Gitiles
Code Review Sign In
gerrit-public.fairphone.software / fp2-dev / platform / external / llvm / c36ed70ec5c3c99f9559cfaa199373f60219a2be / . / docs / LangRef.html
blob: 08d84df4a8fb112cdcefefcb305c8b408d8fb97f [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>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]56 </ol>
57 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]58 <li><a href="#typesystem">Type System</a>
59 <ol>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]60 <li><a href="#t_classifications">Type Classifications</a></li>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]61 <li><a href="#t_primitive">Primitive Types</a>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]62 <ol>
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]63 <li><a href="#t_integer">Integer Type</a></li>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]64 <li><a href="#t_floating">Floating Point Types</a></li>
Dale Johannesen21fe99b2010-10-01 00:48:59 +0000[diff] [blame]65 <li><a href="#t_x86mmx">X86mmx Type</a></li>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]66 <li><a href="#t_void">Void Type</a></li>
67 <li><a href="#t_label">Label Type</a></li>
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]68 <li><a href="#t_metadata">Metadata Type</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]69 </ol>
70 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]71 <li><a href="#t_derived">Derived Types</a>
72 <ol>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]73 <li><a href="#t_aggregate">Aggregate Types</a>
74 <ol>
75 <li><a href="#t_array">Array Type</a></li>
76 <li><a href="#t_struct">Structure Type</a></li>
77 <li><a href="#t_pstruct">Packed Structure Type</a></li>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]78 <li><a href="#t_vector">Vector Type</a></li>
79 </ol>
80 </li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]81 <li><a href="#t_function">Function Type</a></li>
82 <li><a href="#t_pointer">Pointer Type</a></li>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]83 <li><a href="#t_opaque">Opaque Type</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]84 </ol>
85 </li>
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]86 <li><a href="#t_uprefs">Type Up-references</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]87 </ol>
88 </li>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]89 <li><a href="#constants">Constants</a>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]90 <ol>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]91 <li><a href="#simpleconstants">Simple Constants</a></li>
Chris Lattner70882792009-02-28 18:32:25 +0000[diff] [blame]92 <li><a href="#complexconstants">Complex Constants</a></li>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]93 <li><a href="#globalconstants">Global Variable and Function Addresses</a></li>
94 <li><a href="#undefvalues">Undefined Values</a></li>
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]95 <li><a href="#trapvalues">Trap Values</a></li>
Chris Lattnerf9d078e2009-10-27 21:19:13 +0000[diff] [blame]96 <li><a href="#blockaddress">Addresses of Basic Blocks</a></li>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]97 <li><a href="#constantexprs">Constant Expressions</a></li>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]98 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]99 </li>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]100 <li><a href="#othervalues">Other Values</a>
101 <ol>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]102 <li><a href="#inlineasm">Inline Assembler Expressions</a></li>
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]103 <li><a href="#metadata">Metadata Nodes and Metadata Strings</a></li>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]104 </ol>
105 </li>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]106 <li><a href="#intrinsic_globals">Intrinsic Global Variables</a>
107 <ol>
108 <li><a href="#intg_used">The '<tt>llvm.used</tt>' Global Variable</a></li>
Chris Lattner401e10c2009-07-20 06:14:25 +0000[diff] [blame]109 <li><a href="#intg_compiler_used">The '<tt>llvm.compiler.used</tt>'
110 Global Variable</a></li>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]111 <li><a href="#intg_global_ctors">The '<tt>llvm.global_ctors</tt>'
112 Global Variable</a></li>
113 <li><a href="#intg_global_dtors">The '<tt>llvm.global_dtors</tt>'
114 Global Variable</a></li>
115 </ol>
116 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]117 <li><a href="#instref">Instruction Reference</a>
118 <ol>
119 <li><a href="#terminators">Terminator Instructions</a>
120 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]121 <li><a href="#i_ret">'<tt>ret</tt>' Instruction</a></li>
122 <li><a href="#i_br">'<tt>br</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]123 <li><a href="#i_switch">'<tt>switch</tt>' Instruction</a></li>
Chris Lattnerab21db72009-10-28 00:19:10 +0000[diff] [blame]124 <li><a href="#i_indirectbr">'<tt>indirectbr</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]125 <li><a href="#i_invoke">'<tt>invoke</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]126 <li><a href="#i_unwind">'<tt>unwind</tt>' Instruction</a></li>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]127 <li><a href="#i_unreachable">'<tt>unreachable</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]128 </ol>
129 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]130 <li><a href="#binaryops">Binary Operations</a>
131 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]132 <li><a href="#i_add">'<tt>add</tt>' Instruction</a></li>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]133 <li><a href="#i_fadd">'<tt>fadd</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]134 <li><a href="#i_sub">'<tt>sub</tt>' Instruction</a></li>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]135 <li><a href="#i_fsub">'<tt>fsub</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]136 <li><a href="#i_mul">'<tt>mul</tt>' Instruction</a></li>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]137 <li><a href="#i_fmul">'<tt>fmul</tt>' Instruction</a></li>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]138 <li><a href="#i_udiv">'<tt>udiv</tt>' Instruction</a></li>
139 <li><a href="#i_sdiv">'<tt>sdiv</tt>' Instruction</a></li>
140 <li><a href="#i_fdiv">'<tt>fdiv</tt>' Instruction</a></li>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]141 <li><a href="#i_urem">'<tt>urem</tt>' Instruction</a></li>
142 <li><a href="#i_srem">'<tt>srem</tt>' Instruction</a></li>
143 <li><a href="#i_frem">'<tt>frem</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]144 </ol>
145 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]146 <li><a href="#bitwiseops">Bitwise Binary Operations</a>
147 <ol>
Reid Spencer8e11bf82007-02-02 13:57:07 +0000[diff] [blame]148 <li><a href="#i_shl">'<tt>shl</tt>' Instruction</a></li>
149 <li><a href="#i_lshr">'<tt>lshr</tt>' Instruction</a></li>
150 <li><a href="#i_ashr">'<tt>ashr</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]151 <li><a href="#i_and">'<tt>and</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]152 <li><a href="#i_or">'<tt>or</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]153 <li><a href="#i_xor">'<tt>xor</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]154 </ol>
155 </li>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]156 <li><a href="#vectorops">Vector Operations</a>
157 <ol>
158 <li><a href="#i_extractelement">'<tt>extractelement</tt>' Instruction</a></li>
159 <li><a href="#i_insertelement">'<tt>insertelement</tt>' Instruction</a></li>
160 <li><a href="#i_shufflevector">'<tt>shufflevector</tt>' Instruction</a></li>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]161 </ol>
162 </li>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]163 <li><a href="#aggregateops">Aggregate Operations</a>
164 <ol>
165 <li><a href="#i_extractvalue">'<tt>extractvalue</tt>' Instruction</a></li>
166 <li><a href="#i_insertvalue">'<tt>insertvalue</tt>' Instruction</a></li>
167 </ol>
168 </li>
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]169 <li><a href="#memoryops">Memory Access and Addressing Operations</a>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]170 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]171 <li><a href="#i_alloca">'<tt>alloca</tt>' Instruction</a></li>
Robert Bocchino7b81c752006-02-17 21:18:08 +0000[diff] [blame]172 <li><a href="#i_load">'<tt>load</tt>' Instruction</a></li>
173 <li><a href="#i_store">'<tt>store</tt>' Instruction</a></li>
174 <li><a href="#i_getelementptr">'<tt>getelementptr</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]175 </ol>
176 </li>
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]177 <li><a href="#convertops">Conversion Operations</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]178 <ol>
179 <li><a href="#i_trunc">'<tt>trunc .. to</tt>' Instruction</a></li>
180 <li><a href="#i_zext">'<tt>zext .. to</tt>' Instruction</a></li>
181 <li><a href="#i_sext">'<tt>sext .. to</tt>' Instruction</a></li>
182 <li><a href="#i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a></li>
183 <li><a href="#i_fpext">'<tt>fpext .. to</tt>' Instruction</a></li>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]184 <li><a href="#i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a></li>
185 <li><a href="#i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a></li>
186 <li><a href="#i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a></li>
187 <li><a href="#i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a></li>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]188 <li><a href="#i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a></li>
189 <li><a href="#i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a></li>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]190 <li><a href="#i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a></li>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]191 </ol>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]192 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]193 <li><a href="#otherops">Other Operations</a>
194 <ol>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]195 <li><a href="#i_icmp">'<tt>icmp</tt>' Instruction</a></li>
196 <li><a href="#i_fcmp">'<tt>fcmp</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]197 <li><a href="#i_phi">'<tt>phi</tt>' Instruction</a></li>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]198 <li><a href="#i_select">'<tt>select</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]199 <li><a href="#i_call">'<tt>call</tt>' Instruction</a></li>
Chris Lattnerfb6977d2006-01-13 23:26:01 +0000[diff] [blame]200 <li><a href="#i_va_arg">'<tt>va_arg</tt>' Instruction</a></li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]201 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]202 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]203 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]204 </li>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]205 <li><a href="#intrinsics">Intrinsic Functions</a>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]206 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]207 <li><a href="#int_varargs">Variable Argument Handling Intrinsics</a>
208 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]209 <li><a href="#int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a></li>
210 <li><a href="#int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a></li>
211 <li><a href="#int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]212 </ol>
213 </li>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]214 <li><a href="#int_gc">Accurate Garbage Collection Intrinsics</a>
215 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]216 <li><a href="#int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a></li>
217 <li><a href="#int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a></li>
218 <li><a href="#int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a></li>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]219 </ol>
220 </li>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]221 <li><a href="#int_codegen">Code Generator Intrinsics</a>
222 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]223 <li><a href="#int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a></li>
224 <li><a href="#int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a></li>
225 <li><a href="#int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a></li>
226 <li><a href="#int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a></li>
227 <li><a href="#int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a></li>
228 <li><a href="#int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a></li>
Dan Gohman31f1af12010-05-26 21:56:15 +0000[diff] [blame]229 <li><a href="#int_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a></li>
John Criswell7123e272004-04-09 16:43:20 +0000[diff] [blame]230 </ol>
231 </li>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]232 <li><a href="#int_libc">Standard C Library Intrinsics</a>
233 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]234 <li><a href="#int_memcpy">'<tt>llvm.memcpy.*</tt>' Intrinsic</a></li>
235 <li><a href="#int_memmove">'<tt>llvm.memmove.*</tt>' Intrinsic</a></li>
236 <li><a href="#int_memset">'<tt>llvm.memset.*</tt>' Intrinsic</a></li>
237 <li><a href="#int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a></li>
238 <li><a href="#int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a></li>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]239 <li><a href="#int_sin">'<tt>llvm.sin.*</tt>' Intrinsic</a></li>
240 <li><a href="#int_cos">'<tt>llvm.cos.*</tt>' Intrinsic</a></li>
241 <li><a href="#int_pow">'<tt>llvm.pow.*</tt>' Intrinsic</a></li>
Dan Gohman08b280b2011-05-27 00:36:31 +0000[diff] [blame]242 <li><a href="#int_exp">'<tt>llvm.exp.*</tt>' Intrinsic</a></li>
243 <li><a href="#int_log">'<tt>llvm.log.*</tt>' Intrinsic</a></li>
Cameron Zwarich33390842011-07-08 21:39:21 +0000[diff] [blame]244 <li><a href="#int_fma">'<tt>llvm.fma.*</tt>' Intrinsic</a></li>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]245 </ol>
246 </li>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]247 <li><a href="#int_manip">Bit Manipulation Intrinsics</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]248 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]249 <li><a href="#int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a></li>
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]250 <li><a href="#int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic </a></li>
251 <li><a href="#int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic </a></li>
252 <li><a href="#int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic </a></li>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]253 </ol>
254 </li>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]255 <li><a href="#int_overflow">Arithmetic with Overflow Intrinsics</a>
256 <ol>
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]257 <li><a href="#int_sadd_overflow">'<tt>llvm.sadd.with.overflow.*</tt> Intrinsics</a></li>
258 <li><a href="#int_uadd_overflow">'<tt>llvm.uadd.with.overflow.*</tt> Intrinsics</a></li>
259 <li><a href="#int_ssub_overflow">'<tt>llvm.ssub.with.overflow.*</tt> Intrinsics</a></li>
260 <li><a href="#int_usub_overflow">'<tt>llvm.usub.with.overflow.*</tt> Intrinsics</a></li>
261 <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]262 <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]263 </ol>
264 </li>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]265 <li><a href="#int_fp16">Half Precision Floating Point Intrinsics</a>
266 <ol>
Chris Lattner82c3dc62010-03-14 23:03:31 +0000[diff] [blame]267 <li><a href="#int_convert_to_fp16">'<tt>llvm.convert.to.fp16</tt>' Intrinsic</a></li>
268 <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]269 </ol>
270 </li>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]271 <li><a href="#int_debugger">Debugger intrinsics</a></li>
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]272 <li><a href="#int_eh">Exception Handling intrinsics</a></li>
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]273 <li><a href="#int_trampoline">Trampoline Intrinsic</a>
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]274 <ol>
275 <li><a href="#int_it">'<tt>llvm.init.trampoline</tt>' Intrinsic</a></li>
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]276 </ol>
277 </li>
Bill Wendling3c44f5b2008-11-18 22:10:53 +0000[diff] [blame]278 <li><a href="#int_atomics">Atomic intrinsics</a>
279 <ol>
280 <li><a href="#int_memory_barrier"><tt>llvm.memory_barrier</tt></a></li>
281 <li><a href="#int_atomic_cmp_swap"><tt>llvm.atomic.cmp.swap</tt></a></li>
282 <li><a href="#int_atomic_swap"><tt>llvm.atomic.swap</tt></a></li>
283 <li><a href="#int_atomic_load_add"><tt>llvm.atomic.load.add</tt></a></li>
284 <li><a href="#int_atomic_load_sub"><tt>llvm.atomic.load.sub</tt></a></li>
285 <li><a href="#int_atomic_load_and"><tt>llvm.atomic.load.and</tt></a></li>
286 <li><a href="#int_atomic_load_nand"><tt>llvm.atomic.load.nand</tt></a></li>
287 <li><a href="#int_atomic_load_or"><tt>llvm.atomic.load.or</tt></a></li>
288 <li><a href="#int_atomic_load_xor"><tt>llvm.atomic.load.xor</tt></a></li>
289 <li><a href="#int_atomic_load_max"><tt>llvm.atomic.load.max</tt></a></li>
290 <li><a href="#int_atomic_load_min"><tt>llvm.atomic.load.min</tt></a></li>
291 <li><a href="#int_atomic_load_umax"><tt>llvm.atomic.load.umax</tt></a></li>
292 <li><a href="#int_atomic_load_umin"><tt>llvm.atomic.load.umin</tt></a></li>
293 </ol>
294 </li>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]295 <li><a href="#int_memorymarkers">Memory Use Markers</a>
296 <ol>
297 <li><a href="#int_lifetime_start"><tt>llvm.lifetime.start</tt></a></li>
298 <li><a href="#int_lifetime_end"><tt>llvm.lifetime.end</tt></a></li>
299 <li><a href="#int_invariant_start"><tt>llvm.invariant.start</tt></a></li>
300 <li><a href="#int_invariant_end"><tt>llvm.invariant.end</tt></a></li>
301 </ol>
302 </li>
Reid Spencer20677642007-07-20 19:59:11 +0000[diff] [blame]303 <li><a href="#int_general">General intrinsics</a>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]304 <ol>
Reid Spencer20677642007-07-20 19:59:11 +0000[diff] [blame]305 <li><a href="#int_var_annotation">
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]306 '<tt>llvm.var.annotation</tt>' Intrinsic</a></li>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]307 <li><a href="#int_annotation">
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]308 '<tt>llvm.annotation.*</tt>' Intrinsic</a></li>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]309 <li><a href="#int_trap">
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]310 '<tt>llvm.trap</tt>' Intrinsic</a></li>
311 <li><a href="#int_stackprotector">
312 '<tt>llvm.stackprotector</tt>' Intrinsic</a></li>
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]313 <li><a href="#int_objectsize">
314 '<tt>llvm.objectsize</tt>' Intrinsic</a></li>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]315 </ol>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]316 </li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]317 </ol>
318 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]319</ol>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]320
321<div class="doc_author">
322 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>
323 and <a href="mailto:vadve@cs.uiuc.edu">Vikram Adve</a></p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]324</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]325
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]326<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]327<h2><a name="abstract">Abstract</a></h2>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]328<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]329
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]330<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]331
332<p>This document is a reference manual for the LLVM assembly language. LLVM is
333 a Static Single Assignment (SSA) based representation that provides type
334 safety, low-level operations, flexibility, and the capability of representing
335 'all' high-level languages cleanly. It is the common code representation
336 used throughout all phases of the LLVM compilation strategy.</p>
337
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]338</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]339
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]340<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]341<h2><a name="introduction">Introduction</a></h2>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]342<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]343
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]344<div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]345
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]346<p>The LLVM code representation is designed to be used in three different forms:
347 as an in-memory compiler IR, as an on-disk bitcode representation (suitable
348 for fast loading by a Just-In-Time compiler), and as a human readable
349 assembly language representation. This allows LLVM to provide a powerful
350 intermediate representation for efficient compiler transformations and
351 analysis, while providing a natural means to debug and visualize the
352 transformations. The three different forms of LLVM are all equivalent. This
353 document describes the human readable representation and notation.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]354
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]355<p>The LLVM representation aims to be light-weight and low-level while being
356 expressive, typed, and extensible at the same time. It aims to be a
357 "universal IR" of sorts, by being at a low enough level that high-level ideas
358 may be cleanly mapped to it (similar to how microprocessors are "universal
359 IR's", allowing many source languages to be mapped to them). By providing
360 type information, LLVM can be used as the target of optimizations: for
361 example, through pointer analysis, it can be proven that a C automatic
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]362 variable is never accessed outside of the current function, allowing it to
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]363 be promoted to a simple SSA value instead of a memory location.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]364
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]365<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]366<h4>
367 <a name="wellformed">Well-Formedness</a>
368</h4>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]369
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]370<div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]371
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]372<p>It is important to note that this document describes 'well formed' LLVM
373 assembly language. There is a difference between what the parser accepts and
374 what is considered 'well formed'. For example, the following instruction is
375 syntactically okay, but not well formed:</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]376
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]377<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]378%x = <a href="#i_add">add</a> i32 1, %x
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]379</pre>
380
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]381<p>because the definition of <tt>%x</tt> does not dominate all of its uses. The
382 LLVM infrastructure provides a verification pass that may be used to verify
383 that an LLVM module is well formed. This pass is automatically run by the
384 parser after parsing input assembly and by the optimizer before it outputs
385 bitcode. The violations pointed out by the verifier pass indicate bugs in
386 transformation passes or input to the parser.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]387
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]388</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]389
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]390</div>
391
Chris Lattnercc689392007-10-03 17:34:29 +0000[diff] [blame]392<!-- Describe the typesetting conventions here. -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]393
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]394<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]395<h2><a name="identifiers">Identifiers</a></h2>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]396<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]397
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]398<div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]399
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]400<p>LLVM identifiers come in two basic types: global and local. Global
401 identifiers (functions, global variables) begin with the <tt>'@'</tt>
402 character. Local identifiers (register names, types) begin with
403 the <tt>'%'</tt> character. Additionally, there are three different formats
404 for identifiers, for different purposes:</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]405
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]406<ol>
Reid Spencer2c452282007-08-07 14:34:28 +0000[diff] [blame]407 <li>Named values are represented as a string of characters with their prefix.
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]408 For example, <tt>%foo</tt>, <tt>@DivisionByZero</tt>,
409 <tt>%a.really.long.identifier</tt>. The actual regular expression used is
410 '<tt>[%@][a-zA-Z$._][a-zA-Z$._0-9]*</tt>'. Identifiers which require
411 other characters in their names can be surrounded with quotes. Special
412 characters may be escaped using <tt>"\xx"</tt> where <tt>xx</tt> is the
413 ASCII code for the character in hexadecimal. In this way, any character
414 can be used in a name value, even quotes themselves.</li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]415
Reid Spencer2c452282007-08-07 14:34:28 +0000[diff] [blame]416 <li>Unnamed values are represented as an unsigned numeric value with their
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]417 prefix. For example, <tt>%12</tt>, <tt>@2</tt>, <tt>%44</tt>.</li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]418
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]419 <li>Constants, which are described in a <a href="#constants">section about
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]420 constants</a>, below.</li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]421</ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]422
Reid Spencer2c452282007-08-07 14:34:28 +0000[diff] [blame]423<p>LLVM requires that values start with a prefix for two reasons: Compilers
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]424 don't need to worry about name clashes with reserved words, and the set of
425 reserved words may be expanded in the future without penalty. Additionally,
426 unnamed identifiers allow a compiler to quickly come up with a temporary
427 variable without having to avoid symbol table conflicts.</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]428
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]429<p>Reserved words in LLVM are very similar to reserved words in other
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]430 languages. There are keywords for different opcodes
431 ('<tt><a href="#i_add">add</a></tt>',
432 '<tt><a href="#i_bitcast">bitcast</a></tt>',
433 '<tt><a href="#i_ret">ret</a></tt>', etc...), for primitive type names
434 ('<tt><a href="#t_void">void</a></tt>',
435 '<tt><a href="#t_primitive">i32</a></tt>', etc...), and others. These
436 reserved words cannot conflict with variable names, because none of them
437 start with a prefix character (<tt>'%'</tt> or <tt>'@'</tt>).</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]438
439<p>Here is an example of LLVM code to multiply the integer variable
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]440 '<tt>%X</tt>' by 8:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]441
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]442<p>The easy way:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]443
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]444<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]445%result = <a href="#i_mul">mul</a> i32 %X, 8
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]446</pre>
447
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]448<p>After strength reduction:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]449
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]450<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]451%result = <a href="#i_shl">shl</a> i32 %X, i8 3
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]452</pre>
453
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]454<p>And the hard way:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]455
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]456<pre class="doc_code">
Gabor Greifec58f752009-10-28 13:05:07 +0000[diff] [blame]457%0 = <a href="#i_add">add</a> i32 %X, %X <i>; yields {i32}:%0</i>
458%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]459%result = <a href="#i_add">add</a> i32 %1, %1
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]460</pre>
461
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]462<p>This last way of multiplying <tt>%X</tt> by 8 illustrates several important
463 lexical features of LLVM:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]464
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]465<ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]466 <li>Comments are delimited with a '<tt>;</tt>' and go until the end of
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]467 line.</li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]468
469 <li>Unnamed temporaries are created when the result of a computation is not
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]470 assigned to a named value.</li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]471
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]472 <li>Unnamed temporaries are numbered sequentially</li>
473</ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]474
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]475<p>It also shows a convention that we follow in this document. When
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]476 demonstrating instructions, we will follow an instruction with a comment that
477 defines the type and name of value produced. Comments are shown in italic
478 text.</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]479
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]480</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]481
482<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]483<h2><a name="highlevel">High Level Structure</a></h2>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]484<!-- *********************************************************************** -->
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]485<div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]486<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]487<h3>
488 <a name="modulestructure">Module Structure</a>
489</h3>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]490
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]491<div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]492
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]493<p>LLVM programs are composed of "Module"s, each of which is a translation unit
494 of the input programs. Each module consists of functions, global variables,
495 and symbol table entries. Modules may be combined together with the LLVM
496 linker, which merges function (and global variable) definitions, resolves
497 forward declarations, and merges symbol table entries. Here is an example of
498 the "hello world" module:</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]499
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]500<pre class="doc_code">
Chris Lattner63e4ccb2010-08-17 17:13:42 +0000[diff] [blame]501<i>; Declare the string constant as a global constant.</i>&nbsp;
Nick Lewyckydb9cd762011-01-29 01:09:53 +0000[diff] [blame]502<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]503
Chris Lattner63e4ccb2010-08-17 17:13:42 +0000[diff] [blame]504<i>; External declaration of the puts function</i>&nbsp;
505<a href="#functionstructure">declare</a> i32 @puts(i8*) <i>; i32 (i8*)* </i>&nbsp;
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]506
507<i>; Definition of main function</i>
Chris Lattner63e4ccb2010-08-17 17:13:42 +0000[diff] [blame]508define i32 @main() { <i>; i32()* </i>&nbsp;
509 <i>; Convert [13 x i8]* to i8 *...</i>&nbsp;
510 %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]511
Chris Lattner63e4ccb2010-08-17 17:13:42 +0000[diff] [blame]512 <i>; Call puts function to write out the string to stdout.</i>&nbsp;
513 <a href="#i_call">call</a> i32 @puts(i8* %cast210) <i>; i32</i>&nbsp;
514 <a href="#i_ret">ret</a> i32 0&nbsp;
515}
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]516
517<i>; Named metadata</i>
518!1 = metadata !{i32 41}
519!foo = !{!1, null}
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]520</pre>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]521
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]522<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]523 "<tt>.LC0</tt>", an external declaration of the "<tt>puts</tt>" function,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]524 a <a href="#functionstructure">function definition</a> for
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]525 "<tt>main</tt>" and <a href="#namedmetadatastructure">named metadata</a>
526 "<tt>foo"</tt>.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]527
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]528<p>In general, a module is made up of a list of global values, where both
529 functions and global variables are global values. Global values are
530 represented by a pointer to a memory location (in this case, a pointer to an
531 array of char, and a pointer to a function), and have one of the
532 following <a href="#linkage">linkage types</a>.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]533
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]534</div>
535
536<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]537<h3>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]538 <a name="linkage">Linkage Types</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]539</h3>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]540
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]541<div>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]542
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]543<p>All Global Variables and Functions have one of the following types of
544 linkage:</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]545
546<dl>
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]547 <dt><tt><b><a name="linkage_private">private</a></b></tt></dt>
Bill Wendling5e721d72010-07-01 21:55:59 +0000[diff] [blame]548 <dd>Global values with "<tt>private</tt>" linkage are only directly accessible
549 by objects in the current module. In particular, linking code into a
550 module with an private global value may cause the private to be renamed as
551 necessary to avoid collisions. Because the symbol is private to the
552 module, all references can be updated. This doesn't show up in any symbol
553 table in the object file.</dd>
Rafael Espindolabb46f522009-01-15 20:18:42 +0000[diff] [blame]554
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]555 <dt><tt><b><a name="linkage_linker_private">linker_private</a></b></tt></dt>
Bill Wendling5e721d72010-07-01 21:55:59 +0000[diff] [blame]556 <dd>Similar to <tt>private</tt>, but the symbol is passed through the
557 assembler and evaluated by the linker. Unlike normal strong symbols, they
558 are removed by the linker from the final linked image (executable or
559 dynamic library).</dd>
560
561 <dt><tt><b><a name="linkage_linker_private_weak">linker_private_weak</a></b></tt></dt>
562 <dd>Similar to "<tt>linker_private</tt>", but the symbol is weak. Note that
563 <tt>linker_private_weak</tt> symbols are subject to coalescing by the
564 linker. The symbols are removed by the linker from the final linked image
565 (executable or dynamic library).</dd>
Bill Wendling3d10a5a2009-07-20 01:03:30 +0000[diff] [blame]566
Bill Wendling55ae5152010-08-20 22:05:50 +0000[diff] [blame]567 <dt><tt><b><a name="linkage_linker_private_weak_def_auto">linker_private_weak_def_auto</a></b></tt></dt>
568 <dd>Similar to "<tt>linker_private_weak</tt>", but it's known that the address
569 of the object is not taken. For instance, functions that had an inline
570 definition, but the compiler decided not to inline it. Note,
571 unlike <tt>linker_private</tt> and <tt>linker_private_weak</tt>,
572 <tt>linker_private_weak_def_auto</tt> may have only <tt>default</tt>
573 visibility. The symbols are removed by the linker from the final linked
574 image (executable or dynamic library).</dd>
575
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]576 <dt><tt><b><a name="linkage_internal">internal</a></b></tt></dt>
Bill Wendling07d31772010-06-29 22:34:52 +0000[diff] [blame]577 <dd>Similar to private, but the value shows as a local symbol
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]578 (<tt>STB_LOCAL</tt> in the case of ELF) in the object file. This
579 corresponds to the notion of the '<tt>static</tt>' keyword in C.</dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]580
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]581 <dt><tt><b><a name="linkage_available_externally">available_externally</a></b></tt></dt>
Chris Lattner266c7bb2009-04-13 05:44:34 +0000[diff] [blame]582 <dd>Globals with "<tt>available_externally</tt>" linkage are never emitted
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]583 into the object file corresponding to the LLVM module. They exist to
584 allow inlining and other optimizations to take place given knowledge of
585 the definition of the global, which is known to be somewhere outside the
586 module. Globals with <tt>available_externally</tt> linkage are allowed to
587 be discarded at will, and are otherwise the same as <tt>linkonce_odr</tt>.
588 This linkage type is only allowed on definitions, not declarations.</dd>
Chris Lattner266c7bb2009-04-13 05:44:34 +0000[diff] [blame]589
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]590 <dt><tt><b><a name="linkage_linkonce">linkonce</a></b></tt></dt>
Chris Lattner4887bd82007-01-14 06:51:48 +0000[diff] [blame]591 <dd>Globals with "<tt>linkonce</tt>" linkage are merged with other globals of
Chris Lattner873187c2010-01-09 19:15:14 +0000[diff] [blame]592 the same name when linkage occurs. This can be used to implement
593 some forms of inline functions, templates, or other code which must be
594 generated in each translation unit that uses it, but where the body may
595 be overridden with a more definitive definition later. Unreferenced
596 <tt>linkonce</tt> globals are allowed to be discarded. Note that
597 <tt>linkonce</tt> linkage does not actually allow the optimizer to
598 inline the body of this function into callers because it doesn't know if
599 this definition of the function is the definitive definition within the
600 program or whether it will be overridden by a stronger definition.
601 To enable inlining and other optimizations, use "<tt>linkonce_odr</tt>"
602 linkage.</dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]603
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]604 <dt><tt><b><a name="linkage_weak">weak</a></b></tt></dt>
Chris Lattner26d054d2009-08-05 05:21:07 +0000[diff] [blame]605 <dd>"<tt>weak</tt>" linkage has the same merging semantics as
606 <tt>linkonce</tt> linkage, except that unreferenced globals with
607 <tt>weak</tt> linkage may not be discarded. This is used for globals that
608 are declared "weak" in C source code.</dd>
609
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]610 <dt><tt><b><a name="linkage_common">common</a></b></tt></dt>
Chris Lattner26d054d2009-08-05 05:21:07 +0000[diff] [blame]611 <dd>"<tt>common</tt>" linkage is most similar to "<tt>weak</tt>" linkage, but
612 they are used for tentative definitions in C, such as "<tt>int X;</tt>" at
613 global scope.
614 Symbols with "<tt>common</tt>" linkage are merged in the same way as
615 <tt>weak symbols</tt>, and they may not be deleted if unreferenced.
Chris Lattnercd81f5d2009-08-05 05:41:44 +0000[diff] [blame]616 <tt>common</tt> symbols may not have an explicit section,
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]617 must have a zero initializer, and may not be marked '<a
Chris Lattnercd81f5d2009-08-05 05:41:44 +0000[diff] [blame]618 href="#globalvars"><tt>constant</tt></a>'. Functions and aliases may not
619 have common linkage.</dd>
Chris Lattner26d054d2009-08-05 05:21:07 +0000[diff] [blame]620
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]621
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]622 <dt><tt><b><a name="linkage_appending">appending</a></b></tt></dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]623 <dd>"<tt>appending</tt>" linkage may only be applied to global variables of
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]624 pointer to array type. When two global variables with appending linkage
625 are linked together, the two global arrays are appended together. This is
626 the LLVM, typesafe, equivalent of having the system linker append together
627 "sections" with identical names when .o files are linked.</dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]628
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]629 <dt><tt><b><a name="linkage_externweak">extern_weak</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]630 <dd>The semantics of this linkage follow the ELF object file model: the symbol
631 is weak until linked, if not linked, the symbol becomes null instead of
632 being an undefined reference.</dd>
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame]633
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]634 <dt><tt><b><a name="linkage_linkonce_odr">linkonce_odr</a></b></tt></dt>
635 <dt><tt><b><a name="linkage_weak_odr">weak_odr</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]636 <dd>Some languages allow differing globals to be merged, such as two functions
637 with different semantics. Other languages, such as <tt>C++</tt>, ensure
Bill Wendling5e721d72010-07-01 21:55:59 +0000[diff] [blame]638 that only equivalent globals are ever merged (the "one definition rule"
639 &mdash; "ODR"). Such languages can use the <tt>linkonce_odr</tt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]640 and <tt>weak_odr</tt> linkage types to indicate that the global will only
641 be merged with equivalent globals. These linkage types are otherwise the
642 same as their non-<tt>odr</tt> versions.</dd>
Duncan Sands667d4b82009-03-07 15:45:40 +0000[diff] [blame]643
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]644 <dt><tt><b><a name="linkage_external">externally visible</a></b></tt>:</dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]645 <dd>If none of the above identifiers are used, the global is externally
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]646 visible, meaning that it participates in linkage and can be used to
647 resolve external symbol references.</dd>
Reid Spencerc8910842007-04-11 23:49:50 +0000[diff] [blame]648</dl>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]649
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]650<p>The next two types of linkage are targeted for Microsoft Windows platform
651 only. They are designed to support importing (exporting) symbols from (to)
652 DLLs (Dynamic Link Libraries).</p>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]653
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]654<dl>
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]655 <dt><tt><b><a name="linkage_dllimport">dllimport</a></b></tt></dt>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]656 <dd>"<tt>dllimport</tt>" linkage causes the compiler to reference a function
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]657 or variable via a global pointer to a pointer that is set up by the DLL
658 exporting the symbol. On Microsoft Windows targets, the pointer name is
659 formed by combining <code>__imp_</code> and the function or variable
660 name.</dd>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]661
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]662 <dt><tt><b><a name="linkage_dllexport">dllexport</a></b></tt></dt>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]663 <dd>"<tt>dllexport</tt>" linkage causes the compiler to provide a global
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]664 pointer to a pointer in a DLL, so that it can be referenced with the
665 <tt>dllimport</tt> attribute. On Microsoft Windows targets, the pointer
666 name is formed by combining <code>__imp_</code> and the function or
667 variable name.</dd>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]668</dl>
669
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]670<p>For example, since the "<tt>.LC0</tt>" variable is defined to be internal, if
671 another module defined a "<tt>.LC0</tt>" variable and was linked with this
672 one, one of the two would be renamed, preventing a collision. Since
673 "<tt>main</tt>" and "<tt>puts</tt>" are external (i.e., lacking any linkage
674 declarations), they are accessible outside of the current module.</p>
675
676<p>It is illegal for a function <i>declaration</i> to have any linkage type
677 other than "externally visible", <tt>dllimport</tt>
678 or <tt>extern_weak</tt>.</p>
679
Duncan Sands667d4b82009-03-07 15:45:40 +0000[diff] [blame]680<p>Aliases can have only <tt>external</tt>, <tt>internal</tt>, <tt>weak</tt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]681 or <tt>weak_odr</tt> linkages.</p>
682
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]683</div>
684
685<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]686<h3>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]687 <a name="callingconv">Calling Conventions</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]688</h3>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]689
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]690<div>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]691
692<p>LLVM <a href="#functionstructure">functions</a>, <a href="#i_call">calls</a>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]693 and <a href="#i_invoke">invokes</a> can all have an optional calling
694 convention specified for the call. The calling convention of any pair of
695 dynamic caller/callee must match, or the behavior of the program is
696 undefined. The following calling conventions are supported by LLVM, and more
697 may be added in the future:</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]698
699<dl>
700 <dt><b>"<tt>ccc</tt>" - The C calling convention</b>:</dt>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]701 <dd>This calling convention (the default if no other calling convention is
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]702 specified) matches the target C calling conventions. This calling
703 convention supports varargs function calls and tolerates some mismatch in
704 the declared prototype and implemented declaration of the function (as
705 does normal C).</dd>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]706
707 <dt><b>"<tt>fastcc</tt>" - The fast calling convention</b>:</dt>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]708 <dd>This calling convention attempts to make calls as fast as possible
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]709 (e.g. by passing things in registers). This calling convention allows the
710 target to use whatever tricks it wants to produce fast code for the
711 target, without having to conform to an externally specified ABI
Jeffrey Yasskin95fa80a2010-01-09 19:44:16 +0000[diff] [blame]712 (Application Binary Interface).
713 <a href="CodeGenerator.html#tailcallopt">Tail calls can only be optimized
Chris Lattner29689432010-03-11 00:22:57 +0000[diff] [blame]714 when this or the GHC convention is used.</a> This calling convention
715 does not support varargs and requires the prototype of all callees to
716 exactly match the prototype of the function definition.</dd>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]717
718 <dt><b>"<tt>coldcc</tt>" - The cold calling convention</b>:</dt>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]719 <dd>This calling convention attempts to make code in the caller as efficient
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]720 as possible under the assumption that the call is not commonly executed.
721 As such, these calls often preserve all registers so that the call does
722 not break any live ranges in the caller side. This calling convention
723 does not support varargs and requires the prototype of all callees to
724 exactly match the prototype of the function definition.</dd>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]725
Chris Lattner29689432010-03-11 00:22:57 +0000[diff] [blame]726 <dt><b>"<tt>cc <em>10</em></tt>" - GHC convention</b>:</dt>
727 <dd>This calling convention has been implemented specifically for use by the
728 <a href="http://www.haskell.org/ghc">Glasgow Haskell Compiler (GHC)</a>.
729 It passes everything in registers, going to extremes to achieve this by
730 disabling callee save registers. This calling convention should not be
731 used lightly but only for specific situations such as an alternative to
732 the <em>register pinning</em> performance technique often used when
733 implementing functional programming languages.At the moment only X86
734 supports this convention and it has the following limitations:
735 <ul>
736 <li>On <em>X86-32</em> only supports up to 4 bit type parameters. No
737 floating point types are supported.</li>
738 <li>On <em>X86-64</em> only supports up to 10 bit type parameters and
739 6 floating point parameters.</li>
740 </ul>
741 This calling convention supports
742 <a href="CodeGenerator.html#tailcallopt">tail call optimization</a> but
743 requires both the caller and callee are using it.
744 </dd>
745
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]746 <dt><b>"<tt>cc &lt;<em>n</em>&gt;</tt>" - Numbered convention</b>:</dt>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]747 <dd>Any calling convention may be specified by number, allowing
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]748 target-specific calling conventions to be used. Target specific calling
749 conventions start at 64.</dd>
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]750</dl>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]751
752<p>More calling conventions can be added/defined on an as-needed basis, to
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]753 support Pascal conventions or any other well-known target-independent
754 convention.</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]755
756</div>
757
758<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]759<h3>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]760 <a name="visibility">Visibility Styles</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]761</h3>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]762
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]763<div>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]764
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]765<p>All Global Variables and Functions have one of the following visibility
766 styles:</p>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]767
768<dl>
769 <dt><b>"<tt>default</tt>" - Default style</b>:</dt>
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]770 <dd>On targets that use the ELF object file format, default visibility means
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]771 that the declaration is visible to other modules and, in shared libraries,
772 means that the declared entity may be overridden. On Darwin, default
773 visibility means that the declaration is visible to other modules. Default
774 visibility corresponds to "external linkage" in the language.</dd>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]775
776 <dt><b>"<tt>hidden</tt>" - Hidden style</b>:</dt>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]777 <dd>Two declarations of an object with hidden visibility refer to the same
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]778 object if they are in the same shared object. Usually, hidden visibility
779 indicates that the symbol will not be placed into the dynamic symbol
780 table, so no other module (executable or shared library) can reference it
781 directly.</dd>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]782
Anton Korobeynikov6f9896f2007-04-29 18:35:00 +0000[diff] [blame]783 <dt><b>"<tt>protected</tt>" - Protected style</b>:</dt>
Anton Korobeynikov6f9896f2007-04-29 18:35:00 +0000[diff] [blame]784 <dd>On ELF, protected visibility indicates that the symbol will be placed in
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]785 the dynamic symbol table, but that references within the defining module
786 will bind to the local symbol. That is, the symbol cannot be overridden by
787 another module.</dd>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]788</dl>
789
790</div>
791
792<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]793<h3>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]794 <a name="namedtypes">Named Types</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]795</h3>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]796
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]797<div>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]798
799<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]800 it easier to read the IR and make the IR more condensed (particularly when
801 recursive types are involved). An example of a name specification is:</p>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]802
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]803<pre class="doc_code">
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]804%mytype = type { %mytype*, i32 }
805</pre>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]806
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]807<p>You may give a name to any <a href="#typesystem">type</a> except
Chris Lattnerdc65f222010-08-17 23:26:04 +0000[diff] [blame]808 "<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]809 is expected with the syntax "%mytype".</p>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]810
811<p>Note that type names are aliases for the structural type that they indicate,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]812 and that you can therefore specify multiple names for the same type. This
813 often leads to confusing behavior when dumping out a .ll file. Since LLVM IR
814 uses structural typing, the name is not part of the type. When printing out
815 LLVM IR, the printer will pick <em>one name</em> to render all types of a
816 particular shape. This means that if you have code where two different
817 source types end up having the same LLVM type, that the dumper will sometimes
818 print the "wrong" or unexpected type. This is an important design point and
819 isn't going to change.</p>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]820
821</div>
822
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]823<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]824<h3>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]825 <a name="globalvars">Global Variables</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]826</h3>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]827
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]828<div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]829
Chris Lattner3689a342005-02-12 19:30:21 +0000[diff] [blame]830<p>Global variables define regions of memory allocated at compilation time
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]831 instead of run-time. Global variables may optionally be initialized, may
832 have an explicit section to be placed in, and may have an optional explicit
833 alignment specified. A variable may be defined as "thread_local", which
834 means that it will not be shared by threads (each thread will have a
835 separated copy of the variable). A variable may be defined as a global
836 "constant," which indicates that the contents of the variable
837 will <b>never</b> be modified (enabling better optimization, allowing the
838 global data to be placed in the read-only section of an executable, etc).
