Gitiles
Code Review Sign In
gerrit-public.fairphone.software / platform / external / llvm / 43a699a112c7ac988e25111a3849fa1d73e3b51c / . / docs / LangRef.html
blob: a781992f89826c027120e325a3d3ad6ac8e9a799 [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.">
Daniel Dunbaradea4972012-04-19 20:20:34 +0000[diff] [blame]10 <link rel="stylesheet" href="_static/llvm.css" type="text/css">
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]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>
Bill Wendlingf7f06102011-10-11 06:41:28 +0000[diff] [blame]38 <li><a href="#linkage_external">'<tt>external</tt>' Linkage</a></li>
Bill Wendling987e7eb2009-07-20 02:41:50 +0000[diff] [blame]39 <li><a href="#linkage_dllimport">'<tt>dllimport</tt>' Linkage</a></li>
40 <li><a href="#linkage_dllexport">'<tt>dllexport</tt>' Linkage</a></li>
Bill Wendling3d10a5a2009-07-20 01:03:30 +0000[diff] [blame]41 </ol>
42 </li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]43 <li><a href="#callingconv">Calling Conventions</a></li>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]44 <li><a href="#namedtypes">Named Types</a></li>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]45 <li><a href="#globalvars">Global Variables</a></li>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]46 <li><a href="#functionstructure">Functions</a></li>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]47 <li><a href="#aliasstructure">Aliases</a></li>
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]48 <li><a href="#namedmetadatastructure">Named Metadata</a></li>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]49 <li><a href="#paramattrs">Parameter Attributes</a></li>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]50 <li><a href="#fnattrs">Function Attributes</a></li>
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame]51 <li><a href="#gc">Garbage Collector Names</a></li>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]52 <li><a href="#moduleasm">Module-Level Inline Assembly</a></li>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]53 <li><a href="#datalayout">Data Layout</a></li>
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]54 <li><a href="#pointeraliasing">Pointer Aliasing Rules</a></li>
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]55 <li><a href="#volatile">Volatile Memory Accesses</a></li>
Eli Friedman5b60e1b2011-07-20 21:35:53 +0000[diff] [blame]56 <li><a href="#memmodel">Memory Model for Concurrent Operations</a></li>
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]57 <li><a href="#ordering">Atomic Memory Ordering Constraints</a></li>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]58 </ol>
59 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]60 <li><a href="#typesystem">Type System</a>
61 <ol>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]62 <li><a href="#t_classifications">Type Classifications</a></li>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]63 <li><a href="#t_primitive">Primitive Types</a>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]64 <ol>
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]65 <li><a href="#t_integer">Integer Type</a></li>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]66 <li><a href="#t_floating">Floating Point Types</a></li>
Dale Johannesen21fe99b2010-10-01 00:48:59 +0000[diff] [blame]67 <li><a href="#t_x86mmx">X86mmx Type</a></li>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]68 <li><a href="#t_void">Void Type</a></li>
69 <li><a href="#t_label">Label Type</a></li>
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]70 <li><a href="#t_metadata">Metadata Type</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]71 </ol>
72 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]73 <li><a href="#t_derived">Derived Types</a>
74 <ol>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]75 <li><a href="#t_aggregate">Aggregate Types</a>
76 <ol>
77 <li><a href="#t_array">Array Type</a></li>
78 <li><a href="#t_struct">Structure Type</a></li>
Chris Lattner628ed392011-07-23 19:59:08 +0000[diff] [blame]79 <li><a href="#t_opaque">Opaque Structure Types</a></li>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]80 <li><a href="#t_vector">Vector Type</a></li>
81 </ol>
82 </li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]83 <li><a href="#t_function">Function Type</a></li>
84 <li><a href="#t_pointer">Pointer Type</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]85 </ol>
86 </li>
87 </ol>
88 </li>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]89 <li><a href="#constants">Constants</a>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]90 <ol>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]91 <li><a href="#simpleconstants">Simple Constants</a></li>
Chris Lattner70882792009-02-28 18:32:25 +0000[diff] [blame]92 <li><a href="#complexconstants">Complex Constants</a></li>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]93 <li><a href="#globalconstants">Global Variable and Function Addresses</a></li>
94 <li><a href="#undefvalues">Undefined Values</a></li>
Dan Gohmanbfb056d2011-12-06 03:18:47 +0000[diff] [blame]95 <li><a href="#poisonvalues">Poison 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>
Peter Collingbourne249d9532011-10-27 19:19:07 +0000[diff] [blame]103 <li><a href="#metadata">Metadata Nodes and Metadata Strings</a>
104 <ol>
105 <li><a href="#tbaa">'<tt>tbaa</tt>' Metadata</a></li>
Duncan Sands5e5c5f82012-04-14 12:36:06 +0000[diff] [blame]106 <li><a href="#fpmath">'<tt>fpmath</tt>' Metadata</a></li>
Rafael Espindola39dd3282012-03-24 00:14:51 +0000[diff] [blame]107 <li><a href="#range">'<tt>range</tt>' Metadata</a></li>
Peter Collingbourne249d9532011-10-27 19:19:07 +0000[diff] [blame]108 </ol>
109 </li>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]110 </ol>
111 </li>
Bill Wendlingb9d75a92012-02-11 11:59:36 +0000[diff] [blame]112 <li><a href="#module_flags">Module Flags Metadata</a>
113 <ol>
Bill Wendlingf7b367c2012-02-16 01:10:50 +0000[diff] [blame]114 <li><a href="#objc_gc_flags">Objective-C Garbage Collection Module Flags Metadata</a></li>
Bill Wendlingb9d75a92012-02-11 11:59:36 +0000[diff] [blame]115 </ol>
116 </li>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]117 <li><a href="#intrinsic_globals">Intrinsic Global Variables</a>
118 <ol>
119 <li><a href="#intg_used">The '<tt>llvm.used</tt>' Global Variable</a></li>
Chris Lattner401e10c2009-07-20 06:14:25 +0000[diff] [blame]120 <li><a href="#intg_compiler_used">The '<tt>llvm.compiler.used</tt>'
121 Global Variable</a></li>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]122 <li><a href="#intg_global_ctors">The '<tt>llvm.global_ctors</tt>'
123 Global Variable</a></li>
124 <li><a href="#intg_global_dtors">The '<tt>llvm.global_dtors</tt>'
125 Global Variable</a></li>
126 </ol>
127 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]128 <li><a href="#instref">Instruction Reference</a>
129 <ol>
130 <li><a href="#terminators">Terminator Instructions</a>
131 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]132 <li><a href="#i_ret">'<tt>ret</tt>' Instruction</a></li>
133 <li><a href="#i_br">'<tt>br</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]134 <li><a href="#i_switch">'<tt>switch</tt>' Instruction</a></li>
Chris Lattnerab21db72009-10-28 00:19:10 +0000[diff] [blame]135 <li><a href="#i_indirectbr">'<tt>indirectbr</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]136 <li><a href="#i_invoke">'<tt>invoke</tt>' Instruction</a></li>
Bill Wendlingdccc03b2011-07-31 06:30:59 +0000[diff] [blame]137 <li><a href="#i_resume">'<tt>resume</tt>' Instruction</a></li>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]138 <li><a href="#i_unreachable">'<tt>unreachable</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]139 </ol>
140 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]141 <li><a href="#binaryops">Binary Operations</a>
142 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]143 <li><a href="#i_add">'<tt>add</tt>' Instruction</a></li>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]144 <li><a href="#i_fadd">'<tt>fadd</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]145 <li><a href="#i_sub">'<tt>sub</tt>' Instruction</a></li>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]146 <li><a href="#i_fsub">'<tt>fsub</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]147 <li><a href="#i_mul">'<tt>mul</tt>' Instruction</a></li>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]148 <li><a href="#i_fmul">'<tt>fmul</tt>' Instruction</a></li>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]149 <li><a href="#i_udiv">'<tt>udiv</tt>' Instruction</a></li>
150 <li><a href="#i_sdiv">'<tt>sdiv</tt>' Instruction</a></li>
151 <li><a href="#i_fdiv">'<tt>fdiv</tt>' Instruction</a></li>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]152 <li><a href="#i_urem">'<tt>urem</tt>' Instruction</a></li>
153 <li><a href="#i_srem">'<tt>srem</tt>' Instruction</a></li>
154 <li><a href="#i_frem">'<tt>frem</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]155 </ol>
156 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]157 <li><a href="#bitwiseops">Bitwise Binary Operations</a>
158 <ol>
Reid Spencer8e11bf82007-02-02 13:57:07 +0000[diff] [blame]159 <li><a href="#i_shl">'<tt>shl</tt>' Instruction</a></li>
160 <li><a href="#i_lshr">'<tt>lshr</tt>' Instruction</a></li>
161 <li><a href="#i_ashr">'<tt>ashr</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]162 <li><a href="#i_and">'<tt>and</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]163 <li><a href="#i_or">'<tt>or</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]164 <li><a href="#i_xor">'<tt>xor</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]165 </ol>
166 </li>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]167 <li><a href="#vectorops">Vector Operations</a>
168 <ol>
169 <li><a href="#i_extractelement">'<tt>extractelement</tt>' Instruction</a></li>
170 <li><a href="#i_insertelement">'<tt>insertelement</tt>' Instruction</a></li>
171 <li><a href="#i_shufflevector">'<tt>shufflevector</tt>' Instruction</a></li>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]172 </ol>
173 </li>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]174 <li><a href="#aggregateops">Aggregate Operations</a>
175 <ol>
176 <li><a href="#i_extractvalue">'<tt>extractvalue</tt>' Instruction</a></li>
177 <li><a href="#i_insertvalue">'<tt>insertvalue</tt>' Instruction</a></li>
178 </ol>
179 </li>
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]180 <li><a href="#memoryops">Memory Access and Addressing Operations</a>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]181 <ol>
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]182 <li><a href="#i_alloca">'<tt>alloca</tt>' Instruction</a></li>
183 <li><a href="#i_load">'<tt>load</tt>' Instruction</a></li>
184 <li><a href="#i_store">'<tt>store</tt>' Instruction</a></li>
185 <li><a href="#i_fence">'<tt>fence</tt>' Instruction</a></li>
186 <li><a href="#i_cmpxchg">'<tt>cmpxchg</tt>' Instruction</a></li>
187 <li><a href="#i_atomicrmw">'<tt>atomicrmw</tt>' Instruction</a></li>
Robert Bocchino7b81c752006-02-17 21:18:08 +0000[diff] [blame]188 <li><a href="#i_getelementptr">'<tt>getelementptr</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]189 </ol>
190 </li>
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]191 <li><a href="#convertops">Conversion Operations</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]192 <ol>
193 <li><a href="#i_trunc">'<tt>trunc .. to</tt>' Instruction</a></li>
194 <li><a href="#i_zext">'<tt>zext .. to</tt>' Instruction</a></li>
195 <li><a href="#i_sext">'<tt>sext .. to</tt>' Instruction</a></li>
196 <li><a href="#i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a></li>
197 <li><a href="#i_fpext">'<tt>fpext .. to</tt>' Instruction</a></li>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]198 <li><a href="#i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a></li>
199 <li><a href="#i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a></li>
200 <li><a href="#i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a></li>
201 <li><a href="#i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a></li>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]202 <li><a href="#i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a></li>
203 <li><a href="#i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a></li>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]204 <li><a href="#i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a></li>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]205 </ol>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]206 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]207 <li><a href="#otherops">Other Operations</a>
208 <ol>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]209 <li><a href="#i_icmp">'<tt>icmp</tt>' Instruction</a></li>
210 <li><a href="#i_fcmp">'<tt>fcmp</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]211 <li><a href="#i_phi">'<tt>phi</tt>' Instruction</a></li>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]212 <li><a href="#i_select">'<tt>select</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]213 <li><a href="#i_call">'<tt>call</tt>' Instruction</a></li>
Chris Lattnerfb6977d2006-01-13 23:26:01 +0000[diff] [blame]214 <li><a href="#i_va_arg">'<tt>va_arg</tt>' Instruction</a></li>
Bill Wendlingf78faf82011-08-02 21:52:38 +0000[diff] [blame]215 <li><a href="#i_landingpad">'<tt>landingpad</tt>' Instruction</a></li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]216 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]217 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]218 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]219 </li>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]220 <li><a href="#intrinsics">Intrinsic Functions</a>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]221 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]222 <li><a href="#int_varargs">Variable Argument Handling Intrinsics</a>
223 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]224 <li><a href="#int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a></li>
225 <li><a href="#int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a></li>
226 <li><a href="#int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]227 </ol>
228 </li>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]229 <li><a href="#int_gc">Accurate Garbage Collection Intrinsics</a>
230 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]231 <li><a href="#int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a></li>
232 <li><a href="#int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a></li>
233 <li><a href="#int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a></li>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]234 </ol>
235 </li>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]236 <li><a href="#int_codegen">Code Generator Intrinsics</a>
237 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]238 <li><a href="#int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a></li>
239 <li><a href="#int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a></li>
240 <li><a href="#int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a></li>
241 <li><a href="#int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a></li>
242 <li><a href="#int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a></li>
243 <li><a href="#int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a></li>
Dan Gohman31f1af12010-05-26 21:56:15 +0000[diff] [blame]244 <li><a href="#int_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a></li>
John Criswell7123e272004-04-09 16:43:20 +0000[diff] [blame]245 </ol>
246 </li>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]247 <li><a href="#int_libc">Standard C Library Intrinsics</a>
248 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]249 <li><a href="#int_memcpy">'<tt>llvm.memcpy.*</tt>' Intrinsic</a></li>
250 <li><a href="#int_memmove">'<tt>llvm.memmove.*</tt>' Intrinsic</a></li>
251 <li><a href="#int_memset">'<tt>llvm.memset.*</tt>' Intrinsic</a></li>
252 <li><a href="#int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a></li>
253 <li><a href="#int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a></li>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]254 <li><a href="#int_sin">'<tt>llvm.sin.*</tt>' Intrinsic</a></li>
255 <li><a href="#int_cos">'<tt>llvm.cos.*</tt>' Intrinsic</a></li>
256 <li><a href="#int_pow">'<tt>llvm.pow.*</tt>' Intrinsic</a></li>
Dan Gohman08b280b2011-05-27 00:36:31 +0000[diff] [blame]257 <li><a href="#int_exp">'<tt>llvm.exp.*</tt>' Intrinsic</a></li>
258 <li><a href="#int_log">'<tt>llvm.log.*</tt>' Intrinsic</a></li>
Cameron Zwarich33390842011-07-08 21:39:21 +0000[diff] [blame]259 <li><a href="#int_fma">'<tt>llvm.fma.*</tt>' Intrinsic</a></li>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]260 </ol>
261 </li>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]262 <li><a href="#int_manip">Bit Manipulation Intrinsics</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]263 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]264 <li><a href="#int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a></li>
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]265 <li><a href="#int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic </a></li>
266 <li><a href="#int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic </a></li>
267 <li><a href="#int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic </a></li>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]268 </ol>
269 </li>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]270 <li><a href="#int_overflow">Arithmetic with Overflow Intrinsics</a>
271 <ol>
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]272 <li><a href="#int_sadd_overflow">'<tt>llvm.sadd.with.overflow.*</tt> Intrinsics</a></li>
273 <li><a href="#int_uadd_overflow">'<tt>llvm.uadd.with.overflow.*</tt> Intrinsics</a></li>
274 <li><a href="#int_ssub_overflow">'<tt>llvm.ssub.with.overflow.*</tt> Intrinsics</a></li>
275 <li><a href="#int_usub_overflow">'<tt>llvm.usub.with.overflow.*</tt> Intrinsics</a></li>
276 <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]277 <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]278 </ol>
279 </li>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]280 <li><a href="#int_fp16">Half Precision Floating Point Intrinsics</a>
281 <ol>
Chris Lattner82c3dc62010-03-14 23:03:31 +0000[diff] [blame]282 <li><a href="#int_convert_to_fp16">'<tt>llvm.convert.to.fp16</tt>' Intrinsic</a></li>
283 <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]284 </ol>
285 </li>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]286 <li><a href="#int_debugger">Debugger intrinsics</a></li>
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]287 <li><a href="#int_eh">Exception Handling intrinsics</a></li>
Duncan Sands4a544a72011-09-06 13:37:06 +0000[diff] [blame]288 <li><a href="#int_trampoline">Trampoline Intrinsics</a>
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]289 <ol>
290 <li><a href="#int_it">'<tt>llvm.init.trampoline</tt>' Intrinsic</a></li>
Duncan Sands4a544a72011-09-06 13:37:06 +0000[diff] [blame]291 <li><a href="#int_at">'<tt>llvm.adjust.trampoline</tt>' Intrinsic</a></li>
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]292 </ol>
293 </li>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]294 <li><a href="#int_memorymarkers">Memory Use Markers</a>
295 <ol>
Jakub Staszak8e1b12a2011-12-04 20:44:25 +0000[diff] [blame]296 <li><a href="#int_lifetime_start">'<tt>llvm.lifetime.start</tt>' Intrinsic</a></li>
297 <li><a href="#int_lifetime_end">'<tt>llvm.lifetime.end</tt>' Intrinsic</a></li>
298 <li><a href="#int_invariant_start">'<tt>llvm.invariant.start</tt>' Intrinsic</a></li>
299 <li><a href="#int_invariant_end">'<tt>llvm.invariant.end</tt>' Intrinsic</a></li>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]300 </ol>
301 </li>
Reid Spencer20677642007-07-20 19:59:11 +0000[diff] [blame]302 <li><a href="#int_general">General intrinsics</a>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]303 <ol>
Reid Spencer20677642007-07-20 19:59:11 +0000[diff] [blame]304 <li><a href="#int_var_annotation">
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]305 '<tt>llvm.var.annotation</tt>' Intrinsic</a></li>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]306 <li><a href="#int_annotation">
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]307 '<tt>llvm.annotation.*</tt>' Intrinsic</a></li>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]308 <li><a href="#int_trap">
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]309 '<tt>llvm.trap</tt>' Intrinsic</a></li>
Dan Gohmana6063c62012-05-14 18:58:10 +0000[diff] [blame]310 <li><a href="#int_debugtrap">
311 '<tt>llvm.debugtrap</tt>' Intrinsic</a></li>
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]312 <li><a href="#int_stackprotector">
313 '<tt>llvm.stackprotector</tt>' Intrinsic</a></li>
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]314 <li><a href="#int_objectsize">
315 '<tt>llvm.objectsize</tt>' Intrinsic</a></li>
Jakub Staszakb170e2d2011-12-04 18:29:26 +0000[diff] [blame]316 <li><a href="#int_expect">
317 '<tt>llvm.expect</tt>' Intrinsic</a></li>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]318 </ol>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]319 </li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]320 </ol>
321 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]322</ol>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]323
324<div class="doc_author">
325 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>
326 and <a href="mailto:vadve@cs.uiuc.edu">Vikram Adve</a></p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]327</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]328
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]329<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]330<h2><a name="abstract">Abstract</a></h2>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]331<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]332
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]333<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]334
335<p>This document is a reference manual for the LLVM assembly language. LLVM is
336 a Static Single Assignment (SSA) based representation that provides type
337 safety, low-level operations, flexibility, and the capability of representing
338 'all' high-level languages cleanly. It is the common code representation
339 used throughout all phases of the LLVM compilation strategy.</p>
340
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]341</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]342
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]343<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]344<h2><a name="introduction">Introduction</a></h2>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]345<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]346
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]347<div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]348
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]349<p>The LLVM code representation is designed to be used in three different forms:
350 as an in-memory compiler IR, as an on-disk bitcode representation (suitable
351 for fast loading by a Just-In-Time compiler), and as a human readable
352 assembly language representation. This allows LLVM to provide a powerful
353 intermediate representation for efficient compiler transformations and
354 analysis, while providing a natural means to debug and visualize the
355 transformations. The three different forms of LLVM are all equivalent. This
356 document describes the human readable representation and notation.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]357
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]358<p>The LLVM representation aims to be light-weight and low-level while being
359 expressive, typed, and extensible at the same time. It aims to be a
360 "universal IR" of sorts, by being at a low enough level that high-level ideas
361 may be cleanly mapped to it (similar to how microprocessors are "universal
362 IR's", allowing many source languages to be mapped to them). By providing
363 type information, LLVM can be used as the target of optimizations: for
364 example, through pointer analysis, it can be proven that a C automatic
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]365 variable is never accessed outside of the current function, allowing it to
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]366 be promoted to a simple SSA value instead of a memory location.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]367
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]368<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]369<h4>
370 <a name="wellformed">Well-Formedness</a>
371</h4>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]372
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]373<div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]374
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]375<p>It is important to note that this document describes 'well formed' LLVM
376 assembly language. There is a difference between what the parser accepts and
377 what is considered 'well formed'. For example, the following instruction is
378 syntactically okay, but not well formed:</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]379
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]380<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]381%x = <a href="#i_add">add</a> i32 1, %x
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]382</pre>
383
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]384<p>because the definition of <tt>%x</tt> does not dominate all of its uses. The
385 LLVM infrastructure provides a verification pass that may be used to verify
386 that an LLVM module is well formed. This pass is automatically run by the
387 parser after parsing input assembly and by the optimizer before it outputs
388 bitcode. The violations pointed out by the verifier pass indicate bugs in
389 transformation passes or input to the parser.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]390
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]391</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]392
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]393</div>
394
Chris Lattnercc689392007-10-03 17:34:29 +0000[diff] [blame]395<!-- Describe the typesetting conventions here. -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]396
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]397<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]398<h2><a name="identifiers">Identifiers</a></h2>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]399<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]400
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]401<div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]402
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]403<p>LLVM identifiers come in two basic types: global and local. Global
404 identifiers (functions, global variables) begin with the <tt>'@'</tt>
405 character. Local identifiers (register names, types) begin with
406 the <tt>'%'</tt> character. Additionally, there are three different formats
407 for identifiers, for different purposes:</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]408
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]409<ol>
Reid Spencer2c452282007-08-07 14:34:28 +0000[diff] [blame]410 <li>Named values are represented as a string of characters with their prefix.
