
1432
1433<p>A pointer value is <i>based</i> on another pointer value according
1434 to the following rules:</p>
1435
1436<ul>
1437 <li>A pointer value formed from a
1438 <tt><a href="#i_getelementptr">getelementptr</a></tt> operation
1439 is <i>based</i> on the first operand of the <tt>getelementptr</tt>.</li>
1440 <li>The result value of a
1441 <tt><a href="#i_bitcast">bitcast</a></tt> is <i>based</i> on the operand
1442 of the <tt>bitcast</tt>.</li>
1443 <li>A pointer value formed by an
1444 <tt><a href="#i_inttoptr">inttoptr</a></tt> is <i>based</i> on all
1445 pointer values that contribute (directly or indirectly) to the
1446 computation of the pointer's value.</li>
1447 <li>The "<i>based</i> on" relationship is transitive.</li>
1448</ul>
1449
1450<p>Note that this definition of <i>"based"</i> is intentionally
1451 similar to the definition of <i>"based"</i> in C99, though it is
1452 slightly weaker.</p>
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1453
1454<p>LLVM IR does not associate types with memory. The result type of a
Andreas Bolka55e459a2009-07-29 00:02:05 +0000[diff] [blame]1455<tt><a href="#i_load">load</a></tt> merely indicates the size and
1456alignment of the memory from which to load, as well as the
Dan Gohmanc22c0f32010-06-17 19:23:50 +0000[diff] [blame]1457interpretation of the value. The first operand type of a
Andreas Bolka55e459a2009-07-29 00:02:05 +0000[diff] [blame]1458<tt><a href="#i_store">store</a></tt> similarly only indicates the size
1459and alignment of the store.</p>
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1460
1461<p>Consequently, type-based alias analysis, aka TBAA, aka
1462<tt>-fstrict-aliasing</tt>, is not applicable to general unadorned
1463LLVM IR. <a href="#metadata">Metadata</a> may be used to encode
1464additional information which specialized optimization passes may use
1465to implement type-based alias analysis.</p>
1466
1467</div>
1468
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]1469<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1470<h3>
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]1471 <a name="volatile">Volatile Memory Accesses</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1472</h3>
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]1473
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1474<div>
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]1475
1476<p>Certain memory accesses, such as <a href="#i_load"><tt>load</tt></a>s, <a
1477href="#i_store"><tt>store</tt></a>s, and <a
1478href="#int_memcpy"><tt>llvm.memcpy</tt></a>s may be marked <tt>volatile</tt>.
1479The optimizers must not change the number of volatile operations or change their
1480order of execution relative to other volatile operations. The optimizers
1481<i>may</i> change the order of volatile operations relative to non-volatile
1482operations. This is not Java's "volatile" and has no cross-thread
1483synchronization behavior.</p>
1484
1485</div>
1486
Eli Friedman5b60e1b2011-07-20 21:35:53 +0000[diff] [blame]1487<!-- ======================================================================= -->
1488<h3>
1489 <a name="memmodel">Memory Model for Concurrent Operations</a>
1490</h3>
1491
1492<div>
1493
1494<p>The LLVM IR does not define any way to start parallel threads of execution
1495or to register signal handlers. Nonetheless, there are platform-specific
1496ways to create them, and we define LLVM IR's behavior in their presence. This
1497model is inspired by the C++0x memory model.</p>
1498
1499<p>We define a <i>happens-before</i> partial order as the least partial order
1500that</p>
1501<ul>
1502 <li>Is a superset of single-thread program order, and</li>
1503 <li>When a <i>synchronizes-with</i> <tt>b</tt>, includes an edge from
1504 <tt>a</tt> to <tt>b</tt>. <i>Synchronizes-with</i> pairs are introduced
1505 by platform-specific techniques, like pthread locks, thread
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]1506 creation, thread joining, etc., and by atomic instructions.
1507 (See also <a href="#ordering">Atomic Memory Ordering Constraints</a>).
1508 </li>
Eli Friedman5b60e1b2011-07-20 21:35:53 +0000[diff] [blame]1509</ul>
1510
1511<p>Note that program order does not introduce <i>happens-before</i> edges
1512between a thread and signals executing inside that thread.</p>
1513
1514<p>Every (defined) read operation (load instructions, memcpy, atomic
1515loads/read-modify-writes, etc.) <var>R</var> reads a series of bytes written by
1516(defined) write operations (store instructions, atomic
Eli Friedman118973a2011-07-22 03:04:45 +0000[diff] [blame]1517stores/read-modify-writes, memcpy, etc.). For the purposes of this section,
1518initialized globals are considered to have a write of the initializer which is
1519atomic and happens before any other read or write of the memory in question.
1520For each byte of a read <var>R</var>, <var>R<sub>byte</sub></var> may see
1521any write to the same byte, except:</p>
Eli Friedman5b60e1b2011-07-20 21:35:53 +0000[diff] [blame]1522
1523<ul>
1524 <li>If <var>write<sub>1</sub></var> happens before
1525 <var>write<sub>2</sub></var>, and <var>write<sub>2</sub></var> happens
1526 before <var>R<sub>byte</sub></var>, then <var>R<sub>byte</sub></var>
Eli Friedman118973a2011-07-22 03:04:45 +0000[diff] [blame]1527 does not see <var>write<sub>1</sub></var>.
Eli Friedman5b60e1b2011-07-20 21:35:53 +0000[diff] [blame]1528 <li>If <var>R<sub>byte</sub></var> happens before <var>write<sub>3</var>,
Eli Friedman118973a2011-07-22 03:04:45 +0000[diff] [blame]1529 then <var>R<sub>byte</sub></var> does not see
Eli Friedman5b60e1b2011-07-20 21:35:53 +0000[diff] [blame]1530 <var>write<sub>3</sub></var>.
1531</ul>
1532
1533<p>Given that definition, <var>R<sub>byte</sub></var> is defined as follows:
1534<ul>
1535 <li>If there is no write to the same byte that happens before
1536 <var>R<sub>byte</sub></var>, <var>R<sub>byte</sub></var> returns
1537 <tt>undef</tt> for that byte.
Eli Friedman118973a2011-07-22 03:04:45 +0000[diff] [blame]1538 <li>Otherwise, if <var>R<sub>byte</sub></var> may see exactly one write,
Eli Friedman5b60e1b2011-07-20 21:35:53 +0000[diff] [blame]1539 <var>R<sub>byte</sub></var> returns the value written by that
1540 write.</li>
Eli Friedman118973a2011-07-22 03:04:45 +0000[diff] [blame]1541 <li>Otherwise, if <var>R</var> is atomic, and all the writes
1542 <var>R<sub>byte</sub></var> may see are atomic, it chooses one of the
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]1543 values written. See the <a href="#ordering">Atomic Memory Ordering
1544 Constraints</a> section for additional constraints on how the choice
1545 is made.
Eli Friedman5b60e1b2011-07-20 21:35:53 +0000[diff] [blame]1546 <li>Otherwise <var>R<sub>byte</sub></var> returns <tt>undef</tt>.</li>
1547</ul>
1548
1549<p><var>R</var> returns the value composed of the series of bytes it read.
1550This implies that some bytes within the value may be <tt>undef</tt>
1551<b>without</b> the entire value being <tt>undef</tt>. Note that this only
1552defines the semantics of the operation; it doesn't mean that targets will
1553emit more than one instruction to read the series of bytes.</p>
1554
1555<p>Note that in cases where none of the atomic intrinsics are used, this model
1556places only one restriction on IR transformations on top of what is required
1557for single-threaded execution: introducing a store to a byte which might not
Eli Friedman118973a2011-07-22 03:04:45 +0000[diff] [blame]1558otherwise be stored to can introduce undefined behavior. (Specifically, in
1559the case where another thread might write to and read from an address,
1560introducing a store can change a load that may see exactly one write into
1561a load that may see multiple writes.)</p>
Eli Friedman5b60e1b2011-07-20 21:35:53 +0000[diff] [blame]1562
1563<!-- FIXME: This model assumes all targets where concurrency is relevant have
1564a byte-size store which doesn't affect adjacent bytes. As far as I can tell,
1565none of the backends currently in the tree fall into this category; however,
1566there might be targets which care. If there are, we want a paragraph
1567like the following:
1568
1569Targets may specify that stores narrower than a certain width are not
1570available; on such a target, for the purposes of this model, treat any
1571non-atomic write with an alignment or width less than the minimum width
1572as if it writes to the relevant surrounding bytes.
1573-->
1574
1575</div>
1576
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]1577<!-- ======================================================================= -->
1578<div class="doc_subsection">
1579 <a name="ordering">Atomic Memory Ordering Constraints</a>
1580</div>
1581
1582<div class="doc_text">
1583
1584<p>Atomic instructions (<a href="#i_cmpxchg"><code>cmpxchg</code></a>,
1585<a href="#i_atomicrmw"><code>atomicrmw</code></a>, and
1586<a href="#i_fence"><code>fence</code></a>) take an ordering parameter
1587that determines which other atomic instructions on the same address they
1588<i>synchronize with</i>. These semantics are borrowed from Java and C++0x,
1589but are somewhat more colloquial. If these descriptions aren't precise enough,
1590check those specs. <a href="#i_fence"><code>fence</code></a> instructions
1591treat these orderings somewhat differently since they don't take an address.
1592See that instruction's documentation for details.</p>
1593
1594<!-- FIXME Note atomic load+store here once those get added. -->
1595
1596<dl>
1597<!-- FIXME: unordered is intended to be used for atomic load and store;
1598it isn't allowed for any instruction yet. -->
1599<dt><code>unordered</code></dt>
1600<dd>The set of values that can be read is governed by the happens-before
1601partial order. A value cannot be read unless some operation wrote it.
1602This is intended to provide a guarantee strong enough to model Java's
1603non-volatile shared variables. This ordering cannot be specified for
1604read-modify-write operations; it is not strong enough to make them atomic
1605in any interesting way.</dd>
1606<dt><code>monotonic</code></dt>
1607<dd>In addition to the guarantees of <code>unordered</code>, there is a single
1608total order for modifications by <code>monotonic</code> operations on each
1609address. All modification orders must be compatible with the happens-before
1610order. There is no guarantee that the modification orders can be combined to
1611a global total order for the whole program (and this often will not be
1612possible). The read in an atomic read-modify-write operation
1613(<a href="#i_cmpxchg"><code>cmpxchg</code></a> and
1614<a href="#i_atomicrmw"><code>atomicrmw</code></a>)
1615reads the value in the modification order immediately before the value it
1616writes. If one atomic read happens before another atomic read of the same
1617address, the later read must see the same value or a later value in the
1618address's modification order. This disallows reordering of
1619<code>monotonic</code> (or stronger) operations on the same address. If an
1620address is written <code>monotonic</code>ally by one thread, and other threads
1621<code>monotonic</code>ally read that address repeatedly, the other threads must
1622eventually see the write. This is intended to model C++'s relaxed atomic
1623variables.</dd>
1624<dt><code>acquire</code></dt>
1625<dd>In addition to the guarantees of <code>monotonic</code>, if this operation
1626reads a value written by a <code>release</code> atomic operation, it
1627<i>synchronizes-with</i> that operation.</dd>
1628<dt><code>release</code></dt>
1629<dd>In addition to the guarantees of <code>monotonic</code>,
1630a <i>synchronizes-with</i> edge may be formed by an <code>acquire</code>
1631operation.</dd>
1632<dt><code>acq_rel</code> (acquire+release)</dt><dd>Acts as both an
1633<code>acquire</code> and <code>release</code> operation on its address.</dd>
1634<dt><code>seq_cst</code> (sequentially consistent)</dt><dd>
1635<dd>In addition to the guarantees of <code>acq_rel</code>
1636(<code>acquire</code> for an operation which only reads, <code>release</code>
1637for an operation which only writes), there is a global total order on all
1638sequentially-consistent operations on all addresses, which is consistent with
1639the <i>happens-before</i> partial order and with the modification orders of
1640all the affected addresses. Each sequentially-consistent read sees the last
1641preceding write to the same address in this global order. This is intended
1642to model C++'s sequentially-consistent atomic variables and Java's volatile
1643shared variables.</dd>
1644</dl>
1645
1646<p id="singlethread">If an atomic operation is marked <code>singlethread</code>,
1647it only <i>synchronizes with</i> or participates in modification and seq_cst
1648total orderings with other operations running in the same thread (for example,
1649in signal handlers).</p>
1650
1651</div>
1652
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1653</div>
1654
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1655<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1656<h2><a name="typesystem">Type System</a></h2>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1657<!-- *********************************************************************** -->
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]1658
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1659<div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]1660
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1661<p>The LLVM type system is one of the most important features of the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1662 intermediate representation. Being typed enables a number of optimizations
1663 to be performed on the intermediate representation directly, without having
1664 to do extra analyses on the side before the transformation. A strong type
1665 system makes it easier to read the generated code and enables novel analyses
1666 and transformations that are not feasible to perform on normal three address
1667 code representations.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]1668
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1669<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1670<h3>
1671 <a name="t_classifications">Type Classifications</a>
1672</h3>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1673
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1674<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1675
1676<p>The types fall into a few useful classifications:</p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1677
1678<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1679 <tbody>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1680 <tr><th>Classification</th><th>Types</th></tr>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1681 <tr>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1682 <td><a href="#t_integer">integer</a></td>
Reid Spencer2b916312007-05-16 18:44:01 +0000[diff] [blame]1683 <td><tt>i1, i2, i3, ... i8, ... i16, ... i32, ... i64, ... </tt></td>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1684 </tr>
1685 <tr>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1686 <td><a href="#t_floating">floating point</a></td>
1687 <td><tt>float, double, x86_fp80, fp128, ppc_fp128</tt></td>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1688 </tr>
1689 <tr>
1690 <td><a name="t_firstclass">first class</a></td>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1691 <td><a href="#t_integer">integer</a>,
1692 <a href="#t_floating">floating point</a>,
1693 <a href="#t_pointer">pointer</a>,
Dan Gohman0066db62008-06-18 18:42:13 +0000[diff] [blame]1694 <a href="#t_vector">vector</a>,
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]1695 <a href="#t_struct">structure</a>,
1696 <a href="#t_array">array</a>,
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1697 <a href="#t_label">label</a>,
1698 <a href="#t_metadata">metadata</a>.
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1699 </td>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1700 </tr>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1701 <tr>
1702 <td><a href="#t_primitive">primitive</a></td>
1703 <td><a href="#t_label">label</a>,
1704 <a href="#t_void">void</a>,
Tobias Grosser05387292010-12-28 20:29:31 +0000[diff] [blame]1705 <a href="#t_integer">integer</a>,
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1706 <a href="#t_floating">floating point</a>,
Dale Johannesen21fe99b2010-10-01 00:48:59 +0000[diff] [blame]1707 <a href="#t_x86mmx">x86mmx</a>,
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1708 <a href="#t_metadata">metadata</a>.</td>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1709 </tr>
1710 <tr>
1711 <td><a href="#t_derived">derived</a></td>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]1712 <td><a href="#t_array">array</a>,
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1713 <a href="#t_function">function</a>,
1714 <a href="#t_pointer">pointer</a>,
1715 <a href="#t_struct">structure</a>,
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1716 <a href="#t_vector">vector</a>,
1717 <a href="#t_opaque">opaque</a>.
Dan Gohman01ac1012008-10-14 16:32:04 +0000[diff] [blame]1718 </td>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1719 </tr>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1720 </tbody>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1721</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1722
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1723<p>The <a href="#t_firstclass">first class</a> types are perhaps the most
1724 important. Values of these types are the only ones which can be produced by
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]1725 instructions.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1726
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1727</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1728
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1729<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1730<h3>
1731 <a name="t_primitive">Primitive Types</a>
1732</h3>
Chris Lattner8f8c7b72008-01-04 04:34:14 +0000[diff] [blame]1733
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1734<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1735
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1736<p>The primitive types are the fundamental building blocks of the LLVM
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1737 system.</p>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1738
1739<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1740<h4>
1741 <a name="t_integer">Integer Type</a>
1742</h4>
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]1743
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1744<div>
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]1745
1746<h5>Overview:</h5>
1747<p>The integer type is a very simple type that simply specifies an arbitrary
1748 bit width for the integer type desired. Any bit width from 1 bit to
1749 2<sup>23</sup>-1 (about 8 million) can be specified.</p>
1750
1751<h5>Syntax:</h5>
1752<pre>
1753 iN
1754</pre>
1755
1756<p>The number of bits the integer will occupy is specified by the <tt>N</tt>
1757 value.</p>
1758
1759<h5>Examples:</h5>
1760<table class="layout">
1761 <tr class="layout">
1762 <td class="left"><tt>i1</tt></td>
1763 <td class="left">a single-bit integer.</td>
1764 </tr>
1765 <tr class="layout">
1766 <td class="left"><tt>i32</tt></td>
1767 <td class="left">a 32-bit integer.</td>
1768 </tr>
1769 <tr class="layout">
1770 <td class="left"><tt>i1942652</tt></td>
1771 <td class="left">a really big integer of over 1 million bits.</td>
1772 </tr>
1773</table>
1774
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]1775</div>
1776
1777<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1778<h4>
1779 <a name="t_floating">Floating Point Types</a>
1780</h4>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1781
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1782<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1783
1784<table>
1785 <tbody>
1786 <tr><th>Type</th><th>Description</th></tr>
1787 <tr><td><tt>float</tt></td><td>32-bit floating point value</td></tr>
1788 <tr><td><tt>double</tt></td><td>64-bit floating point value</td></tr>
1789 <tr><td><tt>fp128</tt></td><td>128-bit floating point value (112-bit mantissa)</td></tr>
1790 <tr><td><tt>x86_fp80</tt></td><td>80-bit floating point value (X87)</td></tr>
1791 <tr><td><tt>ppc_fp128</tt></td><td>128-bit floating point value (two 64-bits)</td></tr>
1792 </tbody>
1793</table>
1794
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1795</div>
1796
1797<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1798<h4>
1799 <a name="t_x86mmx">X86mmx Type</a>
1800</h4>
Dale Johannesen21fe99b2010-10-01 00:48:59 +0000[diff] [blame]1801
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1802<div>
Dale Johannesen21fe99b2010-10-01 00:48:59 +0000[diff] [blame]1803
1804<h5>Overview:</h5>
1805<p>The x86mmx type represents a value held in an MMX register on an x86 machine. The operations allowed on it are quite limited: parameters and return values, load and store, and bitcast. User-specified MMX instructions are represented as intrinsic or asm calls with arguments and/or results of this type. There are no arrays, vectors or constants of this type.</p>
1806
1807<h5>Syntax:</h5>
1808<pre>
Dale Johannesen473a8c82010-10-01 01:07:02 +0000[diff] [blame]1809 x86mmx
Dale Johannesen21fe99b2010-10-01 00:48:59 +0000[diff] [blame]1810</pre>
1811
1812</div>
1813
1814<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1815<h4>
1816 <a name="t_void">Void Type</a>
1817</h4>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1818
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1819<div>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1820
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1821<h5>Overview:</h5>
1822<p>The void type does not represent any value and has no size.</p>
1823
1824<h5>Syntax:</h5>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1825<pre>
1826 void
1827</pre>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1828
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1829</div>
1830
1831<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1832<h4>
1833 <a name="t_label">Label Type</a>
1834</h4>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1835
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1836<div>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1837
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1838<h5>Overview:</h5>
1839<p>The label type represents code labels.</p>
1840
1841<h5>Syntax:</h5>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1842<pre>
1843 label
1844</pre>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1845
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1846</div>
1847
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1848<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1849<h4>
1850 <a name="t_metadata">Metadata Type</a>
1851</h4>
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1852
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1853<div>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1854
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1855<h5>Overview:</h5>
Nick Lewyckyc261df92009-09-27 23:27:42 +0000[diff] [blame]1856<p>The metadata type represents embedded metadata. No derived types may be
1857 created from metadata except for <a href="#t_function">function</a>
1858 arguments.
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1859
1860<h5>Syntax:</h5>
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1861<pre>
1862 metadata
1863</pre>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1864
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1865</div>
1866
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1867</div>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1868
1869<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1870<h3>
1871 <a name="t_derived">Derived Types</a>
1872</h3>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1873
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1874<div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1875
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1876<p>The real power in LLVM comes from the derived types in the system. This is
1877 what allows a programmer to represent arrays, functions, pointers, and other
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]1878 useful types. Each of these types contain one or more element types which
1879 may be a primitive type, or another derived type. For example, it is
1880 possible to have a two dimensional array, using an array as the element type
1881 of another array.</p>
Dan Gohmand8791e52009-01-24 15:58:40 +0000[diff] [blame]1882
Chris Lattner1afcace2011-07-09 17:41:24 +0000[diff] [blame]1883</div>
1884
1885
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]1886<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1887<h4>
1888 <a name="t_aggregate">Aggregate Types</a>
1889</h4>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]1890
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1891<div>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]1892
1893<p>Aggregate Types are a subset of derived types that can contain multiple
1894 member types. <a href="#t_array">Arrays</a>,
Chris Lattner61c70e92010-08-28 04:09:24 +0000[diff] [blame]1895 <a href="#t_struct">structs</a>, and <a href="#t_vector">vectors</a> are
1896 aggregate types.</p>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]1897
1898</div>
1899
Reid Spencer2b916312007-05-16 18:44:01 +0000[diff] [blame]1900<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1901<h4>
1902 <a name="t_array">Array Type</a>
1903</h4>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1904
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1905<div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1906
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1907<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1908<p>The array type is a very simple derived type that arranges elements
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1909 sequentially in memory. The array type requires a size (number of elements)
1910 and an underlying data type.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1911
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1912<h5>Syntax:</h5>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1913<pre>
1914 [&lt;# elements&gt; x &lt;elementtype&gt;]
1915</pre>
1916
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1917<p>The number of elements is a constant integer value; <tt>elementtype</tt> may
1918 be any type with a size.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1919
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1920<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +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>[40 x i32]</tt></td>
1924 <td class="left">Array of 40 32-bit integer values.</td>
1925 </tr>
1926 <tr class="layout">
1927 <td class="left"><tt>[41 x i32]</tt></td>
1928 <td class="left">Array of 41 32-bit integer values.</td>
1929 </tr>
1930 <tr class="layout">
1931 <td class="left"><tt>[4 x i8]</tt></td>
1932 <td class="left">Array of 4 8-bit integer values.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1933 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1934</table>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1935<p>Here are some examples of multidimensional arrays:</p>
1936<table class="layout">
1937 <tr class="layout">
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]1938 <td class="left"><tt>[3 x [4 x i32]]</tt></td>
1939 <td class="left">3x4 array of 32-bit integer values.</td>
1940 </tr>
1941 <tr class="layout">
1942 <td class="left"><tt>[12 x [10 x float]]</tt></td>
1943 <td class="left">12x10 array of single precision floating point values.</td>
1944 </tr>
1945 <tr class="layout">
1946 <td class="left"><tt>[2 x [3 x [4 x i16]]]</tt></td>
1947 <td class="left">2x3x4 array of 16-bit integer values.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1948 </tr>
1949</table>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]1950
Dan Gohman7657f6b2009-11-09 19:01:53 +0000[diff] [blame]1951<p>There is no restriction on indexing beyond the end of the array implied by
1952 a static type (though there are restrictions on indexing beyond the bounds
1953 of an allocated object in some cases). This means that single-dimension
1954 'variable sized array' addressing can be implemented in LLVM with a zero
1955 length array type. An implementation of 'pascal style arrays' in LLVM could
1956 use the type "<tt>{ i32, [0 x float]}</tt>", for example.</p>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]1957
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1958</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1959
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1960<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1961<h4>
1962 <a name="t_function">Function Type</a>
1963</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1964
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1965<div>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]1966
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1967<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1968<p>The function type can be thought of as a function signature. It consists of
1969 a return type and a list of formal parameter types. The return type of a
Chris Lattner61c70e92010-08-28 04:09:24 +0000[diff] [blame]1970 function type is a first class type or a void type.</p>
Devang Patelc3fc6df2008-03-10 20:49:15 +0000[diff] [blame]1971
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1972<h5>Syntax:</h5>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]1973<pre>
Nick Lewycky51386942009-09-27 07:55:32 +0000[diff] [blame]1974 &lt;returntype&gt; (&lt;parameter list&gt;)
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]1975</pre>
1976
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]1977<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]1978 specifiers. Optionally, the parameter list may include a type <tt>...</tt>,
1979 which indicates that the function takes a variable number of arguments.
1980 Variable argument functions can access their arguments with
1981 the <a href="#int_varargs">variable argument handling intrinsic</a>
Chris Lattner0724fbd2010-03-02 06:36:51 +0000[diff] [blame]1982 functions. '<tt>&lt;returntype&gt;</tt>' is any type except
Nick Lewyckyc261df92009-09-27 23:27:42 +0000[diff] [blame]1983 <a href="#t_label">label</a>.</p>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]1984
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1985<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1986<table class="layout">
1987 <tr class="layout">
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]1988 <td class="left"><tt>i32 (i32)</tt></td>
1989 <td class="left">function taking an <tt>i32</tt>, returning an <tt>i32</tt>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1990 </td>
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]1991 </tr><tr class="layout">
Chris Lattner0724fbd2010-03-02 06:36:51 +0000[diff] [blame]1992 <td class="left"><tt>float&nbsp;(i16,&nbsp;i32&nbsp;*)&nbsp;*
Reid Spencerf17a0b72006-12-31 07:20:23 +0000[diff] [blame]1993 </tt></td>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]1994 <td class="left"><a href="#t_pointer">Pointer</a> to a function that takes
Chris Lattner0724fbd2010-03-02 06:36:51 +0000[diff] [blame]1995 an <tt>i16</tt> and a <a href="#t_pointer">pointer</a> to <tt>i32</tt>,
1996 returning <tt>float</tt>.
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]1997 </td>
1998 </tr><tr class="layout">
1999 <td class="left"><tt>i32 (i8*, ...)</tt></td>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]2000 <td class="left">A vararg function that takes at least one
2001 <a href="#t_pointer">pointer</a> to <tt>i8 </tt> (char in C),
2002 which returns an integer. This is the signature for <tt>printf</tt> in
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]2003 LLVM.
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]2004 </td>
Devang Patela582f402008-03-24 05:35:41 +0000[diff] [blame]2005 </tr><tr class="layout">
2006 <td class="left"><tt>{i32, i32} (i32)</tt></td>
Nick Lewycky51386942009-09-27 07:55:32 +0000[diff] [blame]2007 <td class="left">A function taking an <tt>i32</tt>, returning a
2008 <a href="#t_struct">structure</a> containing two <tt>i32</tt> values
Devang Patela582f402008-03-24 05:35:41 +0000[diff] [blame]2009 </td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]2010 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2011</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]2012
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2013</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2014
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2015<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2016<h4>
2017 <a name="t_struct">Structure Type</a>
2018</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2019
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2020<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2021
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2022<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2023<p>The structure type is used to represent a collection of data members together
Chris Lattner1afcace2011-07-09 17:41:24 +0000[diff] [blame]2024 in memory. The elements of a structure may be any type that has a size.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2025
Jeffrey Yasskin7a088cf2010-01-11 19:19:26 +0000[diff] [blame]2026<p>Structures in memory are accessed using '<tt><a href="#i_load">load</a></tt>'
2027 and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a field
2028 with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.
2029 Structures in registers are accessed using the
2030 '<tt><a href="#i_extractvalue">extractvalue</a></tt>' and
2031 '<tt><a href="#i_insertvalue">insertvalue</a></tt>' instructions.</p>
Chris Lattner1afcace2011-07-09 17:41:24 +0000[diff] [blame]2032
2033<p>Structures may optionally be "packed" structures, which indicate that the
2034 alignment of the struct is one byte, and that there is no padding between
2035 the elements. In non-packed structs, padding between field types is defined
2036 by the target data string to match the underlying processor.</p>
2037
2038<p>Structures can either be "anonymous" or "named". An anonymous structure is
2039 defined inline with other types (e.g. <tt>{i32, i32}*</tt>) and a named types
2040 are always defined at the top level with a name. Anonmyous types are uniqued
2041 by their contents and can never be recursive since there is no way to write
2042 one. Named types can be recursive.
2043</p>
2044
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2045<h5>Syntax:</h5>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]2046<pre>
Chris Lattner1afcace2011-07-09 17:41:24 +0000[diff] [blame]2047 %T1 = type { &lt;type list&gt; } <i>; Named normal struct type</i>
2048 %T2 = type &lt;{ &lt;type list&gt; }&gt; <i>; Named packed struct type</i>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]2049</pre>
Chris Lattner1afcace2011-07-09 17:41:24 +0000[diff] [blame]2050
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2051<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]2052<table class="layout">
2053 <tr class="layout">
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]2054 <td class="left"><tt>{ i32, i32, i32 }</tt></td>
2055 <td class="left">A triple of three <tt>i32</tt> values</td>
Chris Lattner1afcace2011-07-09 17:41:24 +0000[diff] [blame]2056 </tr>
2057 <tr class="layout">
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]2058 <td class="left"><tt>{&nbsp;float,&nbsp;i32&nbsp;(i32)&nbsp;*&nbsp;}</tt></td>
2059 <td class="left">A pair, where the first element is a <tt>float</tt> and the
2060 second element is a <a href="#t_pointer">pointer</a> to a
2061 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
2062 an <tt>i32</tt>.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]2063 </tr>
Chris Lattner1afcace2011-07-09 17:41:24 +0000[diff] [blame]2064 <tr class="layout">
2065 <td class="left"><tt>&lt;{ i8, i32 }&gt;</tt></td>
2066 <td class="left">A packed struct known to be 5 bytes in size.</td>
2067 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2068</table>
Dan Gohmand8791e52009-01-24 15:58:40 +0000[diff] [blame]2069
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2070</div>
Chris Lattner1afcace2011-07-09 17:41:24 +0000[diff] [blame]2071
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2072<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2073<h4>
Chris Lattner628ed392011-07-23 19:59:08 +0000[diff] [blame]2074 <a name="t_opaque">Opaque Structure Types</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2075</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2076
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2077<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2078
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]2079<h5>Overview:</h5>
Chris Lattner628ed392011-07-23 19:59:08 +0000[diff] [blame]2080<p>Opaque structure types are used to represent named structure types that do
2081 not have a body specified. This corresponds (for example) to the C notion of
2082 a forward declared structure.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2083
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]2084<h5>Syntax:</h5>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]2085<pre>
Chris Lattner1afcace2011-07-09 17:41:24 +0000[diff] [blame]2086 %X = type opaque
2087 %52 = type opaque
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]2088</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2089
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]2090<h5>Examples:</h5>
2091<table class="layout">
2092 <tr class="layout">
Chris Lattner1afcace2011-07-09 17:41:24 +0000[diff] [blame]2093 <td class="left"><tt>opaque</tt></td>
2094 <td class="left">An opaque type.</td>
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]2095 </tr>
2096</table>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2097
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]2098</div>
2099
Chris Lattner1afcace2011-07-09 17:41:24 +0000[diff] [blame]2100
2101
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]2102<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2103<h4>
2104 <a name="t_pointer">Pointer Type</a>
2105</h4>
Chris Lattner0fd4a272009-02-08 19:53:29 +0000[diff] [blame]2106
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2107<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2108
2109<h5>Overview:</h5>
Dan Gohmanff3ef322010-02-25 16:50:07 +0000[diff] [blame]2110<p>The pointer type is used to specify memory locations.
2111 Pointers are commonly used to reference objects in memory.</p>
2112
2113<p>Pointer types may have an optional address space attribute defining the
2114 numbered address space where the pointed-to object resides. The default
2115 address space is number zero. The semantics of non-zero address
2116 spaces are target-specific.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2117
2118<p>Note that LLVM does not permit pointers to void (<tt>void*</tt>) nor does it
2119 permit pointers to labels (<tt>label*</tt>). Use <tt>i8*</tt> instead.</p>
Chris Lattner0fd4a272009-02-08 19:53:29 +0000[diff] [blame]2120
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2121<h5>Syntax:</h5>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]2122<pre>
2123 &lt;type&gt; *
2124</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2125
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2126<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]2127<table class="layout">
2128 <tr class="layout">
Dan Gohman2a08c532009-01-04 23:44:43 +0000[diff] [blame]2129 <td class="left"><tt>[4 x i32]*</tt></td>
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]2130 <td class="left">A <a href="#t_pointer">pointer</a> to <a
2131 href="#t_array">array</a> of four <tt>i32</tt> values.</td>
2132 </tr>
2133 <tr class="layout">
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]2134 <td class="left"><tt>i32 (i32*) *</tt></td>
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]2135 <td class="left"> A <a href="#t_pointer">pointer</a> to a <a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2136 href="#t_function">function</a> that takes an <tt>i32*</tt>, returning an
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]2137 <tt>i32</tt>.</td>
2138 </tr>
2139 <tr class="layout">
2140 <td class="left"><tt>i32 addrspace(5)*</tt></td>
2141 <td class="left">A <a href="#t_pointer">pointer</a> to an <tt>i32</tt> value
2142 that resides in address space #5.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]2143 </tr>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2144</table>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2145
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2146</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]2147
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]2148<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2149<h4>
2150 <a name="t_vector">Vector Type</a>
2151</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2152
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2153<div>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]2154
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]2155<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2156<p>A vector type is a simple derived type that represents a vector of elements.
2157 Vector types are used when multiple primitive data are operated in parallel
2158 using a single instruction (SIMD). A vector type requires a size (number of
Duncan Sandsd40d14e2009-11-27 13:38:03 +0000[diff] [blame]2159 elements) and an underlying primitive data type. Vector types are considered
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2160 <a href="#t_firstclass">first class</a>.</p>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]2161
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]2162<h5>Syntax:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]2163<pre>
2164 &lt; &lt;# elements&gt; x &lt;elementtype&gt; &gt;
2165</pre>
2166
Chris Lattner7d2e7be2010-10-10 18:20:35 +0000[diff] [blame]2167<p>The number of elements is a constant integer value larger than 0; elementtype
2168 may be any integer or floating point type. Vectors of size zero are not
2169 allowed, and pointers are not allowed as the element type.</p>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]2170
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]2171<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]2172<table class="layout">
2173 <tr class="layout">
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]2174 <td class="left"><tt>&lt;4 x i32&gt;</tt></td>
2175 <td class="left">Vector of 4 32-bit integer values.</td>
2176 </tr>
2177 <tr class="layout">
2178 <td class="left"><tt>&lt;8 x float&gt;</tt></td>
2179 <td class="left">Vector of 8 32-bit floating-point values.</td>
2180 </tr>
2181 <tr class="layout">
2182 <td class="left"><tt>&lt;2 x i64&gt;</tt></td>
2183 <td class="left">Vector of 2 64-bit integer values.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]2184 </tr>
2185</table>
Dan Gohmand8791e52009-01-24 15:58:40 +0000[diff] [blame]2186
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2187</div>
2188
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2189<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2190<h2><a name="constants">Constants</a></h2>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2191<!-- *********************************************************************** -->
2192
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2193<div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2194
2195<p>LLVM has several different basic types of constants. This section describes
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2196 them all and their syntax.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2197
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2198<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2199<h3>
2200 <a name="simpleconstants">Simple Constants</a>
2201</h3>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2202
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2203<div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2204
2205<dl>
2206 <dt><b>Boolean constants</b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2207 <dd>The two strings '<tt>true</tt>' and '<tt>false</tt>' are both valid
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]2208 constants of the <tt><a href="#t_integer">i1</a></tt> type.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2209
2210 <dt><b>Integer constants</b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2211 <dd>Standard integers (such as '4') are constants of
2212 the <a href="#t_integer">integer</a> type. Negative numbers may be used
2213 with integer types.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2214
2215 <dt><b>Floating point constants</b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2216 <dd>Floating point constants use standard decimal notation (e.g. 123.421),
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2217 exponential notation (e.g. 1.23421e+2), or a more precise hexadecimal
2218 notation (see below). The assembler requires the exact decimal value of a
2219 floating-point constant. For example, the assembler accepts 1.25 but
2220 rejects 1.3 because 1.3 is a repeating decimal in binary. Floating point
2221 constants must have a <a href="#t_floating">floating point</a> type. </dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2222
2223 <dt><b>Null pointer constants</b></dt>
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]2224 <dd>The identifier '<tt>null</tt>' is recognized as a null pointer constant
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2225 and must be of <a href="#t_pointer">pointer type</a>.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2226</dl>
2227
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2228<p>The one non-intuitive notation for constants is the hexadecimal form of
2229 floating point constants. For example, the form '<tt>double
2230 0x432ff973cafa8000</tt>' is equivalent to (but harder to read than)
2231 '<tt>double 4.5e+15</tt>'. The only time hexadecimal floating point
2232 constants are required (and the only time that they are generated by the
2233 disassembler) is when a floating point constant must be emitted but it cannot
2234 be represented as a decimal floating point number in a reasonable number of
2235 digits. For example, NaN's, infinities, and other special values are
2236 represented in their IEEE hexadecimal format so that assembly and disassembly
2237 do not cause any bits to change in the constants.</p>
2238
Dale Johannesenbd5e5a82009-02-11 22:14:51 +0000[diff] [blame]2239<p>When using the hexadecimal form, constants of types float and double are
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2240 represented using the 16-digit form shown above (which matches the IEEE754
2241 representation for double); float values must, however, be exactly
2242 representable as IEE754 single precision. Hexadecimal format is always used
2243 for long double, and there are three forms of long double. The 80-bit format
2244 used by x86 is represented as <tt>0xK</tt> followed by 20 hexadecimal digits.