839 Note that variables that need runtime initialization cannot be marked
840 "constant" as there is a store to the variable.</p>
Chris Lattner3689a342005-02-12 19:30:21 +0000[diff] [blame]841
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]842<p>LLVM explicitly allows <em>declarations</em> of global variables to be marked
843 constant, even if the final definition of the global is not. This capability
844 can be used to enable slightly better optimization of the program, but
845 requires the language definition to guarantee that optimizations based on the
846 'constantness' are valid for the translation units that do not include the
847 definition.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]848
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]849<p>As SSA values, global variables define pointer values that are in scope
850 (i.e. they dominate) all basic blocks in the program. Global variables
851 always define a pointer to their "content" type because they describe a
852 region of memory, and all memory objects in LLVM are accessed through
853 pointers.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]854
Rafael Espindolabea46262011-01-08 16:42:36 +0000[diff] [blame]855<p>Global variables can be marked with <tt>unnamed_addr</tt> which indicates
856 that the address is not significant, only the content. Constants marked
Rafael Espindolaa5eaa862011-01-15 08:20:57 +0000[diff] [blame]857 like this can be merged with other constants if they have the same
858 initializer. Note that a constant with significant address <em>can</em>
859 be merged with a <tt>unnamed_addr</tt> constant, the result being a
860 constant whose address is significant.</p>
Rafael Espindolabea46262011-01-08 16:42:36 +0000[diff] [blame]861
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]862<p>A global variable may be declared to reside in a target-specific numbered
863 address space. For targets that support them, address spaces may affect how
864 optimizations are performed and/or what target instructions are used to
865 access the variable. The default address space is zero. The address space
866 qualifier must precede any other attributes.</p>
Christopher Lamb284d9922007-12-11 09:31:00 +0000[diff] [blame]867
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]868<p>LLVM allows an explicit section to be specified for globals. If the target
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]869 supports it, it will emit globals to the section specified.</p>
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]870
Chris Lattnerce99fa92010-04-28 00:13:42 +0000[diff] [blame]871<p>An explicit alignment may be specified for a global, which must be a power
872 of 2. If not present, or if the alignment is set to zero, the alignment of
873 the global is set by the target to whatever it feels convenient. If an
874 explicit alignment is specified, the global is forced to have exactly that
Chris Lattner2d4b8ee2010-04-28 00:31:12 +0000[diff] [blame]875 alignment. Targets and optimizers are not allowed to over-align the global
876 if the global has an assigned section. In this case, the extra alignment
877 could be observable: for example, code could assume that the globals are
878 densely packed in their section and try to iterate over them as an array,
879 alignment padding would break this iteration.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]880
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]881<p>For example, the following defines a global in a numbered address space with
882 an initializer, section, and alignment:</p>
Chris Lattner68027ea2007-01-14 00:27:09 +0000[diff] [blame]883
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]884<pre class="doc_code">
Dan Gohman398873c2009-01-11 00:40:00 +0000[diff] [blame]885@G = addrspace(5) constant float 1.0, section "foo", align 4
Chris Lattner68027ea2007-01-14 00:27:09 +0000[diff] [blame]886</pre>
887
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]888</div>
889
890
891<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]892<h3>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]893 <a name="functionstructure">Functions</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]894</h3>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]895
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]896<div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]897
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]898<p>LLVM function definitions consist of the "<tt>define</tt>" keyword, an
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]899 optional <a href="#linkage">linkage type</a>, an optional
900 <a href="#visibility">visibility style</a>, an optional
Rafael Espindolabea46262011-01-08 16:42:36 +0000[diff] [blame]901 <a href="#callingconv">calling convention</a>,
902 an optional <tt>unnamed_addr</tt> attribute, a return type, an optional
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]903 <a href="#paramattrs">parameter attribute</a> for the return type, a function
904 name, a (possibly empty) argument list (each with optional
905 <a href="#paramattrs">parameter attributes</a>), optional
906 <a href="#fnattrs">function attributes</a>, an optional section, an optional
907 alignment, an optional <a href="#gc">garbage collector name</a>, an opening
908 curly brace, a list of basic blocks, and a closing curly brace.</p>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]909
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]910<p>LLVM function declarations consist of the "<tt>declare</tt>" keyword, an
911 optional <a href="#linkage">linkage type</a>, an optional
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]912 <a href="#visibility">visibility style</a>, an optional
Rafael Espindolabea46262011-01-08 16:42:36 +0000[diff] [blame]913 <a href="#callingconv">calling convention</a>,
914 an optional <tt>unnamed_addr</tt> attribute, a return type, an optional
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]915 <a href="#paramattrs">parameter attribute</a> for the return type, a function
916 name, a possibly empty list of arguments, an optional alignment, and an
917 optional <a href="#gc">garbage collector name</a>.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]918
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]919<p>A function definition contains a list of basic blocks, forming the CFG
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]920 (Control Flow Graph) for the function. Each basic block may optionally start
921 with a label (giving the basic block a symbol table entry), contains a list
922 of instructions, and ends with a <a href="#terminators">terminator</a>
923 instruction (such as a branch or function return).</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]924
Chris Lattner4a3c9012007-06-08 16:52:14 +0000[diff] [blame]925<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]926 executed on entrance to the function, and it is not allowed to have
927 predecessor basic blocks (i.e. there can not be any branches to the entry
928 block of a function). Because the block can have no predecessors, it also
929 cannot have any <a href="#i_phi">PHI nodes</a>.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]930
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]931<p>LLVM allows an explicit section to be specified for functions. If the target
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]932 supports it, it will emit functions to the section specified.</p>
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]933
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]934<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]935 the alignment is set to zero, the alignment of the function is set by the
936 target to whatever it feels convenient. If an explicit alignment is
937 specified, the function is forced to have at least that much alignment. All
938 alignments must be a power of 2.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]939
Rafael Espindolabea46262011-01-08 16:42:36 +0000[diff] [blame]940<p>If the <tt>unnamed_addr</tt> attribute is given, the address is know to not
941 be significant and two identical functions can be merged</p>.
942
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]943<h5>Syntax:</h5>
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]944<pre class="doc_code">
Chris Lattner50ad45c2008-10-13 16:55:18 +0000[diff] [blame]945define [<a href="#linkage">linkage</a>] [<a href="#visibility">visibility</a>]
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]946 [<a href="#callingconv">cconv</a>] [<a href="#paramattrs">ret attrs</a>]
947 &lt;ResultType&gt; @&lt;FunctionName&gt; ([argument list])
948 [<a href="#fnattrs">fn Attrs</a>] [section "name"] [align N]
949 [<a href="#gc">gc</a>] { ... }
950</pre>
Devang Patel307e8ab2008-10-07 17:48:33 +0000[diff] [blame]951
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]952</div>
953
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]954<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]955<h3>
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]956 <a name="aliasstructure">Aliases</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]957</h3>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]958
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]959<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]960
961<p>Aliases act as "second name" for the aliasee value (which can be either
962 function, global variable, another alias or bitcast of global value). Aliases
963 may have an optional <a href="#linkage">linkage type</a>, and an
964 optional <a href="#visibility">visibility style</a>.</p>
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]965
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]966<h5>Syntax:</h5>
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]967<pre class="doc_code">
Duncan Sands0b23ac12008-09-12 20:48:21 +0000[diff] [blame]968@&lt;Name&gt; = alias [Linkage] [Visibility] &lt;AliaseeTy&gt; @&lt;Aliasee&gt;
Bill Wendlingaac388b2007-05-29 09:42:13 +0000[diff] [blame]969</pre>
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]970
971</div>
972
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]973<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]974<h3>
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]975 <a name="namedmetadatastructure">Named Metadata</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]976</h3>
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]977
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]978<div>
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]979
Chris Lattnere6a5ddd2010-01-15 21:50:19 +0000[diff] [blame]980<p>Named metadata is a collection of metadata. <a href="#metadata">Metadata
Dan Gohman872814a2010-07-21 18:54:18 +0000[diff] [blame]981 nodes</a> (but not metadata strings) are the only valid operands for
Chris Lattnere6a5ddd2010-01-15 21:50:19 +0000[diff] [blame]982 a named metadata.</p>
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]983
984<h5>Syntax:</h5>
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]985<pre class="doc_code">
Dan Gohman872814a2010-07-21 18:54:18 +0000[diff] [blame]986; Some unnamed metadata nodes, which are referenced by the named metadata.
987!0 = metadata !{metadata !"zero"}
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]988!1 = metadata !{metadata !"one"}
Dan Gohman872814a2010-07-21 18:54:18 +0000[diff] [blame]989!2 = metadata !{metadata !"two"}
Dan Gohman1005bc52010-07-13 19:48:13 +0000[diff] [blame]990; A named metadata.
Dan Gohman872814a2010-07-21 18:54:18 +0000[diff] [blame]991!name = !{!0, !1, !2}
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]992</pre>
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]993
994</div>
995
996<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]997<h3>
998 <a name="paramattrs">Parameter Attributes</a>
999</h3>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1000
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1001<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1002
1003<p>The return type and each parameter of a function type may have a set of
1004 <i>parameter attributes</i> associated with them. Parameter attributes are
1005 used to communicate additional information about the result or parameters of
1006 a function. Parameter attributes are considered to be part of the function,
1007 not of the function type, so functions with different parameter attributes
1008 can have the same function type.</p>
1009
1010<p>Parameter attributes are simple keywords that follow the type specified. If
1011 multiple parameter attributes are needed, they are space separated. For
1012 example:</p>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1013
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]1014<pre class="doc_code">
Nick Lewyckyb6a7d252009-02-15 23:06:14 +0000[diff] [blame]1015declare i32 @printf(i8* noalias nocapture, ...)
Chris Lattner66d922c2008-10-04 18:33:34 +0000[diff] [blame]1016declare i32 @atoi(i8 zeroext)
1017declare signext i8 @returns_signed_char()
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1018</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1019
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1020<p>Note that any attributes for the function result (<tt>nounwind</tt>,
1021 <tt>readonly</tt>) come immediately after the argument list.</p>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1022
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1023<p>Currently, only the following parameter attributes are defined:</p>
Chris Lattner47507de2008-01-11 06:20:47 +0000[diff] [blame]1024
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1025<dl>
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1026 <dt><tt><b>zeroext</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1027 <dd>This indicates to the code generator that the parameter or return value
Cameron Zwarichebe81732011-03-16 22:20:18 +0000[diff] [blame]1028 should be zero-extended to the extent required by the target's ABI (which
1029 is usually 32-bits, but is 8-bits for a i1 on x86-64) by the caller (for a
1030 parameter) or the callee (for a return value).</dd>
Chris Lattner47507de2008-01-11 06:20:47 +0000[diff] [blame]1031
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1032 <dt><tt><b>signext</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1033 <dd>This indicates to the code generator that the parameter or return value
Cameron Zwarich9e69ff92011-03-17 14:21:58 +0000[diff] [blame]1034 should be sign-extended to the extent required by the target's ABI (which
1035 is usually 32-bits) by the caller (for a parameter) or the callee (for a
1036 return value).</dd>
Chris Lattner47507de2008-01-11 06:20:47 +0000[diff] [blame]1037
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1038 <dt><tt><b>inreg</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1039 <dd>This indicates that this parameter or return value should be treated in a
1040 special target-dependent fashion during while emitting code for a function
1041 call or return (usually, by putting it in a register as opposed to memory,
1042 though some targets use it to distinguish between two different kinds of
1043 registers). Use of this attribute is target-specific.</dd>
Chris Lattner47507de2008-01-11 06:20:47 +0000[diff] [blame]1044
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1045 <dt><tt><b><a name="byval">byval</a></b></tt></dt>
Chris Lattnera6fd81d2010-11-20 23:49:06 +0000[diff] [blame]1046 <dd><p>This indicates that the pointer parameter should really be passed by
1047 value to the function. The attribute implies that a hidden copy of the
1048 pointee
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1049 is made between the caller and the callee, so the callee is unable to
1050 modify the value in the callee. This attribute is only valid on LLVM
1051 pointer arguments. It is generally used to pass structs and arrays by
1052 value, but is also valid on pointers to scalars. The copy is considered
1053 to belong to the caller not the callee (for example,
1054 <tt><a href="#readonly">readonly</a></tt> functions should not write to
1055 <tt>byval</tt> parameters). This is not a valid attribute for return
Chris Lattnera6fd81d2010-11-20 23:49:06 +0000[diff] [blame]1056 values.</p>
1057
1058 <p>The byval attribute also supports specifying an alignment with
1059 the align attribute. It indicates the alignment of the stack slot to
1060 form and the known alignment of the pointer specified to the call site. If
1061 the alignment is not specified, then the code generator makes a
1062 target-specific assumption.</p></dd>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1063
Dan Gohmanff235352010-07-02 23:18:08 +0000[diff] [blame]1064 <dt><tt><b><a name="sret">sret</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1065 <dd>This indicates that the pointer parameter specifies the address of a
1066 structure that is the return value of the function in the source program.
1067 This pointer must be guaranteed by the caller to be valid: loads and
1068 stores to the structure may be assumed by the callee to not to trap. This
1069 may only be applied to the first parameter. This is not a valid attribute
1070 for return values. </dd>
1071
Dan Gohmanff235352010-07-02 23:18:08 +0000[diff] [blame]1072 <dt><tt><b><a name="noalias">noalias</a></b></tt></dt>
Dan Gohman1e109622010-07-02 18:41:32 +0000[diff] [blame]1073 <dd>This indicates that pointer values
1074 <a href="#pointeraliasing"><i>based</i></a> on the argument or return
Dan Gohmanefca7f92010-07-02 23:46:54 +0000[diff] [blame]1075 value do not alias pointer values which are not <i>based</i> on it,
1076 ignoring certain "irrelevant" dependencies.
1077 For a call to the parent function, dependencies between memory
1078 references from before or after the call and from those during the call
1079 are "irrelevant" to the <tt>noalias</tt> keyword for the arguments and
1080 return value used in that call.
Dan Gohman1e109622010-07-02 18:41:32 +0000[diff] [blame]1081 The caller shares the responsibility with the callee for ensuring that
1082 these requirements are met.
1083 For further details, please see the discussion of the NoAlias response in
Dan Gohmanff70fe42010-07-06 15:26:33 +0000[diff] [blame]1084 <a href="AliasAnalysis.html#MustMayNo">alias analysis</a>.<br>
1085<br>
John McCall191d4ee2010-07-06 21:07:14 +0000[diff] [blame]1086 Note that this definition of <tt>noalias</tt> is intentionally
1087 similar to the definition of <tt>restrict</tt> in C99 for function
Chris Lattner211244a2010-07-06 20:51:35 +0000[diff] [blame]1088 arguments, though it is slightly weaker.
Dan Gohmanff70fe42010-07-06 15:26:33 +0000[diff] [blame]1089<br>
1090 For function return values, C99's <tt>restrict</tt> is not meaningful,
1091 while LLVM's <tt>noalias</tt> is.
1092 </dd>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1093
Dan Gohmanff235352010-07-02 23:18:08 +0000[diff] [blame]1094 <dt><tt><b><a name="nocapture">nocapture</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1095 <dd>This indicates that the callee does not make any copies of the pointer
1096 that outlive the callee itself. This is not a valid attribute for return
1097 values.</dd>
1098
Dan Gohmanff235352010-07-02 23:18:08 +0000[diff] [blame]1099 <dt><tt><b><a name="nest">nest</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1100 <dd>This indicates that the pointer parameter can be excised using the
1101 <a href="#int_trampoline">trampoline intrinsics</a>. This is not a valid
1102 attribute for return values.</dd>
1103</dl>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1104
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1105</div>
1106
1107<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1108<h3>
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame]1109 <a name="gc">Garbage Collector Names</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1110</h3>
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame]1111
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1112<div>
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame]1113
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1114<p>Each function may specify a garbage collector name, which is simply a
1115 string:</p>
1116
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]1117<pre class="doc_code">
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1118define void @f() gc "name" { ... }
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1119</pre>
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame]1120
1121<p>The compiler declares the supported values of <i>name</i>. Specifying a
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1122 collector which will cause the compiler to alter its output in order to
1123 support the named garbage collection algorithm.</p>
1124
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame]1125</div>
1126
1127<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1128<h3>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1129 <a name="fnattrs">Function Attributes</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1130</h3>
Devang Patelf8b94812008-09-04 23:05:13 +0000[diff] [blame]1131
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1132<div>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1133
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1134<p>Function attributes are set to communicate additional information about a
1135 function. Function attributes are considered to be part of the function, not
1136 of the function type, so functions with different parameter attributes can
1137 have the same function type.</p>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1138
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1139<p>Function attributes are simple keywords that follow the type specified. If
1140 multiple attributes are needed, they are space separated. For example:</p>
Devang Patelf8b94812008-09-04 23:05:13 +0000[diff] [blame]1141
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]1142<pre class="doc_code">
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1143define void @f() noinline { ... }
1144define void @f() alwaysinline { ... }
1145define void @f() alwaysinline optsize { ... }
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1146define void @f() optsize { ... }
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1147</pre>
Devang Patelf8b94812008-09-04 23:05:13 +0000[diff] [blame]1148
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1149<dl>
Charles Davis1e063d12010-02-12 00:31:15 +0000[diff] [blame]1150 <dt><tt><b>alignstack(&lt;<em>n</em>&gt;)</b></tt></dt>
1151 <dd>This attribute indicates that, when emitting the prologue and epilogue,
1152 the backend should forcibly align the stack pointer. Specify the
1153 desired alignment, which must be a power of two, in parentheses.
1154
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1155 <dt><tt><b>alwaysinline</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1156 <dd>This attribute indicates that the inliner should attempt to inline this
1157 function into callers whenever possible, ignoring any active inlining size
1158 threshold for this caller.</dd>
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1159
Charles Davis970bfcc2010-10-25 15:37:09 +0000[diff] [blame]1160 <dt><tt><b>hotpatch</b></tt></dt>
Charles Davis6f12e292010-10-25 16:29:03 +0000[diff] [blame]1161 <dd>This attribute indicates that the function should be 'hotpatchable',
Charles Davis0076d202010-10-25 19:07:39 +0000[diff] [blame]1162 meaning the function can be patched and/or hooked even while it is
1163 loaded into memory. On x86, the function prologue will be preceded
1164 by six bytes of padding and will begin with a two-byte instruction.
1165 Most of the functions in the Windows system DLLs in Windows XP SP2 or
1166 higher were compiled in this fashion.</dd>
Charles Davis970bfcc2010-10-25 15:37:09 +0000[diff] [blame]1167
Dan Gohman129bd562011-06-16 16:03:13 +0000[diff] [blame]1168 <dt><tt><b>nonlazybind</b></tt></dt>
1169 <dd>This attribute suppresses lazy symbol binding for the function. This
1170 may make calls to the function faster, at the cost of extra program
1171 startup time if the function is not called during program startup.</dd>
1172
Jakob Stoklund Olesen570a4a52010-02-06 01:16:28 +0000[diff] [blame]1173 <dt><tt><b>inlinehint</b></tt></dt>
1174 <dd>This attribute indicates that the source code contained a hint that inlining
1175 this function is desirable (such as the "inline" keyword in C/C++). It
1176 is just a hint; it imposes no requirements on the inliner.</dd>
1177
Nick Lewycky76ec37a2010-07-06 18:24:09 +0000[diff] [blame]1178 <dt><tt><b>naked</b></tt></dt>
1179 <dd>This attribute disables prologue / epilogue emission for the function.
1180 This can have very system-specific consequences.</dd>
1181
1182 <dt><tt><b>noimplicitfloat</b></tt></dt>
1183 <dd>This attributes disables implicit floating point instructions.</dd>
1184
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1185 <dt><tt><b>noinline</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1186 <dd>This attribute indicates that the inliner should never inline this
1187 function in any situation. This attribute may not be used together with
1188 the <tt>alwaysinline</tt> attribute.</dd>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1189
Nick Lewycky76ec37a2010-07-06 18:24:09 +0000[diff] [blame]1190 <dt><tt><b>noredzone</b></tt></dt>
1191 <dd>This attribute indicates that the code generator should not use a red
1192 zone, even if the target-specific ABI normally permits it.</dd>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1193
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1194 <dt><tt><b>noreturn</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1195 <dd>This function attribute indicates that the function never returns
1196 normally. This produces undefined behavior at runtime if the function
1197 ever does dynamically return.</dd>
Bill Wendling31359ba2008-11-13 01:02:51 +0000[diff] [blame]1198
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1199 <dt><tt><b>nounwind</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1200 <dd>This function attribute indicates that the function never returns with an
1201 unwind or exceptional control flow. If the function does unwind, its
1202 runtime behavior is undefined.</dd>
Bill Wendlingfbaa7ed2008-11-26 19:07:40 +0000[diff] [blame]1203
Nick Lewycky76ec37a2010-07-06 18:24:09 +0000[diff] [blame]1204 <dt><tt><b>optsize</b></tt></dt>
1205 <dd>This attribute suggests that optimization passes and code generator passes
1206 make choices that keep the code size of this function low, and otherwise
1207 do optimizations specifically to reduce code size.</dd>
1208
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1209 <dt><tt><b>readnone</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1210 <dd>This attribute indicates that the function computes its result (or decides
1211 to unwind an exception) based strictly on its arguments, without
1212 dereferencing any pointer arguments or otherwise accessing any mutable
1213 state (e.g. memory, control registers, etc) visible to caller functions.
1214 It does not write through any pointer arguments
1215 (including <tt><a href="#byval">byval</a></tt> arguments) and never
1216 changes any state visible to callers. This means that it cannot unwind
1217 exceptions by calling the <tt>C++</tt> exception throwing methods, but
1218 could use the <tt>unwind</tt> instruction.</dd>
Devang Patel5d96fda2009-06-12 19:45:19 +0000[diff] [blame]1219
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1220 <dt><tt><b><a name="readonly">readonly</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1221 <dd>This attribute indicates that the function does not write through any
1222 pointer arguments (including <tt><a href="#byval">byval</a></tt>
1223 arguments) or otherwise modify any state (e.g. memory, control registers,
1224 etc) visible to caller functions. It may dereference pointer arguments
1225 and read state that may be set in the caller. A readonly function always
1226 returns the same value (or unwinds an exception identically) when called
1227 with the same set of arguments and global state. It cannot unwind an
1228 exception by calling the <tt>C++</tt> exception throwing methods, but may
1229 use the <tt>unwind</tt> instruction.</dd>
Anton Korobeynikovc5ec8a72009-07-17 18:07:26 +0000[diff] [blame]1230
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1231 <dt><tt><b><a name="ssp">ssp</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1232 <dd>This attribute indicates that the function should emit a stack smashing
1233 protector. It is in the form of a "canary"&mdash;a random value placed on
1234 the stack before the local variables that's checked upon return from the
1235 function to see if it has been overwritten. A heuristic is used to
1236 determine if a function needs stack protectors or not.<br>
1237<br>
1238 If a function that has an <tt>ssp</tt> attribute is inlined into a
1239 function that doesn't have an <tt>ssp</tt> attribute, then the resulting
1240 function will have an <tt>ssp</tt> attribute.</dd>
1241
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1242 <dt><tt><b>sspreq</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1243 <dd>This attribute indicates that the function should <em>always</em> emit a
1244 stack smashing protector. This overrides
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1245 the <tt><a href="#ssp">ssp</a></tt> function attribute.<br>
1246<br>
1247 If a function that has an <tt>sspreq</tt> attribute is inlined into a
1248 function that doesn't have an <tt>sspreq</tt> attribute or which has
1249 an <tt>ssp</tt> attribute, then the resulting function will have
1250 an <tt>sspreq</tt> attribute.</dd>
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1251</dl>
1252
Devang Patelf8b94812008-09-04 23:05:13 +0000[diff] [blame]1253</div>
1254
1255<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1256<h3>
Chris Lattner1eeeb0c2006-04-08 04:40:53 +0000[diff] [blame]1257 <a name="moduleasm">Module-Level Inline Assembly</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1258</h3>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]1259
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1260<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1261
1262<p>Modules may contain "module-level inline asm" blocks, which corresponds to
1263 the GCC "file scope inline asm" blocks. These blocks are internally
1264 concatenated by LLVM and treated as a single unit, but may be separated in
1265 the <tt>.ll</tt> file if desired. The syntax is very simple:</p>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]1266
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]1267<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1268module asm "inline asm code goes here"
1269module asm "more can go here"
1270</pre>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]1271
1272<p>The strings can contain any character by escaping non-printable characters.
1273 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]1274 for the number.</p>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]1275
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1276<p>The inline asm code is simply printed to the machine code .s file when
1277 assembly code is generated.</p>
1278
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]1279</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]1280
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1281<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1282<h3>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1283 <a name="datalayout">Data Layout</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1284</h3>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1285
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1286<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1287
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1288<p>A module may specify a target specific data layout string that specifies how
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1289 data is to be laid out in memory. The syntax for the data layout is
1290 simply:</p>
1291
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]1292<pre class="doc_code">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1293target datalayout = "<i>layout specification</i>"
1294</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1295
1296<p>The <i>layout specification</i> consists of a list of specifications
1297 separated by the minus sign character ('-'). Each specification starts with
1298 a letter and may include other information after the letter to define some
1299 aspect of the data layout. The specifications accepted are as follows:</p>
1300
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1301<dl>
1302 <dt><tt>E</tt></dt>
1303 <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]1304 bits with the most significance have the lowest address location.</dd>
1305
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1306 <dt><tt>e</tt></dt>
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]1307 <dd>Specifies that the target lays out data in little-endian form. That is,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1308 the bits with the least significance have the lowest address
1309 location.</dd>
1310
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1311 <dt><tt>p:<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]1312 <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]1313 <i>preferred</i> alignments. All sizes are in bits. Specifying
1314 the <i>pref</i> alignment is optional. If omitted, the
1315 preceding <tt>:</tt> should be omitted too.</dd>
1316
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1317 <dt><tt>i<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1318 <dd>This specifies the alignment for an integer type of a given bit
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1319 <i>size</i>. The value of <i>size</i> must be in the range [1,2^23).</dd>
1320
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1321 <dt><tt>v<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]1322 <dd>This specifies the alignment for a vector type of a given bit
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1323 <i>size</i>.</dd>
1324
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1325 <dt><tt>f<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]1326 <dd>This specifies the alignment for a floating point type of a given bit
Dale Johannesen9d8d2212010-05-28 18:54:47 +0000[diff] [blame]1327 <i>size</i>. Only values of <i>size</i> that are supported by the target
1328 will work. 32 (float) and 64 (double) are supported on all targets;
1329 80 or 128 (different flavors of long double) are also supported on some
1330 targets.
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1331
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1332 <dt><tt>a<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1333 <dd>This specifies the alignment for an aggregate type of a given bit
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1334 <i>size</i>.</dd>
1335
Daniel Dunbar87bde0b2009-06-08 22:17:53 +0000[diff] [blame]1336 <dt><tt>s<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1337 <dd>This specifies the alignment for a stack object of a given bit
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1338 <i>size</i>.</dd>
Chris Lattnere82bdc42009-11-07 09:35:34 +0000[diff] [blame]1339
1340 <dt><tt>n<i>size1</i>:<i>size2</i>:<i>size3</i>...</tt></dt>
1341 <dd>This specifies a set of native integer widths for the target CPU
1342 in bits. For example, it might contain "n32" for 32-bit PowerPC,
1343 "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]1344 this set are considered to support most general arithmetic
Chris Lattnere82bdc42009-11-07 09:35:34 +0000[diff] [blame]1345 operations efficiently.</dd>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1346</dl>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1347
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1348<p>When constructing the data layout for a given target, LLVM starts with a
Dan Gohman1c70c002010-04-28 00:36:01 +0000[diff] [blame]1349 default set of specifications which are then (possibly) overridden by the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1350 specifications in the <tt>datalayout</tt> keyword. The default specifications
1351 are given in this list:</p>
1352
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1353<ul>
1354 <li><tt>E</tt> - big endian</li>
Dan Gohmanfdf2e8c2010-02-23 02:44:03 +0000[diff] [blame]1355 <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]1356 <li><tt>i1:8:8</tt> - i1 is 8-bit (byte) aligned</li>
1357 <li><tt>i8:8:8</tt> - i8 is 8-bit (byte) aligned</li>
1358 <li><tt>i16:16:16</tt> - i16 is 16-bit aligned</li>
1359 <li><tt>i32:32:32</tt> - i32 is 32-bit aligned</li>
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]1360 <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]1361 alignment of 64-bits</li>
1362 <li><tt>f32:32:32</tt> - float is 32-bit aligned</li>
1363 <li><tt>f64:64:64</tt> - double is 64-bit aligned</li>
1364 <li><tt>v64:64:64</tt> - 64-bit vector is 64-bit aligned</li>
1365 <li><tt>v128:128:128</tt> - 128-bit vector is 128-bit aligned</li>
1366 <li><tt>a0:0:1</tt> - aggregates are 8-bit aligned</li>
Daniel Dunbar87bde0b2009-06-08 22:17:53 +0000[diff] [blame]1367 <li><tt>s0:64:64</tt> - stack objects are 64-bit aligned</li>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1368</ul>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1369
1370<p>When LLVM is determining the alignment for a given type, it uses the
1371 following rules:</p>
1372
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1373<ol>
1374 <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]1375 specification is used.</li>
1376
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1377 <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]1378 smallest integer type that is larger than the bitwidth of the sought type
1379 is used. If none of the specifications are larger than the bitwidth then
1380 the the largest integer type is used. For example, given the default
1381 specifications above, the i7 type will use the alignment of i8 (next
1382 largest) while both i65 and i256 will use the alignment of i64 (largest
1383 specified).</li>
1384
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1385 <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]1386 largest vector type that is smaller than the sought vector type will be
1387 used as a fall back. This happens because &lt;128 x double&gt; can be
1388 implemented in terms of 64 &lt;2 x double&gt;, for example.</li>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1389</ol>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1390
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1391</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]1392
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1393<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1394<h3>
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1395 <a name="pointeraliasing">Pointer Aliasing Rules</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1396</h3>
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1397
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1398<div>
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1399
Andreas Bolka55e459a2009-07-29 00:02:05 +0000[diff] [blame]1400<p>Any memory access must be done through a pointer value associated
Andreas Bolka99a82052009-07-27 20:37:10 +0000[diff] [blame]1401with an address range of the memory access, otherwise the behavior
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1402is undefined. Pointer values are associated with address ranges
1403according to the following rules:</p>
1404
1405<ul>
Dan Gohman1e109622010-07-02 18:41:32 +0000[diff] [blame]1406 <li>A pointer value is associated with the addresses associated with
1407 any value it is <i>based</i> on.
Andreas Bolka55e459a2009-07-29 00:02:05 +0000[diff] [blame]1408 <li>An address of a global variable is associated with the address
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1409 range of the variable's storage.</li>
1410 <li>The result value of an allocation instruction is associated with
1411 the address range of the allocated storage.</li>
1412 <li>A null pointer in the default address-space is associated with
Andreas Bolka55e459a2009-07-29 00:02:05 +0000[diff] [blame]1413 no address.</li>
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1414 <li>An integer constant other than zero or a pointer value returned
1415 from a function not defined within LLVM may be associated with address
1416 ranges allocated through mechanisms other than those provided by
Andreas Bolka55e459a2009-07-29 00:02:05 +0000[diff] [blame]1417 LLVM. Such ranges shall not overlap with any ranges of addresses
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1418 allocated by mechanisms provided by LLVM.</li>
Dan Gohman1e109622010-07-02 18:41:32 +0000[diff] [blame]1419</ul>
1420
1421<p>A pointer value is <i>based</i> on another pointer value according
1422 to the following rules:</p>
1423
1424<ul>
1425 <li>A pointer value formed from a
1426 <tt><a href="#i_getelementptr">getelementptr</a></tt> operation
1427 is <i>based</i> on the first operand of the <tt>getelementptr</tt>.</li>
1428 <li>The result value of a
1429 <tt><a href="#i_bitcast">bitcast</a></tt> is <i>based</i> on the operand
1430 of the <tt>bitcast</tt>.</li>
1431 <li>A pointer value formed by an
1432 <tt><a href="#i_inttoptr">inttoptr</a></tt> is <i>based</i> on all
1433 pointer values that contribute (directly or indirectly) to the
1434 computation of the pointer's value.</li>
1435 <li>The "<i>based</i> on" relationship is transitive.</li>
1436</ul>
1437
1438<p>Note that this definition of <i>"based"</i> is intentionally
1439 similar to the definition of <i>"based"</i> in C99, though it is
1440 slightly weaker.</p>
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1441
1442<p>LLVM IR does not associate types with memory. The result type of a
Andreas Bolka55e459a2009-07-29 00:02:05 +0000[diff] [blame]1443<tt><a href="#i_load">load</a></tt> merely indicates the size and
1444alignment of the memory from which to load, as well as the
Dan Gohmanc22c0f32010-06-17 19:23:50 +0000[diff] [blame]1445interpretation of the value. The first operand type of a
Andreas Bolka55e459a2009-07-29 00:02:05 +0000[diff] [blame]1446<tt><a href="#i_store">store</a></tt> similarly only indicates the size
1447and alignment of the store.</p>
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1448
1449<p>Consequently, type-based alias analysis, aka TBAA, aka
1450<tt>-fstrict-aliasing</tt>, is not applicable to general unadorned
1451LLVM IR. <a href="#metadata">Metadata</a> may be used to encode
1452additional information which specialized optimization passes may use
1453to implement type-based alias analysis.</p>
1454
1455</div>
1456
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]1457<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1458<h3>
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]1459 <a name="volatile">Volatile Memory Accesses</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1460</h3>
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]1461
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1462<div>
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]1463
1464<p>Certain memory accesses, such as <a href="#i_load"><tt>load</tt></a>s, <a
1465href="#i_store"><tt>store</tt></a>s, and <a
1466href="#int_memcpy"><tt>llvm.memcpy</tt></a>s may be marked <tt>volatile</tt>.
1467The optimizers must not change the number of volatile operations or change their
1468order of execution relative to other volatile operations. The optimizers
1469<i>may</i> change the order of volatile operations relative to non-volatile
1470operations. This is not Java's "volatile" and has no cross-thread
1471synchronization behavior.</p>
1472
1473</div>
1474
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1475</div>
1476
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1477<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1478<h2><a name="typesystem">Type System</a></h2>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1479<!-- *********************************************************************** -->
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]1480
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1481<div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]1482
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1483<p>The LLVM type system is one of the most important features of the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1484 intermediate representation. Being typed enables a number of optimizations
1485 to be performed on the intermediate representation directly, without having
1486 to do extra analyses on the side before the transformation. A strong type
1487 system makes it easier to read the generated code and enables novel analyses
1488 and transformations that are not feasible to perform on normal three address
1489 code representations.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]1490
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1491<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1492<h3>
1493 <a name="t_classifications">Type Classifications</a>
1494</h3>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1495
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1496<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1497
1498<p>The types fall into a few useful classifications:</p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1499
1500<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1501 <tbody>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1502 <tr><th>Classification</th><th>Types</th></tr>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1503 <tr>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1504 <td><a href="#t_integer">integer</a></td>
Reid Spencer2b916312007-05-16 18:44:01 +0000[diff] [blame]1505 <td><tt>i1, i2, i3, ... i8, ... i16, ... i32, ... i64, ... </tt></td>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1506 </tr>
1507 <tr>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1508 <td><a href="#t_floating">floating point</a></td>
1509 <td><tt>float, double, x86_fp80, fp128, ppc_fp128</tt></td>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1510 </tr>
1511 <tr>
1512 <td><a name="t_firstclass">first class</a></td>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1513 <td><a href="#t_integer">integer</a>,
1514 <a href="#t_floating">floating point</a>,
1515 <a href="#t_pointer">pointer</a>,
Dan Gohman0066db62008-06-18 18:42:13 +0000[diff] [blame]1516 <a href="#t_vector">vector</a>,
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]1517 <a href="#t_struct">structure</a>,
1518 <a href="#t_array">array</a>,
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1519 <a href="#t_label">label</a>,
1520 <a href="#t_metadata">metadata</a>.
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1521 </td>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1522 </tr>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1523 <tr>
1524 <td><a href="#t_primitive">primitive</a></td>
1525 <td><a href="#t_label">label</a>,
1526 <a href="#t_void">void</a>,
Tobias Grosser05387292010-12-28 20:29:31 +0000[diff] [blame]1527 <a href="#t_integer">integer</a>,
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1528 <a href="#t_floating">floating point</a>,
Dale Johannesen21fe99b2010-10-01 00:48:59 +0000[diff] [blame]1529 <a href="#t_x86mmx">x86mmx</a>,
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1530 <a href="#t_metadata">metadata</a>.</td>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1531 </tr>
1532 <tr>
1533 <td><a href="#t_derived">derived</a></td>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]1534 <td><a href="#t_array">array</a>,
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1535 <a href="#t_function">function</a>,
1536 <a href="#t_pointer">pointer</a>,
1537 <a href="#t_struct">structure</a>,
1538 <a href="#t_pstruct">packed structure</a>,
1539 <a href="#t_vector">vector</a>,
1540 <a href="#t_opaque">opaque</a>.
Dan Gohman01ac1012008-10-14 16:32:04 +0000[diff] [blame]1541 </td>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1542 </tr>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1543 </tbody>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1544</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1545
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1546<p>The <a href="#t_firstclass">first class</a> types are perhaps the most
1547 important. Values of these types are the only ones which can be produced by
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]1548 instructions.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1549
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1550</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1551
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1552<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1553<h3>
1554 <a name="t_primitive">Primitive Types</a>
1555</h3>
Chris Lattner8f8c7b72008-01-04 04:34:14 +0000[diff] [blame]1556
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1557<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1558
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1559<p>The primitive types are the fundamental building blocks of the LLVM
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1560 system.</p>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1561
1562<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1563<h4>
1564 <a name="t_integer">Integer Type</a>
1565</h4>
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]1566
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1567<div>
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]1568
1569<h5>Overview:</h5>
1570<p>The integer type is a very simple type that simply specifies an arbitrary
1571 bit width for the integer type desired. Any bit width from 1 bit to
1572 2<sup>23</sup>-1 (about 8 million) can be specified.</p>
1573
1574<h5>Syntax:</h5>
1575<pre>
1576 iN
1577</pre>
1578
1579<p>The number of bits the integer will occupy is specified by the <tt>N</tt>
1580 value.</p>
1581
1582<h5>Examples:</h5>
1583<table class="layout">
1584 <tr class="layout">
1585 <td class="left"><tt>i1</tt></td>
1586 <td class="left">a single-bit integer.</td>
1587 </tr>
1588 <tr class="layout">
1589 <td class="left"><tt>i32</tt></td>
1590 <td class="left">a 32-bit integer.</td>
1591 </tr>
1592 <tr class="layout">
1593 <td class="left"><tt>i1942652</tt></td>
1594 <td class="left">a really big integer of over 1 million bits.</td>
1595 </tr>
1596</table>
1597
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]1598</div>
1599
1600<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1601<h4>
1602 <a name="t_floating">Floating Point Types</a>
1603</h4>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1604
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1605<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1606
1607<table>
1608 <tbody>
1609 <tr><th>Type</th><th>Description</th></tr>
1610 <tr><td><tt>float</tt></td><td>32-bit floating point value</td></tr>
1611 <tr><td><tt>double</tt></td><td>64-bit floating point value</td></tr>
1612 <tr><td><tt>fp128</tt></td><td>128-bit floating point value (112-bit mantissa)</td></tr>
1613 <tr><td><tt>x86_fp80</tt></td><td>80-bit floating point value (X87)</td></tr>
1614 <tr><td><tt>ppc_fp128</tt></td><td>128-bit floating point value (two 64-bits)</td></tr>
1615 </tbody>
1616</table>
1617
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1618</div>
1619
1620<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1621<h4>
1622 <a name="t_x86mmx">X86mmx Type</a>
1623</h4>
Dale Johannesen21fe99b2010-10-01 00:48:59 +0000[diff] [blame]1624
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1625<div>
Dale Johannesen21fe99b2010-10-01 00:48:59 +0000[diff] [blame]1626
1627<h5>Overview:</h5>
1628<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>
1629
1630<h5>Syntax:</h5>
1631<pre>
Dale Johannesen473a8c82010-10-01 01:07:02 +0000[diff] [blame]1632 x86mmx
Dale Johannesen21fe99b2010-10-01 00:48:59 +0000[diff] [blame]1633</pre>
1634
1635</div>
1636
1637<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1638<h4>
1639 <a name="t_void">Void Type</a>
1640</h4>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1641
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1642<div>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1643
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1644<h5>Overview:</h5>
1645<p>The void type does not represent any value and has no size.</p>
1646
1647<h5>Syntax:</h5>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1648<pre>
1649 void
1650</pre>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1651
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1652</div>
1653
1654<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1655<h4>
1656 <a name="t_label">Label Type</a>
1657</h4>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1658
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1659<div>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1660
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1661<h5>Overview:</h5>
1662<p>The label type represents code labels.</p>
1663
1664<h5>Syntax:</h5>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1665<pre>
1666 label
1667</pre>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1668
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1669</div>
1670
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1671<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1672<h4>
1673 <a name="t_metadata">Metadata Type</a>
1674</h4>
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1675
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1676<div>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1677
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1678<h5>Overview:</h5>
Nick Lewyckyc261df92009-09-27 23:27:42 +0000[diff] [blame]1679<p>The metadata type represents embedded metadata. No derived types may be
1680 created from metadata except for <a href="#t_function">function</a>
1681 arguments.
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1682
1683<h5>Syntax:</h5>
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1684<pre>
1685 metadata
1686</pre>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1687
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1688</div>
1689
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1690</div>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1691
1692<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1693<h3>
1694 <a name="t_derived">Derived Types</a>
1695</h3>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1696
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1697<div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1698
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1699<p>The real power in LLVM comes from the derived types in the system. This is
1700 what allows a programmer to represent arrays, functions, pointers, and other
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]1701 useful types. Each of these types contain one or more element types which
1702 may be a primitive type, or another derived type. For example, it is
1703 possible to have a two dimensional array, using an array as the element type
1704 of another array.</p>
Dan Gohmand8791e52009-01-24 15:58:40 +0000[diff] [blame]1705
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]1706
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]1707<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1708<h4>
1709 <a name="t_aggregate">Aggregate Types</a>
1710</h4>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]1711
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1712<div>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]1713
1714<p>Aggregate Types are a subset of derived types that can contain multiple
1715 member types. <a href="#t_array">Arrays</a>,
Chris Lattner61c70e92010-08-28 04:09:24 +0000[diff] [blame]1716 <a href="#t_struct">structs</a>, and <a href="#t_vector">vectors</a> are
1717 aggregate types.</p>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]1718
1719</div>
1720
Reid Spencer2b916312007-05-16 18:44:01 +0000[diff] [blame]1721<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1722<h4>
1723 <a name="t_array">Array Type</a>
1724</h4>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1725
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1726<div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1727
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1728<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1729<p>The array type is a very simple derived type that arranges elements
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1730 sequentially in memory. The array type requires a size (number of elements)
1731 and an underlying data type.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1732
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1733<h5>Syntax:</h5>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1734<pre>
1735 [&lt;# elements&gt; x &lt;elementtype&gt;]
1736</pre>
1737
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1738<p>The number of elements is a constant integer value; <tt>elementtype</tt> may
1739 be any type with a size.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1740
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1741<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1742<table class="layout">
1743 <tr class="layout">
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]1744 <td class="left"><tt>[40 x i32]</tt></td>
1745 <td class="left">Array of 40 32-bit integer values.</td>
1746 </tr>
1747 <tr class="layout">
1748 <td class="left"><tt>[41 x i32]</tt></td>
1749 <td class="left">Array of 41 32-bit integer values.</td>
1750 </tr>
1751 <tr class="layout">
1752 <td class="left"><tt>[4 x i8]</tt></td>
1753 <td class="left">Array of 4 8-bit integer values.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1754 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1755</table>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1756<p>Here are some examples of multidimensional arrays:</p>
1757<table class="layout">
1758 <tr class="layout">
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]1759 <td class="left"><tt>[3 x [4 x i32]]</tt></td>
1760 <td class="left">3x4 array of 32-bit integer values.</td>
1761 </tr>
1762 <tr class="layout">
1763 <td class="left"><tt>[12 x [10 x float]]</tt></td>
1764 <td class="left">12x10 array of single precision floating point values.</td>
1765 </tr>
1766 <tr class="layout">
1767 <td class="left"><tt>[2 x [3 x [4 x i16]]]</tt></td>
1768 <td class="left">2x3x4 array of 16-bit integer values.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1769 </tr>
1770</table>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]1771
Dan Gohman7657f6b2009-11-09 19:01:53 +0000[diff] [blame]1772<p>There is no restriction on indexing beyond the end of the array implied by
1773 a static type (though there are restrictions on indexing beyond the bounds
1774 of an allocated object in some cases). This means that single-dimension
1775 'variable sized array' addressing can be implemented in LLVM with a zero
1776 length array type. An implementation of 'pascal style arrays' in LLVM could
1777 use the type "<tt>{ i32, [0 x float]}</tt>", for example.</p>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]1778
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1779</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1780
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1781<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1782<h4>
1783 <a name="t_function">Function Type</a>
1784</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1785
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1786<div>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]1787
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1788<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1789<p>The function type can be thought of as a function signature. It consists of
1790 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]1791 function type is a first class type or a void type.</p>
Devang Patelc3fc6df2008-03-10 20:49:15 +0000[diff] [blame]1792
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1793<h5>Syntax:</h5>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]1794<pre>
Nick Lewycky51386942009-09-27 07:55:32 +0000[diff] [blame]1795 &lt;returntype&gt; (&lt;parameter list&gt;)
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]1796</pre>
1797
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]1798<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]1799 specifiers. Optionally, the parameter list may include a type <tt>...</tt>,
1800 which indicates that the function takes a variable number of arguments.