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]411 For example, <tt>%foo</tt>, <tt>@DivisionByZero</tt>,
412 <tt>%a.really.long.identifier</tt>. The actual regular expression used is
413 '<tt>[%@][a-zA-Z$._][a-zA-Z$._0-9]*</tt>'. Identifiers which require
414 other characters in their names can be surrounded with quotes. Special
415 characters may be escaped using <tt>"\xx"</tt> where <tt>xx</tt> is the
416 ASCII code for the character in hexadecimal. In this way, any character
417 can be used in a name value, even quotes themselves.</li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]418
Reid Spencer2c452282007-08-07 14:34:28 +0000[diff] [blame]419 <li>Unnamed values are represented as an unsigned numeric value with their
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]420 prefix. For example, <tt>%12</tt>, <tt>@2</tt>, <tt>%44</tt>.</li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]421
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]422 <li>Constants, which are described in a <a href="#constants">section about
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]423 constants</a>, below.</li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]424</ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]425
Reid Spencer2c452282007-08-07 14:34:28 +0000[diff] [blame]426<p>LLVM requires that values start with a prefix for two reasons: Compilers
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]427 don't need to worry about name clashes with reserved words, and the set of
428 reserved words may be expanded in the future without penalty. Additionally,
429 unnamed identifiers allow a compiler to quickly come up with a temporary
430 variable without having to avoid symbol table conflicts.</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]431
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]432<p>Reserved words in LLVM are very similar to reserved words in other
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]433 languages. There are keywords for different opcodes
434 ('<tt><a href="#i_add">add</a></tt>',
435 '<tt><a href="#i_bitcast">bitcast</a></tt>',
436 '<tt><a href="#i_ret">ret</a></tt>', etc...), for primitive type names
437 ('<tt><a href="#t_void">void</a></tt>',
438 '<tt><a href="#t_primitive">i32</a></tt>', etc...), and others. These
439 reserved words cannot conflict with variable names, because none of them
440 start with a prefix character (<tt>'%'</tt> or <tt>'@'</tt>).</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]441
442<p>Here is an example of LLVM code to multiply the integer variable
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]443 '<tt>%X</tt>' by 8:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]444
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]445<p>The easy way:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]446
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]447<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]448%result = <a href="#i_mul">mul</a> i32 %X, 8
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]449</pre>
450
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]451<p>After strength reduction:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]452
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]453<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]454%result = <a href="#i_shl">shl</a> i32 %X, i8 3
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]455</pre>
456
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]457<p>And the hard way:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]458
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]459<pre class="doc_code">
Gabor Greifec58f752009-10-28 13:05:07 +0000[diff] [blame]460%0 = <a href="#i_add">add</a> i32 %X, %X <i>; yields {i32}:%0</i>
461%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]462%result = <a href="#i_add">add</a> i32 %1, %1
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]463</pre>
464
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]465<p>This last way of multiplying <tt>%X</tt> by 8 illustrates several important
466 lexical features of LLVM:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]467
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]468<ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]469 <li>Comments are delimited with a '<tt>;</tt>' and go until the end of
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]470 line.</li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]471
472 <li>Unnamed temporaries are created when the result of a computation is not
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]473 assigned to a named value.</li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]474
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]475 <li>Unnamed temporaries are numbered sequentially</li>
476</ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]477
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]478<p>It also shows a convention that we follow in this document. When
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]479 demonstrating instructions, we will follow an instruction with a comment that
480 defines the type and name of value produced. Comments are shown in italic
481 text.</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]482
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]483</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]484
485<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]486<h2><a name="highlevel">High Level Structure</a></h2>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]487<!-- *********************************************************************** -->
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]488<div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]489<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]490<h3>
491 <a name="modulestructure">Module Structure</a>
492</h3>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]493
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]494<div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]495
Bill Wendling4cc2be62012-03-14 08:07:43 +0000[diff] [blame]496<p>LLVM programs are composed of <tt>Module</tt>s, each of which is a
497 translation unit of the input programs. Each module consists of functions,
498 global variables, and symbol table entries. Modules may be combined together
499 with the LLVM linker, which merges function (and global variable)
500 definitions, resolves forward declarations, and merges symbol table
501 entries. Here is an example of the "hello world" module:</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]502
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]503<pre class="doc_code">
Chris Lattner63e4ccb2010-08-17 17:13:42 +0000[diff] [blame]504<i>; Declare the string constant as a global constant.</i>&nbsp;
Bill Wendling4cc2be62012-03-14 08:07:43 +0000[diff] [blame]505<a href="#identifiers">@.str</a> = <a href="#linkage_private">private</a>&nbsp;<a href="#globalvars">unnamed_addr</a>&nbsp;<a href="#globalvars">constant</a>&nbsp;<a href="#t_array">[13 x i8]</a> c"hello world\0A\00"&nbsp;
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]506
Chris Lattner63e4ccb2010-08-17 17:13:42 +0000[diff] [blame]507<i>; External declaration of the puts function</i>&nbsp;
Bill Wendling4cc2be62012-03-14 08:07:43 +0000[diff] [blame]508<a href="#functionstructure">declare</a> i32 @puts(i8* <a href="#nocapture">nocapture</a>) <a href="#fnattrs">nounwind</a>&nbsp;
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]509
510<i>; Definition of main function</i>
Chris Lattner63e4ccb2010-08-17 17:13:42 +0000[diff] [blame]511define i32 @main() { <i>; i32()* </i>&nbsp;
512 <i>; Convert [13 x i8]* to i8 *...</i>&nbsp;
Bill Wendling4cc2be62012-03-14 08:07:43 +0000[diff] [blame]513 %cast210 = <a href="#i_getelementptr">getelementptr</a> [13 x i8]* @.str, i64 0, i64 0
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]514
Chris Lattner63e4ccb2010-08-17 17:13:42 +0000[diff] [blame]515 <i>; Call puts function to write out the string to stdout.</i>&nbsp;
Bill Wendling4cc2be62012-03-14 08:07:43 +0000[diff] [blame]516 <a href="#i_call">call</a> i32 @puts(i8* %cast210)
Chris Lattner63e4ccb2010-08-17 17:13:42 +0000[diff] [blame]517 <a href="#i_ret">ret</a> i32 0&nbsp;
518}
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]519
520<i>; Named metadata</i>
Bill Wendling4cc2be62012-03-14 08:07:43 +0000[diff] [blame]521!1 = metadata !{i32 42}
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]522!foo = !{!1, null}
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]523</pre>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]524
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]525<p>This example is made up of a <a href="#globalvars">global variable</a> named
Bill Wendling4cc2be62012-03-14 08:07:43 +0000[diff] [blame]526 "<tt>.str</tt>", an external declaration of the "<tt>puts</tt>" function,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]527 a <a href="#functionstructure">function definition</a> for
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]528 "<tt>main</tt>" and <a href="#namedmetadatastructure">named metadata</a>
Bill Wendling4cc2be62012-03-14 08:07:43 +0000[diff] [blame]529 "<tt>foo</tt>".</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]530
Bill Wendling4cc2be62012-03-14 08:07:43 +0000[diff] [blame]531<p>In general, a module is made up of a list of global values (where both
532 functions and global variables are global values). Global values are
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]533 represented by a pointer to a memory location (in this case, a pointer to an
534 array of char, and a pointer to a function), and have one of the
535 following <a href="#linkage">linkage types</a>.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]536
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]537</div>
538
539<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]540<h3>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]541 <a name="linkage">Linkage Types</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]542</h3>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]543
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]544<div>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]545
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]546<p>All Global Variables and Functions have one of the following types of
547 linkage:</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]548
549<dl>
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]550 <dt><tt><b><a name="linkage_private">private</a></b></tt></dt>
Bill Wendling5e721d72010-07-01 21:55:59 +0000[diff] [blame]551 <dd>Global values with "<tt>private</tt>" linkage are only directly accessible
552 by objects in the current module. In particular, linking code into a
553 module with an private global value may cause the private to be renamed as
554 necessary to avoid collisions. Because the symbol is private to the
555 module, all references can be updated. This doesn't show up in any symbol
556 table in the object file.</dd>
Rafael Espindolabb46f522009-01-15 20:18:42 +0000[diff] [blame]557
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]558 <dt><tt><b><a name="linkage_linker_private">linker_private</a></b></tt></dt>
Bill Wendling5e721d72010-07-01 21:55:59 +0000[diff] [blame]559 <dd>Similar to <tt>private</tt>, but the symbol is passed through the
560 assembler and evaluated by the linker. Unlike normal strong symbols, they
561 are removed by the linker from the final linked image (executable or
562 dynamic library).</dd>
563
564 <dt><tt><b><a name="linkage_linker_private_weak">linker_private_weak</a></b></tt></dt>
565 <dd>Similar to "<tt>linker_private</tt>", but the symbol is weak. Note that
566 <tt>linker_private_weak</tt> symbols are subject to coalescing by the
567 linker. The symbols are removed by the linker from the final linked image
568 (executable or dynamic library).</dd>
Bill Wendling3d10a5a2009-07-20 01:03:30 +0000[diff] [blame]569
Bill Wendling55ae5152010-08-20 22:05:50 +0000[diff] [blame]570 <dt><tt><b><a name="linkage_linker_private_weak_def_auto">linker_private_weak_def_auto</a></b></tt></dt>
571 <dd>Similar to "<tt>linker_private_weak</tt>", but it's known that the address
572 of the object is not taken. For instance, functions that had an inline
573 definition, but the compiler decided not to inline it. Note,
574 unlike <tt>linker_private</tt> and <tt>linker_private_weak</tt>,
575 <tt>linker_private_weak_def_auto</tt> may have only <tt>default</tt>
576 visibility. The symbols are removed by the linker from the final linked
577 image (executable or dynamic library).</dd>
578
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]579 <dt><tt><b><a name="linkage_internal">internal</a></b></tt></dt>
Bill Wendling07d31772010-06-29 22:34:52 +0000[diff] [blame]580 <dd>Similar to private, but the value shows as a local symbol
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]581 (<tt>STB_LOCAL</tt> in the case of ELF) in the object file. This
582 corresponds to the notion of the '<tt>static</tt>' keyword in C.</dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]583
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]584 <dt><tt><b><a name="linkage_available_externally">available_externally</a></b></tt></dt>
Chris Lattner266c7bb2009-04-13 05:44:34 +0000[diff] [blame]585 <dd>Globals with "<tt>available_externally</tt>" linkage are never emitted
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]586 into the object file corresponding to the LLVM module. They exist to
587 allow inlining and other optimizations to take place given knowledge of
588 the definition of the global, which is known to be somewhere outside the
589 module. Globals with <tt>available_externally</tt> linkage are allowed to
590 be discarded at will, and are otherwise the same as <tt>linkonce_odr</tt>.
591 This linkage type is only allowed on definitions, not declarations.</dd>
Chris Lattner266c7bb2009-04-13 05:44:34 +0000[diff] [blame]592
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]593 <dt><tt><b><a name="linkage_linkonce">linkonce</a></b></tt></dt>
Chris Lattner4887bd82007-01-14 06:51:48 +0000[diff] [blame]594 <dd>Globals with "<tt>linkonce</tt>" linkage are merged with other globals of
Chris Lattner873187c2010-01-09 19:15:14 +0000[diff] [blame]595 the same name when linkage occurs. This can be used to implement
596 some forms of inline functions, templates, or other code which must be
597 generated in each translation unit that uses it, but where the body may
598 be overridden with a more definitive definition later. Unreferenced
599 <tt>linkonce</tt> globals are allowed to be discarded. Note that
600 <tt>linkonce</tt> linkage does not actually allow the optimizer to
601 inline the body of this function into callers because it doesn't know if
602 this definition of the function is the definitive definition within the
603 program or whether it will be overridden by a stronger definition.
604 To enable inlining and other optimizations, use "<tt>linkonce_odr</tt>"
605 linkage.</dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]606
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]607 <dt><tt><b><a name="linkage_weak">weak</a></b></tt></dt>
Chris Lattner26d054d2009-08-05 05:21:07 +0000[diff] [blame]608 <dd>"<tt>weak</tt>" linkage has the same merging semantics as
609 <tt>linkonce</tt> linkage, except that unreferenced globals with
610 <tt>weak</tt> linkage may not be discarded. This is used for globals that
611 are declared "weak" in C source code.</dd>
612
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]613 <dt><tt><b><a name="linkage_common">common</a></b></tt></dt>
Chris Lattner26d054d2009-08-05 05:21:07 +0000[diff] [blame]614 <dd>"<tt>common</tt>" linkage is most similar to "<tt>weak</tt>" linkage, but
615 they are used for tentative definitions in C, such as "<tt>int X;</tt>" at
616 global scope.
617 Symbols with "<tt>common</tt>" linkage are merged in the same way as
618 <tt>weak symbols</tt>, and they may not be deleted if unreferenced.
Chris Lattnercd81f5d2009-08-05 05:41:44 +0000[diff] [blame]619 <tt>common</tt> symbols may not have an explicit section,
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]620 must have a zero initializer, and may not be marked '<a
Chris Lattnercd81f5d2009-08-05 05:41:44 +0000[diff] [blame]621 href="#globalvars"><tt>constant</tt></a>'. Functions and aliases may not
622 have common linkage.</dd>
Chris Lattner26d054d2009-08-05 05:21:07 +0000[diff] [blame]623
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]624
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]625 <dt><tt><b><a name="linkage_appending">appending</a></b></tt></dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]626 <dd>"<tt>appending</tt>" linkage may only be applied to global variables of
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]627 pointer to array type. When two global variables with appending linkage
628 are linked together, the two global arrays are appended together. This is
629 the LLVM, typesafe, equivalent of having the system linker append together
630 "sections" with identical names when .o files are linked.</dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]631
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]632 <dt><tt><b><a name="linkage_externweak">extern_weak</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]633 <dd>The semantics of this linkage follow the ELF object file model: the symbol
634 is weak until linked, if not linked, the symbol becomes null instead of
635 being an undefined reference.</dd>
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame]636
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]637 <dt><tt><b><a name="linkage_linkonce_odr">linkonce_odr</a></b></tt></dt>
638 <dt><tt><b><a name="linkage_weak_odr">weak_odr</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]639 <dd>Some languages allow differing globals to be merged, such as two functions
640 with different semantics. Other languages, such as <tt>C++</tt>, ensure
Bill Wendling5e721d72010-07-01 21:55:59 +0000[diff] [blame]641 that only equivalent globals are ever merged (the "one definition rule"
642 &mdash; "ODR"). Such languages can use the <tt>linkonce_odr</tt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]643 and <tt>weak_odr</tt> linkage types to indicate that the global will only
644 be merged with equivalent globals. These linkage types are otherwise the
645 same as their non-<tt>odr</tt> versions.</dd>
Duncan Sands667d4b82009-03-07 15:45:40 +0000[diff] [blame]646
Bill Wendling5c3a9f72011-11-04 20:40:41 +0000[diff] [blame]647 <dt><tt><b><a name="linkage_external">external</a></b></tt></dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]648 <dd>If none of the above identifiers are used, the global is externally
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]649 visible, meaning that it participates in linkage and can be used to
650 resolve external symbol references.</dd>
Reid Spencerc8910842007-04-11 23:49:50 +0000[diff] [blame]651</dl>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]652
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]653<p>The next two types of linkage are targeted for Microsoft Windows platform
654 only. They are designed to support importing (exporting) symbols from (to)
655 DLLs (Dynamic Link Libraries).</p>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]656
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]657<dl>
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]658 <dt><tt><b><a name="linkage_dllimport">dllimport</a></b></tt></dt>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]659 <dd>"<tt>dllimport</tt>" linkage causes the compiler to reference a function
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]660 or variable via a global pointer to a pointer that is set up by the DLL
661 exporting the symbol. On Microsoft Windows targets, the pointer name is
662 formed by combining <code>__imp_</code> and the function or variable
663 name.</dd>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]664
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]665 <dt><tt><b><a name="linkage_dllexport">dllexport</a></b></tt></dt>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]666 <dd>"<tt>dllexport</tt>" linkage causes the compiler to provide a global
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]667 pointer to a pointer in a DLL, so that it can be referenced with the
668 <tt>dllimport</tt> attribute. On Microsoft Windows targets, the pointer
669 name is formed by combining <code>__imp_</code> and the function or
670 variable name.</dd>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]671</dl>
672
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]673<p>For example, since the "<tt>.LC0</tt>" variable is defined to be internal, if
674 another module defined a "<tt>.LC0</tt>" variable and was linked with this
675 one, one of the two would be renamed, preventing a collision. Since
676 "<tt>main</tt>" and "<tt>puts</tt>" are external (i.e., lacking any linkage
677 declarations), they are accessible outside of the current module.</p>
678
679<p>It is illegal for a function <i>declaration</i> to have any linkage type
Bill Wendlingf7f06102011-10-11 06:41:28 +0000[diff] [blame]680 other than <tt>external</tt>, <tt>dllimport</tt>
681 or <tt>extern_weak</tt>.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]682
Duncan Sands667d4b82009-03-07 15:45:40 +0000[diff] [blame]683<p>Aliases can have only <tt>external</tt>, <tt>internal</tt>, <tt>weak</tt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]684 or <tt>weak_odr</tt> linkages.</p>
685
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]686</div>
687
688<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]689<h3>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]690 <a name="callingconv">Calling Conventions</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]691</h3>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]692
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]693<div>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]694
695<p>LLVM <a href="#functionstructure">functions</a>, <a href="#i_call">calls</a>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]696 and <a href="#i_invoke">invokes</a> can all have an optional calling
697 convention specified for the call. The calling convention of any pair of
698 dynamic caller/callee must match, or the behavior of the program is
699 undefined. The following calling conventions are supported by LLVM, and more
700 may be added in the future:</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]701
702<dl>
703 <dt><b>"<tt>ccc</tt>" - The C calling convention</b>:</dt>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]704 <dd>This calling convention (the default if no other calling convention is
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]705 specified) matches the target C calling conventions. This calling
706 convention supports varargs function calls and tolerates some mismatch in
707 the declared prototype and implemented declaration of the function (as
708 does normal C).</dd>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]709
710 <dt><b>"<tt>fastcc</tt>" - The fast calling convention</b>:</dt>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]711 <dd>This calling convention attempts to make calls as fast as possible
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]712 (e.g. by passing things in registers). This calling convention allows the
713 target to use whatever tricks it wants to produce fast code for the
714 target, without having to conform to an externally specified ABI
Jeffrey Yasskin95fa80a2010-01-09 19:44:16 +0000[diff] [blame]715 (Application Binary Interface).
716 <a href="CodeGenerator.html#tailcallopt">Tail calls can only be optimized
Chris Lattner29689432010-03-11 00:22:57 +0000[diff] [blame]717 when this or the GHC convention is used.</a> This calling convention
718 does not support varargs and requires the prototype of all callees to
719 exactly match the prototype of the function definition.</dd>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]720
721 <dt><b>"<tt>coldcc</tt>" - The cold calling convention</b>:</dt>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]722 <dd>This calling convention attempts to make code in the caller as efficient
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]723 as possible under the assumption that the call is not commonly executed.
724 As such, these calls often preserve all registers so that the call does
725 not break any live ranges in the caller side. This calling convention
726 does not support varargs and requires the prototype of all callees to
727 exactly match the prototype of the function definition.</dd>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]728
Chris Lattner29689432010-03-11 00:22:57 +0000[diff] [blame]729 <dt><b>"<tt>cc <em>10</em></tt>" - GHC convention</b>:</dt>
730 <dd>This calling convention has been implemented specifically for use by the
731 <a href="http://www.haskell.org/ghc">Glasgow Haskell Compiler (GHC)</a>.
732 It passes everything in registers, going to extremes to achieve this by
733 disabling callee save registers. This calling convention should not be
734 used lightly but only for specific situations such as an alternative to
735 the <em>register pinning</em> performance technique often used when
736 implementing functional programming languages.At the moment only X86
737 supports this convention and it has the following limitations:
738 <ul>
739 <li>On <em>X86-32</em> only supports up to 4 bit type parameters. No
740 floating point types are supported.</li>
741 <li>On <em>X86-64</em> only supports up to 10 bit type parameters and
742 6 floating point parameters.</li>
743 </ul>
744 This calling convention supports
745 <a href="CodeGenerator.html#tailcallopt">tail call optimization</a> but
746 requires both the caller and callee are using it.
747 </dd>
748
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]749 <dt><b>"<tt>cc &lt;<em>n</em>&gt;</tt>" - Numbered convention</b>:</dt>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]750 <dd>Any calling convention may be specified by number, allowing
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]751 target-specific calling conventions to be used. Target specific calling
752 conventions start at 64.</dd>
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]753</dl>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]754
755<p>More calling conventions can be added/defined on an as-needed basis, to
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]756 support Pascal conventions or any other well-known target-independent
757 convention.</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]758
759</div>
760
761<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]762<h3>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]763 <a name="visibility">Visibility Styles</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]764</h3>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]765
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]766<div>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]767
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]768<p>All Global Variables and Functions have one of the following visibility
769 styles:</p>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]770
771<dl>
772 <dt><b>"<tt>default</tt>" - Default style</b>:</dt>
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]773 <dd>On targets that use the ELF object file format, default visibility means
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]774 that the declaration is visible to other modules and, in shared libraries,
775 means that the declared entity may be overridden. On Darwin, default
776 visibility means that the declaration is visible to other modules. Default
777 visibility corresponds to "external linkage" in the language.</dd>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]778
779 <dt><b>"<tt>hidden</tt>" - Hidden style</b>:</dt>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]780 <dd>Two declarations of an object with hidden visibility refer to the same
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]781 object if they are in the same shared object. Usually, hidden visibility
782 indicates that the symbol will not be placed into the dynamic symbol
783 table, so no other module (executable or shared library) can reference it
784 directly.</dd>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]785
Anton Korobeynikov6f9896f2007-04-29 18:35:00 +0000[diff] [blame]786 <dt><b>"<tt>protected</tt>" - Protected style</b>:</dt>
Anton Korobeynikov6f9896f2007-04-29 18:35:00 +0000[diff] [blame]787 <dd>On ELF, protected visibility indicates that the symbol will be placed in
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]788 the dynamic symbol table, but that references within the defining module
789 will bind to the local symbol. That is, the symbol cannot be overridden by
790 another module.</dd>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]791</dl>
792
793</div>
794
795<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]796<h3>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]797 <a name="namedtypes">Named Types</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]798</h3>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]799
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]800<div>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]801
802<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]803 it easier to read the IR and make the IR more condensed (particularly when
804 recursive types are involved). An example of a name specification is:</p>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]805
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]806<pre class="doc_code">
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]807%mytype = type { %mytype*, i32 }
808</pre>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]809
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]810<p>You may give a name to any <a href="#typesystem">type</a> except
Chris Lattnerdc65f222010-08-17 23:26:04 +0000[diff] [blame]811 "<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]812 is expected with the syntax "%mytype".</p>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]813
814<p>Note that type names are aliases for the structural type that they indicate,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]815 and that you can therefore specify multiple names for the same type. This
816 often leads to confusing behavior when dumping out a .ll file. Since LLVM IR
817 uses structural typing, the name is not part of the type. When printing out
818 LLVM IR, the printer will pick <em>one name</em> to render all types of a
819 particular shape. This means that if you have code where two different
820 source types end up having the same LLVM type, that the dumper will sometimes
821 print the "wrong" or unexpected type. This is an important design point and
822 isn't going to change.</p>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]823
824</div>
825
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]826<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]827<h3>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]828 <a name="globalvars">Global Variables</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]829</h3>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]830
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]831<div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]832
Chris Lattner3689a342005-02-12 19:30:21 +0000[diff] [blame]833<p>Global variables define regions of memory allocated at compilation time
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]834 instead of run-time. Global variables may optionally be initialized, may
835 have an explicit section to be placed in, and may have an optional explicit
836 alignment specified. A variable may be defined as "thread_local", which
837 means that it will not be shared by threads (each thread will have a
838 separated copy of the variable). A variable may be defined as a global
839 "constant," which indicates that the contents of the variable
840 will <b>never</b> be modified (enabling better optimization, allowing the
841 global data to be placed in the read-only section of an executable, etc).
842 Note that variables that need runtime initialization cannot be marked
843 "constant" as there is a store to the variable.</p>
Chris Lattner3689a342005-02-12 19:30:21 +0000[diff] [blame]844
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]845<p>LLVM explicitly allows <em>declarations</em> of global variables to be marked
846 constant, even if the final definition of the global is not. This capability
847 can be used to enable slightly better optimization of the program, but
848 requires the language definition to guarantee that optimizations based on the
849 'constantness' are valid for the translation units that do not include the
850 definition.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]851
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]852<p>As SSA values, global variables define pointer values that are in scope
853 (i.e. they dominate) all basic blocks in the program. Global variables
854 always define a pointer to their "content" type because they describe a
855 region of memory, and all memory objects in LLVM are accessed through
856 pointers.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]857
Rafael Espindolabea46262011-01-08 16:42:36 +0000[diff] [blame]858<p>Global variables can be marked with <tt>unnamed_addr</tt> which indicates
859 that the address is not significant, only the content. Constants marked
Rafael Espindolaa5eaa862011-01-15 08:20:57 +0000[diff] [blame]860 like this can be merged with other constants if they have the same
861 initializer. Note that a constant with significant address <em>can</em>
862 be merged with a <tt>unnamed_addr</tt> constant, the result being a
863 constant whose address is significant.</p>
Rafael Espindolabea46262011-01-08 16:42:36 +0000[diff] [blame]864
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]865<p>A global variable may be declared to reside in a target-specific numbered
866 address space. For targets that support them, address spaces may affect how
867 optimizations are performed and/or what target instructions are used to
868 access the variable. The default address space is zero. The address space
869 qualifier must precede any other attributes.</p>
Christopher Lamb284d9922007-12-11 09:31:00 +0000[diff] [blame]870
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]871<p>LLVM allows an explicit section to be specified for globals. If the target
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]872 supports it, it will emit globals to the section specified.</p>
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]873
Chris Lattnerce99fa92010-04-28 00:13:42 +0000[diff] [blame]874<p>An explicit alignment may be specified for a global, which must be a power
875 of 2. If not present, or if the alignment is set to zero, the alignment of
876 the global is set by the target to whatever it feels convenient. If an
877 explicit alignment is specified, the global is forced to have exactly that
Chris Lattner2d4b8ee2010-04-28 00:31:12 +0000[diff] [blame]878 alignment. Targets and optimizers are not allowed to over-align the global
879 if the global has an assigned section. In this case, the extra alignment
880 could be observable: for example, code could assume that the globals are
881 densely packed in their section and try to iterate over them as an array,
882 alignment padding would break this iteration.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]883
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]884<p>For example, the following defines a global in a numbered address space with
885 an initializer, section, and alignment:</p>
Chris Lattner68027ea2007-01-14 00:27:09 +0000[diff] [blame]886
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]887<pre class="doc_code">
Dan Gohman398873c2009-01-11 00:40:00 +0000[diff] [blame]888@G = addrspace(5) constant float 1.0, section "foo", align 4
Chris Lattner68027ea2007-01-14 00:27:09 +0000[diff] [blame]889</pre>
890
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]891</div>
892
893
894<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]895<h3>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]896 <a name="functionstructure">Functions</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]897</h3>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]898
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]899<div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]900
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]901<p>LLVM function definitions consist of the "<tt>define</tt>" keyword, an
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]902 optional <a href="#linkage">linkage type</a>, an optional
903 <a href="#visibility">visibility style</a>, an optional
Rafael Espindolabea46262011-01-08 16:42:36 +0000[diff] [blame]904 <a href="#callingconv">calling convention</a>,
905 an optional <tt>unnamed_addr</tt> attribute, a return type, an optional
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]906 <a href="#paramattrs">parameter attribute</a> for the return type, a function
907 name, a (possibly empty) argument list (each with optional
908 <a href="#paramattrs">parameter attributes</a>), optional
909 <a href="#fnattrs">function attributes</a>, an optional section, an optional
910 alignment, an optional <a href="#gc">garbage collector name</a>, an opening
911 curly brace, a list of basic blocks, and a closing curly brace.</p>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]912
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]913<p>LLVM function declarations consist of the "<tt>declare</tt>" keyword, an
914 optional <a href="#linkage">linkage type</a>, an optional
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]915 <a href="#visibility">visibility style</a>, an optional
Rafael Espindolabea46262011-01-08 16:42:36 +0000[diff] [blame]916 <a href="#callingconv">calling convention</a>,
917 an optional <tt>unnamed_addr</tt> attribute, a return type, an optional
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]918 <a href="#paramattrs">parameter attribute</a> for the return type, a function
919 name, a possibly empty list of arguments, an optional alignment, and an
920 optional <a href="#gc">garbage collector name</a>.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]921
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]922<p>A function definition contains a list of basic blocks, forming the CFG
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]923 (Control Flow Graph) for the function. Each basic block may optionally start
924 with a label (giving the basic block a symbol table entry), contains a list
925 of instructions, and ends with a <a href="#terminators">terminator</a>
926 instruction (such as a branch or function return).</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]927
Chris Lattner4a3c9012007-06-08 16:52:14 +0000[diff] [blame]928<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]929 executed on entrance to the function, and it is not allowed to have
930 predecessor basic blocks (i.e. there can not be any branches to the entry
931 block of a function). Because the block can have no predecessors, it also
932 cannot have any <a href="#i_phi">PHI nodes</a>.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]933
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]934<p>LLVM allows an explicit section to be specified for functions. If the target
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]935 supports it, it will emit functions to the section specified.</p>
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]936
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]937<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]938 the alignment is set to zero, the alignment of the function is set by the
939 target to whatever it feels convenient. If an explicit alignment is
940 specified, the function is forced to have at least that much alignment. All
941 alignments must be a power of 2.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]942
Rafael Espindolabea46262011-01-08 16:42:36 +0000[diff] [blame]943<p>If the <tt>unnamed_addr</tt> attribute is given, the address is know to not
Bill Wendling5c3a9f72011-11-04 20:40:41 +0000[diff] [blame]944 be significant and two identical functions can be merged.</p>
Rafael Espindolabea46262011-01-08 16:42:36 +0000[diff] [blame]945
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]946<h5>Syntax:</h5>
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]947<pre class="doc_code">
Chris Lattner50ad45c2008-10-13 16:55:18 +0000[diff] [blame]948define [<a href="#linkage">linkage</a>] [<a href="#visibility">visibility</a>]
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]949 [<a href="#callingconv">cconv</a>] [<a href="#paramattrs">ret attrs</a>]
950 &lt;ResultType&gt; @&lt;FunctionName&gt; ([argument list])
951 [<a href="#fnattrs">fn Attrs</a>] [section "name"] [align N]
952 [<a href="#gc">gc</a>] { ... }
953</pre>
Devang Patel307e8ab2008-10-07 17:48:33 +0000[diff] [blame]954
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]955</div>
956
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]957<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]958<h3>
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]959 <a name="aliasstructure">Aliases</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]960</h3>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]961
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]962<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]963
964<p>Aliases act as "second name" for the aliasee value (which can be either
965 function, global variable, another alias or bitcast of global value). Aliases
966 may have an optional <a href="#linkage">linkage type</a>, and an
967 optional <a href="#visibility">visibility style</a>.</p>
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]968
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]969<h5>Syntax:</h5>
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]970<pre class="doc_code">
Duncan Sands0b23ac12008-09-12 20:48:21 +0000[diff] [blame]971@&lt;Name&gt; = alias [Linkage] [Visibility] &lt;AliaseeTy&gt; @&lt;Aliasee&gt;
Bill Wendlingaac388b2007-05-29 09:42:13 +0000[diff] [blame]972</pre>
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]973
974</div>
975
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]976<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]977<h3>
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]978 <a name="namedmetadatastructure">Named Metadata</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]979</h3>
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]980
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]981<div>
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]982
Chris Lattnere6a5ddd2010-01-15 21:50:19 +0000[diff] [blame]983<p>Named metadata is a collection of metadata. <a href="#metadata">Metadata
Dan Gohman872814a2010-07-21 18:54:18 +0000[diff] [blame]984 nodes</a> (but not metadata strings) are the only valid operands for
Chris Lattnere6a5ddd2010-01-15 21:50:19 +0000[diff] [blame]985 a named metadata.</p>
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]986
987<h5>Syntax:</h5>
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]988<pre class="doc_code">
Dan Gohman872814a2010-07-21 18:54:18 +0000[diff] [blame]989; Some unnamed metadata nodes, which are referenced by the named metadata.