2245 The 128-bit format used by PowerPC (two adjacent doubles) is represented
2246 by <tt>0xM</tt> followed by 32 hexadecimal digits. The IEEE 128-bit format
2247 is represented by <tt>0xL</tt> followed by 32 hexadecimal digits; no
2248 currently supported target uses this format. Long doubles will only work if
2249 they match the long double format on your target. All hexadecimal formats
2250 are big-endian (sign bit at the left).</p>
2251
Dale Johannesen21fe99b2010-10-01 00:48:59 +0000[diff] [blame]2252<p>There are no constants of type x86mmx.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2253</div>
2254
2255<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2256<h3>
Bill Wendlingd9fe2982009-07-20 02:32:41 +0000[diff] [blame]2257<a name="aggregateconstants"></a> <!-- old anchor -->
2258<a name="complexconstants">Complex Constants</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2259</h3>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2260
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2261<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2262
Chris Lattner70882792009-02-28 18:32:25 +0000[diff] [blame]2263<p>Complex constants are a (potentially recursive) combination of simple
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2264 constants and smaller complex constants.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2265
2266<dl>
2267 <dt><b>Structure constants</b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2268 <dd>Structure constants are represented with notation similar to structure
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2269 type definitions (a comma separated list of elements, surrounded by braces
2270 (<tt>{}</tt>)). For example: "<tt>{ i32 4, float 17.0, i32* @G }</tt>",
2271 where "<tt>@G</tt>" is declared as "<tt>@G = external global i32</tt>".
2272 Structure constants must have <a href="#t_struct">structure type</a>, and
2273 the number and types of elements must match those specified by the
2274 type.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2275
2276 <dt><b>Array constants</b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2277 <dd>Array constants are represented with notation similar to array type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2278 definitions (a comma separated list of elements, surrounded by square
2279 brackets (<tt>[]</tt>)). For example: "<tt>[ i32 42, i32 11, i32 74
2280 ]</tt>". Array constants must have <a href="#t_array">array type</a>, and
2281 the number and types of elements must match those specified by the
2282 type.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2283
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]2284 <dt><b>Vector constants</b></dt>
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]2285 <dd>Vector constants are represented with notation similar to vector type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2286 definitions (a comma separated list of elements, surrounded by
2287 less-than/greater-than's (<tt>&lt;&gt;</tt>)). For example: "<tt>&lt; i32
2288 42, i32 11, i32 74, i32 100 &gt;</tt>". Vector constants must
2289 have <a href="#t_vector">vector type</a>, and the number and types of
2290 elements must match those specified by the type.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2291
2292 <dt><b>Zero initialization</b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2293 <dd>The string '<tt>zeroinitializer</tt>' can be used to zero initialize a
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]2294 value to zero of <em>any</em> type, including scalar and
2295 <a href="#t_aggregate">aggregate</a> types.
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2296 This is often used to avoid having to print large zero initializers
2297 (e.g. for large arrays) and is always exactly equivalent to using explicit
2298 zero initializers.</dd>
Nick Lewycky21cc4462009-04-04 07:22:01 +0000[diff] [blame]2299
2300 <dt><b>Metadata node</b></dt>
Nick Lewycky1e8c7a62009-05-30 16:08:30 +0000[diff] [blame]2301 <dd>A metadata node is a structure-like constant with
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2302 <a href="#t_metadata">metadata type</a>. For example: "<tt>metadata !{
2303 i32 0, metadata !"test" }</tt>". Unlike other constants that are meant to
2304 be interpreted as part of the instruction stream, metadata is a place to
2305 attach additional information such as debug info.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2306</dl>
2307
2308</div>
2309
2310<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2311<h3>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2312 <a name="globalconstants">Global Variable and Function Addresses</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2313</h3>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2314
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2315<div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2316
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2317<p>The addresses of <a href="#globalvars">global variables</a>
2318 and <a href="#functionstructure">functions</a> are always implicitly valid
2319 (link-time) constants. These constants are explicitly referenced when
2320 the <a href="#identifiers">identifier for the global</a> is used and always
2321 have <a href="#t_pointer">pointer</a> type. For example, the following is a
2322 legal LLVM file:</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2323
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2324<pre class="doc_code">
Chris Lattnera18a4242007-06-06 18:28:13 +0000[diff] [blame]2325@X = global i32 17
2326@Y = global i32 42
2327@Z = global [2 x i32*] [ i32* @X, i32* @Y ]
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2328</pre>
2329
2330</div>
2331
2332<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2333<h3>
2334 <a name="undefvalues">Undefined Values</a>
2335</h3>
2336
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2337<div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2338
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2339<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]2340 indicates that the user of the value may receive an unspecified bit-pattern.
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]2341 Undefined values may be of any type (other than '<tt>label</tt>'
2342 or '<tt>void</tt>') and be used anywhere a constant is permitted.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2343
Chris Lattnerc608cb12009-09-11 01:49:31 +0000[diff] [blame]2344<p>Undefined values are useful because they indicate to the compiler that the
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2345 program is well defined no matter what value is used. This gives the
2346 compiler more freedom to optimize. Here are some examples of (potentially
2347 surprising) transformations that are valid (in pseudo IR):</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2348
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2349
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2350<pre class="doc_code">
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2351 %A = add %X, undef
2352 %B = sub %X, undef
2353 %C = xor %X, undef
2354Safe:
2355 %A = undef
2356 %B = undef
2357 %C = undef
2358</pre>
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2359
2360<p>This is safe because all of the output bits are affected by the undef bits.
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]2361 Any output bit can have a zero or one depending on the input bits.</p>
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2362
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2363<pre class="doc_code">
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2364 %A = or %X, undef
2365 %B = and %X, undef
2366Safe:
2367 %A = -1
2368 %B = 0
2369Unsafe:
2370 %A = undef
2371 %B = undef
2372</pre>
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2373
2374<p>These logical operations have bits that are not always affected by the input.
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]2375 For example, if <tt>%X</tt> has a zero bit, then the output of the
2376 '<tt>and</tt>' operation will always be a zero for that bit, no matter what
2377 the corresponding bit from the '<tt>undef</tt>' is. As such, it is unsafe to
2378 optimize or assume that the result of the '<tt>and</tt>' is '<tt>undef</tt>'.
2379 However, it is safe to assume that all bits of the '<tt>undef</tt>' could be
2380 0, and optimize the '<tt>and</tt>' to 0. Likewise, it is safe to assume that
2381 all the bits of the '<tt>undef</tt>' operand to the '<tt>or</tt>' could be
2382 set, allowing the '<tt>or</tt>' to be folded to -1.</p>
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2383
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2384<pre class="doc_code">
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2385 %A = select undef, %X, %Y
2386 %B = select undef, 42, %Y
2387 %C = select %X, %Y, undef
2388Safe:
2389 %A = %X (or %Y)
2390 %B = 42 (or %Y)
2391 %C = %Y
2392Unsafe:
2393 %A = undef
2394 %B = undef
2395 %C = undef
2396</pre>
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2397
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]2398<p>This set of examples shows that undefined '<tt>select</tt>' (and conditional
2399 branch) conditions can go <em>either way</em>, but they have to come from one
2400 of the two operands. In the <tt>%A</tt> example, if <tt>%X</tt> and
2401 <tt>%Y</tt> were both known to have a clear low bit, then <tt>%A</tt> would
2402 have to have a cleared low bit. However, in the <tt>%C</tt> example, the
2403 optimizer is allowed to assume that the '<tt>undef</tt>' operand could be the
2404 same as <tt>%Y</tt>, allowing the whole '<tt>select</tt>' to be
2405 eliminated.</p>
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2406
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2407<pre class="doc_code">
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2408 %A = xor undef, undef
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]2409
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2410 %B = undef
2411 %C = xor %B, %B
2412
2413 %D = undef
2414 %E = icmp lt %D, 4
2415 %F = icmp gte %D, 4
2416
2417Safe:
2418 %A = undef
2419 %B = undef
2420 %C = undef
2421 %D = undef
2422 %E = undef
2423 %F = undef
2424</pre>
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2425
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]2426<p>This example points out that two '<tt>undef</tt>' operands are not
2427 necessarily the same. This can be surprising to people (and also matches C
2428 semantics) where they assume that "<tt>X^X</tt>" is always zero, even
2429 if <tt>X</tt> is undefined. This isn't true for a number of reasons, but the
2430 short answer is that an '<tt>undef</tt>' "variable" can arbitrarily change
2431 its value over its "live range". This is true because the variable doesn't
2432 actually <em>have a live range</em>. Instead, the value is logically read
2433 from arbitrary registers that happen to be around when needed, so the value
2434 is not necessarily consistent over time. In fact, <tt>%A</tt> and <tt>%C</tt>
2435 need to have the same semantics or the core LLVM "replace all uses with"
2436 concept would not hold.</p>
Chris Lattner6e9057b2009-09-07 23:33:52 +0000[diff] [blame]2437
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2438<pre class="doc_code">
Chris Lattner6e9057b2009-09-07 23:33:52 +0000[diff] [blame]2439 %A = fdiv undef, %X
2440 %B = fdiv %X, undef
2441Safe:
2442 %A = undef
2443b: unreachable
2444</pre>
Chris Lattner6e9057b2009-09-07 23:33:52 +0000[diff] [blame]2445
2446<p>These examples show the crucial difference between an <em>undefined
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]2447 value</em> and <em>undefined behavior</em>. An undefined value (like
2448 '<tt>undef</tt>') is allowed to have an arbitrary bit-pattern. This means that
2449 the <tt>%A</tt> operation can be constant folded to '<tt>undef</tt>', because
2450 the '<tt>undef</tt>' could be an SNaN, and <tt>fdiv</tt> is not (currently)
2451 defined on SNaN's. However, in the second example, we can make a more
2452 aggressive assumption: because the <tt>undef</tt> is allowed to be an
2453 arbitrary value, we are allowed to assume that it could be zero. Since a
2454 divide by zero has <em>undefined behavior</em>, we are allowed to assume that
2455 the operation does not execute at all. This allows us to delete the divide and
2456 all code after it. Because the undefined operation "can't happen", the
2457 optimizer can assume that it occurs in dead code.</p>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]2458
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2459<pre class="doc_code">
Chris Lattner6e9057b2009-09-07 23:33:52 +0000[diff] [blame]2460a: store undef -> %X
2461b: store %X -> undef
2462Safe:
2463a: &lt;deleted&gt;
2464b: unreachable
2465</pre>
Chris Lattner6e9057b2009-09-07 23:33:52 +0000[diff] [blame]2466
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]2467<p>These examples reiterate the <tt>fdiv</tt> example: a store <em>of</em> an
2468 undefined value can be assumed to not have any effect; we can assume that the
2469 value is overwritten with bits that happen to match what was already there.
2470 However, a store <em>to</em> an undefined location could clobber arbitrary
2471 memory, therefore, it has undefined behavior.</p>
Chris Lattner6e9057b2009-09-07 23:33:52 +0000[diff] [blame]2472
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2473</div>
2474
2475<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2476<h3>
2477 <a name="trapvalues">Trap Values</a>
2478</h3>
2479
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2480<div>
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]2481
Dan Gohmanc68ce062010-04-26 20:21:21 +0000[diff] [blame]2482<p>Trap values are similar to <a href="#undefvalues">undef values</a>, however
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]2483 instead of representing an unspecified bit pattern, they represent the
2484 fact that an instruction or constant expression which cannot evoke side
2485 effects has nevertheless detected a condition which results in undefined
Dan Gohmanc68ce062010-04-26 20:21:21 +0000[diff] [blame]2486 behavior.</p>
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]2487
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2488<p>There is currently no way of representing a trap value in the IR; they
Dan Gohman855abed2010-05-03 14:51:43 +0000[diff] [blame]2489 only exist when produced by operations such as
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2490 <a href="#i_add"><tt>add</tt></a> with the <tt>nsw</tt> flag.</p>
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2491
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2492<p>Trap value behavior is defined in terms of value <i>dependence</i>:</p>
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2493
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2494<ul>
2495<li>Values other than <a href="#i_phi"><tt>phi</tt></a> nodes depend on
2496 their operands.</li>
2497
2498<li><a href="#i_phi"><tt>Phi</tt></a> nodes depend on the operand corresponding
2499 to their dynamic predecessor basic block.</li>
2500
2501<li>Function arguments depend on the corresponding actual argument values in
2502 the dynamic callers of their functions.</li>
2503
2504<li><a href="#i_call"><tt>Call</tt></a> instructions depend on the
2505 <a href="#i_ret"><tt>ret</tt></a> instructions that dynamically transfer
2506 control back to them.</li>
2507
Dan Gohmanb5328162010-05-03 14:55:22 +0000[diff] [blame]2508<li><a href="#i_invoke"><tt>Invoke</tt></a> instructions depend on the
2509 <a href="#i_ret"><tt>ret</tt></a>, <a href="#i_unwind"><tt>unwind</tt></a>,
2510 or exception-throwing call instructions that dynamically transfer control
2511 back to them.</li>
2512
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2513<li>Non-volatile loads and stores depend on the most recent stores to all of the
2514 referenced memory addresses, following the order in the IR
2515 (including loads and stores implied by intrinsics such as
2516 <a href="#int_memcpy"><tt>@llvm.memcpy</tt></a>.)</li>
2517
Dan Gohman7c24ff12010-05-03 14:59:34 +0000[diff] [blame]2518<!-- TODO: In the case of multiple threads, this only applies if the store
2519 "happens-before" the load or store. -->
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2520
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2521<!-- TODO: floating-point exception state -->
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2522
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2523<li>An instruction with externally visible side effects depends on the most
2524 recent preceding instruction with externally visible side effects, following
Dan Gohmanff70fe42010-07-06 15:26:33 +0000[diff] [blame]2525 the order in the IR. (This includes
2526 <a href="#volatile">volatile operations</a>.)</li>
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]2527
Dan Gohmanb5328162010-05-03 14:55:22 +0000[diff] [blame]2528<li>An instruction <i>control-depends</i> on a
2529 <a href="#terminators">terminator instruction</a>
2530 if the terminator instruction has multiple successors and the instruction
2531 is always executed when control transfers to one of the successors, and
Chris Lattner7a2bdde2011-04-15 05:18:47 +0000[diff] [blame]2532 may not be executed when control is transferred to another.</li>
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2533
Dan Gohmanca4cac42011-04-12 23:05:59 +0000[diff] [blame]2534<li>Additionally, an instruction also <i>control-depends</i> on a terminator
2535 instruction if the set of instructions it otherwise depends on would be
Chris Lattner7a2bdde2011-04-15 05:18:47 +0000[diff] [blame]2536 different if the terminator had transferred control to a different
Dan Gohmanca4cac42011-04-12 23:05:59 +0000[diff] [blame]2537 successor.</li>
2538
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2539<li>Dependence is transitive.</li>
2540
2541</ul>
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2542
2543<p>Whenever a trap value is generated, all values which depend on it evaluate
2544 to trap. If they have side effects, the evoke their side effects as if each
2545 operand with a trap value were undef. If they have externally-visible side
2546 effects, the behavior is undefined.</p>
2547
2548<p>Here are some examples:</p>
Dan Gohmanc30f6e12010-04-26 20:54:53 +0000[diff] [blame]2549
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2550<pre class="doc_code">
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2551entry:
2552 %trap = sub nuw i32 0, 1 ; Results in a trap value.
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2553 %still_trap = and i32 %trap, 0 ; Whereas (and i32 undef, 0) would return 0.
2554 %trap_yet_again = getelementptr i32* @h, i32 %still_trap
2555 store i32 0, i32* %trap_yet_again ; undefined behavior
2556
2557 store i32 %trap, i32* @g ; Trap value conceptually stored to memory.
2558 %trap2 = load i32* @g ; Returns a trap value, not just undef.
2559
2560 volatile store i32 %trap, i32* @g ; External observation; undefined behavior.
2561
2562 %narrowaddr = bitcast i32* @g to i16*
2563 %wideaddr = bitcast i32* @g to i64*
Nick Lewycky64f9fb12011-05-16 19:29:30 +0000[diff] [blame]2564 %trap3 = load i16* %narrowaddr ; Returns a trap value.
2565 %trap4 = load i64* %wideaddr ; Returns a trap value.
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2566
Nick Lewycky64f9fb12011-05-16 19:29:30 +0000[diff] [blame]2567 %cmp = icmp slt i32 %trap, 0 ; Returns a trap value.
2568 br i1 %cmp, label %true, label %end ; Branch to either destination.
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2569
2570true:
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2571 volatile store i32 0, i32* @g ; This is control-dependent on %cmp, so
2572 ; it has undefined behavior.
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2573 br label %end
2574
2575end:
2576 %p = phi i32 [ 0, %entry ], [ 1, %true ]
2577 ; Both edges into this PHI are
2578 ; control-dependent on %cmp, so this
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2579 ; always results in a trap value.
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2580
Dan Gohmanca4cac42011-04-12 23:05:59 +0000[diff] [blame]2581 volatile store i32 0, i32* @g ; This would depend on the store in %true
2582 ; if %cmp is true, or the store in %entry
2583 ; otherwise, so this is undefined behavior.
2584
Nick Lewycky64f9fb12011-05-16 19:29:30 +0000[diff] [blame]2585 br i1 %cmp, label %second_true, label %second_end
Dan Gohmanca4cac42011-04-12 23:05:59 +0000[diff] [blame]2586 ; The same branch again, but this time the
2587 ; true block doesn't have side effects.
2588
2589second_true:
2590 ; No side effects!
Nick Lewycky64f9fb12011-05-16 19:29:30 +0000[diff] [blame]2591 ret void
Dan Gohmanca4cac42011-04-12 23:05:59 +0000[diff] [blame]2592
2593second_end:
2594 volatile store i32 0, i32* @g ; This time, the instruction always depends
2595 ; on the store in %end. Also, it is
2596 ; control-equivalent to %end, so this is
Nick Lewycky64f9fb12011-05-16 19:29:30 +0000[diff] [blame]2597 ; well-defined (again, ignoring earlier
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2598 ; undefined behavior in this example).
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2599</pre>
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]2600
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]2601</div>
2602
2603<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2604<h3>
2605 <a name="blockaddress">Addresses of Basic Blocks</a>
2606</h3>
2607
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2608<div>
Chris Lattnerc6f44362009-10-27 21:01:34 +0000[diff] [blame]2609
Chris Lattnercdfc9402009-11-01 01:27:45 +0000[diff] [blame]2610<p><b><tt>blockaddress(@function, %block)</tt></b></p>
Chris Lattnerc6f44362009-10-27 21:01:34 +0000[diff] [blame]2611
2612<p>The '<tt>blockaddress</tt>' constant computes the address of the specified
Chris Lattner2dfdf2a2009-10-27 21:49:40 +0000[diff] [blame]2613 basic block in the specified function, and always has an i8* type. Taking
Chris Lattnercdfc9402009-11-01 01:27:45 +0000[diff] [blame]2614 the address of the entry block is illegal.</p>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]2615
Chris Lattnerc6f44362009-10-27 21:01:34 +0000[diff] [blame]2616<p>This value only has defined behavior when used as an operand to the
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]2617 '<a href="#i_indirectbr"><tt>indirectbr</tt></a>' instruction, or for
2618 comparisons against null. Pointer equality tests between labels addresses
2619 results in undefined behavior &mdash; though, again, comparison against null
2620 is ok, and no label is equal to the null pointer. This may be passed around
2621 as an opaque pointer sized value as long as the bits are not inspected. This
2622 allows <tt>ptrtoint</tt> and arithmetic to be performed on these values so
2623 long as the original value is reconstituted before the <tt>indirectbr</tt>
2624 instruction.</p>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]2625
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]2626<p>Finally, some targets may provide defined semantics when using the value as
2627 the operand to an inline assembly, but that is target specific.</p>
Chris Lattnerc6f44362009-10-27 21:01:34 +0000[diff] [blame]2628
2629</div>
2630
2631
2632<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2633<h3>
2634 <a name="constantexprs">Constant Expressions</a>
2635</h3>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2636
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2637<div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2638
2639<p>Constant expressions are used to allow expressions involving other constants
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2640 to be used as constants. Constant expressions may be of
2641 any <a href="#t_firstclass">first class</a> type and may involve any LLVM
2642 operation that does not have side effects (e.g. load and call are not
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]2643 supported). The following is the syntax for constant expressions:</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2644
2645<dl>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2646 <dt><b><tt>trunc (CST to TYPE)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2647 <dd>Truncate a constant to another type. The bit size of CST must be larger
2648 than the bit size of TYPE. Both types must be integers.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2649
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2650 <dt><b><tt>zext (CST to TYPE)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2651 <dd>Zero extend a constant to another type. The bit size of CST must be
Duncan Sands28afd432010-07-13 12:06:14 +0000[diff] [blame]2652 smaller than the bit size of TYPE. Both types must be integers.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2653
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2654 <dt><b><tt>sext (CST to TYPE)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2655 <dd>Sign extend a constant to another type. The bit size of CST must be
Duncan Sands28afd432010-07-13 12:06:14 +0000[diff] [blame]2656 smaller than the bit size of TYPE. Both types must be integers.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2657
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2658 <dt><b><tt>fptrunc (CST to TYPE)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2659 <dd>Truncate a floating point constant to another floating point type. The
2660 size of CST must be larger than the size of TYPE. Both types must be
2661 floating point.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2662
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2663 <dt><b><tt>fpext (CST to TYPE)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2664 <dd>Floating point extend a constant to another type. The size of CST must be
2665 smaller or equal to the size of TYPE. Both types must be floating
2666 point.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2667
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2668 <dt><b><tt>fptoui (CST to TYPE)</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2669 <dd>Convert a floating point constant to the corresponding unsigned integer
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2670 constant. TYPE must be a scalar or vector integer type. CST must be of
2671 scalar or vector floating point type. Both CST and TYPE must be scalars,
2672 or vectors of the same number of elements. If the value won't fit in the
2673 integer type, the results are undefined.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2674
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2675 <dt><b><tt>fptosi (CST to TYPE)</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2676 <dd>Convert a floating point constant to the corresponding signed integer
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2677 constant. TYPE must be a scalar or vector integer type. CST must be of
2678 scalar or vector floating point type. Both CST and TYPE must be scalars,
2679 or vectors of the same number of elements. If the value won't fit in the
2680 integer type, the results are undefined.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2681
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2682 <dt><b><tt>uitofp (CST to TYPE)</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2683 <dd>Convert an unsigned integer constant to the corresponding floating point
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2684 constant. TYPE must be a scalar or vector floating point type. CST must be
2685 of scalar or vector integer type. Both CST and TYPE must be scalars, or
2686 vectors of the same number of elements. If the value won't fit in the
2687 floating point type, the results are undefined.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2688
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2689 <dt><b><tt>sitofp (CST to TYPE)</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2690 <dd>Convert a signed integer constant to the corresponding floating point
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2691 constant. TYPE must be a scalar or vector floating point type. CST must be
2692 of scalar or vector integer type. Both CST and TYPE must be scalars, or
2693 vectors of the same number of elements. If the value won't fit in the
2694 floating point type, the results are undefined.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2695
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2696 <dt><b><tt>ptrtoint (CST to TYPE)</tt></b></dt>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]2697 <dd>Convert a pointer typed constant to the corresponding integer constant
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2698 <tt>TYPE</tt> must be an integer type. <tt>CST</tt> must be of pointer
2699 type. The <tt>CST</tt> value is zero extended, truncated, or unchanged to
2700 make it fit in <tt>TYPE</tt>.</dd>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]2701
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2702 <dt><b><tt>inttoptr (CST to TYPE)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2703 <dd>Convert a integer constant to a pointer constant. TYPE must be a pointer
2704 type. CST must be of integer type. The CST value is zero extended,
2705 truncated, or unchanged to make it fit in a pointer size. This one is
2706 <i>really</i> dangerous!</dd>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]2707
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2708 <dt><b><tt>bitcast (CST to TYPE)</tt></b></dt>
Chris Lattner03bbad62009-02-28 18:27:03 +0000[diff] [blame]2709 <dd>Convert a constant, CST, to another TYPE. The constraints of the operands
2710 are the same as those for the <a href="#i_bitcast">bitcast
2711 instruction</a>.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2712
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2713 <dt><b><tt>getelementptr (CSTPTR, IDX0, IDX1, ...)</tt></b></dt>
2714 <dt><b><tt>getelementptr inbounds (CSTPTR, IDX0, IDX1, ...)</tt></b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2715 <dd>Perform the <a href="#i_getelementptr">getelementptr operation</a> on
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2716 constants. As with the <a href="#i_getelementptr">getelementptr</a>
2717 instruction, the index list may have zero or more indexes, which are
2718 required to make sense for the type of "CSTPTR".</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2719
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2720 <dt><b><tt>select (COND, VAL1, VAL2)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2721 <dd>Perform the <a href="#i_select">select operation</a> on constants.</dd>
Reid Spencer01c42592006-12-04 19:23:19 +0000[diff] [blame]2722
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2723 <dt><b><tt>icmp COND (VAL1, VAL2)</tt></b></dt>
Reid Spencer01c42592006-12-04 19:23:19 +0000[diff] [blame]2724 <dd>Performs the <a href="#i_icmp">icmp operation</a> on constants.</dd>
2725
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2726 <dt><b><tt>fcmp COND (VAL1, VAL2)</tt></b></dt>
Reid Spencer01c42592006-12-04 19:23:19 +0000[diff] [blame]2727 <dd>Performs the <a href="#i_fcmp">fcmp operation</a> on constants.</dd>
Robert Bocchino9fbe1452006-01-10 19:31:34 +0000[diff] [blame]2728
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2729 <dt><b><tt>extractelement (VAL, IDX)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2730 <dd>Perform the <a href="#i_extractelement">extractelement operation</a> on
2731 constants.</dd>
Robert Bocchino9fbe1452006-01-10 19:31:34 +0000[diff] [blame]2732
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2733 <dt><b><tt>insertelement (VAL, ELT, IDX)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2734 <dd>Perform the <a href="#i_insertelement">insertelement operation</a> on
2735 constants.</dd>
Chris Lattnerc1989542006-04-08 00:13:41 +0000[diff] [blame]2736
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2737 <dt><b><tt>shufflevector (VEC1, VEC2, IDXMASK)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2738 <dd>Perform the <a href="#i_shufflevector">shufflevector operation</a> on
2739 constants.</dd>
Chris Lattnerc1989542006-04-08 00:13:41 +0000[diff] [blame]2740
Nick Lewycky9e130ce2010-05-29 06:44:15 +0000[diff] [blame]2741 <dt><b><tt>extractvalue (VAL, IDX0, IDX1, ...)</tt></b></dt>
2742 <dd>Perform the <a href="#i_extractvalue">extractvalue operation</a> on
2743 constants. The index list is interpreted in a similar manner as indices in
2744 a '<a href="#i_getelementptr">getelementptr</a>' operation. At least one
2745 index value must be specified.</dd>
2746
2747 <dt><b><tt>insertvalue (VAL, ELT, IDX0, IDX1, ...)</tt></b></dt>
2748 <dd>Perform the <a href="#i_insertvalue">insertvalue operation</a> on
2749 constants. The index list is interpreted in a similar manner as indices in
2750 a '<a href="#i_getelementptr">getelementptr</a>' operation. At least one
2751 index value must be specified.</dd>
2752
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2753 <dt><b><tt>OPCODE (LHS, RHS)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2754 <dd>Perform the specified operation of the LHS and RHS constants. OPCODE may
2755 be any of the <a href="#binaryops">binary</a>
2756 or <a href="#bitwiseops">bitwise binary</a> operations. The constraints
2757 on operands are the same as those for the corresponding instruction
2758 (e.g. no bitwise operations on floating point values are allowed).</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2759</dl>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2760
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2761</div>
Chris Lattner9ee5d222004-03-08 16:49:10 +0000[diff] [blame]2762
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2763</div>
2764
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2765<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2766<h2><a name="othervalues">Other Values</a></h2>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2767<!-- *********************************************************************** -->
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2768<div>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2769<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2770<h3>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2771<a name="inlineasm">Inline Assembler Expressions</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2772</h3>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2773
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2774<div>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2775
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2776<p>LLVM supports inline assembler expressions (as opposed
2777 to <a href="#moduleasm"> Module-Level Inline Assembly</a>) through the use of
2778 a special value. This value represents the inline assembler as a string
2779 (containing the instructions to emit), a list of operand constraints (stored
Dale Johannesen09fed252009-10-13 21:56:55 +0000[diff] [blame]2780 as a string), a flag that indicates whether or not the inline asm
Dale Johannesen8ba2d5b2009-10-21 23:28:00 +0000[diff] [blame]2781 expression has side effects, and a flag indicating whether the function
2782 containing the asm needs to align its stack conservatively. An example
2783 inline assembler expression is:</p>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2784
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2785<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2786i32 (i32) asm "bswap $0", "=r,r"
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2787</pre>
2788
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2789<p>Inline assembler expressions may <b>only</b> be used as the callee operand of
2790 a <a href="#i_call"><tt>call</tt> instruction</a>. Thus, typically we
2791 have:</p>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2792
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2793<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2794%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]2795</pre>
2796
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2797<p>Inline asms with side effects not visible in the constraint list must be
2798 marked as having side effects. This is done through the use of the
2799 '<tt>sideeffect</tt>' keyword, like so:</p>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2800
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2801<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2802call void asm sideeffect "eieio", ""()
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2803</pre>
2804
Dale Johannesen8ba2d5b2009-10-21 23:28:00 +0000[diff] [blame]2805<p>In some cases inline asms will contain code that will not work unless the
2806 stack is aligned in some way, such as calls or SSE instructions on x86,
2807 yet will not contain code that does that alignment within the asm.
2808 The compiler should make conservative assumptions about what the asm might
2809 contain and should generate its usual stack alignment code in the prologue
2810 if the '<tt>alignstack</tt>' keyword is present:</p>
Dale Johannesen09fed252009-10-13 21:56:55 +0000[diff] [blame]2811
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2812<pre class="doc_code">
Dale Johannesen8ba2d5b2009-10-21 23:28:00 +0000[diff] [blame]2813call void asm alignstack "eieio", ""()
Dale Johannesen09fed252009-10-13 21:56:55 +0000[diff] [blame]2814</pre>
Dale Johannesen09fed252009-10-13 21:56:55 +0000[diff] [blame]2815
2816<p>If both keywords appear the '<tt>sideeffect</tt>' keyword must come
2817 first.</p>
2818
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2819<p>TODO: The format of the asm and constraints string still need to be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2820 documented here. Constraints on what can be done (e.g. duplication, moving,
2821 etc need to be documented). This is probably best done by reference to
2822 another document that covers inline asm from a holistic perspective.</p>
Chris Lattnercf9a4152010-04-07 05:38:05 +0000[diff] [blame]2823
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2824<h4>
Chris Lattnercf9a4152010-04-07 05:38:05 +0000[diff] [blame]2825<a name="inlineasm_md">Inline Asm Metadata</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2826</h4>
Chris Lattnercf9a4152010-04-07 05:38:05 +0000[diff] [blame]2827
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2828<div>
Chris Lattnercf9a4152010-04-07 05:38:05 +0000[diff] [blame]2829
2830<p>The call instructions that wrap inline asm nodes may have a "!srcloc" MDNode
Chris Lattnerce1b9ad2010-11-17 08:20:42 +0000[diff] [blame]2831 attached to it that contains a list of constant integers. If present, the
2832 code generator will use the integer as the location cookie value when report
Chris Lattnercf9a4152010-04-07 05:38:05 +0000[diff] [blame]2833 errors through the LLVMContext error reporting mechanisms. This allows a
Dan Gohman1c70c002010-04-28 00:36:01 +0000[diff] [blame]2834 front-end to correlate backend errors that occur with inline asm back to the
Chris Lattnercf9a4152010-04-07 05:38:05 +0000[diff] [blame]2835 source code that produced it. For example:</p>
2836
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2837<pre class="doc_code">
Chris Lattnercf9a4152010-04-07 05:38:05 +0000[diff] [blame]2838call void asm sideeffect "something bad", ""()<b>, !srcloc !42</b>
2839...
2840!42 = !{ i32 1234567 }
2841</pre>
Chris Lattnercf9a4152010-04-07 05:38:05 +0000[diff] [blame]2842
2843<p>It is up to the front-end to make sense of the magic numbers it places in the
Chris Lattnerce1b9ad2010-11-17 08:20:42 +0000[diff] [blame]2844 IR. If the MDNode contains multiple constants, the code generator will use
2845 the one that corresponds to the line of the asm that the error occurs on.</p>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2846
2847</div>
2848
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2849</div>
2850
Chris Lattnere6a5ddd2010-01-15 21:50:19 +0000[diff] [blame]2851<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2852<h3>
2853 <a name="metadata">Metadata Nodes and Metadata Strings</a>
2854</h3>
Chris Lattnere6a5ddd2010-01-15 21:50:19 +0000[diff] [blame]2855
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2856<div>
Chris Lattnere6a5ddd2010-01-15 21:50:19 +0000[diff] [blame]2857
2858<p>LLVM IR allows metadata to be attached to instructions in the program that
2859 can convey extra information about the code to the optimizers and code
2860 generator. One example application of metadata is source-level debug
2861 information. There are two metadata primitives: strings and nodes. All
2862 metadata has the <tt>metadata</tt> type and is identified in syntax by a
2863 preceding exclamation point ('<tt>!</tt>').</p>
2864
2865<p>A metadata string is a string surrounded by double quotes. It can contain
2866 any character by escaping non-printable characters with "\xx" where "xx" is
2867 the two digit hex code. For example: "<tt>!"test\00"</tt>".</p>
2868
2869<p>Metadata nodes are represented with notation similar to structure constants
2870 (a comma separated list of elements, surrounded by braces and preceded by an
2871 exclamation point). For example: "<tt>!{ metadata !"test\00", i32
2872 10}</tt>". Metadata nodes can have any values as their operand.</p>
2873
2874<p>A <a href="#namedmetadatastructure">named metadata</a> is a collection of
2875 metadata nodes, which can be looked up in the module symbol table. For
2876 example: "<tt>!foo = metadata !{!4, !3}</tt>".
2877
Devang Patele1d50cd2010-03-04 23:44:48 +0000[diff] [blame]2878<p>Metadata can be used as function arguments. Here <tt>llvm.dbg.value</tt>
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2879 function is using two metadata arguments.</p>
Devang Patele1d50cd2010-03-04 23:44:48 +0000[diff] [blame]2880
Bill Wendling9ff5de92011-03-02 02:17:11 +0000[diff] [blame]2881<div class="doc_code">
2882<pre>
2883call void @llvm.dbg.value(metadata !24, i64 0, metadata !25)
2884</pre>
2885</div>
Devang Patele1d50cd2010-03-04 23:44:48 +0000[diff] [blame]2886
2887<p>Metadata can be attached with an instruction. Here metadata <tt>!21</tt> is
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2888 attached with <tt>add</tt> instruction using <tt>!dbg</tt> identifier.</p>
Devang Patele1d50cd2010-03-04 23:44:48 +0000[diff] [blame]2889
Bill Wendling9ff5de92011-03-02 02:17:11 +0000[diff] [blame]2890<div class="doc_code">
2891<pre>
2892%indvar.next = add i64 %indvar, 1, !dbg !21
2893</pre>
2894</div>
2895
Chris Lattnere6a5ddd2010-01-15 21:50:19 +0000[diff] [blame]2896</div>
2897
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2898</div>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2899
2900<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2901<h2>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2902 <a name="intrinsic_globals">Intrinsic Global Variables</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2903</h2>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2904<!-- *********************************************************************** -->
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2905<div>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2906<p>LLVM has a number of "magic" global variables that contain data that affect
2907code generation or other IR semantics. These are documented here. All globals
Chris Lattner401e10c2009-07-20 06:14:25 +0000[diff] [blame]2908of this sort should have a section specified as "<tt>llvm.metadata</tt>". This
2909section and all globals that start with "<tt>llvm.</tt>" are reserved for use
2910by LLVM.</p>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2911
2912<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2913<h3>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2914<a name="intg_used">The '<tt>llvm.used</tt>' Global Variable</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2915</h3>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2916
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2917<div>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2918
2919<p>The <tt>@llvm.used</tt> global is an array with i8* element type which has <a
2920href="#linkage_appending">appending linkage</a>. This array contains a list of
2921pointers to global variables and functions which may optionally have a pointer
2922cast formed of bitcast or getelementptr. For example, a legal use of it is:</p>
2923
2924<pre>
2925 @X = global i8 4
2926 @Y = global i32 123
2927
2928 @llvm.used = appending global [2 x i8*] [
2929 i8* @X,
2930 i8* bitcast (i32* @Y to i8*)
2931 ], section "llvm.metadata"
2932</pre>
2933
2934<p>If a global variable appears in the <tt>@llvm.used</tt> list, then the
2935compiler, assembler, and linker are required to treat the symbol as if there is
2936a reference to the global that it cannot see. For example, if a variable has
2937internal linkage and no references other than that from the <tt>@llvm.used</tt>
2938list, it cannot be deleted. This is commonly used to represent references from
2939inline asms and other things the compiler cannot "see", and corresponds to
2940"attribute((used))" in GNU C.</p>
2941
2942<p>On some targets, the code generator must emit a directive to the assembler or
2943object file to prevent the assembler and linker from molesting the symbol.</p>
2944
2945</div>
2946
2947<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2948<h3>
2949 <a name="intg_compiler_used">
2950 The '<tt>llvm.compiler.used</tt>' Global Variable
2951 </a>
2952</h3>
Chris Lattner401e10c2009-07-20 06:14:25 +0000[diff] [blame]2953
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2954<div>
Chris Lattner401e10c2009-07-20 06:14:25 +0000[diff] [blame]2955
2956<p>The <tt>@llvm.compiler.used</tt> directive is the same as the
2957<tt>@llvm.used</tt> directive, except that it only prevents the compiler from
2958touching the symbol. On targets that support it, this allows an intelligent
2959linker to optimize references to the symbol without being impeded as it would be
2960by <tt>@llvm.used</tt>.</p>
2961
2962<p>This is a rare construct that should only be used in rare circumstances, and
2963should not be exposed to source languages.</p>
2964
2965</div>
2966
2967<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2968<h3>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2969<a name="intg_global_ctors">The '<tt>llvm.global_ctors</tt>' Global Variable</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2970</h3>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2971
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2972<div>
David Chisnalle31e9962010-04-30 19:23:49 +0000[diff] [blame]2973<pre>
2974%0 = type { i32, void ()* }
David Chisnall27195a52010-04-30 19:27:35 +0000[diff] [blame]2975@llvm.global_ctors = appending global [1 x %0] [%0 { i32 65535, void ()* @ctor }]
David Chisnalle31e9962010-04-30 19:23:49 +0000[diff] [blame]2976</pre>
2977<p>The <tt>@llvm.global_ctors</tt> array contains a list of constructor functions and associated priorities. The functions referenced by this array will be called in ascending order of priority (i.e. lowest first) when the module is loaded. The order of functions with the same priority is not defined.