1801 Variable argument functions can access their arguments with
1802 the <a href="#int_varargs">variable argument handling intrinsic</a>
Chris Lattner0724fbd2010-03-02 06:36:51 +0000[diff] [blame]1803 functions. '<tt>&lt;returntype&gt;</tt>' is any type except
Nick Lewyckyc261df92009-09-27 23:27:42 +0000[diff] [blame]1804 <a href="#t_label">label</a>.</p>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]1805
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1806<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1807<table class="layout">
1808 <tr class="layout">
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]1809 <td class="left"><tt>i32 (i32)</tt></td>
1810 <td class="left">function taking an <tt>i32</tt>, returning an <tt>i32</tt>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1811 </td>
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]1812 </tr><tr class="layout">
Chris Lattner0724fbd2010-03-02 06:36:51 +0000[diff] [blame]1813 <td class="left"><tt>float&nbsp;(i16,&nbsp;i32&nbsp;*)&nbsp;*
Reid Spencerf17a0b72006-12-31 07:20:23 +0000[diff] [blame]1814 </tt></td>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]1815 <td class="left"><a href="#t_pointer">Pointer</a> to a function that takes
Chris Lattner0724fbd2010-03-02 06:36:51 +0000[diff] [blame]1816 an <tt>i16</tt> and a <a href="#t_pointer">pointer</a> to <tt>i32</tt>,
1817 returning <tt>float</tt>.
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]1818 </td>
1819 </tr><tr class="layout">
1820 <td class="left"><tt>i32 (i8*, ...)</tt></td>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]1821 <td class="left">A vararg function that takes at least one
1822 <a href="#t_pointer">pointer</a> to <tt>i8 </tt> (char in C),
1823 which returns an integer. This is the signature for <tt>printf</tt> in
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]1824 LLVM.
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1825 </td>
Devang Patela582f402008-03-24 05:35:41 +0000[diff] [blame]1826 </tr><tr class="layout">
1827 <td class="left"><tt>{i32, i32} (i32)</tt></td>
Nick Lewycky51386942009-09-27 07:55:32 +0000[diff] [blame]1828 <td class="left">A function taking an <tt>i32</tt>, returning a
1829 <a href="#t_struct">structure</a> containing two <tt>i32</tt> values
Devang Patela582f402008-03-24 05:35:41 +0000[diff] [blame]1830 </td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1831 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1832</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1833
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1834</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1835
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1836<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1837<h4>
1838 <a name="t_struct">Structure Type</a>
1839</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1840
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1841<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1842
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1843<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1844<p>The structure type is used to represent a collection of data members together
1845 in memory. The packing of the field types is defined to match the ABI of the
1846 underlying processor. The elements of a structure may be any type that has a
1847 size.</p>
1848
Jeffrey Yasskin7a088cf2010-01-11 19:19:26 +0000[diff] [blame]1849<p>Structures in memory are accessed using '<tt><a href="#i_load">load</a></tt>'
1850 and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a field
1851 with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.
1852 Structures in registers are accessed using the
1853 '<tt><a href="#i_extractvalue">extractvalue</a></tt>' and
1854 '<tt><a href="#i_insertvalue">insertvalue</a></tt>' instructions.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1855<h5>Syntax:</h5>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1856<pre>
1857 { &lt;type list&gt; }
1858</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1859
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1860<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1861<table class="layout">
1862 <tr class="layout">
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]1863 <td class="left"><tt>{ i32, i32, i32 }</tt></td>
1864 <td class="left">A triple of three <tt>i32</tt> values</td>
1865 </tr><tr class="layout">
1866 <td class="left"><tt>{&nbsp;float,&nbsp;i32&nbsp;(i32)&nbsp;*&nbsp;}</tt></td>
1867 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1868 second element is a <a href="#t_pointer">pointer</a> to a
1869 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1870 an <tt>i32</tt>.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1871 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1872</table>
Dan Gohmand8791e52009-01-24 15:58:40 +0000[diff] [blame]1873
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1874</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1875
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1876<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1877<h4>
1878 <a name="t_pstruct">Packed Structure Type</a>
1879</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1880
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1881<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1882
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]1883<h5>Overview:</h5>
1884<p>The packed structure type is used to represent a collection of data members
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1885 together in memory. There is no padding between fields. Further, the
1886 alignment of a packed structure is 1 byte. The elements of a packed
1887 structure may be any type that has a size.</p>
1888
1889<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt> and
1890 '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a field with
1891 the '<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.</p>
1892
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]1893<h5>Syntax:</h5>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1894<pre>
1895 &lt; { &lt;type list&gt; } &gt;
1896</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1897
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]1898<h5>Examples:</h5>
1899<table class="layout">
1900 <tr class="layout">
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]1901 <td class="left"><tt>&lt; { i32, i32, i32 } &gt;</tt></td>
1902 <td class="left">A triple of three <tt>i32</tt> values</td>
1903 </tr><tr class="layout">
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1904 <td class="left">
1905<tt>&lt;&nbsp;{&nbsp;float,&nbsp;i32&nbsp;(i32)*&nbsp;}&nbsp;&gt;</tt></td>
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]1906 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1907 second element is a <a href="#t_pointer">pointer</a> to a
1908 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1909 an <tt>i32</tt>.</td>
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]1910 </tr>
1911</table>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1912
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]1913</div>
1914
1915<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1916<h4>
1917 <a name="t_pointer">Pointer Type</a>
1918</h4>
Chris Lattner0fd4a272009-02-08 19:53:29 +0000[diff] [blame]1919
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1920<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1921
1922<h5>Overview:</h5>
Dan Gohmanff3ef322010-02-25 16:50:07 +0000[diff] [blame]1923<p>The pointer type is used to specify memory locations.
1924 Pointers are commonly used to reference objects in memory.</p>
1925
1926<p>Pointer types may have an optional address space attribute defining the
1927 numbered address space where the pointed-to object resides. The default
1928 address space is number zero. The semantics of non-zero address
1929 spaces are target-specific.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1930
1931<p>Note that LLVM does not permit pointers to void (<tt>void*</tt>) nor does it
1932 permit pointers to labels (<tt>label*</tt>). Use <tt>i8*</tt> instead.</p>
Chris Lattner0fd4a272009-02-08 19:53:29 +0000[diff] [blame]1933
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1934<h5>Syntax:</h5>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1935<pre>
1936 &lt;type&gt; *
1937</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1938
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1939<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1940<table class="layout">
1941 <tr class="layout">
Dan Gohman2a08c532009-01-04 23:44:43 +0000[diff] [blame]1942 <td class="left"><tt>[4 x i32]*</tt></td>
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]1943 <td class="left">A <a href="#t_pointer">pointer</a> to <a
1944 href="#t_array">array</a> of four <tt>i32</tt> values.</td>
1945 </tr>
1946 <tr class="layout">
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]1947 <td class="left"><tt>i32 (i32*) *</tt></td>
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]1948 <td class="left"> A <a href="#t_pointer">pointer</a> to a <a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1949 href="#t_function">function</a> that takes an <tt>i32*</tt>, returning an
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]1950 <tt>i32</tt>.</td>
1951 </tr>
1952 <tr class="layout">
1953 <td class="left"><tt>i32 addrspace(5)*</tt></td>
1954 <td class="left">A <a href="#t_pointer">pointer</a> to an <tt>i32</tt> value
1955 that resides in address space #5.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1956 </tr>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1957</table>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1958
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1959</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1960
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1961<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1962<h4>
1963 <a name="t_vector">Vector Type</a>
1964</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1965
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1966<div>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1967
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1968<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1969<p>A vector type is a simple derived type that represents a vector of elements.
1970 Vector types are used when multiple primitive data are operated in parallel
1971 using a single instruction (SIMD). A vector type requires a size (number of
Duncan Sandsd40d14e2009-11-27 13:38:03 +0000[diff] [blame]1972 elements) and an underlying primitive data type. Vector types are considered
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1973 <a href="#t_firstclass">first class</a>.</p>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1974
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1975<h5>Syntax:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1976<pre>
1977 &lt; &lt;# elements&gt; x &lt;elementtype&gt; &gt;
1978</pre>
1979
Chris Lattner7d2e7be2010-10-10 18:20:35 +0000[diff] [blame]1980<p>The number of elements is a constant integer value larger than 0; elementtype
1981 may be any integer or floating point type. Vectors of size zero are not
1982 allowed, and pointers are not allowed as the element type.</p>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1983
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1984<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1985<table class="layout">
1986 <tr class="layout">
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]1987 <td class="left"><tt>&lt;4 x i32&gt;</tt></td>
1988 <td class="left">Vector of 4 32-bit integer values.</td>
1989 </tr>
1990 <tr class="layout">
1991 <td class="left"><tt>&lt;8 x float&gt;</tt></td>
1992 <td class="left">Vector of 8 32-bit floating-point values.</td>
1993 </tr>
1994 <tr class="layout">
1995 <td class="left"><tt>&lt;2 x i64&gt;</tt></td>
1996 <td class="left">Vector of 2 64-bit integer values.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1997 </tr>
1998</table>
Dan Gohmand8791e52009-01-24 15:58:40 +0000[diff] [blame]1999
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2000</div>
2001
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]2002<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2003<h4>
2004 <a name="t_opaque">Opaque Type</a>
2005</h4>
2006
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2007<div>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]2008
2009<h5>Overview:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]2010<p>Opaque types are used to represent unknown types in the system. This
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2011 corresponds (for example) to the C notion of a forward declared structure
2012 type. In LLVM, opaque types can eventually be resolved to any type (not just
2013 a structure type).</p>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]2014
2015<h5>Syntax:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]2016<pre>
2017 opaque
2018</pre>
2019
2020<h5>Examples:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]2021<table class="layout">
2022 <tr class="layout">
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]2023 <td class="left"><tt>opaque</tt></td>
2024 <td class="left">An opaque type.</td>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]2025 </tr>
2026</table>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2027
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]2028</div>
2029
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2030</div>
2031
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]2032<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2033<h3>
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]2034 <a name="t_uprefs">Type Up-references</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2035</h3>
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]2036
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2037<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2038
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]2039<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2040<p>An "up reference" allows you to refer to a lexically enclosing type without
2041 requiring it to have a name. For instance, a structure declaration may
2042 contain a pointer to any of the types it is lexically a member of. Example
2043 of up references (with their equivalent as named type declarations)
2044 include:</p>
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]2045
2046<pre>
Chris Lattner3060f5b2009-02-09 10:00:56 +0000[diff] [blame]2047 { \2 * } %x = type { %x* }
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]2048 { \2 }* %y = type { %y }*
2049 \1* %z = type %z*
2050</pre>
2051
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2052<p>An up reference is needed by the asmprinter for printing out cyclic types
2053 when there is no declared name for a type in the cycle. Because the
2054 asmprinter does not want to print out an infinite type string, it needs a
2055 syntax to handle recursive types that have no names (all names are optional
2056 in llvm IR).</p>
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]2057
2058<h5>Syntax:</h5>
2059<pre>
2060 \&lt;level&gt;
2061</pre>
2062
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2063<p>The level is the count of the lexical type that is being referred to.</p>
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]2064
2065<h5>Examples:</h5>
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]2066<table class="layout">
2067 <tr class="layout">
2068 <td class="left"><tt>\1*</tt></td>
2069 <td class="left">Self-referential pointer.</td>
2070 </tr>
2071 <tr class="layout">
2072 <td class="left"><tt>{ { \3*, i8 }, i32 }</tt></td>
2073 <td class="left">Recursive structure where the upref refers to the out-most
2074 structure.</td>
2075 </tr>
2076</table>
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]2077
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2078</div>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]2079
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2080</div>
2081
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2082<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2083<h2><a name="constants">Constants</a></h2>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2084<!-- *********************************************************************** -->
2085
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2086<div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2087
2088<p>LLVM has several different basic types of constants. This section describes
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2089 them all and their syntax.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2090
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2091<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2092<h3>
2093 <a name="simpleconstants">Simple Constants</a>
2094</h3>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2095
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2096<div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2097
2098<dl>
2099 <dt><b>Boolean constants</b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2100 <dd>The two strings '<tt>true</tt>' and '<tt>false</tt>' are both valid
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]2101 constants of the <tt><a href="#t_integer">i1</a></tt> type.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2102
2103 <dt><b>Integer constants</b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2104 <dd>Standard integers (such as '4') are constants of
2105 the <a href="#t_integer">integer</a> type. Negative numbers may be used
2106 with integer types.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2107
2108 <dt><b>Floating point constants</b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2109 <dd>Floating point constants use standard decimal notation (e.g. 123.421),
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2110 exponential notation (e.g. 1.23421e+2), or a more precise hexadecimal
2111 notation (see below). The assembler requires the exact decimal value of a
2112 floating-point constant. For example, the assembler accepts 1.25 but
2113 rejects 1.3 because 1.3 is a repeating decimal in binary. Floating point
2114 constants must have a <a href="#t_floating">floating point</a> type. </dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2115
2116 <dt><b>Null pointer constants</b></dt>
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]2117 <dd>The identifier '<tt>null</tt>' is recognized as a null pointer constant
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2118 and must be of <a href="#t_pointer">pointer type</a>.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2119</dl>
2120
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2121<p>The one non-intuitive notation for constants is the hexadecimal form of
2122 floating point constants. For example, the form '<tt>double
2123 0x432ff973cafa8000</tt>' is equivalent to (but harder to read than)
2124 '<tt>double 4.5e+15</tt>'. The only time hexadecimal floating point
2125 constants are required (and the only time that they are generated by the
2126 disassembler) is when a floating point constant must be emitted but it cannot
2127 be represented as a decimal floating point number in a reasonable number of
2128 digits. For example, NaN's, infinities, and other special values are
2129 represented in their IEEE hexadecimal format so that assembly and disassembly
2130 do not cause any bits to change in the constants.</p>
2131
Dale Johannesenbd5e5a82009-02-11 22:14:51 +0000[diff] [blame]2132<p>When using the hexadecimal form, constants of types float and double are
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2133 represented using the 16-digit form shown above (which matches the IEEE754
2134 representation for double); float values must, however, be exactly
2135 representable as IEE754 single precision. Hexadecimal format is always used
2136 for long double, and there are three forms of long double. The 80-bit format
2137 used by x86 is represented as <tt>0xK</tt> followed by 20 hexadecimal digits.
2138 The 128-bit format used by PowerPC (two adjacent doubles) is represented
2139 by <tt>0xM</tt> followed by 32 hexadecimal digits. The IEEE 128-bit format
2140 is represented by <tt>0xL</tt> followed by 32 hexadecimal digits; no
2141 currently supported target uses this format. Long doubles will only work if
2142 they match the long double format on your target. All hexadecimal formats
2143 are big-endian (sign bit at the left).</p>
2144
Dale Johannesen21fe99b2010-10-01 00:48:59 +0000[diff] [blame]2145<p>There are no constants of type x86mmx.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2146</div>
2147
2148<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2149<h3>
Bill Wendlingd9fe2982009-07-20 02:32:41 +0000[diff] [blame]2150<a name="aggregateconstants"></a> <!-- old anchor -->
2151<a name="complexconstants">Complex Constants</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2152</h3>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2153
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2154<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2155
Chris Lattner70882792009-02-28 18:32:25 +0000[diff] [blame]2156<p>Complex constants are a (potentially recursive) combination of simple
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2157 constants and smaller complex constants.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2158
2159<dl>
2160 <dt><b>Structure constants</b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2161 <dd>Structure constants are represented with notation similar to structure
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2162 type definitions (a comma separated list of elements, surrounded by braces
2163 (<tt>{}</tt>)). For example: "<tt>{ i32 4, float 17.0, i32* @G }</tt>",
2164 where "<tt>@G</tt>" is declared as "<tt>@G = external global i32</tt>".
2165 Structure constants must have <a href="#t_struct">structure type</a>, and
2166 the number and types of elements must match those specified by the
2167 type.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2168
2169 <dt><b>Array constants</b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2170 <dd>Array constants are represented with notation similar to array type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2171 definitions (a comma separated list of elements, surrounded by square
2172 brackets (<tt>[]</tt>)). For example: "<tt>[ i32 42, i32 11, i32 74
2173 ]</tt>". Array constants must have <a href="#t_array">array type</a>, and
2174 the number and types of elements must match those specified by the
2175 type.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2176
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]2177 <dt><b>Vector constants</b></dt>
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]2178 <dd>Vector constants are represented with notation similar to vector type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2179 definitions (a comma separated list of elements, surrounded by
2180 less-than/greater-than's (<tt>&lt;&gt;</tt>)). For example: "<tt>&lt; i32
2181 42, i32 11, i32 74, i32 100 &gt;</tt>". Vector constants must
2182 have <a href="#t_vector">vector type</a>, and the number and types of
2183 elements must match those specified by the type.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2184
2185 <dt><b>Zero initialization</b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2186 <dd>The string '<tt>zeroinitializer</tt>' can be used to zero initialize a
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]2187 value to zero of <em>any</em> type, including scalar and
2188 <a href="#t_aggregate">aggregate</a> types.
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2189 This is often used to avoid having to print large zero initializers
2190 (e.g. for large arrays) and is always exactly equivalent to using explicit
2191 zero initializers.</dd>
Nick Lewycky21cc4462009-04-04 07:22:01 +0000[diff] [blame]2192
2193 <dt><b>Metadata node</b></dt>
Nick Lewycky1e8c7a62009-05-30 16:08:30 +0000[diff] [blame]2194 <dd>A metadata node is a structure-like constant with
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2195 <a href="#t_metadata">metadata type</a>. For example: "<tt>metadata !{
2196 i32 0, metadata !"test" }</tt>". Unlike other constants that are meant to
2197 be interpreted as part of the instruction stream, metadata is a place to
2198 attach additional information such as debug info.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2199</dl>
2200
2201</div>
2202
2203<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2204<h3>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2205 <a name="globalconstants">Global Variable and Function Addresses</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2206</h3>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2207
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2208<div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2209
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2210<p>The addresses of <a href="#globalvars">global variables</a>
2211 and <a href="#functionstructure">functions</a> are always implicitly valid
2212 (link-time) constants. These constants are explicitly referenced when
2213 the <a href="#identifiers">identifier for the global</a> is used and always
2214 have <a href="#t_pointer">pointer</a> type. For example, the following is a
2215 legal LLVM file:</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2216
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2217<pre class="doc_code">
Chris Lattnera18a4242007-06-06 18:28:13 +0000[diff] [blame]2218@X = global i32 17
2219@Y = global i32 42
2220@Z = global [2 x i32*] [ i32* @X, i32* @Y ]
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2221</pre>
2222
2223</div>
2224
2225<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2226<h3>
2227 <a name="undefvalues">Undefined Values</a>
2228</h3>
2229
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2230<div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2231
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2232<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]2233 indicates that the user of the value may receive an unspecified bit-pattern.
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]2234 Undefined values may be of any type (other than '<tt>label</tt>'
2235 or '<tt>void</tt>') and be used anywhere a constant is permitted.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2236
Chris Lattnerc608cb12009-09-11 01:49:31 +0000[diff] [blame]2237<p>Undefined values are useful because they indicate to the compiler that the
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2238 program is well defined no matter what value is used. This gives the
2239 compiler more freedom to optimize. Here are some examples of (potentially
2240 surprising) transformations that are valid (in pseudo IR):</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2241
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2242
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2243<pre class="doc_code">
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2244 %A = add %X, undef
2245 %B = sub %X, undef
2246 %C = xor %X, undef
2247Safe:
2248 %A = undef
2249 %B = undef
2250 %C = undef
2251</pre>
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2252
2253<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]2254 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]2255
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2256<pre class="doc_code">
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2257 %A = or %X, undef
2258 %B = and %X, undef
2259Safe:
2260 %A = -1
2261 %B = 0
2262Unsafe:
2263 %A = undef
2264 %B = undef
2265</pre>
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2266
2267<p>These logical operations have bits that are not always affected by the input.
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]2268 For example, if <tt>%X</tt> has a zero bit, then the output of the
2269 '<tt>and</tt>' operation will always be a zero for that bit, no matter what
2270 the corresponding bit from the '<tt>undef</tt>' is. As such, it is unsafe to
2271 optimize or assume that the result of the '<tt>and</tt>' is '<tt>undef</tt>'.
2272 However, it is safe to assume that all bits of the '<tt>undef</tt>' could be
2273 0, and optimize the '<tt>and</tt>' to 0. Likewise, it is safe to assume that
2274 all the bits of the '<tt>undef</tt>' operand to the '<tt>or</tt>' could be
2275 set, allowing the '<tt>or</tt>' to be folded to -1.</p>
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2276
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2277<pre class="doc_code">
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2278 %A = select undef, %X, %Y
2279 %B = select undef, 42, %Y
2280 %C = select %X, %Y, undef
2281Safe:
2282 %A = %X (or %Y)
2283 %B = 42 (or %Y)
2284 %C = %Y
2285Unsafe:
2286 %A = undef
2287 %B = undef
2288 %C = undef
2289</pre>
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2290
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]2291<p>This set of examples shows that undefined '<tt>select</tt>' (and conditional
2292 branch) conditions can go <em>either way</em>, but they have to come from one
2293 of the two operands. In the <tt>%A</tt> example, if <tt>%X</tt> and
2294 <tt>%Y</tt> were both known to have a clear low bit, then <tt>%A</tt> would
2295 have to have a cleared low bit. However, in the <tt>%C</tt> example, the
2296 optimizer is allowed to assume that the '<tt>undef</tt>' operand could be the
2297 same as <tt>%Y</tt>, allowing the whole '<tt>select</tt>' to be
2298 eliminated.</p>
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2299
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2300<pre class="doc_code">
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2301 %A = xor undef, undef
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]2302
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2303 %B = undef
2304 %C = xor %B, %B
2305
2306 %D = undef
2307 %E = icmp lt %D, 4
2308 %F = icmp gte %D, 4
2309
2310Safe:
2311 %A = undef
2312 %B = undef
2313 %C = undef
2314 %D = undef
2315 %E = undef
2316 %F = undef
2317</pre>
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2318
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]2319<p>This example points out that two '<tt>undef</tt>' operands are not
2320 necessarily the same. This can be surprising to people (and also matches C
2321 semantics) where they assume that "<tt>X^X</tt>" is always zero, even
2322 if <tt>X</tt> is undefined. This isn't true for a number of reasons, but the
2323 short answer is that an '<tt>undef</tt>' "variable" can arbitrarily change
2324 its value over its "live range". This is true because the variable doesn't
2325 actually <em>have a live range</em>. Instead, the value is logically read
2326 from arbitrary registers that happen to be around when needed, so the value
2327 is not necessarily consistent over time. In fact, <tt>%A</tt> and <tt>%C</tt>
2328 need to have the same semantics or the core LLVM "replace all uses with"
2329 concept would not hold.</p>
Chris Lattner6e9057b2009-09-07 23:33:52 +0000[diff] [blame]2330
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2331<pre class="doc_code">
Chris Lattner6e9057b2009-09-07 23:33:52 +0000[diff] [blame]2332 %A = fdiv undef, %X
2333 %B = fdiv %X, undef
2334Safe:
2335 %A = undef
2336b: unreachable
2337</pre>
Chris Lattner6e9057b2009-09-07 23:33:52 +0000[diff] [blame]2338
2339<p>These examples show the crucial difference between an <em>undefined
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]2340 value</em> and <em>undefined behavior</em>. An undefined value (like
2341 '<tt>undef</tt>') is allowed to have an arbitrary bit-pattern. This means that
2342 the <tt>%A</tt> operation can be constant folded to '<tt>undef</tt>', because
2343 the '<tt>undef</tt>' could be an SNaN, and <tt>fdiv</tt> is not (currently)
2344 defined on SNaN's. However, in the second example, we can make a more
2345 aggressive assumption: because the <tt>undef</tt> is allowed to be an
2346 arbitrary value, we are allowed to assume that it could be zero. Since a
2347 divide by zero has <em>undefined behavior</em>, we are allowed to assume that
2348 the operation does not execute at all. This allows us to delete the divide and
2349 all code after it. Because the undefined operation "can't happen", the
2350 optimizer can assume that it occurs in dead code.</p>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]2351
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2352<pre class="doc_code">
Chris Lattner6e9057b2009-09-07 23:33:52 +0000[diff] [blame]2353a: store undef -> %X
2354b: store %X -> undef
2355Safe:
2356a: &lt;deleted&gt;
2357b: unreachable
2358</pre>
Chris Lattner6e9057b2009-09-07 23:33:52 +0000[diff] [blame]2359
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]2360<p>These examples reiterate the <tt>fdiv</tt> example: a store <em>of</em> an
2361 undefined value can be assumed to not have any effect; we can assume that the
2362 value is overwritten with bits that happen to match what was already there.
2363 However, a store <em>to</em> an undefined location could clobber arbitrary
2364 memory, therefore, it has undefined behavior.</p>
Chris Lattner6e9057b2009-09-07 23:33:52 +0000[diff] [blame]2365
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2366</div>
2367
2368<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2369<h3>
2370 <a name="trapvalues">Trap Values</a>
2371</h3>
2372
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2373<div>
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]2374
Dan Gohmanc68ce062010-04-26 20:21:21 +0000[diff] [blame]2375<p>Trap values are similar to <a href="#undefvalues">undef values</a>, however
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]2376 instead of representing an unspecified bit pattern, they represent the
2377 fact that an instruction or constant expression which cannot evoke side
2378 effects has nevertheless detected a condition which results in undefined
Dan Gohmanc68ce062010-04-26 20:21:21 +0000[diff] [blame]2379 behavior.</p>
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]2380
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2381<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]2382 only exist when produced by operations such as
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2383 <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]2384
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2385<p>Trap value behavior is defined in terms of value <i>dependence</i>:</p>
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2386
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2387<ul>
2388<li>Values other than <a href="#i_phi"><tt>phi</tt></a> nodes depend on
2389 their operands.</li>
2390
2391<li><a href="#i_phi"><tt>Phi</tt></a> nodes depend on the operand corresponding
2392 to their dynamic predecessor basic block.</li>
2393
2394<li>Function arguments depend on the corresponding actual argument values in
2395 the dynamic callers of their functions.</li>
2396
2397<li><a href="#i_call"><tt>Call</tt></a> instructions depend on the
2398 <a href="#i_ret"><tt>ret</tt></a> instructions that dynamically transfer
2399 control back to them.</li>
2400
Dan Gohmanb5328162010-05-03 14:55:22 +0000[diff] [blame]2401<li><a href="#i_invoke"><tt>Invoke</tt></a> instructions depend on the
2402 <a href="#i_ret"><tt>ret</tt></a>, <a href="#i_unwind"><tt>unwind</tt></a>,
2403 or exception-throwing call instructions that dynamically transfer control
2404 back to them.</li>
2405
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2406<li>Non-volatile loads and stores depend on the most recent stores to all of the
2407 referenced memory addresses, following the order in the IR
2408 (including loads and stores implied by intrinsics such as
2409 <a href="#int_memcpy"><tt>@llvm.memcpy</tt></a>.)</li>
2410
Dan Gohman7c24ff12010-05-03 14:59:34 +0000[diff] [blame]2411<!-- TODO: In the case of multiple threads, this only applies if the store
2412 "happens-before" the load or store. -->
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2413
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2414<!-- TODO: floating-point exception state -->
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2415
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2416<li>An instruction with externally visible side effects depends on the most
2417 recent preceding instruction with externally visible side effects, following
Dan Gohmanff70fe42010-07-06 15:26:33 +0000[diff] [blame]2418 the order in the IR. (This includes
2419 <a href="#volatile">volatile operations</a>.)</li>
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]2420
Dan Gohmanb5328162010-05-03 14:55:22 +0000[diff] [blame]2421<li>An instruction <i>control-depends</i> on a
2422 <a href="#terminators">terminator instruction</a>
2423 if the terminator instruction has multiple successors and the instruction
2424 is always executed when control transfers to one of the successors, and
Chris Lattner7a2bdde2011-04-15 05:18:47 +0000[diff] [blame]2425 may not be executed when control is transferred to another.</li>
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2426
Dan Gohmanca4cac42011-04-12 23:05:59 +0000[diff] [blame]2427<li>Additionally, an instruction also <i>control-depends</i> on a terminator
2428 instruction if the set of instructions it otherwise depends on would be
Chris Lattner7a2bdde2011-04-15 05:18:47 +0000[diff] [blame]2429 different if the terminator had transferred control to a different
Dan Gohmanca4cac42011-04-12 23:05:59 +0000[diff] [blame]2430 successor.</li>
2431
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2432<li>Dependence is transitive.</li>
2433
2434</ul>
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2435
2436<p>Whenever a trap value is generated, all values which depend on it evaluate
2437 to trap. If they have side effects, the evoke their side effects as if each
2438 operand with a trap value were undef. If they have externally-visible side
2439 effects, the behavior is undefined.</p>
2440
2441<p>Here are some examples:</p>
Dan Gohmanc30f6e12010-04-26 20:54:53 +0000[diff] [blame]2442
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2443<pre class="doc_code">
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2444entry:
2445 %trap = sub nuw i32 0, 1 ; Results in a trap value.
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2446 %still_trap = and i32 %trap, 0 ; Whereas (and i32 undef, 0) would return 0.
2447 %trap_yet_again = getelementptr i32* @h, i32 %still_trap
2448 store i32 0, i32* %trap_yet_again ; undefined behavior
2449
2450 store i32 %trap, i32* @g ; Trap value conceptually stored to memory.
2451 %trap2 = load i32* @g ; Returns a trap value, not just undef.
2452
2453 volatile store i32 %trap, i32* @g ; External observation; undefined behavior.
2454
2455 %narrowaddr = bitcast i32* @g to i16*
2456 %wideaddr = bitcast i32* @g to i64*
Nick Lewycky64f9fb12011-05-16 19:29:30 +0000[diff] [blame]2457 %trap3 = load i16* %narrowaddr ; Returns a trap value.
2458 %trap4 = load i64* %wideaddr ; Returns a trap value.
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2459
Nick Lewycky64f9fb12011-05-16 19:29:30 +0000[diff] [blame]2460 %cmp = icmp slt i32 %trap, 0 ; Returns a trap value.
2461 br i1 %cmp, label %true, label %end ; Branch to either destination.
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2462
2463true:
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2464 volatile store i32 0, i32* @g ; This is control-dependent on %cmp, so
2465 ; it has undefined behavior.
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2466 br label %end
2467
2468end:
2469 %p = phi i32 [ 0, %entry ], [ 1, %true ]
2470 ; Both edges into this PHI are
2471 ; control-dependent on %cmp, so this
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2472 ; always results in a trap value.
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2473
Dan Gohmanca4cac42011-04-12 23:05:59 +0000[diff] [blame]2474 volatile store i32 0, i32* @g ; This would depend on the store in %true
2475 ; if %cmp is true, or the store in %entry
2476 ; otherwise, so this is undefined behavior.
2477
Nick Lewycky64f9fb12011-05-16 19:29:30 +0000[diff] [blame]2478 br i1 %cmp, label %second_true, label %second_end
Dan Gohmanca4cac42011-04-12 23:05:59 +0000[diff] [blame]2479 ; The same branch again, but this time the
2480 ; true block doesn't have side effects.
2481
2482second_true:
2483 ; No side effects!
Nick Lewycky64f9fb12011-05-16 19:29:30 +0000[diff] [blame]2484 ret void
Dan Gohmanca4cac42011-04-12 23:05:59 +0000[diff] [blame]2485
2486second_end:
2487 volatile store i32 0, i32* @g ; This time, the instruction always depends
2488 ; on the store in %end. Also, it is
2489 ; control-equivalent to %end, so this is
Nick Lewycky64f9fb12011-05-16 19:29:30 +0000[diff] [blame]2490 ; well-defined (again, ignoring earlier
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2491 ; undefined behavior in this example).
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2492</pre>
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]2493
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]2494</div>
2495
2496<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2497<h3>
2498 <a name="blockaddress">Addresses of Basic Blocks</a>
2499</h3>
2500
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2501<div>
Chris Lattnerc6f44362009-10-27 21:01:34 +0000[diff] [blame]2502
Chris Lattnercdfc9402009-11-01 01:27:45 +0000[diff] [blame]2503<p><b><tt>blockaddress(@function, %block)</tt></b></p>
Chris Lattnerc6f44362009-10-27 21:01:34 +0000[diff] [blame]2504
2505<p>The '<tt>blockaddress</tt>' constant computes the address of the specified
Chris Lattner2dfdf2a2009-10-27 21:49:40 +0000[diff] [blame]2506 basic block in the specified function, and always has an i8* type. Taking
Chris Lattnercdfc9402009-11-01 01:27:45 +0000[diff] [blame]2507 the address of the entry block is illegal.</p>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]2508
Chris Lattnerc6f44362009-10-27 21:01:34 +0000[diff] [blame]2509<p>This value only has defined behavior when used as an operand to the
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]2510 '<a href="#i_indirectbr"><tt>indirectbr</tt></a>' instruction, or for
2511 comparisons against null. Pointer equality tests between labels addresses
2512 results in undefined behavior &mdash; though, again, comparison against null
2513 is ok, and no label is equal to the null pointer. This may be passed around
2514 as an opaque pointer sized value as long as the bits are not inspected. This
2515 allows <tt>ptrtoint</tt> and arithmetic to be performed on these values so
2516 long as the original value is reconstituted before the <tt>indirectbr</tt>
2517 instruction.</p>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]2518
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]2519<p>Finally, some targets may provide defined semantics when using the value as
2520 the operand to an inline assembly, but that is target specific.</p>
Chris Lattnerc6f44362009-10-27 21:01:34 +0000[diff] [blame]2521
2522</div>
2523
2524
2525<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2526<h3>
2527 <a name="constantexprs">Constant Expressions</a>
2528</h3>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2529
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2530<div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2531
2532<p>Constant expressions are used to allow expressions involving other constants
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2533 to be used as constants. Constant expressions may be of
2534 any <a href="#t_firstclass">first class</a> type and may involve any LLVM
2535 operation that does not have side effects (e.g. load and call are not
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]2536 supported). The following is the syntax for constant expressions:</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2537
2538<dl>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2539 <dt><b><tt>trunc (CST to TYPE)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2540 <dd>Truncate a constant to another type. The bit size of CST must be larger
2541 than the bit size of TYPE. Both types must be integers.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2542
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2543 <dt><b><tt>zext (CST to TYPE)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2544 <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]2545 smaller than the bit size of TYPE. Both types must be integers.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2546
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2547 <dt><b><tt>sext (CST to TYPE)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2548 <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]2549 smaller than the bit size of TYPE. Both types must be integers.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2550
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2551 <dt><b><tt>fptrunc (CST to TYPE)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2552 <dd>Truncate a floating point constant to another floating point type. The
2553 size of CST must be larger than the size of TYPE. Both types must be
2554 floating point.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2555
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2556 <dt><b><tt>fpext (CST to TYPE)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2557 <dd>Floating point extend a constant to another type. The size of CST must be
2558 smaller or equal to the size of TYPE. Both types must be floating
2559 point.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2560
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2561 <dt><b><tt>fptoui (CST to TYPE)</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2562 <dd>Convert a floating point constant to the corresponding unsigned integer
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2563 constant. TYPE must be a scalar or vector integer type. CST must be of
2564 scalar or vector floating point type. Both CST and TYPE must be scalars,
2565 or vectors of the same number of elements. If the value won't fit in the
2566 integer type, the results are undefined.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2567
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2568 <dt><b><tt>fptosi (CST to TYPE)</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2569 <dd>Convert a floating point constant to the corresponding signed integer
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2570 constant. TYPE must be a scalar or vector integer type. CST must be of
2571 scalar or vector floating point type. Both CST and TYPE must be scalars,
2572 or vectors of the same number of elements. If the value won't fit in the
2573 integer type, the results are undefined.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2574
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2575 <dt><b><tt>uitofp (CST to TYPE)</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2576 <dd>Convert an unsigned integer constant to the corresponding floating point
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2577 constant. TYPE must be a scalar or vector floating point type. CST must be
2578 of scalar or vector integer type. Both CST and TYPE must be scalars, or
2579 vectors of the same number of elements. If the value won't fit in the
2580 floating point type, the results are undefined.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2581
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2582 <dt><b><tt>sitofp (CST to TYPE)</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2583 <dd>Convert a signed integer constant to the corresponding floating point
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2584 constant. TYPE must be a scalar or vector floating point type. CST must be
2585 of scalar or vector integer type. Both CST and TYPE must be scalars, or
2586 vectors of the same number of elements. If the value won't fit in the
2587 floating point type, the results are undefined.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2588
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2589 <dt><b><tt>ptrtoint (CST to TYPE)</tt></b></dt>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]2590 <dd>Convert a pointer typed constant to the corresponding integer constant
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2591 <tt>TYPE</tt> must be an integer type. <tt>CST</tt> must be of pointer
2592 type. The <tt>CST</tt> value is zero extended, truncated, or unchanged to
2593 make it fit in <tt>TYPE</tt>.</dd>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]2594
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2595 <dt><b><tt>inttoptr (CST to TYPE)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2596 <dd>Convert a integer constant to a pointer constant. TYPE must be a pointer
2597 type. CST must be of integer type. The CST value is zero extended,
2598 truncated, or unchanged to make it fit in a pointer size. This one is
2599 <i>really</i> dangerous!</dd>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]2600
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2601 <dt><b><tt>bitcast (CST to TYPE)</tt></b></dt>
Chris Lattner03bbad62009-02-28 18:27:03 +0000[diff] [blame]2602 <dd>Convert a constant, CST, to another TYPE. The constraints of the operands
2603 are the same as those for the <a href="#i_bitcast">bitcast
2604 instruction</a>.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2605
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2606 <dt><b><tt>getelementptr (CSTPTR, IDX0, IDX1, ...)</tt></b></dt>
2607 <dt><b><tt>getelementptr inbounds (CSTPTR, IDX0, IDX1, ...)</tt></b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2608 <dd>Perform the <a href="#i_getelementptr">getelementptr operation</a> on
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2609 constants. As with the <a href="#i_getelementptr">getelementptr</a>
2610 instruction, the index list may have zero or more indexes, which are
2611 required to make sense for the type of "CSTPTR".</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2612
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2613 <dt><b><tt>select (COND, VAL1, VAL2)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2614 <dd>Perform the <a href="#i_select">select operation</a> on constants.</dd>
Reid Spencer01c42592006-12-04 19:23:19 +0000[diff] [blame]2615
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2616 <dt><b><tt>icmp COND (VAL1, VAL2)</tt></b></dt>
Reid Spencer01c42592006-12-04 19:23:19 +0000[diff] [blame]2617 <dd>Performs the <a href="#i_icmp">icmp operation</a> on constants.</dd>
2618
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2619 <dt><b><tt>fcmp COND (VAL1, VAL2)</tt></b></dt>
Reid Spencer01c42592006-12-04 19:23:19 +0000[diff] [blame]2620 <dd>Performs the <a href="#i_fcmp">fcmp operation</a> on constants.</dd>
Robert Bocchino9fbe1452006-01-10 19:31:34 +0000[diff] [blame]2621
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2622 <dt><b><tt>extractelement (VAL, IDX)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2623 <dd>Perform the <a href="#i_extractelement">extractelement operation</a> on
2624 constants.</dd>
Robert Bocchino9fbe1452006-01-10 19:31:34 +0000[diff] [blame]2625
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2626 <dt><b><tt>insertelement (VAL, ELT, IDX)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2627 <dd>Perform the <a href="#i_insertelement">insertelement operation</a> on
2628 constants.</dd>
Chris Lattnerc1989542006-04-08 00:13:41 +0000[diff] [blame]2629
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2630 <dt><b><tt>shufflevector (VEC1, VEC2, IDXMASK)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2631 <dd>Perform the <a href="#i_shufflevector">shufflevector operation</a> on
2632 constants.</dd>
Chris Lattnerc1989542006-04-08 00:13:41 +0000[diff] [blame]2633
Nick Lewycky9e130ce2010-05-29 06:44:15 +0000[diff] [blame]2634 <dt><b><tt>extractvalue (VAL, IDX0, IDX1, ...)</tt></b></dt>
2635 <dd>Perform the <a href="#i_extractvalue">extractvalue operation</a> on
2636 constants. The index list is interpreted in a similar manner as indices in
2637 a '<a href="#i_getelementptr">getelementptr</a>' operation. At least one
2638 index value must be specified.</dd>
2639
2640 <dt><b><tt>insertvalue (VAL, ELT, IDX0, IDX1, ...)</tt></b></dt>
2641 <dd>Perform the <a href="#i_insertvalue">insertvalue operation</a> on
2642 constants. The index list is interpreted in a similar manner as indices in
2643 a '<a href="#i_getelementptr">getelementptr</a>' operation. At least one
2644 index value must be specified.</dd>
2645
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2646 <dt><b><tt>OPCODE (LHS, RHS)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2647 <dd>Perform the specified operation of the LHS and RHS constants. OPCODE may
2648 be any of the <a href="#binaryops">binary</a>
2649 or <a href="#bitwiseops">bitwise binary</a> operations. The constraints
2650 on operands are the same as those for the corresponding instruction
2651 (e.g. no bitwise operations on floating point values are allowed).</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2652</dl>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2653
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2654</div>
Chris Lattner9ee5d222004-03-08 16:49:10 +0000[diff] [blame]2655
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2656</div>
2657
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2658<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2659<h2><a name="othervalues">Other Values</a></h2>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2660<!-- *********************************************************************** -->
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2661<div>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2662<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2663<h3>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2664<a name="inlineasm">Inline Assembler Expressions</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2665</h3>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2666
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2667<div>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2668
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2669<p>LLVM supports inline assembler expressions (as opposed
2670 to <a href="#moduleasm"> Module-Level Inline Assembly</a>) through the use of
2671 a special value. This value represents the inline assembler as a string
2672 (containing the instructions to emit), a list of operand constraints (stored
Dale Johannesen09fed252009-10-13 21:56:55 +0000[diff] [blame]2673 as a string), a flag that indicates whether or not the inline asm
Dale Johannesen8ba2d5b2009-10-21 23:28:00 +0000[diff] [blame]2674 expression has side effects, and a flag indicating whether the function
2675 containing the asm needs to align its stack conservatively. An example
2676 inline assembler expression is:</p>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2677
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2678<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2679i32 (i32) asm "bswap $0", "=r,r"
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2680</pre>
2681
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2682<p>Inline assembler expressions may <b>only</b> be used as the callee operand of
2683 a <a href="#i_call"><tt>call</tt> instruction</a>. Thus, typically we
2684 have:</p>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2685
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2686<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2687%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]2688</pre>
2689
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2690<p>Inline asms with side effects not visible in the constraint list must be
2691 marked as having side effects. This is done through the use of the
2692 '<tt>sideeffect</tt>' keyword, like so:</p>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2693
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2694<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2695call void asm sideeffect "eieio", ""()
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2696</pre>
2697
Dale Johannesen8ba2d5b2009-10-21 23:28:00 +0000[diff] [blame]2698<p>In some cases inline asms will contain code that will not work unless the
2699 stack is aligned in some way, such as calls or SSE instructions on x86,
2700 yet will not contain code that does that alignment within the asm.