990!0 = metadata !{metadata !"zero"}
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]991!1 = metadata !{metadata !"one"}
Dan Gohman872814a2010-07-21 18:54:18 +0000[diff] [blame]992!2 = metadata !{metadata !"two"}
Dan Gohman1005bc52010-07-13 19:48:13 +0000[diff] [blame]993; A named metadata.
Dan Gohman872814a2010-07-21 18:54:18 +0000[diff] [blame]994!name = !{!0, !1, !2}
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]995</pre>
Devang Patelcd1fd252010-01-11 19:35:55 +0000[diff] [blame]996
997</div>
998
999<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1000<h3>
1001 <a name="paramattrs">Parameter Attributes</a>
1002</h3>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1003
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1004<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1005
1006<p>The return type and each parameter of a function type may have a set of
1007 <i>parameter attributes</i> associated with them. Parameter attributes are
1008 used to communicate additional information about the result or parameters of
1009 a function. Parameter attributes are considered to be part of the function,
1010 not of the function type, so functions with different parameter attributes
1011 can have the same function type.</p>
1012
1013<p>Parameter attributes are simple keywords that follow the type specified. If
1014 multiple parameter attributes are needed, they are space separated. For
1015 example:</p>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1016
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]1017<pre class="doc_code">
Nick Lewyckyb6a7d252009-02-15 23:06:14 +0000[diff] [blame]1018declare i32 @printf(i8* noalias nocapture, ...)
Chris Lattner66d922c2008-10-04 18:33:34 +0000[diff] [blame]1019declare i32 @atoi(i8 zeroext)
1020declare signext i8 @returns_signed_char()
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1021</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1022
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1023<p>Note that any attributes for the function result (<tt>nounwind</tt>,
1024 <tt>readonly</tt>) come immediately after the argument list.</p>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1025
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1026<p>Currently, only the following parameter attributes are defined:</p>
Chris Lattner47507de2008-01-11 06:20:47 +0000[diff] [blame]1027
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1028<dl>
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1029 <dt><tt><b>zeroext</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1030 <dd>This indicates to the code generator that the parameter or return value
Cameron Zwarichebe81732011-03-16 22:20:18 +0000[diff] [blame]1031 should be zero-extended to the extent required by the target's ABI (which
1032 is usually 32-bits, but is 8-bits for a i1 on x86-64) by the caller (for a
1033 parameter) or the callee (for a return value).</dd>
Chris Lattner47507de2008-01-11 06:20:47 +0000[diff] [blame]1034
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1035 <dt><tt><b>signext</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1036 <dd>This indicates to the code generator that the parameter or return value
Cameron Zwarich9e69ff92011-03-17 14:21:58 +0000[diff] [blame]1037 should be sign-extended to the extent required by the target's ABI (which
1038 is usually 32-bits) by the caller (for a parameter) or the callee (for a
1039 return value).</dd>
Chris Lattner47507de2008-01-11 06:20:47 +0000[diff] [blame]1040
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1041 <dt><tt><b>inreg</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1042 <dd>This indicates that this parameter or return value should be treated in a
1043 special target-dependent fashion during while emitting code for a function
1044 call or return (usually, by putting it in a register as opposed to memory,
1045 though some targets use it to distinguish between two different kinds of
1046 registers). Use of this attribute is target-specific.</dd>
Chris Lattner47507de2008-01-11 06:20:47 +0000[diff] [blame]1047
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1048 <dt><tt><b><a name="byval">byval</a></b></tt></dt>
Chris Lattnera6fd81d2010-11-20 23:49:06 +0000[diff] [blame]1049 <dd><p>This indicates that the pointer parameter should really be passed by
1050 value to the function. The attribute implies that a hidden copy of the
1051 pointee
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1052 is made between the caller and the callee, so the callee is unable to
1053 modify the value in the callee. This attribute is only valid on LLVM
1054 pointer arguments. It is generally used to pass structs and arrays by
1055 value, but is also valid on pointers to scalars. The copy is considered
1056 to belong to the caller not the callee (for example,
1057 <tt><a href="#readonly">readonly</a></tt> functions should not write to
1058 <tt>byval</tt> parameters). This is not a valid attribute for return
Chris Lattnera6fd81d2010-11-20 23:49:06 +0000[diff] [blame]1059 values.</p>
1060
1061 <p>The byval attribute also supports specifying an alignment with
1062 the align attribute. It indicates the alignment of the stack slot to
1063 form and the known alignment of the pointer specified to the call site. If
1064 the alignment is not specified, then the code generator makes a
1065 target-specific assumption.</p></dd>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1066
Dan Gohmanff235352010-07-02 23:18:08 +0000[diff] [blame]1067 <dt><tt><b><a name="sret">sret</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1068 <dd>This indicates that the pointer parameter specifies the address of a
1069 structure that is the return value of the function in the source program.
1070 This pointer must be guaranteed by the caller to be valid: loads and
1071 stores to the structure may be assumed by the callee to not to trap. This
1072 may only be applied to the first parameter. This is not a valid attribute
1073 for return values. </dd>
1074
Dan Gohmanff235352010-07-02 23:18:08 +0000[diff] [blame]1075 <dt><tt><b><a name="noalias">noalias</a></b></tt></dt>
Dan Gohman1e109622010-07-02 18:41:32 +0000[diff] [blame]1076 <dd>This indicates that pointer values
1077 <a href="#pointeraliasing"><i>based</i></a> on the argument or return
Dan Gohmanefca7f92010-07-02 23:46:54 +0000[diff] [blame]1078 value do not alias pointer values which are not <i>based</i> on it,
1079 ignoring certain "irrelevant" dependencies.
1080 For a call to the parent function, dependencies between memory
1081 references from before or after the call and from those during the call
1082 are "irrelevant" to the <tt>noalias</tt> keyword for the arguments and
1083 return value used in that call.
Dan Gohman1e109622010-07-02 18:41:32 +0000[diff] [blame]1084 The caller shares the responsibility with the callee for ensuring that
1085 these requirements are met.
1086 For further details, please see the discussion of the NoAlias response in
Dan Gohmanff70fe42010-07-06 15:26:33 +0000[diff] [blame]1087 <a href="AliasAnalysis.html#MustMayNo">alias analysis</a>.<br>
1088<br>
John McCall191d4ee2010-07-06 21:07:14 +0000[diff] [blame]1089 Note that this definition of <tt>noalias</tt> is intentionally
1090 similar to the definition of <tt>restrict</tt> in C99 for function
Chris Lattner211244a2010-07-06 20:51:35 +0000[diff] [blame]1091 arguments, though it is slightly weaker.
Dan Gohmanff70fe42010-07-06 15:26:33 +0000[diff] [blame]1092<br>
1093 For function return values, C99's <tt>restrict</tt> is not meaningful,
1094 while LLVM's <tt>noalias</tt> is.
1095 </dd>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1096
Dan Gohmanff235352010-07-02 23:18:08 +0000[diff] [blame]1097 <dt><tt><b><a name="nocapture">nocapture</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1098 <dd>This indicates that the callee does not make any copies of the pointer
1099 that outlive the callee itself. This is not a valid attribute for return
1100 values.</dd>
1101
Dan Gohmanff235352010-07-02 23:18:08 +0000[diff] [blame]1102 <dt><tt><b><a name="nest">nest</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1103 <dd>This indicates that the pointer parameter can be excised using the
1104 <a href="#int_trampoline">trampoline intrinsics</a>. This is not a valid
1105 attribute for return values.</dd>
1106</dl>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1107
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1108</div>
1109
1110<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1111<h3>
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame]1112 <a name="gc">Garbage Collector Names</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1113</h3>
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame]1114
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1115<div>
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame]1116
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1117<p>Each function may specify a garbage collector name, which is simply a
1118 string:</p>
1119
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]1120<pre class="doc_code">
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1121define void @f() gc "name" { ... }
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1122</pre>
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame]1123
1124<p>The compiler declares the supported values of <i>name</i>. Specifying a
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1125 collector which will cause the compiler to alter its output in order to
1126 support the named garbage collection algorithm.</p>
1127
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame]1128</div>
1129
1130<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1131<h3>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1132 <a name="fnattrs">Function Attributes</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1133</h3>
Devang Patelf8b94812008-09-04 23:05:13 +0000[diff] [blame]1134
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1135<div>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1136
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1137<p>Function attributes are set to communicate additional information about a
1138 function. Function attributes are considered to be part of the function, not
1139 of the function type, so functions with different parameter attributes can
1140 have the same function type.</p>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1141
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1142<p>Function attributes are simple keywords that follow the type specified. If
1143 multiple attributes are needed, they are space separated. For example:</p>
Devang Patelf8b94812008-09-04 23:05:13 +0000[diff] [blame]1144
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]1145<pre class="doc_code">
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1146define void @f() noinline { ... }
1147define void @f() alwaysinline { ... }
1148define void @f() alwaysinline optsize { ... }
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1149define void @f() optsize { ... }
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1150</pre>
Devang Patelf8b94812008-09-04 23:05:13 +0000[diff] [blame]1151
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1152<dl>
Kostya Serebryany164b86b2012-01-20 17:56:17 +0000[diff] [blame]1153 <dt><tt><b>address_safety</b></tt></dt>
1154 <dd>This attribute indicates that the address safety analysis
1155 is enabled for this function. </dd>
1156
Charles Davis1e063d12010-02-12 00:31:15 +0000[diff] [blame]1157 <dt><tt><b>alignstack(&lt;<em>n</em>&gt;)</b></tt></dt>
1158 <dd>This attribute indicates that, when emitting the prologue and epilogue,
1159 the backend should forcibly align the stack pointer. Specify the
1160 desired alignment, which must be a power of two, in parentheses.
1161
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1162 <dt><tt><b>alwaysinline</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1163 <dd>This attribute indicates that the inliner should attempt to inline this
1164 function into callers whenever possible, ignoring any active inlining size
1165 threshold for this caller.</dd>
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1166
Dan Gohman129bd562011-06-16 16:03:13 +0000[diff] [blame]1167 <dt><tt><b>nonlazybind</b></tt></dt>
1168 <dd>This attribute suppresses lazy symbol binding for the function. This
1169 may make calls to the function faster, at the cost of extra program
1170 startup time if the function is not called during program startup.</dd>
1171
Jakob Stoklund Olesen570a4a52010-02-06 01:16:28 +0000[diff] [blame]1172 <dt><tt><b>inlinehint</b></tt></dt>
1173 <dd>This attribute indicates that the source code contained a hint that inlining
1174 this function is desirable (such as the "inline" keyword in C/C++). It
1175 is just a hint; it imposes no requirements on the inliner.</dd>
1176
Nick Lewycky76ec37a2010-07-06 18:24:09 +0000[diff] [blame]1177 <dt><tt><b>naked</b></tt></dt>
1178 <dd>This attribute disables prologue / epilogue emission for the function.
1179 This can have very system-specific consequences.</dd>
1180
1181 <dt><tt><b>noimplicitfloat</b></tt></dt>
1182 <dd>This attributes disables implicit floating point instructions.</dd>
1183
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1184 <dt><tt><b>noinline</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1185 <dd>This attribute indicates that the inliner should never inline this
1186 function in any situation. This attribute may not be used together with
1187 the <tt>alwaysinline</tt> attribute.</dd>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1188
Nick Lewycky76ec37a2010-07-06 18:24:09 +0000[diff] [blame]1189 <dt><tt><b>noredzone</b></tt></dt>
1190 <dd>This attribute indicates that the code generator should not use a red
1191 zone, even if the target-specific ABI normally permits it.</dd>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1192
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1193 <dt><tt><b>noreturn</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1194 <dd>This function attribute indicates that the function never returns
1195 normally. This produces undefined behavior at runtime if the function
1196 ever does dynamically return.</dd>
Bill Wendling31359ba2008-11-13 01:02:51 +0000[diff] [blame]1197
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1198 <dt><tt><b>nounwind</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1199 <dd>This function attribute indicates that the function never returns with an
1200 unwind or exceptional control flow. If the function does unwind, its
1201 runtime behavior is undefined.</dd>
Bill Wendlingfbaa7ed2008-11-26 19:07:40 +0000[diff] [blame]1202
Nick Lewycky76ec37a2010-07-06 18:24:09 +0000[diff] [blame]1203 <dt><tt><b>optsize</b></tt></dt>
1204 <dd>This attribute suggests that optimization passes and code generator passes
1205 make choices that keep the code size of this function low, and otherwise
1206 do optimizations specifically to reduce code size.</dd>
1207
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1208 <dt><tt><b>readnone</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1209 <dd>This attribute indicates that the function computes its result (or decides
1210 to unwind an exception) based strictly on its arguments, without
1211 dereferencing any pointer arguments or otherwise accessing any mutable
1212 state (e.g. memory, control registers, etc) visible to caller functions.
1213 It does not write through any pointer arguments
1214 (including <tt><a href="#byval">byval</a></tt> arguments) and never
1215 changes any state visible to callers. This means that it cannot unwind
Bill Wendling7b9e5392012-02-06 21:57:33 +0000[diff] [blame]1216 exceptions by calling the <tt>C++</tt> exception throwing methods.</dd>
Devang Patel5d96fda2009-06-12 19:45:19 +0000[diff] [blame]1217
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1218 <dt><tt><b><a name="readonly">readonly</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1219 <dd>This attribute indicates that the function does not write through any
1220 pointer arguments (including <tt><a href="#byval">byval</a></tt>
1221 arguments) or otherwise modify any state (e.g. memory, control registers,
1222 etc) visible to caller functions. It may dereference pointer arguments
1223 and read state that may be set in the caller. A readonly function always
1224 returns the same value (or unwinds an exception identically) when called
1225 with the same set of arguments and global state. It cannot unwind an
Bill Wendling7b9e5392012-02-06 21:57:33 +0000[diff] [blame]1226 exception by calling the <tt>C++</tt> exception throwing methods.</dd>
Anton Korobeynikovc5ec8a72009-07-17 18:07:26 +0000[diff] [blame]1227
Bill Wendling9bd5d042011-12-05 21:27:54 +0000[diff] [blame]1228 <dt><tt><b><a name="returns_twice">returns_twice</a></b></tt></dt>
1229 <dd>This attribute indicates that this function can return twice. The
1230 C <code>setjmp</code> is an example of such a function. The compiler
1231 disables some optimizations (like tail calls) in the caller of these
1232 functions.</dd>
1233
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1234 <dt><tt><b><a name="ssp">ssp</a></b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1235 <dd>This attribute indicates that the function should emit a stack smashing
1236 protector. It is in the form of a "canary"&mdash;a random value placed on
1237 the stack before the local variables that's checked upon return from the
1238 function to see if it has been overwritten. A heuristic is used to
1239 determine if a function needs stack protectors or not.<br>
1240<br>
1241 If a function that has an <tt>ssp</tt> attribute is inlined into a
1242 function that doesn't have an <tt>ssp</tt> attribute, then the resulting
1243 function will have an <tt>ssp</tt> attribute.</dd>
1244
Bill Wendlingf82d40a2009-11-02 00:24:16 +0000[diff] [blame]1245 <dt><tt><b>sspreq</b></tt></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1246 <dd>This attribute indicates that the function should <em>always</em> emit a
1247 stack smashing protector. This overrides
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1248 the <tt><a href="#ssp">ssp</a></tt> function attribute.<br>
1249<br>
1250 If a function that has an <tt>sspreq</tt> attribute is inlined into a
1251 function that doesn't have an <tt>sspreq</tt> attribute or which has
1252 an <tt>ssp</tt> attribute, then the resulting function will have
1253 an <tt>sspreq</tt> attribute.</dd>
Rafael Espindolafbff0ec2011-07-25 15:27:59 +0000[diff] [blame]1254
1255 <dt><tt><b><a name="uwtable">uwtable</a></b></tt></dt>
1256 <dd>This attribute indicates that the ABI being targeted requires that
1257 an unwind table entry be produce for this function even if we can
1258 show that no exceptions passes by it. This is normally the case for
1259 the ELF x86-64 abi, but it can be disabled for some compilation
1260 units.</dd>
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1261</dl>
1262
Devang Patelf8b94812008-09-04 23:05:13 +0000[diff] [blame]1263</div>
1264
1265<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1266<h3>
Chris Lattner1eeeb0c2006-04-08 04:40:53 +0000[diff] [blame]1267 <a name="moduleasm">Module-Level Inline Assembly</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1268</h3>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]1269
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1270<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1271
1272<p>Modules may contain "module-level inline asm" blocks, which corresponds to
1273 the GCC "file scope inline asm" blocks. These blocks are internally
1274 concatenated by LLVM and treated as a single unit, but may be separated in
1275 the <tt>.ll</tt> file if desired. The syntax is very simple:</p>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]1276
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]1277<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1278module asm "inline asm code goes here"
1279module asm "more can go here"
1280</pre>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]1281
1282<p>The strings can contain any character by escaping non-printable characters.
1283 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]1284 for the number.</p>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]1285
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1286<p>The inline asm code is simply printed to the machine code .s file when
1287 assembly code is generated.</p>
1288
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]1289</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]1290
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1291<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1292<h3>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1293 <a name="datalayout">Data Layout</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1294</h3>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1295
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1296<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1297
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1298<p>A module may specify a target specific data layout string that specifies how
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1299 data is to be laid out in memory. The syntax for the data layout is
1300 simply:</p>
1301
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]1302<pre class="doc_code">
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1303target datalayout = "<i>layout specification</i>"
1304</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1305
1306<p>The <i>layout specification</i> consists of a list of specifications
1307 separated by the minus sign character ('-'). Each specification starts with
1308 a letter and may include other information after the letter to define some
1309 aspect of the data layout. The specifications accepted are as follows:</p>
1310
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1311<dl>
1312 <dt><tt>E</tt></dt>
1313 <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]1314 bits with the most significance have the lowest address location.</dd>
1315
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1316 <dt><tt>e</tt></dt>
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]1317 <dd>Specifies that the target lays out data in little-endian form. That is,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1318 the bits with the least significance have the lowest address
1319 location.</dd>
1320
Lang Hamesbb5b3f32011-10-10 23:42:08 +0000[diff] [blame]1321 <dt><tt>S<i>size</i></tt></dt>
1322 <dd>Specifies the natural alignment of the stack in bits. Alignment promotion
1323 of stack variables is limited to the natural stack alignment to avoid
1324 dynamic stack realignment. The stack alignment must be a multiple of
Lang Hames5f119a62011-10-11 17:50:14 +0000[diff] [blame]1325 8-bits. If omitted, the natural stack alignment defaults to "unspecified",
1326 which does not prevent any alignment promotions.</dd>
Lang Hamesbb5b3f32011-10-10 23:42:08 +0000[diff] [blame]1327
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1328 <dt><tt>p:<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]1329 <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]1330 <i>preferred</i> alignments. All sizes are in bits. Specifying
1331 the <i>pref</i> alignment is optional. If omitted, the
1332 preceding <tt>:</tt> should be omitted too.</dd>
1333
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1334 <dt><tt>i<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1335 <dd>This specifies the alignment for an integer type of a given bit
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1336 <i>size</i>. The value of <i>size</i> must be in the range [1,2^23).</dd>
1337
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1338 <dt><tt>v<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]1339 <dd>This specifies the alignment for a vector type of a given bit
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1340 <i>size</i>.</dd>
1341
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1342 <dt><tt>f<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]1343 <dd>This specifies the alignment for a floating point type of a given bit
Dale Johannesen9d8d2212010-05-28 18:54:47 +0000[diff] [blame]1344 <i>size</i>. Only values of <i>size</i> that are supported by the target
1345 will work. 32 (float) and 64 (double) are supported on all targets;
1346 80 or 128 (different flavors of long double) are also supported on some
1347 targets.
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1348
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1349 <dt><tt>a<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1350 <dd>This specifies the alignment for an aggregate type of a given bit
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1351 <i>size</i>.</dd>
1352
Daniel Dunbar87bde0b2009-06-08 22:17:53 +0000[diff] [blame]1353 <dt><tt>s<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1354 <dd>This specifies the alignment for a stack object of a given bit
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1355 <i>size</i>.</dd>
Chris Lattnere82bdc42009-11-07 09:35:34 +0000[diff] [blame]1356
1357 <dt><tt>n<i>size1</i>:<i>size2</i>:<i>size3</i>...</tt></dt>
1358 <dd>This specifies a set of native integer widths for the target CPU
1359 in bits. For example, it might contain "n32" for 32-bit PowerPC,
1360 "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]1361 this set are considered to support most general arithmetic
Chris Lattnere82bdc42009-11-07 09:35:34 +0000[diff] [blame]1362 operations efficiently.</dd>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1363</dl>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1364
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1365<p>When constructing the data layout for a given target, LLVM starts with a
Dan Gohman1c70c002010-04-28 00:36:01 +0000[diff] [blame]1366 default set of specifications which are then (possibly) overridden by the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1367 specifications in the <tt>datalayout</tt> keyword. The default specifications
1368 are given in this list:</p>
1369
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1370<ul>
1371 <li><tt>E</tt> - big endian</li>
Dan Gohmanfdf2e8c2010-02-23 02:44:03 +0000[diff] [blame]1372 <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]1373 <li><tt>i1:8:8</tt> - i1 is 8-bit (byte) aligned</li>
1374 <li><tt>i8:8:8</tt> - i8 is 8-bit (byte) aligned</li>
1375 <li><tt>i16:16:16</tt> - i16 is 16-bit aligned</li>
1376 <li><tt>i32:32:32</tt> - i32 is 32-bit aligned</li>
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]1377 <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]1378 alignment of 64-bits</li>
1379 <li><tt>f32:32:32</tt> - float is 32-bit aligned</li>
1380 <li><tt>f64:64:64</tt> - double is 64-bit aligned</li>
1381 <li><tt>v64:64:64</tt> - 64-bit vector is 64-bit aligned</li>
1382 <li><tt>v128:128:128</tt> - 128-bit vector is 128-bit aligned</li>
1383 <li><tt>a0:0:1</tt> - aggregates are 8-bit aligned</li>
Daniel Dunbar87bde0b2009-06-08 22:17:53 +0000[diff] [blame]1384 <li><tt>s0:64:64</tt> - stack objects are 64-bit aligned</li>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1385</ul>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1386
1387<p>When LLVM is determining the alignment for a given type, it uses the
1388 following rules:</p>
1389
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1390<ol>
1391 <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]1392 specification is used.</li>
1393
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1394 <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]1395 smallest integer type that is larger than the bitwidth of the sought type
1396 is used. If none of the specifications are larger than the bitwidth then
1397 the the largest integer type is used. For example, given the default
1398 specifications above, the i7 type will use the alignment of i8 (next
1399 largest) while both i65 and i256 will use the alignment of i64 (largest
1400 specified).</li>
1401
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1402 <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]1403 largest vector type that is smaller than the sought vector type will be
1404 used as a fall back. This happens because &lt;128 x double&gt; can be
1405 implemented in terms of 64 &lt;2 x double&gt;, for example.</li>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1406</ol>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1407
Chris Lattner6509f502011-10-11 23:01:39 +0000[diff] [blame]1408<p>The function of the data layout string may not be what you expect. Notably,
1409 this is not a specification from the frontend of what alignment the code
1410 generator should use.</p>
1411
1412<p>Instead, if specified, the target data layout is required to match what the
1413 ultimate <em>code generator</em> expects. This string is used by the
1414 mid-level optimizers to
1415 improve code, and this only works if it matches what the ultimate code
1416 generator uses. If you would like to generate IR that does not embed this
1417 target-specific detail into the IR, then you don't have to specify the
1418 string. This will disable some optimizations that require precise layout
1419 information, but this also prevents those optimizations from introducing
1420 target specificity into the IR.</p>
1421
1422
1423
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1424</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]1425
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1426<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1427<h3>
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1428 <a name="pointeraliasing">Pointer Aliasing Rules</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1429</h3>
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1430
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1431<div>
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1432
Andreas Bolka55e459a2009-07-29 00:02:05 +0000[diff] [blame]1433<p>Any memory access must be done through a pointer value associated
Andreas Bolka99a82052009-07-27 20:37:10 +0000[diff] [blame]1434with an address range of the memory access, otherwise the behavior
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1435is undefined. Pointer values are associated with address ranges
1436according to the following rules:</p>
1437
1438<ul>
Dan Gohman1e109622010-07-02 18:41:32 +0000[diff] [blame]1439 <li>A pointer value is associated with the addresses associated with
1440 any value it is <i>based</i> on.