2978</p>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2979
2980</div>
2981
2982<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2983<h3>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2984<a name="intg_global_dtors">The '<tt>llvm.global_dtors</tt>' Global Variable</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2985</h3>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2986
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2987<div>
David Chisnalle31e9962010-04-30 19:23:49 +0000[diff] [blame]2988<pre>
2989%0 = type { i32, void ()* }
David Chisnall27195a52010-04-30 19:27:35 +0000[diff] [blame]2990@llvm.global_dtors = appending global [1 x %0] [%0 { i32 65535, void ()* @dtor }]
David Chisnalle31e9962010-04-30 19:23:49 +0000[diff] [blame]2991</pre>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2992
David Chisnalle31e9962010-04-30 19:23:49 +0000[diff] [blame]2993<p>The <tt>@llvm.global_dtors</tt> array contains a list of destructor functions and associated priorities. The functions referenced by this array will be called in descending order of priority (i.e. highest first) when the module is loaded. The order of functions with the same priority is not defined.
2994</p>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2995
2996</div>
2997
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2998</div>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2999
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]3000<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3001<h2><a name="instref">Instruction Reference</a></h2>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3002<!-- *********************************************************************** -->
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]3003
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3004<div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]3005
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3006<p>The LLVM instruction set consists of several different classifications of
3007 instructions: <a href="#terminators">terminator
3008 instructions</a>, <a href="#binaryops">binary instructions</a>,
3009 <a href="#bitwiseops">bitwise binary instructions</a>,
3010 <a href="#memoryops">memory instructions</a>, and
3011 <a href="#otherops">other instructions</a>.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]3012
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3013<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3014<h3>
3015 <a name="terminators">Terminator Instructions</a>
3016</h3>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]3017
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3018<div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]3019
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3020<p>As mentioned <a href="#functionstructure">previously</a>, every basic block
3021 in a program ends with a "Terminator" instruction, which indicates which
3022 block should be executed after the current block is finished. These
3023 terminator instructions typically yield a '<tt>void</tt>' value: they produce
3024 control flow, not values (the one exception being the
3025 '<a href="#i_invoke"><tt>invoke</tt></a>' instruction).</p>
3026
Bill Wendlingdccc03b2011-07-31 06:30:59 +0000[diff] [blame^]3027<p>There are eight different terminator instructions: the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3028 '<a href="#i_ret"><tt>ret</tt></a>' instruction, the
3029 '<a href="#i_br"><tt>br</tt></a>' instruction, the
3030 '<a href="#i_switch"><tt>switch</tt></a>' instruction, the
Bill Wendling21c346e2009-11-02 00:25:26 +0000[diff] [blame]3031 '<a href="#i_indirectbr">'<tt>indirectbr</tt></a>' Instruction, the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3032 '<a href="#i_invoke"><tt>invoke</tt></a>' instruction, the
Bill Wendlingdccc03b2011-07-31 06:30:59 +0000[diff] [blame^]3033 '<a href="#i_unwind"><tt>unwind</tt></a>' instruction, the
3034 '<a href="#i_resume"><tt>resume</tt></a>' instruction, and the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3035 '<a href="#i_unreachable"><tt>unreachable</tt></a>' instruction.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]3036
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3037<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3038<h4>
3039 <a name="i_ret">'<tt>ret</tt>' Instruction</a>
3040</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3041
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3042<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3043
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3044<h5>Syntax:</h5>
Dan Gohmanb1e6b962008-10-04 19:00:07 +0000[diff] [blame]3045<pre>
3046 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]3047 ret void <i>; Return from void function</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3048</pre>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]3049
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3050<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3051<p>The '<tt>ret</tt>' instruction is used to return control flow (and optionally
3052 a value) from a function back to the caller.</p>
3053
3054<p>There are two forms of the '<tt>ret</tt>' instruction: one that returns a
3055 value and then causes control flow, and one that just causes control flow to
3056 occur.</p>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]3057
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3058<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3059<p>The '<tt>ret</tt>' instruction optionally accepts a single argument, the
3060 return value. The type of the return value must be a
3061 '<a href="#t_firstclass">first class</a>' type.</p>
Dan Gohmanb1e6b962008-10-04 19:00:07 +0000[diff] [blame]3062
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3063<p>A function is not <a href="#wellformed">well formed</a> if it it has a
3064 non-void return type and contains a '<tt>ret</tt>' instruction with no return
3065 value or a return value with a type that does not match its type, or if it
3066 has a void return type and contains a '<tt>ret</tt>' instruction with a
3067 return value.</p>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]3068
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3069<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3070<p>When the '<tt>ret</tt>' instruction is executed, control flow returns back to
3071 the calling function's context. If the caller is a
3072 "<a href="#i_call"><tt>call</tt></a>" instruction, execution continues at the
3073 instruction after the call. If the caller was an
3074 "<a href="#i_invoke"><tt>invoke</tt></a>" instruction, execution continues at
3075 the beginning of the "normal" destination block. If the instruction returns
3076 a value, that value shall set the call or invoke instruction's return
3077 value.</p>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]3078
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3079<h5>Example:</h5>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]3080<pre>
3081 ret i32 5 <i>; Return an integer value of 5</i>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]3082 ret void <i>; Return from a void function</i>
Bill Wendling0a4bbbf2009-02-28 22:12:54 +0000[diff] [blame]3083 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]3084</pre>
Dan Gohmanf3e60bd2009-01-12 23:12:39 +0000[diff] [blame]3085
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3086</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3087<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3088<h4>
3089 <a name="i_br">'<tt>br</tt>' Instruction</a>
3090</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3091
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3092<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3093
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3094<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3095<pre>
Bill Wendlingb3aa4712011-07-26 10:41:15 +0000[diff] [blame]3096 br i1 &lt;cond&gt;, label &lt;iftrue&gt;, label &lt;iffalse&gt;
3097 br label &lt;dest&gt; <i>; Unconditional branch</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3098</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3099
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3100<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3101<p>The '<tt>br</tt>' instruction is used to cause control flow to transfer to a
3102 different basic block in the current function. There are two forms of this
3103 instruction, corresponding to a conditional branch and an unconditional
3104 branch.</p>
3105
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3106<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3107<p>The conditional branch form of the '<tt>br</tt>' instruction takes a single
3108 '<tt>i1</tt>' value and two '<tt>label</tt>' values. The unconditional form
3109 of the '<tt>br</tt>' instruction takes a single '<tt>label</tt>' value as a
3110 target.</p>
3111
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3112<h5>Semantics:</h5>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3113<p>Upon execution of a conditional '<tt>br</tt>' instruction, the '<tt>i1</tt>'
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3114 argument is evaluated. If the value is <tt>true</tt>, control flows to the
3115 '<tt>iftrue</tt>' <tt>label</tt> argument. If "cond" is <tt>false</tt>,
3116 control flows to the '<tt>iffalse</tt>' <tt>label</tt> argument.</p>
3117
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3118<h5>Example:</h5>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]3119<pre>
3120Test:
3121 %cond = <a href="#i_icmp">icmp</a> eq i32 %a, %b
3122 br i1 %cond, label %IfEqual, label %IfUnequal
3123IfEqual:
3124 <a href="#i_ret">ret</a> i32 1
3125IfUnequal:
3126 <a href="#i_ret">ret</a> i32 0
3127</pre>
3128
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3129</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3130
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3131<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3132<h4>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3133 <a name="i_switch">'<tt>switch</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3134</h4>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3135
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3136<div>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3137
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3138<h5>Syntax:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3139<pre>
3140 switch &lt;intty&gt; &lt;value&gt;, label &lt;defaultdest&gt; [ &lt;intty&gt; &lt;val&gt;, label &lt;dest&gt; ... ]
3141</pre>
3142
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3143<h5>Overview:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3144<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]3145 several different places. It is a generalization of the '<tt>br</tt>'
3146 instruction, allowing a branch to occur to one of many possible
3147 destinations.</p>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3148
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3149<h5>Arguments:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3150<p>The '<tt>switch</tt>' instruction uses three parameters: an integer
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3151 comparison value '<tt>value</tt>', a default '<tt>label</tt>' destination,
3152 and an array of pairs of comparison value constants and '<tt>label</tt>'s.
3153 The table is not allowed to contain duplicate constant entries.</p>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3154
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3155<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3156<p>The <tt>switch</tt> instruction specifies a table of values and
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3157 destinations. When the '<tt>switch</tt>' instruction is executed, this table
3158 is searched for the given value. If the value is found, control flow is
Benjamin Kramer8040cd32009-10-12 14:46:08 +0000[diff] [blame]3159 transferred to the corresponding destination; otherwise, control flow is
3160 transferred to the default destination.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3161
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3162<h5>Implementation:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3163<p>Depending on properties of the target machine and the particular
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3164 <tt>switch</tt> instruction, this instruction may be code generated in
3165 different ways. For example, it could be generated as a series of chained
3166 conditional branches or with a lookup table.</p>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3167
3168<h5>Example:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3169<pre>
3170 <i>; Emulate a conditional br instruction</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3171 %Val = <a href="#i_zext">zext</a> i1 %value to i32
Dan Gohman2a08c532009-01-04 23:44:43 +0000[diff] [blame]3172 switch i32 %Val, label %truedest [ i32 0, label %falsedest ]
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3173
3174 <i>; Emulate an unconditional br instruction</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3175 switch i32 0, label %dest [ ]
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3176
3177 <i>; Implement a jump table:</i>
Dan Gohman2a08c532009-01-04 23:44:43 +0000[diff] [blame]3178 switch i32 %val, label %otherwise [ i32 0, label %onzero
3179 i32 1, label %onone
3180 i32 2, label %ontwo ]
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3181</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3182
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3183</div>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3184
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]3185
3186<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3187<h4>
Chris Lattnerab21db72009-10-28 00:19:10 +0000[diff] [blame]3188 <a name="i_indirectbr">'<tt>indirectbr</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3189</h4>
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]3190
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3191<div>
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]3192
3193<h5>Syntax:</h5>
3194<pre>
Chris Lattnerab21db72009-10-28 00:19:10 +0000[diff] [blame]3195 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]3196</pre>
3197
3198<h5>Overview:</h5>
3199
Chris Lattnerab21db72009-10-28 00:19:10 +0000[diff] [blame]3200<p>The '<tt>indirectbr</tt>' instruction implements an indirect branch to a label
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]3201 within the current function, whose address is specified by
Chris Lattnerc6f44362009-10-27 21:01:34 +0000[diff] [blame]3202 "<tt>address</tt>". Address must be derived from a <a
3203 href="#blockaddress">blockaddress</a> constant.</p>
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]3204
3205<h5>Arguments:</h5>
3206
3207<p>The '<tt>address</tt>' argument is the address of the label to jump to. The
3208 rest of the arguments indicate the full set of possible destinations that the
3209 address may point to. Blocks are allowed to occur multiple times in the
3210 destination list, though this isn't particularly useful.</p>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3211
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]3212<p>This destination list is required so that dataflow analysis has an accurate
3213 understanding of the CFG.</p>
3214
3215<h5>Semantics:</h5>
3216
3217<p>Control transfers to the block specified in the address argument. All
3218 possible destination blocks must be listed in the label list, otherwise this
3219 instruction has undefined behavior. This implies that jumps to labels
3220 defined in other functions have undefined behavior as well.</p>
3221
3222<h5>Implementation:</h5>
3223
3224<p>This is typically implemented with a jump through a register.</p>
3225
3226<h5>Example:</h5>
3227<pre>
Chris Lattnerab21db72009-10-28 00:19:10 +0000[diff] [blame]3228 indirectbr i8* %Addr, [ label %bb1, label %bb2, label %bb3 ]
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]3229</pre>
3230
3231</div>
3232
3233
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3234<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3235<h4>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3236 <a name="i_invoke">'<tt>invoke</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3237</h4>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3238
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3239<div>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3240
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3241<h5>Syntax:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3242<pre>
Devang Patel307e8ab2008-10-07 17:48:33 +0000[diff] [blame]3243 &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]3244 to label &lt;normal label&gt; unwind label &lt;exception label&gt;
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3245</pre>
3246
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]3247<h5>Overview:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3248<p>The '<tt>invoke</tt>' instruction causes control to transfer to a specified
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3249 function, with the possibility of control flow transfer to either the
3250 '<tt>normal</tt>' label or the '<tt>exception</tt>' label. If the callee
3251 function returns with the "<tt><a href="#i_ret">ret</a></tt>" instruction,
3252 control flow will return to the "normal" label. If the callee (or any
3253 indirect callees) returns with the "<a href="#i_unwind"><tt>unwind</tt></a>"
3254 instruction, control is interrupted and continued at the dynamically nearest
3255 "exception" label.</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3256
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3257<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3258<p>This instruction requires several arguments:</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3259
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3260<ol>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3261 <li>The optional "cconv" marker indicates which <a href="#callingconv">calling
3262 convention</a> the call should use. If none is specified, the call
3263 defaults to using C calling conventions.</li>
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]3264
3265 <li>The optional <a href="#paramattrs">Parameter Attributes</a> list for
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3266 return values. Only '<tt>zeroext</tt>', '<tt>signext</tt>', and
3267 '<tt>inreg</tt>' attributes are valid here.</li>
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]3268
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3269 <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]3270 function value being invoked. In most cases, this is a direct function
3271 invocation, but indirect <tt>invoke</tt>s are just as possible, branching
3272 off an arbitrary pointer to function value.</li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3273
3274 <li>'<tt>function ptr val</tt>': An LLVM value containing a pointer to a
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3275 function to be invoked. </li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3276
3277 <li>'<tt>function args</tt>': argument list whose types match the function
Chris Lattner0724fbd2010-03-02 06:36:51 +0000[diff] [blame]3278 signature argument types and parameter attributes. All arguments must be
3279 of <a href="#t_firstclass">first class</a> type. If the function
3280 signature indicates the function accepts a variable number of arguments,
3281 the extra arguments can be specified.</li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3282
3283 <li>'<tt>normal label</tt>': the label reached when the called function
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3284 executes a '<tt><a href="#i_ret">ret</a></tt>' instruction. </li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3285
3286 <li>'<tt>exception label</tt>': the label reached when a callee returns with
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3287 the <a href="#i_unwind"><tt>unwind</tt></a> instruction. </li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3288
Devang Patel307e8ab2008-10-07 17:48:33 +0000[diff] [blame]3289 <li>The optional <a href="#fnattrs">function attributes</a> list. Only
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3290 '<tt>noreturn</tt>', '<tt>nounwind</tt>', '<tt>readonly</tt>' and
3291 '<tt>readnone</tt>' attributes are valid here.</li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3292</ol>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3293
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3294<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3295<p>This instruction is designed to operate as a standard
3296 '<tt><a href="#i_call">call</a></tt>' instruction in most regards. The
3297 primary difference is that it establishes an association with a label, which
3298 is used by the runtime library to unwind the stack.</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3299
3300<p>This instruction is used in languages with destructors to ensure that proper
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3301 cleanup is performed in the case of either a <tt>longjmp</tt> or a thrown
3302 exception. Additionally, this is important for implementation of
3303 '<tt>catch</tt>' clauses in high-level languages that support them.</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3304
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3305<p>For the purposes of the SSA form, the definition of the value returned by the
3306 '<tt>invoke</tt>' instruction is deemed to occur on the edge from the current
3307 block to the "normal" label. If the callee unwinds then no return value is
3308 available.</p>
Dan Gohmanf96a4992009-05-22 21:47:08 +0000[diff] [blame]3309
Chris Lattnerdf7a6802010-01-15 18:08:37 +0000[diff] [blame]3310<p>Note that the code generator does not yet completely support unwind, and
3311that the invoke/unwind semantics are likely to change in future versions.</p>
3312
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3313<h5>Example:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3314<pre>
Nick Lewyckyd703f652008-03-16 07:18:12 +0000[diff] [blame]3315 %retval = invoke i32 @Test(i32 15) to label %Continue
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]3316 unwind label %TestCleanup <i>; {i32}:retval set</i>
Nick Lewyckyd703f652008-03-16 07:18:12 +0000[diff] [blame]3317 %retval = invoke <a href="#callingconv">coldcc</a> i32 %Testfnptr(i32 15) to label %Continue
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]3318 unwind label %TestCleanup <i>; {i32}:retval set</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3319</pre>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3320
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3321</div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3322
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]3323<!-- _______________________________________________________________________ -->
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3324
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3325<h4>
3326 <a name="i_unwind">'<tt>unwind</tt>' Instruction</a>
3327</h4>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3328
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3329<div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3330
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]3331<h5>Syntax:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3332<pre>
3333 unwind
3334</pre>
3335
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]3336<h5>Overview:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3337<p>The '<tt>unwind</tt>' instruction unwinds the stack, continuing control flow
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3338 at the first callee in the dynamic call stack which used
3339 an <a href="#i_invoke"><tt>invoke</tt></a> instruction to perform the call.
3340 This is primarily used to implement exception handling.</p>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3341
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]3342<h5>Semantics:</h5>
Chris Lattner72ed2002008-04-19 21:01:16 +0000[diff] [blame]3343<p>The '<tt>unwind</tt>' instruction causes execution of the current function to
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3344 immediately halt. The dynamic call stack is then searched for the
3345 first <a href="#i_invoke"><tt>invoke</tt></a> instruction on the call stack.
3346 Once found, execution continues at the "exceptional" destination block
3347 specified by the <tt>invoke</tt> instruction. If there is no <tt>invoke</tt>
3348 instruction in the dynamic call chain, undefined behavior results.</p>
3349
Chris Lattnerdf7a6802010-01-15 18:08:37 +0000[diff] [blame]3350<p>Note that the code generator does not yet completely support unwind, and
3351that the invoke/unwind semantics are likely to change in future versions.</p>
3352
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3353</div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3354
Bill Wendlingdccc03b2011-07-31 06:30:59 +0000[diff] [blame^]3355 <!-- _______________________________________________________________________ -->
3356
3357<h4>
3358 <a name="i_resume">'<tt>resume</tt>' Instruction</a>
3359</h4>
3360
3361<div>
3362
3363<h5>Syntax:</h5>
3364<pre>
3365 resume &lt;type&gt; &lt;value&gt;
3366</pre>
3367
3368<h5>Overview:</h5>
3369<p>The '<tt>resume</tt>' instruction is a terminator instruction that has no
3370 successors.</p>
3371
3372<h5>Arguments:</h5>
3373<p>The '<tt>resume</tt>' instruction's argument must have the same type as the
3374 result of any '<tt>landingpad</tt>' instruction in the same function.</p>
3375
3376<h5>Semantics:</h5>
3377<p>The '<tt>resume</tt>' instruction resumes propagation of an existing
3378 (in-flight) exception whose unwinding was interrupted with
3379 a landingpad instruction.</p>
3380
3381<h5>Example:</h5>
3382<pre>
3383 resume { i8*, i32 } %exn
3384</pre>
3385
3386</div>
3387
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3388<!-- _______________________________________________________________________ -->
3389
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3390<h4>
3391 <a name="i_unreachable">'<tt>unreachable</tt>' Instruction</a>
3392</h4>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3393
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3394<div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3395
3396<h5>Syntax:</h5>
3397<pre>
3398 unreachable
3399</pre>
3400
3401<h5>Overview:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3402<p>The '<tt>unreachable</tt>' instruction has no defined semantics. This
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3403 instruction is used to inform the optimizer that a particular portion of the
3404 code is not reachable. This can be used to indicate that the code after a
3405 no-return function cannot be reached, and other facts.</p>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3406
3407<h5>Semantics:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3408<p>The '<tt>unreachable</tt>' instruction has no defined semantics.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3409
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3410</div>
3411
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3412</div>
3413
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3414<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3415<h3>
3416 <a name="binaryops">Binary Operations</a>
3417</h3>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3418
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3419<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3420
3421<p>Binary operators are used to do most of the computation in a program. They
3422 require two operands of the same type, execute an operation on them, and
3423 produce a single value. The operands might represent multiple data, as is
3424 the case with the <a href="#t_vector">vector</a> data type. The result value
3425 has the same type as its operands.</p>
3426
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3427<p>There are several different binary operators:</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3428
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3429<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3430<h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3431 <a name="i_add">'<tt>add</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3432</h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3433
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3434<div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3435
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3436<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3437<pre>
Dan Gohmancbb38f22009-07-20 22:41:19 +0000[diff] [blame]3438 &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]3439 &lt;result&gt; = add nuw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3440 &lt;result&gt; = add nsw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3441 &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]3442</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3443
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3444<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3445<p>The '<tt>add</tt>' instruction returns the sum of its two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3446
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3447<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3448<p>The two arguments to the '<tt>add</tt>' instruction must
3449 be <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of
3450 integer values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3451
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3452<h5>Semantics:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3453<p>The value produced is the integer sum of the two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3454
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3455<p>If the sum has unsigned overflow, the result returned is the mathematical
3456 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]3457
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3458<p>Because LLVM integers use a two's complement representation, this instruction
3459 is appropriate for both signed and unsigned integers.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3460
Dan Gohman08d012e2009-07-22 22:44:56 +0000[diff] [blame]3461<p><tt>nuw</tt> and <tt>nsw</tt> stand for &quot;No Unsigned Wrap&quot;
3462 and &quot;No Signed Wrap&quot;, respectively. If the <tt>nuw</tt> and/or
3463 <tt>nsw</tt> keywords are present, the result value of the <tt>add</tt>
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]3464 is a <a href="#trapvalues">trap value</a> if unsigned and/or signed overflow,
3465 respectively, occurs.</p>
Dan Gohmancbb38f22009-07-20 22:41:19 +0000[diff] [blame]3466
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3467<h5>Example:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3468<pre>
3469 &lt;result&gt; = add i32 4, %var <i>; yields {i32}:result = 4 + %var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3470</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3471
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3472</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3473
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3474<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3475<h4>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3476 <a name="i_fadd">'<tt>fadd</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3477</h4>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3478
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3479<div>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3480
3481<h5>Syntax:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3482<pre>
3483 &lt;result&gt; = fadd &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3484</pre>
3485
3486<h5>Overview:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3487<p>The '<tt>fadd</tt>' instruction returns the sum of its two operands.</p>
3488
3489<h5>Arguments:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3490<p>The two arguments to the '<tt>fadd</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3491 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
3492 floating point values. Both arguments must have identical types.</p>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3493
3494<h5>Semantics:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3495<p>The value produced is the floating point sum of the two operands.</p>
3496
3497<h5>Example:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3498<pre>
3499 &lt;result&gt; = fadd float 4.0, %var <i>; yields {float}:result = 4.0 + %var</i>
3500</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3501
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3502</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3503
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3504<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3505<h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3506 <a name="i_sub">'<tt>sub</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3507</h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3508
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3509<div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3510
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3511<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3512<pre>
Dan Gohman08d012e2009-07-22 22:44:56 +0000[diff] [blame]3513 &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]3514 &lt;result&gt; = sub nuw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3515 &lt;result&gt; = sub nsw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3516 &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]3517</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3518
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3519<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3520<p>The '<tt>sub</tt>' instruction returns the difference of its two
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3521 operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3522
3523<p>Note that the '<tt>sub</tt>' instruction is used to represent the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3524 '<tt>neg</tt>' instruction present in most other intermediate
3525 representations.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3526
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3527<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3528<p>The two arguments to the '<tt>sub</tt>' instruction must
3529 be <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of
3530 integer values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3531
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3532<h5>Semantics:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3533<p>The value produced is the integer difference of the two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3534
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3535<p>If the difference has unsigned overflow, the result returned is the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3536 mathematical result modulo 2<sup>n</sup>, where n is the bit width of the
3537 result.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3538
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3539<p>Because LLVM integers use a two's complement representation, this instruction
3540 is appropriate for both signed and unsigned integers.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3541
Dan Gohman08d012e2009-07-22 22:44:56 +0000[diff] [blame]3542<p><tt>nuw</tt> and <tt>nsw</tt> stand for &quot;No Unsigned Wrap&quot;
3543 and &quot;No Signed Wrap&quot;, respectively. If the <tt>nuw</tt> and/or
3544 <tt>nsw</tt> keywords are present, the result value of the <tt>sub</tt>
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]3545 is a <a href="#trapvalues">trap value</a> if unsigned and/or signed overflow,
3546 respectively, occurs.</p>
Dan Gohmancbb38f22009-07-20 22:41:19 +0000[diff] [blame]3547
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3548<h5>Example:</h5>
Bill Wendlingaac388b2007-05-29 09:42:13 +0000[diff] [blame]3549<pre>
3550 &lt;result&gt; = sub i32 4, %var <i>; yields {i32}:result = 4 - %var</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3551 &lt;result&gt; = sub i32 0, %val <i>; yields {i32}:result = -%var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3552</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3553
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3554</div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3555
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3556<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3557<h4>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3558 <a name="i_fsub">'<tt>fsub</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3559</h4>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3560
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3561<div>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3562
3563<h5>Syntax:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3564<pre>
3565 &lt;result&gt; = fsub &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3566</pre>
3567
3568<h5>Overview:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3569<p>The '<tt>fsub</tt>' instruction returns the difference of its two
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3570 operands.</p>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3571
3572<p>Note that the '<tt>fsub</tt>' instruction is used to represent the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3573 '<tt>fneg</tt>' instruction present in most other intermediate
3574 representations.</p>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3575
3576<h5>Arguments:</h5>
Bill Wendlingd9fe2982009-07-20 02:32:41 +0000[diff] [blame]3577<p>The two arguments to the '<tt>fsub</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3578 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
3579 floating point values. Both arguments must have identical types.</p>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3580
3581<h5>Semantics:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3582<p>The value produced is the floating point difference of the two operands.</p>
3583
3584<h5>Example:</h5>
3585<pre>
3586 &lt;result&gt; = fsub float 4.0, %var <i>; yields {float}:result = 4.0 - %var</i>
3587 &lt;result&gt; = fsub float -0.0, %val <i>; yields {float}:result = -%var</i>
3588</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3589
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3590</div>
3591
3592<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3593<h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3594 <a name="i_mul">'<tt>mul</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3595</h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3596
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3597<div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3598
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3599<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3600<pre>
Dan Gohman08d012e2009-07-22 22:44:56 +0000[diff] [blame]3601 &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]3602 &lt;result&gt; = mul nuw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3603 &lt;result&gt; = mul nsw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3604 &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]3605</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3606
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3607<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3608<p>The '<tt>mul</tt>' instruction returns the product of its two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3609
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3610<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3611<p>The two arguments to the '<tt>mul</tt>' instruction must
3612 be <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of
3613 integer values. Both arguments must have identical types.</p>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3614
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3615<h5>Semantics:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3616<p>The value produced is the integer product of the two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3617
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3618<p>If the result of the multiplication has unsigned overflow, the result
3619 returned is the mathematical result modulo 2<sup>n</sup>, where n is the bit
3620 width of the result.</p>
3621
3622<p>Because LLVM integers use a two's complement representation, and the result
3623 is the same width as the operands, this instruction returns the correct
3624 result for both signed and unsigned integers. If a full product
3625 (e.g. <tt>i32</tt>x<tt>i32</tt>-><tt>i64</tt>) is needed, the operands should
3626 be sign-extended or zero-extended as appropriate to the width of the full
3627 product.</p>
3628
Dan Gohman08d012e2009-07-22 22:44:56 +0000[diff] [blame]3629<p><tt>nuw</tt> and <tt>nsw</tt> stand for &quot;No Unsigned Wrap&quot;
3630 and &quot;No Signed Wrap&quot;, respectively. If the <tt>nuw</tt> and/or
3631 <tt>nsw</tt> keywords are present, the result value of the <tt>mul</tt>
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]3632 is a <a href="#trapvalues">trap value</a> if unsigned and/or signed overflow,
3633 respectively, occurs.</p>
Dan Gohmancbb38f22009-07-20 22:41:19 +0000[diff] [blame]3634
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3635<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3636<pre>
3637 &lt;result&gt; = mul i32 4, %var <i>; yields {i32}:result = 4 * %var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3638</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3639
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3640</div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3641
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3642<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3643<h4>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3644 <a name="i_fmul">'<tt>fmul</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3645</h4>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3646
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3647<div>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3648
3649<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3650<pre>
3651 &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]3652</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3653
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3654<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3655<p>The '<tt>fmul</tt>' instruction returns the product of its two operands.</p>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3656
3657<h5>Arguments:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3658<p>The two arguments to the '<tt>fmul</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3659 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
3660 floating point values. Both arguments must have identical types.</p>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3661
3662<h5>Semantics:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3663<p>The value produced is the floating point product of the two operands.</p>
3664
3665<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3666<pre>
3667 &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]3668</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3669
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3670</div>
3671
3672<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3673<h4>
3674 <a name="i_udiv">'<tt>udiv</tt>' Instruction</a>
3675</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3676
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3677<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3678
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3679<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3680<pre>
Chris Lattner35bda892011-02-06 21:44:57 +0000[diff] [blame]3681 &lt;result&gt; = udiv &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3682 &lt;result&gt; = udiv 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]3683</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3684
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3685<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3686<p>The '<tt>udiv</tt>' instruction returns the quotient of its two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3687
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3688<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3689<p>The two arguments to the '<tt>udiv</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3690 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3691 values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3692
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3693<h5>Semantics:</h5>
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3694<p>The value produced is the unsigned integer quotient of the two operands.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3695
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3696<p>Note that unsigned integer division and signed integer division are distinct
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3697 operations; for signed integer division, use '<tt>sdiv</tt>'.</p>
3698
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3699<p>Division by zero leads to undefined behavior.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3700
Chris Lattner35bda892011-02-06 21:44:57 +0000[diff] [blame]3701<p>If the <tt>exact</tt> keyword is present, the result value of the
3702 <tt>udiv</tt> is a <a href="#trapvalues">trap value</a> if %op1 is not a
3703 multiple of %op2 (as such, "((a udiv exact b) mul b) == a").</p>
3704
3705
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3706<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3707<pre>
3708 &lt;result&gt; = udiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3709</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3710
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3711</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3712
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3713<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3714<h4>
3715 <a name="i_sdiv">'<tt>sdiv</tt>' Instruction</a>
3716</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3717
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3718<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3719
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3720<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3721<pre>
Dan Gohmancbb38f22009-07-20 22:41:19 +0000[diff] [blame]3722 &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]3723 &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]3724</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3725
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3726<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3727<p>The '<tt>sdiv</tt>' instruction returns the quotient of its two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3728
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3729<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3730<p>The two arguments to the '<tt>sdiv</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3731 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3732 values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3733
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3734<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3735<p>The value produced is the signed integer quotient of the two operands rounded
3736 towards zero.</p>
3737
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3738<p>Note that signed integer division and unsigned integer division are distinct
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3739 operations; for unsigned integer division, use '<tt>udiv</tt>'.</p>
3740
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3741<p>Division by zero leads to undefined behavior. Overflow also leads to
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3742 undefined behavior; this is a rare case, but can occur, for example, by doing
3743 a 32-bit division of -2147483648 by -1.</p>
3744
Dan Gohman9c5beed2009-07-22 00:04:19 +0000[diff] [blame]3745<p>If the <tt>exact</tt> keyword is present, the result value of the
Dan Gohman27ef9972010-04-23 15:23:32 +0000[diff] [blame]3746 <tt>sdiv</tt> is a <a href="#trapvalues">trap value</a> if the result would
Dan Gohman38da9272010-07-11 00:08:34 +0000[diff] [blame]3747 be rounded.</p>
Dan Gohmancbb38f22009-07-20 22:41:19 +0000[diff] [blame]3748
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3749<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3750<pre>
3751 &lt;result&gt; = sdiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3752</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3753
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3754</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3755
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3756<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3757<h4>
3758 <a name="i_fdiv">'<tt>fdiv</tt>' Instruction</a>
3759</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3760
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3761<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3762
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3763<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3764<pre>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]3765 &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]3766</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3767
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3768<h5>Overview:</h5>
3769<p>The '<tt>fdiv</tt>' instruction returns the quotient of its two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3770
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3771<h5>Arguments:</h5>
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]3772<p>The two arguments to the '<tt>fdiv</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3773 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
3774 floating point values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3775
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3776<h5>Semantics:</h5>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3777<p>The value produced is the floating point quotient of the two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3778
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3779<h5>Example:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3780<pre>
3781 &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]3782</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3783
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3784</div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3785
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3786<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3787<h4>
3788 <a name="i_urem">'<tt>urem</tt>' Instruction</a>
3789</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3790
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3791<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3792
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3793<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3794<pre>
3795 &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]3796</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3797
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3798<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3799<p>The '<tt>urem</tt>' instruction returns the remainder from the unsigned
3800 division of its two arguments.</p>
3801
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3802<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3803<p>The two arguments to the '<tt>urem</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3804 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3805 values. Both arguments must have identical types.</p>
3806
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3807<h5>Semantics:</h5>
3808<p>This instruction returns the unsigned integer <i>remainder</i> of a division.
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3809 This instruction always performs an unsigned division to get the
3810 remainder.</p>
3811
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3812<p>Note that unsigned integer remainder and signed integer remainder are
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3813 distinct operations; for signed integer remainder, use '<tt>srem</tt>'.</p>
3814
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3815<p>Taking the remainder of a division by zero leads to undefined behavior.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3816
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3817<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3818<pre>
3819 &lt;result&gt; = urem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3820</pre>
3821
3822</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3823
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3824<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3825<h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3826 <a name="i_srem">'<tt>srem</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3827</h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3828
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3829<div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3830
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3831<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3832<pre>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]3833 &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]3834</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3835
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3836<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3837<p>The '<tt>srem</tt>' instruction returns the remainder from the signed
3838 division of its two operands. This instruction can also take
3839 <a href="#t_vector">vector</a> versions of the values in which case the
3840 elements must be integers.</p>
Chris Lattnerc7d3ab32008-01-04 04:33:49 +0000[diff] [blame]3841
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3842<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3843<p>The two arguments to the '<tt>srem</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3844 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3845 values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3846
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3847<h5>Semantics:</h5>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3848<p>This instruction returns the <i>remainder</i> of a division (where the result
Duncan Sandsdea3a5e2011-03-07 09:12:24 +0000[diff] [blame]3849 is either zero or has the same sign as the dividend, <tt>op1</tt>), not the
3850 <i>modulo</i> operator (where the result is either zero or has the same sign
3851 as the divisor, <tt>op2</tt>) of a value.
3852 For more information about the difference,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3853 see <a href="http://mathforum.org/dr.math/problems/anne.4.28.99.html">The
3854 Math Forum</a>. For a table of how this is implemented in various languages,
3855 please see <a href="http://en.wikipedia.org/wiki/Modulo_operation">
3856 Wikipedia: modulo operation</a>.</p>
3857
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3858<p>Note that signed integer remainder and unsigned integer remainder are
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3859 distinct operations; for unsigned integer remainder, use '<tt>urem</tt>'.</p>
3860
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3861<p>Taking the remainder of a division by zero leads to undefined behavior.