2701 The compiler should make conservative assumptions about what the asm might
2702 contain and should generate its usual stack alignment code in the prologue
2703 if the '<tt>alignstack</tt>' keyword is present:</p>
Dale Johannesen09fed252009-10-13 21:56:55 +0000[diff] [blame]2704
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2705<pre class="doc_code">
Dale Johannesen8ba2d5b2009-10-21 23:28:00 +0000[diff] [blame]2706call void asm alignstack "eieio", ""()
Dale Johannesen09fed252009-10-13 21:56:55 +0000[diff] [blame]2707</pre>
Dale Johannesen09fed252009-10-13 21:56:55 +0000[diff] [blame]2708
2709<p>If both keywords appear the '<tt>sideeffect</tt>' keyword must come
2710 first.</p>
2711
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2712<p>TODO: The format of the asm and constraints string still need to be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2713 documented here. Constraints on what can be done (e.g. duplication, moving,
2714 etc need to be documented). This is probably best done by reference to
2715 another document that covers inline asm from a holistic perspective.</p>
Chris Lattnercf9a4152010-04-07 05:38:05 +0000[diff] [blame]2716
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2717<h4>
Chris Lattnercf9a4152010-04-07 05:38:05 +0000[diff] [blame]2718<a name="inlineasm_md">Inline Asm Metadata</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2719</h4>
Chris Lattnercf9a4152010-04-07 05:38:05 +0000[diff] [blame]2720
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2721<div>
Chris Lattnercf9a4152010-04-07 05:38:05 +0000[diff] [blame]2722
2723<p>The call instructions that wrap inline asm nodes may have a "!srcloc" MDNode
Chris Lattnerce1b9ad2010-11-17 08:20:42 +0000[diff] [blame]2724 attached to it that contains a list of constant integers. If present, the
2725 code generator will use the integer as the location cookie value when report
Chris Lattnercf9a4152010-04-07 05:38:05 +0000[diff] [blame]2726 errors through the LLVMContext error reporting mechanisms. This allows a
Dan Gohman1c70c002010-04-28 00:36:01 +0000[diff] [blame]2727 front-end to correlate backend errors that occur with inline asm back to the
Chris Lattnercf9a4152010-04-07 05:38:05 +0000[diff] [blame]2728 source code that produced it. For example:</p>
2729
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2730<pre class="doc_code">
Chris Lattnercf9a4152010-04-07 05:38:05 +0000[diff] [blame]2731call void asm sideeffect "something bad", ""()<b>, !srcloc !42</b>
2732...
2733!42 = !{ i32 1234567 }
2734</pre>
Chris Lattnercf9a4152010-04-07 05:38:05 +0000[diff] [blame]2735
2736<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]2737 IR. If the MDNode contains multiple constants, the code generator will use
2738 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]2739
2740</div>
2741
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2742</div>
2743
Chris Lattnere6a5ddd2010-01-15 21:50:19 +0000[diff] [blame]2744<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2745<h3>
2746 <a name="metadata">Metadata Nodes and Metadata Strings</a>
2747</h3>
Chris Lattnere6a5ddd2010-01-15 21:50:19 +0000[diff] [blame]2748
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2749<div>
Chris Lattnere6a5ddd2010-01-15 21:50:19 +0000[diff] [blame]2750
2751<p>LLVM IR allows metadata to be attached to instructions in the program that
2752 can convey extra information about the code to the optimizers and code
2753 generator. One example application of metadata is source-level debug
2754 information. There are two metadata primitives: strings and nodes. All
2755 metadata has the <tt>metadata</tt> type and is identified in syntax by a
2756 preceding exclamation point ('<tt>!</tt>').</p>
2757
2758<p>A metadata string is a string surrounded by double quotes. It can contain
2759 any character by escaping non-printable characters with "\xx" where "xx" is
2760 the two digit hex code. For example: "<tt>!"test\00"</tt>".</p>
2761
2762<p>Metadata nodes are represented with notation similar to structure constants
2763 (a comma separated list of elements, surrounded by braces and preceded by an
2764 exclamation point). For example: "<tt>!{ metadata !"test\00", i32
2765 10}</tt>". Metadata nodes can have any values as their operand.</p>
2766
2767<p>A <a href="#namedmetadatastructure">named metadata</a> is a collection of
2768 metadata nodes, which can be looked up in the module symbol table. For
2769 example: "<tt>!foo = metadata !{!4, !3}</tt>".
2770
Devang Patele1d50cd2010-03-04 23:44:48 +0000[diff] [blame]2771<p>Metadata can be used as function arguments. Here <tt>llvm.dbg.value</tt>
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2772 function is using two metadata arguments.</p>
Devang Patele1d50cd2010-03-04 23:44:48 +0000[diff] [blame]2773
Bill Wendling9ff5de92011-03-02 02:17:11 +0000[diff] [blame]2774<div class="doc_code">
2775<pre>
2776call void @llvm.dbg.value(metadata !24, i64 0, metadata !25)
2777</pre>
2778</div>
Devang Patele1d50cd2010-03-04 23:44:48 +0000[diff] [blame]2779
2780<p>Metadata can be attached with an instruction. Here metadata <tt>!21</tt> is
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2781 attached with <tt>add</tt> instruction using <tt>!dbg</tt> identifier.</p>
Devang Patele1d50cd2010-03-04 23:44:48 +0000[diff] [blame]2782
Bill Wendling9ff5de92011-03-02 02:17:11 +0000[diff] [blame]2783<div class="doc_code">
2784<pre>
2785%indvar.next = add i64 %indvar, 1, !dbg !21
2786</pre>
2787</div>
2788
Chris Lattnere6a5ddd2010-01-15 21:50:19 +0000[diff] [blame]2789</div>
2790
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2791</div>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2792
2793<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2794<h2>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2795 <a name="intrinsic_globals">Intrinsic Global Variables</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2796</h2>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2797<!-- *********************************************************************** -->
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2798<div>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2799<p>LLVM has a number of "magic" global variables that contain data that affect
2800code generation or other IR semantics. These are documented here. All globals
Chris Lattner401e10c2009-07-20 06:14:25 +0000[diff] [blame]2801of this sort should have a section specified as "<tt>llvm.metadata</tt>". This
2802section and all globals that start with "<tt>llvm.</tt>" are reserved for use
2803by LLVM.</p>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2804
2805<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2806<h3>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2807<a name="intg_used">The '<tt>llvm.used</tt>' Global Variable</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2808</h3>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2809
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2810<div>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2811
2812<p>The <tt>@llvm.used</tt> global is an array with i8* element type which has <a
2813href="#linkage_appending">appending linkage</a>. This array contains a list of
2814pointers to global variables and functions which may optionally have a pointer
2815cast formed of bitcast or getelementptr. For example, a legal use of it is:</p>
2816
2817<pre>
2818 @X = global i8 4
2819 @Y = global i32 123
2820
2821 @llvm.used = appending global [2 x i8*] [
2822 i8* @X,
2823 i8* bitcast (i32* @Y to i8*)
2824 ], section "llvm.metadata"
2825</pre>
2826
2827<p>If a global variable appears in the <tt>@llvm.used</tt> list, then the
2828compiler, assembler, and linker are required to treat the symbol as if there is
2829a reference to the global that it cannot see. For example, if a variable has
2830internal linkage and no references other than that from the <tt>@llvm.used</tt>
2831list, it cannot be deleted. This is commonly used to represent references from
2832inline asms and other things the compiler cannot "see", and corresponds to
2833"attribute((used))" in GNU C.</p>
2834
2835<p>On some targets, the code generator must emit a directive to the assembler or
2836object file to prevent the assembler and linker from molesting the symbol.</p>
2837
2838</div>
2839
2840<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2841<h3>
2842 <a name="intg_compiler_used">
2843 The '<tt>llvm.compiler.used</tt>' Global Variable
2844 </a>
2845</h3>
Chris Lattner401e10c2009-07-20 06:14:25 +0000[diff] [blame]2846
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2847<div>
Chris Lattner401e10c2009-07-20 06:14:25 +0000[diff] [blame]2848
2849<p>The <tt>@llvm.compiler.used</tt> directive is the same as the
2850<tt>@llvm.used</tt> directive, except that it only prevents the compiler from
2851touching the symbol. On targets that support it, this allows an intelligent
2852linker to optimize references to the symbol without being impeded as it would be
2853by <tt>@llvm.used</tt>.</p>
2854
2855<p>This is a rare construct that should only be used in rare circumstances, and
2856should not be exposed to source languages.</p>
2857
2858</div>
2859
2860<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2861<h3>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2862<a name="intg_global_ctors">The '<tt>llvm.global_ctors</tt>' Global Variable</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2863</h3>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2864
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2865<div>
David Chisnalle31e9962010-04-30 19:23:49 +0000[diff] [blame]2866<pre>
2867%0 = type { i32, void ()* }
David Chisnall27195a52010-04-30 19:27:35 +0000[diff] [blame]2868@llvm.global_ctors = appending global [1 x %0] [%0 { i32 65535, void ()* @ctor }]
David Chisnalle31e9962010-04-30 19:23:49 +0000[diff] [blame]2869</pre>
2870<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.
2871</p>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2872
2873</div>
2874
2875<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2876<h3>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2877<a name="intg_global_dtors">The '<tt>llvm.global_dtors</tt>' Global Variable</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2878</h3>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2879
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2880<div>
David Chisnalle31e9962010-04-30 19:23:49 +0000[diff] [blame]2881<pre>
2882%0 = type { i32, void ()* }
David Chisnall27195a52010-04-30 19:27:35 +0000[diff] [blame]2883@llvm.global_dtors = appending global [1 x %0] [%0 { i32 65535, void ()* @dtor }]
David Chisnalle31e9962010-04-30 19:23:49 +0000[diff] [blame]2884</pre>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2885
David Chisnalle31e9962010-04-30 19:23:49 +0000[diff] [blame]2886<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.
2887</p>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2888
2889</div>
2890
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2891</div>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]2892
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2893<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2894<h2><a name="instref">Instruction Reference</a></h2>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2895<!-- *********************************************************************** -->
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2896
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2897<div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2898
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2899<p>The LLVM instruction set consists of several different classifications of
2900 instructions: <a href="#terminators">terminator
2901 instructions</a>, <a href="#binaryops">binary instructions</a>,
2902 <a href="#bitwiseops">bitwise binary instructions</a>,
2903 <a href="#memoryops">memory instructions</a>, and
2904 <a href="#otherops">other instructions</a>.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2905
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2906<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2907<h3>
2908 <a name="terminators">Terminator Instructions</a>
2909</h3>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2910
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2911<div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2912
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2913<p>As mentioned <a href="#functionstructure">previously</a>, every basic block
2914 in a program ends with a "Terminator" instruction, which indicates which
2915 block should be executed after the current block is finished. These
2916 terminator instructions typically yield a '<tt>void</tt>' value: they produce
2917 control flow, not values (the one exception being the
2918 '<a href="#i_invoke"><tt>invoke</tt></a>' instruction).</p>
2919
Duncan Sands83821c82010-04-15 20:35:54 +0000[diff] [blame]2920<p>There are seven different terminator instructions: the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2921 '<a href="#i_ret"><tt>ret</tt></a>' instruction, the
2922 '<a href="#i_br"><tt>br</tt></a>' instruction, the
2923 '<a href="#i_switch"><tt>switch</tt></a>' instruction, the
Bill Wendling21c346e2009-11-02 00:25:26 +0000[diff] [blame]2924 '<a href="#i_indirectbr">'<tt>indirectbr</tt></a>' Instruction, the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2925 '<a href="#i_invoke"><tt>invoke</tt></a>' instruction, the
2926 '<a href="#i_unwind"><tt>unwind</tt></a>' instruction, and the
2927 '<a href="#i_unreachable"><tt>unreachable</tt></a>' instruction.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2928
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2929<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2930<h4>
2931 <a name="i_ret">'<tt>ret</tt>' Instruction</a>
2932</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2933
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2934<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2935
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2936<h5>Syntax:</h5>
Dan Gohmanb1e6b962008-10-04 19:00:07 +0000[diff] [blame]2937<pre>
2938 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]2939 ret void <i>; Return from void function</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2940</pre>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]2941
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2942<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2943<p>The '<tt>ret</tt>' instruction is used to return control flow (and optionally
2944 a value) from a function back to the caller.</p>
2945
2946<p>There are two forms of the '<tt>ret</tt>' instruction: one that returns a
2947 value and then causes control flow, and one that just causes control flow to
2948 occur.</p>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]2949
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2950<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2951<p>The '<tt>ret</tt>' instruction optionally accepts a single argument, the
2952 return value. The type of the return value must be a
2953 '<a href="#t_firstclass">first class</a>' type.</p>
Dan Gohmanb1e6b962008-10-04 19:00:07 +0000[diff] [blame]2954
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2955<p>A function is not <a href="#wellformed">well formed</a> if it it has a
2956 non-void return type and contains a '<tt>ret</tt>' instruction with no return
2957 value or a return value with a type that does not match its type, or if it
2958 has a void return type and contains a '<tt>ret</tt>' instruction with a
2959 return value.</p>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]2960
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2961<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2962<p>When the '<tt>ret</tt>' instruction is executed, control flow returns back to
2963 the calling function's context. If the caller is a
2964 "<a href="#i_call"><tt>call</tt></a>" instruction, execution continues at the
2965 instruction after the call. If the caller was an
2966 "<a href="#i_invoke"><tt>invoke</tt></a>" instruction, execution continues at
2967 the beginning of the "normal" destination block. If the instruction returns
2968 a value, that value shall set the call or invoke instruction's return
2969 value.</p>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]2970
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2971<h5>Example:</h5>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]2972<pre>
2973 ret i32 5 <i>; Return an integer value of 5</i>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2974 ret void <i>; Return from a void function</i>
Bill Wendling0a4bbbf2009-02-28 22:12:54 +0000[diff] [blame]2975 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]2976</pre>
Dan Gohmanf3e60bd2009-01-12 23:12:39 +0000[diff] [blame]2977
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2978</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2979<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2980<h4>
2981 <a name="i_br">'<tt>br</tt>' Instruction</a>
2982</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2983
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2984<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2985
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2986<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2987<pre>
2988 br i1 &lt;cond&gt;, label &lt;iftrue&gt;, label &lt;iffalse&gt;<br> br label &lt;dest&gt; <i>; Unconditional branch</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2989</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2990
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2991<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2992<p>The '<tt>br</tt>' instruction is used to cause control flow to transfer to a
2993 different basic block in the current function. There are two forms of this
2994 instruction, corresponding to a conditional branch and an unconditional
2995 branch.</p>
2996
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2997<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2998<p>The conditional branch form of the '<tt>br</tt>' instruction takes a single
2999 '<tt>i1</tt>' value and two '<tt>label</tt>' values. The unconditional form
3000 of the '<tt>br</tt>' instruction takes a single '<tt>label</tt>' value as a
3001 target.</p>
3002
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3003<h5>Semantics:</h5>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3004<p>Upon execution of a conditional '<tt>br</tt>' instruction, the '<tt>i1</tt>'
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3005 argument is evaluated. If the value is <tt>true</tt>, control flows to the
3006 '<tt>iftrue</tt>' <tt>label</tt> argument. If "cond" is <tt>false</tt>,
3007 control flows to the '<tt>iffalse</tt>' <tt>label</tt> argument.</p>
3008
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3009<h5>Example:</h5>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]3010<pre>
3011Test:
3012 %cond = <a href="#i_icmp">icmp</a> eq i32 %a, %b
3013 br i1 %cond, label %IfEqual, label %IfUnequal
3014IfEqual:
3015 <a href="#i_ret">ret</a> i32 1
3016IfUnequal:
3017 <a href="#i_ret">ret</a> i32 0
3018</pre>
3019
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3020</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3021
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3022<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3023<h4>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3024 <a name="i_switch">'<tt>switch</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3025</h4>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3026
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3027<div>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3028
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3029<h5>Syntax:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3030<pre>
3031 switch &lt;intty&gt; &lt;value&gt;, label &lt;defaultdest&gt; [ &lt;intty&gt; &lt;val&gt;, label &lt;dest&gt; ... ]
3032</pre>
3033
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3034<h5>Overview:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3035<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]3036 several different places. It is a generalization of the '<tt>br</tt>'
3037 instruction, allowing a branch to occur to one of many possible
3038 destinations.</p>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3039
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3040<h5>Arguments:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3041<p>The '<tt>switch</tt>' instruction uses three parameters: an integer
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3042 comparison value '<tt>value</tt>', a default '<tt>label</tt>' destination,
3043 and an array of pairs of comparison value constants and '<tt>label</tt>'s.
3044 The table is not allowed to contain duplicate constant entries.</p>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3045
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3046<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3047<p>The <tt>switch</tt> instruction specifies a table of values and
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3048 destinations. When the '<tt>switch</tt>' instruction is executed, this table
3049 is searched for the given value. If the value is found, control flow is
Benjamin Kramer8040cd32009-10-12 14:46:08 +0000[diff] [blame]3050 transferred to the corresponding destination; otherwise, control flow is
3051 transferred to the default destination.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3052
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3053<h5>Implementation:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3054<p>Depending on properties of the target machine and the particular
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3055 <tt>switch</tt> instruction, this instruction may be code generated in
3056 different ways. For example, it could be generated as a series of chained
3057 conditional branches or with a lookup table.</p>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3058
3059<h5>Example:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3060<pre>
3061 <i>; Emulate a conditional br instruction</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3062 %Val = <a href="#i_zext">zext</a> i1 %value to i32
Dan Gohman2a08c532009-01-04 23:44:43 +0000[diff] [blame]3063 switch i32 %Val, label %truedest [ i32 0, label %falsedest ]
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3064
3065 <i>; Emulate an unconditional br instruction</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3066 switch i32 0, label %dest [ ]
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3067
3068 <i>; Implement a jump table:</i>
Dan Gohman2a08c532009-01-04 23:44:43 +0000[diff] [blame]3069 switch i32 %val, label %otherwise [ i32 0, label %onzero
3070 i32 1, label %onone
3071 i32 2, label %ontwo ]
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3072</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3073
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3074</div>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3075
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]3076
3077<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3078<h4>
Chris Lattnerab21db72009-10-28 00:19:10 +0000[diff] [blame]3079 <a name="i_indirectbr">'<tt>indirectbr</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3080</h4>
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]3081
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3082<div>
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]3083
3084<h5>Syntax:</h5>
3085<pre>
Chris Lattnerab21db72009-10-28 00:19:10 +0000[diff] [blame]3086 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]3087</pre>
3088
3089<h5>Overview:</h5>
3090
Chris Lattnerab21db72009-10-28 00:19:10 +0000[diff] [blame]3091<p>The '<tt>indirectbr</tt>' instruction implements an indirect branch to a label
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]3092 within the current function, whose address is specified by
Chris Lattnerc6f44362009-10-27 21:01:34 +0000[diff] [blame]3093 "<tt>address</tt>". Address must be derived from a <a
3094 href="#blockaddress">blockaddress</a> constant.</p>
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]3095
3096<h5>Arguments:</h5>
3097
3098<p>The '<tt>address</tt>' argument is the address of the label to jump to. The
3099 rest of the arguments indicate the full set of possible destinations that the
3100 address may point to. Blocks are allowed to occur multiple times in the
3101 destination list, though this isn't particularly useful.</p>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3102
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]3103<p>This destination list is required so that dataflow analysis has an accurate
3104 understanding of the CFG.</p>
3105
3106<h5>Semantics:</h5>
3107
3108<p>Control transfers to the block specified in the address argument. All
3109 possible destination blocks must be listed in the label list, otherwise this
3110 instruction has undefined behavior. This implies that jumps to labels
3111 defined in other functions have undefined behavior as well.</p>
3112
3113<h5>Implementation:</h5>
3114
3115<p>This is typically implemented with a jump through a register.</p>
3116
3117<h5>Example:</h5>
3118<pre>
Chris Lattnerab21db72009-10-28 00:19:10 +0000[diff] [blame]3119 indirectbr i8* %Addr, [ label %bb1, label %bb2, label %bb3 ]
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]3120</pre>
3121
3122</div>
3123
3124
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3125<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3126<h4>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3127 <a name="i_invoke">'<tt>invoke</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3128</h4>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3129
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3130<div>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3131
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3132<h5>Syntax:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3133<pre>
Devang Patel307e8ab2008-10-07 17:48:33 +0000[diff] [blame]3134 &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]3135 to label &lt;normal label&gt; unwind label &lt;exception label&gt;
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3136</pre>
3137
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]3138<h5>Overview:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3139<p>The '<tt>invoke</tt>' instruction causes control to transfer to a specified
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3140 function, with the possibility of control flow transfer to either the
3141 '<tt>normal</tt>' label or the '<tt>exception</tt>' label. If the callee
3142 function returns with the "<tt><a href="#i_ret">ret</a></tt>" instruction,
3143 control flow will return to the "normal" label. If the callee (or any
3144 indirect callees) returns with the "<a href="#i_unwind"><tt>unwind</tt></a>"
3145 instruction, control is interrupted and continued at the dynamically nearest
3146 "exception" label.</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3147
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3148<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3149<p>This instruction requires several arguments:</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3150
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3151<ol>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3152 <li>The optional "cconv" marker indicates which <a href="#callingconv">calling
3153 convention</a> the call should use. If none is specified, the call
3154 defaults to using C calling conventions.</li>
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]3155
3156 <li>The optional <a href="#paramattrs">Parameter Attributes</a> list for
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3157 return values. Only '<tt>zeroext</tt>', '<tt>signext</tt>', and
3158 '<tt>inreg</tt>' attributes are valid here.</li>
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]3159
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3160 <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]3161 function value being invoked. In most cases, this is a direct function
3162 invocation, but indirect <tt>invoke</tt>s are just as possible, branching
3163 off an arbitrary pointer to function value.</li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3164
3165 <li>'<tt>function ptr val</tt>': An LLVM value containing a pointer to a
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3166 function to be invoked. </li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3167
3168 <li>'<tt>function args</tt>': argument list whose types match the function
Chris Lattner0724fbd2010-03-02 06:36:51 +0000[diff] [blame]3169 signature argument types and parameter attributes. All arguments must be
3170 of <a href="#t_firstclass">first class</a> type. If the function
3171 signature indicates the function accepts a variable number of arguments,
3172 the extra arguments can be specified.</li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3173
3174 <li>'<tt>normal label</tt>': the label reached when the called function
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3175 executes a '<tt><a href="#i_ret">ret</a></tt>' instruction. </li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3176
3177 <li>'<tt>exception label</tt>': the label reached when a callee returns with
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3178 the <a href="#i_unwind"><tt>unwind</tt></a> instruction. </li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3179
Devang Patel307e8ab2008-10-07 17:48:33 +0000[diff] [blame]3180 <li>The optional <a href="#fnattrs">function attributes</a> list. Only
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3181 '<tt>noreturn</tt>', '<tt>nounwind</tt>', '<tt>readonly</tt>' and
3182 '<tt>readnone</tt>' attributes are valid here.</li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3183</ol>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3184
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3185<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3186<p>This instruction is designed to operate as a standard
3187 '<tt><a href="#i_call">call</a></tt>' instruction in most regards. The
3188 primary difference is that it establishes an association with a label, which
3189 is used by the runtime library to unwind the stack.</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3190
3191<p>This instruction is used in languages with destructors to ensure that proper
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3192 cleanup is performed in the case of either a <tt>longjmp</tt> or a thrown
3193 exception. Additionally, this is important for implementation of
3194 '<tt>catch</tt>' clauses in high-level languages that support them.</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3195
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3196<p>For the purposes of the SSA form, the definition of the value returned by the
3197 '<tt>invoke</tt>' instruction is deemed to occur on the edge from the current
3198 block to the "normal" label. If the callee unwinds then no return value is
3199 available.</p>
Dan Gohmanf96a4992009-05-22 21:47:08 +0000[diff] [blame]3200
Chris Lattnerdf7a6802010-01-15 18:08:37 +0000[diff] [blame]3201<p>Note that the code generator does not yet completely support unwind, and
3202that the invoke/unwind semantics are likely to change in future versions.</p>
3203
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3204<h5>Example:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3205<pre>
Nick Lewyckyd703f652008-03-16 07:18:12 +0000[diff] [blame]3206 %retval = invoke i32 @Test(i32 15) to label %Continue
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]3207 unwind label %TestCleanup <i>; {i32}:retval set</i>
Nick Lewyckyd703f652008-03-16 07:18:12 +0000[diff] [blame]3208 %retval = invoke <a href="#callingconv">coldcc</a> i32 %Testfnptr(i32 15) to label %Continue
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]3209 unwind label %TestCleanup <i>; {i32}:retval set</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3210</pre>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3211
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3212</div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3213
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]3214<!-- _______________________________________________________________________ -->
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3215
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3216<h4>
3217 <a name="i_unwind">'<tt>unwind</tt>' Instruction</a>
3218</h4>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3219
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3220<div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3221
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]3222<h5>Syntax:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3223<pre>
3224 unwind
3225</pre>
3226
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]3227<h5>Overview:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3228<p>The '<tt>unwind</tt>' instruction unwinds the stack, continuing control flow
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3229 at the first callee in the dynamic call stack which used
3230 an <a href="#i_invoke"><tt>invoke</tt></a> instruction to perform the call.
3231 This is primarily used to implement exception handling.</p>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3232
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]3233<h5>Semantics:</h5>
Chris Lattner72ed2002008-04-19 21:01:16 +0000[diff] [blame]3234<p>The '<tt>unwind</tt>' instruction causes execution of the current function to
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3235 immediately halt. The dynamic call stack is then searched for the
3236 first <a href="#i_invoke"><tt>invoke</tt></a> instruction on the call stack.
3237 Once found, execution continues at the "exceptional" destination block
3238 specified by the <tt>invoke</tt> instruction. If there is no <tt>invoke</tt>
3239 instruction in the dynamic call chain, undefined behavior results.</p>
3240
Chris Lattnerdf7a6802010-01-15 18:08:37 +0000[diff] [blame]3241<p>Note that the code generator does not yet completely support unwind, and
3242that the invoke/unwind semantics are likely to change in future versions.</p>
3243
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3244</div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3245
3246<!-- _______________________________________________________________________ -->
3247
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3248<h4>
3249 <a name="i_unreachable">'<tt>unreachable</tt>' Instruction</a>
3250</h4>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3251
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3252<div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3253
3254<h5>Syntax:</h5>
3255<pre>
3256 unreachable
3257</pre>
3258
3259<h5>Overview:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3260<p>The '<tt>unreachable</tt>' instruction has no defined semantics. This
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3261 instruction is used to inform the optimizer that a particular portion of the
3262 code is not reachable. This can be used to indicate that the code after a
3263 no-return function cannot be reached, and other facts.</p>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3264
3265<h5>Semantics:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3266<p>The '<tt>unreachable</tt>' instruction has no defined semantics.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3267
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3268</div>
3269
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3270</div>
3271
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3272<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3273<h3>
3274 <a name="binaryops">Binary Operations</a>
3275</h3>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3276
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3277<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3278
3279<p>Binary operators are used to do most of the computation in a program. They
3280 require two operands of the same type, execute an operation on them, and
3281 produce a single value. The operands might represent multiple data, as is
3282 the case with the <a href="#t_vector">vector</a> data type. The result value
3283 has the same type as its operands.</p>
3284
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3285<p>There are several different binary operators:</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3286
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3287<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3288<h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3289 <a name="i_add">'<tt>add</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3290</h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3291
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3292<div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3293
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3294<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3295<pre>
Dan Gohmancbb38f22009-07-20 22:41:19 +0000[diff] [blame]3296 &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]3297 &lt;result&gt; = add nuw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3298 &lt;result&gt; = add nsw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3299 &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]3300</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3301
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3302<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3303<p>The '<tt>add</tt>' instruction returns the sum of its two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3304
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3305<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3306<p>The two arguments to the '<tt>add</tt>' instruction must
3307 be <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of
3308 integer values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3309
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3310<h5>Semantics:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3311<p>The value produced is the integer sum of the two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3312
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3313<p>If the sum has unsigned overflow, the result returned is the mathematical
3314 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]3315
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3316<p>Because LLVM integers use a two's complement representation, this instruction
3317 is appropriate for both signed and unsigned integers.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3318
Dan Gohman08d012e2009-07-22 22:44:56 +0000[diff] [blame]3319<p><tt>nuw</tt> and <tt>nsw</tt> stand for &quot;No Unsigned Wrap&quot;
3320 and &quot;No Signed Wrap&quot;, respectively. If the <tt>nuw</tt> and/or
3321 <tt>nsw</tt> keywords are present, the result value of the <tt>add</tt>
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]3322 is a <a href="#trapvalues">trap value</a> if unsigned and/or signed overflow,
3323 respectively, occurs.</p>
Dan Gohmancbb38f22009-07-20 22:41:19 +0000[diff] [blame]3324
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3325<h5>Example:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3326<pre>
3327 &lt;result&gt; = add i32 4, %var <i>; yields {i32}:result = 4 + %var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3328</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3329
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3330</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3331
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3332<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3333<h4>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3334 <a name="i_fadd">'<tt>fadd</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3335</h4>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3336
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3337<div>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3338
3339<h5>Syntax:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3340<pre>
3341 &lt;result&gt; = fadd &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3342</pre>
3343
3344<h5>Overview:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3345<p>The '<tt>fadd</tt>' instruction returns the sum of its two operands.</p>
3346
3347<h5>Arguments:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3348<p>The two arguments to the '<tt>fadd</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3349 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
3350 floating point values. Both arguments must have identical types.</p>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3351
3352<h5>Semantics:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3353<p>The value produced is the floating point sum of the two operands.</p>
3354
3355<h5>Example:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3356<pre>
3357 &lt;result&gt; = fadd float 4.0, %var <i>; yields {float}:result = 4.0 + %var</i>
3358</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3359
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3360</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3361
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3362<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3363<h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3364 <a name="i_sub">'<tt>sub</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3365</h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3366
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3367<div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3368
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3369<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3370<pre>
Dan Gohman08d012e2009-07-22 22:44:56 +0000[diff] [blame]3371 &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]3372 &lt;result&gt; = sub nuw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3373 &lt;result&gt; = sub nsw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3374 &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]3375</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3376
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3377<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3378<p>The '<tt>sub</tt>' instruction returns the difference of its two
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3379 operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3380
3381<p>Note that the '<tt>sub</tt>' instruction is used to represent the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3382 '<tt>neg</tt>' instruction present in most other intermediate
3383 representations.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3384
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3385<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3386<p>The two arguments to the '<tt>sub</tt>' instruction must
3387 be <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of
3388 integer values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3389
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3390<h5>Semantics:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3391<p>The value produced is the integer difference of the two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3392
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3393<p>If the difference has unsigned overflow, the result returned is the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3394 mathematical result modulo 2<sup>n</sup>, where n is the bit width of the
3395 result.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3396
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3397<p>Because LLVM integers use a two's complement representation, this instruction
3398 is appropriate for both signed and unsigned integers.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3399
Dan Gohman08d012e2009-07-22 22:44:56 +0000[diff] [blame]3400<p><tt>nuw</tt> and <tt>nsw</tt> stand for &quot;No Unsigned Wrap&quot;
3401 and &quot;No Signed Wrap&quot;, respectively. If the <tt>nuw</tt> and/or
3402 <tt>nsw</tt> keywords are present, the result value of the <tt>sub</tt>
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]3403 is a <a href="#trapvalues">trap value</a> if unsigned and/or signed overflow,
3404 respectively, occurs.</p>
Dan Gohmancbb38f22009-07-20 22:41:19 +0000[diff] [blame]3405
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3406<h5>Example:</h5>
Bill Wendlingaac388b2007-05-29 09:42:13 +0000[diff] [blame]3407<pre>
3408 &lt;result&gt; = sub i32 4, %var <i>; yields {i32}:result = 4 - %var</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3409 &lt;result&gt; = sub i32 0, %val <i>; yields {i32}:result = -%var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3410</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3411
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3412</div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3413
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3414<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3415<h4>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3416 <a name="i_fsub">'<tt>fsub</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3417</h4>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3418
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3419<div>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3420
3421<h5>Syntax:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3422<pre>
3423 &lt;result&gt; = fsub &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3424</pre>
3425
3426<h5>Overview:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3427<p>The '<tt>fsub</tt>' instruction returns the difference of its two
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3428 operands.</p>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3429
3430<p>Note that the '<tt>fsub</tt>' instruction is used to represent the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3431 '<tt>fneg</tt>' instruction present in most other intermediate
3432 representations.</p>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3433
3434<h5>Arguments:</h5>
Bill Wendlingd9fe2982009-07-20 02:32:41 +0000[diff] [blame]3435<p>The two arguments to the '<tt>fsub</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3436 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
3437 floating point values. Both arguments must have identical types.</p>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3438
3439<h5>Semantics:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3440<p>The value produced is the floating point difference of the two operands.</p>
3441
3442<h5>Example:</h5>
3443<pre>
3444 &lt;result&gt; = fsub float 4.0, %var <i>; yields {float}:result = 4.0 - %var</i>
3445 &lt;result&gt; = fsub float -0.0, %val <i>; yields {float}:result = -%var</i>
3446</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3447
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3448</div>
3449
3450<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3451<h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3452 <a name="i_mul">'<tt>mul</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3453</h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3454
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3455<div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3456
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3457<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3458<pre>
Dan Gohman08d012e2009-07-22 22:44:56 +0000[diff] [blame]3459 &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]3460 &lt;result&gt; = mul nuw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3461 &lt;result&gt; = mul nsw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3462 &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]3463</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3464
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3465<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3466<p>The '<tt>mul</tt>' instruction returns the product of its two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3467
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3468<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3469<p>The two arguments to the '<tt>mul</tt>' instruction must
3470 be <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of
3471 integer values. Both arguments must have identical types.</p>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3472
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3473<h5>Semantics:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3474<p>The value produced is the integer product of the two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3475
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3476<p>If the result of the multiplication has unsigned overflow, the result
3477 returned is the mathematical result modulo 2<sup>n</sup>, where n is the bit
3478 width of the result.</p>
3479
3480<p>Because LLVM integers use a two's complement representation, and the result
3481 is the same width as the operands, this instruction returns the correct
3482 result for both signed and unsigned integers. If a full product
3483 (e.g. <tt>i32</tt>x<tt>i32</tt>-><tt>i64</tt>) is needed, the operands should
3484 be sign-extended or zero-extended as appropriate to the width of the full
3485 product.</p>
3486
Dan Gohman08d012e2009-07-22 22:44:56 +0000[diff] [blame]3487<p><tt>nuw</tt> and <tt>nsw</tt> stand for &quot;No Unsigned Wrap&quot;
3488 and &quot;No Signed Wrap&quot;, respectively. If the <tt>nuw</tt> and/or
3489 <tt>nsw</tt> keywords are present, the result value of the <tt>mul</tt>
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]3490 is a <a href="#trapvalues">trap value</a> if unsigned and/or signed overflow,
3491 respectively, occurs.</p>
Dan Gohmancbb38f22009-07-20 22:41:19 +0000[diff] [blame]3492
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3493<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3494<pre>
3495 &lt;result&gt; = mul i32 4, %var <i>; yields {i32}:result = 4 * %var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3496</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3497
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3498</div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3499
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3500<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3501<h4>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3502 <a name="i_fmul">'<tt>fmul</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3503</h4>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3504
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3505<div>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3506
3507<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3508<pre>
3509 &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]3510</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3511
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3512<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3513<p>The '<tt>fmul</tt>' instruction returns the product of its two operands.</p>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3514
3515<h5>Arguments:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3516<p>The two arguments to the '<tt>fmul</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3517 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
3518 floating point values. Both arguments must have identical types.</p>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3519
3520<h5>Semantics:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3521<p>The value produced is the floating point product of the two operands.</p>
3522
3523<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3524<pre>
3525 &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]3526</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3527
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3528</div>
3529
3530<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3531<h4>
3532 <a name="i_udiv">'<tt>udiv</tt>' Instruction</a>
3533</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3534
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3535<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3536
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3537<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3538<pre>
Chris Lattner35bda892011-02-06 21:44:57 +0000[diff] [blame]3539 &lt;result&gt; = udiv &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3540 &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]3541</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3542
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3543<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3544<p>The '<tt>udiv</tt>' instruction returns the quotient of its two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3545
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3546<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3547<p>The two arguments to the '<tt>udiv</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3548 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3549 values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3550
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3551<h5>Semantics:</h5>
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3552<p>The value produced is the unsigned integer quotient of the two operands.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3553
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3554<p>Note that unsigned integer division and signed integer division are distinct
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3555 operations; for signed integer division, use '<tt>sdiv</tt>'.</p>
3556
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3557<p>Division by zero leads to undefined behavior.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3558
Chris Lattner35bda892011-02-06 21:44:57 +0000[diff] [blame]3559<p>If the <tt>exact</tt> keyword is present, the result value of the
3560 <tt>udiv</tt> is a <a href="#trapvalues">trap value</a> if %op1 is not a
3561 multiple of %op2 (as such, "((a udiv exact b) mul b) == a").</p>
3562
3563
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3564<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3565<pre>
3566 &lt;result&gt; = udiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3567</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3568
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3569</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3570
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3571<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3572<h4>
3573 <a name="i_sdiv">'<tt>sdiv</tt>' Instruction</a>
3574</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3575
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3576<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3577
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3578<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3579<pre>
Dan Gohmancbb38f22009-07-20 22:41:19 +0000[diff] [blame]3580 &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]3581 &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]3582</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3583
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3584<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3585<p>The '<tt>sdiv</tt>' instruction returns the quotient of its two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3586
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3587<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3588<p>The two arguments to the '<tt>sdiv</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3589 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3590 values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3591
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3592<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3593<p>The value produced is the signed integer quotient of the two operands rounded
3594 towards zero.</p>
3595
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3596<p>Note that signed integer division and unsigned integer division are distinct
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3597 operations; for unsigned integer division, use '<tt>udiv</tt>'.</p>
3598
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3599<p>Division by zero leads to undefined behavior. Overflow also leads to
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3600 undefined behavior; this is a rare case, but can occur, for example, by doing
3601 a 32-bit division of -2147483648 by -1.</p>
3602
Dan Gohman9c5beed2009-07-22 00:04:19 +0000[diff] [blame]3603<p>If the <tt>exact</tt> keyword is present, the result value of the
Dan Gohman27ef9972010-04-23 15:23:32 +0000[diff] [blame]3604 <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]3605 be rounded.</p>
Dan Gohmancbb38f22009-07-20 22:41:19 +0000[diff] [blame]3606
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3607<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3608<pre>
3609 &lt;result&gt; = sdiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3610</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3611
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3612</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3613
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3614<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3615<h4>
3616 <a name="i_fdiv">'<tt>fdiv</tt>' Instruction</a>
3617</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3618
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3619<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3620
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3621<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3622<pre>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]3623 &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]3624</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3625
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3626<h5>Overview:</h5>
3627<p>The '<tt>fdiv</tt>' instruction returns the quotient of its two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3628
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3629<h5>Arguments:</h5>
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]3630<p>The two arguments to the '<tt>fdiv</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3631 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
3632 floating point values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3633
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3634<h5>Semantics:</h5>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]3635<p>The value produced is the floating point quotient of the two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3636
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3637<h5>Example:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3638<pre>
3639 &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]3640</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3641
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3642</div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3643
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3644<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3645<h4>
3646 <a name="i_urem">'<tt>urem</tt>' Instruction</a>
3647</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3648
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3649<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3650
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3651<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3652<pre>
3653 &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]3654</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3655
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3656<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3657<p>The '<tt>urem</tt>' instruction returns the remainder from the unsigned
3658 division of its two arguments.</p>
3659
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3660<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3661<p>The two arguments to the '<tt>urem</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3662 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3663 values. Both arguments must have identical types.</p>
3664
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3665<h5>Semantics:</h5>
3666<p>This instruction returns the unsigned integer <i>remainder</i> of a division.
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3667 This instruction always performs an unsigned division to get the
3668 remainder.</p>
3669
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3670<p>Note that unsigned integer remainder and signed integer remainder are
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3671 distinct operations; for signed integer remainder, use '<tt>srem</tt>'.</p>
3672
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3673<p>Taking the remainder of a division by zero leads to undefined behavior.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3674
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3675<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3676<pre>
3677 &lt;result&gt; = urem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3678</pre>
3679
3680</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3681
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3682<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3683<h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3684 <a name="i_srem">'<tt>srem</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3685</h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3686
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3687<div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3688
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3689<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3690<pre>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]3691 &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]3692</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3693
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3694<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3695<p>The '<tt>srem</tt>' instruction returns the remainder from the signed
3696 division of its two operands. This instruction can also take
3697 <a href="#t_vector">vector</a> versions of the values in which case the
3698 elements must be integers.</p>
Chris Lattnerc7d3ab32008-01-04 04:33:49 +0000[diff] [blame]3699
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3700<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3701<p>The two arguments to the '<tt>srem</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3702 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3703 values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3704
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3705<h5>Semantics:</h5>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3706<p>This instruction returns the <i>remainder</i> of a division (where the result
Duncan Sandsdea3a5e2011-03-07 09:12:24 +0000[diff] [blame]3707 is either zero or has the same sign as the dividend, <tt>op1</tt>), not the
3708 <i>modulo</i> operator (where the result is either zero or has the same sign
3709 as the divisor, <tt>op2</tt>) of a value.
3710 For more information about the difference,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3711 see <a href="http://mathforum.org/dr.math/problems/anne.4.28.99.html">The
3712 Math Forum</a>. For a table of how this is implemented in various languages,
3713 please see <a href="http://en.wikipedia.org/wiki/Modulo_operation">
3714 Wikipedia: modulo operation</a>.</p>
3715
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3716<p>Note that signed integer remainder and unsigned integer remainder are
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3717 distinct operations; for unsigned integer remainder, use '<tt>urem</tt>'.</p>
3718
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]3719<p>Taking the remainder of a division by zero leads to undefined behavior.