Andreas Bolka55e459a2009-07-29 00:02:05 +0000[diff] [blame]1441 <li>An address of a global variable is associated with the address
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1442 range of the variable's storage.</li>
1443 <li>The result value of an allocation instruction is associated with
1444 the address range of the allocated storage.</li>
1445 <li>A null pointer in the default address-space is associated with
Andreas Bolka55e459a2009-07-29 00:02:05 +0000[diff] [blame]1446 no address.</li>
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1447 <li>An integer constant other than zero or a pointer value returned
1448 from a function not defined within LLVM may be associated with address
1449 ranges allocated through mechanisms other than those provided by
Andreas Bolka55e459a2009-07-29 00:02:05 +0000[diff] [blame]1450 LLVM. Such ranges shall not overlap with any ranges of addresses
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1451 allocated by mechanisms provided by LLVM.</li>
Dan Gohman1e109622010-07-02 18:41:32 +0000[diff] [blame]1452</ul>
1453
1454<p>A pointer value is <i>based</i> on another pointer value according
1455 to the following rules:</p>
1456
1457<ul>
1458 <li>A pointer value formed from a
1459 <tt><a href="#i_getelementptr">getelementptr</a></tt> operation
1460 is <i>based</i> on the first operand of the <tt>getelementptr</tt>.</li>
1461 <li>The result value of a
1462 <tt><a href="#i_bitcast">bitcast</a></tt> is <i>based</i> on the operand
1463 of the <tt>bitcast</tt>.</li>
1464 <li>A pointer value formed by an
1465 <tt><a href="#i_inttoptr">inttoptr</a></tt> is <i>based</i> on all
1466 pointer values that contribute (directly or indirectly) to the
1467 computation of the pointer's value.</li>
1468 <li>The "<i>based</i> on" relationship is transitive.</li>
1469</ul>
1470
1471<p>Note that this definition of <i>"based"</i> is intentionally
1472 similar to the definition of <i>"based"</i> in C99, though it is
1473 slightly weaker.</p>
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1474
1475<p>LLVM IR does not associate types with memory. The result type of a
Andreas Bolka55e459a2009-07-29 00:02:05 +0000[diff] [blame]1476<tt><a href="#i_load">load</a></tt> merely indicates the size and
1477alignment of the memory from which to load, as well as the
Dan Gohmanc22c0f32010-06-17 19:23:50 +0000[diff] [blame]1478interpretation of the value. The first operand type of a
Andreas Bolka55e459a2009-07-29 00:02:05 +0000[diff] [blame]1479<tt><a href="#i_store">store</a></tt> similarly only indicates the size
1480and alignment of the store.</p>
Dan Gohman556ca272009-07-27 18:07:55 +0000[diff] [blame]1481
1482<p>Consequently, type-based alias analysis, aka TBAA, aka
1483<tt>-fstrict-aliasing</tt>, is not applicable to general unadorned
1484LLVM IR. <a href="#metadata">Metadata</a> may be used to encode
1485additional information which specialized optimization passes may use
1486to implement type-based alias analysis.</p>
1487
1488</div>
1489
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]1490<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1491<h3>
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]1492 <a name="volatile">Volatile Memory Accesses</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1493</h3>
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]1494
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1495<div>
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]1496
1497<p>Certain memory accesses, such as <a href="#i_load"><tt>load</tt></a>s, <a
1498href="#i_store"><tt>store</tt></a>s, and <a
1499href="#int_memcpy"><tt>llvm.memcpy</tt></a>s may be marked <tt>volatile</tt>.
1500The optimizers must not change the number of volatile operations or change their
1501order of execution relative to other volatile operations. The optimizers
1502<i>may</i> change the order of volatile operations relative to non-volatile
1503operations. This is not Java's "volatile" and has no cross-thread
1504synchronization behavior.</p>
1505
1506</div>
1507
Eli Friedman5b60e1b2011-07-20 21:35:53 +0000[diff] [blame]1508<!-- ======================================================================= -->
1509<h3>
1510 <a name="memmodel">Memory Model for Concurrent Operations</a>
1511</h3>
1512
1513<div>
1514
1515<p>The LLVM IR does not define any way to start parallel threads of execution
1516or to register signal handlers. Nonetheless, there are platform-specific
1517ways to create them, and we define LLVM IR's behavior in their presence. This
1518model is inspired by the C++0x memory model.</p>
1519
Eli Friedman234bccd2011-08-22 21:35:27 +0000[diff] [blame]1520<p>For a more informal introduction to this model, see the
1521<a href="Atomics.html">LLVM Atomic Instructions and Concurrency Guide</a>.
1522
Eli Friedman5b60e1b2011-07-20 21:35:53 +0000[diff] [blame]1523<p>We define a <i>happens-before</i> partial order as the least partial order
1524that</p>
1525<ul>
1526 <li>Is a superset of single-thread program order, and</li>
1527 <li>When a <i>synchronizes-with</i> <tt>b</tt>, includes an edge from
1528 <tt>a</tt> to <tt>b</tt>. <i>Synchronizes-with</i> pairs are introduced
1529 by platform-specific techniques, like pthread locks, thread
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]1530 creation, thread joining, etc., and by atomic instructions.
1531 (See also <a href="#ordering">Atomic Memory Ordering Constraints</a>).
1532 </li>
Eli Friedman5b60e1b2011-07-20 21:35:53 +0000[diff] [blame]1533</ul>
1534
1535<p>Note that program order does not introduce <i>happens-before</i> edges
1536between a thread and signals executing inside that thread.</p>
1537
1538<p>Every (defined) read operation (load instructions, memcpy, atomic
1539loads/read-modify-writes, etc.) <var>R</var> reads a series of bytes written by
1540(defined) write operations (store instructions, atomic
Eli Friedman118973a2011-07-22 03:04:45 +0000[diff] [blame]1541stores/read-modify-writes, memcpy, etc.). For the purposes of this section,
1542initialized globals are considered to have a write of the initializer which is
1543atomic and happens before any other read or write of the memory in question.
1544For each byte of a read <var>R</var>, <var>R<sub>byte</sub></var> may see
1545any write to the same byte, except:</p>
Eli Friedman5b60e1b2011-07-20 21:35:53 +0000[diff] [blame]1546
1547<ul>
1548 <li>If <var>write<sub>1</sub></var> happens before
1549 <var>write<sub>2</sub></var>, and <var>write<sub>2</sub></var> happens
1550 before <var>R<sub>byte</sub></var>, then <var>R<sub>byte</sub></var>
Eli Friedman118973a2011-07-22 03:04:45 +0000[diff] [blame]1551 does not see <var>write<sub>1</sub></var>.
Bill Wendling0246bb72011-07-31 06:45:03 +0000[diff] [blame]1552 <li>If <var>R<sub>byte</sub></var> happens before
1553 <var>write<sub>3</sub></var>, then <var>R<sub>byte</sub></var> does not
1554 see <var>write<sub>3</sub></var>.
Eli Friedman5b60e1b2011-07-20 21:35:53 +0000[diff] [blame]1555</ul>
1556
1557<p>Given that definition, <var>R<sub>byte</sub></var> is defined as follows:
1558<ul>
Eli Friedman234bccd2011-08-22 21:35:27 +0000[diff] [blame]1559 <li>If <var>R</var> is volatile, the result is target-dependent. (Volatile
1560 is supposed to give guarantees which can support
1561 <code>sig_atomic_t</code> in C/C++, and may be used for accesses to
1562 addresses which do not behave like normal memory. It does not generally
1563 provide cross-thread synchronization.)
1564 <li>Otherwise, if there is no write to the same byte that happens before
Eli Friedman5b60e1b2011-07-20 21:35:53 +0000[diff] [blame]1565 <var>R<sub>byte</sub></var>, <var>R<sub>byte</sub></var> returns
1566 <tt>undef</tt> for that byte.
Eli Friedman118973a2011-07-22 03:04:45 +0000[diff] [blame]1567 <li>Otherwise, if <var>R<sub>byte</sub></var> may see exactly one write,
Eli Friedman5b60e1b2011-07-20 21:35:53 +0000[diff] [blame]1568 <var>R<sub>byte</sub></var> returns the value written by that
1569 write.</li>
Eli Friedman118973a2011-07-22 03:04:45 +0000[diff] [blame]1570 <li>Otherwise, if <var>R</var> is atomic, and all the writes
1571 <var>R<sub>byte</sub></var> may see are atomic, it chooses one of the
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]1572 values written. See the <a href="#ordering">Atomic Memory Ordering
1573 Constraints</a> section for additional constraints on how the choice
1574 is made.
Eli Friedman5b60e1b2011-07-20 21:35:53 +0000[diff] [blame]1575 <li>Otherwise <var>R<sub>byte</sub></var> returns <tt>undef</tt>.</li>
1576</ul>
1577
1578<p><var>R</var> returns the value composed of the series of bytes it read.
1579This implies that some bytes within the value may be <tt>undef</tt>
1580<b>without</b> the entire value being <tt>undef</tt>. Note that this only
1581defines the semantics of the operation; it doesn't mean that targets will
1582emit more than one instruction to read the series of bytes.</p>
1583
1584<p>Note that in cases where none of the atomic intrinsics are used, this model
1585places only one restriction on IR transformations on top of what is required
1586for single-threaded execution: introducing a store to a byte which might not
Eli Friedman101c81d2011-08-02 01:15:34 +0000[diff] [blame]1587otherwise be stored is not allowed in general. (Specifically, in the case
1588where another thread might write to and read from an address, introducing a
1589store can change a load that may see exactly one write into a load that may
1590see multiple writes.)</p>
Eli Friedman5b60e1b2011-07-20 21:35:53 +0000[diff] [blame]1591
1592<!-- FIXME: This model assumes all targets where concurrency is relevant have
1593a byte-size store which doesn't affect adjacent bytes. As far as I can tell,
1594none of the backends currently in the tree fall into this category; however,
1595there might be targets which care. If there are, we want a paragraph
1596like the following:
1597
1598Targets may specify that stores narrower than a certain width are not
1599available; on such a target, for the purposes of this model, treat any
1600non-atomic write with an alignment or width less than the minimum width
1601as if it writes to the relevant surrounding bytes.
1602-->
1603
1604</div>
1605
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]1606<!-- ======================================================================= -->
NAKAMURA Takumib5bb29c2011-08-12 06:17:17 +0000[diff] [blame]1607<h3>
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]1608 <a name="ordering">Atomic Memory Ordering Constraints</a>
NAKAMURA Takumib5bb29c2011-08-12 06:17:17 +0000[diff] [blame]1609</h3>
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]1610
NAKAMURA Takumib5bb29c2011-08-12 06:17:17 +0000[diff] [blame]1611<div>
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]1612
1613<p>Atomic instructions (<a href="#i_cmpxchg"><code>cmpxchg</code></a>,
Eli Friedman21006d42011-08-09 23:02:53 +0000[diff] [blame]1614<a href="#i_atomicrmw"><code>atomicrmw</code></a>,
1615<a href="#i_fence"><code>fence</code></a>,
1616<a href="#i_load"><code>atomic load</code></a>, and
Eli Friedman8fa281a2011-08-09 23:26:12 +0000[diff] [blame]1617<a href="#i_store"><code>atomic store</code></a>) take an ordering parameter
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]1618that determines which other atomic instructions on the same address they
1619<i>synchronize with</i>. These semantics are borrowed from Java and C++0x,
1620but are somewhat more colloquial. If these descriptions aren't precise enough,
Eli Friedman234bccd2011-08-22 21:35:27 +0000[diff] [blame]1621check those specs (see spec references in the
Nick Lewycky300a2632012-01-23 08:47:21 +0000[diff] [blame]1622<a href="Atomics.html#introduction">atomics guide</a>).
Eli Friedman234bccd2011-08-22 21:35:27 +0000[diff] [blame]1623<a href="#i_fence"><code>fence</code></a> instructions
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]1624treat these orderings somewhat differently since they don't take an address.
1625See that instruction's documentation for details.</p>
1626
Eli Friedman234bccd2011-08-22 21:35:27 +0000[diff] [blame]1627<p>For a simpler introduction to the ordering constraints, see the
1628<a href="Atomics.html">LLVM Atomic Instructions and Concurrency Guide</a>.</p>
1629
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]1630<dl>
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]1631<dt><code>unordered</code></dt>
1632<dd>The set of values that can be read is governed by the happens-before
1633partial order. A value cannot be read unless some operation wrote it.
1634This is intended to provide a guarantee strong enough to model Java's
1635non-volatile shared variables. This ordering cannot be specified for
1636read-modify-write operations; it is not strong enough to make them atomic
1637in any interesting way.</dd>
1638<dt><code>monotonic</code></dt>
1639<dd>In addition to the guarantees of <code>unordered</code>, there is a single
1640total order for modifications by <code>monotonic</code> operations on each
1641address. All modification orders must be compatible with the happens-before
1642order. There is no guarantee that the modification orders can be combined to
1643a global total order for the whole program (and this often will not be
1644possible). The read in an atomic read-modify-write operation
1645(<a href="#i_cmpxchg"><code>cmpxchg</code></a> and
1646<a href="#i_atomicrmw"><code>atomicrmw</code></a>)
1647reads the value in the modification order immediately before the value it
1648writes. If one atomic read happens before another atomic read of the same
1649address, the later read must see the same value or a later value in the
1650address's modification order. This disallows reordering of
1651<code>monotonic</code> (or stronger) operations on the same address. If an
1652address is written <code>monotonic</code>ally by one thread, and other threads
1653<code>monotonic</code>ally read that address repeatedly, the other threads must
Eli Friedman234bccd2011-08-22 21:35:27 +0000[diff] [blame]1654eventually see the write. This corresponds to the C++0x/C1x
1655<code>memory_order_relaxed</code>.</dd>
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]1656<dt><code>acquire</code></dt>
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]1657<dd>In addition to the guarantees of <code>monotonic</code>,
Eli Friedmanc264b2f2011-08-24 20:28:39 +0000[diff] [blame]1658a <i>synchronizes-with</i> edge may be formed with a <code>release</code>
1659operation. This is intended to model C++'s <code>memory_order_acquire</code>.</dd>
1660<dt><code>release</code></dt>
1661<dd>In addition to the guarantees of <code>monotonic</code>, if this operation
1662writes a value which is subsequently read by an <code>acquire</code> operation,
1663it <i>synchronizes-with</i> that operation. (This isn't a complete
1664description; see the C++0x definition of a release sequence.) This corresponds
1665to the C++0x/C1x <code>memory_order_release</code>.</dd>
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]1666<dt><code>acq_rel</code> (acquire+release)</dt><dd>Acts as both an
Eli Friedman234bccd2011-08-22 21:35:27 +0000[diff] [blame]1667<code>acquire</code> and <code>release</code> operation on its address.
1668This corresponds to the C++0x/C1x <code>memory_order_acq_rel</code>.</dd>
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]1669<dt><code>seq_cst</code> (sequentially consistent)</dt><dd>
1670<dd>In addition to the guarantees of <code>acq_rel</code>
1671(<code>acquire</code> for an operation which only reads, <code>release</code>
1672for an operation which only writes), there is a global total order on all
1673sequentially-consistent operations on all addresses, which is consistent with
1674the <i>happens-before</i> partial order and with the modification orders of
1675all the affected addresses. Each sequentially-consistent read sees the last
Eli Friedman234bccd2011-08-22 21:35:27 +0000[diff] [blame]1676preceding write to the same address in this global order. This corresponds
1677to the C++0x/C1x <code>memory_order_seq_cst</code> and Java volatile.</dd>
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]1678</dl>
1679
1680<p id="singlethread">If an atomic operation is marked <code>singlethread</code>,
1681it only <i>synchronizes with</i> or participates in modification and seq_cst
1682total orderings with other operations running in the same thread (for example,
1683in signal handlers).</p>
1684
1685</div>
1686
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1687</div>
1688
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1689<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1690<h2><a name="typesystem">Type System</a></h2>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1691<!-- *********************************************************************** -->
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]1692
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1693<div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]1694
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1695<p>The LLVM type system is one of the most important features of the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1696 intermediate representation. Being typed enables a number of optimizations
1697 to be performed on the intermediate representation directly, without having
1698 to do extra analyses on the side before the transformation. A strong type
1699 system makes it easier to read the generated code and enables novel analyses
1700 and transformations that are not feasible to perform on normal three address
1701 code representations.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]1702
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1703<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1704<h3>
1705 <a name="t_classifications">Type Classifications</a>
1706</h3>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1707
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1708<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1709
1710<p>The types fall into a few useful classifications:</p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1711
1712<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1713 <tbody>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1714 <tr><th>Classification</th><th>Types</th></tr>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1715 <tr>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1716 <td><a href="#t_integer">integer</a></td>
Reid Spencer2b916312007-05-16 18:44:01 +0000[diff] [blame]1717 <td><tt>i1, i2, i3, ... i8, ... i16, ... i32, ... i64, ... </tt></td>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1718 </tr>
1719 <tr>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1720 <td><a href="#t_floating">floating point</a></td>
Dan Gohmance163392011-12-17 00:04:22 +0000[diff] [blame]1721 <td><tt>half, float, double, x86_fp80, fp128, ppc_fp128</tt></td>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1722 </tr>
1723 <tr>
1724 <td><a name="t_firstclass">first class</a></td>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1725 <td><a href="#t_integer">integer</a>,
1726 <a href="#t_floating">floating point</a>,
1727 <a href="#t_pointer">pointer</a>,
Dan Gohman0066db62008-06-18 18:42:13 +0000[diff] [blame]1728 <a href="#t_vector">vector</a>,
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]1729 <a href="#t_struct">structure</a>,
1730 <a href="#t_array">array</a>,
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1731 <a href="#t_label">label</a>,
1732 <a href="#t_metadata">metadata</a>.
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1733 </td>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1734 </tr>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1735 <tr>
1736 <td><a href="#t_primitive">primitive</a></td>
1737 <td><a href="#t_label">label</a>,
1738 <a href="#t_void">void</a>,
Tobias Grosser05387292010-12-28 20:29:31 +0000[diff] [blame]1739 <a href="#t_integer">integer</a>,
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1740 <a href="#t_floating">floating point</a>,
Dale Johannesen21fe99b2010-10-01 00:48:59 +0000[diff] [blame]1741 <a href="#t_x86mmx">x86mmx</a>,
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1742 <a href="#t_metadata">metadata</a>.</td>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1743 </tr>
1744 <tr>
1745 <td><a href="#t_derived">derived</a></td>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]1746 <td><a href="#t_array">array</a>,
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1747 <a href="#t_function">function</a>,
1748 <a href="#t_pointer">pointer</a>,
1749 <a href="#t_struct">structure</a>,
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1750 <a href="#t_vector">vector</a>,
1751 <a href="#t_opaque">opaque</a>.
Dan Gohman01ac1012008-10-14 16:32:04 +0000[diff] [blame]1752 </td>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1753 </tr>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1754 </tbody>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1755</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1756
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1757<p>The <a href="#t_firstclass">first class</a> types are perhaps the most
1758 important. Values of these types are the only ones which can be produced by
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]1759 instructions.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1760
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1761</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1762
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1763<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1764<h3>
1765 <a name="t_primitive">Primitive Types</a>
1766</h3>
Chris Lattner8f8c7b72008-01-04 04:34:14 +0000[diff] [blame]1767
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1768<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1769
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1770<p>The primitive types are the fundamental building blocks of the LLVM
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1771 system.</p>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1772
1773<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1774<h4>
1775 <a name="t_integer">Integer Type</a>
1776</h4>
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]1777
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1778<div>
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]1779
1780<h5>Overview:</h5>
1781<p>The integer type is a very simple type that simply specifies an arbitrary
1782 bit width for the integer type desired. Any bit width from 1 bit to
1783 2<sup>23</sup>-1 (about 8 million) can be specified.</p>
1784
1785<h5>Syntax:</h5>
1786<pre>
1787 iN
1788</pre>
1789
1790<p>The number of bits the integer will occupy is specified by the <tt>N</tt>
1791 value.</p>
1792
1793<h5>Examples:</h5>
1794<table class="layout">
1795 <tr class="layout">
1796 <td class="left"><tt>i1</tt></td>
1797 <td class="left">a single-bit integer.</td>
1798 </tr>
1799 <tr class="layout">
1800 <td class="left"><tt>i32</tt></td>
1801 <td class="left">a 32-bit integer.</td>
1802 </tr>
1803 <tr class="layout">
1804 <td class="left"><tt>i1942652</tt></td>
1805 <td class="left">a really big integer of over 1 million bits.</td>
1806 </tr>
1807</table>
1808
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]1809</div>
1810
1811<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1812<h4>
1813 <a name="t_floating">Floating Point Types</a>
1814</h4>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1815
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1816<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1817
1818<table>
1819 <tbody>
1820 <tr><th>Type</th><th>Description</th></tr>
Dan Gohmance163392011-12-17 00:04:22 +0000[diff] [blame]1821 <tr><td><tt>half</tt></td><td>16-bit floating point value</td></tr>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1822 <tr><td><tt>float</tt></td><td>32-bit floating point value</td></tr>
1823 <tr><td><tt>double</tt></td><td>64-bit floating point value</td></tr>
1824 <tr><td><tt>fp128</tt></td><td>128-bit floating point value (112-bit mantissa)</td></tr>
1825 <tr><td><tt>x86_fp80</tt></td><td>80-bit floating point value (X87)</td></tr>
1826 <tr><td><tt>ppc_fp128</tt></td><td>128-bit floating point value (two 64-bits)</td></tr>
1827 </tbody>
1828</table>
1829
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1830</div>
1831
1832<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1833<h4>
1834 <a name="t_x86mmx">X86mmx Type</a>
1835</h4>
Dale Johannesen21fe99b2010-10-01 00:48:59 +0000[diff] [blame]1836
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1837<div>
Dale Johannesen21fe99b2010-10-01 00:48:59 +0000[diff] [blame]1838
1839<h5>Overview:</h5>
1840<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>
1841
1842<h5>Syntax:</h5>
1843<pre>
Dale Johannesen473a8c82010-10-01 01:07:02 +0000[diff] [blame]1844 x86mmx
Dale Johannesen21fe99b2010-10-01 00:48:59 +0000[diff] [blame]1845</pre>
1846
1847</div>
1848
1849<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1850<h4>
1851 <a name="t_void">Void Type</a>
1852</h4>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1853
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1854<div>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1855
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1856<h5>Overview:</h5>
1857<p>The void type does not represent any value and has no size.</p>
1858
1859<h5>Syntax:</h5>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1860<pre>
1861 void
1862</pre>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1863
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1864</div>
1865
1866<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1867<h4>
1868 <a name="t_label">Label Type</a>
1869</h4>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1870
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1871<div>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1872
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1873<h5>Overview:</h5>
1874<p>The label type represents code labels.</p>
1875
1876<h5>Syntax:</h5>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1877<pre>
1878 label
1879</pre>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1880
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1881</div>
1882
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1883<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1884<h4>
1885 <a name="t_metadata">Metadata Type</a>
1886</h4>
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1887
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1888<div>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1889
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1890<h5>Overview:</h5>
Nick Lewyckyc261df92009-09-27 23:27:42 +0000[diff] [blame]1891<p>The metadata type represents embedded metadata. No derived types may be
1892 created from metadata except for <a href="#t_function">function</a>
1893 arguments.
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1894
1895<h5>Syntax:</h5>
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1896<pre>
1897 metadata
1898</pre>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]1899
Nick Lewycky7a0370f2009-05-30 05:06:04 +0000[diff] [blame]1900</div>
1901
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1902</div>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1903
1904<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1905<h3>
1906 <a name="t_derived">Derived Types</a>
1907</h3>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1908
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1909<div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1910
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1911<p>The real power in LLVM comes from the derived types in the system. This is
1912 what allows a programmer to represent arrays, functions, pointers, and other
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]1913 useful types. Each of these types contain one or more element types which
1914 may be a primitive type, or another derived type. For example, it is
1915 possible to have a two dimensional array, using an array as the element type
1916 of another array.</p>
Dan Gohmand8791e52009-01-24 15:58:40 +0000[diff] [blame]1917
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]1918<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1919<h4>
1920 <a name="t_aggregate">Aggregate Types</a>
1921</h4>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]1922
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1923<div>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]1924
1925<p>Aggregate Types are a subset of derived types that can contain multiple
Duncan Sands20536b52011-12-14 15:44:20 +0000[diff] [blame]1926 member types. <a href="#t_array">Arrays</a> and
1927 <a href="#t_struct">structs</a> are aggregate types.