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3862 Overflow also leads to undefined behavior; this is a rare case, but can
3863 occur, for example, by taking the remainder of a 32-bit division of
3864 -2147483648 by -1. (The remainder doesn't actually overflow, but this rule
3865 lets srem be implemented using instructions that return both the result of
3866 the division and the remainder.)</p>
3867
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3868<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3869<pre>
3870 &lt;result&gt; = srem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3871</pre>
3872
3873</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3874
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3875<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3876<h4>
3877 <a name="i_frem">'<tt>frem</tt>' Instruction</a>
3878</h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3879
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3880<div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3881
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3882<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3883<pre>
3884 &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]3885</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3886
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3887<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3888<p>The '<tt>frem</tt>' instruction returns the remainder from the division of
3889 its two operands.</p>
3890
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3891<h5>Arguments:</h5>
3892<p>The two arguments to the '<tt>frem</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3893 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
3894 floating point values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3895
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3896<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3897<p>This instruction returns the <i>remainder</i> of a division. The remainder
3898 has the same sign as the dividend.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3899
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3900<h5>Example:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3901<pre>
3902 &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]3903</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3904
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3905</div>
Robert Bocchino7b81c752006-02-17 21:18:08 +0000[diff] [blame]3906
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3907</div>
3908
Reid Spencer8e11bf82007-02-02 13:57:07 +0000[diff] [blame]3909<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3910<h3>
3911 <a name="bitwiseops">Bitwise Binary Operations</a>
3912</h3>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3913
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3914<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3915
3916<p>Bitwise binary operators are used to do various forms of bit-twiddling in a
3917 program. They are generally very efficient instructions and can commonly be
3918 strength reduced from other instructions. They require two operands of the
3919 same type, execute an operation on them, and produce a single value. The
3920 resulting value is the same type as its operands.</p>
3921
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3922<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3923<h4>
3924 <a name="i_shl">'<tt>shl</tt>' Instruction</a>
3925</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3926
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3927<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3928
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3929<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3930<pre>
Chris Lattnerf067d582011-02-07 16:40:21 +0000[diff] [blame]3931 &lt;result&gt; = shl &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3932 &lt;result&gt; = shl nuw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3933 &lt;result&gt; = shl nsw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3934 &lt;result&gt; = shl nuw nsw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3935</pre>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]3936
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3937<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3938<p>The '<tt>shl</tt>' instruction returns the first operand shifted to the left
3939 a specified number of bits.</p>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]3940
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3941<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3942<p>Both arguments to the '<tt>shl</tt>' instruction must be the
3943 same <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of
3944 integer type. '<tt>op2</tt>' is treated as an unsigned value.</p>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3945
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3946<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3947<p>The value produced is <tt>op1</tt> * 2<sup><tt>op2</tt></sup> mod
3948 2<sup>n</sup>, where <tt>n</tt> is the width of the result. If <tt>op2</tt>
3949 is (statically or dynamically) negative or equal to or larger than the number
3950 of bits in <tt>op1</tt>, the result is undefined. If the arguments are
3951 vectors, each vector element of <tt>op1</tt> is shifted by the corresponding
3952 shift amount in <tt>op2</tt>.</p>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]3953
Chris Lattnerf067d582011-02-07 16:40:21 +0000[diff] [blame]3954<p>If the <tt>nuw</tt> keyword is present, then the shift produces a
3955 <a href="#trapvalues">trap value</a> if it shifts out any non-zero bits. If
Chris Lattner66298c12011-02-09 16:44:44 +0000[diff] [blame]3956 the <tt>nsw</tt> keyword is present, then the shift produces a
Chris Lattnerf067d582011-02-07 16:40:21 +0000[diff] [blame]3957 <a href="#trapvalues">trap value</a> if it shifts out any bits that disagree
3958 with the resultant sign bit. As such, NUW/NSW have the same semantics as
3959 they would if the shift were expressed as a mul instruction with the same
3960 nsw/nuw bits in (mul %op1, (shl 1, %op2)).</p>
3961
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3962<h5>Example:</h5>
3963<pre>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3964 &lt;result&gt; = shl i32 4, %var <i>; yields {i32}: 4 &lt;&lt; %var</i>
3965 &lt;result&gt; = shl i32 4, 2 <i>; yields {i32}: 16</i>
3966 &lt;result&gt; = shl i32 1, 10 <i>; yields {i32}: 1024</i>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]3967 &lt;result&gt; = shl i32 1, 32 <i>; undefined</i>
Mon P Wange9f10152008-12-09 05:46:39 +0000[diff] [blame]3968 &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]3969</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3970
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3971</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3972
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3973<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3974<h4>
3975 <a name="i_lshr">'<tt>lshr</tt>' Instruction</a>
3976</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3977
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3978<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3979
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3980<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3981<pre>
Chris Lattnerf067d582011-02-07 16:40:21 +0000[diff] [blame]3982 &lt;result&gt; = lshr &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3983 &lt;result&gt; = lshr exact &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3984</pre>
3985
3986<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3987<p>The '<tt>lshr</tt>' instruction (logical shift right) returns the first
3988 operand shifted to the right a specified number of bits with zero fill.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3989
3990<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3991<p>Both arguments to the '<tt>lshr</tt>' instruction must be the same
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3992 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3993 type. '<tt>op2</tt>' is treated as an unsigned value.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3994
3995<h5>Semantics:</h5>
3996<p>This instruction always performs a logical shift right operation. The most
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3997 significant bits of the result will be filled with zero bits after the shift.
3998 If <tt>op2</tt> is (statically or dynamically) equal to or larger than the
3999 number of bits in <tt>op1</tt>, the result is undefined. If the arguments are
4000 vectors, each vector element of <tt>op1</tt> is shifted by the corresponding
4001 shift amount in <tt>op2</tt>.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]4002
Chris Lattnerf067d582011-02-07 16:40:21 +0000[diff] [blame]4003<p>If the <tt>exact</tt> keyword is present, the result value of the
4004 <tt>lshr</tt> is a <a href="#trapvalues">trap value</a> if any of the bits
4005 shifted out are non-zero.</p>
4006
4007
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]4008<h5>Example:</h5>
4009<pre>
4010 &lt;result&gt; = lshr i32 4, 1 <i>; yields {i32}:result = 2</i>
4011 &lt;result&gt; = lshr i32 4, 2 <i>; yields {i32}:result = 1</i>
4012 &lt;result&gt; = lshr i8 4, 3 <i>; yields {i8}:result = 0</i>
4013 &lt;result&gt; = lshr i8 -2, 1 <i>; yields {i8}:result = 0x7FFFFFFF </i>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]4014 &lt;result&gt; = lshr i32 1, 32 <i>; undefined</i>
Mon P Wange9f10152008-12-09 05:46:39 +0000[diff] [blame]4015 &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]4016</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4017
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]4018</div>
4019
Reid Spencer8e11bf82007-02-02 13:57:07 +0000[diff] [blame]4020<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4021<h4>
4022 <a name="i_ashr">'<tt>ashr</tt>' Instruction</a>
4023</h4>
4024
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4025<div>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]4026
4027<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4028<pre>
Chris Lattnerf067d582011-02-07 16:40:21 +0000[diff] [blame]4029 &lt;result&gt; = ashr &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
4030 &lt;result&gt; = ashr exact &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]4031</pre>
4032
4033<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4034<p>The '<tt>ashr</tt>' instruction (arithmetic shift right) returns the first
4035 operand shifted to the right a specified number of bits with sign
4036 extension.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]4037
4038<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4039<p>Both arguments to the '<tt>ashr</tt>' instruction must be the same
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4040 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
4041 type. '<tt>op2</tt>' is treated as an unsigned value.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]4042
4043<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4044<p>This instruction always performs an arithmetic shift right operation, The
4045 most significant bits of the result will be filled with the sign bit
4046 of <tt>op1</tt>. If <tt>op2</tt> is (statically or dynamically) equal to or
4047 larger than the number of bits in <tt>op1</tt>, the result is undefined. If
4048 the arguments are vectors, each vector element of <tt>op1</tt> is shifted by
4049 the corresponding shift amount in <tt>op2</tt>.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]4050
Chris Lattnerf067d582011-02-07 16:40:21 +0000[diff] [blame]4051<p>If the <tt>exact</tt> keyword is present, the result value of the
4052 <tt>ashr</tt> is a <a href="#trapvalues">trap value</a> if any of the bits
4053 shifted out are non-zero.</p>
4054
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]4055<h5>Example:</h5>
4056<pre>
4057 &lt;result&gt; = ashr i32 4, 1 <i>; yields {i32}:result = 2</i>
4058 &lt;result&gt; = ashr i32 4, 2 <i>; yields {i32}:result = 1</i>
4059 &lt;result&gt; = ashr i8 4, 3 <i>; yields {i8}:result = 0</i>
4060 &lt;result&gt; = ashr i8 -2, 1 <i>; yields {i8}:result = -1</i>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]4061 &lt;result&gt; = ashr i32 1, 32 <i>; undefined</i>
Mon P Wange9f10152008-12-09 05:46:39 +0000[diff] [blame]4062 &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]4063</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4064
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]4065</div>
4066
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4067<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4068<h4>
4069 <a name="i_and">'<tt>and</tt>' Instruction</a>
4070</h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4071
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4072<div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4073
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4074<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4075<pre>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4076 &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]4077</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4078
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4079<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4080<p>The '<tt>and</tt>' instruction returns the bitwise logical and of its two
4081 operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4082
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4083<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4084<p>The two arguments to the '<tt>and</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4085 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
4086 values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4087
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4088<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4089<p>The truth table used for the '<tt>and</tt>' instruction is:</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4090
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4091<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4092 <tbody>
4093 <tr>
4094 <td>In0</td>
4095 <td>In1</td>
4096 <td>Out</td>
4097 </tr>
4098 <tr>
4099 <td>0</td>
4100 <td>0</td>
4101 <td>0</td>
4102 </tr>
4103 <tr>
4104 <td>0</td>
4105 <td>1</td>
4106 <td>0</td>
4107 </tr>
4108 <tr>
4109 <td>1</td>
4110 <td>0</td>
4111 <td>0</td>
4112 </tr>
4113 <tr>
4114 <td>1</td>
4115 <td>1</td>
4116 <td>1</td>
4117 </tr>
4118 </tbody>
4119</table>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4120
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4121<h5>Example:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4122<pre>
4123 &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]4124 &lt;result&gt; = and i32 15, 40 <i>; yields {i32}:result = 8</i>
4125 &lt;result&gt; = and i32 4, 8 <i>; yields {i32}:result = 0</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4126</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4127</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4128<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4129<h4>
4130 <a name="i_or">'<tt>or</tt>' Instruction</a>
4131</h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4132
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4133<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4134
4135<h5>Syntax:</h5>
4136<pre>
4137 &lt;result&gt; = or &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
4138</pre>
4139
4140<h5>Overview:</h5>
4141<p>The '<tt>or</tt>' instruction returns the bitwise logical inclusive or of its
4142 two operands.</p>
4143
4144<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4145<p>The two arguments to the '<tt>or</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4146 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
4147 values. Both arguments must have identical types.</p>
4148
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4149<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4150<p>The truth table used for the '<tt>or</tt>' instruction is:</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4151
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4152<table border="1" cellspacing="0" cellpadding="4">
4153 <tbody>
4154 <tr>
4155 <td>In0</td>
4156 <td>In1</td>
4157 <td>Out</td>
4158 </tr>
4159 <tr>
4160 <td>0</td>
4161 <td>0</td>
4162 <td>0</td>
4163 </tr>
4164 <tr>
4165 <td>0</td>
4166 <td>1</td>
4167 <td>1</td>
4168 </tr>
4169 <tr>
4170 <td>1</td>
4171 <td>0</td>
4172 <td>1</td>
4173 </tr>
4174 <tr>
4175 <td>1</td>
4176 <td>1</td>
4177 <td>1</td>
4178 </tr>
4179 </tbody>
4180</table>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4181
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4182<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4183<pre>
4184 &lt;result&gt; = or i32 4, %var <i>; yields {i32}:result = 4 | %var</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4185 &lt;result&gt; = or i32 15, 40 <i>; yields {i32}:result = 47</i>
4186 &lt;result&gt; = or i32 4, 8 <i>; yields {i32}:result = 12</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4187</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4188
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4189</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4190
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4191<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4192<h4>
4193 <a name="i_xor">'<tt>xor</tt>' Instruction</a>
4194</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4195
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4196<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4197
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4198<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4199<pre>
4200 &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]4201</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4202
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4203<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4204<p>The '<tt>xor</tt>' instruction returns the bitwise logical exclusive or of
4205 its two operands. The <tt>xor</tt> is used to implement the "one's
4206 complement" operation, which is the "~" operator in C.</p>
4207
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4208<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4209<p>The two arguments to the '<tt>xor</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4210 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
4211 values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4212
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4213<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4214<p>The truth table used for the '<tt>xor</tt>' instruction is:</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4215
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4216<table border="1" cellspacing="0" cellpadding="4">
4217 <tbody>
4218 <tr>
4219 <td>In0</td>
4220 <td>In1</td>
4221 <td>Out</td>
4222 </tr>
4223 <tr>
4224 <td>0</td>
4225 <td>0</td>
4226 <td>0</td>
4227 </tr>
4228 <tr>
4229 <td>0</td>
4230 <td>1</td>
4231 <td>1</td>
4232 </tr>
4233 <tr>
4234 <td>1</td>
4235 <td>0</td>
4236 <td>1</td>
4237 </tr>
4238 <tr>
4239 <td>1</td>
4240 <td>1</td>
4241 <td>0</td>
4242 </tr>
4243 </tbody>
4244</table>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4245
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4246<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4247<pre>
4248 &lt;result&gt; = xor i32 4, %var <i>; yields {i32}:result = 4 ^ %var</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4249 &lt;result&gt; = xor i32 15, 40 <i>; yields {i32}:result = 39</i>
4250 &lt;result&gt; = xor i32 4, 8 <i>; yields {i32}:result = 12</i>
4251 &lt;result&gt; = xor i32 %V, -1 <i>; yields {i32}:result = ~%V</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4252</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4253
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4254</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4255
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4256</div>
4257
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4258<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4259<h3>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4260 <a name="vectorops">Vector Operations</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4261</h3>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4262
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4263<div>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4264
4265<p>LLVM supports several instructions to represent vector operations in a
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4266 target-independent manner. These instructions cover the element-access and
4267 vector-specific operations needed to process vectors effectively. While LLVM
4268 does directly support these vector operations, many sophisticated algorithms
4269 will want to use target-specific intrinsics to take full advantage of a
4270 specific target.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4271
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4272<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4273<h4>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4274 <a name="i_extractelement">'<tt>extractelement</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4275</h4>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4276
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4277<div>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4278
4279<h5>Syntax:</h5>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4280<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4281 &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]4282</pre>
4283
4284<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4285<p>The '<tt>extractelement</tt>' instruction extracts a single scalar element
4286 from a vector at a specified index.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4287
4288
4289<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4290<p>The first operand of an '<tt>extractelement</tt>' instruction is a value
4291 of <a href="#t_vector">vector</a> type. The second operand is an index
4292 indicating the position from which to extract the element. The index may be
4293 a variable.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4294
4295<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4296<p>The result is a scalar of the same type as the element type of
4297 <tt>val</tt>. Its value is the value at position <tt>idx</tt> of
4298 <tt>val</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
4299 results are undefined.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4300
4301<h5>Example:</h5>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4302<pre>
Gabor Greifa5b6f452009-10-28 13:14:50 +0000[diff] [blame]4303 &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]4304</pre>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4305
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4306</div>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4307
4308<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4309<h4>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4310 <a name="i_insertelement">'<tt>insertelement</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4311</h4>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4312
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4313<div>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4314
4315<h5>Syntax:</h5>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4316<pre>
Dan Gohmanf3480b92008-05-12 23:38:42 +0000[diff] [blame]4317 &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]4318</pre>
4319
4320<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4321<p>The '<tt>insertelement</tt>' instruction inserts a scalar element into a
4322 vector at a specified index.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4323
4324<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4325<p>The first operand of an '<tt>insertelement</tt>' instruction is a value
4326 of <a href="#t_vector">vector</a> type. The second operand is a scalar value
4327 whose type must equal the element type of the first operand. The third
4328 operand is an index indicating the position at which to insert the value.
4329 The index may be a variable.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4330
4331<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4332<p>The result is a vector of the same type as <tt>val</tt>. Its element values
4333 are those of <tt>val</tt> except at position <tt>idx</tt>, where it gets the
4334 value <tt>elt</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
4335 results are undefined.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4336
4337<h5>Example:</h5>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4338<pre>
Gabor Greifa5b6f452009-10-28 13:14:50 +0000[diff] [blame]4339 &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]4340</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4341
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4342</div>
4343
4344<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4345<h4>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4346 <a name="i_shufflevector">'<tt>shufflevector</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4347</h4>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4348
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4349<div>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4350
4351<h5>Syntax:</h5>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4352<pre>
Mon P Wangaeb06d22008-11-10 04:46:22 +0000[diff] [blame]4353 &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]4354</pre>
4355
4356<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4357<p>The '<tt>shufflevector</tt>' instruction constructs a permutation of elements
4358 from two input vectors, returning a vector with the same element type as the
4359 input and length that is the same as the shuffle mask.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4360
4361<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4362<p>The first two operands of a '<tt>shufflevector</tt>' instruction are vectors
4363 with types that match each other. The third argument is a shuffle mask whose
4364 element type is always 'i32'. The result of the instruction is a vector
4365 whose length is the same as the shuffle mask and whose element type is the
4366 same as the element type of the first two operands.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4367
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4368<p>The shuffle mask operand is required to be a constant vector with either
4369 constant integer or undef values.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4370
4371<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4372<p>The elements of the two input vectors are numbered from left to right across
4373 both of the vectors. The shuffle mask operand specifies, for each element of
4374 the result vector, which element of the two input vectors the result element
4375 gets. The element selector may be undef (meaning "don't care") and the
4376 second operand may be undef if performing a shuffle from only one vector.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4377
4378<h5>Example:</h5>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4379<pre>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4380 &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]4381 &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]4382 &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]4383 &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]4384 &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]4385 &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]4386 &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]4387 &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]4388</pre>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4389
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4390</div>
Tanya Lattner09474292006-04-14 19:24:33 +0000[diff] [blame]4391
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4392</div>
4393
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4394<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4395<h3>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4396 <a name="aggregateops">Aggregate Operations</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4397</h3>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4398
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4399<div>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4400
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4401<p>LLVM supports several instructions for working with
4402 <a href="#t_aggregate">aggregate</a> values.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4403
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4404<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4405<h4>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4406 <a name="i_extractvalue">'<tt>extractvalue</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4407</h4>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4408
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4409<div>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4410
4411<h5>Syntax:</h5>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4412<pre>
4413 &lt;result&gt; = extractvalue &lt;aggregate type&gt; &lt;val&gt;, &lt;idx&gt;{, &lt;idx&gt;}*
4414</pre>
4415
4416<h5>Overview:</h5>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4417<p>The '<tt>extractvalue</tt>' instruction extracts the value of a member field
4418 from an <a href="#t_aggregate">aggregate</a> value.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4419
4420<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4421<p>The first operand of an '<tt>extractvalue</tt>' instruction is a value
Chris Lattner61c70e92010-08-28 04:09:24 +0000[diff] [blame]4422 of <a href="#t_struct">struct</a> or
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4423 <a href="#t_array">array</a> type. The operands are constant indices to
4424 specify which value to extract in a similar manner as indices in a
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4425 '<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.</p>
Frits van Bommel13242892010-12-05 20:54:38 +0000[diff] [blame]4426 <p>The major differences to <tt>getelementptr</tt> indexing are:</p>
4427 <ul>
4428 <li>Since the value being indexed is not a pointer, the first index is
4429 omitted and assumed to be zero.</li>
4430 <li>At least one index must be specified.</li>
4431 <li>Not only struct indices but also array indices must be in
4432 bounds.</li>
4433 </ul>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4434
4435<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4436<p>The result is the value at the position in the aggregate specified by the
4437 index operands.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4438
4439<h5>Example:</h5>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4440<pre>
Gabor Greifa5b6f452009-10-28 13:14:50 +0000[diff] [blame]4441 &lt;result&gt; = extractvalue {i32, float} %agg, 0 <i>; yields i32</i>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4442</pre>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4443
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4444</div>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4445
4446<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4447<h4>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4448 <a name="i_insertvalue">'<tt>insertvalue</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4449</h4>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4450
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4451<div>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4452
4453<h5>Syntax:</h5>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4454<pre>
Bill Wendling194229e2011-07-26 20:42:28 +0000[diff] [blame]4455 &lt;result&gt; = insertvalue &lt;aggregate type&gt; &lt;val&gt;, &lt;ty&gt; &lt;elt&gt;, &lt;idx&gt;{, &lt;idx&gt;}* <i>; yields &lt;aggregate type&gt;</i>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4456</pre>
4457
4458<h5>Overview:</h5>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4459<p>The '<tt>insertvalue</tt>' instruction inserts a value into a member field
4460 in an <a href="#t_aggregate">aggregate</a> value.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4461
4462<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4463<p>The first operand of an '<tt>insertvalue</tt>' instruction is a value
Chris Lattner61c70e92010-08-28 04:09:24 +0000[diff] [blame]4464 of <a href="#t_struct">struct</a> or
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4465 <a href="#t_array">array</a> type. The second operand is a first-class
4466 value to insert. The following operands are constant indices indicating
4467 the position at which to insert the value in a similar manner as indices in a
Frits van Bommel13242892010-12-05 20:54:38 +0000[diff] [blame]4468 '<tt><a href="#i_extractvalue">extractvalue</a></tt>' instruction. The
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4469 value to insert must have the same type as the value identified by the
4470 indices.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4471
4472<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4473<p>The result is an aggregate of the same type as <tt>val</tt>. Its value is
4474 that of <tt>val</tt> except that the value at the position specified by the
4475 indices is that of <tt>elt</tt>.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4476
4477<h5>Example:</h5>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4478<pre>
Chris Lattner8645d1a2011-05-22 07:18:08 +0000[diff] [blame]4479 %agg1 = insertvalue {i32, float} undef, i32 1, 0 <i>; yields {i32 1, float undef}</i>
4480 %agg2 = insertvalue {i32, float} %agg1, float %val, 1 <i>; yields {i32 1, float %val}</i>
4481 %agg3 = insertvalue {i32, {float}} %agg1, float %val, 1, 0 <i>; yields {i32 1, float %val}</i>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4482</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4483
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4484</div>
4485
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4486</div>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4487
4488<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4489<h3>
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]4490 <a name="memoryops">Memory Access and Addressing Operations</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4491</h3>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4492
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4493<div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4494
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4495<p>A key design point of an SSA-based representation is how it represents
4496 memory. In LLVM, no memory locations are in SSA form, which makes things
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]4497 very simple. This section describes how to read, write, and allocate
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4498 memory in LLVM.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4499
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4500<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4501<h4>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4502 <a name="i_alloca">'<tt>alloca</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4503</h4>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4504
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4505<div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4506
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4507<h5>Syntax:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4508<pre>
Dan Gohmanf75a7d32010-05-28 01:14:11 +0000[diff] [blame]4509 &lt;result&gt; = alloca &lt;type&gt;[, &lt;ty&gt; &lt;NumElements&gt;][, align &lt;alignment&gt;] <i>; yields {type*}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4510</pre>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4511
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4512<h5>Overview:</h5>
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]4513<p>The '<tt>alloca</tt>' instruction allocates memory on the stack frame of the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4514 currently executing function, to be automatically released when this function
4515 returns to its caller. The object is always allocated in the generic address
4516 space (address space zero).</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4517
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4518<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4519<p>The '<tt>alloca</tt>' instruction
4520 allocates <tt>sizeof(&lt;type&gt;)*NumElements</tt> bytes of memory on the
4521 runtime stack, returning a pointer of the appropriate type to the program.
4522 If "NumElements" is specified, it is the number of elements allocated,
4523 otherwise "NumElements" is defaulted to be one. If a constant alignment is
4524 specified, the value result of the allocation is guaranteed to be aligned to
4525 at least that boundary. If not specified, or if zero, the target can choose
4526 to align the allocation on any convenient boundary compatible with the
4527 type.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4528
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4529<p>'<tt>type</tt>' may be any sized type.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4530
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4531<h5>Semantics:</h5>
Bill Wendling871eb0a2009-05-08 20:49:29 +0000[diff] [blame]4532<p>Memory is allocated; a pointer is returned. The operation is undefined if
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4533 there is insufficient stack space for the allocation. '<tt>alloca</tt>'d
4534 memory is automatically released when the function returns. The
4535 '<tt>alloca</tt>' instruction is commonly used to represent automatic
4536 variables that must have an address available. When the function returns
4537 (either with the <tt><a href="#i_ret">ret</a></tt>
4538 or <tt><a href="#i_unwind">unwind</a></tt> instructions), the memory is
4539 reclaimed. Allocating zero bytes is legal, but the result is undefined.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4540
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4541<h5>Example:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4542<pre>
Dan Gohman81e21672009-01-04 23:49:44 +0000[diff] [blame]4543 %ptr = alloca i32 <i>; yields {i32*}:ptr</i>
4544 %ptr = alloca i32, i32 4 <i>; yields {i32*}:ptr</i>
4545 %ptr = alloca i32, i32 4, align 1024 <i>; yields {i32*}:ptr</i>
4546 %ptr = alloca i32, align 1024 <i>; yields {i32*}:ptr</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4547</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4548
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4549</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4550
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4551<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4552<h4>
4553 <a name="i_load">'<tt>load</tt>' Instruction</a>
4554</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4555
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4556<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4557
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4558<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4559<pre>
Bill Wendling7c78dbb2010-02-25 21:23:24 +0000[diff] [blame]4560 &lt;result&gt; = load &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;][, !nontemporal !&lt;index&gt;]
4561 &lt;result&gt; = volatile load &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;][, !nontemporal !&lt;index&gt;]
4562 !&lt;index&gt; = !{ i32 1 }
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4563</pre>
4564
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4565<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4566<p>The '<tt>load</tt>' instruction is used to read from memory.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4567
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4568<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4569<p>The argument to the '<tt>load</tt>' instruction specifies the memory address
4570 from which to load. The pointer must point to
4571 a <a href="#t_firstclass">first class</a> type. If the <tt>load</tt> is
4572 marked as <tt>volatile</tt>, then the optimizer is not allowed to modify the
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]4573 number or order of execution of this <tt>load</tt> with other <a
4574 href="#volatile">volatile operations</a>.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4575
Bill Wendling7c78dbb2010-02-25 21:23:24 +0000[diff] [blame]4576<p>The optional constant <tt>align</tt> argument specifies the alignment of the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4577 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]4578 omitted <tt>align</tt> argument means that the operation has the preferential
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4579 alignment for the target. It is the responsibility of the code emitter to
4580 ensure that the alignment information is correct. Overestimating the
Bill Wendling7c78dbb2010-02-25 21:23:24 +0000[diff] [blame]4581 alignment results in undefined behavior. Underestimating the alignment may
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4582 produce less efficient code. An alignment of 1 is always safe.</p>
4583
Bill Wendling7c78dbb2010-02-25 21:23:24 +0000[diff] [blame]4584<p>The optional <tt>!nontemporal</tt> metadata must reference a single
4585 metatadata name &lt;index&gt; corresponding to a metadata node with
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]4586 one <tt>i32</tt> entry of value 1. The existence of
Bill Wendling7c78dbb2010-02-25 21:23:24 +0000[diff] [blame]4587 the <tt>!nontemporal</tt> metatadata on the instruction tells the optimizer
4588 and code generator that this load is not expected to be reused in the cache.
4589 The code generator may select special instructions to save cache bandwidth,
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]4590 such as the <tt>MOVNT</tt> instruction on x86.</p>
David Greene8939b0d2010-02-16 20:50:18 +0000[diff] [blame]4591
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4592<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4593<p>The location of memory pointed to is loaded. If the value being loaded is of
4594 scalar type then the number of bytes read does not exceed the minimum number
4595 of bytes needed to hold all bits of the type. For example, loading an
4596 <tt>i24</tt> reads at most three bytes. When loading a value of a type like
4597 <tt>i20</tt> with a size that is not an integral number of bytes, the result
4598 is undefined if the value was not originally written using a store of the
4599 same type.</p>
4600
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4601<h5>Examples:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4602<pre>
4603 %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
4604 <a href="#i_store">store</a> i32 3, i32* %ptr <i>; yields {void}</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4605 %val = load i32* %ptr <i>; yields {i32}:val = i32 3</i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4606</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4607
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4608</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4609
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4610<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4611<h4>
4612 <a name="i_store">'<tt>store</tt>' Instruction</a>
4613</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4614
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4615<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4616
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4617<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4618<pre>
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]4619 store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;][, !nontemporal !&lt;index&gt;] <i>; yields {void}</i>
4620 volatile store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;][, !nontemporal !&lt;index&gt;] <i>; yields {void}</i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4621</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4622
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4623<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4624<p>The '<tt>store</tt>' instruction is used to write to memory.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4625
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4626<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4627<p>There are two arguments to the '<tt>store</tt>' instruction: a value to store
4628 and an address at which to store it. The type of the
4629 '<tt>&lt;pointer&gt;</tt>' operand must be a pointer to
4630 the <a href="#t_firstclass">first class</a> type of the
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]4631 '<tt>&lt;value&gt;</tt>' operand. If the <tt>store</tt> is marked as
4632 <tt>volatile</tt>, then the optimizer is not allowed to modify the number or
4633 order of execution of this <tt>store</tt> with other <a
4634 href="#volatile">volatile operations</a>.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4635
4636<p>The optional constant "align" argument specifies the alignment of the
4637 operation (that is, the alignment of the memory address). A value of 0 or an
4638 omitted "align" argument means that the operation has the preferential
4639 alignment for the target. It is the responsibility of the code emitter to
4640 ensure that the alignment information is correct. Overestimating the
4641 alignment results in an undefined behavior. Underestimating the alignment may
4642 produce less efficient code. An alignment of 1 is always safe.</p>
4643
David Greene8939b0d2010-02-16 20:50:18 +0000[diff] [blame]4644<p>The optional !nontemporal metadata must reference a single metatadata
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]4645 name &lt;index&gt; corresponding to a metadata node with one i32 entry of
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]4646 value 1. The existence of the !nontemporal metatadata on the
David Greene8939b0d2010-02-16 20:50:18 +0000[diff] [blame]4647 instruction tells the optimizer and code generator that this load is
4648 not expected to be reused in the cache. The code generator may
4649 select special instructions to save cache bandwidth, such as the
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]4650 MOVNT instruction on x86.</p>
David Greene8939b0d2010-02-16 20:50:18 +0000[diff] [blame]4651
4652
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4653<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4654<p>The contents of memory are updated to contain '<tt>&lt;value&gt;</tt>' at the
4655 location specified by the '<tt>&lt;pointer&gt;</tt>' operand. If
4656 '<tt>&lt;value&gt;</tt>' is of scalar type then the number of bytes written
4657 does not exceed the minimum number of bytes needed to hold all bits of the
4658 type. For example, storing an <tt>i24</tt> writes at most three bytes. When
4659 writing a value of a type like <tt>i20</tt> with a size that is not an
4660 integral number of bytes, it is unspecified what happens to the extra bits
4661 that do not belong to the type, but they will typically be overwritten.</p>
4662
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4663<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4664<pre>
4665 %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
Bill Wendling8c6c72d2007-10-22 05:10:05 +0000[diff] [blame]4666 store i32 3, i32* %ptr <i>; yields {void}</i>
4667 %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]4668</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4669
Reid Spencer47ce1792006-11-09 21:15:49 +0000[diff] [blame]4670</div>
4671
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4672<!-- _______________________________________________________________________ -->
Eli Friedman47f35132011-07-25 23:16:38 +0000[diff] [blame]4673<div class="doc_subsubsection"> <a name="i_fence">'<tt>fence</tt>'
4674Instruction</a> </div>
4675
4676<div class="doc_text">
4677
4678<h5>Syntax:</h5>
4679<pre>
4680 fence [singlethread] &lt;ordering&gt; <i>; yields {void}</i>
4681</pre>
4682
4683<h5>Overview:</h5>
4684<p>The '<tt>fence</tt>' instruction is used to introduce happens-before edges
4685between operations.</p>
4686
4687<h5>Arguments:</h5> <p>'<code>fence</code>' instructions take an <a
4688href="#ordering">ordering</a> argument which defines what
4689<i>synchronizes-with</i> edges they add. They can only be given
4690<code>acquire</code>, <code>release</code>, <code>acq_rel</code>, and
4691<code>seq_cst</code> orderings.</p>
4692
4693<h5>Semantics:</h5>
4694<p>A fence <var>A</var> which has (at least) <code>release</code> ordering
4695semantics <i>synchronizes with</i> a fence <var>B</var> with (at least)
4696<code>acquire</code> ordering semantics if and only if there exist atomic
4697operations <var>X</var> and <var>Y</var>, both operating on some atomic object
4698<var>M</var>, such that <var>A</var> is sequenced before <var>X</var>,
4699<var>X</var> modifies <var>M</var> (either directly or through some side effect
4700of a sequence headed by <var>X</var>), <var>Y</var> is sequenced before
4701<var>B</var>, and <var>Y</var> observes <var>M</var>. This provides a
4702<i>happens-before</i> dependency between <var>A</var> and <var>B</var>. Rather
4703than an explicit <code>fence</code>, one (but not both) of the atomic operations
4704<var>X</var> or <var>Y</var> might provide a <code>release</code> or
4705<code>acquire</code> (resp.) ordering constraint and still
4706<i>synchronize-with</i> the explicit <code>fence</code> and establish the
4707<i>happens-before</i> edge.</p>
4708
4709<p>A <code>fence</code> which has <code>seq_cst</code> ordering, in addition to
4710having both <code>acquire</code> and <code>release</code> semantics specified
4711above, participates in the global program order of other <code>seq_cst</code>
4712operations and/or fences.</p>
4713
4714<p>The optional "<a href="#singlethread"><code>singlethread</code></a>" argument
4715specifies that the fence only synchronizes with other fences in the same
4716thread. (This is useful for interacting with signal handlers.)</p>
4717
4718<p>FIXME: This instruction is a work in progress; until it is finished, use
4719 llvm.memory.barrier.
4720
4721<h5>Example:</h5>
4722<pre>
4723 fence acquire <i>; yields {void}</i>
4724 fence singlethread seq_cst <i>; yields {void}</i>
4725</pre>
4726
4727</div>
4728
4729<!-- _______________________________________________________________________ -->
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]4730<div class="doc_subsubsection"> <a name="i_cmpxchg">'<tt>cmpxchg</tt>'
4731Instruction</a> </div>
4732
4733<div class="doc_text">
4734
4735<h5>Syntax:</h5>
4736<pre>
4737 [volatile] cmpxchg &lt;ty&gt;* &lt;pointer&gt;, &lt;ty&gt; &lt;cmp&gt;, &lt;ty&gt; &lt;new&gt; [singlethread] &lt;ordering&gt; <i>; yields {ty}</i>
4738</pre>
4739
4740<h5>Overview:</h5>
4741<p>The '<tt>cmpxchg</tt>' instruction is used to atomically modify memory.
4742It loads a value in memory and compares it to a given value. If they are
4743equal, it stores a new value into the memory.</p>
4744
4745<h5>Arguments:</h5>
4746<p>There are three arguments to the '<code>cmpxchg</code>' instruction: an
4747address to operate on, a value to compare to the value currently be at that
4748address, and a new value to place at that address if the compared values are
4749equal. The type of '<var>&lt;cmp&gt;</var>' must be an integer type whose
4750bit width is a power of two greater than or equal to eight and less than
4751or equal to a target-specific size limit. '<var>&lt;cmp&gt;</var>' and
4752'<var>&lt;new&gt;</var>' must have the same type, and the type of
4753'<var>&lt;pointer&gt;</var>' must be a pointer to that type. If the
4754<code>cmpxchg</code> is marked as <code>volatile</code>, then the
4755optimizer is not allowed to modify the number or order of execution
4756of this <code>cmpxchg</code> with other <a href="#volatile">volatile
4757operations</a>.</p>
4758
4759<!-- FIXME: Extend allowed types. -->
4760
4761<p>The <a href="#ordering"><var>ordering</var></a> argument specifies how this
4762<code>cmpxchg</code> synchronizes with other atomic operations.</p>
4763
4764<p>The optional "<code>singlethread</code>" argument declares that the
4765<code>cmpxchg</code> is only atomic with respect to code (usually signal
4766handlers) running in the same thread as the <code>cmpxchg</code>. Otherwise the
4767cmpxchg is atomic with respect to all other code in the system.</p>
4768
4769<p>The pointer passed into cmpxchg must have alignment greater than or equal to
4770the size in memory of the operand.
4771
4772<h5>Semantics:</h5>
4773<p>The contents of memory at the location specified by the
4774'<tt>&lt;pointer&gt;</tt>' operand is read and compared to
4775'<tt>&lt;cmp&gt;</tt>'; if the read value is the equal,
4776'<tt>&lt;new&gt;</tt>' is written. The original value at the location
4777is returned.
4778
4779<p>A successful <code>cmpxchg</code> is a read-modify-write instruction for the
4780purpose of identifying <a href="#release_sequence">release sequences</a>. A
4781failed <code>cmpxchg</code> is equivalent to an atomic load with an ordering
4782parameter determined by dropping any <code>release</code> part of the
4783<code>cmpxchg</code>'s ordering.</p>
4784
4785<!--
4786FIXME: Is compare_exchange_weak() necessary? (Consider after we've done
4787optimization work on ARM.)
4788
4789FIXME: Is a weaker ordering constraint on failure helpful in practice?
4790-->
4791
4792<h5>Example:</h5>
4793<pre>
4794entry:
4795 %orig = atomic <a href="#i_load">load</a> i32* %ptr unordered <i>; yields {i32}</i>
4796 <a href="#i_br">br</a> label %loop
4797
4798loop:
4799 %cmp = <a href="#i_phi">phi</a> i32 [ %orig, %entry ], [%old, %loop]
4800 %squared = <a href="#i_mul">mul</a> i32 %cmp, %cmp
4801 %old = cmpxchg i32* %ptr, i32 %cmp, i32 %squared <i>; yields {i32}</i>
4802 %success = <a href="#i_icmp">icmp</a> eq i32 %cmp, %old
4803 <a href="#i_br">br</a> i1 %success, label %done, label %loop
4804
4805done:
4806 ...
4807</pre>
4808
4809</div>
4810
4811<!-- _______________________________________________________________________ -->
4812<div class="doc_subsubsection"> <a name="i_atomicrmw">'<tt>atomicrmw</tt>'
4813Instruction</a> </div>
4814
4815<div class="doc_text">
4816
4817<h5>Syntax:</h5>
4818<pre>
4819 [volatile] atomicrmw &lt;operation&gt; &lt;ty&gt;* &lt;pointer&gt;, &lt;ty&gt; &lt;value&gt; [singlethread] &lt;ordering&gt; <i>; yields {ty}</i>
4820</pre>
4821
4822<h5>Overview:</h5>
4823<p>The '<tt>atomicrmw</tt>' instruction is used to atomically modify memory.</p>
4824
4825<h5>Arguments:</h5>
4826<p>There are three arguments to the '<code>atomicrmw</code>' instruction: an
4827operation to apply, an address whose value to modify, an argument to the
4828operation. The operation must be one of the following keywords:</p>
4829<ul>
4830 <li>xchg</li>
4831 <li>add</li>
4832 <li>sub</li>
4833 <li>and</li>
4834 <li>nand</li>
4835 <li>or</li>
4836 <li>xor</li>
4837 <li>max</li>
4838 <li>min</li>
4839 <li>umax</li>
4840 <li>umin</li>
4841</ul>
4842
4843<p>The type of '<var>&lt;value&gt;</var>' must be an integer type whose
4844bit width is a power of two greater than or equal to eight and less than
4845or equal to a target-specific size limit. The type of the
4846'<code>&lt;pointer&gt;</code>' operand must be a pointer to that type.