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3720 Overflow also leads to undefined behavior; this is a rare case, but can
3721 occur, for example, by taking the remainder of a 32-bit division of
3722 -2147483648 by -1. (The remainder doesn't actually overflow, but this rule
3723 lets srem be implemented using instructions that return both the result of
3724 the division and the remainder.)</p>
3725
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3726<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3727<pre>
3728 &lt;result&gt; = srem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3729</pre>
3730
3731</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3732
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3733<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3734<h4>
3735 <a name="i_frem">'<tt>frem</tt>' Instruction</a>
3736</h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3737
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3738<div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3739
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3740<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3741<pre>
3742 &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]3743</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3744
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3745<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3746<p>The '<tt>frem</tt>' instruction returns the remainder from the division of
3747 its two operands.</p>
3748
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3749<h5>Arguments:</h5>
3750<p>The two arguments to the '<tt>frem</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3751 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
3752 floating point values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3753
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3754<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3755<p>This instruction returns the <i>remainder</i> of a division. The remainder
3756 has the same sign as the dividend.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3757
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]3758<h5>Example:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3759<pre>
3760 &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]3761</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3762
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3763</div>
Robert Bocchino7b81c752006-02-17 21:18:08 +0000[diff] [blame]3764
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3765</div>
3766
Reid Spencer8e11bf82007-02-02 13:57:07 +0000[diff] [blame]3767<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3768<h3>
3769 <a name="bitwiseops">Bitwise Binary Operations</a>
3770</h3>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3771
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3772<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3773
3774<p>Bitwise binary operators are used to do various forms of bit-twiddling in a
3775 program. They are generally very efficient instructions and can commonly be
3776 strength reduced from other instructions. They require two operands of the
3777 same type, execute an operation on them, and produce a single value. The
3778 resulting value is the same type as its operands.</p>
3779
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3780<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3781<h4>
3782 <a name="i_shl">'<tt>shl</tt>' Instruction</a>
3783</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3784
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3785<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3786
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3787<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3788<pre>
Chris Lattnerf067d582011-02-07 16:40:21 +0000[diff] [blame]3789 &lt;result&gt; = shl &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3790 &lt;result&gt; = shl nuw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3791 &lt;result&gt; = shl nsw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3792 &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]3793</pre>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]3794
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3795<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3796<p>The '<tt>shl</tt>' instruction returns the first operand shifted to the left
3797 a specified number of bits.</p>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]3798
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3799<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3800<p>Both arguments to the '<tt>shl</tt>' instruction must be the
3801 same <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of
3802 integer type. '<tt>op2</tt>' is treated as an unsigned value.</p>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3803
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3804<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3805<p>The value produced is <tt>op1</tt> * 2<sup><tt>op2</tt></sup> mod
3806 2<sup>n</sup>, where <tt>n</tt> is the width of the result. If <tt>op2</tt>
3807 is (statically or dynamically) negative or equal to or larger than the number
3808 of bits in <tt>op1</tt>, the result is undefined. If the arguments are
3809 vectors, each vector element of <tt>op1</tt> is shifted by the corresponding
3810 shift amount in <tt>op2</tt>.</p>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]3811
Chris Lattnerf067d582011-02-07 16:40:21 +0000[diff] [blame]3812<p>If the <tt>nuw</tt> keyword is present, then the shift produces a
3813 <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]3814 the <tt>nsw</tt> keyword is present, then the shift produces a
Chris Lattnerf067d582011-02-07 16:40:21 +0000[diff] [blame]3815 <a href="#trapvalues">trap value</a> if it shifts out any bits that disagree
3816 with the resultant sign bit. As such, NUW/NSW have the same semantics as
3817 they would if the shift were expressed as a mul instruction with the same
3818 nsw/nuw bits in (mul %op1, (shl 1, %op2)).</p>
3819
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3820<h5>Example:</h5>
3821<pre>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3822 &lt;result&gt; = shl i32 4, %var <i>; yields {i32}: 4 &lt;&lt; %var</i>
3823 &lt;result&gt; = shl i32 4, 2 <i>; yields {i32}: 16</i>
3824 &lt;result&gt; = shl i32 1, 10 <i>; yields {i32}: 1024</i>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]3825 &lt;result&gt; = shl i32 1, 32 <i>; undefined</i>
Mon P Wange9f10152008-12-09 05:46:39 +0000[diff] [blame]3826 &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]3827</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3828
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3829</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3830
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3831<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3832<h4>
3833 <a name="i_lshr">'<tt>lshr</tt>' Instruction</a>
3834</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3835
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3836<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3837
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3838<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3839<pre>
Chris Lattnerf067d582011-02-07 16:40:21 +0000[diff] [blame]3840 &lt;result&gt; = lshr &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3841 &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]3842</pre>
3843
3844<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3845<p>The '<tt>lshr</tt>' instruction (logical shift right) returns the first
3846 operand shifted to the right a specified number of bits with zero fill.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3847
3848<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3849<p>Both arguments to the '<tt>lshr</tt>' instruction must be the same
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3850 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3851 type. '<tt>op2</tt>' is treated as an unsigned value.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3852
3853<h5>Semantics:</h5>
3854<p>This instruction always performs a logical shift right operation. The most
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3855 significant bits of the result will be filled with zero bits after the shift.
3856 If <tt>op2</tt> is (statically or dynamically) equal to or larger than the
3857 number of bits in <tt>op1</tt>, the result is undefined. If the arguments are
3858 vectors, each vector element of <tt>op1</tt> is shifted by the corresponding
3859 shift amount in <tt>op2</tt>.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3860
Chris Lattnerf067d582011-02-07 16:40:21 +0000[diff] [blame]3861<p>If the <tt>exact</tt> keyword is present, the result value of the
3862 <tt>lshr</tt> is a <a href="#trapvalues">trap value</a> if any of the bits
3863 shifted out are non-zero.</p>
3864
3865
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3866<h5>Example:</h5>
3867<pre>
3868 &lt;result&gt; = lshr i32 4, 1 <i>; yields {i32}:result = 2</i>
3869 &lt;result&gt; = lshr i32 4, 2 <i>; yields {i32}:result = 1</i>
3870 &lt;result&gt; = lshr i8 4, 3 <i>; yields {i8}:result = 0</i>
3871 &lt;result&gt; = lshr i8 -2, 1 <i>; yields {i8}:result = 0x7FFFFFFF </i>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]3872 &lt;result&gt; = lshr i32 1, 32 <i>; undefined</i>
Mon P Wange9f10152008-12-09 05:46:39 +0000[diff] [blame]3873 &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]3874</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3875
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3876</div>
3877
Reid Spencer8e11bf82007-02-02 13:57:07 +0000[diff] [blame]3878<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3879<h4>
3880 <a name="i_ashr">'<tt>ashr</tt>' Instruction</a>
3881</h4>
3882
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3883<div>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3884
3885<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3886<pre>
Chris Lattnerf067d582011-02-07 16:40:21 +0000[diff] [blame]3887 &lt;result&gt; = ashr &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3888 &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]3889</pre>
3890
3891<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3892<p>The '<tt>ashr</tt>' instruction (arithmetic shift right) returns the first
3893 operand shifted to the right a specified number of bits with sign
3894 extension.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3895
3896<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3897<p>Both arguments to the '<tt>ashr</tt>' instruction must be the same
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3898 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3899 type. '<tt>op2</tt>' is treated as an unsigned value.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3900
3901<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3902<p>This instruction always performs an arithmetic shift right operation, The
3903 most significant bits of the result will be filled with the sign bit
3904 of <tt>op1</tt>. If <tt>op2</tt> is (statically or dynamically) equal to or
3905 larger than the number of bits in <tt>op1</tt>, the result is undefined. If
3906 the arguments are vectors, each vector element of <tt>op1</tt> is shifted by
3907 the corresponding shift amount in <tt>op2</tt>.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3908
Chris Lattnerf067d582011-02-07 16:40:21 +0000[diff] [blame]3909<p>If the <tt>exact</tt> keyword is present, the result value of the
3910 <tt>ashr</tt> is a <a href="#trapvalues">trap value</a> if any of the bits
3911 shifted out are non-zero.</p>
3912
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3913<h5>Example:</h5>
3914<pre>
3915 &lt;result&gt; = ashr i32 4, 1 <i>; yields {i32}:result = 2</i>
3916 &lt;result&gt; = ashr i32 4, 2 <i>; yields {i32}:result = 1</i>
3917 &lt;result&gt; = ashr i8 4, 3 <i>; yields {i8}:result = 0</i>
3918 &lt;result&gt; = ashr i8 -2, 1 <i>; yields {i8}:result = -1</i>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]3919 &lt;result&gt; = ashr i32 1, 32 <i>; undefined</i>
Mon P Wange9f10152008-12-09 05:46:39 +0000[diff] [blame]3920 &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]3921</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3922
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]3923</div>
3924
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3925<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3926<h4>
3927 <a name="i_and">'<tt>and</tt>' Instruction</a>
3928</h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3929
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3930<div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3931
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3932<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3933<pre>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]3934 &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]3935</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3936
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3937<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3938<p>The '<tt>and</tt>' instruction returns the bitwise logical and of its two
3939 operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3940
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3941<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3942<p>The two arguments to the '<tt>and</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3943 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3944 values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3945
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3946<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3947<p>The truth table used for the '<tt>and</tt>' instruction is:</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3948
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3949<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3950 <tbody>
3951 <tr>
3952 <td>In0</td>
3953 <td>In1</td>
3954 <td>Out</td>
3955 </tr>
3956 <tr>
3957 <td>0</td>
3958 <td>0</td>
3959 <td>0</td>
3960 </tr>
3961 <tr>
3962 <td>0</td>
3963 <td>1</td>
3964 <td>0</td>
3965 </tr>
3966 <tr>
3967 <td>1</td>
3968 <td>0</td>
3969 <td>0</td>
3970 </tr>
3971 <tr>
3972 <td>1</td>
3973 <td>1</td>
3974 <td>1</td>
3975 </tr>
3976 </tbody>
3977</table>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3978
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3979<h5>Example:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3980<pre>
3981 &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]3982 &lt;result&gt; = and i32 15, 40 <i>; yields {i32}:result = 8</i>
3983 &lt;result&gt; = and i32 4, 8 <i>; yields {i32}:result = 0</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3984</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3985</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3986<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3987<h4>
3988 <a name="i_or">'<tt>or</tt>' Instruction</a>
3989</h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3990
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3991<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3992
3993<h5>Syntax:</h5>
3994<pre>
3995 &lt;result&gt; = or &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3996</pre>
3997
3998<h5>Overview:</h5>
3999<p>The '<tt>or</tt>' instruction returns the bitwise logical inclusive or of its
4000 two operands.</p>
4001
4002<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4003<p>The two arguments to the '<tt>or</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4004 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
4005 values. Both arguments must have identical types.</p>
4006
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4007<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4008<p>The truth table used for the '<tt>or</tt>' instruction is:</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4009
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4010<table border="1" cellspacing="0" cellpadding="4">
4011 <tbody>
4012 <tr>
4013 <td>In0</td>
4014 <td>In1</td>
4015 <td>Out</td>
4016 </tr>
4017 <tr>
4018 <td>0</td>
4019 <td>0</td>
4020 <td>0</td>
4021 </tr>
4022 <tr>
4023 <td>0</td>
4024 <td>1</td>
4025 <td>1</td>
4026 </tr>
4027 <tr>
4028 <td>1</td>
4029 <td>0</td>
4030 <td>1</td>
4031 </tr>
4032 <tr>
4033 <td>1</td>
4034 <td>1</td>
4035 <td>1</td>
4036 </tr>
4037 </tbody>
4038</table>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4039
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4040<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4041<pre>
4042 &lt;result&gt; = or i32 4, %var <i>; yields {i32}:result = 4 | %var</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4043 &lt;result&gt; = or i32 15, 40 <i>; yields {i32}:result = 47</i>
4044 &lt;result&gt; = or i32 4, 8 <i>; yields {i32}:result = 12</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4045</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4046
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4047</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4048
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4049<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4050<h4>
4051 <a name="i_xor">'<tt>xor</tt>' Instruction</a>
4052</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4053
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4054<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4055
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4056<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4057<pre>
4058 &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]4059</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4060
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4061<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4062<p>The '<tt>xor</tt>' instruction returns the bitwise logical exclusive or of
4063 its two operands. The <tt>xor</tt> is used to implement the "one's
4064 complement" operation, which is the "~" operator in C.</p>
4065
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4066<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4067<p>The two arguments to the '<tt>xor</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4068 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
4069 values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4070
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4071<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4072<p>The truth table used for the '<tt>xor</tt>' instruction is:</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4073
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4074<table border="1" cellspacing="0" cellpadding="4">
4075 <tbody>
4076 <tr>
4077 <td>In0</td>
4078 <td>In1</td>
4079 <td>Out</td>
4080 </tr>
4081 <tr>
4082 <td>0</td>
4083 <td>0</td>
4084 <td>0</td>
4085 </tr>
4086 <tr>
4087 <td>0</td>
4088 <td>1</td>
4089 <td>1</td>
4090 </tr>
4091 <tr>
4092 <td>1</td>
4093 <td>0</td>
4094 <td>1</td>
4095 </tr>
4096 <tr>
4097 <td>1</td>
4098 <td>1</td>
4099 <td>0</td>
4100 </tr>
4101 </tbody>
4102</table>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4103
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4104<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4105<pre>
4106 &lt;result&gt; = xor i32 4, %var <i>; yields {i32}:result = 4 ^ %var</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4107 &lt;result&gt; = xor i32 15, 40 <i>; yields {i32}:result = 39</i>
4108 &lt;result&gt; = xor i32 4, 8 <i>; yields {i32}:result = 12</i>
4109 &lt;result&gt; = xor i32 %V, -1 <i>; yields {i32}:result = ~%V</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4110</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4111
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4112</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4113
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4114</div>
4115
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4116<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4117<h3>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4118 <a name="vectorops">Vector Operations</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4119</h3>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4120
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4121<div>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4122
4123<p>LLVM supports several instructions to represent vector operations in a
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4124 target-independent manner. These instructions cover the element-access and
4125 vector-specific operations needed to process vectors effectively. While LLVM
4126 does directly support these vector operations, many sophisticated algorithms
4127 will want to use target-specific intrinsics to take full advantage of a
4128 specific target.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4129
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4130<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4131<h4>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4132 <a name="i_extractelement">'<tt>extractelement</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4133</h4>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4134
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4135<div>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4136
4137<h5>Syntax:</h5>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4138<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4139 &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]4140</pre>
4141
4142<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4143<p>The '<tt>extractelement</tt>' instruction extracts a single scalar element
4144 from a vector at a specified index.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4145
4146
4147<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4148<p>The first operand of an '<tt>extractelement</tt>' instruction is a value
4149 of <a href="#t_vector">vector</a> type. The second operand is an index
4150 indicating the position from which to extract the element. The index may be
4151 a variable.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4152
4153<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4154<p>The result is a scalar of the same type as the element type of
4155 <tt>val</tt>. Its value is the value at position <tt>idx</tt> of
4156 <tt>val</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
4157 results are undefined.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4158
4159<h5>Example:</h5>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4160<pre>
Gabor Greifa5b6f452009-10-28 13:14:50 +0000[diff] [blame]4161 &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]4162</pre>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4163
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4164</div>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4165
4166<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4167<h4>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4168 <a name="i_insertelement">'<tt>insertelement</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4169</h4>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4170
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4171<div>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4172
4173<h5>Syntax:</h5>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4174<pre>
Dan Gohmanf3480b92008-05-12 23:38:42 +0000[diff] [blame]4175 &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]4176</pre>
4177
4178<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4179<p>The '<tt>insertelement</tt>' instruction inserts a scalar element into a
4180 vector at a specified index.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4181
4182<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4183<p>The first operand of an '<tt>insertelement</tt>' instruction is a value
4184 of <a href="#t_vector">vector</a> type. The second operand is a scalar value
4185 whose type must equal the element type of the first operand. The third
4186 operand is an index indicating the position at which to insert the value.
4187 The index may be a variable.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4188
4189<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4190<p>The result is a vector of the same type as <tt>val</tt>. Its element values
4191 are those of <tt>val</tt> except at position <tt>idx</tt>, where it gets the
4192 value <tt>elt</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
4193 results are undefined.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4194
4195<h5>Example:</h5>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4196<pre>
Gabor Greifa5b6f452009-10-28 13:14:50 +0000[diff] [blame]4197 &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]4198</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4199
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4200</div>
4201
4202<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4203<h4>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4204 <a name="i_shufflevector">'<tt>shufflevector</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4205</h4>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4206
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4207<div>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4208
4209<h5>Syntax:</h5>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4210<pre>
Mon P Wangaeb06d22008-11-10 04:46:22 +0000[diff] [blame]4211 &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]4212</pre>
4213
4214<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4215<p>The '<tt>shufflevector</tt>' instruction constructs a permutation of elements
4216 from two input vectors, returning a vector with the same element type as the
4217 input and length that is the same as the shuffle mask.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4218
4219<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4220<p>The first two operands of a '<tt>shufflevector</tt>' instruction are vectors
4221 with types that match each other. The third argument is a shuffle mask whose
4222 element type is always 'i32'. The result of the instruction is a vector
4223 whose length is the same as the shuffle mask and whose element type is the
4224 same as the element type of the first two operands.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4225
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4226<p>The shuffle mask operand is required to be a constant vector with either
4227 constant integer or undef values.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4228
4229<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4230<p>The elements of the two input vectors are numbered from left to right across
4231 both of the vectors. The shuffle mask operand specifies, for each element of
4232 the result vector, which element of the two input vectors the result element
4233 gets. The element selector may be undef (meaning "don't care") and the
4234 second operand may be undef if performing a shuffle from only one vector.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4235
4236<h5>Example:</h5>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4237<pre>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4238 &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]4239 &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]4240 &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]4241 &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]4242 &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]4243 &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]4244 &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]4245 &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]4246</pre>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4247
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4248</div>
Tanya Lattner09474292006-04-14 19:24:33 +0000[diff] [blame]4249
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4250</div>
4251
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4252<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4253<h3>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4254 <a name="aggregateops">Aggregate Operations</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4255</h3>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4256
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4257<div>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4258
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4259<p>LLVM supports several instructions for working with
4260 <a href="#t_aggregate">aggregate</a> values.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4261
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4262<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4263<h4>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4264 <a name="i_extractvalue">'<tt>extractvalue</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4265</h4>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4266
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4267<div>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4268
4269<h5>Syntax:</h5>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4270<pre>
4271 &lt;result&gt; = extractvalue &lt;aggregate type&gt; &lt;val&gt;, &lt;idx&gt;{, &lt;idx&gt;}*
4272</pre>
4273
4274<h5>Overview:</h5>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4275<p>The '<tt>extractvalue</tt>' instruction extracts the value of a member field
4276 from an <a href="#t_aggregate">aggregate</a> value.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4277
4278<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4279<p>The first operand of an '<tt>extractvalue</tt>' instruction is a value
Chris Lattner61c70e92010-08-28 04:09:24 +0000[diff] [blame]4280 of <a href="#t_struct">struct</a> or
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4281 <a href="#t_array">array</a> type. The operands are constant indices to
4282 specify which value to extract in a similar manner as indices in a
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4283 '<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.</p>
Frits van Bommel13242892010-12-05 20:54:38 +0000[diff] [blame]4284 <p>The major differences to <tt>getelementptr</tt> indexing are:</p>
4285 <ul>
4286 <li>Since the value being indexed is not a pointer, the first index is
4287 omitted and assumed to be zero.</li>
4288 <li>At least one index must be specified.</li>
4289 <li>Not only struct indices but also array indices must be in
4290 bounds.</li>
4291 </ul>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4292
4293<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4294<p>The result is the value at the position in the aggregate specified by the
4295 index operands.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4296
4297<h5>Example:</h5>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4298<pre>
Gabor Greifa5b6f452009-10-28 13:14:50 +0000[diff] [blame]4299 &lt;result&gt; = extractvalue {i32, float} %agg, 0 <i>; yields i32</i>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4300</pre>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4301
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4302</div>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4303
4304<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4305<h4>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4306 <a name="i_insertvalue">'<tt>insertvalue</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4307</h4>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4308
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4309<div>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4310
4311<h5>Syntax:</h5>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4312<pre>
Chris Lattner8645d1a2011-05-22 07:18:08 +0000[diff] [blame]4313 &lt;result&gt; = insertvalue &lt;aggregate type&gt; &lt;val&gt;, &lt;ty&gt; &lt;elt&gt;, &lt;idx&gt;{, <idx>}* <i>; yields &lt;aggregate type&gt;</i>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4314</pre>
4315
4316<h5>Overview:</h5>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4317<p>The '<tt>insertvalue</tt>' instruction inserts a value into a member field
4318 in an <a href="#t_aggregate">aggregate</a> value.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4319
4320<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4321<p>The first operand of an '<tt>insertvalue</tt>' instruction is a value
Chris Lattner61c70e92010-08-28 04:09:24 +0000[diff] [blame]4322 of <a href="#t_struct">struct</a> or
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4323 <a href="#t_array">array</a> type. The second operand is a first-class
4324 value to insert. The following operands are constant indices indicating
4325 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]4326 '<tt><a href="#i_extractvalue">extractvalue</a></tt>' instruction. The
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4327 value to insert must have the same type as the value identified by the
4328 indices.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4329
4330<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4331<p>The result is an aggregate of the same type as <tt>val</tt>. Its value is
4332 that of <tt>val</tt> except that the value at the position specified by the
4333 indices is that of <tt>elt</tt>.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4334
4335<h5>Example:</h5>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4336<pre>
Chris Lattner8645d1a2011-05-22 07:18:08 +0000[diff] [blame]4337 %agg1 = insertvalue {i32, float} undef, i32 1, 0 <i>; yields {i32 1, float undef}</i>
4338 %agg2 = insertvalue {i32, float} %agg1, float %val, 1 <i>; yields {i32 1, float %val}</i>
4339 %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]4340</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4341
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4342</div>
4343
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4344</div>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4345
4346<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4347<h3>
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]4348 <a name="memoryops">Memory Access and Addressing Operations</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4349</h3>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4350
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4351<div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4352
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4353<p>A key design point of an SSA-based representation is how it represents
4354 memory. In LLVM, no memory locations are in SSA form, which makes things
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]4355 very simple. This section describes how to read, write, and allocate
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4356 memory in LLVM.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4357
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4358<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4359<h4>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4360 <a name="i_alloca">'<tt>alloca</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4361</h4>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4362
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4363<div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4364
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4365<h5>Syntax:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4366<pre>
Dan Gohmanf75a7d32010-05-28 01:14:11 +0000[diff] [blame]4367 &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]4368</pre>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4369
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4370<h5>Overview:</h5>
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]4371<p>The '<tt>alloca</tt>' instruction allocates memory on the stack frame of the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4372 currently executing function, to be automatically released when this function
4373 returns to its caller. The object is always allocated in the generic address
4374 space (address space zero).</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4375
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4376<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4377<p>The '<tt>alloca</tt>' instruction
4378 allocates <tt>sizeof(&lt;type&gt;)*NumElements</tt> bytes of memory on the
4379 runtime stack, returning a pointer of the appropriate type to the program.
4380 If "NumElements" is specified, it is the number of elements allocated,
4381 otherwise "NumElements" is defaulted to be one. If a constant alignment is
4382 specified, the value result of the allocation is guaranteed to be aligned to
4383 at least that boundary. If not specified, or if zero, the target can choose
4384 to align the allocation on any convenient boundary compatible with the
4385 type.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4386
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4387<p>'<tt>type</tt>' may be any sized type.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4388
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4389<h5>Semantics:</h5>
Bill Wendling871eb0a2009-05-08 20:49:29 +0000[diff] [blame]4390<p>Memory is allocated; a pointer is returned. The operation is undefined if
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4391 there is insufficient stack space for the allocation. '<tt>alloca</tt>'d
4392 memory is automatically released when the function returns. The
4393 '<tt>alloca</tt>' instruction is commonly used to represent automatic
4394 variables that must have an address available. When the function returns
4395 (either with the <tt><a href="#i_ret">ret</a></tt>
4396 or <tt><a href="#i_unwind">unwind</a></tt> instructions), the memory is
4397 reclaimed. Allocating zero bytes is legal, but the result is undefined.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4398
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4399<h5>Example:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4400<pre>
Dan Gohman81e21672009-01-04 23:49:44 +0000[diff] [blame]4401 %ptr = alloca i32 <i>; yields {i32*}:ptr</i>
4402 %ptr = alloca i32, i32 4 <i>; yields {i32*}:ptr</i>
4403 %ptr = alloca i32, i32 4, align 1024 <i>; yields {i32*}:ptr</i>
4404 %ptr = alloca i32, align 1024 <i>; yields {i32*}:ptr</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4405</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4406
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4407</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4408
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4409<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4410<h4>
4411 <a name="i_load">'<tt>load</tt>' Instruction</a>
4412</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4413
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4414<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4415
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4416<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4417<pre>
Bill Wendling7c78dbb2010-02-25 21:23:24 +0000[diff] [blame]4418 &lt;result&gt; = load &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;][, !nontemporal !&lt;index&gt;]
4419 &lt;result&gt; = volatile load &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;][, !nontemporal !&lt;index&gt;]
4420 !&lt;index&gt; = !{ i32 1 }
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4421</pre>
4422
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4423<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4424<p>The '<tt>load</tt>' instruction is used to read from memory.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4425
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4426<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4427<p>The argument to the '<tt>load</tt>' instruction specifies the memory address
4428 from which to load. The pointer must point to
4429 a <a href="#t_firstclass">first class</a> type. If the <tt>load</tt> is
4430 marked as <tt>volatile</tt>, then the optimizer is not allowed to modify the
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]4431 number or order of execution of this <tt>load</tt> with other <a
4432 href="#volatile">volatile operations</a>.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4433
Bill Wendling7c78dbb2010-02-25 21:23:24 +0000[diff] [blame]4434<p>The optional constant <tt>align</tt> argument specifies the alignment of the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4435 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]4436 omitted <tt>align</tt> argument means that the operation has the preferential
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4437 alignment for the target. It is the responsibility of the code emitter to
4438 ensure that the alignment information is correct. Overestimating the
Bill Wendling7c78dbb2010-02-25 21:23:24 +0000[diff] [blame]4439 alignment results in undefined behavior. Underestimating the alignment may
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4440 produce less efficient code. An alignment of 1 is always safe.</p>
4441
Bill Wendling7c78dbb2010-02-25 21:23:24 +0000[diff] [blame]4442<p>The optional <tt>!nontemporal</tt> metadata must reference a single
4443 metatadata name &lt;index&gt; corresponding to a metadata node with
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]4444 one <tt>i32</tt> entry of value 1. The existence of
Bill Wendling7c78dbb2010-02-25 21:23:24 +0000[diff] [blame]4445 the <tt>!nontemporal</tt> metatadata on the instruction tells the optimizer
4446 and code generator that this load is not expected to be reused in the cache.
4447 The code generator may select special instructions to save cache bandwidth,
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]4448 such as the <tt>MOVNT</tt> instruction on x86.</p>
David Greene8939b0d2010-02-16 20:50:18 +0000[diff] [blame]4449
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4450<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4451<p>The location of memory pointed to is loaded. If the value being loaded is of
4452 scalar type then the number of bytes read does not exceed the minimum number
4453 of bytes needed to hold all bits of the type. For example, loading an
4454 <tt>i24</tt> reads at most three bytes. When loading a value of a type like
4455 <tt>i20</tt> with a size that is not an integral number of bytes, the result
4456 is undefined if the value was not originally written using a store of the
4457 same type.</p>
4458
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4459<h5>Examples:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4460<pre>
4461 %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
4462 <a href="#i_store">store</a> i32 3, i32* %ptr <i>; yields {void}</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4463 %val = load i32* %ptr <i>; yields {i32}:val = i32 3</i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4464</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4465
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4466</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4467
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4468<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4469<h4>
4470 <a name="i_store">'<tt>store</tt>' Instruction</a>
4471</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4472
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4473<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4474
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4475<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4476<pre>
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]4477 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>
4478 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]4479</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4480
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4481<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4482<p>The '<tt>store</tt>' instruction is used to write to memory.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4483
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4484<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4485<p>There are two arguments to the '<tt>store</tt>' instruction: a value to store
4486 and an address at which to store it. The type of the
4487 '<tt>&lt;pointer&gt;</tt>' operand must be a pointer to
4488 the <a href="#t_firstclass">first class</a> type of the
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]4489 '<tt>&lt;value&gt;</tt>' operand. If the <tt>store</tt> is marked as
4490 <tt>volatile</tt>, then the optimizer is not allowed to modify the number or
4491 order of execution of this <tt>store</tt> with other <a
4492 href="#volatile">volatile operations</a>.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4493
4494<p>The optional constant "align" argument specifies the alignment of the
4495 operation (that is, the alignment of the memory address). A value of 0 or an
4496 omitted "align" argument means that the operation has the preferential
4497 alignment for the target. It is the responsibility of the code emitter to
4498 ensure that the alignment information is correct. Overestimating the
4499 alignment results in an undefined behavior. Underestimating the alignment may
4500 produce less efficient code. An alignment of 1 is always safe.</p>
4501
David Greene8939b0d2010-02-16 20:50:18 +0000[diff] [blame]4502<p>The optional !nontemporal metadata must reference a single metatadata
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]4503 name &lt;index&gt; corresponding to a metadata node with one i32 entry of
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]4504 value 1. The existence of the !nontemporal metatadata on the
David Greene8939b0d2010-02-16 20:50:18 +0000[diff] [blame]4505 instruction tells the optimizer and code generator that this load is
4506 not expected to be reused in the cache. The code generator may
4507 select special instructions to save cache bandwidth, such as the
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]4508 MOVNT instruction on x86.</p>
David Greene8939b0d2010-02-16 20:50:18 +0000[diff] [blame]4509
4510
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4511<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4512<p>The contents of memory are updated to contain '<tt>&lt;value&gt;</tt>' at the
4513 location specified by the '<tt>&lt;pointer&gt;</tt>' operand. If
4514 '<tt>&lt;value&gt;</tt>' is of scalar type then the number of bytes written
4515 does not exceed the minimum number of bytes needed to hold all bits of the
4516 type. For example, storing an <tt>i24</tt> writes at most three bytes. When
4517 writing a value of a type like <tt>i20</tt> with a size that is not an
4518 integral number of bytes, it is unspecified what happens to the extra bits
4519 that do not belong to the type, but they will typically be overwritten.</p>
4520
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4521<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4522<pre>
4523 %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
Bill Wendling8c6c72d2007-10-22 05:10:05 +0000[diff] [blame]4524 store i32 3, i32* %ptr <i>; yields {void}</i>
4525 %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]4526</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4527
Reid Spencer47ce1792006-11-09 21:15:49 +0000[diff] [blame]4528</div>
4529
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4530<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4531<h4>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4532 <a name="i_getelementptr">'<tt>getelementptr</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4533</h4>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4534
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4535<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4536
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]4537<h5>Syntax:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4538<pre>
Matthijs Kooijmane49d0bc2008-10-13 13:44:15 +0000[diff] [blame]4539 &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]4540 &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]4541</pre>
4542
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]4543<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4544<p>The '<tt>getelementptr</tt>' instruction is used to get the address of a
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4545 subelement of an <a href="#t_aggregate">aggregate</a> data structure.
4546 It performs address calculation only and does not access memory.</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4547
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]4548<h5>Arguments:</h5>
Matthijs Kooijmane49d0bc2008-10-13 13:44:15 +0000[diff] [blame]4549<p>The first argument is always a pointer, and forms the basis of the
Chris Lattnerc8eef442009-07-29 06:44:13 +0000[diff] [blame]4550 calculation. The remaining arguments are indices that indicate which of the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4551 elements of the aggregate object are indexed. The interpretation of each
4552 index is dependent on the type being indexed into. The first index always
4553 indexes the pointer value given as the first argument, the second index
4554 indexes a value of the type pointed to (not necessarily the value directly
4555 pointed to, since the first index can be non-zero), etc. The first type
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4556 indexed into must be a pointer value, subsequent types can be arrays,
Chris Lattner61c70e92010-08-28 04:09:24 +0000[diff] [blame]4557 vectors, and structs. Note that subsequent types being indexed into
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4558 can never be pointers, since that would require loading the pointer before
4559 continuing calculation.</p>
Matthijs Kooijmane49d0bc2008-10-13 13:44:15 +0000[diff] [blame]4560
4561<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]4562 When indexing into a (optionally packed) structure, only <tt>i32</tt>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4563 integer <b>constants</b> are allowed. When indexing into an array, pointer
4564 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]4565 constant.</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4566
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4567<p>For example, let's consider a C code fragment and how it gets compiled to
4568 LLVM:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4569
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]4570<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]4571struct RT {
4572 char A;
Chris Lattnercabc8462007-05-29 15:43:56 +0000[diff] [blame]4573 int B[10][20];
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]4574 char C;
4575};
4576struct ST {
Chris Lattnercabc8462007-05-29 15:43:56 +0000[diff] [blame]4577 int X;
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]4578 double Y;
4579 struct RT Z;
4580};
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4581
Chris Lattnercabc8462007-05-29 15:43:56 +0000[diff] [blame]4582int *foo(struct ST *s) {
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]4583 return &amp;s[1].Z.B[5][13];
4584}
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4585</pre>
4586
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4587<p>The LLVM code generated by the GCC frontend is:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4588
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]4589<pre class="doc_code">
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]4590%RT = <a href="#namedtypes">type</a> { i8 , [10 x [20 x i32]], i8 }
4591%ST = <a href="#namedtypes">type</a> { i32, double, %RT }
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4592
Dan Gohman4df605b2009-07-25 02:23:48 +0000[diff] [blame]4593define i32* @foo(%ST* %s) {
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]4594entry:
4595 %reg = getelementptr %ST* %s, i32 1, i32 2, i32 1, i32 5, i32 13
4596 ret i32* %reg
4597}
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4598</pre>
4599
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]4600<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4601<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]4602 type, which is a pointer, yielding a '<tt>%ST</tt>' = '<tt>{ i32, double, %RT
4603 }</tt>' type, a structure. The second index indexes into the third element
4604 of the structure, yielding a '<tt>%RT</tt>' = '<tt>{ i8 , [10 x [20 x i32]],
4605 i8 }</tt>' type, another structure. The third index indexes into the second
4606 element of the structure, yielding a '<tt>[10 x [20 x i32]]</tt>' type, an
4607 array. The two dimensions of the array are subscripted into, yielding an
4608 '<tt>i32</tt>' type. The '<tt>getelementptr</tt>' instruction returns a
4609 pointer to this element, thus computing a value of '<tt>i32*</tt>' type.</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4610
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4611<p>Note that it is perfectly legal to index partially through a structure,
4612 returning a pointer to an inner element. Because of this, the LLVM code for
4613 the given testcase is equivalent to:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4614
4615<pre>
Dan Gohman4df605b2009-07-25 02:23:48 +0000[diff] [blame]4616 define i32* @foo(%ST* %s) {
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4617 %t1 = getelementptr %ST* %s, i32 1 <i>; yields %ST*:%t1</i>
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]4618 %t2 = getelementptr %ST* %t1, i32 0, i32 2 <i>; yields %RT*:%t2</i>
4619 %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]4620 %t4 = getelementptr [10 x [20 x i32]]* %t3, i32 0, i32 5 <i>; yields [20 x i32]*:%t4</i>
4621 %t5 = getelementptr [20 x i32]* %t4, i32 0, i32 13 <i>; yields i32*:%t5</i>
4622 ret i32* %t5
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4623 }
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]4624</pre>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]4625
Dan Gohmandd8004d2009-07-27 21:53:46 +0000[diff] [blame]4626<p>If the <tt>inbounds</tt> keyword is present, the result value of the
Dan Gohman27ef9972010-04-23 15:23:32 +0000[diff] [blame]4627 <tt>getelementptr</tt> is a <a href="#trapvalues">trap value</a> if the
4628 base pointer is not an <i>in bounds</i> address of an allocated object,
4629 or if any of the addresses that would be formed by successive addition of
4630 the offsets implied by the indices to the base address with infinitely
4631 precise arithmetic are not an <i>in bounds</i> address of that allocated
4632 object. The <i>in bounds</i> addresses for an allocated object are all
4633 the addresses that point into the object, plus the address one byte past
4634 the end.</p>
Dan Gohmandd8004d2009-07-27 21:53:46 +0000[diff] [blame]4635
4636<p>If the <tt>inbounds</tt> keyword is not present, the offsets are added to
4637 the base address with silently-wrapping two's complement arithmetic, and
4638 the result value of the <tt>getelementptr</tt> may be outside the object
4639 pointed to by the base pointer. The result value may not necessarily be
4640 used to access memory though, even if it happens to point into allocated
4641 storage. See the <a href="#pointeraliasing">Pointer Aliasing Rules</a>
4642 section for more information.</p>
4643
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4644<p>The getelementptr instruction is often confusing. For some more insight into
4645 how it works, see <a href="GetElementPtr.html">the getelementptr FAQ</a>.</p>
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]4646
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]4647<h5>Example:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4648<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4649 <i>; yields [12 x i8]*:aptr</i>
Matthijs Kooijmane49d0bc2008-10-13 13:44:15 +0000[diff] [blame]4650 %aptr = getelementptr {i32, [12 x i8]}* %saptr, i64 0, i32 1
4651 <i>; yields i8*:vptr</i>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]4652 %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]4653 <i>; yields i8*:eptr</i>
4654 %eptr = getelementptr [12 x i8]* %aptr, i64 0, i32 1
Sanjiv Gupta9f805c22009-04-25 07:27:44 +0000[diff] [blame]4655 <i>; yields i32*:iptr</i>
Sanjiv Gupta16ffa802009-04-24 16:38:13 +0000[diff] [blame]4656 %iptr = getelementptr [10 x i32]* @arr, i16 0, i16 0
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4657</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4658
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]4659</div>
Reid Spencer47ce1792006-11-09 21:15:49 +0000[diff] [blame]4660
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4661</div>
4662
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4663<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4664<h3>
4665 <a name="convertops">Conversion Operations</a>
4666</h3>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4667
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4668<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4669
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]4670<p>The instructions in this category are the conversion instructions (casting)
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4671 which all take a single operand and a type. They perform various bit
4672 conversions on the operand.</p>
4673
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]4674<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4675<h4>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4676 <a name="i_trunc">'<tt>trunc .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4677</h4>
4678
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4679<div>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4680
4681<h5>Syntax:</h5>
4682<pre>
4683 &lt;result&gt; = trunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4684</pre>
4685
4686<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4687<p>The '<tt>trunc</tt>' instruction truncates its operand to the
4688 type <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4689
4690<h5>Arguments:</h5>
Nadav Rotem8c20ec52011-02-24 21:01:34 +0000[diff] [blame]4691<p>The '<tt>trunc</tt>' instruction takes a value to trunc, and a type to trunc it to.
4692 Both types must be of <a href="#t_integer">integer</a> types, or vectors
4693 of the same number of integers.
4694 The bit size of the <tt>value</tt> must be larger than
4695 the bit size of the destination type, <tt>ty2</tt>.
4696 Equal sized types are not allowed.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4697
4698<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4699<p>The '<tt>trunc</tt>' instruction truncates the high order bits
4700 in <tt>value</tt> and converts the remaining bits to <tt>ty2</tt>. Since the
4701 source size must be larger than the destination size, <tt>trunc</tt> cannot
4702 be a <i>no-op cast</i>. It will always truncate bits.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4703
4704<h5>Example:</h5>
4705<pre>
Nadav Rotem8c20ec52011-02-24 21:01:34 +0000[diff] [blame]4706 %X = trunc i32 257 to i8 <i>; yields i8:1</i>
4707 %Y = trunc i32 123 to i1 <i>; yields i1:true</i>
4708 %Z = trunc i32 122 to i1 <i>; yields i1:false</i>
4709 %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]4710</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4711
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4712</div>
4713
4714<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4715<h4>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4716 <a name="i_zext">'<tt>zext .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4717</h4>
4718
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4719<div>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4720
4721<h5>Syntax:</h5>
4722<pre>
4723 &lt;result&gt; = zext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4724</pre>
4725
4726<h5>Overview:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4727<p>The '<tt>zext</tt>' instruction zero extends its operand to type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4728 <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4729
4730
4731<h5>Arguments:</h5>
Nadav Rotemed9b9342011-02-20 12:37:50 +0000[diff] [blame]4732<p>The '<tt>zext</tt>' instruction takes a value to cast, and a type to cast it to.
4733 Both types must be of <a href="#t_integer">integer</a> types, or vectors
4734 of the same number of integers.
4735 The bit size of the <tt>value</tt> must be smaller than
4736 the bit size of the destination type,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4737 <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4738
4739<h5>Semantics:</h5>
4740<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]4741 bits until it reaches the size of the destination type, <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4742
Reid Spencerb5929522007-01-12 15:46:11 +0000[diff] [blame]4743<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]4744
4745<h5>Example:</h5>
4746<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4747 %X = zext i32 257 to i64 <i>; yields i64:257</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]4748 %Y = zext i1 true to i32 <i>; yields i32:1</i>
Nadav Rotemed9b9342011-02-20 12:37:50 +0000[diff] [blame]4749 %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]4750</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4751
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4752</div>
4753
4754<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4755<h4>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4756 <a name="i_sext">'<tt>sext .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4757</h4>
4758
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4759<div>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4760
4761<h5>Syntax:</h5>
4762<pre>
4763 &lt;result&gt; = sext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4764</pre>
4765
4766<h5>Overview:</h5>
4767<p>The '<tt>sext</tt>' sign extends <tt>value</tt> to the type <tt>ty2</tt>.</p>
4768
4769<h5>Arguments:</h5>
Nadav Rotem8c20ec52011-02-24 21:01:34 +0000[diff] [blame]4770<p>The '<tt>sext</tt>' instruction takes a value to cast, and a type to cast it to.
4771 Both types must be of <a href="#t_integer">integer</a> types, or vectors
4772 of the same number of integers.