1928 <a href="#t_vector">Vectors</a> are not considered to be aggregate types.</p>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]1929
1930</div>
1931
Reid Spencer2b916312007-05-16 18:44:01 +0000[diff] [blame]1932<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1933<h4>
1934 <a name="t_array">Array Type</a>
1935</h4>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1936
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1937<div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1938
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1939<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1940<p>The array type is a very simple derived type that arranges elements
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1941 sequentially in memory. The array type requires a size (number of elements)
1942 and an underlying data type.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1943
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1944<h5>Syntax:</h5>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1945<pre>
1946 [&lt;# elements&gt; x &lt;elementtype&gt;]
1947</pre>
1948
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1949<p>The number of elements is a constant integer value; <tt>elementtype</tt> may
1950 be any type with a size.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1951
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1952<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1953<table class="layout">
1954 <tr class="layout">
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]1955 <td class="left"><tt>[40 x i32]</tt></td>
1956 <td class="left">Array of 40 32-bit integer values.</td>
1957 </tr>
1958 <tr class="layout">
1959 <td class="left"><tt>[41 x i32]</tt></td>
1960 <td class="left">Array of 41 32-bit integer values.</td>
1961 </tr>
1962 <tr class="layout">
1963 <td class="left"><tt>[4 x i8]</tt></td>
1964 <td class="left">Array of 4 8-bit integer values.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1965 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1966</table>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1967<p>Here are some examples of multidimensional arrays:</p>
1968<table class="layout">
1969 <tr class="layout">
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]1970 <td class="left"><tt>[3 x [4 x i32]]</tt></td>
1971 <td class="left">3x4 array of 32-bit integer values.</td>
1972 </tr>
1973 <tr class="layout">
1974 <td class="left"><tt>[12 x [10 x float]]</tt></td>
1975 <td class="left">12x10 array of single precision floating point values.</td>
1976 </tr>
1977 <tr class="layout">
1978 <td class="left"><tt>[2 x [3 x [4 x i16]]]</tt></td>
1979 <td class="left">2x3x4 array of 16-bit integer values.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1980 </tr>
1981</table>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]1982
Dan Gohman7657f6b2009-11-09 19:01:53 +0000[diff] [blame]1983<p>There is no restriction on indexing beyond the end of the array implied by
1984 a static type (though there are restrictions on indexing beyond the bounds
1985 of an allocated object in some cases). This means that single-dimension
1986 'variable sized array' addressing can be implemented in LLVM with a zero
1987 length array type. An implementation of 'pascal style arrays' in LLVM could
1988 use the type "<tt>{ i32, [0 x float]}</tt>", for example.</p>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]1989
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1990</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1991
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1992<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1993<h4>
1994 <a name="t_function">Function Type</a>
1995</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]1996
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1997<div>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]1998
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1999<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2000<p>The function type can be thought of as a function signature. It consists of
2001 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]2002 function type is a first class type or a void type.</p>
Devang Patelc3fc6df2008-03-10 20:49:15 +0000[diff] [blame]2003
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2004<h5>Syntax:</h5>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]2005<pre>
Nick Lewycky51386942009-09-27 07:55:32 +0000[diff] [blame]2006 &lt;returntype&gt; (&lt;parameter list&gt;)
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]2007</pre>
2008
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]2009<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]2010 specifiers. Optionally, the parameter list may include a type <tt>...</tt>,
2011 which indicates that the function takes a variable number of arguments.
2012 Variable argument functions can access their arguments with
2013 the <a href="#int_varargs">variable argument handling intrinsic</a>
Chris Lattner0724fbd2010-03-02 06:36:51 +0000[diff] [blame]2014 functions. '<tt>&lt;returntype&gt;</tt>' is any type except
Nick Lewyckyc261df92009-09-27 23:27:42 +0000[diff] [blame]2015 <a href="#t_label">label</a>.</p>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]2016
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2017<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]2018<table class="layout">
2019 <tr class="layout">
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]2020 <td class="left"><tt>i32 (i32)</tt></td>
2021 <td class="left">function taking an <tt>i32</tt>, returning an <tt>i32</tt>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]2022 </td>
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]2023 </tr><tr class="layout">
Chris Lattner0724fbd2010-03-02 06:36:51 +0000[diff] [blame]2024 <td class="left"><tt>float&nbsp;(i16,&nbsp;i32&nbsp;*)&nbsp;*
Reid Spencerf17a0b72006-12-31 07:20:23 +0000[diff] [blame]2025 </tt></td>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]2026 <td class="left"><a href="#t_pointer">Pointer</a> to a function that takes
Chris Lattner0724fbd2010-03-02 06:36:51 +0000[diff] [blame]2027 an <tt>i16</tt> and a <a href="#t_pointer">pointer</a> to <tt>i32</tt>,
2028 returning <tt>float</tt>.
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]2029 </td>
2030 </tr><tr class="layout">
2031 <td class="left"><tt>i32 (i8*, ...)</tt></td>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]2032 <td class="left">A vararg function that takes at least one
2033 <a href="#t_pointer">pointer</a> to <tt>i8 </tt> (char in C),
2034 which returns an integer. This is the signature for <tt>printf</tt> in
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]2035 LLVM.
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]2036 </td>
Devang Patela582f402008-03-24 05:35:41 +0000[diff] [blame]2037 </tr><tr class="layout">
2038 <td class="left"><tt>{i32, i32} (i32)</tt></td>
Nick Lewycky51386942009-09-27 07:55:32 +0000[diff] [blame]2039 <td class="left">A function taking an <tt>i32</tt>, returning a
2040 <a href="#t_struct">structure</a> containing two <tt>i32</tt> values
Devang Patela582f402008-03-24 05:35:41 +0000[diff] [blame]2041 </td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]2042 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2043</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]2044
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2045</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2046
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2047<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2048<h4>
2049 <a name="t_struct">Structure Type</a>
2050</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2051
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2052<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2053
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2054<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2055<p>The structure type is used to represent a collection of data members together
Chris Lattner1afcace2011-07-09 17:41:24 +0000[diff] [blame]2056 in memory. The elements of a structure may be any type that has a size.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2057
Jeffrey Yasskin7a088cf2010-01-11 19:19:26 +0000[diff] [blame]2058<p>Structures in memory are accessed using '<tt><a href="#i_load">load</a></tt>'
2059 and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a field
2060 with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.
2061 Structures in registers are accessed using the
2062 '<tt><a href="#i_extractvalue">extractvalue</a></tt>' and
2063 '<tt><a href="#i_insertvalue">insertvalue</a></tt>' instructions.</p>
Chris Lattner1afcace2011-07-09 17:41:24 +0000[diff] [blame]2064
2065<p>Structures may optionally be "packed" structures, which indicate that the
2066 alignment of the struct is one byte, and that there is no padding between
Chris Lattner2c38d652011-08-12 17:31:02 +0000[diff] [blame]2067 the elements. In non-packed structs, padding between field types is inserted
2068 as defined by the TargetData string in the module, which is required to match
Chris Lattnere4617b02011-10-11 23:02:17 +0000[diff] [blame]2069 what the underlying code generator expects.</p>
Chris Lattner1afcace2011-07-09 17:41:24 +0000[diff] [blame]2070
Chris Lattner2c38d652011-08-12 17:31:02 +0000[diff] [blame]2071<p>Structures can either be "literal" or "identified". A literal structure is
2072 defined inline with other types (e.g. <tt>{i32, i32}*</tt>) whereas identified
2073 types are always defined at the top level with a name. Literal types are
2074 uniqued by their contents and can never be recursive or opaque since there is
Chris Lattneraa175c32011-08-12 18:12:40 +0000[diff] [blame]2075 no way to write one. Identified types can be recursive, can be opaqued, and are
Chris Lattner2c38d652011-08-12 17:31:02 +0000[diff] [blame]2076 never uniqued.
Chris Lattner1afcace2011-07-09 17:41:24 +0000[diff] [blame]2077</p>
2078
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2079<h5>Syntax:</h5>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]2080<pre>
Chris Lattner2c38d652011-08-12 17:31:02 +0000[diff] [blame]2081 %T1 = type { &lt;type list&gt; } <i>; Identified normal struct type</i>
2082 %T2 = type &lt;{ &lt;type list&gt; }&gt; <i>; Identified packed struct type</i>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]2083</pre>
Chris Lattner1afcace2011-07-09 17:41:24 +0000[diff] [blame]2084
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2085<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]2086<table class="layout">
2087 <tr class="layout">
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]2088 <td class="left"><tt>{ i32, i32, i32 }</tt></td>
2089 <td class="left">A triple of three <tt>i32</tt> values</td>
Chris Lattner1afcace2011-07-09 17:41:24 +0000[diff] [blame]2090 </tr>
2091 <tr class="layout">
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]2092 <td class="left"><tt>{&nbsp;float,&nbsp;i32&nbsp;(i32)&nbsp;*&nbsp;}</tt></td>
2093 <td class="left">A pair, where the first element is a <tt>float</tt> and the
2094 second element is a <a href="#t_pointer">pointer</a> to a
2095 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
2096 an <tt>i32</tt>.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]2097 </tr>
Chris Lattner1afcace2011-07-09 17:41:24 +0000[diff] [blame]2098 <tr class="layout">
2099 <td class="left"><tt>&lt;{ i8, i32 }&gt;</tt></td>
2100 <td class="left">A packed struct known to be 5 bytes in size.</td>
2101 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2102</table>
Dan Gohmand8791e52009-01-24 15:58:40 +0000[diff] [blame]2103
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2104</div>
Chris Lattner1afcace2011-07-09 17:41:24 +0000[diff] [blame]2105
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2106<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2107<h4>
Chris Lattner628ed392011-07-23 19:59:08 +0000[diff] [blame]2108 <a name="t_opaque">Opaque Structure Types</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2109</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2110
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2111<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2112
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]2113<h5>Overview:</h5>
Chris Lattner628ed392011-07-23 19:59:08 +0000[diff] [blame]2114<p>Opaque structure types are used to represent named structure types that do
2115 not have a body specified. This corresponds (for example) to the C notion of
2116 a forward declared structure.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2117
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]2118<h5>Syntax:</h5>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]2119<pre>
Chris Lattner1afcace2011-07-09 17:41:24 +0000[diff] [blame]2120 %X = type opaque
2121 %52 = type opaque
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]2122</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2123
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]2124<h5>Examples:</h5>
2125<table class="layout">
2126 <tr class="layout">
Chris Lattner1afcace2011-07-09 17:41:24 +0000[diff] [blame]2127 <td class="left"><tt>opaque</tt></td>
2128 <td class="left">An opaque type.</td>
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]2129 </tr>
2130</table>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2131
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]2132</div>
2133
Chris Lattner1afcace2011-07-09 17:41:24 +0000[diff] [blame]2134
2135
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]2136<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2137<h4>
2138 <a name="t_pointer">Pointer Type</a>
2139</h4>
Chris Lattner0fd4a272009-02-08 19:53:29 +0000[diff] [blame]2140
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2141<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2142
2143<h5>Overview:</h5>
Dan Gohmanff3ef322010-02-25 16:50:07 +0000[diff] [blame]2144<p>The pointer type is used to specify memory locations.
2145 Pointers are commonly used to reference objects in memory.</p>
2146
2147<p>Pointer types may have an optional address space attribute defining the
2148 numbered address space where the pointed-to object resides. The default
2149 address space is number zero. The semantics of non-zero address
2150 spaces are target-specific.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2151
2152<p>Note that LLVM does not permit pointers to void (<tt>void*</tt>) nor does it
2153 permit pointers to labels (<tt>label*</tt>). Use <tt>i8*</tt> instead.</p>
Chris Lattner0fd4a272009-02-08 19:53:29 +0000[diff] [blame]2154
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2155<h5>Syntax:</h5>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]2156<pre>
2157 &lt;type&gt; *
2158</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2159
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2160<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]2161<table class="layout">
2162 <tr class="layout">
Dan Gohman2a08c532009-01-04 23:44:43 +0000[diff] [blame]2163 <td class="left"><tt>[4 x i32]*</tt></td>
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]2164 <td class="left">A <a href="#t_pointer">pointer</a> to <a
2165 href="#t_array">array</a> of four <tt>i32</tt> values.</td>
2166 </tr>
2167 <tr class="layout">
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]2168 <td class="left"><tt>i32 (i32*) *</tt></td>
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]2169 <td class="left"> A <a href="#t_pointer">pointer</a> to a <a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2170 href="#t_function">function</a> that takes an <tt>i32*</tt>, returning an
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]2171 <tt>i32</tt>.</td>
2172 </tr>
2173 <tr class="layout">
2174 <td class="left"><tt>i32 addrspace(5)*</tt></td>
2175 <td class="left">A <a href="#t_pointer">pointer</a> to an <tt>i32</tt> value
2176 that resides in address space #5.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]2177 </tr>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2178</table>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2179
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2180</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]2181
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]2182<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2183<h4>
2184 <a name="t_vector">Vector Type</a>
2185</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2186
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2187<div>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]2188
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]2189<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2190<p>A vector type is a simple derived type that represents a vector of elements.
2191 Vector types are used when multiple primitive data are operated in parallel
2192 using a single instruction (SIMD). A vector type requires a size (number of
Duncan Sandsd40d14e2009-11-27 13:38:03 +0000[diff] [blame]2193 elements) and an underlying primitive data type. Vector types are considered
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2194 <a href="#t_firstclass">first class</a>.</p>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]2195
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]2196<h5>Syntax:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]2197<pre>
2198 &lt; &lt;# elements&gt; x &lt;elementtype&gt; &gt;
2199</pre>
2200
Chris Lattner7d2e7be2010-10-10 18:20:35 +0000[diff] [blame]2201<p>The number of elements is a constant integer value larger than 0; elementtype
Nadav Rotem16087692011-12-05 06:29:09 +0000[diff] [blame]2202 may be any integer or floating point type, or a pointer to these types.
2203 Vectors of size zero are not allowed. </p>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]2204
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]2205<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]2206<table class="layout">
2207 <tr class="layout">
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]2208 <td class="left"><tt>&lt;4 x i32&gt;</tt></td>
2209 <td class="left">Vector of 4 32-bit integer values.</td>
2210 </tr>
2211 <tr class="layout">
2212 <td class="left"><tt>&lt;8 x float&gt;</tt></td>
2213 <td class="left">Vector of 8 32-bit floating-point values.</td>
2214 </tr>
2215 <tr class="layout">
2216 <td class="left"><tt>&lt;2 x i64&gt;</tt></td>
2217 <td class="left">Vector of 2 64-bit integer values.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]2218 </tr>
Nadav Rotem16087692011-12-05 06:29:09 +0000[diff] [blame]2219 <tr class="layout">
2220 <td class="left"><tt>&lt;4 x i64*&gt;</tt></td>
2221 <td class="left">Vector of 4 pointers to 64-bit integer values.</td>
2222 </tr>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]2223</table>
Dan Gohmand8791e52009-01-24 15:58:40 +0000[diff] [blame]2224
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2225</div>
2226
Bill Wendlingaf75f0c2011-07-31 06:47:33 +0000[diff] [blame]2227</div>
2228
NAKAMURA Takumi4b2e07a2011-10-31 13:04:26 +0000[diff] [blame]2229</div>
2230
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2231<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2232<h2><a name="constants">Constants</a></h2>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2233<!-- *********************************************************************** -->
2234
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2235<div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2236
2237<p>LLVM has several different basic types of constants. This section describes
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2238 them all and their syntax.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2239
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2240<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2241<h3>
2242 <a name="simpleconstants">Simple Constants</a>
2243</h3>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2244
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2245<div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2246
2247<dl>
2248 <dt><b>Boolean constants</b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2249 <dd>The two strings '<tt>true</tt>' and '<tt>false</tt>' are both valid
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]2250 constants of the <tt><a href="#t_integer">i1</a></tt> type.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2251
2252 <dt><b>Integer constants</b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2253 <dd>Standard integers (such as '4') are constants of
2254 the <a href="#t_integer">integer</a> type. Negative numbers may be used
2255 with integer types.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2256
2257 <dt><b>Floating point constants</b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2258 <dd>Floating point constants use standard decimal notation (e.g. 123.421),
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2259 exponential notation (e.g. 1.23421e+2), or a more precise hexadecimal
2260 notation (see below). The assembler requires the exact decimal value of a
2261 floating-point constant. For example, the assembler accepts 1.25 but
2262 rejects 1.3 because 1.3 is a repeating decimal in binary. Floating point
2263 constants must have a <a href="#t_floating">floating point</a> type. </dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2264
2265 <dt><b>Null pointer constants</b></dt>
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]2266 <dd>The identifier '<tt>null</tt>' is recognized as a null pointer constant
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2267 and must be of <a href="#t_pointer">pointer type</a>.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2268</dl>
2269
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2270<p>The one non-intuitive notation for constants is the hexadecimal form of
2271 floating point constants. For example, the form '<tt>double
2272 0x432ff973cafa8000</tt>' is equivalent to (but harder to read than)
2273 '<tt>double 4.5e+15</tt>'. The only time hexadecimal floating point
2274 constants are required (and the only time that they are generated by the
2275 disassembler) is when a floating point constant must be emitted but it cannot
2276 be represented as a decimal floating point number in a reasonable number of
2277 digits. For example, NaN's, infinities, and other special values are
2278 represented in their IEEE hexadecimal format so that assembly and disassembly
2279 do not cause any bits to change in the constants.</p>
2280
Dan Gohmance163392011-12-17 00:04:22 +0000[diff] [blame]2281<p>When using the hexadecimal form, constants of types half, float, and double are
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2282 represented using the 16-digit form shown above (which matches the IEEE754
Dan Gohmance163392011-12-17 00:04:22 +0000[diff] [blame]2283 representation for double); half and float values must, however, be exactly
2284 representable as IEE754 half and single precision, respectively.
2285 Hexadecimal format is always used
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2286 for long double, and there are three forms of long double. The 80-bit format
2287 used by x86 is represented as <tt>0xK</tt> followed by 20 hexadecimal digits.
2288 The 128-bit format used by PowerPC (two adjacent doubles) is represented
2289 by <tt>0xM</tt> followed by 32 hexadecimal digits. The IEEE 128-bit format
2290 is represented by <tt>0xL</tt> followed by 32 hexadecimal digits; no
2291 currently supported target uses this format. Long doubles will only work if
Tobias Grosser057beb82012-05-24 15:59:06 +0000[diff] [blame]2292 they match the long double format on your target. The IEEE 16-bit format
2293 (half precision) is represented by <tt>0xH</tt> followed by 4 hexadecimal
2294 digits. All hexadecimal formats are big-endian (sign bit at the left).</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2295
Dale Johannesen21fe99b2010-10-01 00:48:59 +0000[diff] [blame]2296<p>There are no constants of type x86mmx.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2297</div>
2298
2299<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2300<h3>
Bill Wendlingd9fe2982009-07-20 02:32:41 +0000[diff] [blame]2301<a name="aggregateconstants"></a> <!-- old anchor -->
2302<a name="complexconstants">Complex Constants</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2303</h3>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2304
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2305<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2306
Chris Lattner70882792009-02-28 18:32:25 +0000[diff] [blame]2307<p>Complex constants are a (potentially recursive) combination of simple
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2308 constants and smaller complex constants.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2309
2310<dl>
2311 <dt><b>Structure constants</b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2312 <dd>Structure constants are represented with notation similar to structure
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2313 type definitions (a comma separated list of elements, surrounded by braces
2314 (<tt>{}</tt>)). For example: "<tt>{ i32 4, float 17.0, i32* @G }</tt>",
2315 where "<tt>@G</tt>" is declared as "<tt>@G = external global i32</tt>".
2316 Structure constants must have <a href="#t_struct">structure type</a>, and
2317 the number and types of elements must match those specified by the
2318 type.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2319
2320 <dt><b>Array constants</b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2321 <dd>Array constants are represented with notation similar to array type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2322 definitions (a comma separated list of elements, surrounded by square
2323 brackets (<tt>[]</tt>)). For example: "<tt>[ i32 42, i32 11, i32 74
2324 ]</tt>". Array constants must have <a href="#t_array">array type</a>, and
2325 the number and types of elements must match those specified by the
2326 type.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2327
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]2328 <dt><b>Vector constants</b></dt>
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]2329 <dd>Vector constants are represented with notation similar to vector type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2330 definitions (a comma separated list of elements, surrounded by
2331 less-than/greater-than's (<tt>&lt;&gt;</tt>)). For example: "<tt>&lt; i32
2332 42, i32 11, i32 74, i32 100 &gt;</tt>". Vector constants must
2333 have <a href="#t_vector">vector type</a>, and the number and types of
2334 elements must match those specified by the type.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2335
2336 <dt><b>Zero initialization</b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2337 <dd>The string '<tt>zeroinitializer</tt>' can be used to zero initialize a
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]2338 value to zero of <em>any</em> type, including scalar and
2339 <a href="#t_aggregate">aggregate</a> types.
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2340 This is often used to avoid having to print large zero initializers
2341 (e.g. for large arrays) and is always exactly equivalent to using explicit
2342 zero initializers.</dd>
Nick Lewycky21cc4462009-04-04 07:22:01 +0000[diff] [blame]2343
2344 <dt><b>Metadata node</b></dt>
Nick Lewycky1e8c7a62009-05-30 16:08:30 +0000[diff] [blame]2345 <dd>A metadata node is a structure-like constant with
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2346 <a href="#t_metadata">metadata type</a>. For example: "<tt>metadata !{
2347 i32 0, metadata !"test" }</tt>". Unlike other constants that are meant to
2348 be interpreted as part of the instruction stream, metadata is a place to
2349 attach additional information such as debug info.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2350</dl>
2351
2352</div>
2353
2354<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2355<h3>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2356 <a name="globalconstants">Global Variable and Function Addresses</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2357</h3>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2358
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2359<div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2360
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2361<p>The addresses of <a href="#globalvars">global variables</a>
2362 and <a href="#functionstructure">functions</a> are always implicitly valid
2363 (link-time) constants. These constants are explicitly referenced when
2364 the <a href="#identifiers">identifier for the global</a> is used and always
2365 have <a href="#t_pointer">pointer</a> type. For example, the following is a
2366 legal LLVM file:</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2367
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2368<pre class="doc_code">
Chris Lattnera18a4242007-06-06 18:28:13 +0000[diff] [blame]2369@X = global i32 17
2370@Y = global i32 42
2371@Z = global [2 x i32*] [ i32* @X, i32* @Y ]
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2372</pre>
2373
2374</div>
2375
2376<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2377<h3>
2378 <a name="undefvalues">Undefined Values</a>
2379</h3>
2380
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2381<div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2382
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2383<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]2384 indicates that the user of the value may receive an unspecified bit-pattern.
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]2385 Undefined values may be of any type (other than '<tt>label</tt>'
2386 or '<tt>void</tt>') and be used anywhere a constant is permitted.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2387
Chris Lattnerc608cb12009-09-11 01:49:31 +0000[diff] [blame]2388<p>Undefined values are useful because they indicate to the compiler that the
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2389 program is well defined no matter what value is used. This gives the
2390 compiler more freedom to optimize. Here are some examples of (potentially
2391 surprising) transformations that are valid (in pseudo IR):</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2392
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2393
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2394<pre class="doc_code">
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2395 %A = add %X, undef
2396 %B = sub %X, undef
2397 %C = xor %X, undef
2398Safe:
2399 %A = undef
2400 %B = undef
2401 %C = undef
2402</pre>
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2403
2404<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]2405 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]2406
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2407<pre class="doc_code">
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2408 %A = or %X, undef
2409 %B = and %X, undef
2410Safe:
2411 %A = -1
2412 %B = 0
2413Unsafe:
2414 %A = undef
2415 %B = undef
2416</pre>
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2417
2418<p>These logical operations have bits that are not always affected by the input.
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]2419 For example, if <tt>%X</tt> has a zero bit, then the output of the
2420 '<tt>and</tt>' operation will always be a zero for that bit, no matter what
2421 the corresponding bit from the '<tt>undef</tt>' is. As such, it is unsafe to
2422 optimize or assume that the result of the '<tt>and</tt>' is '<tt>undef</tt>'.