4847If the <code>atomicrmw</code> is marked as <code>volatile</code>, then the
4848optimizer is not allowed to modify the number or order of execution of this
4849<code>atomicrmw</code> with other <a href="#volatile">volatile
4850 operations</a>.</p>
4851
4852<!-- FIXME: Extend allowed types. -->
4853
4854<h5>Semantics:</h5>
4855<p>The contents of memory at the location specified by the
4856'<tt>&lt;pointer&gt;</tt>' operand are atomically read, modified, and written
4857back. The original value at the location is returned. The modification is
4858specified by the <var>operation</var> argument:</p>
4859
4860<ul>
4861 <li>xchg: <code>*ptr = val</code></li>
4862 <li>add: <code>*ptr = *ptr + val</code></li>
4863 <li>sub: <code>*ptr = *ptr - val</code></li>
4864 <li>and: <code>*ptr = *ptr &amp; val</code></li>
4865 <li>nand: <code>*ptr = ~(*ptr &amp; val)</code></li>
4866 <li>or: <code>*ptr = *ptr | val</code></li>
4867 <li>xor: <code>*ptr = *ptr ^ val</code></li>
4868 <li>max: <code>*ptr = *ptr &gt; val ? *ptr : val</code> (using a signed comparison)</li>
4869 <li>min: <code>*ptr = *ptr &lt; val ? *ptr : val</code> (using a signed comparison)</li>
4870 <li>umax: <code>*ptr = *ptr &gt; val ? *ptr : val</code> (using an unsigned comparison)</li>
4871 <li>umin: <code>*ptr = *ptr &lt; val ? *ptr : val</code> (using an unsigned comparison)</li>
4872</ul>
4873
4874<h5>Example:</h5>
4875<pre>
4876 %old = atomicrmw add i32* %ptr, i32 1 acquire <i>; yields {i32}</i>
4877</pre>
4878
4879</div>
4880
4881<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4882<h4>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4883 <a name="i_getelementptr">'<tt>getelementptr</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4884</h4>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4885
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4886<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4887
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]4888<h5>Syntax:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4889<pre>
Matthijs Kooijmane49d0bc2008-10-13 13:44:15 +0000[diff] [blame]4890 &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]4891 &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]4892</pre>
4893
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]4894<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4895<p>The '<tt>getelementptr</tt>' instruction is used to get the address of a
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4896 subelement of an <a href="#t_aggregate">aggregate</a> data structure.
4897 It performs address calculation only and does not access memory.</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4898
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]4899<h5>Arguments:</h5>
Matthijs Kooijmane49d0bc2008-10-13 13:44:15 +0000[diff] [blame]4900<p>The first argument is always a pointer, and forms the basis of the
Chris Lattnerc8eef442009-07-29 06:44:13 +0000[diff] [blame]4901 calculation. The remaining arguments are indices that indicate which of the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4902 elements of the aggregate object are indexed. The interpretation of each
4903 index is dependent on the type being indexed into. The first index always
4904 indexes the pointer value given as the first argument, the second index
4905 indexes a value of the type pointed to (not necessarily the value directly
4906 pointed to, since the first index can be non-zero), etc. The first type
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4907 indexed into must be a pointer value, subsequent types can be arrays,
Chris Lattner61c70e92010-08-28 04:09:24 +0000[diff] [blame]4908 vectors, and structs. Note that subsequent types being indexed into
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4909 can never be pointers, since that would require loading the pointer before
4910 continuing calculation.</p>
Matthijs Kooijmane49d0bc2008-10-13 13:44:15 +0000[diff] [blame]4911
4912<p>The type of each index argument depends on the type it is indexing into.
Chris Lattner61c70e92010-08-28 04:09:24 +0000[diff] [blame]4913 When indexing into a (optionally packed) structure, only <tt>i32</tt>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4914 integer <b>constants</b> are allowed. When indexing into an array, pointer
4915 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]4916 constant.</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4917
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4918<p>For example, let's consider a C code fragment and how it gets compiled to
4919 LLVM:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4920
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]4921<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]4922struct RT {
4923 char A;
Chris Lattnercabc8462007-05-29 15:43:56 +0000[diff] [blame]4924 int B[10][20];
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]4925 char C;
4926};
4927struct ST {
Chris Lattnercabc8462007-05-29 15:43:56 +0000[diff] [blame]4928 int X;
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]4929 double Y;
4930 struct RT Z;
4931};
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4932
Chris Lattnercabc8462007-05-29 15:43:56 +0000[diff] [blame]4933int *foo(struct ST *s) {
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]4934 return &amp;s[1].Z.B[5][13];
4935}
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4936</pre>
4937
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4938<p>The LLVM code generated by the GCC frontend is:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4939
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]4940<pre class="doc_code">
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]4941%RT = <a href="#namedtypes">type</a> { i8 , [10 x [20 x i32]], i8 }
4942%ST = <a href="#namedtypes">type</a> { i32, double, %RT }
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4943
Dan Gohman4df605b2009-07-25 02:23:48 +0000[diff] [blame]4944define i32* @foo(%ST* %s) {
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]4945entry:
4946 %reg = getelementptr %ST* %s, i32 1, i32 2, i32 1, i32 5, i32 13
4947 ret i32* %reg
4948}
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4949</pre>
4950
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]4951<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4952<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]4953 type, which is a pointer, yielding a '<tt>%ST</tt>' = '<tt>{ i32, double, %RT
4954 }</tt>' type, a structure. The second index indexes into the third element
4955 of the structure, yielding a '<tt>%RT</tt>' = '<tt>{ i8 , [10 x [20 x i32]],
4956 i8 }</tt>' type, another structure. The third index indexes into the second
4957 element of the structure, yielding a '<tt>[10 x [20 x i32]]</tt>' type, an
4958 array. The two dimensions of the array are subscripted into, yielding an
4959 '<tt>i32</tt>' type. The '<tt>getelementptr</tt>' instruction returns a
4960 pointer to this element, thus computing a value of '<tt>i32*</tt>' type.</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4961
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4962<p>Note that it is perfectly legal to index partially through a structure,
4963 returning a pointer to an inner element. Because of this, the LLVM code for
4964 the given testcase is equivalent to:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4965
4966<pre>
Dan Gohman4df605b2009-07-25 02:23:48 +0000[diff] [blame]4967 define i32* @foo(%ST* %s) {
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4968 %t1 = getelementptr %ST* %s, i32 1 <i>; yields %ST*:%t1</i>
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]4969 %t2 = getelementptr %ST* %t1, i32 0, i32 2 <i>; yields %RT*:%t2</i>
4970 %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]4971 %t4 = getelementptr [10 x [20 x i32]]* %t3, i32 0, i32 5 <i>; yields [20 x i32]*:%t4</i>
4972 %t5 = getelementptr [20 x i32]* %t4, i32 0, i32 13 <i>; yields i32*:%t5</i>
4973 ret i32* %t5
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4974 }
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]4975</pre>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]4976
Dan Gohmandd8004d2009-07-27 21:53:46 +0000[diff] [blame]4977<p>If the <tt>inbounds</tt> keyword is present, the result value of the
Dan Gohman27ef9972010-04-23 15:23:32 +0000[diff] [blame]4978 <tt>getelementptr</tt> is a <a href="#trapvalues">trap value</a> if the
4979 base pointer is not an <i>in bounds</i> address of an allocated object,
4980 or if any of the addresses that would be formed by successive addition of
4981 the offsets implied by the indices to the base address with infinitely
4982 precise arithmetic are not an <i>in bounds</i> address of that allocated
4983 object. The <i>in bounds</i> addresses for an allocated object are all
4984 the addresses that point into the object, plus the address one byte past
4985 the end.</p>
Dan Gohmandd8004d2009-07-27 21:53:46 +0000[diff] [blame]4986
4987<p>If the <tt>inbounds</tt> keyword is not present, the offsets are added to
4988 the base address with silently-wrapping two's complement arithmetic, and
4989 the result value of the <tt>getelementptr</tt> may be outside the object
4990 pointed to by the base pointer. The result value may not necessarily be
4991 used to access memory though, even if it happens to point into allocated
4992 storage. See the <a href="#pointeraliasing">Pointer Aliasing Rules</a>
4993 section for more information.</p>
4994
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4995<p>The getelementptr instruction is often confusing. For some more insight into
4996 how it works, see <a href="GetElementPtr.html">the getelementptr FAQ</a>.</p>
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]4997
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]4998<h5>Example:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4999<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5000 <i>; yields [12 x i8]*:aptr</i>
Matthijs Kooijmane49d0bc2008-10-13 13:44:15 +0000[diff] [blame]5001 %aptr = getelementptr {i32, [12 x i8]}* %saptr, i64 0, i32 1
5002 <i>; yields i8*:vptr</i>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5003 %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]5004 <i>; yields i8*:eptr</i>
5005 %eptr = getelementptr [12 x i8]* %aptr, i64 0, i32 1
Sanjiv Gupta9f805c22009-04-25 07:27:44 +0000[diff] [blame]5006 <i>; yields i32*:iptr</i>
Sanjiv Gupta16ffa802009-04-24 16:38:13 +0000[diff] [blame]5007 %iptr = getelementptr [10 x i32]* @arr, i16 0, i16 0
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]5008</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5009
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]5010</div>
Reid Spencer47ce1792006-11-09 21:15:49 +0000[diff] [blame]5011
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5012</div>
5013
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]5014<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5015<h3>
5016 <a name="convertops">Conversion Operations</a>
5017</h3>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5018
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5019<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5020
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5021<p>The instructions in this category are the conversion instructions (casting)
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5022 which all take a single operand and a type. They perform various bit
5023 conversions on the operand.</p>
5024
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]5025<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5026<h4>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5027 <a name="i_trunc">'<tt>trunc .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5028</h4>
5029
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5030<div>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5031
5032<h5>Syntax:</h5>
5033<pre>
5034 &lt;result&gt; = trunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
5035</pre>
5036
5037<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5038<p>The '<tt>trunc</tt>' instruction truncates its operand to the
5039 type <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5040
5041<h5>Arguments:</h5>
Nadav Rotem8c20ec52011-02-24 21:01:34 +0000[diff] [blame]5042<p>The '<tt>trunc</tt>' instruction takes a value to trunc, and a type to trunc it to.
5043 Both types must be of <a href="#t_integer">integer</a> types, or vectors
5044 of the same number of integers.
5045 The bit size of the <tt>value</tt> must be larger than
5046 the bit size of the destination type, <tt>ty2</tt>.
5047 Equal sized types are not allowed.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5048
5049<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5050<p>The '<tt>trunc</tt>' instruction truncates the high order bits
5051 in <tt>value</tt> and converts the remaining bits to <tt>ty2</tt>. Since the
5052 source size must be larger than the destination size, <tt>trunc</tt> cannot
5053 be a <i>no-op cast</i>. It will always truncate bits.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5054
5055<h5>Example:</h5>
5056<pre>
Nadav Rotem8c20ec52011-02-24 21:01:34 +0000[diff] [blame]5057 %X = trunc i32 257 to i8 <i>; yields i8:1</i>
5058 %Y = trunc i32 123 to i1 <i>; yields i1:true</i>
5059 %Z = trunc i32 122 to i1 <i>; yields i1:false</i>
5060 %W = trunc &lt;2 x i16&gt; &lt;i16 8, i16 7&gt; to &lt;2 x i8&gt; <i>; yields &lt;i8 8, i8 7&gt;</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5061</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5062
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5063</div>
5064
5065<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5066<h4>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5067 <a name="i_zext">'<tt>zext .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5068</h4>
5069
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5070<div>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5071
5072<h5>Syntax:</h5>
5073<pre>
5074 &lt;result&gt; = zext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
5075</pre>
5076
5077<h5>Overview:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5078<p>The '<tt>zext</tt>' instruction zero extends its operand to type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5079 <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5080
5081
5082<h5>Arguments:</h5>
Nadav Rotemed9b9342011-02-20 12:37:50 +0000[diff] [blame]5083<p>The '<tt>zext</tt>' instruction takes a value to cast, and a type to cast it to.
5084 Both types must be of <a href="#t_integer">integer</a> types, or vectors
5085 of the same number of integers.
5086 The bit size of the <tt>value</tt> must be smaller than
5087 the bit size of the destination type,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5088 <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5089
5090<h5>Semantics:</h5>
5091<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]5092 bits until it reaches the size of the destination type, <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5093
Reid Spencerb5929522007-01-12 15:46:11 +0000[diff] [blame]5094<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]5095
5096<h5>Example:</h5>
5097<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5098 %X = zext i32 257 to i64 <i>; yields i64:257</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]5099 %Y = zext i1 true to i32 <i>; yields i32:1</i>
Nadav Rotemed9b9342011-02-20 12:37:50 +0000[diff] [blame]5100 %Z = zext &lt;2 x i16&gt; &lt;i16 8, i16 7&gt; to &lt;2 x i32&gt; <i>; yields &lt;i32 8, i32 7&gt;</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5101</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5102
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5103</div>
5104
5105<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5106<h4>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5107 <a name="i_sext">'<tt>sext .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5108</h4>
5109
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5110<div>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5111
5112<h5>Syntax:</h5>
5113<pre>
5114 &lt;result&gt; = sext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
5115</pre>
5116
5117<h5>Overview:</h5>
5118<p>The '<tt>sext</tt>' sign extends <tt>value</tt> to the type <tt>ty2</tt>.</p>
5119
5120<h5>Arguments:</h5>
Nadav Rotem8c20ec52011-02-24 21:01:34 +0000[diff] [blame]5121<p>The '<tt>sext</tt>' instruction takes a value to cast, and a type to cast it to.
5122 Both types must be of <a href="#t_integer">integer</a> types, or vectors
5123 of the same number of integers.
5124 The bit size of the <tt>value</tt> must be smaller than
5125 the bit size of the destination type,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5126 <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5127
5128<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5129<p>The '<tt>sext</tt>' instruction performs a sign extension by copying the sign
5130 bit (highest order bit) of the <tt>value</tt> until it reaches the bit size
5131 of the type <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5132
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]5133<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]5134
5135<h5>Example:</h5>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5136<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5137 %X = sext i8 -1 to i16 <i>; yields i16 :65535</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]5138 %Y = sext i1 true to i32 <i>; yields i32:-1</i>
Nadav Rotem8c20ec52011-02-24 21:01:34 +0000[diff] [blame]5139 %Z = sext &lt;2 x i16&gt; &lt;i16 8, i16 7&gt; to &lt;2 x i32&gt; <i>; yields &lt;i32 8, i32 7&gt;</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5140</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5141
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5142</div>
5143
5144<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5145<h4>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]5146 <a name="i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5147</h4>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]5148
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5149<div>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]5150
5151<h5>Syntax:</h5>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]5152<pre>
5153 &lt;result&gt; = fptrunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
5154</pre>
5155
5156<h5>Overview:</h5>
5157<p>The '<tt>fptrunc</tt>' instruction truncates <tt>value</tt> to type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5158 <tt>ty2</tt>.</p>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]5159
5160<h5>Arguments:</h5>
5161<p>The '<tt>fptrunc</tt>' instruction takes a <a href="#t_floating">floating
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5162 point</a> value to cast and a <a href="#t_floating">floating point</a> type
5163 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]5164 <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]5165 <i>no-op cast</i>.</p>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]5166
5167<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5168<p>The '<tt>fptrunc</tt>' instruction truncates a <tt>value</tt> from a larger
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5169 <a href="#t_floating">floating point</a> type to a smaller
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5170 <a href="#t_floating">floating point</a> type. If the value cannot fit
5171 within the destination type, <tt>ty2</tt>, then the results are
5172 undefined.</p>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]5173
5174<h5>Example:</h5>
5175<pre>
5176 %X = fptrunc double 123.0 to float <i>; yields float:123.0</i>
5177 %Y = fptrunc double 1.0E+300 to float <i>; yields undefined</i>
5178</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5179
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]5180</div>
5181
5182<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5183<h4>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5184 <a name="i_fpext">'<tt>fpext .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5185</h4>
5186
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5187<div>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5188
5189<h5>Syntax:</h5>
5190<pre>
5191 &lt;result&gt; = fpext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
5192</pre>
5193
5194<h5>Overview:</h5>
5195<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]5196 floating point value.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5197
5198<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5199<p>The '<tt>fpext</tt>' instruction takes a
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5200 <a href="#t_floating">floating point</a> <tt>value</tt> to cast, and
5201 a <a href="#t_floating">floating point</a> type to cast it to. The source
5202 type must be smaller than the destination type.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5203
5204<h5>Semantics:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]5205<p>The '<tt>fpext</tt>' instruction extends the <tt>value</tt> from a smaller
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5206 <a href="#t_floating">floating point</a> type to a larger
5207 <a href="#t_floating">floating point</a> type. The <tt>fpext</tt> cannot be
5208 used to make a <i>no-op cast</i> because it always changes bits. Use
5209 <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]5210
5211<h5>Example:</h5>
5212<pre>
Nick Lewycky5bb3ece2011-03-31 18:20:19 +0000[diff] [blame]5213 %X = fpext float 3.125 to double <i>; yields double:3.125000e+00</i>
5214 %Y = fpext double %X to fp128 <i>; yields fp128:0xL00000000000000004000900000000000</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5215</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5216
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5217</div>
5218
5219<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5220<h4>
Reid Spencer24d6da52007-01-21 00:29:26 +0000[diff] [blame]5221 <a name="i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5222</h4>
5223
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5224<div>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5225
5226<h5>Syntax:</h5>
5227<pre>
Reid Spencer1539a1c2007-07-31 14:40:14 +0000[diff] [blame]5228 &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]5229</pre>
5230
5231<h5>Overview:</h5>
Reid Spencer1539a1c2007-07-31 14:40:14 +0000[diff] [blame]5232<p>The '<tt>fptoui</tt>' converts a floating point <tt>value</tt> to its
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5233 unsigned integer equivalent of type <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5234
5235<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5236<p>The '<tt>fptoui</tt>' instruction takes a value to cast, which must be a
5237 scalar or vector <a href="#t_floating">floating point</a> value, and a type
5238 to cast it to <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a>
5239 type. If <tt>ty</tt> is a vector floating point type, <tt>ty2</tt> must be a
5240 vector integer type with the same number of elements as <tt>ty</tt></p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5241
5242<h5>Semantics:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5243<p>The '<tt>fptoui</tt>' instruction converts its
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5244 <a href="#t_floating">floating point</a> operand into the nearest (rounding
5245 towards zero) unsigned integer value. If the value cannot fit
5246 in <tt>ty2</tt>, the results are undefined.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5247
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5248<h5>Example:</h5>
5249<pre>
Reid Spencer1539a1c2007-07-31 14:40:14 +0000[diff] [blame]5250 %X = fptoui double 123.0 to i32 <i>; yields i32:123</i>
Chris Lattner88519042007-09-22 03:17:52 +0000[diff] [blame]5251 %Y = fptoui float 1.0E+300 to i1 <i>; yields undefined:1</i>
Gabor Greif6a292012009-10-28 09:21:30 +0000[diff] [blame]5252 %Z = fptoui float 1.04E+17 to i8 <i>; yields undefined:1</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5253</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5254
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5255</div>
5256
5257<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5258<h4>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]5259 <a name="i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5260</h4>
5261
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5262<div>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5263
5264<h5>Syntax:</h5>
5265<pre>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]5266 &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]5267</pre>
5268
5269<h5>Overview:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5270<p>The '<tt>fptosi</tt>' instruction converts
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5271 <a href="#t_floating">floating point</a> <tt>value</tt> to
5272 type <tt>ty2</tt>.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5273
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]5274<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5275<p>The '<tt>fptosi</tt>' instruction takes a value to cast, which must be a
5276 scalar or vector <a href="#t_floating">floating point</a> value, and a type
5277 to cast it to <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a>
5278 type. If <tt>ty</tt> is a vector floating point type, <tt>ty2</tt> must be a
5279 vector integer type with the same number of elements as <tt>ty</tt></p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5280
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]5281<h5>Semantics:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5282<p>The '<tt>fptosi</tt>' instruction converts its
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5283 <a href="#t_floating">floating point</a> operand into the nearest (rounding
5284 towards zero) signed integer value. If the value cannot fit in <tt>ty2</tt>,
5285 the results are undefined.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5286
Chris Lattner33ba0d92001-07-09 00:26:23 +0000[diff] [blame]5287<h5>Example:</h5>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5288<pre>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]5289 %X = fptosi double -123.0 to i32 <i>; yields i32:-123</i>
Chris Lattner88519042007-09-22 03:17:52 +0000[diff] [blame]5290 %Y = fptosi float 1.0E-247 to i1 <i>; yields undefined:1</i>
Gabor Greif6a292012009-10-28 09:21:30 +0000[diff] [blame]5291 %Z = fptosi float 1.04E+17 to i8 <i>; yields undefined:1</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5292</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5293
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5294</div>
5295
5296<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5297<h4>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]5298 <a name="i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5299</h4>
5300
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5301<div>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5302
5303<h5>Syntax:</h5>
5304<pre>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]5305 &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]5306</pre>
5307
5308<h5>Overview:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]5309<p>The '<tt>uitofp</tt>' instruction regards <tt>value</tt> as an unsigned
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5310 integer and converts that value to the <tt>ty2</tt> type.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5311
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5312<h5>Arguments:</h5>
Nate Begemanb348d182007-11-17 03:58:34 +0000[diff] [blame]5313<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]5314 scalar or vector <a href="#t_integer">integer</a> value, and a type to cast
5315 it to <tt>ty2</tt>, which must be an <a href="#t_floating">floating point</a>
5316 type. If <tt>ty</tt> is a vector integer type, <tt>ty2</tt> must be a vector
5317 floating point type with the same number of elements as <tt>ty</tt></p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5318
5319<h5>Semantics:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]5320<p>The '<tt>uitofp</tt>' instruction interprets its operand as an unsigned
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5321 integer quantity and converts it to the corresponding floating point
5322 value. If the value cannot fit in the floating point value, the results are
5323 undefined.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5324
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5325<h5>Example:</h5>
5326<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5327 %X = uitofp i32 257 to float <i>; yields float:257.0</i>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5328 %Y = uitofp i8 -1 to double <i>; yields double:255.0</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5329</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5330
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5331</div>
5332
5333<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5334<h4>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]5335 <a name="i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5336</h4>
5337
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5338<div>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5339
5340<h5>Syntax:</h5>
5341<pre>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]5342 &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]5343</pre>
5344
5345<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5346<p>The '<tt>sitofp</tt>' instruction regards <tt>value</tt> as a signed integer
5347 and converts that value to the <tt>ty2</tt> type.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5348
5349<h5>Arguments:</h5>
Nate Begemanb348d182007-11-17 03:58:34 +0000[diff] [blame]5350<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]5351 scalar or vector <a href="#t_integer">integer</a> value, and a type to cast
5352 it to <tt>ty2</tt>, which must be an <a href="#t_floating">floating point</a>
5353 type. If <tt>ty</tt> is a vector integer type, <tt>ty2</tt> must be a vector
5354 floating point type with the same number of elements as <tt>ty</tt></p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5355
5356<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5357<p>The '<tt>sitofp</tt>' instruction interprets its operand as a signed integer
5358 quantity and converts it to the corresponding floating point value. If the
5359 value cannot fit in the floating point value, the results are undefined.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5360
5361<h5>Example:</h5>
5362<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5363 %X = sitofp i32 257 to float <i>; yields float:257.0</i>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5364 %Y = sitofp i8 -1 to double <i>; yields double:-1.0</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5365</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5366
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5367</div>
5368
5369<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5370<h4>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5371 <a name="i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5372</h4>
5373
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5374<div>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5375
5376<h5>Syntax:</h5>
5377<pre>
5378 &lt;result&gt; = ptrtoint &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
5379</pre>
5380
5381<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5382<p>The '<tt>ptrtoint</tt>' instruction converts the pointer <tt>value</tt> to
5383 the integer type <tt>ty2</tt>.</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5384
5385<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5386<p>The '<tt>ptrtoint</tt>' instruction takes a <tt>value</tt> to cast, which
5387 must be a <a href="#t_pointer">pointer</a> value, and a type to cast it to
5388 <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]5389
5390<h5>Semantics:</h5>
5391<p>The '<tt>ptrtoint</tt>' instruction converts <tt>value</tt> to integer type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5392 <tt>ty2</tt> by interpreting the pointer value as an integer and either
5393 truncating or zero extending that value to the size of the integer type. If
5394 <tt>value</tt> is smaller than <tt>ty2</tt> then a zero extension is done. If
5395 <tt>value</tt> is larger than <tt>ty2</tt> then a truncation is done. If they
5396 are the same size, then nothing is done (<i>no-op cast</i>) other than a type
5397 change.</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5398
5399<h5>Example:</h5>
5400<pre>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5401 %X = ptrtoint i32* %X to i8 <i>; yields truncation on 32-bit architecture</i>
5402 %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]5403</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5404
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5405</div>
5406
5407<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5408<h4>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5409 <a name="i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5410</h4>
5411
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5412<div>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5413
5414<h5>Syntax:</h5>
5415<pre>
5416 &lt;result&gt; = inttoptr &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
5417</pre>
5418
5419<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5420<p>The '<tt>inttoptr</tt>' instruction converts an integer <tt>value</tt> to a
5421 pointer type, <tt>ty2</tt>.</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5422
5423<h5>Arguments:</h5>
Duncan Sands8036ca42007-03-30 12:22:09 +0000[diff] [blame]5424<p>The '<tt>inttoptr</tt>' instruction takes an <a href="#t_integer">integer</a>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5425 value to cast, and a type to cast it to, which must be a
5426 <a href="#t_pointer">pointer</a> type.</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5427
5428<h5>Semantics:</h5>
5429<p>The '<tt>inttoptr</tt>' instruction converts <tt>value</tt> to type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5430 <tt>ty2</tt> by applying either a zero extension or a truncation depending on
5431 the size of the integer <tt>value</tt>. If <tt>value</tt> is larger than the
5432 size of a pointer then a truncation is done. If <tt>value</tt> is smaller
5433 than the size of a pointer then a zero extension is done. If they are the
5434 same size, nothing is done (<i>no-op cast</i>).</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5435
5436<h5>Example:</h5>
5437<pre>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5438 %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]5439 %Y = inttoptr i32 255 to i32* <i>; yields no-op on 32-bit architecture</i>
5440 %Z = inttoptr i64 0 to i32* <i>; yields truncation on 32-bit architecture</i>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5441</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5442
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5443</div>
5444
5445<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5446<h4>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]5447 <a name="i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5448</h4>
5449
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5450<div>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5451
5452<h5>Syntax:</h5>
5453<pre>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]5454 &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]5455</pre>
5456
5457<h5>Overview:</h5>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]5458<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5459 <tt>ty2</tt> without changing any bits.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5460
5461<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5462<p>The '<tt>bitcast</tt>' instruction takes a value to cast, which must be a
5463 non-aggregate first class value, and a type to cast it to, which must also be
5464 a non-aggregate <a href="#t_firstclass">first class</a> type. The bit sizes
5465 of <tt>value</tt> and the destination type, <tt>ty2</tt>, must be
5466 identical. If the source type is a pointer, the destination type must also be
5467 a pointer. This instruction supports bitwise conversion of vectors to
5468 integers and to vectors of other types (as long as they have the same
5469 size).</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5470
5471<h5>Semantics:</h5>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]5472<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5473 <tt>ty2</tt>. It is always a <i>no-op cast</i> because no bits change with
5474 this conversion. The conversion is done as if the <tt>value</tt> had been
5475 stored to memory and read back as type <tt>ty2</tt>. Pointer types may only
5476 be converted to other pointer types with this instruction. To convert
5477 pointers to other types, use the <a href="#i_inttoptr">inttoptr</a> or
5478 <a href="#i_ptrtoint">ptrtoint</a> instructions first.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5479
5480<h5>Example:</h5>
5481<pre>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5482 %X = bitcast i8 255 to i8 <i>; yields i8 :-1</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5483 %Y = bitcast i32* %x to sint* <i>; yields sint*:%x</i>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5484 %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]5485</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5486
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5487</div>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5488
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5489</div>
5490
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5491<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5492<h3>
5493 <a name="otherops">Other Operations</a>
5494</h3>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5495
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5496<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5497
5498<p>The instructions in this category are the "miscellaneous" instructions, which
5499 defy better classification.</p>
5500
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5501<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5502<h4>
5503 <a name="i_icmp">'<tt>icmp</tt>' Instruction</a>
5504</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5505
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5506<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5507
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5508<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5509<pre>
5510 &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]5511</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5512
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5513<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5514<p>The '<tt>icmp</tt>' instruction returns a boolean value or a vector of
5515 boolean values based on comparison of its two integer, integer vector, or
5516 pointer operands.</p>
5517
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5518<h5>Arguments:</h5>
5519<p>The '<tt>icmp</tt>' instruction takes three operands. The first operand is
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5520 the condition code indicating the kind of comparison to perform. It is not a
5521 value, just a keyword. The possible condition code are:</p>
5522
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5523<ol>
5524 <li><tt>eq</tt>: equal</li>
5525 <li><tt>ne</tt>: not equal </li>
5526 <li><tt>ugt</tt>: unsigned greater than</li>
5527 <li><tt>uge</tt>: unsigned greater or equal</li>
5528 <li><tt>ult</tt>: unsigned less than</li>
5529 <li><tt>ule</tt>: unsigned less or equal</li>
5530 <li><tt>sgt</tt>: signed greater than</li>
5531 <li><tt>sge</tt>: signed greater or equal</li>
5532 <li><tt>slt</tt>: signed less than</li>
5533 <li><tt>sle</tt>: signed less or equal</li>
5534</ol>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5535
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]5536<p>The remaining two arguments must be <a href="#t_integer">integer</a> or
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5537 <a href="#t_pointer">pointer</a> or integer <a href="#t_vector">vector</a>
5538 typed. They must also be identical types.</p>
5539
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5540<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5541<p>The '<tt>icmp</tt>' compares <tt>op1</tt> and <tt>op2</tt> according to the
5542 condition code given as <tt>cond</tt>. The comparison performed always yields
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]5543 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]5544 result, as follows:</p>
5545
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5546<ol>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5547 <li><tt>eq</tt>: yields <tt>true</tt> if the operands are equal,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5548 <tt>false</tt> otherwise. No sign interpretation is necessary or
5549 performed.</li>
5550
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5551 <li><tt>ne</tt>: yields <tt>true</tt> if the operands are unequal,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5552 <tt>false</tt> otherwise. No sign interpretation is necessary or
5553 performed.</li>
5554
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5555 <li><tt>ugt</tt>: interprets the operands as unsigned values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5556 <tt>true</tt> if <tt>op1</tt> is greater than <tt>op2</tt>.</li>
5557
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5558 <li><tt>uge</tt>: interprets the operands as unsigned values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5559 <tt>true</tt> if <tt>op1</tt> is greater than or equal
5560 to <tt>op2</tt>.</li>
5561
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5562 <li><tt>ult</tt>: interprets the operands as unsigned values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5563 <tt>true</tt> if <tt>op1</tt> is less than <tt>op2</tt>.</li>
5564
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5565 <li><tt>ule</tt>: interprets the operands as unsigned values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5566 <tt>true</tt> if <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
5567
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5568 <li><tt>sgt</tt>: interprets the operands as signed values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5569 <tt>true</tt> if <tt>op1</tt> is greater than <tt>op2</tt>.</li>
5570
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5571 <li><tt>sge</tt>: interprets the operands as signed values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5572 <tt>true</tt> if <tt>op1</tt> is greater than or equal
5573 to <tt>op2</tt>.</li>
5574
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5575 <li><tt>slt</tt>: interprets the operands as signed values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5576 <tt>true</tt> if <tt>op1</tt> is less than <tt>op2</tt>.</li>
5577
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5578 <li><tt>sle</tt>: interprets the operands as signed values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5579 <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]5580</ol>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5581
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5582<p>If the operands are <a href="#t_pointer">pointer</a> typed, the pointer
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5583 values are compared as if they were integers.</p>
5584
5585<p>If the operands are integer vectors, then they are compared element by
5586 element. The result is an <tt>i1</tt> vector with the same number of elements
5587 as the values being compared. Otherwise, the result is an <tt>i1</tt>.</p>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5588
5589<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5590<pre>
5591 &lt;result&gt; = icmp eq i32 4, 5 <i>; yields: result=false</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5592 &lt;result&gt; = icmp ne float* %X, %X <i>; yields: result=false</i>
5593 &lt;result&gt; = icmp ult i16 4, 5 <i>; yields: result=true</i>
5594 &lt;result&gt; = icmp sgt i16 4, 5 <i>; yields: result=false</i>
5595 &lt;result&gt; = icmp ule i16 -4, 5 <i>; yields: result=false</i>
5596 &lt;result&gt; = icmp sge i16 4, 5 <i>; yields: result=false</i>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5597</pre>
Dan Gohmanc766f722009-01-22 01:39:38 +0000[diff] [blame]5598
5599<p>Note that the code generator does not yet support vector types with
5600 the <tt>icmp</tt> instruction.</p>
5601
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5602</div>
5603
5604<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5605<h4>
5606 <a name="i_fcmp">'<tt>fcmp</tt>' Instruction</a>
5607</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5608
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5609<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5610
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5611<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5612<pre>
5613 &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]5614</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5615
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5616<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5617<p>The '<tt>fcmp</tt>' instruction returns a boolean value or vector of boolean
5618 values based on comparison of its operands.</p>
5619
5620<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]5621(<a href="#t_integer"><tt>i1</tt></a>).</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5622
5623<p>If the operands are floating point vectors, then the result type is a vector
5624 of boolean with the same number of elements as the operands being
5625 compared.</p>
5626
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5627<h5>Arguments:</h5>
5628<p>The '<tt>fcmp</tt>' instruction takes three operands. The first operand is
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5629 the condition code indicating the kind of comparison to perform. It is not a
5630 value, just a keyword. The possible condition code are:</p>
5631
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5632<ol>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]5633 <li><tt>false</tt>: no comparison, always returns false</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5634 <li><tt>oeq</tt>: ordered and equal</li>
5635 <li><tt>ogt</tt>: ordered and greater than </li>
5636 <li><tt>oge</tt>: ordered and greater than or equal</li>
5637 <li><tt>olt</tt>: ordered and less than </li>
5638 <li><tt>ole</tt>: ordered and less than or equal</li>
5639 <li><tt>one</tt>: ordered and not equal</li>
5640 <li><tt>ord</tt>: ordered (no nans)</li>
5641 <li><tt>ueq</tt>: unordered or equal</li>
5642 <li><tt>ugt</tt>: unordered or greater than </li>
5643 <li><tt>uge</tt>: unordered or greater than or equal</li>
5644 <li><tt>ult</tt>: unordered or less than </li>
5645 <li><tt>ule</tt>: unordered or less than or equal</li>
5646 <li><tt>une</tt>: unordered or not equal</li>
5647 <li><tt>uno</tt>: unordered (either nans)</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]5648 <li><tt>true</tt>: no comparison, always returns true</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5649</ol>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5650
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5651<p><i>Ordered</i> means that neither operand is a QNAN while
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5652 <i>unordered</i> means that either operand may be a QNAN.</p>
5653
5654<p>Each of <tt>val1</tt> and <tt>val2</tt> arguments must be either
5655 a <a href="#t_floating">floating point</a> type or
5656 a <a href="#t_vector">vector</a> of floating point type. They must have
5657 identical types.</p>
5658
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5659<h5>Semantics:</h5>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]5660<p>The '<tt>fcmp</tt>' instruction compares <tt>op1</tt> and <tt>op2</tt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5661 according to the condition code given as <tt>cond</tt>. If the operands are
5662 vectors, then the vectors are compared element by element. Each comparison
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]5663 performed always yields an <a href="#t_integer">i1</a> result, as
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5664 follows:</p>
5665
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5666<ol>
5667 <li><tt>false</tt>: always yields <tt>false</tt>, regardless of operands.</li>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5668
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5669 <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]5670 <tt>op1</tt> is equal to <tt>op2</tt>.</li>
5671
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]5672 <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]5673 <tt>op1</tt> is greater than <tt>op2</tt>.</li>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5674
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5675 <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]5676 <tt>op1</tt> is greater than or equal to <tt>op2</tt>.</li>
5677
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5678 <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]5679 <tt>op1</tt> is less than <tt>op2</tt>.</li>
5680
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5681 <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]5682 <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
5683
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5684 <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]5685 <tt>op1</tt> is not equal to <tt>op2</tt>.</li>
5686
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]5687 <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]5688
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5689 <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]5690 <tt>op1</tt> is equal to <tt>op2</tt>.</li>
5691
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5692 <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]5693 <tt>op1</tt> is greater than <tt>op2</tt>.</li>
5694
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5695 <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]5696 <tt>op1</tt> is greater than or equal to <tt>op2</tt>.</li>
5697
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5698 <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]5699 <tt>op1</tt> is less than <tt>op2</tt>.</li>
5700
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5701 <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]5702 <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
5703
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5704 <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]5705 <tt>op1</tt> is not equal to <tt>op2</tt>.</li>
5706
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]5707 <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]5708
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5709 <li><tt>true</tt>: always yields <tt>true</tt>, regardless of operands.</li>
5710</ol>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5711
5712<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5713<pre>
5714 &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]5715 &lt;result&gt; = fcmp one float 4.0, 5.0 <i>; yields: result=true</i>
5716 &lt;result&gt; = fcmp olt float 4.0, 5.0 <i>; yields: result=true</i>
5717 &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]5718</pre>
Dan Gohmanc766f722009-01-22 01:39:38 +0000[diff] [blame]5719
5720<p>Note that the code generator does not yet support vector types with
5721 the <tt>fcmp</tt> instruction.</p>
5722
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5723</div>
5724
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5725<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5726<h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]5727 <a name="i_phi">'<tt>phi</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5728</h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]5729
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5730<div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]5731
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5732<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5733<pre>
5734 &lt;result&gt; = phi &lt;ty&gt; [ &lt;val0&gt;, &lt;label0&gt;], ...