4773 The bit size of the <tt>value</tt> must be smaller than
4774 the bit size of the destination type,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4775 <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4776
4777<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4778<p>The '<tt>sext</tt>' instruction performs a sign extension by copying the sign
4779 bit (highest order bit) of the <tt>value</tt> until it reaches the bit size
4780 of the type <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4781
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]4782<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]4783
4784<h5>Example:</h5>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4785<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4786 %X = sext i8 -1 to i16 <i>; yields i16 :65535</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]4787 %Y = sext i1 true to i32 <i>; yields i32:-1</i>
Nadav Rotem8c20ec52011-02-24 21:01:34 +0000[diff] [blame]4788 %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]4789</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4790
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4791</div>
4792
4793<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4794<h4>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]4795 <a name="i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4796</h4>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]4797
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4798<div>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]4799
4800<h5>Syntax:</h5>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]4801<pre>
4802 &lt;result&gt; = fptrunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4803</pre>
4804
4805<h5>Overview:</h5>
4806<p>The '<tt>fptrunc</tt>' instruction truncates <tt>value</tt> to type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4807 <tt>ty2</tt>.</p>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]4808
4809<h5>Arguments:</h5>
4810<p>The '<tt>fptrunc</tt>' instruction takes a <a href="#t_floating">floating
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4811 point</a> value to cast and a <a href="#t_floating">floating point</a> type
4812 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]4813 <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]4814 <i>no-op cast</i>.</p>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]4815
4816<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4817<p>The '<tt>fptrunc</tt>' instruction truncates a <tt>value</tt> from a larger
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4818 <a href="#t_floating">floating point</a> type to a smaller
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4819 <a href="#t_floating">floating point</a> type. If the value cannot fit
4820 within the destination type, <tt>ty2</tt>, then the results are
4821 undefined.</p>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]4822
4823<h5>Example:</h5>
4824<pre>
4825 %X = fptrunc double 123.0 to float <i>; yields float:123.0</i>
4826 %Y = fptrunc double 1.0E+300 to float <i>; yields undefined</i>
4827</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4828
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]4829</div>
4830
4831<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4832<h4>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4833 <a name="i_fpext">'<tt>fpext .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4834</h4>
4835
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4836<div>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4837
4838<h5>Syntax:</h5>
4839<pre>
4840 &lt;result&gt; = fpext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4841</pre>
4842
4843<h5>Overview:</h5>
4844<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]4845 floating point value.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4846
4847<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4848<p>The '<tt>fpext</tt>' instruction takes a
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4849 <a href="#t_floating">floating point</a> <tt>value</tt> to cast, and
4850 a <a href="#t_floating">floating point</a> type to cast it to. The source
4851 type must be smaller than the destination type.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4852
4853<h5>Semantics:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]4854<p>The '<tt>fpext</tt>' instruction extends the <tt>value</tt> from a smaller
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4855 <a href="#t_floating">floating point</a> type to a larger
4856 <a href="#t_floating">floating point</a> type. The <tt>fpext</tt> cannot be
4857 used to make a <i>no-op cast</i> because it always changes bits. Use
4858 <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]4859
4860<h5>Example:</h5>
4861<pre>
Nick Lewycky5bb3ece2011-03-31 18:20:19 +0000[diff] [blame]4862 %X = fpext float 3.125 to double <i>; yields double:3.125000e+00</i>
4863 %Y = fpext double %X to fp128 <i>; yields fp128:0xL00000000000000004000900000000000</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4864</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4865
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4866</div>
4867
4868<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4869<h4>
Reid Spencer24d6da52007-01-21 00:29:26 +0000[diff] [blame]4870 <a name="i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4871</h4>
4872
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4873<div>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4874
4875<h5>Syntax:</h5>
4876<pre>
Reid Spencer1539a1c2007-07-31 14:40:14 +0000[diff] [blame]4877 &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]4878</pre>
4879
4880<h5>Overview:</h5>
Reid Spencer1539a1c2007-07-31 14:40:14 +0000[diff] [blame]4881<p>The '<tt>fptoui</tt>' converts a floating point <tt>value</tt> to its
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4882 unsigned integer equivalent of type <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4883
4884<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4885<p>The '<tt>fptoui</tt>' instruction takes a value to cast, which must be a
4886 scalar or vector <a href="#t_floating">floating point</a> value, and a type
4887 to cast it to <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a>
4888 type. If <tt>ty</tt> is a vector floating point type, <tt>ty2</tt> must be a
4889 vector integer type with the same number of elements as <tt>ty</tt></p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4890
4891<h5>Semantics:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4892<p>The '<tt>fptoui</tt>' instruction converts its
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4893 <a href="#t_floating">floating point</a> operand into the nearest (rounding
4894 towards zero) unsigned integer value. If the value cannot fit
4895 in <tt>ty2</tt>, the results are undefined.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4896
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4897<h5>Example:</h5>
4898<pre>
Reid Spencer1539a1c2007-07-31 14:40:14 +0000[diff] [blame]4899 %X = fptoui double 123.0 to i32 <i>; yields i32:123</i>
Chris Lattner88519042007-09-22 03:17:52 +0000[diff] [blame]4900 %Y = fptoui float 1.0E+300 to i1 <i>; yields undefined:1</i>
Gabor Greif6a292012009-10-28 09:21:30 +0000[diff] [blame]4901 %Z = fptoui float 1.04E+17 to i8 <i>; yields undefined:1</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4902</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4903
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4904</div>
4905
4906<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4907<h4>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]4908 <a name="i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4909</h4>
4910
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4911<div>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4912
4913<h5>Syntax:</h5>
4914<pre>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]4915 &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]4916</pre>
4917
4918<h5>Overview:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4919<p>The '<tt>fptosi</tt>' instruction converts
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4920 <a href="#t_floating">floating point</a> <tt>value</tt> to
4921 type <tt>ty2</tt>.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]4922
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]4923<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4924<p>The '<tt>fptosi</tt>' instruction takes a value to cast, which must be a
4925 scalar or vector <a href="#t_floating">floating point</a> value, and a type
4926 to cast it to <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a>
4927 type. If <tt>ty</tt> is a vector floating point type, <tt>ty2</tt> must be a
4928 vector integer type with the same number of elements as <tt>ty</tt></p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]4929
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]4930<h5>Semantics:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4931<p>The '<tt>fptosi</tt>' instruction converts its
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4932 <a href="#t_floating">floating point</a> operand into the nearest (rounding
4933 towards zero) signed integer value. If the value cannot fit in <tt>ty2</tt>,
4934 the results are undefined.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]4935
Chris Lattner33ba0d92001-07-09 00:26:23 +0000[diff] [blame]4936<h5>Example:</h5>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]4937<pre>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]4938 %X = fptosi double -123.0 to i32 <i>; yields i32:-123</i>
Chris Lattner88519042007-09-22 03:17:52 +0000[diff] [blame]4939 %Y = fptosi float 1.0E-247 to i1 <i>; yields undefined:1</i>
Gabor Greif6a292012009-10-28 09:21:30 +0000[diff] [blame]4940 %Z = fptosi float 1.04E+17 to i8 <i>; yields undefined:1</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4941</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4942
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4943</div>
4944
4945<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4946<h4>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]4947 <a name="i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4948</h4>
4949
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4950<div>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4951
4952<h5>Syntax:</h5>
4953<pre>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]4954 &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]4955</pre>
4956
4957<h5>Overview:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]4958<p>The '<tt>uitofp</tt>' instruction regards <tt>value</tt> as an unsigned
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4959 integer and converts that value to the <tt>ty2</tt> type.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4960
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4961<h5>Arguments:</h5>
Nate Begemanb348d182007-11-17 03:58:34 +0000[diff] [blame]4962<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]4963 scalar or vector <a href="#t_integer">integer</a> value, and a type to cast
4964 it to <tt>ty2</tt>, which must be an <a href="#t_floating">floating point</a>
4965 type. If <tt>ty</tt> is a vector integer type, <tt>ty2</tt> must be a vector
4966 floating point type with the same number of elements as <tt>ty</tt></p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4967
4968<h5>Semantics:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]4969<p>The '<tt>uitofp</tt>' instruction interprets its operand as an unsigned
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4970 integer quantity and converts it to the corresponding floating point
4971 value. If the value cannot fit in the floating point value, the results are
4972 undefined.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4973
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4974<h5>Example:</h5>
4975<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4976 %X = uitofp i32 257 to float <i>; yields float:257.0</i>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]4977 %Y = uitofp i8 -1 to double <i>; yields double:255.0</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4978</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4979
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4980</div>
4981
4982<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4983<h4>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]4984 <a name="i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4985</h4>
4986
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4987<div>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4988
4989<h5>Syntax:</h5>
4990<pre>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]4991 &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]4992</pre>
4993
4994<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4995<p>The '<tt>sitofp</tt>' instruction regards <tt>value</tt> as a signed integer
4996 and converts that value to the <tt>ty2</tt> type.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4997
4998<h5>Arguments:</h5>
Nate Begemanb348d182007-11-17 03:58:34 +0000[diff] [blame]4999<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]5000 scalar or vector <a href="#t_integer">integer</a> value, and a type to cast
5001 it to <tt>ty2</tt>, which must be an <a href="#t_floating">floating point</a>
5002 type. If <tt>ty</tt> is a vector integer type, <tt>ty2</tt> must be a vector
5003 floating point type with the same number of elements as <tt>ty</tt></p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5004
5005<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5006<p>The '<tt>sitofp</tt>' instruction interprets its operand as a signed integer
5007 quantity and converts it to the corresponding floating point value. If the
5008 value cannot fit in the floating point value, the results are undefined.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5009
5010<h5>Example:</h5>
5011<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5012 %X = sitofp i32 257 to float <i>; yields float:257.0</i>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5013 %Y = sitofp i8 -1 to double <i>; yields double:-1.0</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5014</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5015
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5016</div>
5017
5018<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5019<h4>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5020 <a name="i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5021</h4>
5022
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5023<div>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5024
5025<h5>Syntax:</h5>
5026<pre>
5027 &lt;result&gt; = ptrtoint &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
5028</pre>
5029
5030<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5031<p>The '<tt>ptrtoint</tt>' instruction converts the pointer <tt>value</tt> to
5032 the integer type <tt>ty2</tt>.</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5033
5034<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5035<p>The '<tt>ptrtoint</tt>' instruction takes a <tt>value</tt> to cast, which
5036 must be a <a href="#t_pointer">pointer</a> value, and a type to cast it to
5037 <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]5038
5039<h5>Semantics:</h5>
5040<p>The '<tt>ptrtoint</tt>' instruction converts <tt>value</tt> to integer type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5041 <tt>ty2</tt> by interpreting the pointer value as an integer and either
5042 truncating or zero extending that value to the size of the integer type. If
5043 <tt>value</tt> is smaller than <tt>ty2</tt> then a zero extension is done. If
5044 <tt>value</tt> is larger than <tt>ty2</tt> then a truncation is done. If they
5045 are the same size, then nothing is done (<i>no-op cast</i>) other than a type
5046 change.</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5047
5048<h5>Example:</h5>
5049<pre>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5050 %X = ptrtoint i32* %X to i8 <i>; yields truncation on 32-bit architecture</i>
5051 %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]5052</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5053
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5054</div>
5055
5056<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5057<h4>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5058 <a name="i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5059</h4>
5060
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5061<div>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5062
5063<h5>Syntax:</h5>
5064<pre>
5065 &lt;result&gt; = inttoptr &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
5066</pre>
5067
5068<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5069<p>The '<tt>inttoptr</tt>' instruction converts an integer <tt>value</tt> to a
5070 pointer type, <tt>ty2</tt>.</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5071
5072<h5>Arguments:</h5>
Duncan Sands8036ca42007-03-30 12:22:09 +0000[diff] [blame]5073<p>The '<tt>inttoptr</tt>' instruction takes an <a href="#t_integer">integer</a>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5074 value to cast, and a type to cast it to, which must be a
5075 <a href="#t_pointer">pointer</a> type.</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5076
5077<h5>Semantics:</h5>
5078<p>The '<tt>inttoptr</tt>' instruction converts <tt>value</tt> to type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5079 <tt>ty2</tt> by applying either a zero extension or a truncation depending on
5080 the size of the integer <tt>value</tt>. If <tt>value</tt> is larger than the
5081 size of a pointer then a truncation is done. If <tt>value</tt> is smaller
5082 than the size of a pointer then a zero extension is done. If they are the
5083 same size, nothing is done (<i>no-op cast</i>).</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5084
5085<h5>Example:</h5>
5086<pre>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5087 %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]5088 %Y = inttoptr i32 255 to i32* <i>; yields no-op on 32-bit architecture</i>
5089 %Z = inttoptr i64 0 to i32* <i>; yields truncation on 32-bit architecture</i>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5090</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5091
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5092</div>
5093
5094<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5095<h4>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]5096 <a name="i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5097</h4>
5098
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5099<div>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5100
5101<h5>Syntax:</h5>
5102<pre>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]5103 &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]5104</pre>
5105
5106<h5>Overview:</h5>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]5107<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5108 <tt>ty2</tt> without changing any bits.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5109
5110<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5111<p>The '<tt>bitcast</tt>' instruction takes a value to cast, which must be a
5112 non-aggregate first class value, and a type to cast it to, which must also be
5113 a non-aggregate <a href="#t_firstclass">first class</a> type. The bit sizes
5114 of <tt>value</tt> and the destination type, <tt>ty2</tt>, must be
5115 identical. If the source type is a pointer, the destination type must also be
5116 a pointer. This instruction supports bitwise conversion of vectors to
5117 integers and to vectors of other types (as long as they have the same
5118 size).</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5119
5120<h5>Semantics:</h5>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]5121<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5122 <tt>ty2</tt>. It is always a <i>no-op cast</i> because no bits change with
5123 this conversion. The conversion is done as if the <tt>value</tt> had been
5124 stored to memory and read back as type <tt>ty2</tt>. Pointer types may only
5125 be converted to other pointer types with this instruction. To convert
5126 pointers to other types, use the <a href="#i_inttoptr">inttoptr</a> or
5127 <a href="#i_ptrtoint">ptrtoint</a> instructions first.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5128
5129<h5>Example:</h5>
5130<pre>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5131 %X = bitcast i8 255 to i8 <i>; yields i8 :-1</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5132 %Y = bitcast i32* %x to sint* <i>; yields sint*:%x</i>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5133 %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]5134</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5135
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5136</div>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5137
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5138</div>
5139
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5140<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5141<h3>
5142 <a name="otherops">Other Operations</a>
5143</h3>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5144
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5145<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5146
5147<p>The instructions in this category are the "miscellaneous" instructions, which
5148 defy better classification.</p>
5149
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5150<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5151<h4>
5152 <a name="i_icmp">'<tt>icmp</tt>' Instruction</a>
5153</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5154
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5155<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5156
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5157<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5158<pre>
5159 &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]5160</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5161
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5162<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5163<p>The '<tt>icmp</tt>' instruction returns a boolean value or a vector of
5164 boolean values based on comparison of its two integer, integer vector, or
5165 pointer operands.</p>
5166
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5167<h5>Arguments:</h5>
5168<p>The '<tt>icmp</tt>' instruction takes three operands. The first operand is
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5169 the condition code indicating the kind of comparison to perform. It is not a
5170 value, just a keyword. The possible condition code are:</p>
5171
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5172<ol>
5173 <li><tt>eq</tt>: equal</li>
5174 <li><tt>ne</tt>: not equal </li>
5175 <li><tt>ugt</tt>: unsigned greater than</li>
5176 <li><tt>uge</tt>: unsigned greater or equal</li>
5177 <li><tt>ult</tt>: unsigned less than</li>
5178 <li><tt>ule</tt>: unsigned less or equal</li>
5179 <li><tt>sgt</tt>: signed greater than</li>
5180 <li><tt>sge</tt>: signed greater or equal</li>
5181 <li><tt>slt</tt>: signed less than</li>
5182 <li><tt>sle</tt>: signed less or equal</li>
5183</ol>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5184
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]5185<p>The remaining two arguments must be <a href="#t_integer">integer</a> or
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5186 <a href="#t_pointer">pointer</a> or integer <a href="#t_vector">vector</a>
5187 typed. They must also be identical types.</p>
5188
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5189<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5190<p>The '<tt>icmp</tt>' compares <tt>op1</tt> and <tt>op2</tt> according to the
5191 condition code given as <tt>cond</tt>. The comparison performed always yields
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]5192 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]5193 result, as follows:</p>
5194
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5195<ol>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5196 <li><tt>eq</tt>: yields <tt>true</tt> if the operands are equal,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5197 <tt>false</tt> otherwise. No sign interpretation is necessary or
5198 performed.</li>
5199
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5200 <li><tt>ne</tt>: yields <tt>true</tt> if the operands are unequal,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5201 <tt>false</tt> otherwise. No sign interpretation is necessary or
5202 performed.</li>
5203
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5204 <li><tt>ugt</tt>: interprets the operands as unsigned values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5205 <tt>true</tt> if <tt>op1</tt> is greater than <tt>op2</tt>.</li>
5206
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5207 <li><tt>uge</tt>: interprets the operands as unsigned values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5208 <tt>true</tt> if <tt>op1</tt> is greater than or equal
5209 to <tt>op2</tt>.</li>
5210
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5211 <li><tt>ult</tt>: interprets the operands as unsigned values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5212 <tt>true</tt> if <tt>op1</tt> is less than <tt>op2</tt>.</li>
5213
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5214 <li><tt>ule</tt>: interprets the operands as unsigned values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5215 <tt>true</tt> if <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
5216
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5217 <li><tt>sgt</tt>: interprets the operands as signed values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5218 <tt>true</tt> if <tt>op1</tt> is greater than <tt>op2</tt>.</li>
5219
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5220 <li><tt>sge</tt>: interprets the operands as signed values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5221 <tt>true</tt> if <tt>op1</tt> is greater than or equal
5222 to <tt>op2</tt>.</li>
5223
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5224 <li><tt>slt</tt>: interprets the operands as signed values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5225 <tt>true</tt> if <tt>op1</tt> is less than <tt>op2</tt>.</li>
5226
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5227 <li><tt>sle</tt>: interprets the operands as signed values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5228 <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]5229</ol>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5230
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5231<p>If the operands are <a href="#t_pointer">pointer</a> typed, the pointer
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5232 values are compared as if they were integers.</p>
5233
5234<p>If the operands are integer vectors, then they are compared element by
5235 element. The result is an <tt>i1</tt> vector with the same number of elements
5236 as the values being compared. Otherwise, the result is an <tt>i1</tt>.</p>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5237
5238<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5239<pre>
5240 &lt;result&gt; = icmp eq i32 4, 5 <i>; yields: result=false</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5241 &lt;result&gt; = icmp ne float* %X, %X <i>; yields: result=false</i>
5242 &lt;result&gt; = icmp ult i16 4, 5 <i>; yields: result=true</i>
5243 &lt;result&gt; = icmp sgt i16 4, 5 <i>; yields: result=false</i>
5244 &lt;result&gt; = icmp ule i16 -4, 5 <i>; yields: result=false</i>
5245 &lt;result&gt; = icmp sge i16 4, 5 <i>; yields: result=false</i>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5246</pre>
Dan Gohmanc766f722009-01-22 01:39:38 +0000[diff] [blame]5247
5248<p>Note that the code generator does not yet support vector types with
5249 the <tt>icmp</tt> instruction.</p>
5250
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5251</div>
5252
5253<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5254<h4>
5255 <a name="i_fcmp">'<tt>fcmp</tt>' Instruction</a>
5256</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5257
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5258<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5259
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5260<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5261<pre>
5262 &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]5263</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5264
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5265<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5266<p>The '<tt>fcmp</tt>' instruction returns a boolean value or vector of boolean
5267 values based on comparison of its operands.</p>
5268
5269<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]5270(<a href="#t_integer"><tt>i1</tt></a>).</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5271
5272<p>If the operands are floating point vectors, then the result type is a vector
5273 of boolean with the same number of elements as the operands being
5274 compared.</p>
5275
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5276<h5>Arguments:</h5>
5277<p>The '<tt>fcmp</tt>' instruction takes three operands. The first operand is
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5278 the condition code indicating the kind of comparison to perform. It is not a
5279 value, just a keyword. The possible condition code are:</p>
5280
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5281<ol>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]5282 <li><tt>false</tt>: no comparison, always returns false</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5283 <li><tt>oeq</tt>: ordered and equal</li>
5284 <li><tt>ogt</tt>: ordered and greater than </li>
5285 <li><tt>oge</tt>: ordered and greater than or equal</li>
5286 <li><tt>olt</tt>: ordered and less than </li>
5287 <li><tt>ole</tt>: ordered and less than or equal</li>
5288 <li><tt>one</tt>: ordered and not equal</li>
5289 <li><tt>ord</tt>: ordered (no nans)</li>
5290 <li><tt>ueq</tt>: unordered or equal</li>
5291 <li><tt>ugt</tt>: unordered or greater than </li>
5292 <li><tt>uge</tt>: unordered or greater than or equal</li>
5293 <li><tt>ult</tt>: unordered or less than </li>
5294 <li><tt>ule</tt>: unordered or less than or equal</li>
5295 <li><tt>une</tt>: unordered or not equal</li>
5296 <li><tt>uno</tt>: unordered (either nans)</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]5297 <li><tt>true</tt>: no comparison, always returns true</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5298</ol>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5299
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5300<p><i>Ordered</i> means that neither operand is a QNAN while
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5301 <i>unordered</i> means that either operand may be a QNAN.</p>
5302
5303<p>Each of <tt>val1</tt> and <tt>val2</tt> arguments must be either
5304 a <a href="#t_floating">floating point</a> type or
5305 a <a href="#t_vector">vector</a> of floating point type. They must have
5306 identical types.</p>
5307
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5308<h5>Semantics:</h5>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]5309<p>The '<tt>fcmp</tt>' instruction compares <tt>op1</tt> and <tt>op2</tt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5310 according to the condition code given as <tt>cond</tt>. If the operands are
5311 vectors, then the vectors are compared element by element. Each comparison
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]5312 performed always yields an <a href="#t_integer">i1</a> result, as
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5313 follows:</p>
5314
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5315<ol>
5316 <li><tt>false</tt>: always yields <tt>false</tt>, regardless of operands.</li>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5317
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5318 <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]5319 <tt>op1</tt> is equal to <tt>op2</tt>.</li>
5320
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]5321 <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]5322 <tt>op1</tt> is greater than <tt>op2</tt>.</li>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5323
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5324 <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]5325 <tt>op1</tt> is greater than or equal to <tt>op2</tt>.</li>
5326
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5327 <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]5328 <tt>op1</tt> is less than <tt>op2</tt>.</li>
5329
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5330 <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]5331 <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
5332
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5333 <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]5334 <tt>op1</tt> is not equal to <tt>op2</tt>.</li>
5335
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]5336 <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]5337
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5338 <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]5339 <tt>op1</tt> is equal to <tt>op2</tt>.</li>
5340
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5341 <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]5342 <tt>op1</tt> is greater than <tt>op2</tt>.</li>
5343
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5344 <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]5345 <tt>op1</tt> is greater than or equal to <tt>op2</tt>.</li>
5346
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5347 <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]5348 <tt>op1</tt> is less than <tt>op2</tt>.</li>
5349
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5350 <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]5351 <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
5352
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5353 <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]5354 <tt>op1</tt> is not equal to <tt>op2</tt>.</li>
5355
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]5356 <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]5357
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5358 <li><tt>true</tt>: always yields <tt>true</tt>, regardless of operands.</li>
5359</ol>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5360
5361<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5362<pre>
5363 &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]5364 &lt;result&gt; = fcmp one float 4.0, 5.0 <i>; yields: result=true</i>
5365 &lt;result&gt; = fcmp olt float 4.0, 5.0 <i>; yields: result=true</i>
5366 &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]5367</pre>
Dan Gohmanc766f722009-01-22 01:39:38 +0000[diff] [blame]5368
5369<p>Note that the code generator does not yet support vector types with
5370 the <tt>fcmp</tt> instruction.</p>
5371
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5372</div>
5373
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5374<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5375<h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]5376 <a name="i_phi">'<tt>phi</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5377</h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]5378
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5379<div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]5380
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5381<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5382<pre>
5383 &lt;result&gt; = phi &lt;ty&gt; [ &lt;val0&gt;, &lt;label0&gt;], ...
5384</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]5385
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5386<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5387<p>The '<tt>phi</tt>' instruction is used to implement the &#966; node in the
5388 SSA graph representing the function.</p>
5389
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5390<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5391<p>The type of the incoming values is specified with the first type field. After
5392 this, the '<tt>phi</tt>' instruction takes a list of pairs as arguments, with
5393 one pair for each predecessor basic block of the current block. Only values
5394 of <a href="#t_firstclass">first class</a> type may be used as the value
5395 arguments to the PHI node. Only labels may be used as the label
5396 arguments.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]5397
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5398<p>There must be no non-phi instructions between the start of a basic block and
5399 the PHI instructions: i.e. PHI instructions must be first in a basic
5400 block.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]5401
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5402<p>For the purposes of the SSA form, the use of each incoming value is deemed to
5403 occur on the edge from the corresponding predecessor block to the current
5404 block (but after any definition of an '<tt>invoke</tt>' instruction's return
5405 value on the same edge).</p>
Jay Foadd2449092009-06-03 10:20:10 +0000[diff] [blame]5406
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5407<h5>Semantics:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5408<p>At runtime, the '<tt>phi</tt>' instruction logically takes on the value
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5409 specified by the pair corresponding to the predecessor basic block that
5410 executed just prior to the current block.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]5411
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5412<h5>Example:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]5413<pre>
5414Loop: ; Infinite loop that counts from 0 on up...
5415 %indvar = phi i32 [ 0, %LoopHeader ], [ %nextindvar, %Loop ]
5416 %nextindvar = add i32 %indvar, 1
5417 br label %Loop
5418</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5419
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5420</div>
5421
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5422<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5423<h4>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5424 <a name="i_select">'<tt>select</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5425</h4>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5426
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5427<div>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5428
5429<h5>Syntax:</h5>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5430<pre>
Dan Gohmanf72fb672008-09-09 01:02:47 +0000[diff] [blame]5431 &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>
5432
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5433 <i>selty</i> is either i1 or {&lt;N x i1&gt;}
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5434</pre>
5435
5436<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5437<p>The '<tt>select</tt>' instruction is used to choose one value based on a
5438 condition, without branching.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5439
5440
5441<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5442<p>The '<tt>select</tt>' instruction requires an 'i1' value or a vector of 'i1'
5443 values indicating the condition, and two values of the
5444 same <a href="#t_firstclass">first class</a> type. If the val1/val2 are
5445 vectors and the condition is a scalar, then entire vectors are selected, not
5446 individual elements.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5447
5448<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5449<p>If the condition is an i1 and it evaluates to 1, the instruction returns the
5450 first value argument; otherwise, it returns the second value argument.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5451
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5452<p>If the condition is a vector of i1, then the value arguments must be vectors
5453 of the same size, and the selection is done element by element.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5454
5455<h5>Example:</h5>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5456<pre>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]5457 %X = select i1 true, i8 17, i8 42 <i>; yields i8:17</i>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5458</pre>
Dan Gohmanc766f722009-01-22 01:39:38 +0000[diff] [blame]5459
5460<p>Note that the code generator does not yet support conditions
5461 with vector type.</p>
5462
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5463</div>
5464
Robert Bocchino05ccd702006-01-15 20:48:27 +0000[diff] [blame]5465<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5466<h4>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5467 <a name="i_call">'<tt>call</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5468</h4>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5469
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5470<div>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5471
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]5472<h5>Syntax:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5473<pre>
Devang Patel307e8ab2008-10-07 17:48:33 +0000[diff] [blame]5474 &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]5475</pre>
5476
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]5477<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5478<p>The '<tt>call</tt>' instruction represents a simple function call.</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5479
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]5480<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5481<p>This instruction requires several arguments:</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5482
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]5483<ol>
Jeffrey Yasskin95fa80a2010-01-09 19:44:16 +0000[diff] [blame]5484 <li>The optional "tail" marker indicates that the callee function does not
5485 access any allocas or varargs in the caller. Note that calls may be
5486 marked "tail" even if they do not occur before
5487 a <a href="#i_ret"><tt>ret</tt></a> instruction. If the "tail" marker is
5488 present, the function call is eligible for tail call optimization,
5489 but <a href="CodeGenerator.html#tailcallopt">might not in fact be
Evan Chengdc444e92010-03-08 21:05:02 +0000[diff] [blame]5490 optimized into a jump</a>. The code generator may optimize calls marked
5491 "tail" with either 1) automatic <a href="CodeGenerator.html#sibcallopt">
5492 sibling call optimization</a> when the caller and callee have
5493 matching signatures, or 2) forced tail call optimization when the
5494 following extra requirements are met:
Jeffrey Yasskin95fa80a2010-01-09 19:44:16 +0000[diff] [blame]5495 <ul>
5496 <li>Caller and callee both have the calling
5497 convention <tt>fastcc</tt>.</li>
5498 <li>The call is in tail position (ret immediately follows call and ret
5499 uses value of call or is void).</li>
5500 <li>Option <tt>-tailcallopt</tt> is enabled,
Dan Gohmanfbbee8d2010-03-02 01:08:11 +0000[diff] [blame]5501 or <code>llvm::GuaranteedTailCallOpt</code> is <code>true</code>.</li>
Jeffrey Yasskin95fa80a2010-01-09 19:44:16 +0000[diff] [blame]5502 <li><a href="CodeGenerator.html#tailcallopt">Platform specific
5503 constraints are met.</a></li>
5504 </ul>
5505 </li>
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]5506
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5507 <li>The optional "cconv" marker indicates which <a href="#callingconv">calling
5508 convention</a> the call should use. If none is specified, the call
Jeffrey Yasskin95fa80a2010-01-09 19:44:16 +0000[diff] [blame]5509 defaults to using C calling conventions. The calling convention of the
5510 call must match the calling convention of the target function, or else the
5511 behavior is undefined.</li>
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]5512
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5513 <li>The optional <a href="#paramattrs">Parameter Attributes</a> list for
5514 return values. Only '<tt>zeroext</tt>', '<tt>signext</tt>', and
5515 '<tt>inreg</tt>' attributes are valid here.</li>
5516
5517 <li>'<tt>ty</tt>': the type of the call instruction itself which is also the
5518 type of the return value. Functions that return no value are marked
5519 <tt><a href="#t_void">void</a></tt>.</li>
5520
5521 <li>'<tt>fnty</tt>': shall be the signature of the pointer to function value
5522 being invoked. The argument types must match the types implied by this
5523 signature. This type can be omitted if the function is not varargs and if
5524 the function type does not return a pointer to a function.</li>
5525
5526 <li>'<tt>fnptrval</tt>': An LLVM value containing a pointer to a function to
5527 be invoked. In most cases, this is a direct function invocation, but
5528 indirect <tt>call</tt>s are just as possible, calling an arbitrary pointer
5529 to function value.</li>
5530
5531 <li>'<tt>function args</tt>': argument list whose types match the function
Chris Lattner0724fbd2010-03-02 06:36:51 +0000[diff] [blame]5532 signature argument types and parameter attributes. All arguments must be
5533 of <a href="#t_firstclass">first class</a> type. If the function
5534 signature indicates the function accepts a variable number of arguments,
5535 the extra arguments can be specified.</li>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5536
5537 <li>The optional <a href="#fnattrs">function attributes</a> list. Only
5538 '<tt>noreturn</tt>', '<tt>nounwind</tt>', '<tt>readonly</tt>' and
5539 '<tt>readnone</tt>' attributes are valid here.</li>
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]5540</ol>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5541
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]5542<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5543<p>The '<tt>call</tt>' instruction is used to cause control flow to transfer to
5544 a specified function, with its incoming arguments bound to the specified
5545 values. Upon a '<tt><a href="#i_ret">ret</a></tt>' instruction in the called
5546 function, control flow continues with the instruction after the function
5547 call, and the return value of the function is bound to the result
5548 argument.</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5549
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]5550<h5>Example:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5551<pre>
Nick Lewyckydb7e3c92007-09-08 13:57:50 +0000[diff] [blame]5552 %retval = call i32 @test(i32 %argc)
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]5553 call i32 (i8*, ...)* @printf(i8* %msg, i32 12, i8 42) <i>; yields i32</i>
Chris Lattner772fccf2008-03-21 17:24:17 +0000[diff] [blame]5554 %X = tail call i32 @foo() <i>; yields i32</i>
5555 %Y = tail call <a href="#callingconv">fastcc</a> i32 @foo() <i>; yields i32</i>
5556 call void %foo(i8 97 signext)
Devang Patelc3fc6df2008-03-10 20:49:15 +0000[diff] [blame]5557
5558 %struct.A = type { i32, i8 }
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]5559 %r = call %struct.A @foo() <i>; yields { 32, i8 }</i>
Dan Gohmanb1e6b962008-10-04 19:00:07 +0000[diff] [blame]5560 %gr = extractvalue %struct.A %r, 0 <i>; yields i32</i>
5561 %gr1 = extractvalue %struct.A %r, 1 <i>; yields i8</i>
Chris Lattner85a350f2008-10-08 06:26:11 +0000[diff] [blame]5562 %Z = call void @foo() noreturn <i>; indicates that %foo never returns normally</i>
Matthijs Kooijmancb73d192008-10-07 10:03:45 +0000[diff] [blame]5563 %ZZ = call zeroext i32 @bar() <i>; Return value is %zero extended</i>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]5564</pre>
5565
Dale Johannesen07de8d12009-09-24 18:38:21 +0000[diff] [blame]5566<p>llvm treats calls to some functions with names and arguments that match the
Dale Johannesen9f8380b2009-09-25 17:04:42 +0000[diff] [blame]5567standard C99 library as being the C99 library functions, and may perform
5568optimizations or generate code for them under that assumption. This is
5569something we'd like to change in the future to provide better support for
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]5570freestanding environments and non-C-based languages.</p>
Dale Johannesen07de8d12009-09-24 18:38:21 +0000[diff] [blame]5571
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5572</div>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5573
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5574<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5575<h4>
Chris Lattnerfb6977d2006-01-13 23:26:01 +0000[diff] [blame]5576 <a name="i_va_arg">'<tt>va_arg</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5577</h4>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5578
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5579<div>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5580
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]5581<h5>Syntax:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5582<pre>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]5583 &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]5584</pre>
5585
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]5586<h5>Overview:</h5>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]5587<p>The '<tt>va_arg</tt>' instruction is used to access arguments passed through
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5588 the "variable argument" area of a function call. It is used to implement the
5589 <tt>va_arg</tt> macro in C.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5590
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]5591<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5592<p>This instruction takes a <tt>va_list*</tt> value and the type of the
5593 argument. It returns a value of the specified argument type and increments
5594 the <tt>va_list</tt> to point to the next argument. The actual type
5595 of <tt>va_list</tt> is target specific.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5596
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]5597<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5598<p>The '<tt>va_arg</tt>' instruction loads an argument of the specified type
5599 from the specified <tt>va_list</tt> and causes the <tt>va_list</tt> to point
5600 to the next argument. For more information, see the variable argument
5601 handling <a href="#int_varargs">Intrinsic Functions</a>.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5602
5603<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]5604 take a variable number of arguments, for example, the <tt>vfprintf</tt>
5605 function.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5606
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5607<p><tt>va_arg</tt> is an LLVM instruction instead of
5608 an <a href="#intrinsics">intrinsic function</a> because it takes a type as an
5609 argument.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5610
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]5611<h5>Example:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]5612<p>See the <a href="#int_varargs">variable argument processing</a> section.</p>
5613
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5614<p>Note that the code generator does not yet fully support va_arg on many
5615 targets. Also, it does not currently support va_arg with aggregate types on
5616 any target.</p>
Dan Gohmanf3e60bd2009-01-12 23:12:39 +0000[diff] [blame]5617
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5618</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5619
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5620</div>
5621
5622</div>
5623
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5624<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5625<h2><a name="intrinsics">Intrinsic Functions</a></h2>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]5626<!-- *********************************************************************** -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5627
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5628<div>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5629
5630<p>LLVM supports the notion of an "intrinsic function". These functions have
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5631 well known names and semantics and are required to follow certain
5632 restrictions. Overall, these intrinsics represent an extension mechanism for
5633 the LLVM language that does not require changing all of the transformations
5634 in LLVM when adding to the language (or the bitcode reader/writer, the
5635 parser, etc...).</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5636
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]5637<p>Intrinsic function names must all start with an "<tt>llvm.</tt>" prefix. This
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5638 prefix is reserved in LLVM for intrinsic names; thus, function names may not
5639 begin with this prefix. Intrinsic functions must always be external
5640 functions: you cannot define the body of intrinsic functions. Intrinsic
5641 functions may only be used in call or invoke instructions: it is illegal to
5642 take the address of an intrinsic function. Additionally, because intrinsic
5643 functions are part of the LLVM language, it is required if any are added that
5644 they be documented here.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5645
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5646<p>Some intrinsic functions can be overloaded, i.e., the intrinsic represents a
5647 family of functions that perform the same operation but on different data
5648 types. Because LLVM can represent over 8 million different integer types,
5649 overloading is used commonly to allow an intrinsic function to operate on any
5650 integer type. One or more of the argument types or the result type can be
5651 overloaded to accept any integer type. Argument types may also be defined as
5652 exactly matching a previous argument's type or the result type. This allows
5653 an intrinsic function which accepts multiple arguments, but needs all of them
5654 to be of the same type, to only be overloaded with respect to a single
5655 argument or the result.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5656
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5657<p>Overloaded intrinsics will have the names of its overloaded argument types
5658 encoded into its function name, each preceded by a period. Only those types
5659 which are overloaded result in a name suffix. Arguments whose type is matched
5660 against another type do not. For example, the <tt>llvm.ctpop</tt> function
5661 can take an integer of any width and returns an integer of exactly the same
5662 integer width. This leads to a family of functions such as
5663 <tt>i8 @llvm.ctpop.i8(i8 %val)</tt> and <tt>i29 @llvm.ctpop.i29(i29
5664 %val)</tt>. Only one type, the return type, is overloaded, and only one type
5665 suffix is required. Because the argument's type is matched against the return
5666 type, it does not require its own name suffix.</p>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]5667
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5668<p>To learn how to add an intrinsic function, please see the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5669 <a href="ExtendingLLVM.html">Extending LLVM Guide</a>.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5670
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5671<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5672<h3>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5673 <a name="int_varargs">Variable Argument Handling Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5674</h3>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5675
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5676<div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5677
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5678<p>Variable argument support is defined in LLVM with
5679 the <a href="#i_va_arg"><tt>va_arg</tt></a> instruction and these three
5680 intrinsic functions. These functions are related to the similarly named
5681 macros defined in the <tt>&lt;stdarg.h&gt;</tt> header file.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5682
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5683<p>All of these functions operate on arguments that use a target-specific value
5684 type "<tt>va_list</tt>". The LLVM assembly language reference manual does
5685 not define what this type is, so all transformations should be prepared to
5686 handle these functions regardless of the type used.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5687
Chris Lattner374ab302006-05-15 17:26:46 +0000[diff] [blame]5688<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]5689 instruction and the variable argument handling intrinsic functions are
5690 used.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5691
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]5692<pre class="doc_code">
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]5693define i32 @test(i32 %X, ...) {
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5694 ; Initialize variable argument processing
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5695 %ap = alloca i8*
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]5696 %ap2 = bitcast i8** %ap to i8*
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]5697 call void @llvm.va_start(i8* %ap2)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5698
5699 ; Read a single integer argument
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5700 %tmp = va_arg i8** %ap, i32
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5701
5702 ; Demonstrate usage of llvm.va_copy and llvm.va_end
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5703 %aq = alloca i8*
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]5704 %aq2 = bitcast i8** %aq to i8*
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5705 call void @llvm.va_copy(i8* %aq2, i8* %ap2)
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]5706 call void @llvm.va_end(i8* %aq2)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5707
5708 ; Stop processing of arguments.
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]5709 call void @llvm.va_end(i8* %ap2)
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5710 ret i32 %tmp
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5711}
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]5712
5713declare void @llvm.va_start(i8*)
5714declare void @llvm.va_copy(i8*, i8*)
5715declare void @llvm.va_end(i8*)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5716</pre>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5717
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5718<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5719<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5720 <a name="int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5721</h4>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5722
5723
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5724<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5725
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5726<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5727<pre>
5728 declare void %llvm.va_start(i8* &lt;arglist&gt;)
5729</pre>
5730
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5731<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5732<p>The '<tt>llvm.va_start</tt>' intrinsic initializes <tt>*&lt;arglist&gt;</tt>
5733 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]5734
5735<h5>Arguments:</h5>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5736<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]5737
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5738<h5>Semantics:</h5>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5739<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]5740 macro available in C. In a target-dependent way, it initializes
5741 the <tt>va_list</tt> element to which the argument points, so that the next
5742 call to <tt>va_arg</tt> will produce the first variable argument passed to
5743 the function. Unlike the C <tt>va_start</tt> macro, this intrinsic does not
5744 need to know the last argument of the function as the compiler can figure
5745 that out.</p>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]5746
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5747</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5748
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5749<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5750<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5751 <a name="int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5752</h4>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5753
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5754<div>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]5755
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5756<h5>Syntax:</h5>
5757<pre>
5758 declare void @llvm.va_end(i8* &lt;arglist&gt;)
5759</pre>
5760
5761<h5>Overview:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5762<p>The '<tt>llvm.va_end</tt>' intrinsic destroys <tt>*&lt;arglist&gt;</tt>,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5763 which has been initialized previously
5764 with <tt><a href="#int_va_start">llvm.va_start</a></tt>
5765 or <tt><a href="#i_va_copy">llvm.va_copy</a></tt>.</p>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]5766
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5767<h5>Arguments:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5768<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]5769
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5770<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5771<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]5772 macro available in C. In a target-dependent way, it destroys
5773 the <tt>va_list</tt> element to which the argument points. Calls
5774 to <a href="#int_va_start"><tt>llvm.va_start</tt></a>
5775 and <a href="#int_va_copy"> <tt>llvm.va_copy</tt></a> must be matched exactly
5776 with calls to <tt>llvm.va_end</tt>.</p>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]5777
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5778</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5779
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5780<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5781<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5782 <a name="int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5783</h4>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5784
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5785<div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5786
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5787<h5>Syntax:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5788<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]5789 declare void @llvm.va_copy(i8* &lt;destarglist&gt;, i8* &lt;srcarglist&gt;)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5790</pre>
5791
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5792<h5>Overview:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5793<p>The '<tt>llvm.va_copy</tt>' intrinsic copies the current argument position
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5794 from the source argument list to the destination argument list.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5795
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5796<h5>Arguments:</h5>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]5797<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]5798 The second argument is a pointer to a <tt>va_list</tt> element to copy
5799 from.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5800
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]5801<h5>Semantics:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5802<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]5803 macro available in C. In a target-dependent way, it copies the
5804 source <tt>va_list</tt> element into the destination <tt>va_list</tt>
5805 element. This intrinsic is necessary because
5806 the <tt><a href="#int_va_start"> llvm.va_start</a></tt> intrinsic may be
5807 arbitrarily complex and require, for example, memory allocation.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5808
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5809</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5810
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5811</div>
5812
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5813<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5814<h3>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5815 <a name="int_gc">Accurate Garbage Collection Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5816</h3>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5817
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5818<div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5819
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5820<p>LLVM support for <a href="GarbageCollection.html">Accurate Garbage
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]5821Collection</a> (GC) requires the implementation and generation of these
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5822intrinsics. These intrinsics allow identification of <a href="#int_gcroot">GC
5823roots on the stack</a>, as well as garbage collector implementations that
5824require <a href="#int_gcread">read</a> and <a href="#int_gcwrite">write</a>
5825barriers. Front-ends for type-safe garbage collected languages should generate
5826these intrinsics to make use of the LLVM garbage collectors. For more details,
5827see <a href="GarbageCollection.html">Accurate Garbage Collection with
5828LLVM</a>.</p>
Christopher Lamb303dae92007-12-17 01:00:21 +0000[diff] [blame]5829
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5830<p>The garbage collection intrinsics only operate on objects in the generic
5831 address space (address space zero).</p>
Christopher Lamb303dae92007-12-17 01:00:21 +0000[diff] [blame]5832
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5833<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5834<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5835 <a name="int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5836</h4>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5837
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5838<div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5839
5840<h5>Syntax:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5841<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]5842 declare void @llvm.gcroot(i8** %ptrloc, i8* %metadata)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5843</pre>
5844
5845<h5>Overview:</h5>
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]5846<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]5847 the code generator, and allows some metadata to be associated with it.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5848
5849<h5>Arguments:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5850<p>The first argument specifies the address of a stack object that contains the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5851 root pointer. The second pointer (which must be either a constant or a
5852 global value address) contains the meta-data to be associated with the
5853 root.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5854
5855<h5>Semantics:</h5>
Chris Lattner05d67092008-04-24 05:59:56 +0000[diff] [blame]5856<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]5857 location. At compile-time, the code generator generates information to allow
5858 the runtime to find the pointer at GC safe points. The '<tt>llvm.gcroot</tt>'
5859 intrinsic may only be used in a function which <a href="#gc">specifies a GC
5860 algorithm</a>.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5861
5862</div>
5863
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5864<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5865<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5866 <a name="int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5867</h4>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5868
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5869<div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5870
5871<h5>Syntax:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5872<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]5873 declare i8* @llvm.gcread(i8* %ObjPtr, i8** %Ptr)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5874</pre>
5875
5876<h5>Overview:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5877<p>The '<tt>llvm.gcread</tt>' intrinsic identifies reads of references from heap
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5878 locations, allowing garbage collector implementations that require read
5879 barriers.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5880
5881<h5>Arguments:</h5>
Chris Lattner80626e92006-03-14 20:02:51 +0000[diff] [blame]5882<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]5883 allocated from the garbage collector. The first object is a pointer to the
5884 start of the referenced object, if needed by the language runtime (otherwise
5885 null).</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5886
5887<h5>Semantics:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5888<p>The '<tt>llvm.gcread</tt>' intrinsic has the same semantics as a load
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5889 instruction, but may be replaced with substantially more complex code by the
5890 garbage collector runtime, as needed. The '<tt>llvm.gcread</tt>' intrinsic
5891 may only be used in a function which <a href="#gc">specifies a GC
5892 algorithm</a>.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5893
5894</div>
5895
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5896<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5897<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5898 <a name="int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5899</h4>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5900
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5901<div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5902
5903<h5>Syntax:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5904<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]5905 declare void @llvm.gcwrite(i8* %P1, i8* %Obj, i8** %P2)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5906</pre>
5907
5908<h5>Overview:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5909<p>The '<tt>llvm.gcwrite</tt>' intrinsic identifies writes of references to heap
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5910 locations, allowing garbage collector implementations that require write
5911 barriers (such as generational or reference counting collectors).</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5912
5913<h5>Arguments:</h5>
Chris Lattner80626e92006-03-14 20:02:51 +0000[diff] [blame]5914<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]5915 object to store it to, and the third is the address of the field of Obj to
5916 store to. If the runtime does not require a pointer to the object, Obj may
5917 be null.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5918
5919<h5>Semantics:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5920<p>The '<tt>llvm.gcwrite</tt>' intrinsic has the same semantics as a store
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5921 instruction, but may be replaced with substantially more complex code by the
5922 garbage collector runtime, as needed. The '<tt>llvm.gcwrite</tt>' intrinsic
5923 may only be used in a function which <a href="#gc">specifies a GC
5924 algorithm</a>.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5925
5926</div>
5927
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5928</div>
5929
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]5930<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5931<h3>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5932 <a name="int_codegen">Code Generator Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5933</h3>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5934
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5935<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5936
5937<p>These intrinsics are provided by LLVM to expose special features that may
5938 only be implemented with code generator support.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5939
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5940<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5941<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5942 <a name="int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5943</h4>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5944
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5945<div>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5946
5947<h5>Syntax:</h5>
5948<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]5949 declare i8 *@llvm.returnaddress(i32 &lt;level&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5950</pre>
5951
5952<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5953<p>The '<tt>llvm.returnaddress</tt>' intrinsic attempts to compute a
5954 target-specific value indicating the return address of the current function
5955 or one of its callers.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5956
5957<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5958<p>The argument to this intrinsic indicates which function to return the address
5959 for. Zero indicates the calling function, one indicates its caller, etc.