2423 However, it is safe to assume that all bits of the '<tt>undef</tt>' could be
2424 0, and optimize the '<tt>and</tt>' to 0. Likewise, it is safe to assume that
2425 all the bits of the '<tt>undef</tt>' operand to the '<tt>or</tt>' could be
2426 set, allowing the '<tt>or</tt>' to be folded to -1.</p>
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2427
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2428<pre class="doc_code">
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2429 %A = select undef, %X, %Y
2430 %B = select undef, 42, %Y
2431 %C = select %X, %Y, undef
2432Safe:
2433 %A = %X (or %Y)
2434 %B = 42 (or %Y)
2435 %C = %Y
2436Unsafe:
2437 %A = undef
2438 %B = undef
2439 %C = undef
2440</pre>
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2441
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]2442<p>This set of examples shows that undefined '<tt>select</tt>' (and conditional
2443 branch) conditions can go <em>either way</em>, but they have to come from one
2444 of the two operands. In the <tt>%A</tt> example, if <tt>%X</tt> and
2445 <tt>%Y</tt> were both known to have a clear low bit, then <tt>%A</tt> would
2446 have to have a cleared low bit. However, in the <tt>%C</tt> example, the
2447 optimizer is allowed to assume that the '<tt>undef</tt>' operand could be the
2448 same as <tt>%Y</tt>, allowing the whole '<tt>select</tt>' to be
2449 eliminated.</p>
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2450
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2451<pre class="doc_code">
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2452 %A = xor undef, undef
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]2453
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2454 %B = undef
2455 %C = xor %B, %B
2456
2457 %D = undef
2458 %E = icmp lt %D, 4
2459 %F = icmp gte %D, 4
2460
2461Safe:
2462 %A = undef
2463 %B = undef
2464 %C = undef
2465 %D = undef
2466 %E = undef
2467 %F = undef
2468</pre>
Chris Lattner48a109c2009-09-07 22:52:39 +0000[diff] [blame]2469
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]2470<p>This example points out that two '<tt>undef</tt>' operands are not
2471 necessarily the same. This can be surprising to people (and also matches C
2472 semantics) where they assume that "<tt>X^X</tt>" is always zero, even
2473 if <tt>X</tt> is undefined. This isn't true for a number of reasons, but the
2474 short answer is that an '<tt>undef</tt>' "variable" can arbitrarily change
2475 its value over its "live range". This is true because the variable doesn't
2476 actually <em>have a live range</em>. Instead, the value is logically read
2477 from arbitrary registers that happen to be around when needed, so the value
2478 is not necessarily consistent over time. In fact, <tt>%A</tt> and <tt>%C</tt>
2479 need to have the same semantics or the core LLVM "replace all uses with"
2480 concept would not hold.</p>
Chris Lattner6e9057b2009-09-07 23:33:52 +0000[diff] [blame]2481
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2482<pre class="doc_code">
Chris Lattner6e9057b2009-09-07 23:33:52 +0000[diff] [blame]2483 %A = fdiv undef, %X
2484 %B = fdiv %X, undef
2485Safe:
2486 %A = undef
2487b: unreachable
2488</pre>
Chris Lattner6e9057b2009-09-07 23:33:52 +0000[diff] [blame]2489
2490<p>These examples show the crucial difference between an <em>undefined
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]2491 value</em> and <em>undefined behavior</em>. An undefined value (like
2492 '<tt>undef</tt>') is allowed to have an arbitrary bit-pattern. This means that
2493 the <tt>%A</tt> operation can be constant folded to '<tt>undef</tt>', because
2494 the '<tt>undef</tt>' could be an SNaN, and <tt>fdiv</tt> is not (currently)
2495 defined on SNaN's. However, in the second example, we can make a more
2496 aggressive assumption: because the <tt>undef</tt> is allowed to be an
2497 arbitrary value, we are allowed to assume that it could be zero. Since a
2498 divide by zero has <em>undefined behavior</em>, we are allowed to assume that
2499 the operation does not execute at all. This allows us to delete the divide and
2500 all code after it. Because the undefined operation "can't happen", the
2501 optimizer can assume that it occurs in dead code.</p>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]2502
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2503<pre class="doc_code">
Chris Lattner6e9057b2009-09-07 23:33:52 +0000[diff] [blame]2504a: store undef -> %X
2505b: store %X -> undef
2506Safe:
2507a: &lt;deleted&gt;
2508b: unreachable
2509</pre>
Chris Lattner6e9057b2009-09-07 23:33:52 +0000[diff] [blame]2510
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]2511<p>These examples reiterate the <tt>fdiv</tt> example: a store <em>of</em> an
2512 undefined value can be assumed to not have any effect; we can assume that the
2513 value is overwritten with bits that happen to match what was already there.
2514 However, a store <em>to</em> an undefined location could clobber arbitrary
2515 memory, therefore, it has undefined behavior.</p>
Chris Lattner6e9057b2009-09-07 23:33:52 +0000[diff] [blame]2516
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2517</div>
2518
2519<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2520<h3>
Dan Gohmanbfb056d2011-12-06 03:18:47 +0000[diff] [blame]2521 <a name="poisonvalues">Poison Values</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2522</h3>
2523
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2524<div>
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]2525
Dan Gohmanbfb056d2011-12-06 03:18:47 +0000[diff] [blame]2526<p>Poison values are similar to <a href="#undefvalues">undef values</a>, however
Dan Gohmane1a29842011-12-06 03:35:58 +0000[diff] [blame]2527 they also represent the fact that an instruction or constant expression which
2528 cannot evoke side effects has nevertheless detected a condition which results
2529 in undefined behavior.</p>
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]2530
Dan Gohmanbfb056d2011-12-06 03:18:47 +0000[diff] [blame]2531<p>There is currently no way of representing a poison value in the IR; they
Dan Gohman855abed2010-05-03 14:51:43 +0000[diff] [blame]2532 only exist when produced by operations such as
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2533 <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]2534
Dan Gohmanbfb056d2011-12-06 03:18:47 +0000[diff] [blame]2535<p>Poison value behavior is defined in terms of value <i>dependence</i>:</p>
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2536
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2537<ul>
2538<li>Values other than <a href="#i_phi"><tt>phi</tt></a> nodes depend on
2539 their operands.</li>
2540
2541<li><a href="#i_phi"><tt>Phi</tt></a> nodes depend on the operand corresponding
2542 to their dynamic predecessor basic block.</li>
2543
2544<li>Function arguments depend on the corresponding actual argument values in
2545 the dynamic callers of their functions.</li>
2546
2547<li><a href="#i_call"><tt>Call</tt></a> instructions depend on the
2548 <a href="#i_ret"><tt>ret</tt></a> instructions that dynamically transfer
2549 control back to them.</li>
2550
Dan Gohmanb5328162010-05-03 14:55:22 +0000[diff] [blame]2551<li><a href="#i_invoke"><tt>Invoke</tt></a> instructions depend on the
Bill Wendling7b9e5392012-02-06 21:57:33 +0000[diff] [blame]2552 <a href="#i_ret"><tt>ret</tt></a>, <a href="#i_resume"><tt>resume</tt></a>,
Dan Gohmanb5328162010-05-03 14:55:22 +0000[diff] [blame]2553 or exception-throwing call instructions that dynamically transfer control
2554 back to them.</li>
2555
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2556<li>Non-volatile loads and stores depend on the most recent stores to all of the
2557 referenced memory addresses, following the order in the IR
2558 (including loads and stores implied by intrinsics such as
2559 <a href="#int_memcpy"><tt>@llvm.memcpy</tt></a>.)</li>
2560
Dan Gohman7c24ff12010-05-03 14:59:34 +0000[diff] [blame]2561<!-- TODO: In the case of multiple threads, this only applies if the store
2562 "happens-before" the load or store. -->
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2563
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2564<!-- TODO: floating-point exception state -->
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2565
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2566<li>An instruction with externally visible side effects depends on the most
2567 recent preceding instruction with externally visible side effects, following
Dan Gohmanff70fe42010-07-06 15:26:33 +0000[diff] [blame]2568 the order in the IR. (This includes
2569 <a href="#volatile">volatile operations</a>.)</li>
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]2570
Dan Gohmanb5328162010-05-03 14:55:22 +0000[diff] [blame]2571<li>An instruction <i>control-depends</i> on a
2572 <a href="#terminators">terminator instruction</a>
2573 if the terminator instruction has multiple successors and the instruction
2574 is always executed when control transfers to one of the successors, and
Chris Lattner7a2bdde2011-04-15 05:18:47 +0000[diff] [blame]2575 may not be executed when control is transferred to another.</li>
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2576
Dan Gohmanca4cac42011-04-12 23:05:59 +0000[diff] [blame]2577<li>Additionally, an instruction also <i>control-depends</i> on a terminator
2578 instruction if the set of instructions it otherwise depends on would be
Chris Lattner7a2bdde2011-04-15 05:18:47 +0000[diff] [blame]2579 different if the terminator had transferred control to a different
Dan Gohmanca4cac42011-04-12 23:05:59 +0000[diff] [blame]2580 successor.</li>
2581
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2582<li>Dependence is transitive.</li>
2583
2584</ul>
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2585
Dan Gohmane1a29842011-12-06 03:35:58 +0000[diff] [blame]2586<p>Poison Values have the same behavior as <a href="#undefvalues">undef values</a>,
2587 with the additional affect that any instruction which has a <i>dependence</i>
2588 on a poison value has undefined behavior.</p>
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2589
2590<p>Here are some examples:</p>
Dan Gohmanc30f6e12010-04-26 20:54:53 +0000[diff] [blame]2591
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2592<pre class="doc_code">
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2593entry:
Dan Gohmanbfb056d2011-12-06 03:18:47 +0000[diff] [blame]2594 %poison = sub nuw i32 0, 1 ; Results in a poison value.
Dan Gohmane1a29842011-12-06 03:35:58 +0000[diff] [blame]2595 %still_poison = and i32 %poison, 0 ; 0, but also poison.
Dan Gohmanbfb056d2011-12-06 03:18:47 +0000[diff] [blame]2596 %poison_yet_again = getelementptr i32* @h, i32 %still_poison
Dan Gohmane1a29842011-12-06 03:35:58 +0000[diff] [blame]2597 store i32 0, i32* %poison_yet_again ; memory at @h[0] is poisoned
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2598
Dan Gohmane1a29842011-12-06 03:35:58 +0000[diff] [blame]2599 store i32 %poison, i32* @g ; Poison value stored to memory.
2600 %poison2 = load i32* @g ; Poison value loaded back from memory.
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2601
Dan Gohmanbfb056d2011-12-06 03:18:47 +0000[diff] [blame]2602 store volatile i32 %poison, i32* @g ; External observation; undefined behavior.
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2603
2604 %narrowaddr = bitcast i32* @g to i16*
2605 %wideaddr = bitcast i32* @g to i64*
Dan Gohmanbfb056d2011-12-06 03:18:47 +0000[diff] [blame]2606 %poison3 = load i16* %narrowaddr ; Returns a poison value.
2607 %poison4 = load i64* %wideaddr ; Returns a poison value.
Dan Gohman34b3d992010-04-28 00:49:41 +0000[diff] [blame]2608
Dan Gohman5cdc51e2011-12-06 03:31:14 +0000[diff] [blame]2609 %cmp = icmp slt i32 %poison, 0 ; Returns a poison value.
2610 br i1 %cmp, label %true, label %end ; Branch to either destination.
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2611
2612true:
Dan Gohman5cdc51e2011-12-06 03:31:14 +0000[diff] [blame]2613 store volatile i32 0, i32* @g ; This is control-dependent on %cmp, so
2614 ; it has undefined behavior.
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2615 br label %end
2616
2617end:
2618 %p = phi i32 [ 0, %entry ], [ 1, %true ]
Dan Gohman5cdc51e2011-12-06 03:31:14 +0000[diff] [blame]2619 ; Both edges into this PHI are
2620 ; control-dependent on %cmp, so this
2621 ; always results in a poison value.
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2622
Dan Gohman5cdc51e2011-12-06 03:31:14 +0000[diff] [blame]2623 store volatile i32 0, i32* @g ; This would depend on the store in %true
2624 ; if %cmp is true, or the store in %entry
2625 ; otherwise, so this is undefined behavior.
Dan Gohmanca4cac42011-04-12 23:05:59 +0000[diff] [blame]2626
Nick Lewycky64f9fb12011-05-16 19:29:30 +0000[diff] [blame]2627 br i1 %cmp, label %second_true, label %second_end
Dan Gohman5cdc51e2011-12-06 03:31:14 +0000[diff] [blame]2628 ; The same branch again, but this time the
2629 ; true block doesn't have side effects.
Dan Gohmanca4cac42011-04-12 23:05:59 +0000[diff] [blame]2630
2631second_true:
2632 ; No side effects!
Nick Lewycky64f9fb12011-05-16 19:29:30 +0000[diff] [blame]2633 ret void
Dan Gohmanca4cac42011-04-12 23:05:59 +0000[diff] [blame]2634
2635second_end:
Dan Gohman5cdc51e2011-12-06 03:31:14 +0000[diff] [blame]2636 store volatile i32 0, i32* @g ; This time, the instruction always depends
2637 ; on the store in %end. Also, it is
2638 ; control-equivalent to %end, so this is
Dan Gohmane1a29842011-12-06 03:35:58 +0000[diff] [blame]2639 ; well-defined (ignoring earlier undefined
2640 ; behavior in this example).
Dan Gohmanae11c3f2010-04-26 23:36:52 +0000[diff] [blame]2641</pre>
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]2642
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]2643</div>
2644
2645<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2646<h3>
2647 <a name="blockaddress">Addresses of Basic Blocks</a>
2648</h3>
2649
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2650<div>
Chris Lattnerc6f44362009-10-27 21:01:34 +0000[diff] [blame]2651
Chris Lattnercdfc9402009-11-01 01:27:45 +0000[diff] [blame]2652<p><b><tt>blockaddress(@function, %block)</tt></b></p>
Chris Lattnerc6f44362009-10-27 21:01:34 +0000[diff] [blame]2653
2654<p>The '<tt>blockaddress</tt>' constant computes the address of the specified
Chris Lattner2dfdf2a2009-10-27 21:49:40 +0000[diff] [blame]2655 basic block in the specified function, and always has an i8* type. Taking
Chris Lattnercdfc9402009-11-01 01:27:45 +0000[diff] [blame]2656 the address of the entry block is illegal.</p>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]2657
Chris Lattnerc6f44362009-10-27 21:01:34 +0000[diff] [blame]2658<p>This value only has defined behavior when used as an operand to the
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]2659 '<a href="#i_indirectbr"><tt>indirectbr</tt></a>' instruction, or for
2660 comparisons against null. Pointer equality tests between labels addresses
2661 results in undefined behavior &mdash; though, again, comparison against null
2662 is ok, and no label is equal to the null pointer. This may be passed around
2663 as an opaque pointer sized value as long as the bits are not inspected. This
2664 allows <tt>ptrtoint</tt> and arithmetic to be performed on these values so
2665 long as the original value is reconstituted before the <tt>indirectbr</tt>
2666 instruction.</p>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]2667
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]2668<p>Finally, some targets may provide defined semantics when using the value as
2669 the operand to an inline assembly, but that is target specific.</p>
Chris Lattnerc6f44362009-10-27 21:01:34 +0000[diff] [blame]2670
2671</div>
2672
2673
2674<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2675<h3>
2676 <a name="constantexprs">Constant Expressions</a>
2677</h3>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2678
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2679<div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2680
2681<p>Constant expressions are used to allow expressions involving other constants
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2682 to be used as constants. Constant expressions may be of
2683 any <a href="#t_firstclass">first class</a> type and may involve any LLVM
2684 operation that does not have side effects (e.g. load and call are not
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]2685 supported). The following is the syntax for constant expressions:</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2686
2687<dl>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2688 <dt><b><tt>trunc (CST to TYPE)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2689 <dd>Truncate a constant to another type. The bit size of CST must be larger
2690 than the bit size of TYPE. Both types must be integers.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2691
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2692 <dt><b><tt>zext (CST to TYPE)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2693 <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]2694 smaller than the bit size of TYPE. Both types must be integers.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2695
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2696 <dt><b><tt>sext (CST to TYPE)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2697 <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]2698 smaller than the bit size of TYPE. Both types must be integers.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2699
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2700 <dt><b><tt>fptrunc (CST to TYPE)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2701 <dd>Truncate a floating point constant to another floating point type. The
2702 size of CST must be larger than the size of TYPE. Both types must be
2703 floating point.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2704
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2705 <dt><b><tt>fpext (CST to TYPE)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2706 <dd>Floating point extend a constant to another type. The size of CST must be
2707 smaller or equal to the size of TYPE. Both types must be floating
2708 point.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2709
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2710 <dt><b><tt>fptoui (CST to TYPE)</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2711 <dd>Convert a floating point constant to the corresponding unsigned integer
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2712 constant. TYPE must be a scalar or vector integer type. CST must be of
2713 scalar or vector floating point type. Both CST and TYPE must be scalars,
2714 or vectors of the same number of elements. If the value won't fit in the
2715 integer type, the results are undefined.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2716
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2717 <dt><b><tt>fptosi (CST to TYPE)</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2718 <dd>Convert a floating point constant to the corresponding signed integer
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2719 constant. TYPE must be a scalar or vector integer type. CST must be of
2720 scalar or vector floating point type. Both CST and TYPE must be scalars,
2721 or vectors of the same number of elements. If the value won't fit in the
2722 integer type, the results are undefined.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2723
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2724 <dt><b><tt>uitofp (CST to TYPE)</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2725 <dd>Convert an unsigned integer constant to the corresponding floating point
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2726 constant. TYPE must be a scalar or vector floating point type. CST must be
2727 of scalar or vector integer type. Both CST and TYPE must be scalars, or
2728 vectors of the same number of elements. If the value won't fit in the
2729 floating point type, the results are undefined.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2730
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2731 <dt><b><tt>sitofp (CST to TYPE)</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2732 <dd>Convert a signed integer constant to the corresponding floating point
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2733 constant. TYPE must be a scalar or vector floating point type. CST must be
2734 of scalar or vector integer type. Both CST and TYPE must be scalars, or
2735 vectors of the same number of elements. If the value won't fit in the
2736 floating point type, the results are undefined.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2737
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2738 <dt><b><tt>ptrtoint (CST to TYPE)</tt></b></dt>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]2739 <dd>Convert a pointer typed constant to the corresponding integer constant
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2740 <tt>TYPE</tt> must be an integer type. <tt>CST</tt> must be of pointer
2741 type. The <tt>CST</tt> value is zero extended, truncated, or unchanged to
2742 make it fit in <tt>TYPE</tt>.</dd>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]2743
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2744 <dt><b><tt>inttoptr (CST to TYPE)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2745 <dd>Convert a integer constant to a pointer constant. TYPE must be a pointer
2746 type. CST must be of integer type. The CST value is zero extended,
2747 truncated, or unchanged to make it fit in a pointer size. This one is
2748 <i>really</i> dangerous!</dd>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]2749
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2750 <dt><b><tt>bitcast (CST to TYPE)</tt></b></dt>
Chris Lattner03bbad62009-02-28 18:27:03 +0000[diff] [blame]2751 <dd>Convert a constant, CST, to another TYPE. The constraints of the operands
2752 are the same as those for the <a href="#i_bitcast">bitcast
2753 instruction</a>.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2754
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2755 <dt><b><tt>getelementptr (CSTPTR, IDX0, IDX1, ...)</tt></b></dt>
2756 <dt><b><tt>getelementptr inbounds (CSTPTR, IDX0, IDX1, ...)</tt></b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2757 <dd>Perform the <a href="#i_getelementptr">getelementptr operation</a> on
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2758 constants. As with the <a href="#i_getelementptr">getelementptr</a>
2759 instruction, the index list may have zero or more indexes, which are
2760 required to make sense for the type of "CSTPTR".</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2761
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2762 <dt><b><tt>select (COND, VAL1, VAL2)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2763 <dd>Perform the <a href="#i_select">select operation</a> on constants.</dd>
Reid Spencer01c42592006-12-04 19:23:19 +0000[diff] [blame]2764
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2765 <dt><b><tt>icmp COND (VAL1, VAL2)</tt></b></dt>
Reid Spencer01c42592006-12-04 19:23:19 +0000[diff] [blame]2766 <dd>Performs the <a href="#i_icmp">icmp operation</a> on constants.</dd>
2767
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2768 <dt><b><tt>fcmp COND (VAL1, VAL2)</tt></b></dt>
Reid Spencer01c42592006-12-04 19:23:19 +0000[diff] [blame]2769 <dd>Performs the <a href="#i_fcmp">fcmp operation</a> on constants.</dd>
Robert Bocchino9fbe1452006-01-10 19:31:34 +0000[diff] [blame]2770
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2771 <dt><b><tt>extractelement (VAL, IDX)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2772 <dd>Perform the <a href="#i_extractelement">extractelement operation</a> on
2773 constants.</dd>
Robert Bocchino9fbe1452006-01-10 19:31:34 +0000[diff] [blame]2774
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2775 <dt><b><tt>insertelement (VAL, ELT, IDX)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2776 <dd>Perform the <a href="#i_insertelement">insertelement operation</a> on
2777 constants.</dd>
Chris Lattnerc1989542006-04-08 00:13:41 +0000[diff] [blame]2778
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2779 <dt><b><tt>shufflevector (VEC1, VEC2, IDXMASK)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2780 <dd>Perform the <a href="#i_shufflevector">shufflevector operation</a> on
2781 constants.</dd>
Chris Lattnerc1989542006-04-08 00:13:41 +0000[diff] [blame]2782
Nick Lewycky9e130ce2010-05-29 06:44:15 +0000[diff] [blame]2783 <dt><b><tt>extractvalue (VAL, IDX0, IDX1, ...)</tt></b></dt>
2784 <dd>Perform the <a href="#i_extractvalue">extractvalue operation</a> on
2785 constants. The index list is interpreted in a similar manner as indices in
2786 a '<a href="#i_getelementptr">getelementptr</a>' operation. At least one
2787 index value must be specified.</dd>
2788
2789 <dt><b><tt>insertvalue (VAL, ELT, IDX0, IDX1, ...)</tt></b></dt>
2790 <dd>Perform the <a href="#i_insertvalue">insertvalue operation</a> on
2791 constants. The index list is interpreted in a similar manner as indices in
2792 a '<a href="#i_getelementptr">getelementptr</a>' operation. At least one
2793 index value must be specified.</dd>
2794
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]2795 <dt><b><tt>OPCODE (LHS, RHS)</tt></b></dt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2796 <dd>Perform the specified operation of the LHS and RHS constants. OPCODE may
2797 be any of the <a href="#binaryops">binary</a>
2798 or <a href="#bitwiseops">bitwise binary</a> operations. The constraints
2799 on operands are the same as those for the corresponding instruction
2800 (e.g. no bitwise operations on floating point values are allowed).</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2801</dl>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2802
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2803</div>
Chris Lattner9ee5d222004-03-08 16:49:10 +0000[diff] [blame]2804
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2805</div>
2806
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2807<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2808<h2><a name="othervalues">Other Values</a></h2>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2809<!-- *********************************************************************** -->
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2810<div>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2811<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2812<h3>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2813<a name="inlineasm">Inline Assembler Expressions</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2814</h3>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2815
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2816<div>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2817
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2818<p>LLVM supports inline assembler expressions (as opposed
Bill Wendlingaee0f452011-11-30 21:52:43 +0000[diff] [blame]2819 to <a href="#moduleasm">Module-Level Inline Assembly</a>) through the use of
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2820 a special value. This value represents the inline assembler as a string
2821 (containing the instructions to emit), a list of operand constraints (stored
Dale Johannesen09fed252009-10-13 21:56:55 +0000[diff] [blame]2822 as a string), a flag that indicates whether or not the inline asm
Dale Johannesen8ba2d5b2009-10-21 23:28:00 +0000[diff] [blame]2823 expression has side effects, and a flag indicating whether the function
2824 containing the asm needs to align its stack conservatively. An example
2825 inline assembler expression is:</p>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2826
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2827<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2828i32 (i32) asm "bswap $0", "=r,r"
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2829</pre>
2830
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2831<p>Inline assembler expressions may <b>only</b> be used as the callee operand of
2832 a <a href="#i_call"><tt>call</tt> instruction</a>. Thus, typically we
2833 have:</p>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2834
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2835<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2836%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]2837</pre>
2838
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2839<p>Inline asms with side effects not visible in the constraint list must be
2840 marked as having side effects. This is done through the use of the
2841 '<tt>sideeffect</tt>' keyword, like so:</p>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2842
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2843<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2844call void asm sideeffect "eieio", ""()
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2845</pre>
2846
Dale Johannesen8ba2d5b2009-10-21 23:28:00 +0000[diff] [blame]2847<p>In some cases inline asms will contain code that will not work unless the
2848 stack is aligned in some way, such as calls or SSE instructions on x86,
2849 yet will not contain code that does that alignment within the asm.
2850 The compiler should make conservative assumptions about what the asm might
2851 contain and should generate its usual stack alignment code in the prologue
2852 if the '<tt>alignstack</tt>' keyword is present:</p>
Dale Johannesen09fed252009-10-13 21:56:55 +0000[diff] [blame]2853
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2854<pre class="doc_code">
Dale Johannesen8ba2d5b2009-10-21 23:28:00 +0000[diff] [blame]2855call void asm alignstack "eieio", ""()
Dale Johannesen09fed252009-10-13 21:56:55 +0000[diff] [blame]2856</pre>
Dale Johannesen09fed252009-10-13 21:56:55 +0000[diff] [blame]2857
2858<p>If both keywords appear the '<tt>sideeffect</tt>' keyword must come
2859 first.</p>
2860
Bill Wendlingaee0f452011-11-30 21:52:43 +0000[diff] [blame]2861<!--
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2862<p>TODO: The format of the asm and constraints string still need to be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]2863 documented here. Constraints on what can be done (e.g. duplication, moving,
2864 etc need to be documented). This is probably best done by reference to
2865 another document that covers inline asm from a holistic perspective.</p>
Bill Wendlingaee0f452011-11-30 21:52:43 +0000[diff] [blame]2866 -->
Chris Lattnercf9a4152010-04-07 05:38:05 +0000[diff] [blame]2867
Bill Wendlingaee0f452011-11-30 21:52:43 +0000[diff] [blame]2868<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2869<h4>
Bill Wendlingaee0f452011-11-30 21:52:43 +0000[diff] [blame]2870 <a name="inlineasm_md">Inline Asm Metadata</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2871</h4>
Chris Lattnercf9a4152010-04-07 05:38:05 +0000[diff] [blame]2872
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2873<div>
Chris Lattnercf9a4152010-04-07 05:38:05 +0000[diff] [blame]2874
Bill Wendlingaee0f452011-11-30 21:52:43 +0000[diff] [blame]2875<p>The call instructions that wrap inline asm nodes may have a
2876 "<tt>!srcloc</tt>" MDNode attached to it that contains a list of constant
2877 integers. If present, the code generator will use the integer as the
2878 location cookie value when report errors through the <tt>LLVMContext</tt>
2879 error reporting mechanisms. This allows a front-end to correlate backend
2880 errors that occur with inline asm back to the source code that produced it.
2881 For example:</p>
Chris Lattnercf9a4152010-04-07 05:38:05 +0000[diff] [blame]2882
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]2883<pre class="doc_code">
Chris Lattnercf9a4152010-04-07 05:38:05 +0000[diff] [blame]2884call void asm sideeffect "something bad", ""()<b>, !srcloc !42</b>
2885...