5735</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]5736
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5737<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5738<p>The '<tt>phi</tt>' instruction is used to implement the &#966; node in the
5739 SSA graph representing the function.</p>
5740
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5741<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5742<p>The type of the incoming values is specified with the first type field. After
5743 this, the '<tt>phi</tt>' instruction takes a list of pairs as arguments, with
5744 one pair for each predecessor basic block of the current block. Only values
5745 of <a href="#t_firstclass">first class</a> type may be used as the value
5746 arguments to the PHI node. Only labels may be used as the label
5747 arguments.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]5748
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5749<p>There must be no non-phi instructions between the start of a basic block and
5750 the PHI instructions: i.e. PHI instructions must be first in a basic
5751 block.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]5752
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5753<p>For the purposes of the SSA form, the use of each incoming value is deemed to
5754 occur on the edge from the corresponding predecessor block to the current
5755 block (but after any definition of an '<tt>invoke</tt>' instruction's return
5756 value on the same edge).</p>
Jay Foadd2449092009-06-03 10:20:10 +0000[diff] [blame]5757
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5758<h5>Semantics:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5759<p>At runtime, the '<tt>phi</tt>' instruction logically takes on the value
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5760 specified by the pair corresponding to the predecessor basic block that
5761 executed just prior to the current block.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]5762
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5763<h5>Example:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]5764<pre>
5765Loop: ; Infinite loop that counts from 0 on up...
5766 %indvar = phi i32 [ 0, %LoopHeader ], [ %nextindvar, %Loop ]
5767 %nextindvar = add i32 %indvar, 1
5768 br label %Loop
5769</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5770
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5771</div>
5772
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5773<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5774<h4>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5775 <a name="i_select">'<tt>select</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5776</h4>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5777
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5778<div>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5779
5780<h5>Syntax:</h5>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5781<pre>
Dan Gohmanf72fb672008-09-09 01:02:47 +0000[diff] [blame]5782 &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>
5783
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5784 <i>selty</i> is either i1 or {&lt;N x i1&gt;}
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5785</pre>
5786
5787<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5788<p>The '<tt>select</tt>' instruction is used to choose one value based on a
5789 condition, without branching.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5790
5791
5792<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5793<p>The '<tt>select</tt>' instruction requires an 'i1' value or a vector of 'i1'
5794 values indicating the condition, and two values of the
5795 same <a href="#t_firstclass">first class</a> type. If the val1/val2 are
5796 vectors and the condition is a scalar, then entire vectors are selected, not
5797 individual elements.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5798
5799<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5800<p>If the condition is an i1 and it evaluates to 1, the instruction returns the
5801 first value argument; otherwise, it returns the second value argument.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5802
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5803<p>If the condition is a vector of i1, then the value arguments must be vectors
5804 of the same size, and the selection is done element by element.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5805
5806<h5>Example:</h5>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5807<pre>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]5808 %X = select i1 true, i8 17, i8 42 <i>; yields i8:17</i>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5809</pre>
Dan Gohmanc766f722009-01-22 01:39:38 +0000[diff] [blame]5810
5811<p>Note that the code generator does not yet support conditions
5812 with vector type.</p>
5813
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5814</div>
5815
Robert Bocchino05ccd702006-01-15 20:48:27 +0000[diff] [blame]5816<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5817<h4>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5818 <a name="i_call">'<tt>call</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5819</h4>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5820
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5821<div>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5822
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]5823<h5>Syntax:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5824<pre>
Devang Patel307e8ab2008-10-07 17:48:33 +0000[diff] [blame]5825 &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]5826</pre>
5827
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]5828<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5829<p>The '<tt>call</tt>' instruction represents a simple function call.</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5830
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]5831<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5832<p>This instruction requires several arguments:</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5833
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]5834<ol>
Jeffrey Yasskin95fa80a2010-01-09 19:44:16 +0000[diff] [blame]5835 <li>The optional "tail" marker indicates that the callee function does not
5836 access any allocas or varargs in the caller. Note that calls may be
5837 marked "tail" even if they do not occur before
5838 a <a href="#i_ret"><tt>ret</tt></a> instruction. If the "tail" marker is
5839 present, the function call is eligible for tail call optimization,
5840 but <a href="CodeGenerator.html#tailcallopt">might not in fact be
Evan Chengdc444e92010-03-08 21:05:02 +0000[diff] [blame]5841 optimized into a jump</a>. The code generator may optimize calls marked
5842 "tail" with either 1) automatic <a href="CodeGenerator.html#sibcallopt">
5843 sibling call optimization</a> when the caller and callee have
5844 matching signatures, or 2) forced tail call optimization when the
5845 following extra requirements are met:
Jeffrey Yasskin95fa80a2010-01-09 19:44:16 +0000[diff] [blame]5846 <ul>
5847 <li>Caller and callee both have the calling
5848 convention <tt>fastcc</tt>.</li>
5849 <li>The call is in tail position (ret immediately follows call and ret
5850 uses value of call or is void).</li>
5851 <li>Option <tt>-tailcallopt</tt> is enabled,
Dan Gohmanfbbee8d2010-03-02 01:08:11 +0000[diff] [blame]5852 or <code>llvm::GuaranteedTailCallOpt</code> is <code>true</code>.</li>
Jeffrey Yasskin95fa80a2010-01-09 19:44:16 +0000[diff] [blame]5853 <li><a href="CodeGenerator.html#tailcallopt">Platform specific
5854 constraints are met.</a></li>
5855 </ul>
5856 </li>
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]5857
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5858 <li>The optional "cconv" marker indicates which <a href="#callingconv">calling
5859 convention</a> the call should use. If none is specified, the call
Jeffrey Yasskin95fa80a2010-01-09 19:44:16 +0000[diff] [blame]5860 defaults to using C calling conventions. The calling convention of the
5861 call must match the calling convention of the target function, or else the
5862 behavior is undefined.</li>
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]5863
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5864 <li>The optional <a href="#paramattrs">Parameter Attributes</a> list for
5865 return values. Only '<tt>zeroext</tt>', '<tt>signext</tt>', and
5866 '<tt>inreg</tt>' attributes are valid here.</li>
5867
5868 <li>'<tt>ty</tt>': the type of the call instruction itself which is also the
5869 type of the return value. Functions that return no value are marked
5870 <tt><a href="#t_void">void</a></tt>.</li>
5871
5872 <li>'<tt>fnty</tt>': shall be the signature of the pointer to function value
5873 being invoked. The argument types must match the types implied by this
5874 signature. This type can be omitted if the function is not varargs and if
5875 the function type does not return a pointer to a function.</li>
5876
5877 <li>'<tt>fnptrval</tt>': An LLVM value containing a pointer to a function to
5878 be invoked. In most cases, this is a direct function invocation, but
5879 indirect <tt>call</tt>s are just as possible, calling an arbitrary pointer
5880 to function value.</li>
5881
5882 <li>'<tt>function args</tt>': argument list whose types match the function
Chris Lattner0724fbd2010-03-02 06:36:51 +0000[diff] [blame]5883 signature argument types and parameter attributes. All arguments must be
5884 of <a href="#t_firstclass">first class</a> type. If the function
5885 signature indicates the function accepts a variable number of arguments,
5886 the extra arguments can be specified.</li>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5887
5888 <li>The optional <a href="#fnattrs">function attributes</a> list. Only
5889 '<tt>noreturn</tt>', '<tt>nounwind</tt>', '<tt>readonly</tt>' and
5890 '<tt>readnone</tt>' attributes are valid here.</li>
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]5891</ol>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5892
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]5893<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5894<p>The '<tt>call</tt>' instruction is used to cause control flow to transfer to
5895 a specified function, with its incoming arguments bound to the specified
5896 values. Upon a '<tt><a href="#i_ret">ret</a></tt>' instruction in the called
5897 function, control flow continues with the instruction after the function
5898 call, and the return value of the function is bound to the result
5899 argument.</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5900
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]5901<h5>Example:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5902<pre>
Nick Lewyckydb7e3c92007-09-08 13:57:50 +0000[diff] [blame]5903 %retval = call i32 @test(i32 %argc)
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]5904 call i32 (i8*, ...)* @printf(i8* %msg, i32 12, i8 42) <i>; yields i32</i>
Chris Lattner772fccf2008-03-21 17:24:17 +0000[diff] [blame]5905 %X = tail call i32 @foo() <i>; yields i32</i>
5906 %Y = tail call <a href="#callingconv">fastcc</a> i32 @foo() <i>; yields i32</i>
5907 call void %foo(i8 97 signext)
Devang Patelc3fc6df2008-03-10 20:49:15 +0000[diff] [blame]5908
5909 %struct.A = type { i32, i8 }
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]5910 %r = call %struct.A @foo() <i>; yields { 32, i8 }</i>
Dan Gohmanb1e6b962008-10-04 19:00:07 +0000[diff] [blame]5911 %gr = extractvalue %struct.A %r, 0 <i>; yields i32</i>
5912 %gr1 = extractvalue %struct.A %r, 1 <i>; yields i8</i>
Chris Lattner85a350f2008-10-08 06:26:11 +0000[diff] [blame]5913 %Z = call void @foo() noreturn <i>; indicates that %foo never returns normally</i>
Matthijs Kooijmancb73d192008-10-07 10:03:45 +0000[diff] [blame]5914 %ZZ = call zeroext i32 @bar() <i>; Return value is %zero extended</i>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5915</pre>
5916
Dale Johannesen07de8d12009-09-24 18:38:21 +0000[diff] [blame]5917<p>llvm treats calls to some functions with names and arguments that match the
Dale Johannesen9f8380b2009-09-25 17:04:42 +0000[diff] [blame]5918standard C99 library as being the C99 library functions, and may perform
5919optimizations or generate code for them under that assumption. This is
5920something we'd like to change in the future to provide better support for
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]5921freestanding environments and non-C-based languages.</p>
Dale Johannesen07de8d12009-09-24 18:38:21 +0000[diff] [blame]5922
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5923</div>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5924
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5925<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5926<h4>
Chris Lattnerfb6977d2006-01-13 23:26:01 +0000[diff] [blame]5927 <a name="i_va_arg">'<tt>va_arg</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5928</h4>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5929
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5930<div>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5931
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]5932<h5>Syntax:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5933<pre>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]5934 &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]5935</pre>
5936
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]5937<h5>Overview:</h5>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]5938<p>The '<tt>va_arg</tt>' instruction is used to access arguments passed through
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5939 the "variable argument" area of a function call. It is used to implement the
5940 <tt>va_arg</tt> macro in C.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5941
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]5942<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5943<p>This instruction takes a <tt>va_list*</tt> value and the type of the
5944 argument. It returns a value of the specified argument type and increments
5945 the <tt>va_list</tt> to point to the next argument. The actual type
5946 of <tt>va_list</tt> is target specific.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5947
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]5948<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5949<p>The '<tt>va_arg</tt>' instruction loads an argument of the specified type
5950 from the specified <tt>va_list</tt> and causes the <tt>va_list</tt> to point
5951 to the next argument. For more information, see the variable argument
5952 handling <a href="#int_varargs">Intrinsic Functions</a>.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5953
5954<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]5955 take a variable number of arguments, for example, the <tt>vfprintf</tt>
5956 function.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5957
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5958<p><tt>va_arg</tt> is an LLVM instruction instead of
5959 an <a href="#intrinsics">intrinsic function</a> because it takes a type as an
5960 argument.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5961
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]5962<h5>Example:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5963<p>See the <a href="#int_varargs">variable argument processing</a> section.</p>
5964
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5965<p>Note that the code generator does not yet fully support va_arg on many
5966 targets. Also, it does not currently support va_arg with aggregate types on
5967 any target.</p>
Dan Gohmanf3e60bd2009-01-12 23:12:39 +0000[diff] [blame]5968
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5969</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5970
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5971</div>
5972
5973</div>
5974
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5975<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5976<h2><a name="intrinsics">Intrinsic Functions</a></h2>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]5977<!-- *********************************************************************** -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5978
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5979<div>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5980
5981<p>LLVM supports the notion of an "intrinsic function". These functions have
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5982 well known names and semantics and are required to follow certain
5983 restrictions. Overall, these intrinsics represent an extension mechanism for
5984 the LLVM language that does not require changing all of the transformations
5985 in LLVM when adding to the language (or the bitcode reader/writer, the
5986 parser, etc...).</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5987
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]5988<p>Intrinsic function names must all start with an "<tt>llvm.</tt>" prefix. This
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5989 prefix is reserved in LLVM for intrinsic names; thus, function names may not
5990 begin with this prefix. Intrinsic functions must always be external
5991 functions: you cannot define the body of intrinsic functions. Intrinsic
5992 functions may only be used in call or invoke instructions: it is illegal to
5993 take the address of an intrinsic function. Additionally, because intrinsic
5994 functions are part of the LLVM language, it is required if any are added that
5995 they be documented here.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5996
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5997<p>Some intrinsic functions can be overloaded, i.e., the intrinsic represents a
5998 family of functions that perform the same operation but on different data
5999 types. Because LLVM can represent over 8 million different integer types,
6000 overloading is used commonly to allow an intrinsic function to operate on any
6001 integer type. One or more of the argument types or the result type can be
6002 overloaded to accept any integer type. Argument types may also be defined as
6003 exactly matching a previous argument's type or the result type. This allows
6004 an intrinsic function which accepts multiple arguments, but needs all of them
6005 to be of the same type, to only be overloaded with respect to a single
6006 argument or the result.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6007
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6008<p>Overloaded intrinsics will have the names of its overloaded argument types
6009 encoded into its function name, each preceded by a period. Only those types
6010 which are overloaded result in a name suffix. Arguments whose type is matched
6011 against another type do not. For example, the <tt>llvm.ctpop</tt> function
6012 can take an integer of any width and returns an integer of exactly the same
6013 integer width. This leads to a family of functions such as
6014 <tt>i8 @llvm.ctpop.i8(i8 %val)</tt> and <tt>i29 @llvm.ctpop.i29(i29
6015 %val)</tt>. Only one type, the return type, is overloaded, and only one type
6016 suffix is required. Because the argument's type is matched against the return
6017 type, it does not require its own name suffix.</p>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]6018
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]6019<p>To learn how to add an intrinsic function, please see the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6020 <a href="ExtendingLLVM.html">Extending LLVM Guide</a>.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6021
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]6022<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6023<h3>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]6024 <a name="int_varargs">Variable Argument Handling Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6025</h3>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]6026
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6027<div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6028
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6029<p>Variable argument support is defined in LLVM with
6030 the <a href="#i_va_arg"><tt>va_arg</tt></a> instruction and these three
6031 intrinsic functions. These functions are related to the similarly named
6032 macros defined in the <tt>&lt;stdarg.h&gt;</tt> header file.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6033
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6034<p>All of these functions operate on arguments that use a target-specific value
6035 type "<tt>va_list</tt>". The LLVM assembly language reference manual does
6036 not define what this type is, so all transformations should be prepared to
6037 handle these functions regardless of the type used.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6038
Chris Lattner374ab302006-05-15 17:26:46 +0000[diff] [blame]6039<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]6040 instruction and the variable argument handling intrinsic functions are
6041 used.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6042
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]6043<pre class="doc_code">
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]6044define i32 @test(i32 %X, ...) {
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6045 ; Initialize variable argument processing
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]6046 %ap = alloca i8*
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]6047 %ap2 = bitcast i8** %ap to i8*
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]6048 call void @llvm.va_start(i8* %ap2)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6049
6050 ; Read a single integer argument
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]6051 %tmp = va_arg i8** %ap, i32
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6052
6053 ; Demonstrate usage of llvm.va_copy and llvm.va_end
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]6054 %aq = alloca i8*
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]6055 %aq2 = bitcast i8** %aq to i8*
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]6056 call void @llvm.va_copy(i8* %aq2, i8* %ap2)
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]6057 call void @llvm.va_end(i8* %aq2)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6058
6059 ; Stop processing of arguments.
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]6060 call void @llvm.va_end(i8* %ap2)
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]6061 ret i32 %tmp
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6062}
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]6063
6064declare void @llvm.va_start(i8*)
6065declare void @llvm.va_copy(i8*, i8*)
6066declare void @llvm.va_end(i8*)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6067</pre>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]6068
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]6069<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6070<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6071 <a name="int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6072</h4>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]6073
6074
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6075<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6076
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]6077<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6078<pre>
6079 declare void %llvm.va_start(i8* &lt;arglist&gt;)
6080</pre>
6081
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]6082<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6083<p>The '<tt>llvm.va_start</tt>' intrinsic initializes <tt>*&lt;arglist&gt;</tt>
6084 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]6085
6086<h5>Arguments:</h5>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]6087<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]6088
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]6089<h5>Semantics:</h5>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]6090<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]6091 macro available in C. In a target-dependent way, it initializes
6092 the <tt>va_list</tt> element to which the argument points, so that the next
6093 call to <tt>va_arg</tt> will produce the first variable argument passed to
6094 the function. Unlike the C <tt>va_start</tt> macro, this intrinsic does not
6095 need to know the last argument of the function as the compiler can figure
6096 that out.</p>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]6097
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]6098</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]6099
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]6100<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6101<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6102 <a name="int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6103</h4>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]6104
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6105<div>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]6106
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6107<h5>Syntax:</h5>
6108<pre>
6109 declare void @llvm.va_end(i8* &lt;arglist&gt;)
6110</pre>
6111
6112<h5>Overview:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]6113<p>The '<tt>llvm.va_end</tt>' intrinsic destroys <tt>*&lt;arglist&gt;</tt>,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6114 which has been initialized previously
6115 with <tt><a href="#int_va_start">llvm.va_start</a></tt>
6116 or <tt><a href="#i_va_copy">llvm.va_copy</a></tt>.</p>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]6117
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]6118<h5>Arguments:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]6119<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]6120
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]6121<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]6122<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]6123 macro available in C. In a target-dependent way, it destroys
6124 the <tt>va_list</tt> element to which the argument points. Calls
6125 to <a href="#int_va_start"><tt>llvm.va_start</tt></a>
6126 and <a href="#int_va_copy"> <tt>llvm.va_copy</tt></a> must be matched exactly
6127 with calls to <tt>llvm.va_end</tt>.</p>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]6128
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]6129</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]6130
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]6131<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6132<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6133 <a name="int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6134</h4>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]6135
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6136<div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6137
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]6138<h5>Syntax:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6139<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]6140 declare void @llvm.va_copy(i8* &lt;destarglist&gt;, i8* &lt;srcarglist&gt;)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6141</pre>
6142
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]6143<h5>Overview:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]6144<p>The '<tt>llvm.va_copy</tt>' intrinsic copies the current argument position
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6145 from the source argument list to the destination argument list.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6146
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]6147<h5>Arguments:</h5>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]6148<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]6149 The second argument is a pointer to a <tt>va_list</tt> element to copy
6150 from.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6151
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]6152<h5>Semantics:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]6153<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]6154 macro available in C. In a target-dependent way, it copies the
6155 source <tt>va_list</tt> element into the destination <tt>va_list</tt>
6156 element. This intrinsic is necessary because
6157 the <tt><a href="#int_va_start"> llvm.va_start</a></tt> intrinsic may be
6158 arbitrarily complex and require, for example, memory allocation.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6159
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]6160</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]6161
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6162</div>
6163
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6164<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6165<h3>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6166 <a name="int_gc">Accurate Garbage Collection Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6167</h3>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6168
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6169<div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6170
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6171<p>LLVM support for <a href="GarbageCollection.html">Accurate Garbage
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]6172Collection</a> (GC) requires the implementation and generation of these
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6173intrinsics. These intrinsics allow identification of <a href="#int_gcroot">GC
6174roots on the stack</a>, as well as garbage collector implementations that
6175require <a href="#int_gcread">read</a> and <a href="#int_gcwrite">write</a>
6176barriers. Front-ends for type-safe garbage collected languages should generate
6177these intrinsics to make use of the LLVM garbage collectors. For more details,
6178see <a href="GarbageCollection.html">Accurate Garbage Collection with
6179LLVM</a>.</p>
Christopher Lamb303dae92007-12-17 01:00:21 +0000[diff] [blame]6180
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6181<p>The garbage collection intrinsics only operate on objects in the generic
6182 address space (address space zero).</p>
Christopher Lamb303dae92007-12-17 01:00:21 +0000[diff] [blame]6183
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6184<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6185<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6186 <a name="int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6187</h4>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6188
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6189<div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6190
6191<h5>Syntax:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6192<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]6193 declare void @llvm.gcroot(i8** %ptrloc, i8* %metadata)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6194</pre>
6195
6196<h5>Overview:</h5>
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]6197<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]6198 the code generator, and allows some metadata to be associated with it.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6199
6200<h5>Arguments:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6201<p>The first argument specifies the address of a stack object that contains the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6202 root pointer. The second pointer (which must be either a constant or a
6203 global value address) contains the meta-data to be associated with the
6204 root.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6205
6206<h5>Semantics:</h5>
Chris Lattner05d67092008-04-24 05:59:56 +0000[diff] [blame]6207<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]6208 location. At compile-time, the code generator generates information to allow
6209 the runtime to find the pointer at GC safe points. The '<tt>llvm.gcroot</tt>'
6210 intrinsic may only be used in a function which <a href="#gc">specifies a GC
6211 algorithm</a>.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6212
6213</div>
6214
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6215<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6216<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6217 <a name="int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6218</h4>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6219
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6220<div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6221
6222<h5>Syntax:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6223<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]6224 declare i8* @llvm.gcread(i8* %ObjPtr, i8** %Ptr)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6225</pre>
6226
6227<h5>Overview:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6228<p>The '<tt>llvm.gcread</tt>' intrinsic identifies reads of references from heap
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6229 locations, allowing garbage collector implementations that require read
6230 barriers.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6231
6232<h5>Arguments:</h5>
Chris Lattner80626e92006-03-14 20:02:51 +0000[diff] [blame]6233<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]6234 allocated from the garbage collector. The first object is a pointer to the
6235 start of the referenced object, if needed by the language runtime (otherwise
6236 null).</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6237
6238<h5>Semantics:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6239<p>The '<tt>llvm.gcread</tt>' intrinsic has the same semantics as a load
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6240 instruction, but may be replaced with substantially more complex code by the
6241 garbage collector runtime, as needed. The '<tt>llvm.gcread</tt>' intrinsic
6242 may only be used in a function which <a href="#gc">specifies a GC
6243 algorithm</a>.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6244
6245</div>
6246
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6247<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6248<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6249 <a name="int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6250</h4>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6251
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6252<div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6253
6254<h5>Syntax:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6255<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]6256 declare void @llvm.gcwrite(i8* %P1, i8* %Obj, i8** %P2)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6257</pre>
6258
6259<h5>Overview:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6260<p>The '<tt>llvm.gcwrite</tt>' intrinsic identifies writes of references to heap
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6261 locations, allowing garbage collector implementations that require write
6262 barriers (such as generational or reference counting collectors).</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6263
6264<h5>Arguments:</h5>
Chris Lattner80626e92006-03-14 20:02:51 +0000[diff] [blame]6265<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]6266 object to store it to, and the third is the address of the field of Obj to
6267 store to. If the runtime does not require a pointer to the object, Obj may
6268 be null.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6269
6270<h5>Semantics:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6271<p>The '<tt>llvm.gcwrite</tt>' intrinsic has the same semantics as a store
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6272 instruction, but may be replaced with substantially more complex code by the
6273 garbage collector runtime, as needed. The '<tt>llvm.gcwrite</tt>' intrinsic
6274 may only be used in a function which <a href="#gc">specifies a GC
6275 algorithm</a>.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6276
6277</div>
6278
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6279</div>
6280
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6281<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6282<h3>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6283 <a name="int_codegen">Code Generator Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6284</h3>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6285
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6286<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6287
6288<p>These intrinsics are provided by LLVM to expose special features that may
6289 only be implemented with code generator support.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6290
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6291<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6292<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6293 <a name="int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6294</h4>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6295
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6296<div>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6297
6298<h5>Syntax:</h5>
6299<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]6300 declare i8 *@llvm.returnaddress(i32 &lt;level&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6301</pre>
6302
6303<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6304<p>The '<tt>llvm.returnaddress</tt>' intrinsic attempts to compute a
6305 target-specific value indicating the return address of the current function
6306 or one of its callers.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6307
6308<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6309<p>The argument to this intrinsic indicates which function to return the address
6310 for. Zero indicates the calling function, one indicates its caller, etc.
6311 The argument is <b>required</b> to be a constant integer value.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6312
6313<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6314<p>The '<tt>llvm.returnaddress</tt>' intrinsic either returns a pointer
6315 indicating the return address of the specified call frame, or zero if it
6316 cannot be identified. The value returned by this intrinsic is likely to be
6317 incorrect or 0 for arguments other than zero, so it should only be used for
6318 debugging purposes.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6319
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6320<p>Note that calling this intrinsic does not prevent function inlining or other
6321 aggressive transformations, so the value returned may not be that of the
6322 obvious source-language caller.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6323
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6324</div>
6325
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6326<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6327<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6328 <a name="int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6329</h4>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6330
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6331<div>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6332
6333<h5>Syntax:</h5>
6334<pre>
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]6335 declare i8* @llvm.frameaddress(i32 &lt;level&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6336</pre>
6337
6338<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6339<p>The '<tt>llvm.frameaddress</tt>' intrinsic attempts to return the
6340 target-specific frame pointer value for the specified stack frame.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6341
6342<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6343<p>The argument to this intrinsic indicates which function to return the frame
6344 pointer for. Zero indicates the calling function, one indicates its caller,
6345 etc. The argument is <b>required</b> to be a constant integer value.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6346
6347<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6348<p>The '<tt>llvm.frameaddress</tt>' intrinsic either returns a pointer
6349 indicating the frame address of the specified call frame, or zero if it
6350 cannot be identified. The value returned by this intrinsic is likely to be
6351 incorrect or 0 for arguments other than zero, so it should only be used for
6352 debugging purposes.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6353
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6354<p>Note that calling this intrinsic does not prevent function inlining or other
6355 aggressive transformations, so the value returned may not be that of the
6356 obvious source-language caller.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6357
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6358</div>
6359
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]6360<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6361<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6362 <a name="int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6363</h4>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6364
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6365<div>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6366
6367<h5>Syntax:</h5>
6368<pre>
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]6369 declare i8* @llvm.stacksave()
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6370</pre>
6371
6372<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6373<p>The '<tt>llvm.stacksave</tt>' intrinsic is used to remember the current state
6374 of the function stack, for use
6375 with <a href="#int_stackrestore"> <tt>llvm.stackrestore</tt></a>. This is
6376 useful for implementing language features like scoped automatic variable
6377 sized arrays in C99.</p>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6378
6379<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6380<p>This intrinsic returns a opaque pointer value that can be passed
6381 to <a href="#int_stackrestore"><tt>llvm.stackrestore</tt></a>. When
6382 an <tt>llvm.stackrestore</tt> intrinsic is executed with a value saved
6383 from <tt>llvm.stacksave</tt>, it effectively restores the state of the stack
6384 to the state it was in when the <tt>llvm.stacksave</tt> intrinsic executed.
6385 In practice, this pops any <a href="#i_alloca">alloca</a> blocks from the
6386 stack that were allocated after the <tt>llvm.stacksave</tt> was executed.</p>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6387
6388</div>
6389
6390<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6391<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6392 <a name="int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6393</h4>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6394
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6395<div>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6396
6397<h5>Syntax:</h5>
6398<pre>
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]6399 declare void @llvm.stackrestore(i8* %ptr)
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6400</pre>
6401
6402<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6403<p>The '<tt>llvm.stackrestore</tt>' intrinsic is used to restore the state of
6404 the function stack to the state it was in when the
6405 corresponding <a href="#int_stacksave"><tt>llvm.stacksave</tt></a> intrinsic
6406 executed. This is useful for implementing language features like scoped
6407 automatic variable sized arrays in C99.</p>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6408
6409<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6410<p>See the description
6411 for <a href="#int_stacksave"><tt>llvm.stacksave</tt></a>.</p>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6412
6413</div>
6414
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6415<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6416<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6417 <a name="int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6418</h4>
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]6419
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6420<div>
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]6421
6422<h5>Syntax:</h5>
6423<pre>
Bruno Cardoso Lopes9a767332011-06-14 04:58:37 +0000[diff] [blame]6424 declare void @llvm.prefetch(i8* &lt;address&gt;, i32 &lt;rw&gt;, i32 &lt;locality&gt;, i32 &lt;cache type&gt;)
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]6425</pre>
6426
6427<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6428<p>The '<tt>llvm.prefetch</tt>' intrinsic is a hint to the code generator to
6429 insert a prefetch instruction if supported; otherwise, it is a noop.
6430 Prefetches have no effect on the behavior of the program but can change its
6431 performance characteristics.</p>
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]6432
6433<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6434<p><tt>address</tt> is the address to be prefetched, <tt>rw</tt> is the
6435 specifier determining if the fetch should be for a read (0) or write (1),
6436 and <tt>locality</tt> is a temporal locality specifier ranging from (0) - no
Bruno Cardoso Lopes9a767332011-06-14 04:58:37 +0000[diff] [blame]6437 locality, to (3) - extremely local keep in cache. The <tt>cache type</tt>
6438 specifies whether the prefetch is performed on the data (1) or instruction (0)
6439 cache. The <tt>rw</tt>, <tt>locality</tt> and <tt>cache type</tt> arguments
6440 must be constant integers.</p>
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]6441
6442<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6443<p>This intrinsic does not modify the behavior of the program. In particular,
6444 prefetches cannot trap and do not produce a value. On targets that support
6445 this intrinsic, the prefetch can provide hints to the processor cache for
6446 better performance.</p>
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]6447
6448</div>
6449
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]6450<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6451<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6452 <a name="int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6453</h4>
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]6454
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6455<div>
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]6456
6457<h5>Syntax:</h5>
6458<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]6459 declare void @llvm.pcmarker(i32 &lt;id&gt;)
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]6460</pre>
6461
6462<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6463<p>The '<tt>llvm.pcmarker</tt>' intrinsic is a method to export a Program
6464 Counter (PC) in a region of code to simulators and other tools. The method
6465 is target specific, but it is expected that the marker will use exported
6466 symbols to transmit the PC of the marker. The marker makes no guarantees
6467 that it will remain with any specific instruction after optimizations. It is
6468 possible that the presence of a marker will inhibit optimizations. The
6469 intended use is to be inserted after optimizations to allow correlations of
6470 simulation runs.</p>
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]6471
6472<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6473<p><tt>id</tt> is a numerical id identifying the marker.</p>
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]6474
6475<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6476<p>This intrinsic does not modify the behavior of the program. Backends that do
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]6477 not support this intrinsic may ignore it.</p>
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]6478
6479</div>
6480
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]6481<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6482<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6483 <a name="int_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6484</h4>
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]6485
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6486<div>
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]6487
6488<h5>Syntax:</h5>
6489<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]6490 declare i64 @llvm.readcyclecounter()
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]6491</pre>
6492
6493<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6494<p>The '<tt>llvm.readcyclecounter</tt>' intrinsic provides access to the cycle
6495 counter register (or similar low latency, high accuracy clocks) on those
6496 targets that support it. On X86, it should map to RDTSC. On Alpha, it
6497 should map to RPCC. As the backing counters overflow quickly (on the order
6498 of 9 seconds on alpha), this should only be used for small timings.</p>
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]6499
6500<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6501<p>When directly supported, reading the cycle counter should not modify any
6502 memory. Implementations are allowed to either return a application specific
6503 value or a system wide value. On backends without support, this is lowered
6504 to a constant 0.</p>
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]6505
6506</div>
6507
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6508</div>
6509
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6510<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6511<h3>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6512 <a name="int_libc">Standard C Library Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6513</h3>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6514
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6515<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6516
6517<p>LLVM provides intrinsics for a few important standard C library functions.
6518 These intrinsics allow source-language front-ends to pass information about
6519 the alignment of the pointer arguments to the code generator, providing
6520 opportunity for more efficient code generation.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6521
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6522<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6523<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6524 <a name="int_memcpy">'<tt>llvm.memcpy</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6525</h4>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6526
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6527<div>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6528
6529<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6530<p>This is an overloaded intrinsic. You can use <tt>llvm.memcpy</tt> on any
Mon P Wange88909b2010-04-07 06:35:53 +0000[diff] [blame]6531 integer bit width and for different address spaces. Not all targets support
6532 all bit widths however.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6533
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6534<pre>
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]6535 declare void @llvm.memcpy.p0i8.p0i8.i32(i8* &lt;dest&gt;, i8* &lt;src&gt;,
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6536 i32 &lt;len&gt;, i32 &lt;align&gt;, i1 &lt;isvolatile&gt;)
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]6537 declare void @llvm.memcpy.p0i8.p0i8.i64(i8* &lt;dest&gt;, i8* &lt;src&gt;,
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6538 i64 &lt;len&gt;, i32 &lt;align&gt;, i1 &lt;isvolatile&gt;)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6539</pre>
6540
6541<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6542<p>The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the
6543 source location to the destination location.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6544
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6545<p>Note that, unlike the standard libc function, the <tt>llvm.memcpy.*</tt>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6546 intrinsics do not return a value, takes extra alignment/isvolatile arguments
6547 and the pointers can be in specified address spaces.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6548
6549<h5>Arguments:</h5>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6550
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6551<p>The first argument is a pointer to the destination, the second is a pointer
6552 to the source. The third argument is an integer argument specifying the
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6553 number of bytes to copy, the fourth argument is the alignment of the
6554 source and destination locations, and the fifth is a boolean indicating a
6555 volatile access.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6556
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]6557<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]6558 then the caller guarantees that both the source and destination pointers are
6559 aligned to that boundary.</p>
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]6560
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]6561<p>If the <tt>isvolatile</tt> parameter is <tt>true</tt>, the
6562 <tt>llvm.memcpy</tt> call is a <a href="#volatile">volatile operation</a>.
6563 The detailed access behavior is not very cleanly specified and it is unwise
6564 to depend on it.</p>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6565
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6566<h5>Semantics:</h5>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6567
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6568<p>The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the
6569 source location to the destination location, which are not allowed to
6570 overlap. It copies "len" bytes of memory over. If the argument is known to
6571 be aligned to some boundary, this can be specified as the fourth argument,
6572 otherwise it should be set to 0 or 1.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6573
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6574</div>
6575
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6576<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6577<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6578 <a name="int_memmove">'<tt>llvm.memmove</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6579</h4>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6580
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6581<div>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6582
6583<h5>Syntax:</h5>
Chris Lattner824b9582008-11-21 16:42:48 +0000[diff] [blame]6584<p>This is an overloaded intrinsic. You can use llvm.memmove on any integer bit
Mon P Wange88909b2010-04-07 06:35:53 +0000[diff] [blame]6585 width and for different address space. Not all targets support all bit
6586 widths however.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6587
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6588<pre>
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]6589 declare void @llvm.memmove.p0i8.p0i8.i32(i8* &lt;dest&gt;, i8* &lt;src&gt;,
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6590 i32 &lt;len&gt;, i32 &lt;align&gt;, i1 &lt;isvolatile&gt;)
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]6591 declare void @llvm.memmove.p0i8.p0i8.i64(i8* &lt;dest&gt;, i8* &lt;src&gt;,
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6592 i64 &lt;len&gt;, i32 &lt;align&gt;, i1 &lt;isvolatile&gt;)
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6593</pre>
6594
6595<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6596<p>The '<tt>llvm.memmove.*</tt>' intrinsics move a block of memory from the
6597 source location to the destination location. It is similar to the
6598 '<tt>llvm.memcpy</tt>' intrinsic but allows the two memory locations to
6599 overlap.</p>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6600
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6601<p>Note that, unlike the standard libc function, the <tt>llvm.memmove.*</tt>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6602 intrinsics do not return a value, takes extra alignment/isvolatile arguments
6603 and the pointers can be in specified address spaces.</p>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6604
6605<h5>Arguments:</h5>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6606
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6607<p>The first argument is a pointer to the destination, the second is a pointer
6608 to the source. The third argument is an integer argument specifying the
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6609 number of bytes to copy, the fourth argument is the alignment of the
6610 source and destination locations, and the fifth is a boolean indicating a
6611 volatile access.</p>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6612
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]6613<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]6614 then the caller guarantees that the source and destination pointers are
6615 aligned to that boundary.</p>
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]6616
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]6617<p>If the <tt>isvolatile</tt> parameter is <tt>true</tt>, the
6618 <tt>llvm.memmove</tt> call is a <a href="#volatile">volatile operation</a>.