5960 The argument is <b>required</b> to be a constant integer value.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5961
5962<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5963<p>The '<tt>llvm.returnaddress</tt>' intrinsic either returns a pointer
5964 indicating the return address of the specified call frame, or zero if it
5965 cannot be identified. The value returned by this intrinsic is likely to be
5966 incorrect or 0 for arguments other than zero, so it should only be used for
5967 debugging purposes.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5968
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5969<p>Note that calling this intrinsic does not prevent function inlining or other
5970 aggressive transformations, so the value returned may not be that of the
5971 obvious source-language caller.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5972
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5973</div>
5974
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5975<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5976<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5977 <a name="int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5978</h4>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5979
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5980<div>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5981
5982<h5>Syntax:</h5>
5983<pre>
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]5984 declare i8* @llvm.frameaddress(i32 &lt;level&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5985</pre>
5986
5987<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5988<p>The '<tt>llvm.frameaddress</tt>' intrinsic attempts to return the
5989 target-specific frame pointer value for the specified stack frame.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5990
5991<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5992<p>The argument to this intrinsic indicates which function to return the frame
5993 pointer for. Zero indicates the calling function, one indicates its caller,
5994 etc. The argument is <b>required</b> to be a constant integer value.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5995
5996<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5997<p>The '<tt>llvm.frameaddress</tt>' intrinsic either returns a pointer
5998 indicating the frame address of the specified call frame, or zero if it
5999 cannot be identified. The value returned by this intrinsic is likely to be
6000 incorrect or 0 for arguments other than zero, so it should only be used for
6001 debugging purposes.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6002
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6003<p>Note that calling this intrinsic does not prevent function inlining or other
6004 aggressive transformations, so the value returned may not be that of the
6005 obvious source-language caller.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6006
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6007</div>
6008
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]6009<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6010<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6011 <a name="int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6012</h4>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6013
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6014<div>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6015
6016<h5>Syntax:</h5>
6017<pre>
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]6018 declare i8* @llvm.stacksave()
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6019</pre>
6020
6021<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6022<p>The '<tt>llvm.stacksave</tt>' intrinsic is used to remember the current state
6023 of the function stack, for use
6024 with <a href="#int_stackrestore"> <tt>llvm.stackrestore</tt></a>. This is
6025 useful for implementing language features like scoped automatic variable
6026 sized arrays in C99.</p>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6027
6028<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6029<p>This intrinsic returns a opaque pointer value that can be passed
6030 to <a href="#int_stackrestore"><tt>llvm.stackrestore</tt></a>. When
6031 an <tt>llvm.stackrestore</tt> intrinsic is executed with a value saved
6032 from <tt>llvm.stacksave</tt>, it effectively restores the state of the stack
6033 to the state it was in when the <tt>llvm.stacksave</tt> intrinsic executed.
6034 In practice, this pops any <a href="#i_alloca">alloca</a> blocks from the
6035 stack that were allocated after the <tt>llvm.stacksave</tt> was executed.</p>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6036
6037</div>
6038
6039<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6040<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6041 <a name="int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6042</h4>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6043
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6044<div>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6045
6046<h5>Syntax:</h5>
6047<pre>
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]6048 declare void @llvm.stackrestore(i8* %ptr)
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6049</pre>
6050
6051<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6052<p>The '<tt>llvm.stackrestore</tt>' intrinsic is used to restore the state of
6053 the function stack to the state it was in when the
6054 corresponding <a href="#int_stacksave"><tt>llvm.stacksave</tt></a> intrinsic
6055 executed. This is useful for implementing language features like scoped
6056 automatic variable sized arrays in C99.</p>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6057
6058<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6059<p>See the description
6060 for <a href="#int_stacksave"><tt>llvm.stacksave</tt></a>.</p>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6061
6062</div>
6063
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6064<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6065<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6066 <a name="int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6067</h4>
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]6068
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6069<div>
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]6070
6071<h5>Syntax:</h5>
6072<pre>
Bruno Cardoso Lopes9a767332011-06-14 04:58:37 +0000[diff] [blame]6073 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]6074</pre>
6075
6076<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6077<p>The '<tt>llvm.prefetch</tt>' intrinsic is a hint to the code generator to
6078 insert a prefetch instruction if supported; otherwise, it is a noop.
6079 Prefetches have no effect on the behavior of the program but can change its
6080 performance characteristics.</p>
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]6081
6082<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6083<p><tt>address</tt> is the address to be prefetched, <tt>rw</tt> is the
6084 specifier determining if the fetch should be for a read (0) or write (1),
6085 and <tt>locality</tt> is a temporal locality specifier ranging from (0) - no
Bruno Cardoso Lopes9a767332011-06-14 04:58:37 +0000[diff] [blame]6086 locality, to (3) - extremely local keep in cache. The <tt>cache type</tt>
6087 specifies whether the prefetch is performed on the data (1) or instruction (0)
6088 cache. The <tt>rw</tt>, <tt>locality</tt> and <tt>cache type</tt> arguments
6089 must be constant integers.</p>
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]6090
6091<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6092<p>This intrinsic does not modify the behavior of the program. In particular,
6093 prefetches cannot trap and do not produce a value. On targets that support
6094 this intrinsic, the prefetch can provide hints to the processor cache for
6095 better performance.</p>
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]6096
6097</div>
6098
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]6099<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6100<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6101 <a name="int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6102</h4>
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]6103
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6104<div>
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]6105
6106<h5>Syntax:</h5>
6107<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]6108 declare void @llvm.pcmarker(i32 &lt;id&gt;)
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]6109</pre>
6110
6111<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6112<p>The '<tt>llvm.pcmarker</tt>' intrinsic is a method to export a Program
6113 Counter (PC) in a region of code to simulators and other tools. The method
6114 is target specific, but it is expected that the marker will use exported
6115 symbols to transmit the PC of the marker. The marker makes no guarantees
6116 that it will remain with any specific instruction after optimizations. It is
6117 possible that the presence of a marker will inhibit optimizations. The
6118 intended use is to be inserted after optimizations to allow correlations of
6119 simulation runs.</p>
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]6120
6121<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6122<p><tt>id</tt> is a numerical id identifying the marker.</p>
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]6123
6124<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6125<p>This intrinsic does not modify the behavior of the program. Backends that do
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]6126 not support this intrinsic may ignore it.</p>
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]6127
6128</div>
6129
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]6130<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6131<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6132 <a name="int_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6133</h4>
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]6134
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6135<div>
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]6136
6137<h5>Syntax:</h5>
6138<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]6139 declare i64 @llvm.readcyclecounter()
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]6140</pre>
6141
6142<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6143<p>The '<tt>llvm.readcyclecounter</tt>' intrinsic provides access to the cycle
6144 counter register (or similar low latency, high accuracy clocks) on those
6145 targets that support it. On X86, it should map to RDTSC. On Alpha, it
6146 should map to RPCC. As the backing counters overflow quickly (on the order
6147 of 9 seconds on alpha), this should only be used for small timings.</p>
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]6148
6149<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6150<p>When directly supported, reading the cycle counter should not modify any
6151 memory. Implementations are allowed to either return a application specific
6152 value or a system wide value. On backends without support, this is lowered
6153 to a constant 0.</p>
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]6154
6155</div>
6156
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6157</div>
6158
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6159<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6160<h3>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6161 <a name="int_libc">Standard C Library Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6162</h3>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6163
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6164<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6165
6166<p>LLVM provides intrinsics for a few important standard C library functions.
6167 These intrinsics allow source-language front-ends to pass information about
6168 the alignment of the pointer arguments to the code generator, providing
6169 opportunity for more efficient code generation.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6170
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6171<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6172<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6173 <a name="int_memcpy">'<tt>llvm.memcpy</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6174</h4>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6175
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6176<div>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6177
6178<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6179<p>This is an overloaded intrinsic. You can use <tt>llvm.memcpy</tt> on any
Mon P Wange88909b2010-04-07 06:35:53 +0000[diff] [blame]6180 integer bit width and for different address spaces. Not all targets support
6181 all bit widths however.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6182
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6183<pre>
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]6184 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]6185 i32 &lt;len&gt;, i32 &lt;align&gt;, i1 &lt;isvolatile&gt;)
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]6186 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]6187 i64 &lt;len&gt;, i32 &lt;align&gt;, i1 &lt;isvolatile&gt;)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6188</pre>
6189
6190<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6191<p>The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the
6192 source location to the destination location.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6193
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6194<p>Note that, unlike the standard libc function, the <tt>llvm.memcpy.*</tt>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6195 intrinsics do not return a value, takes extra alignment/isvolatile arguments
6196 and the pointers can be in specified address spaces.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6197
6198<h5>Arguments:</h5>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6199
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6200<p>The first argument is a pointer to the destination, the second is a pointer
6201 to the source. The third argument is an integer argument specifying the
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6202 number of bytes to copy, the fourth argument is the alignment of the
6203 source and destination locations, and the fifth is a boolean indicating a
6204 volatile access.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6205
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]6206<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]6207 then the caller guarantees that both the source and destination pointers are
6208 aligned to that boundary.</p>
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]6209
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]6210<p>If the <tt>isvolatile</tt> parameter is <tt>true</tt>, the
6211 <tt>llvm.memcpy</tt> call is a <a href="#volatile">volatile operation</a>.
6212 The detailed access behavior is not very cleanly specified and it is unwise
6213 to depend on it.</p>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6214
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6215<h5>Semantics:</h5>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6216
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6217<p>The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the
6218 source location to the destination location, which are not allowed to
6219 overlap. It copies "len" bytes of memory over. If the argument is known to
6220 be aligned to some boundary, this can be specified as the fourth argument,
6221 otherwise it should be set to 0 or 1.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6222
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6223</div>
6224
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6225<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6226<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6227 <a name="int_memmove">'<tt>llvm.memmove</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6228</h4>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6229
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6230<div>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6231
6232<h5>Syntax:</h5>
Chris Lattner824b9582008-11-21 16:42:48 +0000[diff] [blame]6233<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]6234 width and for different address space. Not all targets support all bit
6235 widths however.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6236
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6237<pre>
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]6238 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]6239 i32 &lt;len&gt;, i32 &lt;align&gt;, i1 &lt;isvolatile&gt;)
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]6240 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]6241 i64 &lt;len&gt;, i32 &lt;align&gt;, i1 &lt;isvolatile&gt;)
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6242</pre>
6243
6244<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6245<p>The '<tt>llvm.memmove.*</tt>' intrinsics move a block of memory from the
6246 source location to the destination location. It is similar to the
6247 '<tt>llvm.memcpy</tt>' intrinsic but allows the two memory locations to
6248 overlap.</p>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6249
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6250<p>Note that, unlike the standard libc function, the <tt>llvm.memmove.*</tt>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6251 intrinsics do not return a value, takes extra alignment/isvolatile arguments
6252 and the pointers can be in specified address spaces.</p>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6253
6254<h5>Arguments:</h5>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6255
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6256<p>The first argument is a pointer to the destination, the second is a pointer
6257 to the source. The third argument is an integer argument specifying the
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6258 number of bytes to copy, the fourth argument is the alignment of the
6259 source and destination locations, and the fifth is a boolean indicating a
6260 volatile access.</p>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6261
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]6262<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]6263 then the caller guarantees that the source and destination pointers are
6264 aligned to that boundary.</p>
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]6265
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]6266<p>If the <tt>isvolatile</tt> parameter is <tt>true</tt>, the
6267 <tt>llvm.memmove</tt> call is a <a href="#volatile">volatile operation</a>.
6268 The detailed access behavior is not very cleanly specified and it is unwise
6269 to depend on it.</p>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6270
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6271<h5>Semantics:</h5>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6272
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6273<p>The '<tt>llvm.memmove.*</tt>' intrinsics copy a block of memory from the
6274 source location to the destination location, which may overlap. It copies
6275 "len" bytes of memory over. If the argument is known to be aligned to some
6276 boundary, this can be specified as the fourth argument, otherwise it should
6277 be set to 0 or 1.</p>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6278
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]6279</div>
6280
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6281<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6282<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6283 <a name="int_memset">'<tt>llvm.memset.*</tt>' Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6284</h4>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6285
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6286<div>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6287
6288<h5>Syntax:</h5>
Chris Lattner824b9582008-11-21 16:42:48 +0000[diff] [blame]6289<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]6290 width and for different address spaces. However, not all targets support all
6291 bit widths.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6292
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6293<pre>
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]6294 declare void @llvm.memset.p0i8.i32(i8* &lt;dest&gt;, i8 &lt;val&gt;,
Chris Lattnerff35c3f2010-04-08 00:54:34 +0000[diff] [blame]6295 i32 &lt;len&gt;, i32 &lt;align&gt;, i1 &lt;isvolatile&gt;)
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]6296 declare void @llvm.memset.p0i8.i64(i8* &lt;dest&gt;, i8 &lt;val&gt;,
Chris Lattnerff35c3f2010-04-08 00:54:34 +0000[diff] [blame]6297 i64 &lt;len&gt;, i32 &lt;align&gt;, i1 &lt;isvolatile&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6298</pre>
6299
6300<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6301<p>The '<tt>llvm.memset.*</tt>' intrinsics fill a block of memory with a
6302 particular byte value.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6303
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6304<p>Note that, unlike the standard libc function, the <tt>llvm.memset</tt>
John Criswellcdcbbfc2010-07-30 16:30:28 +0000[diff] [blame]6305 intrinsic does not return a value and takes extra alignment/volatile
6306 arguments. Also, the destination can be in an arbitrary address space.</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 first argument is a pointer to the destination to fill, the second is the
John Criswellcdcbbfc2010-07-30 16:30:28 +0000[diff] [blame]6310 byte value with which to fill it, the third argument is an integer argument
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6311 specifying the number of bytes to fill, and the fourth argument is the known
John Criswellcdcbbfc2010-07-30 16:30:28 +0000[diff] [blame]6312 alignment of the destination location.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6313
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]6314<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]6315 then the caller guarantees that the destination pointer is aligned to that
6316 boundary.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6317
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]6318<p>If the <tt>isvolatile</tt> parameter is <tt>true</tt>, the
6319 <tt>llvm.memset</tt> call is a <a href="#volatile">volatile operation</a>.
6320 The detailed access behavior is not very cleanly specified and it is unwise
6321 to depend on it.</p>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]6322
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6323<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6324<p>The '<tt>llvm.memset.*</tt>' intrinsics fill "len" bytes of memory starting
6325 at the destination location. If the argument is known to be aligned to some
6326 boundary, this can be specified as the fourth argument, otherwise it should
6327 be set to 0 or 1.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6328
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6329</div>
6330
Chris Lattner32006282004-06-11 02:28:03 +0000[diff] [blame]6331<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6332<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6333 <a name="int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6334</h4>
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]6335
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6336<div>
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]6337
6338<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6339<p>This is an overloaded intrinsic. You can use <tt>llvm.sqrt</tt> on any
6340 floating point or vector of floating point type. Not all targets support all
6341 types however.</p>
6342
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]6343<pre>
Dale Johannesen408f9c12007-10-02 17:47:38 +0000[diff] [blame]6344 declare float @llvm.sqrt.f32(float %Val)
6345 declare double @llvm.sqrt.f64(double %Val)
6346 declare x86_fp80 @llvm.sqrt.f80(x86_fp80 %Val)
6347 declare fp128 @llvm.sqrt.f128(fp128 %Val)
6348 declare ppc_fp128 @llvm.sqrt.ppcf128(ppc_fp128 %Val)
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]6349</pre>
6350
6351<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6352<p>The '<tt>llvm.sqrt</tt>' intrinsics return the sqrt of the specified operand,
6353 returning the same value as the libm '<tt>sqrt</tt>' functions would.
6354 Unlike <tt>sqrt</tt> in libm, however, <tt>llvm.sqrt</tt> has undefined
6355 behavior for negative numbers other than -0.0 (which allows for better
6356 optimization, because there is no need to worry about errno being
6357 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]6358
6359<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6360<p>The argument and return value are floating point numbers of the same
6361 type.</p>
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]6362
6363<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6364<p>This function returns the sqrt of the specified operand if it is a
6365 nonnegative floating point number.</p>
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]6366
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]6367</div>
6368
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6369<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6370<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6371 <a name="int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6372</h4>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6373
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6374<div>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6375
6376<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6377<p>This is an overloaded intrinsic. You can use <tt>llvm.powi</tt> on any
6378 floating point or vector of floating point type. Not all targets support all
6379 types however.</p>
6380
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6381<pre>
Dale Johannesen408f9c12007-10-02 17:47:38 +0000[diff] [blame]6382 declare float @llvm.powi.f32(float %Val, i32 %power)
6383 declare double @llvm.powi.f64(double %Val, i32 %power)
6384 declare x86_fp80 @llvm.powi.f80(x86_fp80 %Val, i32 %power)
6385 declare fp128 @llvm.powi.f128(fp128 %Val, i32 %power)
6386 declare ppc_fp128 @llvm.powi.ppcf128(ppc_fp128 %Val, i32 %power)
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6387</pre>
6388
6389<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6390<p>The '<tt>llvm.powi.*</tt>' intrinsics return the first operand raised to the
6391 specified (positive or negative) power. The order of evaluation of
6392 multiplications is not defined. When a vector of floating point type is
6393 used, the second argument remains a scalar integer value.</p>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6394
6395<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6396<p>The second argument is an integer power, and the first is a value to raise to
6397 that power.</p>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6398
6399<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6400<p>This function returns the first value raised to the second power with an
6401 unspecified sequence of rounding operations.</p>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6402
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]6403</div>
6404
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6405<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6406<h4>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6407 <a name="int_sin">'<tt>llvm.sin.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6408</h4>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6409
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6410<div>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6411
6412<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6413<p>This is an overloaded intrinsic. You can use <tt>llvm.sin</tt> on any
6414 floating point or vector of floating point type. Not all targets support all
6415 types however.</p>
6416
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6417<pre>
6418 declare float @llvm.sin.f32(float %Val)
6419 declare double @llvm.sin.f64(double %Val)
6420 declare x86_fp80 @llvm.sin.f80(x86_fp80 %Val)
6421 declare fp128 @llvm.sin.f128(fp128 %Val)
6422 declare ppc_fp128 @llvm.sin.ppcf128(ppc_fp128 %Val)
6423</pre>
6424
6425<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6426<p>The '<tt>llvm.sin.*</tt>' intrinsics return the sine of the operand.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6427
6428<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6429<p>The argument and return value are floating point numbers of the same
6430 type.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6431
6432<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6433<p>This function returns the sine of the specified operand, returning the same
6434 values as the libm <tt>sin</tt> functions would, and handles error conditions
6435 in the same way.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6436
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6437</div>
6438
6439<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6440<h4>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6441 <a name="int_cos">'<tt>llvm.cos.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6442</h4>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6443
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6444<div>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6445
6446<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6447<p>This is an overloaded intrinsic. You can use <tt>llvm.cos</tt> on any
6448 floating point or vector of floating point type. Not all targets support all
6449 types however.</p>
6450
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6451<pre>
6452 declare float @llvm.cos.f32(float %Val)
6453 declare double @llvm.cos.f64(double %Val)
6454 declare x86_fp80 @llvm.cos.f80(x86_fp80 %Val)
6455 declare fp128 @llvm.cos.f128(fp128 %Val)
6456 declare ppc_fp128 @llvm.cos.ppcf128(ppc_fp128 %Val)
6457</pre>
6458
6459<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6460<p>The '<tt>llvm.cos.*</tt>' intrinsics return the cosine of the operand.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6461
6462<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6463<p>The argument and return value are floating point numbers of the same
6464 type.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6465
6466<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6467<p>This function returns the cosine of the specified operand, returning the same
6468 values as the libm <tt>cos</tt> functions would, and handles error conditions
6469 in the same way.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6470
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6471</div>
6472
6473<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6474<h4>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6475 <a name="int_pow">'<tt>llvm.pow.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6476</h4>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6477
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6478<div>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6479
6480<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6481<p>This is an overloaded intrinsic. You can use <tt>llvm.pow</tt> on any
6482 floating point or vector of floating point type. Not all targets support all
6483 types however.</p>
6484
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6485<pre>
6486 declare float @llvm.pow.f32(float %Val, float %Power)
6487 declare double @llvm.pow.f64(double %Val, double %Power)
6488 declare x86_fp80 @llvm.pow.f80(x86_fp80 %Val, x86_fp80 %Power)
6489 declare fp128 @llvm.pow.f128(fp128 %Val, fp128 %Power)
6490 declare ppc_fp128 @llvm.pow.ppcf128(ppc_fp128 %Val, ppc_fp128 Power)
6491</pre>
6492
6493<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6494<p>The '<tt>llvm.pow.*</tt>' intrinsics return the first operand raised to the
6495 specified (positive or negative) power.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6496
6497<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6498<p>The second argument is a floating point power, and the first is a value to
6499 raise to that power.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6500
6501<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6502<p>This function returns the first value raised to the second power, returning
6503 the same values as the libm <tt>pow</tt> functions would, and handles error
6504 conditions in the same way.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6505
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]6506</div>
6507
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6508</div>
6509
Dan Gohman4e9011c2011-05-23 21:13:03 +0000[diff] [blame]6510<!-- _______________________________________________________________________ -->
6511<h4>
6512 <a name="int_exp">'<tt>llvm.exp.*</tt>' Intrinsic</a>
6513</h4>
6514
6515<div>
6516
6517<h5>Syntax:</h5>
6518<p>This is an overloaded intrinsic. You can use <tt>llvm.exp</tt> on any
6519 floating point or vector of floating point type. Not all targets support all
6520 types however.</p>
6521
6522<pre>
6523 declare float @llvm.exp.f32(float %Val)
6524 declare double @llvm.exp.f64(double %Val)
6525 declare x86_fp80 @llvm.exp.f80(x86_fp80 %Val)
6526 declare fp128 @llvm.exp.f128(fp128 %Val)
6527 declare ppc_fp128 @llvm.exp.ppcf128(ppc_fp128 %Val)
6528</pre>
6529
6530<h5>Overview:</h5>
6531<p>The '<tt>llvm.exp.*</tt>' intrinsics perform the exp function.</p>
6532
6533<h5>Arguments:</h5>
6534<p>The argument and return value are floating point numbers of the same
6535 type.</p>
6536
6537<h5>Semantics:</h5>
6538<p>This function returns the same values as the libm <tt>exp</tt> functions
6539 would, and handles error conditions in the same way.</p>
6540
6541</div>
6542
6543<!-- _______________________________________________________________________ -->
6544<h4>
6545 <a name="int_log">'<tt>llvm.log.*</tt>' Intrinsic</a>
6546</h4>
6547
6548<div>
6549
6550<h5>Syntax:</h5>
6551<p>This is an overloaded intrinsic. You can use <tt>llvm.log</tt> on any
6552 floating point or vector of floating point type. Not all targets support all
6553 types however.</p>
6554
6555<pre>
6556 declare float @llvm.log.f32(float %Val)
6557 declare double @llvm.log.f64(double %Val)
6558 declare x86_fp80 @llvm.log.f80(x86_fp80 %Val)
6559 declare fp128 @llvm.log.f128(fp128 %Val)
6560 declare ppc_fp128 @llvm.log.ppcf128(ppc_fp128 %Val)
6561</pre>
6562
6563<h5>Overview:</h5>
6564<p>The '<tt>llvm.log.*</tt>' intrinsics perform the log function.</p>
6565
6566<h5>Arguments:</h5>
6567<p>The argument and return value are floating point numbers of the same
6568 type.</p>
6569
6570<h5>Semantics:</h5>
6571<p>This function returns the same values as the libm <tt>log</tt> functions
6572 would, and handles error conditions in the same way.</p>
6573
Cameron Zwarich33390842011-07-08 21:39:21 +0000[diff] [blame]6574<h4>
6575 <a name="int_fma">'<tt>llvm.fma.*</tt>' Intrinsic</a>
6576</h4>
6577
6578<div>
6579
6580<h5>Syntax:</h5>
6581<p>This is an overloaded intrinsic. You can use <tt>llvm.fma</tt> on any
6582 floating point or vector of floating point type. Not all targets support all
6583 types however.</p>
6584
6585<pre>
6586 declare float @llvm.fma.f32(float %a, float %b, float %c)
6587 declare double @llvm.fma.f64(double %a, double %b, double %c)
6588 declare x86_fp80 @llvm.fma.f80(x86_fp80 %a, x86_fp80 %b, x86_fp80 %c)
6589 declare fp128 @llvm.fma.f128(fp128 %a, fp128 %b, fp128 %c)
6590 declare ppc_fp128 @llvm.fma.ppcf128(ppc_fp128 %a, ppc_fp128 %b, ppc_fp128 %c)
6591</pre>
6592
6593<h5>Overview:</h5>
Cameron Zwarichabc43e62011-07-08 22:13:55 +0000[diff] [blame]6594<p>The '<tt>llvm.fma.*</tt>' intrinsics perform the fused multiply-add
Cameron Zwarich33390842011-07-08 21:39:21 +0000[diff] [blame]6595 operation.</p>
6596
6597<h5>Arguments:</h5>
6598<p>The argument and return value are floating point numbers of the same
6599 type.</p>
6600
6601<h5>Semantics:</h5>
6602<p>This function returns the same values as the libm <tt>fma</tt> functions
6603 would.</p>
6604
Dan Gohman4e9011c2011-05-23 21:13:03 +0000[diff] [blame]6605</div>
6606
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6607<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6608<h3>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]6609 <a name="int_manip">Bit Manipulation Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6610</h3>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6611
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6612<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6613
6614<p>LLVM provides intrinsics for a few important bit manipulation operations.
6615 These allow efficient code generation for some algorithms.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6616
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6617<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6618<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6619 <a name="int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6620</h4>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]6621
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6622<div>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]6623
6624<h5>Syntax:</h5>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]6625<p>This is an overloaded intrinsic function. You can use bswap on any integer
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6626 type that is an even number of bytes (i.e. BitWidth % 16 == 0).</p>
6627
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]6628<pre>
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]6629 declare i16 @llvm.bswap.i16(i16 &lt;id&gt;)
6630 declare i32 @llvm.bswap.i32(i32 &lt;id&gt;)
6631 declare i64 @llvm.bswap.i64(i64 &lt;id&gt;)
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]6632</pre>
6633
6634<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6635<p>The '<tt>llvm.bswap</tt>' family of intrinsics is used to byte swap integer
6636 values with an even number of bytes (positive multiple of 16 bits). These
6637 are useful for performing operations on data that is not in the target's
6638 native byte order.</p>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]6639
6640<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6641<p>The <tt>llvm.bswap.i16</tt> intrinsic returns an i16 value that has the high
6642 and low byte of the input i16 swapped. Similarly,
6643 the <tt>llvm.bswap.i32</tt> intrinsic returns an i32 value that has the four
6644 bytes of the input i32 swapped, so that if the input bytes are numbered 0, 1,
6645 2, 3 then the returned i32 will have its bytes in 3, 2, 1, 0 order.
6646 The <tt>llvm.bswap.i48</tt>, <tt>llvm.bswap.i64</tt> and other intrinsics
6647 extend this concept to additional even-byte lengths (6 bytes, 8 bytes and
6648 more, respectively).</p>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]6649
6650</div>
6651
6652<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6653<h4>
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]6654 <a name="int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6655</h4>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6656
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6657<div>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6658
6659<h5>Syntax:</h5>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]6660<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]6661 width, or on any vector with integer elements. Not all targets support all
6662 bit widths or vector types, however.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6663
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6664<pre>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6665 declare i8 @llvm.ctpop.i8(i8 &lt;src&gt;)
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]6666 declare i16 @llvm.ctpop.i16(i16 &lt;src&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]6667 declare i32 @llvm.ctpop.i32(i32 &lt;src&gt;)
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]6668 declare i64 @llvm.ctpop.i64(i64 &lt;src&gt;)
6669 declare i256 @llvm.ctpop.i256(i256 &lt;src&gt;)
Owen Andersonf1ac4652011-07-01 21:52:38 +0000[diff] [blame]6670 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]6671</pre>
6672
6673<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6674<p>The '<tt>llvm.ctpop</tt>' family of intrinsics counts the number of bits set
6675 in a value.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6676
6677<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6678<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]6679 integer type, or a vector with integer elements.
6680 The return type must match the argument type.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6681
6682<h5>Semantics:</h5>
Owen Andersonf1ac4652011-07-01 21:52:38 +0000[diff] [blame]6683<p>The '<tt>llvm.ctpop</tt>' intrinsic counts the 1's in a variable, or within each
6684 element of a vector.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6685
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6686</div>
6687
6688<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6689<h4>
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]6690 <a name="int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6691</h4>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6692
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6693<div>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6694
6695<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6696<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]6697 integer bit width, or any vector whose elements are integers. Not all
6698 targets support all bit widths or vector types, however.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6699
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6700<pre>
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]6701 declare i8 @llvm.ctlz.i8 (i8 &lt;src&gt;)
6702 declare i16 @llvm.ctlz.i16(i16 &lt;src&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]6703 declare i32 @llvm.ctlz.i32(i32 &lt;src&gt;)
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]6704 declare i64 @llvm.ctlz.i64(i64 &lt;src&gt;)
6705 declare i256 @llvm.ctlz.i256(i256 &lt;src&gt;)
Owen Andersonf1ac4652011-07-01 21:52:38 +0000[diff] [blame]6706 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]6707</pre>
6708
6709<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6710<p>The '<tt>llvm.ctlz</tt>' family of intrinsic functions counts the number of
6711 leading zeros in a variable.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6712
6713<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6714<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]6715 integer type, or any vector type with integer element type.
6716 The return type must match the argument type.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6717
6718<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6719<p>The '<tt>llvm.ctlz</tt>' intrinsic counts the leading (most significant)
Owen Andersonf1ac4652011-07-01 21:52:38 +0000[diff] [blame]6720 zeros in a variable, or within each element of the vector if the operation
6721 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]6722 the type of src. For example, <tt>llvm.ctlz(i32 2) = 30</tt>.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6723
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]6724</div>
Chris Lattner32006282004-06-11 02:28:03 +0000[diff] [blame]6725
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]6726<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6727<h4>
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]6728 <a name="int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6729</h4>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]6730
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6731<div>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]6732
6733<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6734<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]6735 integer bit width, or any vector of integer elements. Not all targets
6736 support all bit widths or vector types, however.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6737
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]6738<pre>
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]6739 declare i8 @llvm.cttz.i8 (i8 &lt;src&gt;)
6740 declare i16 @llvm.cttz.i16(i16 &lt;src&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]6741 declare i32 @llvm.cttz.i32(i32 &lt;src&gt;)
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]6742 declare i64 @llvm.cttz.i64(i64 &lt;src&gt;)
6743 declare i256 @llvm.cttz.i256(i256 &lt;src&gt;)
Owen Andersonf1ac4652011-07-01 21:52:38 +0000[diff] [blame]6744 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]6745</pre>
6746
6747<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6748<p>The '<tt>llvm.cttz</tt>' family of intrinsic functions counts the number of
6749 trailing zeros.</p>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]6750
6751<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6752<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]6753 integer type, or a vectory with integer element type.. The return type
6754 must match the argument type.</p>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]6755
6756<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6757<p>The '<tt>llvm.cttz</tt>' intrinsic counts the trailing (least significant)
Owen Andersonf1ac4652011-07-01 21:52:38 +0000[diff] [blame]6758 zeros in a variable, or within each element of a vector.
6759 If the src == 0 then the result is the size in bits of
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6760 the type of src. For example, <tt>llvm.cttz(2) = 1</tt>.</p>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]6761
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]6762</div>
6763
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6764</div>
6765
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]6766<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6767<h3>
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]6768 <a name="int_overflow">Arithmetic with Overflow Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6769</h3>
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]6770
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6771<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6772
6773<p>LLVM provides intrinsics for some arithmetic with overflow operations.</p>
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]6774
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6775<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6776<h4>
6777 <a name="int_sadd_overflow">
6778 '<tt>llvm.sadd.with.overflow.*</tt>' Intrinsics
6779 </a>
6780</h4>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6781
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6782<div>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6783
6784<h5>Syntax:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6785<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]6786 on any integer bit width.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6787
6788<pre>
6789 declare {i16, i1} @llvm.sadd.with.overflow.i16(i16 %a, i16 %b)
6790 declare {i32, i1} @llvm.sadd.with.overflow.i32(i32 %a, i32 %b)
6791 declare {i64, i1} @llvm.sadd.with.overflow.i64(i64 %a, i64 %b)
6792</pre>
6793
6794<h5>Overview:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6795<p>The '<tt>llvm.sadd.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6796 a signed addition of the two arguments, and indicate whether an overflow
6797 occurred during the signed summation.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6798
6799<h5>Arguments:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6800<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]6801 be of integer types of any bit width, but they must have the same bit
6802 width. The second element of the result structure must be of
6803 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
6804 undergo signed addition.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6805
6806<h5>Semantics:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6807<p>The '<tt>llvm.sadd.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6808 a signed addition of the two variables. They return a structure &mdash; the
6809 first element of which is the signed summation, and the second element of
6810 which is a bit specifying if the signed summation resulted in an
6811 overflow.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6812
6813<h5>Examples:</h5>
6814<pre>
6815 %res = call {i32, i1} @llvm.sadd.with.overflow.i32(i32 %a, i32 %b)
6816 %sum = extractvalue {i32, i1} %res, 0
6817 %obit = extractvalue {i32, i1} %res, 1
6818 br i1 %obit, label %overflow, label %normal
6819</pre>
6820
6821</div>
6822
6823<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6824<h4>
6825 <a name="int_uadd_overflow">
6826 '<tt>llvm.uadd.with.overflow.*</tt>' Intrinsics
6827 </a>
6828</h4>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6829
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6830<div>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6831
6832<h5>Syntax:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6833<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]6834 on any integer bit width.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6835
6836<pre>
6837 declare {i16, i1} @llvm.uadd.with.overflow.i16(i16 %a, i16 %b)
6838 declare {i32, i1} @llvm.uadd.with.overflow.i32(i32 %a, i32 %b)
6839 declare {i64, i1} @llvm.uadd.with.overflow.i64(i64 %a, i64 %b)
6840</pre>
6841
6842<h5>Overview:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6843<p>The '<tt>llvm.uadd.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6844 an unsigned addition of the two arguments, and indicate whether a carry
6845 occurred during the unsigned summation.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6846
6847<h5>Arguments:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6848<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]6849 be of integer types of any bit width, but they must have the same bit
6850 width. The second element of the result structure must be of
6851 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
6852 undergo unsigned addition.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6853
6854<h5>Semantics:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6855<p>The '<tt>llvm.uadd.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6856 an unsigned addition of the two arguments. They return a structure &mdash;
6857 the first element of which is the sum, and the second element of which is a
6858 bit specifying if the unsigned summation resulted in a carry.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6859
6860<h5>Examples:</h5>
6861<pre>
6862 %res = call {i32, i1} @llvm.uadd.with.overflow.i32(i32 %a, i32 %b)
6863 %sum = extractvalue {i32, i1} %res, 0
6864 %obit = extractvalue {i32, i1} %res, 1
6865 br i1 %obit, label %carry, label %normal
6866</pre>
6867
6868</div>
6869
6870<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6871<h4>
6872 <a name="int_ssub_overflow">
6873 '<tt>llvm.ssub.with.overflow.*</tt>' Intrinsics
6874 </a>
6875</h4>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6876
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6877<div>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6878
6879<h5>Syntax:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6880<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]6881 on any integer bit width.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6882
6883<pre>
6884 declare {i16, i1} @llvm.ssub.with.overflow.i16(i16 %a, i16 %b)
6885 declare {i32, i1} @llvm.ssub.with.overflow.i32(i32 %a, i32 %b)
6886 declare {i64, i1} @llvm.ssub.with.overflow.i64(i64 %a, i64 %b)
6887</pre>
6888
6889<h5>Overview:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6890<p>The '<tt>llvm.ssub.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6891 a signed subtraction of the two arguments, and indicate whether an overflow
6892 occurred during the signed subtraction.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6893
6894<h5>Arguments:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6895<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]6896 be of integer types of any bit width, but they must have the same bit
6897 width. The second element of the result structure must be of
6898 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
6899 undergo signed subtraction.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6900
6901<h5>Semantics:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6902<p>The '<tt>llvm.ssub.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6903 a signed subtraction of the two arguments. They return a structure &mdash;
6904 the first element of which is the subtraction, and the second element of
6905 which is a bit specifying if the signed subtraction resulted in an
6906 overflow.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6907
6908<h5>Examples:</h5>
6909<pre>
6910 %res = call {i32, i1} @llvm.ssub.with.overflow.i32(i32 %a, i32 %b)
6911 %sum = extractvalue {i32, i1} %res, 0
6912 %obit = extractvalue {i32, i1} %res, 1
6913 br i1 %obit, label %overflow, label %normal
6914</pre>
6915
6916</div>
6917
6918<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6919<h4>
6920 <a name="int_usub_overflow">
6921 '<tt>llvm.usub.with.overflow.*</tt>' Intrinsics
6922 </a>
6923</h4>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6924
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6925<div>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6926
6927<h5>Syntax:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6928<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]6929 on any integer bit width.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6930
6931<pre>
6932 declare {i16, i1} @llvm.usub.with.overflow.i16(i16 %a, i16 %b)
6933 declare {i32, i1} @llvm.usub.with.overflow.i32(i32 %a, i32 %b)
6934 declare {i64, i1} @llvm.usub.with.overflow.i64(i64 %a, i64 %b)
6935</pre>
6936
6937<h5>Overview:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6938<p>The '<tt>llvm.usub.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6939 an unsigned subtraction of the two arguments, and indicate whether an
6940 overflow occurred during the unsigned subtraction.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6941
6942<h5>Arguments:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6943<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]6944 be of integer types of any bit width, but they must have the same bit
6945 width. The second element of the result structure must be of
6946 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
6947 undergo unsigned subtraction.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6948
6949<h5>Semantics:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6950<p>The '<tt>llvm.usub.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6951 an unsigned subtraction of the two arguments. They return a structure &mdash;
6952 the first element of which is the subtraction, and the second element of
6953 which is a bit specifying if the unsigned subtraction resulted in an
6954 overflow.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6955
6956<h5>Examples:</h5>
6957<pre>
6958 %res = call {i32, i1} @llvm.usub.with.overflow.i32(i32 %a, i32 %b)
6959 %sum = extractvalue {i32, i1} %res, 0
6960 %obit = extractvalue {i32, i1} %res, 1
6961 br i1 %obit, label %overflow, label %normal
6962</pre>
6963
6964</div>
6965
6966<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6967<h4>
6968 <a name="int_smul_overflow">
6969 '<tt>llvm.smul.with.overflow.*</tt>' Intrinsics
6970 </a>
6971</h4>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6972
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6973<div>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6974
6975<h5>Syntax:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6976<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]6977 on any integer bit width.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6978
6979<pre>
6980 declare {i16, i1} @llvm.smul.with.overflow.i16(i16 %a, i16 %b)
6981 declare {i32, i1} @llvm.smul.with.overflow.i32(i32 %a, i32 %b)
6982 declare {i64, i1} @llvm.smul.with.overflow.i64(i64 %a, i64 %b)
6983</pre>
6984
6985<h5>Overview:</h5>
6986
6987<p>The '<tt>llvm.smul.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6988 a signed multiplication of the two arguments, and indicate whether an
6989 overflow occurred during the signed multiplication.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6990
6991<h5>Arguments:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6992<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]6993 be of integer types of any bit width, but they must have the same bit
6994 width. The second element of the result structure must be of
6995 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
6996 undergo signed multiplication.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6997
6998<h5>Semantics:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6999<p>The '<tt>llvm.smul.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7000 a signed multiplication of the two arguments. They return a structure &mdash;
7001 the first element of which is the multiplication, and the second element of
7002 which is a bit specifying if the signed multiplication resulted in an
7003 overflow.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7004
7005<h5>Examples:</h5>
7006<pre>
7007 %res = call {i32, i1} @llvm.smul.with.overflow.i32(i32 %a, i32 %b)
7008 %sum = extractvalue {i32, i1} %res, 0
7009 %obit = extractvalue {i32, i1} %res, 1
7010 br i1 %obit, label %overflow, label %normal
7011</pre>
7012
Reid Spencerf86037f2007-04-11 23:23:49 +0000[diff] [blame]7013</div>
7014
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7015<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7016<h4>
7017 <a name="int_umul_overflow">
7018 '<tt>llvm.umul.with.overflow.*</tt>' Intrinsics
7019 </a>
7020</h4>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7021
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7022<div>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7023
7024<h5>Syntax:</h5>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7025<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]7026 on any integer bit width.</p>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7027
7028<pre>
7029 declare {i16, i1} @llvm.umul.with.overflow.i16(i16 %a, i16 %b)
7030 declare {i32, i1} @llvm.umul.with.overflow.i32(i32 %a, i32 %b)
7031 declare {i64, i1} @llvm.umul.with.overflow.i64(i64 %a, i64 %b)
7032</pre>
7033
7034<h5>Overview:</h5>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7035<p>The '<tt>llvm.umul.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7036 a unsigned multiplication of the two arguments, and indicate whether an
7037 overflow occurred during the unsigned multiplication.</p>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7038
7039<h5>Arguments:</h5>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7040<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]7041 be of integer types of any bit width, but they must have the same bit
7042 width. The second element of the result structure must be of
7043 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
7044 undergo unsigned multiplication.</p>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7045
7046<h5>Semantics:</h5>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7047<p>The '<tt>llvm.umul.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7048 an unsigned multiplication of the two arguments. They return a structure
7049 &mdash; the first element of which is the multiplication, and the second
7050 element of which is a bit specifying if the unsigned multiplication resulted
7051 in an overflow.</p>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7052
7053<h5>Examples:</h5>
7054<pre>
7055 %res = call {i32, i1} @llvm.umul.with.overflow.i32(i32 %a, i32 %b)
7056 %sum = extractvalue {i32, i1} %res, 0
7057 %obit = extractvalue {i32, i1} %res, 1
7058 br i1 %obit, label %overflow, label %normal
7059</pre>
7060
7061</div>
7062
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7063</div>
7064
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]7065<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7066<h3>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]7067 <a name="int_fp16">Half Precision Floating Point Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7068</h3>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]7069
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7070<div>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]7071
Chris Lattner0cec9c82010-03-15 04:12:21 +0000[diff] [blame]7072<p>Half precision floating point is a storage-only format. This means that it is
7073 a dense encoding (in memory) but does not support computation in the
7074 format.</p>
Chris Lattner82c3dc62010-03-14 23:03:31 +0000[diff] [blame]7075
Chris Lattner0cec9c82010-03-15 04:12:21 +0000[diff] [blame]7076<p>This means that code must first load the half-precision floating point
Chris Lattner82c3dc62010-03-14 23:03:31 +0000[diff] [blame]7077 value as an i16, then convert it to float with <a
7078 href="#int_convert_from_fp16"><tt>llvm.convert.from.fp16</tt></a>.