2886!42 = !{ i32 1234567 }
2887</pre>
Chris Lattnercf9a4152010-04-07 05:38:05 +0000[diff] [blame]2888
2889<p>It is up to the front-end to make sense of the magic numbers it places in the
Bill Wendlingaee0f452011-11-30 21:52:43 +0000[diff] [blame]2890 IR. If the MDNode contains multiple constants, the code generator will use
Chris Lattnerce1b9ad2010-11-17 08:20:42 +0000[diff] [blame]2891 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]2892
2893</div>
2894
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2895</div>
2896
Chris Lattnere6a5ddd2010-01-15 21:50:19 +0000[diff] [blame]2897<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]2898<h3>
2899 <a name="metadata">Metadata Nodes and Metadata Strings</a>
2900</h3>
Chris Lattnere6a5ddd2010-01-15 21:50:19 +0000[diff] [blame]2901
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]2902<div>
Chris Lattnere6a5ddd2010-01-15 21:50:19 +0000[diff] [blame]2903
2904<p>LLVM IR allows metadata to be attached to instructions in the program that
2905 can convey extra information about the code to the optimizers and code
2906 generator. One example application of metadata is source-level debug
2907 information. There are two metadata primitives: strings and nodes. All
2908 metadata has the <tt>metadata</tt> type and is identified in syntax by a
2909 preceding exclamation point ('<tt>!</tt>').</p>
2910
2911<p>A metadata string is a string surrounded by double quotes. It can contain
Bill Wendlingf6cc4c22011-11-30 21:43:43 +0000[diff] [blame]2912 any character by escaping non-printable characters with "<tt>\xx</tt>" where
2913 "<tt>xx</tt>" is the two digit hex code. For example:
2914 "<tt>!"test\00"</tt>".</p>
Chris Lattnere6a5ddd2010-01-15 21:50:19 +0000[diff] [blame]2915
2916<p>Metadata nodes are represented with notation similar to structure constants
2917 (a comma separated list of elements, surrounded by braces and preceded by an
Bill Wendlingf6cc4c22011-11-30 21:43:43 +0000[diff] [blame]2918 exclamation point). Metadata nodes can have any values as their operand. For
2919 example:</p>
2920
2921<div class="doc_code">
2922<pre>
2923!{ metadata !"test\00", i32 10}
2924</pre>
2925</div>
Chris Lattnere6a5ddd2010-01-15 21:50:19 +0000[diff] [blame]2926
2927<p>A <a href="#namedmetadatastructure">named metadata</a> is a collection of
2928 metadata nodes, which can be looked up in the module symbol table. For
Bill Wendlingf6cc4c22011-11-30 21:43:43 +0000[diff] [blame]2929 example:</p>
2930
2931<div class="doc_code">
2932<pre>
2933!foo = metadata !{!4, !3}
2934</pre>
2935</div>
Chris Lattnere6a5ddd2010-01-15 21:50:19 +0000[diff] [blame]2936
Devang Patele1d50cd2010-03-04 23:44:48 +0000[diff] [blame]2937<p>Metadata can be used as function arguments. Here <tt>llvm.dbg.value</tt>
Bill Wendlingf6cc4c22011-11-30 21:43:43 +0000[diff] [blame]2938 function is using two metadata arguments:</p>
Devang Patele1d50cd2010-03-04 23:44:48 +0000[diff] [blame]2939
Bill Wendling9ff5de92011-03-02 02:17:11 +0000[diff] [blame]2940<div class="doc_code">
2941<pre>
2942call void @llvm.dbg.value(metadata !24, i64 0, metadata !25)
2943</pre>
2944</div>
Devang Patele1d50cd2010-03-04 23:44:48 +0000[diff] [blame]2945
2946<p>Metadata can be attached with an instruction. Here metadata <tt>!21</tt> is
Bill Wendlingf6cc4c22011-11-30 21:43:43 +0000[diff] [blame]2947 attached to the <tt>add</tt> instruction using the <tt>!dbg</tt>
2948 identifier:</p>
Devang Patele1d50cd2010-03-04 23:44:48 +0000[diff] [blame]2949
Bill Wendling9ff5de92011-03-02 02:17:11 +0000[diff] [blame]2950<div class="doc_code">
2951<pre>
2952%indvar.next = add i64 %indvar, 1, !dbg !21
2953</pre>
2954</div>
2955
Peter Collingbourne249d9532011-10-27 19:19:07 +0000[diff] [blame]2956<p>More information about specific metadata nodes recognized by the optimizers
2957 and code generator is found below.</p>
2958
Bill Wendlingf6cc4c22011-11-30 21:43:43 +0000[diff] [blame]2959<!-- _______________________________________________________________________ -->
Peter Collingbourne249d9532011-10-27 19:19:07 +0000[diff] [blame]2960<h4>
2961 <a name="tbaa">'<tt>tbaa</tt>' Metadata</a>
2962</h4>
2963
2964<div>
2965
2966<p>In LLVM IR, memory does not have types, so LLVM's own type system is not
2967 suitable for doing TBAA. Instead, metadata is added to the IR to describe
2968 a type system of a higher level language. This can be used to implement
2969 typical C/C++ TBAA, but it can also be used to implement custom alias
2970 analysis behavior for other languages.</p>
2971
2972<p>The current metadata format is very simple. TBAA metadata nodes have up to
2973 three fields, e.g.:</p>
2974
2975<div class="doc_code">
2976<pre>
2977!0 = metadata !{ metadata !"an example type tree" }
2978!1 = metadata !{ metadata !"int", metadata !0 }
2979!2 = metadata !{ metadata !"float", metadata !0 }
2980!3 = metadata !{ metadata !"const float", metadata !2, i64 1 }
2981</pre>
2982</div>
2983
2984<p>The first field is an identity field. It can be any value, usually
2985 a metadata string, which uniquely identifies the type. The most important
2986 name in the tree is the name of the root node. Two trees with
2987 different root node names are entirely disjoint, even if they
2988 have leaves with common names.</p>
2989
2990<p>The second field identifies the type's parent node in the tree, or
2991 is null or omitted for a root node. A type is considered to alias
2992 all of its descendants and all of its ancestors in the tree. Also,
2993 a type is considered to alias all types in other trees, so that
2994 bitcode produced from multiple front-ends is handled conservatively.</p>
2995
2996<p>If the third field is present, it's an integer which if equal to 1
2997 indicates that the type is "constant" (meaning
2998 <tt>pointsToConstantMemory</tt> should return true; see
2999 <a href="AliasAnalysis.html#OtherItfs">other useful
3000 <tt>AliasAnalysis</tt> methods</a>).</p>
3001
3002</div>
3003
Bill Wendlingf6cc4c22011-11-30 21:43:43 +0000[diff] [blame]3004<!-- _______________________________________________________________________ -->
Peter Collingbourne999f90b2011-10-27 19:19:14 +0000[diff] [blame]3005<h4>
Duncan Sands5e5c5f82012-04-14 12:36:06 +0000[diff] [blame]3006 <a name="fpmath">'<tt>fpmath</tt>' Metadata</a>
Peter Collingbourne999f90b2011-10-27 19:19:14 +0000[diff] [blame]3007</h4>
3008
3009<div>
3010
Duncan Sands5e5c5f82012-04-14 12:36:06 +0000[diff] [blame]3011<p><tt>fpmath</tt> metadata may be attached to any instruction of floating point
Duncan Sands8883c432012-04-16 16:28:59 +0000[diff] [blame]3012 type. It can be used to express the maximum acceptable error in the result of
3013 that instruction, in ULPs, thus potentially allowing the compiler to use a
Duncan Sands2867c852012-04-16 19:39:33 +0000[diff] [blame]3014 more efficient but less accurate method of computing it. ULP is defined as
3015 follows:</p>
Peter Collingbourne999f90b2011-10-27 19:19:14 +0000[diff] [blame]3016
Bill Wendling0656e252011-11-09 19:33:56 +0000[diff] [blame]3017<blockquote>
3018
3019<p>If <tt>x</tt> is a real number that lies between two finite consecutive
3020 floating-point numbers <tt>a</tt> and <tt>b</tt>, without being equal to one
3021 of them, then <tt>ulp(x) = |b - a|</tt>, otherwise <tt>ulp(x)</tt> is the
3022 distance between the two non-equal finite floating-point numbers nearest
3023 <tt>x</tt>. Moreover, <tt>ulp(NaN)</tt> is <tt>NaN</tt>.</p>
3024
3025</blockquote>
Peter Collingbourne999f90b2011-10-27 19:19:14 +0000[diff] [blame]3026
Duncan Sands8883c432012-04-16 16:28:59 +0000[diff] [blame]3027<p>The metadata node shall consist of a single positive floating point number
Duncan Sands2867c852012-04-16 19:39:33 +0000[diff] [blame]3028 representing the maximum relative error, for example:</p>
Peter Collingbourne999f90b2011-10-27 19:19:14 +0000[diff] [blame]3029
3030<div class="doc_code">
3031<pre>
Duncan Sands8883c432012-04-16 16:28:59 +0000[diff] [blame]3032!0 = metadata !{ float 2.5 } ; maximum acceptable inaccuracy is 2.5 ULPs
Peter Collingbourne999f90b2011-10-27 19:19:14 +0000[diff] [blame]3033</pre>
3034</div>
3035
NAKAMURA Takumi9c55f592012-03-27 11:25:16 +0000[diff] [blame]3036</div>
3037
Rafael Espindola39dd3282012-03-24 00:14:51 +0000[diff] [blame]3038<!-- _______________________________________________________________________ -->
3039<h4>
3040 <a name="range">'<tt>range</tt>' Metadata</a>
3041</h4>
3042
3043<div>
3044<p><tt>range</tt> metadata may be attached only to loads of integer types. It
3045 expresses the possible ranges the loaded value is in. The ranges are
3046 represented with a flattened list of integers. The loaded value is known to
3047 be in the union of the ranges defined by each consecutive pair. Each pair
3048 has the following properties:</p>
3049<ul>
3050 <li>The type must match the type loaded by the instruction.</li>
3051 <li>The pair <tt>a,b</tt> represents the range <tt>[a,b)</tt>.</li>
3052 <li>Both <tt>a</tt> and <tt>b</tt> are constants.</li>
3053 <li>The range is allowed to wrap.</li>
3054 <li>The range should not represent the full or empty set. That is,
3055 <tt>a!=b</tt>. </li>
3056</ul>
3057
3058<p>Examples:</p>
3059<div class="doc_code">
3060<pre>
3061 %a = load i8* %x, align 1, !range !0 ; Can only be 0 or 1
3062 %b = load i8* %y, align 1, !range !1 ; Can only be 255 (-1), 0 or 1
3063 %c = load i8* %z, align 1, !range !2 ; Can only be 0, 1, 3, 4 or 5
3064...
3065!0 = metadata !{ i8 0, i8 2 }
3066!1 = metadata !{ i8 255, i8 2 }
3067!2 = metadata !{ i8 0, i8 2, i8 3, i8 6 }
3068</pre>
3069</div>
3070</div>
Peter Collingbourne999f90b2011-10-27 19:19:14 +0000[diff] [blame]3071</div>
3072
Chris Lattnere6a5ddd2010-01-15 21:50:19 +0000[diff] [blame]3073</div>
3074
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]3075<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3076<h2>
Bill Wendlingb9d75a92012-02-11 11:59:36 +0000[diff] [blame]3077 <a name="module_flags">Module Flags Metadata</a>
3078</h2>
3079<!-- *********************************************************************** -->
3080
3081<div>
3082
3083<p>Information about the module as a whole is difficult to convey to LLVM's
3084 subsystems. The LLVM IR isn't sufficient to transmit this
3085 information. The <tt>llvm.module.flags</tt> named metadata exists in order to
3086 facilitate this. These flags are in the form of key / value pairs &mdash;
3087 much like a dictionary &mdash; making it easy for any subsystem who cares
3088 about a flag to look it up.</p>
3089
3090<p>The <tt>llvm.module.flags</tt> metadata contains a list of metadata
3091 triplets. Each triplet has the following form:</p>
3092
3093<ul>
3094 <li>The first element is a <i>behavior</i> flag, which specifies the behavior
3095 when two (or more) modules are merged together, and it encounters two (or
3096 more) metadata with the same ID. The supported behaviors are described
3097 below.</li>
3098
3099 <li>The second element is a metadata string that is a unique ID for the
3100 metadata. How each ID is interpreted is documented below.</li>
3101
3102 <li>The third element is the value of the flag.</li>
3103</ul>
3104
3105<p>When two (or more) modules are merged together, the resulting
3106 <tt>llvm.module.flags</tt> metadata is the union of the
3107 modules' <tt>llvm.module.flags</tt> metadata. The only exception being a flag
3108 with the <i>Override</i> behavior, which may override another flag's value
3109 (see below).</p>
3110
3111<p>The following behaviors are supported:</p>
3112
3113<table border="1" cellspacing="0" cellpadding="4">
3114 <tbody>
3115 <tr>
3116 <th>Value</th>
3117 <th>Behavior</th>
3118 </tr>
3119 <tr>
3120 <td>1</td>
3121 <td align="left">
Bill Wendlinga0edecf2012-03-06 09:17:04 +0000[diff] [blame]3122 <dl>
3123 <dt><b>Error</b></dt>
3124 <dd>Emits an error if two values disagree. It is an error to have an ID
3125 with both an Error and a Warning behavior.</dd>
3126 </dl>
Bill Wendlingb9d75a92012-02-11 11:59:36 +0000[diff] [blame]3127 </td>
3128 </tr>
3129 <tr>
3130 <td>2</td>
3131 <td align="left">
Bill Wendlinga0edecf2012-03-06 09:17:04 +0000[diff] [blame]3132 <dl>
3133 <dt><b>Warning</b></dt>
3134 <dd>Emits a warning if two values disagree.</dd>
3135 </dl>
Bill Wendlingb9d75a92012-02-11 11:59:36 +0000[diff] [blame]3136 </td>
3137 </tr>
3138 <tr>
3139 <td>3</td>
3140 <td align="left">
Bill Wendlinga0edecf2012-03-06 09:17:04 +0000[diff] [blame]3141 <dl>
3142 <dt><b>Require</b></dt>
3143 <dd>Emits an error when the specified value is not present or doesn't
3144 have the specified value. It is an error for two (or more)
3145 <tt>llvm.module.flags</tt> with the same ID to have the Require
3146 behavior but different values. There may be multiple Require flags
3147 per ID.</dd>
3148 </dl>
Bill Wendlingb9d75a92012-02-11 11:59:36 +0000[diff] [blame]3149 </td>
3150 </tr>
3151 <tr>
3152 <td>4</td>
3153 <td align="left">
Bill Wendlinga0edecf2012-03-06 09:17:04 +0000[diff] [blame]3154 <dl>
3155 <dt><b>Override</b></dt>
3156 <dd>Uses the specified value if the two values disagree. It is an
3157 error for two (or more) <tt>llvm.module.flags</tt> with the same
3158 ID to have the Override behavior but different values.</dd>
3159 </dl>
Bill Wendlingb9d75a92012-02-11 11:59:36 +0000[diff] [blame]3160 </td>
3161 </tr>
3162 </tbody>
3163</table>
3164
3165<p>An example of module flags:</p>
3166
3167<pre class="doc_code">
3168!0 = metadata !{ i32 1, metadata !"foo", i32 1 }
3169!1 = metadata !{ i32 4, metadata !"bar", i32 37 }
3170!2 = metadata !{ i32 2, metadata !"qux", i32 42 }
3171!3 = metadata !{ i32 3, metadata !"qux",
3172 metadata !{
3173 metadata !"foo", i32 1
3174 }
3175}
3176!llvm.module.flags = !{ !0, !1, !2, !3 }
3177</pre>
3178
3179<ul>
3180 <li><p>Metadata <tt>!0</tt> has the ID <tt>!"foo"</tt> and the value '1'. The
3181 behavior if two or more <tt>!"foo"</tt> flags are seen is to emit an
3182 error if their values are not equal.</p></li>
3183
3184 <li><p>Metadata <tt>!1</tt> has the ID <tt>!"bar"</tt> and the value '37'. The
3185 behavior if two or more <tt>!"bar"</tt> flags are seen is to use the
3186 value '37' if their values are not equal.</p></li>
3187
3188 <li><p>Metadata <tt>!2</tt> has the ID <tt>!"qux"</tt> and the value '42'. The
3189 behavior if two or more <tt>!"qux"</tt> flags are seen is to emit a
3190 warning if their values are not equal.</p></li>
3191
3192 <li><p>Metadata <tt>!3</tt> has the ID <tt>!"qux"</tt> and the value:</p>
3193
3194<pre class="doc_code">
3195metadata !{ metadata !"foo", i32 1 }
3196</pre>
Bill Wendlingf7b367c2012-02-16 01:10:50 +0000[diff] [blame]3197
Bill Wendlingb9d75a92012-02-11 11:59:36 +0000[diff] [blame]3198 <p>The behavior is to emit an error if the <tt>llvm.module.flags</tt> does
3199 not contain a flag with the ID <tt>!"foo"</tt> that has the value
3200 '1'. If two or more <tt>!"qux"</tt> flags exist, then they must have
3201 the same value or an error will be issued.</p></li>
3202</ul>
3203
Bill Wendlingf7b367c2012-02-16 01:10:50 +0000[diff] [blame]3204
3205<!-- ======================================================================= -->
3206<h3>
3207<a name="objc_gc_flags">Objective-C Garbage Collection Module Flags Metadata</a>
3208</h3>
3209
3210<div>
3211
3212<p>On the Mach-O platform, Objective-C stores metadata about garbage collection
3213 in a special section called "image info". The metadata consists of a version
3214 number and a bitmask specifying what types of garbage collection are
3215 supported (if any) by the file. If two or more modules are linked together
3216 their garbage collection metadata needs to be merged rather than appended
3217 together.</p>
3218
3219<p>The Objective-C garbage collection module flags metadata consists of the
3220 following key-value pairs:</p>
3221
3222<table border="1" cellspacing="0" cellpadding="4">
Bill Wendlingb3ef2232012-03-06 09:23:25 +0000[diff] [blame]3223 <col width="30%">
Bill Wendlingf7b367c2012-02-16 01:10:50 +0000[diff] [blame]3224 <tbody>
3225 <tr>
Bill Wendlinga0edecf2012-03-06 09:17:04 +0000[diff] [blame]3226 <th>Key</th>
Bill Wendlingf7b367c2012-02-16 01:10:50 +0000[diff] [blame]3227 <th>Value</th>
3228 </tr>
3229 <tr>
3230 <td><tt>Objective-C&nbsp;Version</tt></td>
3231 <td align="left"><b>[Required]</b> &mdash; The Objective-C ABI
3232 version. Valid values are 1 and 2.</td>
3233 </tr>
3234 <tr>
3235 <td><tt>Objective-C&nbsp;Image&nbsp;Info&nbsp;Version</tt></td>
3236 <td align="left"><b>[Required]</b> &mdash; The version of the image info
3237 section. Currently always 0.</td>
3238 </tr>
3239 <tr>
3240 <td><tt>Objective-C&nbsp;Image&nbsp;Info&nbsp;Section</tt></td>
3241 <td align="left"><b>[Required]</b> &mdash; The section to place the
3242 metadata. Valid values are <tt>"__OBJC, __image_info, regular"</tt> for
3243 Objective-C ABI version 1, and <tt>"__DATA,__objc_imageinfo, regular,
3244 no_dead_strip"</tt> for Objective-C ABI version 2.</td>
3245 </tr>
3246 <tr>
3247 <td><tt>Objective-C&nbsp;Garbage&nbsp;Collection</tt></td>
3248 <td align="left"><b>[Required]</b> &mdash; Specifies whether garbage
3249 collection is supported or not. Valid values are 0, for no garbage
3250 collection, and 2, for garbage collection supported.</td>
3251 </tr>
3252 <tr>
3253 <td><tt>Objective-C&nbsp;GC&nbsp;Only</tt></td>
3254 <td align="left"><b>[Optional]</b> &mdash; Specifies that only garbage
3255 collection is supported. If present, its value must be 6. This flag
3256 requires that the <tt>Objective-C Garbage Collection</tt> flag have the
3257 value 2.</td>
3258 </tr>
3259 </tbody>
3260</table>
3261
3262<p>Some important flag interactions:</p>
3263
3264<ul>
3265 <li>If a module with <tt>Objective-C Garbage Collection</tt> set to 0 is
3266 merged with a module with <tt>Objective-C Garbage Collection</tt> set to
3267 2, then the resulting module has the <tt>Objective-C Garbage
3268 Collection</tt> flag set to 0.</li>
3269
3270 <li>A module with <tt>Objective-C Garbage Collection</tt> set to 0 cannot be
3271 merged with a module with <tt>Objective-C GC Only</tt> set to 6.</li>
3272</ul>
3273
3274</div>
3275
Bill Wendlingb9d75a92012-02-11 11:59:36 +0000[diff] [blame]3276</div>
3277
3278<!-- *********************************************************************** -->
3279<h2>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]3280 <a name="intrinsic_globals">Intrinsic Global Variables</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3281</h2>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]3282<!-- *********************************************************************** -->
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3283<div>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]3284<p>LLVM has a number of "magic" global variables that contain data that affect
3285code generation or other IR semantics. These are documented here. All globals
Chris Lattner401e10c2009-07-20 06:14:25 +0000[diff] [blame]3286of this sort should have a section specified as "<tt>llvm.metadata</tt>". This
3287section and all globals that start with "<tt>llvm.</tt>" are reserved for use
3288by LLVM.</p>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]3289
3290<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3291<h3>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]3292<a name="intg_used">The '<tt>llvm.used</tt>' Global Variable</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3293</h3>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]3294
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3295<div>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]3296
3297<p>The <tt>@llvm.used</tt> global is an array with i8* element type which has <a
3298href="#linkage_appending">appending linkage</a>. This array contains a list of
3299pointers to global variables and functions which may optionally have a pointer
3300cast formed of bitcast or getelementptr. For example, a legal use of it is:</p>
3301
Bill Wendling9ae75632011-11-08 00:32:45 +0000[diff] [blame]3302<div class="doc_code">
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]3303<pre>
Bill Wendling9ae75632011-11-08 00:32:45 +0000[diff] [blame]3304@X = global i8 4
3305@Y = global i32 123
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]3306
Bill Wendling9ae75632011-11-08 00:32:45 +0000[diff] [blame]3307@llvm.used = appending global [2 x i8*] [
3308 i8* @X,
3309 i8* bitcast (i32* @Y to i8*)
3310], section "llvm.metadata"
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]3311</pre>
Bill Wendling9ae75632011-11-08 00:32:45 +0000[diff] [blame]3312</div>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]3313
3314<p>If a global variable appears in the <tt>@llvm.used</tt> list, then the
Bill Wendling9ae75632011-11-08 00:32:45 +0000[diff] [blame]3315 compiler, assembler, and linker are required to treat the symbol as if there
3316 is a reference to the global that it cannot see. For example, if a variable
3317 has internal linkage and no references other than that from
3318 the <tt>@llvm.used</tt> list, it cannot be deleted. This is commonly used to
3319 represent references from inline asms and other things the compiler cannot
3320 "see", and corresponds to "<tt>attribute((used))</tt>" in GNU C.</p>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]3321
3322<p>On some targets, the code generator must emit a directive to the assembler or
Bill Wendling9ae75632011-11-08 00:32:45 +0000[diff] [blame]3323 object file to prevent the assembler and linker from molesting the
3324 symbol.</p>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]3325
3326</div>
3327
3328<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3329<h3>
3330 <a name="intg_compiler_used">
3331 The '<tt>llvm.compiler.used</tt>' Global Variable
3332 </a>
3333</h3>
Chris Lattner401e10c2009-07-20 06:14:25 +0000[diff] [blame]3334
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3335<div>
Chris Lattner401e10c2009-07-20 06:14:25 +0000[diff] [blame]3336
3337<p>The <tt>@llvm.compiler.used</tt> directive is the same as the
Bill Wendling9ae75632011-11-08 00:32:45 +0000[diff] [blame]3338 <tt>@llvm.used</tt> directive, except that it only prevents the compiler from
3339 touching the symbol. On targets that support it, this allows an intelligent
3340 linker to optimize references to the symbol without being impeded as it would
3341 be by <tt>@llvm.used</tt>.</p>
Chris Lattner401e10c2009-07-20 06:14:25 +0000[diff] [blame]3342
3343<p>This is a rare construct that should only be used in rare circumstances, and
Bill Wendling9ae75632011-11-08 00:32:45 +0000[diff] [blame]3344 should not be exposed to source languages.</p>
Chris Lattner401e10c2009-07-20 06:14:25 +0000[diff] [blame]3345
3346</div>
3347
3348<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3349<h3>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]3350<a name="intg_global_ctors">The '<tt>llvm.global_ctors</tt>' Global Variable</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3351</h3>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]3352
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3353<div>
Bill Wendling9ae75632011-11-08 00:32:45 +0000[diff] [blame]3354
3355<div class="doc_code">
David Chisnalle31e9962010-04-30 19:23:49 +0000[diff] [blame]3356<pre>
3357%0 = type { i32, void ()* }
David Chisnall27195a52010-04-30 19:27:35 +0000[diff] [blame]3358@llvm.global_ctors = appending global [1 x %0] [%0 { i32 65535, void ()* @ctor }]
David Chisnalle31e9962010-04-30 19:23:49 +0000[diff] [blame]3359</pre>
Bill Wendling9ae75632011-11-08 00:32:45 +0000[diff] [blame]3360</div>
3361
3362<p>The <tt>@llvm.global_ctors</tt> array contains a list of constructor
3363 functions and associated priorities. The functions referenced by this array
3364 will be called in ascending order of priority (i.e. lowest first) when the
3365 module is loaded. The order of functions with the same priority is not
3366 defined.</p>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]3367
3368</div>
3369
3370<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3371<h3>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]3372<a name="intg_global_dtors">The '<tt>llvm.global_dtors</tt>' Global Variable</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3373</h3>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]3374
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3375<div>
Bill Wendling9ae75632011-11-08 00:32:45 +0000[diff] [blame]3376
3377<div class="doc_code">
David Chisnalle31e9962010-04-30 19:23:49 +0000[diff] [blame]3378<pre>
3379%0 = type { i32, void ()* }
David Chisnall27195a52010-04-30 19:27:35 +0000[diff] [blame]3380@llvm.global_dtors = appending global [1 x %0] [%0 { i32 65535, void ()* @dtor }]
David Chisnalle31e9962010-04-30 19:23:49 +0000[diff] [blame]3381</pre>
Bill Wendling9ae75632011-11-08 00:32:45 +0000[diff] [blame]3382</div>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]3383
Bill Wendling9ae75632011-11-08 00:32:45 +0000[diff] [blame]3384<p>The <tt>@llvm.global_dtors</tt> array contains a list of destructor functions
3385 and associated priorities. The functions referenced by this array will be
3386 called in descending order of priority (i.e. highest first) when the module
3387 is loaded. The order of functions with the same priority is not defined.</p>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]3388
3389</div>
3390
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3391</div>
Chris Lattner857755c2009-07-20 05:55:19 +0000[diff] [blame]3392
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]3393<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3394<h2><a name="instref">Instruction Reference</a></h2>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3395<!-- *********************************************************************** -->
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]3396
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3397<div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]3398
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3399<p>The LLVM instruction set consists of several different classifications of
3400 instructions: <a href="#terminators">terminator
3401 instructions</a>, <a href="#binaryops">binary instructions</a>,
3402 <a href="#bitwiseops">bitwise binary instructions</a>,
3403 <a href="#memoryops">memory instructions</a>, and
3404 <a href="#otherops">other instructions</a>.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]3405
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3406<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3407<h3>
3408 <a name="terminators">Terminator Instructions</a>
3409</h3>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]3410
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3411<div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]3412
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3413<p>As mentioned <a href="#functionstructure">previously</a>, every basic block
3414 in a program ends with a "Terminator" instruction, which indicates which
3415 block should be executed after the current block is finished. These
3416 terminator instructions typically yield a '<tt>void</tt>' value: they produce
3417 control flow, not values (the one exception being the
3418 '<a href="#i_invoke"><tt>invoke</tt></a>' instruction).</p>
3419
Chris Lattner6445ecb2011-08-02 20:29:13 +0000[diff] [blame]3420<p>The terminator instructions are:
3421 '<a href="#i_ret"><tt>ret</tt></a>',
3422 '<a href="#i_br"><tt>br</tt></a>',
3423 '<a href="#i_switch"><tt>switch</tt></a>',
3424 '<a href="#i_indirectbr"><tt>indirectbr</tt></a>',
3425 '<a href="#i_invoke"><tt>invoke</tt></a>',
Chris Lattner6445ecb2011-08-02 20:29:13 +0000[diff] [blame]3426 '<a href="#i_resume"><tt>resume</tt></a>', and
3427 '<a href="#i_unreachable"><tt>unreachable</tt></a>'.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]3428
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3429<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3430<h4>
3431 <a name="i_ret">'<tt>ret</tt>' Instruction</a>
3432</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3433
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3434<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3435
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3436<h5>Syntax:</h5>
Dan Gohmanb1e6b962008-10-04 19:00:07 +0000[diff] [blame]3437<pre>
3438 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]3439 ret void <i>; Return from void function</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3440</pre>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]3441
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3442<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3443<p>The '<tt>ret</tt>' instruction is used to return control flow (and optionally
3444 a value) from a function back to the caller.</p>
3445
3446<p>There are two forms of the '<tt>ret</tt>' instruction: one that returns a
3447 value and then causes control flow, and one that just causes control flow to
3448 occur.</p>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]3449
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3450<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3451<p>The '<tt>ret</tt>' instruction optionally accepts a single argument, the
3452 return value. The type of the return value must be a
3453 '<a href="#t_firstclass">first class</a>' type.</p>
Dan Gohmanb1e6b962008-10-04 19:00:07 +0000[diff] [blame]3454
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3455<p>A function is not <a href="#wellformed">well formed</a> if it it has a
3456 non-void return type and contains a '<tt>ret</tt>' instruction with no return
3457 value or a return value with a type that does not match its type, or if it
3458 has a void return type and contains a '<tt>ret</tt>' instruction with a
3459 return value.</p>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]3460
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3461<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3462<p>When the '<tt>ret</tt>' instruction is executed, control flow returns back to
3463 the calling function's context. If the caller is a
3464 "<a href="#i_call"><tt>call</tt></a>" instruction, execution continues at the
3465 instruction after the call. If the caller was an
3466 "<a href="#i_invoke"><tt>invoke</tt></a>" instruction, execution continues at
3467 the beginning of the "normal" destination block. If the instruction returns
3468 a value, that value shall set the call or invoke instruction's return
3469 value.</p>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]3470
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3471<h5>Example:</h5>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]3472<pre>
3473 ret i32 5 <i>; Return an integer value of 5</i>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]3474 ret void <i>; Return from a void function</i>