6619 The detailed access behavior is not very cleanly specified and it is unwise
6620 to depend on it.</p>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6621
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6622<h5>Semantics:</h5>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6623
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6624<p>The '<tt>llvm.memmove.*</tt>' intrinsics copy a block of memory from the
6625 source location to the destination location, which may overlap. It copies
6626 "len" bytes of memory over. If the argument is known to be aligned to some
6627 boundary, this can be specified as the fourth argument, otherwise it should
6628 be set to 0 or 1.</p>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6629
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6630</div>
6631
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6632<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6633<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6634 <a name="int_memset">'<tt>llvm.memset.*</tt>' Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6635</h4>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6636
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6637<div>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6638
6639<h5>Syntax:</h5>
Chris Lattner824b9582008-11-21 16:42:48 +0000[diff] [blame]6640<p>This is an overloaded intrinsic. You can use llvm.memset on any integer bit
John Criswellcdcbbfc2010-07-30 16:30:28 +0000[diff] [blame]6641 width and for different address spaces. However, not all targets support all
6642 bit widths.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6643
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6644<pre>
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]6645 declare void @llvm.memset.p0i8.i32(i8* &lt;dest&gt;, i8 &lt;val&gt;,
Chris Lattnerff35c3f2010-04-08 00:54:34 +0000[diff] [blame]6646 i32 &lt;len&gt;, i32 &lt;align&gt;, i1 &lt;isvolatile&gt;)
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]6647 declare void @llvm.memset.p0i8.i64(i8* &lt;dest&gt;, i8 &lt;val&gt;,
Chris Lattnerff35c3f2010-04-08 00:54:34 +0000[diff] [blame]6648 i64 &lt;len&gt;, i32 &lt;align&gt;, i1 &lt;isvolatile&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6649</pre>
6650
6651<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6652<p>The '<tt>llvm.memset.*</tt>' intrinsics fill a block of memory with a
6653 particular byte value.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6654
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6655<p>Note that, unlike the standard libc function, the <tt>llvm.memset</tt>
John Criswellcdcbbfc2010-07-30 16:30:28 +0000[diff] [blame]6656 intrinsic does not return a value and takes extra alignment/volatile
6657 arguments. Also, the destination can be in an arbitrary address space.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6658
6659<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6660<p>The first argument is a pointer to the destination to fill, the second is the
John Criswellcdcbbfc2010-07-30 16:30:28 +0000[diff] [blame]6661 byte value with which to fill it, the third argument is an integer argument
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6662 specifying the number of bytes to fill, and the fourth argument is the known
John Criswellcdcbbfc2010-07-30 16:30:28 +0000[diff] [blame]6663 alignment of the destination location.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6664
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]6665<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]6666 then the caller guarantees that the destination pointer is aligned to that
6667 boundary.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6668
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]6669<p>If the <tt>isvolatile</tt> parameter is <tt>true</tt>, the
6670 <tt>llvm.memset</tt> call is a <a href="#volatile">volatile operation</a>.
6671 The detailed access behavior is not very cleanly specified and it is unwise
6672 to depend on it.</p>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6673
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6674<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6675<p>The '<tt>llvm.memset.*</tt>' intrinsics fill "len" bytes of memory starting
6676 at the destination location. If the argument is known to be aligned to some
6677 boundary, this can be specified as the fourth argument, otherwise it should
6678 be set to 0 or 1.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6679
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6680</div>
6681
Chris Lattner32006282004-06-11 02:28:03 +0000[diff] [blame]6682<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6683<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6684 <a name="int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6685</h4>
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]6686
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6687<div>
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]6688
6689<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6690<p>This is an overloaded intrinsic. You can use <tt>llvm.sqrt</tt> on any
6691 floating point or vector of floating point type. Not all targets support all
6692 types however.</p>
6693
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]6694<pre>
Dale Johannesen408f9c12007-10-02 17:47:38 +0000[diff] [blame]6695 declare float @llvm.sqrt.f32(float %Val)
6696 declare double @llvm.sqrt.f64(double %Val)
6697 declare x86_fp80 @llvm.sqrt.f80(x86_fp80 %Val)
6698 declare fp128 @llvm.sqrt.f128(fp128 %Val)
6699 declare ppc_fp128 @llvm.sqrt.ppcf128(ppc_fp128 %Val)
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]6700</pre>
6701
6702<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6703<p>The '<tt>llvm.sqrt</tt>' intrinsics return the sqrt of the specified operand,
6704 returning the same value as the libm '<tt>sqrt</tt>' functions would.
6705 Unlike <tt>sqrt</tt> in libm, however, <tt>llvm.sqrt</tt> has undefined
6706 behavior for negative numbers other than -0.0 (which allows for better
6707 optimization, because there is no need to worry about errno being
6708 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]6709
6710<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6711<p>The argument and return value are floating point numbers of the same
6712 type.</p>
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]6713
6714<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6715<p>This function returns the sqrt of the specified operand if it is a
6716 nonnegative floating point number.</p>
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]6717
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]6718</div>
6719
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6720<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6721<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6722 <a name="int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6723</h4>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6724
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6725<div>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6726
6727<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6728<p>This is an overloaded intrinsic. You can use <tt>llvm.powi</tt> on any
6729 floating point or vector of floating point type. Not all targets support all
6730 types however.</p>
6731
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6732<pre>
Dale Johannesen408f9c12007-10-02 17:47:38 +0000[diff] [blame]6733 declare float @llvm.powi.f32(float %Val, i32 %power)
6734 declare double @llvm.powi.f64(double %Val, i32 %power)
6735 declare x86_fp80 @llvm.powi.f80(x86_fp80 %Val, i32 %power)
6736 declare fp128 @llvm.powi.f128(fp128 %Val, i32 %power)
6737 declare ppc_fp128 @llvm.powi.ppcf128(ppc_fp128 %Val, i32 %power)
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6738</pre>
6739
6740<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6741<p>The '<tt>llvm.powi.*</tt>' intrinsics return the first operand raised to the
6742 specified (positive or negative) power. The order of evaluation of
6743 multiplications is not defined. When a vector of floating point type is
6744 used, the second argument remains a scalar integer value.</p>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6745
6746<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6747<p>The second argument is an integer power, and the first is a value to raise to
6748 that power.</p>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6749
6750<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6751<p>This function returns the first value raised to the second power with an
6752 unspecified sequence of rounding operations.</p>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6753
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6754</div>
6755
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6756<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6757<h4>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6758 <a name="int_sin">'<tt>llvm.sin.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6759</h4>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6760
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6761<div>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6762
6763<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6764<p>This is an overloaded intrinsic. You can use <tt>llvm.sin</tt> on any
6765 floating point or vector of floating point type. Not all targets support all
6766 types however.</p>
6767
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6768<pre>
6769 declare float @llvm.sin.f32(float %Val)
6770 declare double @llvm.sin.f64(double %Val)
6771 declare x86_fp80 @llvm.sin.f80(x86_fp80 %Val)
6772 declare fp128 @llvm.sin.f128(fp128 %Val)
6773 declare ppc_fp128 @llvm.sin.ppcf128(ppc_fp128 %Val)
6774</pre>
6775
6776<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6777<p>The '<tt>llvm.sin.*</tt>' intrinsics return the sine of the operand.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6778
6779<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6780<p>The argument and return value are floating point numbers of the same
6781 type.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6782
6783<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6784<p>This function returns the sine of the specified operand, returning the same
6785 values as the libm <tt>sin</tt> functions would, and handles error conditions
6786 in the same way.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6787
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6788</div>
6789
6790<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6791<h4>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6792 <a name="int_cos">'<tt>llvm.cos.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6793</h4>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6794
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6795<div>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6796
6797<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6798<p>This is an overloaded intrinsic. You can use <tt>llvm.cos</tt> on any
6799 floating point or vector of floating point type. Not all targets support all
6800 types however.</p>
6801
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6802<pre>
6803 declare float @llvm.cos.f32(float %Val)
6804 declare double @llvm.cos.f64(double %Val)
6805 declare x86_fp80 @llvm.cos.f80(x86_fp80 %Val)
6806 declare fp128 @llvm.cos.f128(fp128 %Val)
6807 declare ppc_fp128 @llvm.cos.ppcf128(ppc_fp128 %Val)
6808</pre>
6809
6810<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6811<p>The '<tt>llvm.cos.*</tt>' intrinsics return the cosine of the operand.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6812
6813<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6814<p>The argument and return value are floating point numbers of the same
6815 type.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6816
6817<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6818<p>This function returns the cosine of the specified operand, returning the same
6819 values as the libm <tt>cos</tt> functions would, and handles error conditions
6820 in the same way.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6821
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6822</div>
6823
6824<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6825<h4>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6826 <a name="int_pow">'<tt>llvm.pow.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6827</h4>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6828
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6829<div>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6830
6831<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6832<p>This is an overloaded intrinsic. You can use <tt>llvm.pow</tt> on any
6833 floating point or vector of floating point type. Not all targets support all
6834 types however.</p>
6835
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6836<pre>
6837 declare float @llvm.pow.f32(float %Val, float %Power)
6838 declare double @llvm.pow.f64(double %Val, double %Power)
6839 declare x86_fp80 @llvm.pow.f80(x86_fp80 %Val, x86_fp80 %Power)
6840 declare fp128 @llvm.pow.f128(fp128 %Val, fp128 %Power)
6841 declare ppc_fp128 @llvm.pow.ppcf128(ppc_fp128 %Val, ppc_fp128 Power)
6842</pre>
6843
6844<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6845<p>The '<tt>llvm.pow.*</tt>' intrinsics return the first operand raised to the
6846 specified (positive or negative) power.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6847
6848<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6849<p>The second argument is a floating point power, and the first is a value to
6850 raise to that power.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6851
6852<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6853<p>This function returns the first value raised to the second power, returning
6854 the same values as the libm <tt>pow</tt> functions would, and handles error
6855 conditions in the same way.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6856
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6857</div>
6858
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6859</div>
6860
Dan Gohman4e9011c2011-05-23 21:13:03 +0000[diff] [blame]6861<!-- _______________________________________________________________________ -->
6862<h4>
6863 <a name="int_exp">'<tt>llvm.exp.*</tt>' Intrinsic</a>
6864</h4>
6865
6866<div>
6867
6868<h5>Syntax:</h5>
6869<p>This is an overloaded intrinsic. You can use <tt>llvm.exp</tt> on any
6870 floating point or vector of floating point type. Not all targets support all
6871 types however.</p>
6872
6873<pre>
6874 declare float @llvm.exp.f32(float %Val)
6875 declare double @llvm.exp.f64(double %Val)
6876 declare x86_fp80 @llvm.exp.f80(x86_fp80 %Val)
6877 declare fp128 @llvm.exp.f128(fp128 %Val)
6878 declare ppc_fp128 @llvm.exp.ppcf128(ppc_fp128 %Val)
6879</pre>
6880
6881<h5>Overview:</h5>
6882<p>The '<tt>llvm.exp.*</tt>' intrinsics perform the exp function.</p>
6883
6884<h5>Arguments:</h5>
6885<p>The argument and return value are floating point numbers of the same
6886 type.</p>
6887
6888<h5>Semantics:</h5>
6889<p>This function returns the same values as the libm <tt>exp</tt> functions
6890 would, and handles error conditions in the same way.</p>
6891
6892</div>
6893
6894<!-- _______________________________________________________________________ -->
6895<h4>
6896 <a name="int_log">'<tt>llvm.log.*</tt>' Intrinsic</a>
6897</h4>
6898
6899<div>
6900
6901<h5>Syntax:</h5>
6902<p>This is an overloaded intrinsic. You can use <tt>llvm.log</tt> on any
6903 floating point or vector of floating point type. Not all targets support all
6904 types however.</p>
6905
6906<pre>
6907 declare float @llvm.log.f32(float %Val)
6908 declare double @llvm.log.f64(double %Val)
6909 declare x86_fp80 @llvm.log.f80(x86_fp80 %Val)
6910 declare fp128 @llvm.log.f128(fp128 %Val)
6911 declare ppc_fp128 @llvm.log.ppcf128(ppc_fp128 %Val)
6912</pre>
6913
6914<h5>Overview:</h5>
6915<p>The '<tt>llvm.log.*</tt>' intrinsics perform the log function.</p>
6916
6917<h5>Arguments:</h5>
6918<p>The argument and return value are floating point numbers of the same
6919 type.</p>
6920
6921<h5>Semantics:</h5>
6922<p>This function returns the same values as the libm <tt>log</tt> functions
6923 would, and handles error conditions in the same way.</p>
6924
Cameron Zwarich33390842011-07-08 21:39:21 +0000[diff] [blame]6925<h4>
6926 <a name="int_fma">'<tt>llvm.fma.*</tt>' Intrinsic</a>
6927</h4>
6928
6929<div>
6930
6931<h5>Syntax:</h5>
6932<p>This is an overloaded intrinsic. You can use <tt>llvm.fma</tt> on any
6933 floating point or vector of floating point type. Not all targets support all
6934 types however.</p>
6935
6936<pre>
6937 declare float @llvm.fma.f32(float %a, float %b, float %c)
6938 declare double @llvm.fma.f64(double %a, double %b, double %c)
6939 declare x86_fp80 @llvm.fma.f80(x86_fp80 %a, x86_fp80 %b, x86_fp80 %c)
6940 declare fp128 @llvm.fma.f128(fp128 %a, fp128 %b, fp128 %c)
6941 declare ppc_fp128 @llvm.fma.ppcf128(ppc_fp128 %a, ppc_fp128 %b, ppc_fp128 %c)
6942</pre>
6943
6944<h5>Overview:</h5>
Cameron Zwarichabc43e62011-07-08 22:13:55 +0000[diff] [blame]6945<p>The '<tt>llvm.fma.*</tt>' intrinsics perform the fused multiply-add
Cameron Zwarich33390842011-07-08 21:39:21 +0000[diff] [blame]6946 operation.</p>
6947
6948<h5>Arguments:</h5>
6949<p>The argument and return value are floating point numbers of the same
6950 type.</p>
6951
6952<h5>Semantics:</h5>
6953<p>This function returns the same values as the libm <tt>fma</tt> functions
6954 would.</p>
6955
Dan Gohman4e9011c2011-05-23 21:13:03 +0000[diff] [blame]6956</div>
6957
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6958<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6959<h3>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]6960 <a name="int_manip">Bit Manipulation Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6961</h3>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6962
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6963<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6964
6965<p>LLVM provides intrinsics for a few important bit manipulation operations.
6966 These allow efficient code generation for some algorithms.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6967
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6968<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6969<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6970 <a name="int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6971</h4>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]6972
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6973<div>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]6974
6975<h5>Syntax:</h5>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]6976<p>This is an overloaded intrinsic function. You can use bswap on any integer
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6977 type that is an even number of bytes (i.e. BitWidth % 16 == 0).</p>
6978
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]6979<pre>
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]6980 declare i16 @llvm.bswap.i16(i16 &lt;id&gt;)
6981 declare i32 @llvm.bswap.i32(i32 &lt;id&gt;)
6982 declare i64 @llvm.bswap.i64(i64 &lt;id&gt;)
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]6983</pre>
6984
6985<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6986<p>The '<tt>llvm.bswap</tt>' family of intrinsics is used to byte swap integer
6987 values with an even number of bytes (positive multiple of 16 bits). These
6988 are useful for performing operations on data that is not in the target's
6989 native byte order.</p>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]6990
6991<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6992<p>The <tt>llvm.bswap.i16</tt> intrinsic returns an i16 value that has the high
6993 and low byte of the input i16 swapped. Similarly,
6994 the <tt>llvm.bswap.i32</tt> intrinsic returns an i32 value that has the four
6995 bytes of the input i32 swapped, so that if the input bytes are numbered 0, 1,
6996 2, 3 then the returned i32 will have its bytes in 3, 2, 1, 0 order.
6997 The <tt>llvm.bswap.i48</tt>, <tt>llvm.bswap.i64</tt> and other intrinsics
6998 extend this concept to additional even-byte lengths (6 bytes, 8 bytes and
6999 more, respectively).</p>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]7000
7001</div>
7002
7003<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7004<h4>
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]7005 <a name="int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7006</h4>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7007
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7008<div>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7009
7010<h5>Syntax:</h5>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]7011<p>This is an overloaded intrinsic. You can use llvm.ctpop on any integer bit
Owen Andersonf1ac4652011-07-01 21:52:38 +0000[diff] [blame]7012 width, or on any vector with integer elements. Not all targets support all
7013 bit widths or vector types, however.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7014
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7015<pre>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7016 declare i8 @llvm.ctpop.i8(i8 &lt;src&gt;)
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]7017 declare i16 @llvm.ctpop.i16(i16 &lt;src&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]7018 declare i32 @llvm.ctpop.i32(i32 &lt;src&gt;)
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]7019 declare i64 @llvm.ctpop.i64(i64 &lt;src&gt;)
7020 declare i256 @llvm.ctpop.i256(i256 &lt;src&gt;)
Owen Andersonf1ac4652011-07-01 21:52:38 +0000[diff] [blame]7021 declare &lt;2 x i32&gt; @llvm.ctpop.v2i32(&lt;2 x i32&gt; &lt;src&gt;)
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7022</pre>
7023
7024<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7025<p>The '<tt>llvm.ctpop</tt>' family of intrinsics counts the number of bits set
7026 in a value.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7027
7028<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7029<p>The only argument is the value to be counted. The argument may be of any
Owen Andersonf1ac4652011-07-01 21:52:38 +0000[diff] [blame]7030 integer type, or a vector with integer elements.
7031 The return type must match the argument type.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7032
7033<h5>Semantics:</h5>
Owen Andersonf1ac4652011-07-01 21:52:38 +0000[diff] [blame]7034<p>The '<tt>llvm.ctpop</tt>' intrinsic counts the 1's in a variable, or within each
7035 element of a vector.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7036
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7037</div>
7038
7039<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7040<h4>
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]7041 <a name="int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7042</h4>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7043
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7044<div>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7045
7046<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7047<p>This is an overloaded intrinsic. You can use <tt>llvm.ctlz</tt> on any
Owen Andersonf1ac4652011-07-01 21:52:38 +0000[diff] [blame]7048 integer bit width, or any vector whose elements are integers. Not all
7049 targets support all bit widths or vector types, however.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7050
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7051<pre>
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]7052 declare i8 @llvm.ctlz.i8 (i8 &lt;src&gt;)
7053 declare i16 @llvm.ctlz.i16(i16 &lt;src&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]7054 declare i32 @llvm.ctlz.i32(i32 &lt;src&gt;)
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]7055 declare i64 @llvm.ctlz.i64(i64 &lt;src&gt;)
7056 declare i256 @llvm.ctlz.i256(i256 &lt;src&gt;)
Owen Andersonf1ac4652011-07-01 21:52:38 +0000[diff] [blame]7057 declare &lt;2 x i32&gt; @llvm.ctlz.v2i32(&lt;2 x i32&gt; &lt;src;gt)
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7058</pre>
7059
7060<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7061<p>The '<tt>llvm.ctlz</tt>' family of intrinsic functions counts the number of
7062 leading zeros in a variable.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7063
7064<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7065<p>The only argument is the value to be counted. The argument may be of any
Owen Andersonf1ac4652011-07-01 21:52:38 +0000[diff] [blame]7066 integer type, or any vector type with integer element type.
7067 The return type must match the argument type.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7068
7069<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7070<p>The '<tt>llvm.ctlz</tt>' intrinsic counts the leading (most significant)
Owen Andersonf1ac4652011-07-01 21:52:38 +0000[diff] [blame]7071 zeros in a variable, or within each element of the vector if the operation
7072 is of vector type. If the src == 0 then the result is the size in bits of
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7073 the type of src. For example, <tt>llvm.ctlz(i32 2) = 30</tt>.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7074
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7075</div>
Chris Lattner32006282004-06-11 02:28:03 +0000[diff] [blame]7076
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]7077<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7078<h4>
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]7079 <a name="int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7080</h4>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]7081
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7082<div>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]7083
7084<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7085<p>This is an overloaded intrinsic. You can use <tt>llvm.cttz</tt> on any
Owen Andersonf1ac4652011-07-01 21:52:38 +0000[diff] [blame]7086 integer bit width, or any vector of integer elements. Not all targets
7087 support all bit widths or vector types, however.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7088
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]7089<pre>
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]7090 declare i8 @llvm.cttz.i8 (i8 &lt;src&gt;)
7091 declare i16 @llvm.cttz.i16(i16 &lt;src&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]7092 declare i32 @llvm.cttz.i32(i32 &lt;src&gt;)
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]7093 declare i64 @llvm.cttz.i64(i64 &lt;src&gt;)
7094 declare i256 @llvm.cttz.i256(i256 &lt;src&gt;)
Owen Andersonf1ac4652011-07-01 21:52:38 +0000[diff] [blame]7095 declase &lt;2 x i32&gt; @llvm.cttz.v2i32(&lt;2 x i32&gt; &lt;src&gt;)
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]7096</pre>
7097
7098<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7099<p>The '<tt>llvm.cttz</tt>' family of intrinsic functions counts the number of
7100 trailing zeros.</p>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]7101
7102<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7103<p>The only argument is the value to be counted. The argument may be of any
Owen Andersonf1ac4652011-07-01 21:52:38 +0000[diff] [blame]7104 integer type, or a vectory with integer element type.. The return type
7105 must match the argument type.</p>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]7106
7107<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7108<p>The '<tt>llvm.cttz</tt>' intrinsic counts the trailing (least significant)
Owen Andersonf1ac4652011-07-01 21:52:38 +0000[diff] [blame]7109 zeros in a variable, or within each element of a vector.
7110 If the src == 0 then the result is the size in bits of
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7111 the type of src. For example, <tt>llvm.cttz(2) = 1</tt>.</p>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]7112
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]7113</div>
7114
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7115</div>
7116
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]7117<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7118<h3>
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]7119 <a name="int_overflow">Arithmetic with Overflow Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7120</h3>
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]7121
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7122<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7123
7124<p>LLVM provides intrinsics for some arithmetic with overflow operations.</p>
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]7125
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7126<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7127<h4>
7128 <a name="int_sadd_overflow">
7129 '<tt>llvm.sadd.with.overflow.*</tt>' Intrinsics
7130 </a>
7131</h4>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7132
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7133<div>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7134
7135<h5>Syntax:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7136<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]7137 on any integer bit width.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7138
7139<pre>
7140 declare {i16, i1} @llvm.sadd.with.overflow.i16(i16 %a, i16 %b)
7141 declare {i32, i1} @llvm.sadd.with.overflow.i32(i32 %a, i32 %b)
7142 declare {i64, i1} @llvm.sadd.with.overflow.i64(i64 %a, i64 %b)
7143</pre>
7144
7145<h5>Overview:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7146<p>The '<tt>llvm.sadd.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7147 a signed addition of the two arguments, and indicate whether an overflow
7148 occurred during the signed summation.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7149
7150<h5>Arguments:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7151<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]7152 be of integer types of any bit width, but they must have the same bit
7153 width. The second element of the result structure must be of
7154 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
7155 undergo signed addition.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7156
7157<h5>Semantics:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7158<p>The '<tt>llvm.sadd.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7159 a signed addition of the two variables. They return a structure &mdash; the
7160 first element of which is the signed summation, and the second element of
7161 which is a bit specifying if the signed summation resulted in an
7162 overflow.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7163
7164<h5>Examples:</h5>
7165<pre>
7166 %res = call {i32, i1} @llvm.sadd.with.overflow.i32(i32 %a, i32 %b)
7167 %sum = extractvalue {i32, i1} %res, 0
7168 %obit = extractvalue {i32, i1} %res, 1
7169 br i1 %obit, label %overflow, label %normal
7170</pre>
7171
7172</div>
7173
7174<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7175<h4>
7176 <a name="int_uadd_overflow">
7177 '<tt>llvm.uadd.with.overflow.*</tt>' Intrinsics
7178 </a>
7179</h4>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7180
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7181<div>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7182
7183<h5>Syntax:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7184<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]7185 on any integer bit width.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7186
7187<pre>
7188 declare {i16, i1} @llvm.uadd.with.overflow.i16(i16 %a, i16 %b)
7189 declare {i32, i1} @llvm.uadd.with.overflow.i32(i32 %a, i32 %b)
7190 declare {i64, i1} @llvm.uadd.with.overflow.i64(i64 %a, i64 %b)
7191</pre>
7192
7193<h5>Overview:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7194<p>The '<tt>llvm.uadd.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7195 an unsigned addition of the two arguments, and indicate whether a carry
7196 occurred during the unsigned summation.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7197
7198<h5>Arguments:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7199<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]7200 be of integer types of any bit width, but they must have the same bit
7201 width. The second element of the result structure must be of
7202 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
7203 undergo unsigned addition.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7204
7205<h5>Semantics:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7206<p>The '<tt>llvm.uadd.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7207 an unsigned addition of the two arguments. They return a structure &mdash;
7208 the first element of which is the sum, and the second element of which is a
7209 bit specifying if the unsigned summation resulted in a carry.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7210
7211<h5>Examples:</h5>
7212<pre>
7213 %res = call {i32, i1} @llvm.uadd.with.overflow.i32(i32 %a, i32 %b)
7214 %sum = extractvalue {i32, i1} %res, 0
7215 %obit = extractvalue {i32, i1} %res, 1
7216 br i1 %obit, label %carry, label %normal
7217</pre>
7218
7219</div>
7220
7221<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7222<h4>
7223 <a name="int_ssub_overflow">
7224 '<tt>llvm.ssub.with.overflow.*</tt>' Intrinsics
7225 </a>
7226</h4>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7227
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7228<div>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7229
7230<h5>Syntax:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7231<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]7232 on any integer bit width.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7233
7234<pre>
7235 declare {i16, i1} @llvm.ssub.with.overflow.i16(i16 %a, i16 %b)
7236 declare {i32, i1} @llvm.ssub.with.overflow.i32(i32 %a, i32 %b)
7237 declare {i64, i1} @llvm.ssub.with.overflow.i64(i64 %a, i64 %b)
7238</pre>
7239
7240<h5>Overview:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7241<p>The '<tt>llvm.ssub.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7242 a signed subtraction of the two arguments, and indicate whether an overflow
7243 occurred during the signed subtraction.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7244
7245<h5>Arguments:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7246<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]7247 be of integer types of any bit width, but they must have the same bit
7248 width. The second element of the result structure must be of
7249 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
7250 undergo signed subtraction.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7251
7252<h5>Semantics:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7253<p>The '<tt>llvm.ssub.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7254 a signed subtraction of the two arguments. They return a structure &mdash;
7255 the first element of which is the subtraction, and the second element of
7256 which is a bit specifying if the signed subtraction resulted in an
7257 overflow.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7258
7259<h5>Examples:</h5>
7260<pre>
7261 %res = call {i32, i1} @llvm.ssub.with.overflow.i32(i32 %a, i32 %b)
7262 %sum = extractvalue {i32, i1} %res, 0
7263 %obit = extractvalue {i32, i1} %res, 1
7264 br i1 %obit, label %overflow, label %normal
7265</pre>
7266
7267</div>
7268
7269<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7270<h4>
7271 <a name="int_usub_overflow">
7272 '<tt>llvm.usub.with.overflow.*</tt>' Intrinsics
7273 </a>
7274</h4>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7275
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7276<div>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7277
7278<h5>Syntax:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7279<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]7280 on any integer bit width.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7281
7282<pre>
7283 declare {i16, i1} @llvm.usub.with.overflow.i16(i16 %a, i16 %b)
7284 declare {i32, i1} @llvm.usub.with.overflow.i32(i32 %a, i32 %b)
7285 declare {i64, i1} @llvm.usub.with.overflow.i64(i64 %a, i64 %b)
7286</pre>
7287
7288<h5>Overview:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7289<p>The '<tt>llvm.usub.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7290 an unsigned subtraction of the two arguments, and indicate whether an
7291 overflow occurred during the unsigned subtraction.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7292
7293<h5>Arguments:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7294<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]7295 be of integer types of any bit width, but they must have the same bit
7296 width. The second element of the result structure must be of
7297 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
7298 undergo unsigned subtraction.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7299
7300<h5>Semantics:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7301<p>The '<tt>llvm.usub.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7302 an unsigned subtraction of the two arguments. They return a structure &mdash;
7303 the first element of which is the subtraction, and the second element of
7304 which is a bit specifying if the unsigned subtraction resulted in an
7305 overflow.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7306
7307<h5>Examples:</h5>
7308<pre>
7309 %res = call {i32, i1} @llvm.usub.with.overflow.i32(i32 %a, i32 %b)
7310 %sum = extractvalue {i32, i1} %res, 0
7311 %obit = extractvalue {i32, i1} %res, 1
7312 br i1 %obit, label %overflow, label %normal
7313</pre>
7314
7315</div>
7316
7317<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7318<h4>
7319 <a name="int_smul_overflow">
7320 '<tt>llvm.smul.with.overflow.*</tt>' Intrinsics
7321 </a>
7322</h4>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7323
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7324<div>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7325
7326<h5>Syntax:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7327<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]7328 on any integer bit width.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7329
7330<pre>
7331 declare {i16, i1} @llvm.smul.with.overflow.i16(i16 %a, i16 %b)
7332 declare {i32, i1} @llvm.smul.with.overflow.i32(i32 %a, i32 %b)
7333 declare {i64, i1} @llvm.smul.with.overflow.i64(i64 %a, i64 %b)
7334</pre>
7335
7336<h5>Overview:</h5>
7337
7338<p>The '<tt>llvm.smul.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7339 a signed multiplication of the two arguments, and indicate whether an
7340 overflow occurred during the signed multiplication.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7341
7342<h5>Arguments:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7343<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]7344 be of integer types of any bit width, but they must have the same bit
7345 width. The second element of the result structure must be of
7346 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
7347 undergo signed multiplication.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7348
7349<h5>Semantics:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7350<p>The '<tt>llvm.smul.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7351 a signed multiplication of the two arguments. They return a structure &mdash;
7352 the first element of which is the multiplication, and the second element of
7353 which is a bit specifying if the signed multiplication resulted in an
7354 overflow.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7355
7356<h5>Examples:</h5>
7357<pre>
7358 %res = call {i32, i1} @llvm.smul.with.overflow.i32(i32 %a, i32 %b)
7359 %sum = extractvalue {i32, i1} %res, 0
7360 %obit = extractvalue {i32, i1} %res, 1
7361 br i1 %obit, label %overflow, label %normal
7362</pre>
7363
Reid Spencerf86037f2007-04-11 23:23:49 +0000[diff] [blame]7364</div>
7365
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7366<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7367<h4>
7368 <a name="int_umul_overflow">
7369 '<tt>llvm.umul.with.overflow.*</tt>' Intrinsics
7370 </a>
7371</h4>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7372
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7373<div>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7374
7375<h5>Syntax:</h5>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7376<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]7377 on any integer bit width.</p>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7378
7379<pre>
7380 declare {i16, i1} @llvm.umul.with.overflow.i16(i16 %a, i16 %b)
7381 declare {i32, i1} @llvm.umul.with.overflow.i32(i32 %a, i32 %b)
7382 declare {i64, i1} @llvm.umul.with.overflow.i64(i64 %a, i64 %b)
7383</pre>
7384
7385<h5>Overview:</h5>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7386<p>The '<tt>llvm.umul.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7387 a unsigned multiplication of the two arguments, and indicate whether an
7388 overflow occurred during the unsigned multiplication.</p>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7389
7390<h5>Arguments:</h5>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7391<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]7392 be of integer types of any bit width, but they must have the same bit
7393 width. The second element of the result structure must be of
7394 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
7395 undergo unsigned multiplication.</p>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7396
7397<h5>Semantics:</h5>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7398<p>The '<tt>llvm.umul.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7399 an unsigned multiplication of the two arguments. They return a structure
7400 &mdash; the first element of which is the multiplication, and the second
7401 element of which is a bit specifying if the unsigned multiplication resulted
7402 in an overflow.</p>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7403
7404<h5>Examples:</h5>
7405<pre>
7406 %res = call {i32, i1} @llvm.umul.with.overflow.i32(i32 %a, i32 %b)
7407 %sum = extractvalue {i32, i1} %res, 0
7408 %obit = extractvalue {i32, i1} %res, 1
7409 br i1 %obit, label %overflow, label %normal
7410</pre>
7411
7412</div>
7413
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7414</div>
7415
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]7416<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7417<h3>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]7418 <a name="int_fp16">Half Precision Floating Point Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7419</h3>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]7420
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7421<div>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]7422
Chris Lattner0cec9c82010-03-15 04:12:21 +0000[diff] [blame]7423<p>Half precision floating point is a storage-only format. This means that it is
7424 a dense encoding (in memory) but does not support computation in the
7425 format.</p>
Chris Lattner82c3dc62010-03-14 23:03:31 +0000[diff] [blame]7426
Chris Lattner0cec9c82010-03-15 04:12:21 +0000[diff] [blame]7427<p>This means that code must first load the half-precision floating point
Chris Lattner82c3dc62010-03-14 23:03:31 +0000[diff] [blame]7428 value as an i16, then convert it to float with <a
7429 href="#int_convert_from_fp16"><tt>llvm.convert.from.fp16</tt></a>.
7430 Computation can then be performed on the float value (including extending to
Chris Lattner0cec9c82010-03-15 04:12:21 +0000[diff] [blame]7431 double etc). To store the value back to memory, it is first converted to
7432 float if needed, then converted to i16 with
Chris Lattner82c3dc62010-03-14 23:03:31 +0000[diff] [blame]7433 <a href="#int_convert_to_fp16"><tt>llvm.convert.to.fp16</tt></a>, then
7434 storing as an i16 value.</p>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]7435
7436<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7437<h4>
7438 <a name="int_convert_to_fp16">
7439 '<tt>llvm.convert.to.fp16</tt>' Intrinsic
7440 </a>
7441</h4>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]7442
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7443<div>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]7444
7445<h5>Syntax:</h5>
7446<pre>
7447 declare i16 @llvm.convert.to.fp16(f32 %a)
7448</pre>
7449
7450<h5>Overview:</h5>
7451<p>The '<tt>llvm.convert.to.fp16</tt>' intrinsic function performs
7452 a conversion from single precision floating point format to half precision
7453 floating point format.</p>
7454
7455<h5>Arguments:</h5>
7456<p>The intrinsic function contains single argument - the value to be
7457 converted.</p>
7458
7459<h5>Semantics:</h5>
7460<p>The '<tt>llvm.convert.to.fp16</tt>' intrinsic function performs
7461 a conversion from single precision floating point format to half precision
Chris Lattner0cec9c82010-03-15 04:12:21 +0000[diff] [blame]7462 floating point format. The return value is an <tt>i16</tt> which
Chris Lattner82c3dc62010-03-14 23:03:31 +0000[diff] [blame]7463 contains the converted number.</p>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]7464
7465<h5>Examples:</h5>
7466<pre>
7467 %res = call i16 @llvm.convert.to.fp16(f32 %a)
7468 store i16 %res, i16* @x, align 2
7469</pre>
7470
7471</div>
7472
7473<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7474<h4>
7475 <a name="int_convert_from_fp16">
7476 '<tt>llvm.convert.from.fp16</tt>' Intrinsic
7477 </a>
7478</h4>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]7479
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7480<div>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]7481
7482<h5>Syntax:</h5>
7483<pre>
7484 declare f32 @llvm.convert.from.fp16(i16 %a)
7485</pre>
7486
7487<h5>Overview:</h5>
7488<p>The '<tt>llvm.convert.from.fp16</tt>' intrinsic function performs
7489 a conversion from half precision floating point format to single precision
7490 floating point format.</p>
7491
7492<h5>Arguments:</h5>
7493<p>The intrinsic function contains single argument - the value to be
7494 converted.</p>
7495
7496<h5>Semantics:</h5>
7497<p>The '<tt>llvm.convert.from.fp16</tt>' intrinsic function performs a
Chris Lattner0cec9c82010-03-15 04:12:21 +0000[diff] [blame]7498 conversion from half single precision floating point format to single
Chris Lattner82c3dc62010-03-14 23:03:31 +0000[diff] [blame]7499 precision floating point format. The input half-float value is represented by
7500 an <tt>i16</tt> value.</p>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]7501
7502<h5>Examples:</h5>
7503<pre>
7504 %a = load i16* @x, align 2
7505 %res = call f32 @llvm.convert.from.fp16(i16 %a)
7506</pre>
7507
7508</div>
7509
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7510</div>
7511
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]7512<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7513<h3>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]7514 <a name="int_debugger">Debugger Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7515</h3>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]7516
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7517<div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]7518
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7519<p>The LLVM debugger intrinsics (which all start with <tt>llvm.dbg.</tt>
7520 prefix), are described in
7521 the <a href="SourceLevelDebugging.html#format_common_intrinsics">LLVM Source
7522 Level Debugging</a> document.</p>
7523
7524</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]7525
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]7526<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7527<h3>
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]7528 <a name="int_eh">Exception Handling Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7529</h3>
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]7530
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7531<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7532
7533<p>The LLVM exception handling intrinsics (which all start with
7534 <tt>llvm.eh.</tt> prefix), are described in
7535 the <a href="ExceptionHandling.html#format_common_intrinsics">LLVM Exception
7536 Handling</a> document.</p>
7537
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]7538</div>
7539
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]7540<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7541<h3>
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]7542 <a name="int_trampoline">Trampoline Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7543</h3>
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]7544
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7545<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7546
7547<p>This intrinsic makes it possible to excise one parameter, marked with
Dan Gohmanff235352010-07-02 23:18:08 +0000[diff] [blame]7548 the <a href="#nest"><tt>nest</tt></a> attribute, from a function.