7079 Computation can then be performed on the float value (including extending to
Chris Lattner0cec9c82010-03-15 04:12:21 +0000[diff] [blame]7080 double etc). To store the value back to memory, it is first converted to
7081 float if needed, then converted to i16 with
Chris Lattner82c3dc62010-03-14 23:03:31 +0000[diff] [blame]7082 <a href="#int_convert_to_fp16"><tt>llvm.convert.to.fp16</tt></a>, then
7083 storing as an i16 value.</p>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]7084
7085<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7086<h4>
7087 <a name="int_convert_to_fp16">
7088 '<tt>llvm.convert.to.fp16</tt>' Intrinsic
7089 </a>
7090</h4>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]7091
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7092<div>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]7093
7094<h5>Syntax:</h5>
7095<pre>
7096 declare i16 @llvm.convert.to.fp16(f32 %a)
7097</pre>
7098
7099<h5>Overview:</h5>
7100<p>The '<tt>llvm.convert.to.fp16</tt>' intrinsic function performs
7101 a conversion from single precision floating point format to half precision
7102 floating point format.</p>
7103
7104<h5>Arguments:</h5>
7105<p>The intrinsic function contains single argument - the value to be
7106 converted.</p>
7107
7108<h5>Semantics:</h5>
7109<p>The '<tt>llvm.convert.to.fp16</tt>' intrinsic function performs
7110 a conversion from single precision floating point format to half precision
Chris Lattner0cec9c82010-03-15 04:12:21 +0000[diff] [blame]7111 floating point format. The return value is an <tt>i16</tt> which
Chris Lattner82c3dc62010-03-14 23:03:31 +0000[diff] [blame]7112 contains the converted number.</p>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]7113
7114<h5>Examples:</h5>
7115<pre>
7116 %res = call i16 @llvm.convert.to.fp16(f32 %a)
7117 store i16 %res, i16* @x, align 2
7118</pre>
7119
7120</div>
7121
7122<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7123<h4>
7124 <a name="int_convert_from_fp16">
7125 '<tt>llvm.convert.from.fp16</tt>' Intrinsic
7126 </a>
7127</h4>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]7128
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7129<div>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]7130
7131<h5>Syntax:</h5>
7132<pre>
7133 declare f32 @llvm.convert.from.fp16(i16 %a)
7134</pre>
7135
7136<h5>Overview:</h5>
7137<p>The '<tt>llvm.convert.from.fp16</tt>' intrinsic function performs
7138 a conversion from half precision floating point format to single precision
7139 floating point format.</p>
7140
7141<h5>Arguments:</h5>
7142<p>The intrinsic function contains single argument - the value to be
7143 converted.</p>
7144
7145<h5>Semantics:</h5>
7146<p>The '<tt>llvm.convert.from.fp16</tt>' intrinsic function performs a
Chris Lattner0cec9c82010-03-15 04:12:21 +0000[diff] [blame]7147 conversion from half single precision floating point format to single
Chris Lattner82c3dc62010-03-14 23:03:31 +0000[diff] [blame]7148 precision floating point format. The input half-float value is represented by
7149 an <tt>i16</tt> value.</p>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]7150
7151<h5>Examples:</h5>
7152<pre>
7153 %a = load i16* @x, align 2
7154 %res = call f32 @llvm.convert.from.fp16(i16 %a)
7155</pre>
7156
7157</div>
7158
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7159</div>
7160
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]7161<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7162<h3>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]7163 <a name="int_debugger">Debugger Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7164</h3>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]7165
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7166<div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]7167
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7168<p>The LLVM debugger intrinsics (which all start with <tt>llvm.dbg.</tt>
7169 prefix), are described in
7170 the <a href="SourceLevelDebugging.html#format_common_intrinsics">LLVM Source
7171 Level Debugging</a> document.</p>
7172
7173</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]7174
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]7175<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7176<h3>
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]7177 <a name="int_eh">Exception Handling Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7178</h3>
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]7179
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7180<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7181
7182<p>The LLVM exception handling intrinsics (which all start with
7183 <tt>llvm.eh.</tt> prefix), are described in
7184 the <a href="ExceptionHandling.html#format_common_intrinsics">LLVM Exception
7185 Handling</a> document.</p>
7186
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]7187</div>
7188
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]7189<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7190<h3>
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]7191 <a name="int_trampoline">Trampoline Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7192</h3>
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]7193
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7194<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7195
7196<p>This intrinsic makes it possible to excise one parameter, marked with
Dan Gohmanff235352010-07-02 23:18:08 +0000[diff] [blame]7197 the <a href="#nest"><tt>nest</tt></a> attribute, from a function.
7198 The result is a callable
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7199 function pointer lacking the nest parameter - the caller does not need to
7200 provide a value for it. Instead, the value to use is stored in advance in a
7201 "trampoline", a block of memory usually allocated on the stack, which also
7202 contains code to splice the nest value into the argument list. This is used
7203 to implement the GCC nested function address extension.</p>
7204
7205<p>For example, if the function is
7206 <tt>i32 f(i8* nest %c, i32 %x, i32 %y)</tt> then the resulting function
7207 pointer has signature <tt>i32 (i32, i32)*</tt>. It can be created as
7208 follows:</p>
7209
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]7210<pre class="doc_code">
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]7211 %tramp = alloca [10 x i8], align 4 ; size and alignment only correct for X86
7212 %tramp1 = getelementptr [10 x i8]* %tramp, i32 0, i32 0
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7213 %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]7214 %fp = bitcast i8* %p to i32 (i32, i32)*
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]7215</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7216
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7217<p>The call <tt>%val = call i32 %fp(i32 %x, i32 %y)</tt> is then equivalent
7218 to <tt>%val = call i32 %f(i8* %nval, i32 %x, i32 %y)</tt>.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7219
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]7220<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7221<h4>
7222 <a name="int_it">
7223 '<tt>llvm.init.trampoline</tt>' Intrinsic
7224 </a>
7225</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7226
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7227<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7228
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]7229<h5>Syntax:</h5>
7230<pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7231 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]7232</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7233
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]7234<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7235<p>This fills the memory pointed to by <tt>tramp</tt> with code and returns a
7236 function pointer suitable for executing it.</p>
7237
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]7238<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7239<p>The <tt>llvm.init.trampoline</tt> intrinsic takes three arguments, all
7240 pointers. The <tt>tramp</tt> argument must point to a sufficiently large and
7241 sufficiently aligned block of memory; this memory is written to by the
7242 intrinsic. Note that the size and the alignment are target-specific - LLVM
7243 currently provides no portable way of determining them, so a front-end that
7244 generates this intrinsic needs to have some target-specific knowledge.
7245 The <tt>func</tt> argument must hold a function bitcast to
7246 an <tt>i8*</tt>.</p>
7247
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]7248<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7249<p>The block of memory pointed to by <tt>tramp</tt> is filled with target
7250 dependent code, turning it into a function. A pointer to this function is
7251 returned, but needs to be bitcast to an <a href="#int_trampoline">appropriate
7252 function pointer type</a> before being called. The new function's signature
7253 is the same as that of <tt>func</tt> with any arguments marked with
7254 the <tt>nest</tt> attribute removed. At most one such <tt>nest</tt> argument
7255 is allowed, and it must be of pointer type. Calling the new function is
7256 equivalent to calling <tt>func</tt> with the same argument list, but
7257 with <tt>nval</tt> used for the missing <tt>nest</tt> argument. If, after
7258 calling <tt>llvm.init.trampoline</tt>, the memory pointed to
7259 by <tt>tramp</tt> is modified, then the effect of any later call to the
7260 returned function pointer is undefined.</p>
7261
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]7262</div>
7263
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7264</div>
7265
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]7266<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7267<h3>
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7268 <a name="int_atomics">Atomic Operations and Synchronization Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7269</h3>
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7270
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7271<div>
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7272
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7273<p>These intrinsic functions expand the "universal IR" of LLVM to represent
7274 hardware constructs for atomic operations and memory synchronization. This
7275 provides an interface to the hardware, not an interface to the programmer. It
7276 is aimed at a low enough level to allow any programming models or APIs
7277 (Application Programming Interfaces) which need atomic behaviors to map
7278 cleanly onto it. It is also modeled primarily on hardware behavior. Just as
7279 hardware provides a "universal IR" for source languages, it also provides a
7280 starting point for developing a "universal" atomic operation and
7281 synchronization IR.</p>
7282
7283<p>These do <em>not</em> form an API such as high-level threading libraries,
7284 software transaction memory systems, atomic primitives, and intrinsic
7285 functions as found in BSD, GNU libc, atomic_ops, APR, and other system and
7286 application libraries. The hardware interface provided by LLVM should allow
7287 a clean implementation of all of these APIs and parallel programming models.
7288 No one model or paradigm should be selected above others unless the hardware
7289 itself ubiquitously does so.</p>
7290
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7291<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7292<h4>
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7293 <a name="int_memory_barrier">'<tt>llvm.memory.barrier</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7294</h4>
7295
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7296<div>
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7297<h5>Syntax:</h5>
7298<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7299 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]7300</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7301
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7302<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7303<p>The <tt>llvm.memory.barrier</tt> intrinsic guarantees ordering between
7304 specific pairs of memory access types.</p>
7305
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7306<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7307<p>The <tt>llvm.memory.barrier</tt> intrinsic requires five boolean arguments.
7308 The first four arguments enables a specific barrier as listed below. The
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]7309 fifth argument specifies that the barrier applies to io or device or uncached
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7310 memory.</p>
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7311
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7312<ul>
7313 <li><tt>ll</tt>: load-load barrier</li>
7314 <li><tt>ls</tt>: load-store barrier</li>
7315 <li><tt>sl</tt>: store-load barrier</li>
7316 <li><tt>ss</tt>: store-store barrier</li>
7317 <li><tt>device</tt>: barrier applies to device and uncached memory also.</li>
7318</ul>
7319
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7320<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7321<p>This intrinsic causes the system to enforce some ordering constraints upon
7322 the loads and stores of the program. This barrier does not
7323 indicate <em>when</em> any events will occur, it only enforces
7324 an <em>order</em> in which they occur. For any of the specified pairs of load
7325 and store operations (f.ex. load-load, or store-load), all of the first
7326 operations preceding the barrier will complete before any of the second
7327 operations succeeding the barrier begin. Specifically the semantics for each
7328 pairing is as follows:</p>
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7329
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7330<ul>
7331 <li><tt>ll</tt>: All loads before the barrier must complete before any load
7332 after the barrier begins.</li>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]7333 <li><tt>ls</tt>: All loads before the barrier must complete before any
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7334 store after the barrier begins.</li>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]7335 <li><tt>ss</tt>: All stores before the barrier must complete before any
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7336 store after the barrier begins.</li>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]7337 <li><tt>sl</tt>: All stores before the barrier must complete before any
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7338 load after the barrier begins.</li>
7339</ul>
7340
7341<p>These semantics are applied with a logical "and" behavior when more than one
7342 is enabled in a single memory barrier intrinsic.</p>
7343
7344<p>Backends may implement stronger barriers than those requested when they do
7345 not support as fine grained a barrier as requested. Some architectures do
7346 not need all types of barriers and on such architectures, these become
7347 noops.</p>
7348
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7349<h5>Example:</h5>
7350<pre>
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]7351%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
7352%ptr = bitcast i8* %mallocP to i32*
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7353 store i32 4, %ptr
7354
7355%result1 = load i32* %ptr <i>; yields {i32}:result1 = 4</i>
Evan Cheng0b0669a2011-06-29 17:14:00 +0000[diff] [blame]7356 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]7357 <i>; guarantee the above finishes</i>
7358 store i32 8, %ptr <i>; before this begins</i>
7359</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7360
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7361</div>
7362
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7363<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7364<h4>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7365 <a name="int_atomic_cmp_swap">'<tt>llvm.atomic.cmp.swap.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7366</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7367
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7368<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7369
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7370<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7371<p>This is an overloaded intrinsic. You can use <tt>llvm.atomic.cmp.swap</tt> on
7372 any integer bit width and for different address spaces. Not all targets
7373 support all bit widths however.</p>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7374
7375<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7376 declare i8 @llvm.atomic.cmp.swap.i8.p0i8(i8* &lt;ptr&gt;, i8 &lt;cmp&gt;, i8 &lt;val&gt;)
7377 declare i16 @llvm.atomic.cmp.swap.i16.p0i16(i16* &lt;ptr&gt;, i16 &lt;cmp&gt;, i16 &lt;val&gt;)
7378 declare i32 @llvm.atomic.cmp.swap.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;cmp&gt;, i32 &lt;val&gt;)
7379 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]7380</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7381
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7382<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7383<p>This loads a value in memory and compares it to a given value. If they are
7384 equal, it stores a new value into the memory.</p>
7385
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7386<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7387<p>The <tt>llvm.atomic.cmp.swap</tt> intrinsic takes three arguments. The result
7388 as well as both <tt>cmp</tt> and <tt>val</tt> must be integer values with the
7389 same bit width. The <tt>ptr</tt> argument must be a pointer to a value of
7390 this integer type. While any bit width integer may be used, targets may only
7391 lower representations they support in hardware.</p>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7392
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7393<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7394<p>This entire intrinsic must be executed atomically. It first loads the value
7395 in memory pointed to by <tt>ptr</tt> and compares it with the
7396 value <tt>cmp</tt>. If they are equal, <tt>val</tt> is stored into the
7397 memory. The loaded value is yielded in all cases. This provides the
7398 equivalent of an atomic compare-and-swap operation within the SSA
7399 framework.</p>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7400
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7401<h5>Examples:</h5>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7402<pre>
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]7403%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
7404%ptr = bitcast i8* %mallocP to i32*
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7405 store i32 4, %ptr
7406
7407%val1 = add i32 4, 4
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7408%result1 = call i32 @llvm.atomic.cmp.swap.i32.p0i32(i32* %ptr, i32 4, %val1)
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7409 <i>; yields {i32}:result1 = 4</i>
7410%stored1 = icmp eq i32 %result1, 4 <i>; yields {i1}:stored1 = true</i>
7411%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 8</i>
7412
7413%val2 = add i32 1, 1
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7414%result2 = call i32 @llvm.atomic.cmp.swap.i32.p0i32(i32* %ptr, i32 5, %val2)
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7415 <i>; yields {i32}:result2 = 8</i>
7416%stored2 = icmp eq i32 %result2, 5 <i>; yields {i1}:stored2 = false</i>
7417
7418%memval2 = load i32* %ptr <i>; yields {i32}:memval2 = 8</i>
7419</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7420
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7421</div>
7422
7423<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7424<h4>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7425 <a name="int_atomic_swap">'<tt>llvm.atomic.swap.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7426</h4>
7427
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7428<div>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7429<h5>Syntax:</h5>
7430
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7431<p>This is an overloaded intrinsic. You can use <tt>llvm.atomic.swap</tt> on any
7432 integer bit width. Not all targets support all bit widths however.</p>
7433
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7434<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7435 declare i8 @llvm.atomic.swap.i8.p0i8(i8* &lt;ptr&gt;, i8 &lt;val&gt;)
7436 declare i16 @llvm.atomic.swap.i16.p0i16(i16* &lt;ptr&gt;, i16 &lt;val&gt;)
7437 declare i32 @llvm.atomic.swap.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;val&gt;)
7438 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]7439</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7440
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7441<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7442<p>This intrinsic loads the value stored in memory at <tt>ptr</tt> and yields
7443 the value from memory. It then stores the value in <tt>val</tt> in the memory
7444 at <tt>ptr</tt>.</p>
7445
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7446<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7447<p>The <tt>llvm.atomic.swap</tt> intrinsic takes two arguments. Both
7448 the <tt>val</tt> argument and the result must be integers of the same bit
7449 width. The first argument, <tt>ptr</tt>, must be a pointer to a value of this
7450 integer type. The targets may only lower integer representations they
7451 support.</p>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7452
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7453<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7454<p>This intrinsic loads the value pointed to by <tt>ptr</tt>, yields it, and
7455 stores <tt>val</tt> back into <tt>ptr</tt> atomically. This provides the
7456 equivalent of an atomic swap operation within the SSA framework.</p>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7457
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7458<h5>Examples:</h5>
7459<pre>
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]7460%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
7461%ptr = bitcast i8* %mallocP to i32*
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7462 store i32 4, %ptr
7463
7464%val1 = add i32 4, 4
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7465%result1 = call i32 @llvm.atomic.swap.i32.p0i32(i32* %ptr, i32 %val1)
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7466 <i>; yields {i32}:result1 = 4</i>
7467%stored1 = icmp eq i32 %result1, 4 <i>; yields {i1}:stored1 = true</i>
7468%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 8</i>
7469
7470%val2 = add i32 1, 1
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7471%result2 = call i32 @llvm.atomic.swap.i32.p0i32(i32* %ptr, i32 %val2)
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7472 <i>; yields {i32}:result2 = 8</i>
7473
7474%stored2 = icmp eq i32 %result2, 8 <i>; yields {i1}:stored2 = true</i>
7475%memval2 = load i32* %ptr <i>; yields {i32}:memval2 = 2</i>
7476</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7477
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7478</div>
7479
7480<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7481<h4>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7482 <a name="int_atomic_load_add">'<tt>llvm.atomic.load.add.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7483</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7484
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7485<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7486
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7487<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7488<p>This is an overloaded intrinsic. You can use <tt>llvm.atomic.load.add</tt> on
7489 any integer bit width. Not all targets support all bit widths however.</p>
7490
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7491<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7492 declare i8 @llvm.atomic.load.add.i8.p0i8(i8* &lt;ptr&gt;, i8 &lt;delta&gt;)
7493 declare i16 @llvm.atomic.load.add.i16.p0i16(i16* &lt;ptr&gt;, i16 &lt;delta&gt;)
7494 declare i32 @llvm.atomic.load.add.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;delta&gt;)
7495 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]7496</pre>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7497
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7498<h5>Overview:</h5>
7499<p>This intrinsic adds <tt>delta</tt> to the value stored in memory
7500 at <tt>ptr</tt>. It yields the original value at <tt>ptr</tt>.</p>
7501
7502<h5>Arguments:</h5>
7503<p>The intrinsic takes two arguments, the first a pointer to an integer value
7504 and the second an integer value. The result is also an integer value. These
7505 integer types can have any bit width, but they must all have the same bit
7506 width. The targets may only lower integer representations they support.</p>
7507
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7508<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7509<p>This intrinsic does a series of operations atomically. It first loads the
7510 value stored at <tt>ptr</tt>. It then adds <tt>delta</tt>, stores the result
7511 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]7512
7513<h5>Examples:</h5>
7514<pre>
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]7515%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
7516%ptr = bitcast i8* %mallocP to i32*
7517 store i32 4, %ptr
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7518%result1 = call i32 @llvm.atomic.load.add.i32.p0i32(i32* %ptr, i32 4)
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7519 <i>; yields {i32}:result1 = 4</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7520%result2 = call i32 @llvm.atomic.load.add.i32.p0i32(i32* %ptr, i32 2)
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7521 <i>; yields {i32}:result2 = 8</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7522%result3 = call i32 @llvm.atomic.load.add.i32.p0i32(i32* %ptr, i32 5)
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7523 <i>; yields {i32}:result3 = 10</i>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7524%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 15</i>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7525</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7526
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]7527</div>
7528
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7529<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7530<h4>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7531 <a name="int_atomic_load_sub">'<tt>llvm.atomic.load.sub.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7532</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7533
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7534<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7535
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7536<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7537<p>This is an overloaded intrinsic. You can use <tt>llvm.atomic.load.sub</tt> on
7538 any integer bit width and for different address spaces. Not all targets
7539 support all bit widths however.</p>
7540
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7541<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7542 declare i8 @llvm.atomic.load.sub.i8.p0i32(i8* &lt;ptr&gt;, i8 &lt;delta&gt;)
7543 declare i16 @llvm.atomic.load.sub.i16.p0i32(i16* &lt;ptr&gt;, i16 &lt;delta&gt;)
7544 declare i32 @llvm.atomic.load.sub.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;delta&gt;)
7545 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]7546</pre>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7547
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7548<h5>Overview:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]7549<p>This intrinsic subtracts <tt>delta</tt> to the value stored in memory at
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7550 <tt>ptr</tt>. It yields the original value at <tt>ptr</tt>.</p>
7551
7552<h5>Arguments:</h5>
7553<p>The intrinsic takes two arguments, the first a pointer to an integer value
7554 and the second an integer value. The result is also an integer value. These
7555 integer types can have any bit width, but they must all have the same bit
7556 width. The targets may only lower integer representations they support.</p>
7557
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7558<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7559<p>This intrinsic does a series of operations atomically. It first loads the
7560 value stored at <tt>ptr</tt>. It then subtracts <tt>delta</tt>, stores the
7561 result to <tt>ptr</tt>. It yields the original value stored
7562 at <tt>ptr</tt>.</p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7563
7564<h5>Examples:</h5>
7565<pre>
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]7566%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
7567%ptr = bitcast i8* %mallocP to i32*
7568 store i32 8, %ptr
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7569%result1 = call i32 @llvm.atomic.load.sub.i32.p0i32(i32* %ptr, i32 4)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7570 <i>; yields {i32}:result1 = 8</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7571%result2 = call i32 @llvm.atomic.load.sub.i32.p0i32(i32* %ptr, i32 2)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7572 <i>; yields {i32}:result2 = 4</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7573%result3 = call i32 @llvm.atomic.load.sub.i32.p0i32(i32* %ptr, i32 5)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7574 <i>; yields {i32}:result3 = 2</i>
7575%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = -3</i>
7576</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7577
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7578</div>
7579
7580<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7581<h4>
7582 <a name="int_atomic_load_and">
7583 '<tt>llvm.atomic.load.and.*</tt>' Intrinsic
7584 </a>
7585 <br>
7586 <a name="int_atomic_load_nand">
7587 '<tt>llvm.atomic.load.nand.*</tt>' Intrinsic
7588 </a>
7589 <br>
7590 <a name="int_atomic_load_or">
7591 '<tt>llvm.atomic.load.or.*</tt>' Intrinsic
7592 </a>
7593 <br>
7594 <a name="int_atomic_load_xor">
7595 '<tt>llvm.atomic.load.xor.*</tt>' Intrinsic
7596 </a>
7597</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7598
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7599<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7600
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7601<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7602<p>These are overloaded intrinsics. You can
7603 use <tt>llvm.atomic.load_and</tt>, <tt>llvm.atomic.load_nand</tt>,
7604 <tt>llvm.atomic.load_or</tt>, and <tt>llvm.atomic.load_xor</tt> on any integer
7605 bit width and for different address spaces. Not all targets support all bit
7606 widths however.</p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7607
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7608<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7609 declare i8 @llvm.atomic.load.and.i8.p0i8(i8* &lt;ptr&gt;, i8 &lt;delta&gt;)
7610 declare i16 @llvm.atomic.load.and.i16.p0i16(i16* &lt;ptr&gt;, i16 &lt;delta&gt;)
7611 declare i32 @llvm.atomic.load.and.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;delta&gt;)
7612 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]7613</pre>
7614
7615<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7616 declare i8 @llvm.atomic.load.or.i8.p0i8(i8* &lt;ptr&gt;, i8 &lt;delta&gt;)
7617 declare i16 @llvm.atomic.load.or.i16.p0i16(i16* &lt;ptr&gt;, i16 &lt;delta&gt;)
7618 declare i32 @llvm.atomic.load.or.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;delta&gt;)
7619 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]7620</pre>
7621
7622<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7623 declare i8 @llvm.atomic.load.nand.i8.p0i32(i8* &lt;ptr&gt;, i8 &lt;delta&gt;)
7624 declare i16 @llvm.atomic.load.nand.i16.p0i32(i16* &lt;ptr&gt;, i16 &lt;delta&gt;)
7625 declare i32 @llvm.atomic.load.nand.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;delta&gt;)
7626 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]7627</pre>
7628
7629<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7630 declare i8 @llvm.atomic.load.xor.i8.p0i32(i8* &lt;ptr&gt;, i8 &lt;delta&gt;)
7631 declare i16 @llvm.atomic.load.xor.i16.p0i32(i16* &lt;ptr&gt;, i16 &lt;delta&gt;)
7632 declare i32 @llvm.atomic.load.xor.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;delta&gt;)
7633 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]7634</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7635
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7636<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7637<p>These intrinsics bitwise the operation (and, nand, or, xor) <tt>delta</tt> to
7638 the value stored in memory at <tt>ptr</tt>. It yields the original value
7639 at <tt>ptr</tt>.</p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7640
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7641<h5>Arguments:</h5>
7642<p>These intrinsics take two arguments, the first a pointer to an integer value
7643 and the second an integer value. The result is also an integer value. These
7644 integer types can have any bit width, but they must all have the same bit
7645 width. The targets may only lower integer representations they support.</p>
7646
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7647<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7648<p>These intrinsics does a series of operations atomically. They first load the
7649 value stored at <tt>ptr</tt>. They then do the bitwise
7650 operation <tt>delta</tt>, store the result to <tt>ptr</tt>. They yield the
7651 original value stored at <tt>ptr</tt>.</p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7652
7653<h5>Examples:</h5>
7654<pre>
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]7655%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
7656%ptr = bitcast i8* %mallocP to i32*
7657 store i32 0x0F0F, %ptr
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7658%result0 = call i32 @llvm.atomic.load.nand.i32.p0i32(i32* %ptr, i32 0xFF)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7659 <i>; yields {i32}:result0 = 0x0F0F</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7660%result1 = call i32 @llvm.atomic.load.and.i32.p0i32(i32* %ptr, i32 0xFF)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7661 <i>; yields {i32}:result1 = 0xFFFFFFF0</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7662%result2 = call i32 @llvm.atomic.load.or.i32.p0i32(i32* %ptr, i32 0F)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7663 <i>; yields {i32}:result2 = 0xF0</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7664%result3 = call i32 @llvm.atomic.load.xor.i32.p0i32(i32* %ptr, i32 0F)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7665 <i>; yields {i32}:result3 = FF</i>
7666%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = F0</i>
7667</pre>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7668
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7669</div>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7670
7671<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7672<h4>
7673 <a name="int_atomic_load_max">
7674 '<tt>llvm.atomic.load.max.*</tt>' Intrinsic
7675 </a>
7676 <br>
7677 <a name="int_atomic_load_min">
7678 '<tt>llvm.atomic.load.min.*</tt>' Intrinsic
7679 </a>
7680 <br>
7681 <a name="int_atomic_load_umax">
7682 '<tt>llvm.atomic.load.umax.*</tt>' Intrinsic
7683 </a>
7684 <br>
7685 <a name="int_atomic_load_umin">
7686 '<tt>llvm.atomic.load.umin.*</tt>' Intrinsic
7687 </a>
7688</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7689
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7690<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7691
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7692<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7693<p>These are overloaded intrinsics. You can use <tt>llvm.atomic.load_max</tt>,
7694 <tt>llvm.atomic.load_min</tt>, <tt>llvm.atomic.load_umax</tt>, and
7695 <tt>llvm.atomic.load_umin</tt> on any integer bit width and for different
7696 address spaces. Not all targets support all bit widths however.</p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7697
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7698<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7699 declare i8 @llvm.atomic.load.max.i8.p0i8(i8* &lt;ptr&gt;, i8 &lt;delta&gt;)
7700 declare i16 @llvm.atomic.load.max.i16.p0i16(i16* &lt;ptr&gt;, i16 &lt;delta&gt;)
7701 declare i32 @llvm.atomic.load.max.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;delta&gt;)
7702 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]7703</pre>
7704
7705<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7706 declare i8 @llvm.atomic.load.min.i8.p0i8(i8* &lt;ptr&gt;, i8 &lt;delta&gt;)
7707 declare i16 @llvm.atomic.load.min.i16.p0i16(i16* &lt;ptr&gt;, i16 &lt;delta&gt;)
7708 declare i32 @llvm.atomic.load.min.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;delta&gt;)
7709 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]7710</pre>
7711
7712<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7713 declare i8 @llvm.atomic.load.umax.i8.p0i8(i8* &lt;ptr&gt;, i8 &lt;delta&gt;)
7714 declare i16 @llvm.atomic.load.umax.i16.p0i16(i16* &lt;ptr&gt;, i16 &lt;delta&gt;)
7715 declare i32 @llvm.atomic.load.umax.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;delta&gt;)
7716 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]7717</pre>
7718
7719<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7720 declare i8 @llvm.atomic.load.umin.i8.p0i8(i8* &lt;ptr&gt;, i8 &lt;delta&gt;)
7721 declare i16 @llvm.atomic.load.umin.i16.p0i16(i16* &lt;ptr&gt;, i16 &lt;delta&gt;)
7722 declare i32 @llvm.atomic.load.umin.i32.p0i32(i32* &lt;ptr&gt;, i32 &lt;delta&gt;)
7723 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]7724</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7725
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7726<h5>Overview:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]7727<p>These intrinsics takes the signed or unsigned minimum or maximum of
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7728 <tt>delta</tt> and the value stored in memory at <tt>ptr</tt>. It yields the
7729 original value at <tt>ptr</tt>.</p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7730
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7731<h5>Arguments:</h5>
7732<p>These intrinsics take two arguments, the first a pointer to an integer value
7733 and the second an integer value. The result is also an integer value. These
7734 integer types can have any bit width, but they must all have the same bit
7735 width. The targets may only lower integer representations they support.</p>
7736
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7737<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7738<p>These intrinsics does a series of operations atomically. They first load the
7739 value stored at <tt>ptr</tt>. They then do the signed or unsigned min or
7740 max <tt>delta</tt> and the value, store the result to <tt>ptr</tt>. They
7741 yield the original value stored at <tt>ptr</tt>.</p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7742
7743<h5>Examples:</h5>
7744<pre>
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]7745%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
7746%ptr = bitcast i8* %mallocP to i32*
7747 store i32 7, %ptr
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7748%result0 = call i32 @llvm.atomic.load.min.i32.p0i32(i32* %ptr, i32 -2)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7749 <i>; yields {i32}:result0 = 7</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7750%result1 = call i32 @llvm.atomic.load.max.i32.p0i32(i32* %ptr, i32 8)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7751 <i>; yields {i32}:result1 = -2</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7752%result2 = call i32 @llvm.atomic.load.umin.i32.p0i32(i32* %ptr, i32 10)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7753 <i>; yields {i32}:result2 = 8</i>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7754%result3 = call i32 @llvm.atomic.load.umax.i32.p0i32(i32* %ptr, i32 30)
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7755 <i>; yields {i32}:result3 = 8</i>
7756%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 30</i>
7757</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7758
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]7759</div>
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7760
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7761</div>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7762
7763<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7764<h3>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7765 <a name="int_memorymarkers">Memory Use Markers</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7766</h3>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7767
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7768<div>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7769
7770<p>This class of intrinsics exists to information about the lifetime of memory
7771 objects and ranges where variables are immutable.</p>
7772
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7773<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7774<h4>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7775 <a name="int_lifetime_start">'<tt>llvm.lifetime.start</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7776</h4>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7777
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7778<div>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7779
7780<h5>Syntax:</h5>
7781<pre>
7782 declare void @llvm.lifetime.start(i64 &lt;size&gt;, i8* nocapture &lt;ptr&gt;)
7783</pre>
7784
7785<h5>Overview:</h5>
7786<p>The '<tt>llvm.lifetime.start</tt>' intrinsic specifies the start of a memory
7787 object's lifetime.</p>
7788
7789<h5>Arguments:</h5>
Nick Lewycky321333e2009-10-13 07:57:33 +0000[diff] [blame]7790<p>The first argument is a constant integer representing the size of the
7791 object, or -1 if it is variable sized. The second argument is a pointer to
7792 the object.</p>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7793
7794<h5>Semantics:</h5>
7795<p>This intrinsic indicates that before this point in the code, the value of the
7796 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]7797 never be used and has an undefined value. A load from the pointer that
7798 precedes this intrinsic can be replaced with
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7799 <tt>'<a href="#undefvalues">undef</a>'</tt>.</p>
7800
7801</div>
7802
7803<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7804<h4>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7805 <a name="int_lifetime_end">'<tt>llvm.lifetime.end</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7806</h4>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7807
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7808<div>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7809
7810<h5>Syntax:</h5>
7811<pre>
7812 declare void @llvm.lifetime.end(i64 &lt;size&gt;, i8* nocapture &lt;ptr&gt;)
7813</pre>
7814
7815<h5>Overview:</h5>
7816<p>The '<tt>llvm.lifetime.end</tt>' intrinsic specifies the end of a memory
7817 object's lifetime.</p>
7818
7819<h5>Arguments:</h5>
Nick Lewycky321333e2009-10-13 07:57:33 +0000[diff] [blame]7820<p>The first argument is a constant integer representing the size of the
7821 object, or -1 if it is variable sized. The second argument is a pointer to
7822 the object.</p>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7823
7824<h5>Semantics:</h5>
7825<p>This intrinsic indicates that after this point in the code, the value of the
7826 memory pointed to by <tt>ptr</tt> is dead. This means that it is known to
7827 never be used and has an undefined value. Any stores into the memory object
7828 following this intrinsic may be removed as dead.
7829
7830</div>
7831
7832<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7833<h4>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7834 <a name="int_invariant_start">'<tt>llvm.invariant.start</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7835</h4>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7836
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7837<div>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7838
7839<h5>Syntax:</h5>
7840<pre>
Nick Lewycky29b6cb42010-11-30 04:13:41 +0000[diff] [blame]7841 declare {}* @llvm.invariant.start(i64 &lt;size&gt;, i8* nocapture &lt;ptr&gt;)
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7842</pre>
7843
7844<h5>Overview:</h5>
7845<p>The '<tt>llvm.invariant.start</tt>' intrinsic specifies that the contents of
7846 a memory object will not change.</p>
7847
7848<h5>Arguments:</h5>
Nick Lewycky321333e2009-10-13 07:57:33 +0000[diff] [blame]7849<p>The first argument is a constant integer representing the size of the
7850 object, or -1 if it is variable sized. The second argument is a pointer to
7851 the object.</p>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7852
7853<h5>Semantics:</h5>
7854<p>This intrinsic indicates that until an <tt>llvm.invariant.end</tt> that uses
7855 the return value, the referenced memory location is constant and
7856 unchanging.</p>
7857
7858</div>
7859
7860<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7861<h4>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7862 <a name="int_invariant_end">'<tt>llvm.invariant.end</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7863</h4>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7864
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7865<div>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7866
7867<h5>Syntax:</h5>
7868<pre>
7869 declare void @llvm.invariant.end({}* &lt;start&gt;, i64 &lt;size&gt;, i8* nocapture &lt;ptr&gt;)
7870</pre>
7871
7872<h5>Overview:</h5>
7873<p>The '<tt>llvm.invariant.end</tt>' intrinsic specifies that the contents of
7874 a memory object are mutable.</p>
7875
7876<h5>Arguments:</h5>
7877<p>The first argument is the matching <tt>llvm.invariant.start</tt> intrinsic.
Nick Lewycky321333e2009-10-13 07:57:33 +0000[diff] [blame]7878 The second argument is a constant integer representing the size of the
7879 object, or -1 if it is variable sized and the third argument is a pointer
7880 to the object.</p>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]7881
7882<h5>Semantics:</h5>
7883<p>This intrinsic indicates that the memory is mutable again.</p>
7884
7885</div>
7886
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7887</div>
7888
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]7889<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7890<h3>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]7891 <a name="int_general">General Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7892</h3>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]7893
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7894<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7895
7896<p>This class of intrinsics is designed to be generic and has no specific
7897 purpose.</p>
7898
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]7899<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7900<h4>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]7901 <a name="int_var_annotation">'<tt>llvm.var.annotation</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7902</h4>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]7903
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7904<div>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]7905
7906<h5>Syntax:</h5>
7907<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7908 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]7909</pre>
7910
7911<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7912<p>The '<tt>llvm.var.annotation</tt>' intrinsic.</p>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]7913
7914<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7915<p>The first argument is a pointer to a value, the second is a pointer to a
7916 global string, the third is a pointer to a global string which is the source
7917 file name, and the last argument is the line number.</p>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]7918
7919<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7920<p>This intrinsic allows annotation of local variables with arbitrary strings.
7921 This can be useful for special purpose optimizations that want to look for
7922 these annotations. These have no other defined use, they are ignored by code
7923 generation and optimization.</p>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]7924
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]7925</div>
7926
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]7927<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7928<h4>
Tanya Lattnere1a8da02007-09-21 23:57:59 +0000[diff] [blame]7929 <a name="int_annotation">'<tt>llvm.annotation.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7930</h4>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]7931
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7932<div>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]7933
7934<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7935<p>This is an overloaded intrinsic. You can use '<tt>llvm.annotation</tt>' on
7936 any integer bit width.</p>
7937
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]7938<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7939 declare i8 @llvm.annotation.i8(i8 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt;)
7940 declare i16 @llvm.annotation.i16(i16 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt;)
7941 declare i32 @llvm.annotation.i32(i32 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt;)
7942 declare i64 @llvm.annotation.i64(i64 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt;)
7943 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]7944</pre>
7945
7946<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7947<p>The '<tt>llvm.annotation</tt>' intrinsic.</p>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]7948
7949<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7950<p>The first argument is an integer value (result of some expression), the
7951 second is a pointer to a global string, the third is a pointer to a global
7952 string which is the source file name, and the last argument is the line
7953 number. It returns the value of the first argument.</p>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]7954
7955<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7956<p>This intrinsic allows annotations to be put on arbitrary expressions with
7957 arbitrary strings. This can be useful for special purpose optimizations that
7958 want to look for these annotations. These have no other defined use, they
7959 are ignored by code generation and optimization.</p>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]7960
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]7961</div>
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]7962
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]7963<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7964<h4>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]7965 <a name="int_trap">'<tt>llvm.trap</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7966</h4>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]7967
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7968<div>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]7969
7970<h5>Syntax:</h5>
7971<pre>
7972 declare void @llvm.trap()
7973</pre>
7974
7975<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7976<p>The '<tt>llvm.trap</tt>' intrinsic.</p>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]7977
7978<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7979<p>None.</p>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]7980
7981<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7982<p>This intrinsics is lowered to the target dependent trap instruction. If the
7983 target does not have a trap instruction, this intrinsic will be lowered to
7984 the call of the <tt>abort()</tt> function.</p>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]7985
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]7986</div>
7987
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]7988<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7989<h4>
Misha Brukmandccb0252008-11-22 23:55:29 +0000[diff] [blame]7990 <a name="int_stackprotector">'<tt>llvm.stackprotector</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7991</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7992
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7993<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7994
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]7995<h5>Syntax:</h5>
7996<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7997 declare void @llvm.stackprotector(i8* &lt;guard&gt;, i8** &lt;slot&gt;)
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]7998</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7999
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]8000<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8001<p>The <tt>llvm.stackprotector</tt> intrinsic takes the <tt>guard</tt> and
8002 stores it onto the stack at <tt>slot</tt>. The stack slot is adjusted to
8003 ensure that it is placed on the stack before local variables.</p>
8004
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]8005<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8006<p>The <tt>llvm.stackprotector</tt> intrinsic requires two pointer
8007 arguments. The first argument is the value loaded from the stack
8008 guard <tt>@__stack_chk_guard</tt>. The second variable is an <tt>alloca</tt>
8009 that has enough space to hold the value of the guard.</p>
8010
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]8011<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8012<p>This intrinsic causes the prologue/epilogue inserter to force the position of
8013 the <tt>AllocaInst</tt> stack slot to be before local variables on the
8014 stack. This is to ensure that if a local variable on the stack is
8015 overwritten, it will destroy the value of the guard. When the function exits,
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]8016 the guard on the stack is checked against the original guard. If they are
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8017 different, then the program aborts by calling the <tt>__stack_chk_fail()</tt>
8018 function.</p>
8019
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]8020</div>
8021
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]8022<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8023<h4>
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]8024 <a name="int_objectsize">'<tt>llvm.objectsize</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8025</h4>
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]8026
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8027<div>
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]8028
8029<h5>Syntax:</h5>
8030<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]8031 declare i32 @llvm.objectsize.i32(i8* &lt;object&gt;, i1 &lt;type&gt;)
8032 declare i64 @llvm.objectsize.i64(i8* &lt;object&gt;, i1 &lt;type&gt;)
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]8033</pre>
8034
8035<h5>Overview:</h5>
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]8036<p>The <tt>llvm.objectsize</tt> intrinsic is designed to provide information to
8037 the optimizers to determine at compile time whether a) an operation (like
8038 memcpy) will overflow a buffer that corresponds to an object, or b) that a
8039 runtime check for overflow isn't necessary. An object in this context means
8040 an allocation of a specific class, structure, array, or other object.</p>
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]8041
8042<h5>Arguments:</h5>
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]8043<p>The <tt>llvm.objectsize</tt> intrinsic takes two arguments. The first
Eric Christopher8295a0a2009-12-23 00:29:49 +0000[diff] [blame]8044 argument is a pointer to or into the <tt>object</tt>. The second argument
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]8045 is a boolean 0 or 1. This argument determines whether you want the
8046 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]8047 1, variables are not allowed.</p>
8048
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]8049<h5>Semantics:</h5>
8050<p>The <tt>llvm.objectsize</tt> intrinsic is lowered to either a constant
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]8051 representing the size of the object concerned, or <tt>i32/i64 -1 or 0</tt>,
8052 depending on the <tt>type</tt> argument, if the size cannot be determined at
8053 compile time.</p>
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]8054
8055</div>
8056
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8057</div>
8058
8059</div>
8060
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]8061<!-- *********************************************************************** -->
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]8062<hr>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]8063<address>
8064 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
Misha Brukman44408702008-12-11 17:34:48 +0000[diff] [blame]8065 src="http://jigsaw.w3.org/css-validator/images/vcss-blue" alt="Valid CSS"></a>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]8066 <a href="http://validator.w3.org/check/referer"><img
Misha Brukman44408702008-12-11 17:34:48 +0000[diff] [blame]8067 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]8068
8069 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
NAKAMURA Takumib9a33632011-04-09 02:13:37 +0000[diff] [blame]8070 <a href="http://llvm.org/">The LLVM Compiler Infrastructure</a><br>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]8071 Last modified: $Date$
8072</address>
Chris Lattnerc7d3ab32008-01-04 04:33:49 +0000[diff] [blame]8073
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]8074</body>
8075</html>