7549 The result is a callable
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7550 function pointer lacking the nest parameter - the caller does not need to
7551 provide a value for it. Instead, the value to use is stored in advance in a
7552 "trampoline", a block of memory usually allocated on the stack, which also
7553 contains code to splice the nest value into the argument list. This is used
7554 to implement the GCC nested function address extension.</p>
7555
7556<p>For example, if the function is
7557 <tt>i32 f(i8* nest %c, i32 %x, i32 %y)</tt> then the resulting function
7558 pointer has signature <tt>i32 (i32, i32)*</tt>. It can be created as
7559 follows:</p>
7560
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]7561<pre class="doc_code">
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]7562 %tramp = alloca [10 x i8], align 4 ; size and alignment only correct for X86
7563 %tramp1 = getelementptr [10 x i8]* %tramp, i32 0, i32 0
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7564 %p = call i8* @llvm.init.trampoline(i8* %tramp1, i8* bitcast (i32 (i8* nest , i32, i32)* @f to i8*), i8* %nval)
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]7565 %fp = bitcast i8* %p to i32 (i32, i32)*
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]7566</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7567
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7568<p>The call <tt>%val = call i32 %fp(i32 %x, i32 %y)</tt> is then equivalent
7569 to <tt>%val = call i32 %f(i8* %nval, i32 %x, i32 %y)</tt>.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7570
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]7571<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7572<h4>
7573 <a name="int_it">
7574 '<tt>llvm.init.trampoline</tt>' Intrinsic
7575 </a>
7576</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7577
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7578<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7579
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]7580<h5>Syntax:</h5>
7581<pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7582 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]7583</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7584
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]7585<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7586<p>This fills the memory pointed to by <tt>tramp</tt> with code and returns a
7587 function pointer suitable for executing it.</p>
7588
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]7589<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7590<p>The <tt>llvm.init.trampoline</tt> intrinsic takes three arguments, all
7591 pointers. The <tt>tramp</tt> argument must point to a sufficiently large and
7592 sufficiently aligned block of memory; this memory is written to by the
7593 intrinsic. Note that the size and the alignment are target-specific - LLVM
7594 currently provides no portable way of determining them, so a front-end that
7595 generates this intrinsic needs to have some target-specific knowledge.
7596 The <tt>func</tt> argument must hold a function bitcast to
7597 an <tt>i8*</tt>.</p>
7598
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]7599<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7600<p>The block of memory pointed to by <tt>tramp</tt> is filled with target
7601 dependent code, turning it into a function. A pointer to this function is
7602 returned, but needs to be bitcast to an <a href="#int_trampoline">appropriate
7603 function pointer type</a> before being called. The new function's signature
7604 is the same as that of <tt>func</tt> with any arguments marked with
7605 the <tt>nest</tt> attribute removed. At most one such <tt>nest</tt> argument
7606 is allowed, and it must be of pointer type. Calling the new function is
7607 equivalent to calling <tt>func</tt> with the same argument list, but
7608 with <tt>nval</tt> used for the missing <tt>nest</tt> argument. If, after
7609 calling <tt>llvm.init.trampoline</tt>, the memory pointed to
7610 by <tt>tramp</tt> is modified, then the effect of any later call to the
7611 returned function pointer is undefined.</p>
7612
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]7613</div>
7614
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7615</div>
7616
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]7617<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7618<h3>
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7619 <a name="int_atomics">Atomic Operations and Synchronization Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7620</h3>
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7621
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7622<div>
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7623
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7624<p>These intrinsic functions expand the "universal IR" of LLVM to represent
7625 hardware constructs for atomic operations and memory synchronization. This
7626 provides an interface to the hardware, not an interface to the programmer. It
7627 is aimed at a low enough level to allow any programming models or APIs
7628 (Application Programming Interfaces) which need atomic behaviors to map
7629 cleanly onto it. It is also modeled primarily on hardware behavior. Just as
7630 hardware provides a "universal IR" for source languages, it also provides a
7631 starting point for developing a "universal" atomic operation and
7632 synchronization IR.</p>
7633
7634<p>These do <em>not</em> form an API such as high-level threading libraries,
7635 software transaction memory systems, atomic primitives, and intrinsic
7636 functions as found in BSD, GNU libc, atomic_ops, APR, and other system and
7637 application libraries. The hardware interface provided by LLVM should allow
7638 a clean implementation of all of these APIs and parallel programming models.
7639 No one model or paradigm should be selected above others unless the hardware
7640 itself ubiquitously does so.</p>
7641
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7642<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7643<h4>
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7644 <a name="int_memory_barrier">'<tt>llvm.memory.barrier</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7645</h4>
7646
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7647<div>
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7648<h5>Syntax:</h5>
7649<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7650 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]7651</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7652
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7653<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7654<p>The <tt>llvm.memory.barrier</tt> intrinsic guarantees ordering between
7655 specific pairs of memory access types.</p>
7656
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7657<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7658<p>The <tt>llvm.memory.barrier</tt> intrinsic requires five boolean arguments.
7659 The first four arguments enables a specific barrier as listed below. The
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]7660 fifth argument specifies that the barrier applies to io or device or uncached
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7661 memory.</p>
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7662
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7663<ul>
7664 <li><tt>ll</tt>: load-load barrier</li>
7665 <li><tt>ls</tt>: load-store barrier</li>
7666 <li><tt>sl</tt>: store-load barrier</li>
7667 <li><tt>ss</tt>: store-store barrier</li>
7668 <li><tt>device</tt>: barrier applies to device and uncached memory also.</li>
7669</ul>
7670
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7671<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7672<p>This intrinsic causes the system to enforce some ordering constraints upon
7673 the loads and stores of the program. This barrier does not
7674 indicate <em>when</em> any events will occur, it only enforces
7675 an <em>order</em> in which they occur. For any of the specified pairs of load
7676 and store operations (f.ex. load-load, or store-load), all of the first
7677 operations preceding the barrier will complete before any of the second
7678 operations succeeding the barrier begin. Specifically the semantics for each
7679 pairing is as follows:</p>
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7680
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7681<ul>
7682 <li><tt>ll</tt>: All loads before the barrier must complete before any load
7683 after the barrier begins.</li>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]7684 <li><tt>ls</tt>: All loads before the barrier must complete before any
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7685 store after the barrier begins.</li>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]7686 <li><tt>ss</tt>: All stores before the barrier must complete before any
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7687 store after the barrier begins.</li>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]7688 <li><tt>sl</tt>: All stores before the barrier must complete before any
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7689 load after the barrier begins.</li>
7690</ul>
7691
7692<p>These semantics are applied with a logical "and" behavior when more than one
7693 is enabled in a single memory barrier intrinsic.</p>
7694
7695<p>Backends may implement stronger barriers than those requested when they do
7696 not support as fine grained a barrier as requested. Some architectures do
7697 not need all types of barriers and on such architectures, these become
7698 noops.</p>
7699
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7700<h5>Example:</h5>
7701<pre>
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]7702%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
7703%ptr = bitcast i8* %mallocP to i32*
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7704 store i32 4, %ptr
7705
7706%result1 = load i32* %ptr <i>; yields {i32}:result1 = 4</i>
Evan Cheng0b0669a2011-06-29 17:14:00 +0000[diff] [blame]7707 call void @llvm.memory.barrier(i1 false, i1 true, i1 false, i1 false, i1 true)
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7708 <i>; guarantee the above finishes</i>
7709 store i32 8, %ptr <i>; before this begins</i>
7710</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7711
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7712</div>
7713
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7714<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7715<h4>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7716 <a name="int_atomic_cmp_swap">'<tt>llvm.atomic.cmp.swap.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7717</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7718
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7719<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7720
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7721<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7722<p>This is an overloaded intrinsic. You can use <tt>llvm.atomic.cmp.swap</tt> on
7723 any integer bit width and for different address spaces. Not all targets
7724 support all bit widths however.</p>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7725
7726<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7727 declare i8 @llvm.atomic.cmp.swap.i8.p0i8(i8* &lt;ptr&gt;, i8 &lt;cmp&gt;, i8 &lt;val&gt;)
7728 declare i16 @llvm.atomic.cmp.swap.i16.p0i16(i16* &lt;ptr&gt;, i16 &lt;cmp&gt;, i16 &lt;val&gt;)
7729 declare i32 @llvm.atomic.cmp.swap.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;cmp&gt;, i32 &lt;val&gt;)
7730 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]7731</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7732
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7733<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7734<p>This loads a value in memory and compares it to a given value. If they are
7735 equal, it stores a new value into the memory.</p>
7736
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7737<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7738<p>The <tt>llvm.atomic.cmp.swap</tt> intrinsic takes three arguments. The result
7739 as well as both <tt>cmp</tt> and <tt>val</tt> must be integer values with the
7740 same bit width. The <tt>ptr</tt> argument must be a pointer to a value of
7741 this integer type. While any bit width integer may be used, targets may only
7742 lower representations they support in hardware.</p>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7743
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7744<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7745<p>This entire intrinsic must be executed atomically. It first loads the value
7746 in memory pointed to by <tt>ptr</tt> and compares it with the
7747 value <tt>cmp</tt>. If they are equal, <tt>val</tt> is stored into the
7748 memory. The loaded value is yielded in all cases. This provides the
7749 equivalent of an atomic compare-and-swap operation within the SSA
7750 framework.</p>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7751
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7752<h5>Examples:</h5>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7753<pre>
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]7754%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
7755%ptr = bitcast i8* %mallocP to i32*
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7756 store i32 4, %ptr
7757
7758%val1 = add i32 4, 4
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7759%result1 = call i32 @llvm.atomic.cmp.swap.i32.p0i32(i32* %ptr, i32 4, %val1)
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7760 <i>; yields {i32}:result1 = 4</i>
7761%stored1 = icmp eq i32 %result1, 4 <i>; yields {i1}:stored1 = true</i>
7762%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 8</i>
7763
7764%val2 = add i32 1, 1
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7765%result2 = call i32 @llvm.atomic.cmp.swap.i32.p0i32(i32* %ptr, i32 5, %val2)
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7766 <i>; yields {i32}:result2 = 8</i>
7767%stored2 = icmp eq i32 %result2, 5 <i>; yields {i1}:stored2 = false</i>
7768
7769%memval2 = load i32* %ptr <i>; yields {i32}:memval2 = 8</i>
7770</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7771
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7772</div>
7773
7774<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7775<h4>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7776 <a name="int_atomic_swap">'<tt>llvm.atomic.swap.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7777</h4>
7778
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7779<div>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7780<h5>Syntax:</h5>
7781
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7782<p>This is an overloaded intrinsic. You can use <tt>llvm.atomic.swap</tt> on any
7783 integer bit width. Not all targets support all bit widths however.</p>
7784
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7785<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7786 declare i8 @llvm.atomic.swap.i8.p0i8(i8* &lt;ptr&gt;, i8 &lt;val&gt;)
7787 declare i16 @llvm.atomic.swap.i16.p0i16(i16* &lt;ptr&gt;, i16 &lt;val&gt;)
7788 declare i32 @llvm.atomic.swap.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;val&gt;)
7789 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]7790</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7791
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7792<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7793<p>This intrinsic loads the value stored in memory at <tt>ptr</tt> and yields
7794 the value from memory. It then stores the value in <tt>val</tt> in the memory
7795 at <tt>ptr</tt>.</p>
7796
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7797<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7798<p>The <tt>llvm.atomic.swap</tt> intrinsic takes two arguments. Both
7799 the <tt>val</tt> argument and the result must be integers of the same bit
7800 width. The first argument, <tt>ptr</tt>, must be a pointer to a value of this
7801 integer type. The targets may only lower integer representations they
7802 support.</p>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7803
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7804<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7805<p>This intrinsic loads the value pointed to by <tt>ptr</tt>, yields it, and
7806 stores <tt>val</tt> back into <tt>ptr</tt> atomically. This provides the
7807 equivalent of an atomic swap operation within the SSA framework.</p>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7808
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7809<h5>Examples:</h5>
7810<pre>
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]7811%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
7812%ptr = bitcast i8* %mallocP to i32*
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7813 store i32 4, %ptr
7814
7815%val1 = add i32 4, 4
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7816%result1 = call i32 @llvm.atomic.swap.i32.p0i32(i32* %ptr, i32 %val1)
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7817 <i>; yields {i32}:result1 = 4</i>
7818%stored1 = icmp eq i32 %result1, 4 <i>; yields {i1}:stored1 = true</i>
7819%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 8</i>
7820
7821%val2 = add i32 1, 1
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7822%result2 = call i32 @llvm.atomic.swap.i32.p0i32(i32* %ptr, i32 %val2)
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7823 <i>; yields {i32}:result2 = 8</i>
7824
7825%stored2 = icmp eq i32 %result2, 8 <i>; yields {i1}:stored2 = true</i>
7826%memval2 = load i32* %ptr <i>; yields {i32}:memval2 = 2</i>
7827</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7828
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7829</div>
7830
7831<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7832<h4>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7833 <a name="int_atomic_load_add">'<tt>llvm.atomic.load.add.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7834</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7835
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7836<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7837
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7838<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7839<p>This is an overloaded intrinsic. You can use <tt>llvm.atomic.load.add</tt> on
7840 any integer bit width. Not all targets support all bit widths however.</p>
7841
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7842<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7843 declare i8 @llvm.atomic.load.add.i8.p0i8(i8* &lt;ptr&gt;, i8 &lt;delta&gt;)
7844 declare i16 @llvm.atomic.load.add.i16.p0i16(i16* &lt;ptr&gt;, i16 &lt;delta&gt;)
7845 declare i32 @llvm.atomic.load.add.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;delta&gt;)
7846 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]7847</pre>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7848
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7849<h5>Overview:</h5>
7850<p>This intrinsic adds <tt>delta</tt> to the value stored in memory
7851 at <tt>ptr</tt>. It yields the original value at <tt>ptr</tt>.</p>
7852
7853<h5>Arguments:</h5>
7854<p>The intrinsic takes two arguments, the first a pointer to an integer value
7855 and the second an integer value. The result is also an integer value. These
7856 integer types can have any bit width, but they must all have the same bit
7857 width. The targets may only lower integer representations they support.</p>
7858
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7859<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7860<p>This intrinsic does a series of operations atomically. It first loads the
7861 value stored at <tt>ptr</tt>. It then adds <tt>delta</tt>, stores the result
7862 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]7863
7864<h5>Examples:</h5>
7865<pre>
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]7866%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
7867%ptr = bitcast i8* %mallocP to i32*
7868 store i32 4, %ptr
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7869%result1 = call i32 @llvm.atomic.load.add.i32.p0i32(i32* %ptr, i32 4)
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7870 <i>; yields {i32}:result1 = 4</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7871%result2 = call i32 @llvm.atomic.load.add.i32.p0i32(i32* %ptr, i32 2)
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7872 <i>; yields {i32}:result2 = 8</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7873%result3 = call i32 @llvm.atomic.load.add.i32.p0i32(i32* %ptr, i32 5)
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7874 <i>; yields {i32}:result3 = 10</i>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7875%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 15</i>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7876</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7877
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7878</div>
7879
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7880<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7881<h4>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7882 <a name="int_atomic_load_sub">'<tt>llvm.atomic.load.sub.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7883</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7884
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7885<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7886
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7887<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7888<p>This is an overloaded intrinsic. You can use <tt>llvm.atomic.load.sub</tt> on
7889 any integer bit width and for different address spaces. Not all targets
7890 support all bit widths however.</p>
7891
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7892<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7893 declare i8 @llvm.atomic.load.sub.i8.p0i32(i8* &lt;ptr&gt;, i8 &lt;delta&gt;)
7894 declare i16 @llvm.atomic.load.sub.i16.p0i32(i16* &lt;ptr&gt;, i16 &lt;delta&gt;)
7895 declare i32 @llvm.atomic.load.sub.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;delta&gt;)
7896 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]7897</pre>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7898
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7899<h5>Overview:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]7900<p>This intrinsic subtracts <tt>delta</tt> to the value stored in memory at
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7901 <tt>ptr</tt>. It yields the original value at <tt>ptr</tt>.</p>
7902
7903<h5>Arguments:</h5>
7904<p>The intrinsic takes two arguments, the first a pointer to an integer value
7905 and the second an integer value. The result is also an integer value. These
7906 integer types can have any bit width, but they must all have the same bit
7907 width. The targets may only lower integer representations they support.</p>
7908
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7909<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7910<p>This intrinsic does a series of operations atomically. It first loads the
7911 value stored at <tt>ptr</tt>. It then subtracts <tt>delta</tt>, stores the
7912 result to <tt>ptr</tt>. It yields the original value stored
7913 at <tt>ptr</tt>.</p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7914
7915<h5>Examples:</h5>
7916<pre>
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]7917%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
7918%ptr = bitcast i8* %mallocP to i32*
7919 store i32 8, %ptr
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7920%result1 = call i32 @llvm.atomic.load.sub.i32.p0i32(i32* %ptr, i32 4)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7921 <i>; yields {i32}:result1 = 8</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7922%result2 = call i32 @llvm.atomic.load.sub.i32.p0i32(i32* %ptr, i32 2)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7923 <i>; yields {i32}:result2 = 4</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7924%result3 = call i32 @llvm.atomic.load.sub.i32.p0i32(i32* %ptr, i32 5)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7925 <i>; yields {i32}:result3 = 2</i>
7926%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = -3</i>
7927</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7928
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7929</div>
7930
7931<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7932<h4>
7933 <a name="int_atomic_load_and">
7934 '<tt>llvm.atomic.load.and.*</tt>' Intrinsic
7935 </a>
7936 <br>
7937 <a name="int_atomic_load_nand">
7938 '<tt>llvm.atomic.load.nand.*</tt>' Intrinsic
7939 </a>
7940 <br>
7941 <a name="int_atomic_load_or">
7942 '<tt>llvm.atomic.load.or.*</tt>' Intrinsic
7943 </a>
7944 <br>
7945 <a name="int_atomic_load_xor">
7946 '<tt>llvm.atomic.load.xor.*</tt>' Intrinsic
7947 </a>
7948</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7949
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7950<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7951
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7952<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7953<p>These are overloaded intrinsics. You can
7954 use <tt>llvm.atomic.load_and</tt>, <tt>llvm.atomic.load_nand</tt>,
7955 <tt>llvm.atomic.load_or</tt>, and <tt>llvm.atomic.load_xor</tt> on any integer
7956 bit width and for different address spaces. Not all targets support all bit
7957 widths however.</p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7958
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7959<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7960 declare i8 @llvm.atomic.load.and.i8.p0i8(i8* &lt;ptr&gt;, i8 &lt;delta&gt;)
7961 declare i16 @llvm.atomic.load.and.i16.p0i16(i16* &lt;ptr&gt;, i16 &lt;delta&gt;)
7962 declare i32 @llvm.atomic.load.and.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;delta&gt;)
7963 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]7964</pre>
7965
7966<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7967 declare i8 @llvm.atomic.load.or.i8.p0i8(i8* &lt;ptr&gt;, i8 &lt;delta&gt;)
7968 declare i16 @llvm.atomic.load.or.i16.p0i16(i16* &lt;ptr&gt;, i16 &lt;delta&gt;)
7969 declare i32 @llvm.atomic.load.or.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;delta&gt;)
7970 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]7971</pre>
7972
7973<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7974 declare i8 @llvm.atomic.load.nand.i8.p0i32(i8* &lt;ptr&gt;, i8 &lt;delta&gt;)
7975 declare i16 @llvm.atomic.load.nand.i16.p0i32(i16* &lt;ptr&gt;, i16 &lt;delta&gt;)
7976 declare i32 @llvm.atomic.load.nand.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;delta&gt;)
7977 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]7978</pre>
7979
7980<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7981 declare i8 @llvm.atomic.load.xor.i8.p0i32(i8* &lt;ptr&gt;, i8 &lt;delta&gt;)
7982 declare i16 @llvm.atomic.load.xor.i16.p0i32(i16* &lt;ptr&gt;, i16 &lt;delta&gt;)
7983 declare i32 @llvm.atomic.load.xor.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;delta&gt;)
7984 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]7985</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7986
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7987<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7988<p>These intrinsics bitwise the operation (and, nand, or, xor) <tt>delta</tt> to
7989 the value stored in memory at <tt>ptr</tt>. It yields the original value
7990 at <tt>ptr</tt>.</p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7991
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7992<h5>Arguments:</h5>
7993<p>These intrinsics take two arguments, the first a pointer to an integer value
7994 and the second an integer value. The result is also an integer value. These
7995 integer types can have any bit width, but they must all have the same bit
7996 width. The targets may only lower integer representations they support.</p>
7997
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7998<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7999<p>These intrinsics does a series of operations atomically. They first load the
8000 value stored at <tt>ptr</tt>. They then do the bitwise
8001 operation <tt>delta</tt>, store the result to <tt>ptr</tt>. They yield the
8002 original value stored at <tt>ptr</tt>.</p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]8003
8004<h5>Examples:</h5>
8005<pre>
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]8006%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
8007%ptr = bitcast i8* %mallocP to i32*
8008 store i32 0x0F0F, %ptr
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]8009%result0 = call i32 @llvm.atomic.load.nand.i32.p0i32(i32* %ptr, i32 0xFF)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]8010 <i>; yields {i32}:result0 = 0x0F0F</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]8011%result1 = call i32 @llvm.atomic.load.and.i32.p0i32(i32* %ptr, i32 0xFF)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]8012 <i>; yields {i32}:result1 = 0xFFFFFFF0</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]8013%result2 = call i32 @llvm.atomic.load.or.i32.p0i32(i32* %ptr, i32 0F)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]8014 <i>; yields {i32}:result2 = 0xF0</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]8015%result3 = call i32 @llvm.atomic.load.xor.i32.p0i32(i32* %ptr, i32 0F)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]8016 <i>; yields {i32}:result3 = FF</i>
8017%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = F0</i>
8018</pre>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]8019
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8020</div>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]8021
8022<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8023<h4>
8024 <a name="int_atomic_load_max">
8025 '<tt>llvm.atomic.load.max.*</tt>' Intrinsic
8026 </a>
8027 <br>
8028 <a name="int_atomic_load_min">
8029 '<tt>llvm.atomic.load.min.*</tt>' Intrinsic
8030 </a>
8031 <br>
8032 <a name="int_atomic_load_umax">
8033 '<tt>llvm.atomic.load.umax.*</tt>' Intrinsic
8034 </a>
8035 <br>
8036 <a name="int_atomic_load_umin">
8037 '<tt>llvm.atomic.load.umin.*</tt>' Intrinsic
8038 </a>
8039</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8040
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8041<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8042
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]8043<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8044<p>These are overloaded intrinsics. You can use <tt>llvm.atomic.load_max</tt>,
8045 <tt>llvm.atomic.load_min</tt>, <tt>llvm.atomic.load_umax</tt>, and
8046 <tt>llvm.atomic.load_umin</tt> on any integer bit width and for different
8047 address spaces. Not all targets support all bit widths however.</p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]8048
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8049<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]8050 declare i8 @llvm.atomic.load.max.i8.p0i8(i8* &lt;ptr&gt;, i8 &lt;delta&gt;)
8051 declare i16 @llvm.atomic.load.max.i16.p0i16(i16* &lt;ptr&gt;, i16 &lt;delta&gt;)
8052 declare i32 @llvm.atomic.load.max.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;delta&gt;)
8053 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]8054</pre>
8055
8056<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]8057 declare i8 @llvm.atomic.load.min.i8.p0i8(i8* &lt;ptr&gt;, i8 &lt;delta&gt;)
8058 declare i16 @llvm.atomic.load.min.i16.p0i16(i16* &lt;ptr&gt;, i16 &lt;delta&gt;)
8059 declare i32 @llvm.atomic.load.min.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;delta&gt;)
8060 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]8061</pre>
8062
8063<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]8064 declare i8 @llvm.atomic.load.umax.i8.p0i8(i8* &lt;ptr&gt;, i8 &lt;delta&gt;)
8065 declare i16 @llvm.atomic.load.umax.i16.p0i16(i16* &lt;ptr&gt;, i16 &lt;delta&gt;)
8066 declare i32 @llvm.atomic.load.umax.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;delta&gt;)
8067 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]8068</pre>
8069
8070<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]8071 declare i8 @llvm.atomic.load.umin.i8.p0i8(i8* &lt;ptr&gt;, i8 &lt;delta&gt;)
8072 declare i16 @llvm.atomic.load.umin.i16.p0i16(i16* &lt;ptr&gt;, i16 &lt;delta&gt;)
8073 declare i32 @llvm.atomic.load.umin.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;delta&gt;)
8074 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]8075</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8076
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]8077<h5>Overview:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]8078<p>These intrinsics takes the signed or unsigned minimum or maximum of
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8079 <tt>delta</tt> and the value stored in memory at <tt>ptr</tt>. It yields the
8080 original value at <tt>ptr</tt>.</p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]8081
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8082<h5>Arguments:</h5>
8083<p>These intrinsics take two arguments, the first a pointer to an integer value
8084 and the second an integer value. The result is also an integer value. These
8085 integer types can have any bit width, but they must all have the same bit
8086 width. The targets may only lower integer representations they support.</p>
8087
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]8088<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8089<p>These intrinsics does a series of operations atomically. They first load the
8090 value stored at <tt>ptr</tt>. They then do the signed or unsigned min or
8091 max <tt>delta</tt> and the value, store the result to <tt>ptr</tt>. They
8092 yield the original value stored at <tt>ptr</tt>.</p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]8093
8094<h5>Examples:</h5>
8095<pre>
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]8096%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
8097%ptr = bitcast i8* %mallocP to i32*
8098 store i32 7, %ptr
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]8099%result0 = call i32 @llvm.atomic.load.min.i32.p0i32(i32* %ptr, i32 -2)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]8100 <i>; yields {i32}:result0 = 7</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]8101%result1 = call i32 @llvm.atomic.load.max.i32.p0i32(i32* %ptr, i32 8)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]8102 <i>; yields {i32}:result1 = -2</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]8103%result2 = call i32 @llvm.atomic.load.umin.i32.p0i32(i32* %ptr, i32 10)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]8104 <i>; yields {i32}:result2 = 8</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]8105%result3 = call i32 @llvm.atomic.load.umax.i32.p0i32(i32* %ptr, i32 30)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]8106 <i>; yields {i32}:result3 = 8</i>
8107%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 30</i>
8108</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8109
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]8110</div>
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]8111
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8112</div>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8113
8114<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8115<h3>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8116 <a name="int_memorymarkers">Memory Use Markers</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8117</h3>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8118
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8119<div>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8120
8121<p>This class of intrinsics exists to information about the lifetime of memory
8122 objects and ranges where variables are immutable.</p>
8123
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8124<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8125<h4>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8126 <a name="int_lifetime_start">'<tt>llvm.lifetime.start</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8127</h4>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8128
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8129<div>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8130
8131<h5>Syntax:</h5>
8132<pre>
8133 declare void @llvm.lifetime.start(i64 &lt;size&gt;, i8* nocapture &lt;ptr&gt;)
8134</pre>
8135
8136<h5>Overview:</h5>
8137<p>The '<tt>llvm.lifetime.start</tt>' intrinsic specifies the start of a memory
8138 object's lifetime.</p>
8139
8140<h5>Arguments:</h5>
Nick Lewycky321333e2009-10-13 07:57:33 +0000[diff] [blame]8141<p>The first argument is a constant integer representing the size of the
8142 object, or -1 if it is variable sized. The second argument is a pointer to
8143 the object.</p>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8144
8145<h5>Semantics:</h5>
8146<p>This intrinsic indicates that before this point in the code, the value of the
8147 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]8148 never be used and has an undefined value. A load from the pointer that
8149 precedes this intrinsic can be replaced with
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8150 <tt>'<a href="#undefvalues">undef</a>'</tt>.</p>
8151
8152</div>
8153
8154<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8155<h4>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8156 <a name="int_lifetime_end">'<tt>llvm.lifetime.end</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8157</h4>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8158
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8159<div>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8160
8161<h5>Syntax:</h5>
8162<pre>
8163 declare void @llvm.lifetime.end(i64 &lt;size&gt;, i8* nocapture &lt;ptr&gt;)
8164</pre>
8165
8166<h5>Overview:</h5>
8167<p>The '<tt>llvm.lifetime.end</tt>' intrinsic specifies the end of a memory
8168 object's lifetime.</p>
8169
8170<h5>Arguments:</h5>
Nick Lewycky321333e2009-10-13 07:57:33 +0000[diff] [blame]8171<p>The first argument is a constant integer representing the size of the
8172 object, or -1 if it is variable sized. The second argument is a pointer to
8173 the object.</p>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8174
8175<h5>Semantics:</h5>
8176<p>This intrinsic indicates that after this point in the code, the value of the
8177 memory pointed to by <tt>ptr</tt> is dead. This means that it is known to
8178 never be used and has an undefined value. Any stores into the memory object
8179 following this intrinsic may be removed as dead.
8180
8181</div>
8182
8183<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8184<h4>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8185 <a name="int_invariant_start">'<tt>llvm.invariant.start</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8186</h4>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8187
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8188<div>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8189
8190<h5>Syntax:</h5>
8191<pre>
Nick Lewycky29b6cb42010-11-30 04:13:41 +0000[diff] [blame]8192 declare {}* @llvm.invariant.start(i64 &lt;size&gt;, i8* nocapture &lt;ptr&gt;)
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8193</pre>
8194
8195<h5>Overview:</h5>
8196<p>The '<tt>llvm.invariant.start</tt>' intrinsic specifies that the contents of
8197 a memory object will not change.</p>
8198
8199<h5>Arguments:</h5>
Nick Lewycky321333e2009-10-13 07:57:33 +0000[diff] [blame]8200<p>The first argument is a constant integer representing the size of the
8201 object, or -1 if it is variable sized. The second argument is a pointer to
8202 the object.</p>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8203
8204<h5>Semantics:</h5>
8205<p>This intrinsic indicates that until an <tt>llvm.invariant.end</tt> that uses
8206 the return value, the referenced memory location is constant and
8207 unchanging.</p>
8208
8209</div>
8210
8211<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8212<h4>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8213 <a name="int_invariant_end">'<tt>llvm.invariant.end</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8214</h4>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8215
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8216<div>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8217
8218<h5>Syntax:</h5>
8219<pre>
8220 declare void @llvm.invariant.end({}* &lt;start&gt;, i64 &lt;size&gt;, i8* nocapture &lt;ptr&gt;)
8221</pre>
8222
8223<h5>Overview:</h5>
8224<p>The '<tt>llvm.invariant.end</tt>' intrinsic specifies that the contents of
8225 a memory object are mutable.</p>
8226
8227<h5>Arguments:</h5>
8228<p>The first argument is the matching <tt>llvm.invariant.start</tt> intrinsic.
Nick Lewycky321333e2009-10-13 07:57:33 +0000[diff] [blame]8229 The second argument is a constant integer representing the size of the
8230 object, or -1 if it is variable sized and the third argument is a pointer
8231 to the object.</p>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8232
8233<h5>Semantics:</h5>
8234<p>This intrinsic indicates that the memory is mutable again.</p>
8235
8236</div>
8237
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8238</div>
8239
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]8240<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8241<h3>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]8242 <a name="int_general">General Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8243</h3>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]8244
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8245<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8246
8247<p>This class of intrinsics is designed to be generic and has no specific
8248 purpose.</p>
8249
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]8250<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8251<h4>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]8252 <a name="int_var_annotation">'<tt>llvm.var.annotation</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8253</h4>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]8254
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8255<div>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]8256
8257<h5>Syntax:</h5>
8258<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]8259 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]8260</pre>
8261
8262<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8263<p>The '<tt>llvm.var.annotation</tt>' intrinsic.</p>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]8264
8265<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8266<p>The first argument is a pointer to a value, the second is a pointer to a
8267 global string, the third is a pointer to a global string which is the source
8268 file name, and the last argument is the line number.</p>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]8269
8270<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8271<p>This intrinsic allows annotation of local variables with arbitrary strings.
8272 This can be useful for special purpose optimizations that want to look for
8273 these annotations. These have no other defined use, they are ignored by code
8274 generation and optimization.</p>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]8275
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]8276</div>
8277
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]8278<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8279<h4>
Tanya Lattnere1a8da02007-09-21 23:57:59 +0000[diff] [blame]8280 <a name="int_annotation">'<tt>llvm.annotation.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8281</h4>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]8282
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8283<div>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]8284
8285<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8286<p>This is an overloaded intrinsic. You can use '<tt>llvm.annotation</tt>' on
8287 any integer bit width.</p>
8288
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]8289<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]8290 declare i8 @llvm.annotation.i8(i8 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt;)
8291 declare i16 @llvm.annotation.i16(i16 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt;)
8292 declare i32 @llvm.annotation.i32(i32 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt;)
8293 declare i64 @llvm.annotation.i64(i64 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt;)
8294 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]8295</pre>
8296
8297<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8298<p>The '<tt>llvm.annotation</tt>' intrinsic.</p>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]8299
8300<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8301<p>The first argument is an integer value (result of some expression), the
8302 second is a pointer to a global string, the third is a pointer to a global
8303 string which is the source file name, and the last argument is the line
8304 number. It returns the value of the first argument.</p>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]8305
8306<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8307<p>This intrinsic allows annotations to be put on arbitrary expressions with
8308 arbitrary strings. This can be useful for special purpose optimizations that
8309 want to look for these annotations. These have no other defined use, they
8310 are ignored by code generation and optimization.</p>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]8311
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]8312</div>
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]8313
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]8314<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8315<h4>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]8316 <a name="int_trap">'<tt>llvm.trap</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8317</h4>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]8318
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8319<div>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]8320
8321<h5>Syntax:</h5>
8322<pre>
8323 declare void @llvm.trap()
8324</pre>
8325
8326<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8327<p>The '<tt>llvm.trap</tt>' intrinsic.</p>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]8328
8329<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8330<p>None.</p>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]8331
8332<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8333<p>This intrinsics is lowered to the target dependent trap instruction. If the
8334 target does not have a trap instruction, this intrinsic will be lowered to
8335 the call of the <tt>abort()</tt> function.</p>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]8336
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]8337</div>
8338
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]8339<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8340<h4>
Misha Brukmandccb0252008-11-22 23:55:29 +0000[diff] [blame]8341 <a name="int_stackprotector">'<tt>llvm.stackprotector</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8342</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8343
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8344<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8345
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]8346<h5>Syntax:</h5>
8347<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]8348 declare void @llvm.stackprotector(i8* &lt;guard&gt;, i8** &lt;slot&gt;)
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]8349</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8350
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]8351<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8352<p>The <tt>llvm.stackprotector</tt> intrinsic takes the <tt>guard</tt> and
8353 stores it onto the stack at <tt>slot</tt>. The stack slot is adjusted to
8354 ensure that it is placed on the stack before local variables.</p>
8355
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]8356<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8357<p>The <tt>llvm.stackprotector</tt> intrinsic requires two pointer
8358 arguments. The first argument is the value loaded from the stack
8359 guard <tt>@__stack_chk_guard</tt>. The second variable is an <tt>alloca</tt>
8360 that has enough space to hold the value of the guard.</p>
8361
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]8362<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8363<p>This intrinsic causes the prologue/epilogue inserter to force the position of
8364 the <tt>AllocaInst</tt> stack slot to be before local variables on the
8365 stack. This is to ensure that if a local variable on the stack is
8366 overwritten, it will destroy the value of the guard. When the function exits,
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]8367 the guard on the stack is checked against the original guard. If they are
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8368 different, then the program aborts by calling the <tt>__stack_chk_fail()</tt>
8369 function.</p>
8370
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]8371</div>
8372
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]8373<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8374<h4>
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]8375 <a name="int_objectsize">'<tt>llvm.objectsize</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8376</h4>
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]8377
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8378<div>
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]8379
8380<h5>Syntax:</h5>
8381<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]8382 declare i32 @llvm.objectsize.i32(i8* &lt;object&gt;, i1 &lt;type&gt;)
8383 declare i64 @llvm.objectsize.i64(i8* &lt;object&gt;, i1 &lt;type&gt;)
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]8384</pre>
8385
8386<h5>Overview:</h5>
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]8387<p>The <tt>llvm.objectsize</tt> intrinsic is designed to provide information to
8388 the optimizers to determine at compile time whether a) an operation (like
8389 memcpy) will overflow a buffer that corresponds to an object, or b) that a
8390 runtime check for overflow isn't necessary. An object in this context means
8391 an allocation of a specific class, structure, array, or other object.</p>
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]8392
8393<h5>Arguments:</h5>
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]8394<p>The <tt>llvm.objectsize</tt> intrinsic takes two arguments. The first
Eric Christopher8295a0a2009-12-23 00:29:49 +0000[diff] [blame]8395 argument is a pointer to or into the <tt>object</tt>. The second argument
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]8396 is a boolean 0 or 1. This argument determines whether you want the
8397 maximum (0) or minimum (1) bytes remaining. This needs to be a literal 0 or
Eric Christopher8295a0a2009-12-23 00:29:49 +0000[diff] [blame]8398 1, variables are not allowed.</p>
8399
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]8400<h5>Semantics:</h5>
8401<p>The <tt>llvm.objectsize</tt> intrinsic is lowered to either a constant
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]8402 representing the size of the object concerned, or <tt>i32/i64 -1 or 0</tt>,
8403 depending on the <tt>type</tt> argument, if the size cannot be determined at
8404 compile time.</p>
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]8405
8406</div>
8407
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8408</div>
8409
8410</div>
8411
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]8412<!-- *********************************************************************** -->
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]8413<hr>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]8414<address>
8415 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
Misha Brukman44408702008-12-11 17:34:48 +0000[diff] [blame]8416 src="http://jigsaw.w3.org/css-validator/images/vcss-blue" alt="Valid CSS"></a>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]8417 <a href="http://validator.w3.org/check/referer"><img
Misha Brukman44408702008-12-11 17:34:48 +0000[diff] [blame]8418 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]8419
8420 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
NAKAMURA Takumib9a33632011-04-09 02:13:37 +0000[diff] [blame]8421 <a href="http://llvm.org/">The LLVM Compiler Infrastructure</a><br>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]8422 Last modified: $Date$
8423</address>
Chris Lattnerc7d3ab32008-01-04 04:33:49 +0000[diff] [blame]8424
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]8425</body>
8426</html>