Bill Wendling0a4bbbf2009-02-28 22:12:54 +0000[diff] [blame]3475 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]3476</pre>
Dan Gohmanf3e60bd2009-01-12 23:12:39 +0000[diff] [blame]3477
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3478</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3479<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3480<h4>
3481 <a name="i_br">'<tt>br</tt>' Instruction</a>
3482</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3483
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3484<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3485
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3486<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3487<pre>
Bill Wendlingb3aa4712011-07-26 10:41:15 +0000[diff] [blame]3488 br i1 &lt;cond&gt;, label &lt;iftrue&gt;, label &lt;iffalse&gt;
3489 br label &lt;dest&gt; <i>; Unconditional branch</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3490</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3491
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3492<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3493<p>The '<tt>br</tt>' instruction is used to cause control flow to transfer to a
3494 different basic block in the current function. There are two forms of this
3495 instruction, corresponding to a conditional branch and an unconditional
3496 branch.</p>
3497
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3498<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3499<p>The conditional branch form of the '<tt>br</tt>' instruction takes a single
3500 '<tt>i1</tt>' value and two '<tt>label</tt>' values. The unconditional form
3501 of the '<tt>br</tt>' instruction takes a single '<tt>label</tt>' value as a
3502 target.</p>
3503
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3504<h5>Semantics:</h5>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3505<p>Upon execution of a conditional '<tt>br</tt>' instruction, the '<tt>i1</tt>'
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3506 argument is evaluated. If the value is <tt>true</tt>, control flows to the
3507 '<tt>iftrue</tt>' <tt>label</tt> argument. If "cond" is <tt>false</tt>,
3508 control flows to the '<tt>iffalse</tt>' <tt>label</tt> argument.</p>
3509
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3510<h5>Example:</h5>
Bill Wendlingc39e3e02009-07-20 02:39:26 +0000[diff] [blame]3511<pre>
3512Test:
3513 %cond = <a href="#i_icmp">icmp</a> eq i32 %a, %b
3514 br i1 %cond, label %IfEqual, label %IfUnequal
3515IfEqual:
3516 <a href="#i_ret">ret</a> i32 1
3517IfUnequal:
3518 <a href="#i_ret">ret</a> i32 0
3519</pre>
3520
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3521</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3522
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3523<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3524<h4>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3525 <a name="i_switch">'<tt>switch</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3526</h4>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3527
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3528<div>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3529
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3530<h5>Syntax:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3531<pre>
3532 switch &lt;intty&gt; &lt;value&gt;, label &lt;defaultdest&gt; [ &lt;intty&gt; &lt;val&gt;, label &lt;dest&gt; ... ]
3533</pre>
3534
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3535<h5>Overview:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3536<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]3537 several different places. It is a generalization of the '<tt>br</tt>'
3538 instruction, allowing a branch to occur to one of many possible
3539 destinations.</p>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3540
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3541<h5>Arguments:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3542<p>The '<tt>switch</tt>' instruction uses three parameters: an integer
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3543 comparison value '<tt>value</tt>', a default '<tt>label</tt>' destination,
3544 and an array of pairs of comparison value constants and '<tt>label</tt>'s.
3545 The table is not allowed to contain duplicate constant entries.</p>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3546
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3547<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3548<p>The <tt>switch</tt> instruction specifies a table of values and
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3549 destinations. When the '<tt>switch</tt>' instruction is executed, this table
3550 is searched for the given value. If the value is found, control flow is
Benjamin Kramer8040cd32009-10-12 14:46:08 +0000[diff] [blame]3551 transferred to the corresponding destination; otherwise, control flow is
3552 transferred to the default destination.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3553
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3554<h5>Implementation:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3555<p>Depending on properties of the target machine and the particular
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3556 <tt>switch</tt> instruction, this instruction may be code generated in
3557 different ways. For example, it could be generated as a series of chained
3558 conditional branches or with a lookup table.</p>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3559
3560<h5>Example:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3561<pre>
3562 <i>; Emulate a conditional br instruction</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3563 %Val = <a href="#i_zext">zext</a> i1 %value to i32
Dan Gohman2a08c532009-01-04 23:44:43 +0000[diff] [blame]3564 switch i32 %Val, label %truedest [ i32 0, label %falsedest ]
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3565
3566 <i>; Emulate an unconditional br instruction</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3567 switch i32 0, label %dest [ ]
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]3568
3569 <i>; Implement a jump table:</i>
Dan Gohman2a08c532009-01-04 23:44:43 +0000[diff] [blame]3570 switch i32 %val, label %otherwise [ i32 0, label %onzero
3571 i32 1, label %onone
3572 i32 2, label %ontwo ]
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3573</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3574
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3575</div>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3576
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]3577
3578<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3579<h4>
Chris Lattnerab21db72009-10-28 00:19:10 +0000[diff] [blame]3580 <a name="i_indirectbr">'<tt>indirectbr</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3581</h4>
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]3582
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3583<div>
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]3584
3585<h5>Syntax:</h5>
3586<pre>
Chris Lattnerab21db72009-10-28 00:19:10 +0000[diff] [blame]3587 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]3588</pre>
3589
3590<h5>Overview:</h5>
3591
Chris Lattnerab21db72009-10-28 00:19:10 +0000[diff] [blame]3592<p>The '<tt>indirectbr</tt>' instruction implements an indirect branch to a label
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]3593 within the current function, whose address is specified by
Chris Lattnerc6f44362009-10-27 21:01:34 +0000[diff] [blame]3594 "<tt>address</tt>". Address must be derived from a <a
3595 href="#blockaddress">blockaddress</a> constant.</p>
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]3596
3597<h5>Arguments:</h5>
3598
3599<p>The '<tt>address</tt>' argument is the address of the label to jump to. The
3600 rest of the arguments indicate the full set of possible destinations that the
3601 address may point to. Blocks are allowed to occur multiple times in the
3602 destination list, though this isn't particularly useful.</p>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3603
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]3604<p>This destination list is required so that dataflow analysis has an accurate
3605 understanding of the CFG.</p>
3606
3607<h5>Semantics:</h5>
3608
3609<p>Control transfers to the block specified in the address argument. All
3610 possible destination blocks must be listed in the label list, otherwise this
3611 instruction has undefined behavior. This implies that jumps to labels
3612 defined in other functions have undefined behavior as well.</p>
3613
3614<h5>Implementation:</h5>
3615
3616<p>This is typically implemented with a jump through a register.</p>
3617
3618<h5>Example:</h5>
3619<pre>
Chris Lattnerab21db72009-10-28 00:19:10 +0000[diff] [blame]3620 indirectbr i8* %Addr, [ label %bb1, label %bb2, label %bb3 ]
Chris Lattnerf9be95f2009-10-27 19:13:16 +0000[diff] [blame]3621</pre>
3622
3623</div>
3624
3625
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3626<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3627<h4>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3628 <a name="i_invoke">'<tt>invoke</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3629</h4>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3630
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3631<div>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3632
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3633<h5>Syntax:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3634<pre>
Devang Patel307e8ab2008-10-07 17:48:33 +0000[diff] [blame]3635 &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]3636 to label &lt;normal label&gt; unwind label &lt;exception label&gt;
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3637</pre>
3638
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]3639<h5>Overview:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3640<p>The '<tt>invoke</tt>' instruction causes control to transfer to a specified
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3641 function, with the possibility of control flow transfer to either the
3642 '<tt>normal</tt>' label or the '<tt>exception</tt>' label. If the callee
3643 function returns with the "<tt><a href="#i_ret">ret</a></tt>" instruction,
3644 control flow will return to the "normal" label. If the callee (or any
Bill Wendling7b9e5392012-02-06 21:57:33 +0000[diff] [blame]3645 indirect callees) returns via the "<a href="#i_resume"><tt>resume</tt></a>"
3646 instruction or other exception handling mechanism, control is interrupted and
3647 continued at the dynamically nearest "exception" label.</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3648
Bill Wendlingf78faf82011-08-02 21:52:38 +0000[diff] [blame]3649<p>The '<tt>exception</tt>' label is a
3650 <i><a href="ExceptionHandling.html#overview">landing pad</a></i> for the
3651 exception. As such, '<tt>exception</tt>' label is required to have the
3652 "<a href="#i_landingpad"><tt>landingpad</tt></a>" instruction, which contains
Chad Rosier85f5a1a2011-12-09 02:00:44 +0000[diff] [blame]3653 the information about the behavior of the program after unwinding
Bill Wendlingf78faf82011-08-02 21:52:38 +0000[diff] [blame]3654 happens, as its first non-PHI instruction. The restrictions on the
3655 "<tt>landingpad</tt>" instruction's tightly couples it to the
3656 "<tt>invoke</tt>" instruction, so that the important information contained
3657 within the "<tt>landingpad</tt>" instruction can't be lost through normal
3658 code motion.</p>
3659
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3660<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3661<p>This instruction requires several arguments:</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3662
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3663<ol>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3664 <li>The optional "cconv" marker indicates which <a href="#callingconv">calling
3665 convention</a> the call should use. If none is specified, the call
3666 defaults to using C calling conventions.</li>
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]3667
3668 <li>The optional <a href="#paramattrs">Parameter Attributes</a> list for
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3669 return values. Only '<tt>zeroext</tt>', '<tt>signext</tt>', and
3670 '<tt>inreg</tt>' attributes are valid here.</li>
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]3671
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3672 <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]3673 function value being invoked. In most cases, this is a direct function
3674 invocation, but indirect <tt>invoke</tt>s are just as possible, branching
3675 off an arbitrary pointer to function value.</li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3676
3677 <li>'<tt>function ptr val</tt>': An LLVM value containing a pointer to a
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3678 function to be invoked. </li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3679
3680 <li>'<tt>function args</tt>': argument list whose types match the function
Chris Lattner0724fbd2010-03-02 06:36:51 +0000[diff] [blame]3681 signature argument types and parameter attributes. All arguments must be
3682 of <a href="#t_firstclass">first class</a> type. If the function
3683 signature indicates the function accepts a variable number of arguments,
3684 the extra arguments can be specified.</li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3685
3686 <li>'<tt>normal label</tt>': the label reached when the called function
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3687 executes a '<tt><a href="#i_ret">ret</a></tt>' instruction. </li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3688
Bill Wendling7b9e5392012-02-06 21:57:33 +0000[diff] [blame]3689 <li>'<tt>exception label</tt>': the label reached when a callee returns via
3690 the <a href="#i_resume"><tt>resume</tt></a> instruction or other exception
3691 handling mechanism.</li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3692
Devang Patel307e8ab2008-10-07 17:48:33 +0000[diff] [blame]3693 <li>The optional <a href="#fnattrs">function attributes</a> list. Only
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3694 '<tt>noreturn</tt>', '<tt>nounwind</tt>', '<tt>readonly</tt>' and
3695 '<tt>readnone</tt>' attributes are valid here.</li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3696</ol>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3697
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3698<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3699<p>This instruction is designed to operate as a standard
3700 '<tt><a href="#i_call">call</a></tt>' instruction in most regards. The
3701 primary difference is that it establishes an association with a label, which
3702 is used by the runtime library to unwind the stack.</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3703
3704<p>This instruction is used in languages with destructors to ensure that proper
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3705 cleanup is performed in the case of either a <tt>longjmp</tt> or a thrown
3706 exception. Additionally, this is important for implementation of
3707 '<tt>catch</tt>' clauses in high-level languages that support them.</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3708
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3709<p>For the purposes of the SSA form, the definition of the value returned by the
3710 '<tt>invoke</tt>' instruction is deemed to occur on the edge from the current
3711 block to the "normal" label. If the callee unwinds then no return value is
3712 available.</p>
Dan Gohmanf96a4992009-05-22 21:47:08 +0000[diff] [blame]3713
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3714<h5>Example:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3715<pre>
Nick Lewyckyd703f652008-03-16 07:18:12 +0000[diff] [blame]3716 %retval = invoke i32 @Test(i32 15) to label %Continue
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]3717 unwind label %TestCleanup <i>; {i32}:retval set</i>
Nick Lewyckyd703f652008-03-16 07:18:12 +0000[diff] [blame]3718 %retval = invoke <a href="#callingconv">coldcc</a> i32 %Testfnptr(i32 15) to label %Continue
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]3719 unwind label %TestCleanup <i>; {i32}:retval set</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3720</pre>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3721
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3722</div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3723
Bill Wendlingdccc03b2011-07-31 06:30:59 +0000[diff] [blame]3724 <!-- _______________________________________________________________________ -->
3725
3726<h4>
3727 <a name="i_resume">'<tt>resume</tt>' Instruction</a>
3728</h4>
3729
3730<div>
3731
3732<h5>Syntax:</h5>
3733<pre>
3734 resume &lt;type&gt; &lt;value&gt;
3735</pre>
3736
3737<h5>Overview:</h5>
3738<p>The '<tt>resume</tt>' instruction is a terminator instruction that has no
3739 successors.</p>
3740
3741<h5>Arguments:</h5>
Bill Wendlingf78faf82011-08-02 21:52:38 +0000[diff] [blame]3742<p>The '<tt>resume</tt>' instruction requires one argument, which must have the
Bill Wendlinge4ad50b2011-08-03 18:37:32 +0000[diff] [blame]3743 same type as the result of any '<tt>landingpad</tt>' instruction in the same
3744 function.</p>
Bill Wendlingdccc03b2011-07-31 06:30:59 +0000[diff] [blame]3745
3746<h5>Semantics:</h5>
3747<p>The '<tt>resume</tt>' instruction resumes propagation of an existing
3748 (in-flight) exception whose unwinding was interrupted with
Bill Wendlingf78faf82011-08-02 21:52:38 +0000[diff] [blame]3749 a <a href="#i_landingpad"><tt>landingpad</tt></a> instruction.</p>
Bill Wendlingdccc03b2011-07-31 06:30:59 +0000[diff] [blame]3750
3751<h5>Example:</h5>
3752<pre>
Bill Wendlingf78faf82011-08-02 21:52:38 +0000[diff] [blame]3753 resume { i8*, i32 } %exn
Bill Wendlingdccc03b2011-07-31 06:30:59 +0000[diff] [blame]3754</pre>
3755
3756</div>
3757
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3758<!-- _______________________________________________________________________ -->
3759
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3760<h4>
3761 <a name="i_unreachable">'<tt>unreachable</tt>' Instruction</a>
3762</h4>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3763
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3764<div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3765
3766<h5>Syntax:</h5>
3767<pre>
3768 unreachable
3769</pre>
3770
3771<h5>Overview:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3772<p>The '<tt>unreachable</tt>' instruction has no defined semantics. This
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3773 instruction is used to inform the optimizer that a particular portion of the
3774 code is not reachable. This can be used to indicate that the code after a
3775 no-return function cannot be reached, and other facts.</p>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3776
3777<h5>Semantics:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3778<p>The '<tt>unreachable</tt>' instruction has no defined semantics.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3779
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]3780</div>
3781
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3782</div>
3783
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3784<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3785<h3>
3786 <a name="binaryops">Binary Operations</a>
3787</h3>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3788
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3789<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3790
3791<p>Binary operators are used to do most of the computation in a program. They
3792 require two operands of the same type, execute an operation on them, and
3793 produce a single value. The operands might represent multiple data, as is
3794 the case with the <a href="#t_vector">vector</a> data type. The result value
3795 has the same type as its operands.</p>
3796
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3797<p>There are several different binary operators:</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3798
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3799<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3800<h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3801 <a name="i_add">'<tt>add</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3802</h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3803
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3804<div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3805
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3806<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3807<pre>
Dan Gohmancbb38f22009-07-20 22:41:19 +0000[diff] [blame]3808 &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]3809 &lt;result&gt; = add nuw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3810 &lt;result&gt; = add nsw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3811 &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]3812</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3813
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3814<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3815<p>The '<tt>add</tt>' instruction returns the sum of its two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3816
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3817<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3818<p>The two arguments to the '<tt>add</tt>' instruction must
3819 be <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of
3820 integer values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3821
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3822<h5>Semantics:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3823<p>The value produced is the integer sum of the two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3824
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3825<p>If the sum has unsigned overflow, the result returned is the mathematical
3826 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]3827
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3828<p>Because LLVM integers use a two's complement representation, this instruction
3829 is appropriate for both signed and unsigned integers.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3830
Dan Gohman08d012e2009-07-22 22:44:56 +0000[diff] [blame]3831<p><tt>nuw</tt> and <tt>nsw</tt> stand for &quot;No Unsigned Wrap&quot;
3832 and &quot;No Signed Wrap&quot;, respectively. If the <tt>nuw</tt> and/or
3833 <tt>nsw</tt> keywords are present, the result value of the <tt>add</tt>
Dan Gohmanbfb056d2011-12-06 03:18:47 +0000[diff] [blame]3834 is a <a href="#poisonvalues">poison value</a> if unsigned and/or signed overflow,
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]3835 respectively, occurs.</p>
Dan Gohmancbb38f22009-07-20 22:41:19 +0000[diff] [blame]3836
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3837<h5>Example:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3838<pre>
3839 &lt;result&gt; = add i32 4, %var <i>; yields {i32}:result = 4 + %var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3840</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3841
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3842</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3843
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3844<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3845<h4>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3846 <a name="i_fadd">'<tt>fadd</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3847</h4>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3848
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3849<div>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3850
3851<h5>Syntax:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3852<pre>
3853 &lt;result&gt; = fadd &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3854</pre>
3855
3856<h5>Overview:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3857<p>The '<tt>fadd</tt>' instruction returns the sum of its two operands.</p>
3858
3859<h5>Arguments:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3860<p>The two arguments to the '<tt>fadd</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3861 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
3862 floating point values. Both arguments must have identical types.</p>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3863
3864<h5>Semantics:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3865<p>The value produced is the floating point sum of the two operands.</p>
3866
3867<h5>Example:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3868<pre>
3869 &lt;result&gt; = fadd float 4.0, %var <i>; yields {float}:result = 4.0 + %var</i>
3870</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3871
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3872</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3873
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3874<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3875<h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3876 <a name="i_sub">'<tt>sub</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3877</h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3878
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3879<div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3880
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3881<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3882<pre>
Dan Gohman08d012e2009-07-22 22:44:56 +0000[diff] [blame]3883 &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]3884 &lt;result&gt; = sub nuw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3885 &lt;result&gt; = sub nsw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3886 &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]3887</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3888
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3889<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3890<p>The '<tt>sub</tt>' instruction returns the difference of its two
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3891 operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3892
3893<p>Note that the '<tt>sub</tt>' instruction is used to represent the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3894 '<tt>neg</tt>' instruction present in most other intermediate
3895 representations.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3896
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3897<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3898<p>The two arguments to the '<tt>sub</tt>' instruction must
3899 be <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of
3900 integer values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3901
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3902<h5>Semantics:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3903<p>The value produced is the integer difference of the two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3904
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3905<p>If the difference has unsigned overflow, the result returned is the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3906 mathematical result modulo 2<sup>n</sup>, where n is the bit width of the
3907 result.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3908
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3909<p>Because LLVM integers use a two's complement representation, this instruction
3910 is appropriate for both signed and unsigned integers.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3911
Dan Gohman08d012e2009-07-22 22:44:56 +0000[diff] [blame]3912<p><tt>nuw</tt> and <tt>nsw</tt> stand for &quot;No Unsigned Wrap&quot;
3913 and &quot;No Signed Wrap&quot;, respectively. If the <tt>nuw</tt> and/or
3914 <tt>nsw</tt> keywords are present, the result value of the <tt>sub</tt>
Dan Gohmanbfb056d2011-12-06 03:18:47 +0000[diff] [blame]3915 is a <a href="#poisonvalues">poison value</a> if unsigned and/or signed overflow,
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]3916 respectively, occurs.</p>
Dan Gohmancbb38f22009-07-20 22:41:19 +0000[diff] [blame]3917
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3918<h5>Example:</h5>
Bill Wendlingaac388b2007-05-29 09:42:13 +0000[diff] [blame]3919<pre>
3920 &lt;result&gt; = sub i32 4, %var <i>; yields {i32}:result = 4 - %var</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3921 &lt;result&gt; = sub i32 0, %val <i>; yields {i32}:result = -%var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3922</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3923
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3924</div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3925
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3926<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3927<h4>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3928 <a name="i_fsub">'<tt>fsub</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3929</h4>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3930
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3931<div>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3932
3933<h5>Syntax:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3934<pre>
3935 &lt;result&gt; = fsub &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3936</pre>
3937
3938<h5>Overview:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3939<p>The '<tt>fsub</tt>' instruction returns the difference of its two
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3940 operands.</p>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3941
3942<p>Note that the '<tt>fsub</tt>' instruction is used to represent the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3943 '<tt>fneg</tt>' instruction present in most other intermediate
3944 representations.</p>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3945
3946<h5>Arguments:</h5>
Bill Wendlingd9fe2982009-07-20 02:32:41 +0000[diff] [blame]3947<p>The two arguments to the '<tt>fsub</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3948 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
3949 floating point values. Both arguments must have identical types.</p>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3950
3951<h5>Semantics:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3952<p>The value produced is the floating point difference of the two operands.</p>
3953
3954<h5>Example:</h5>
3955<pre>
3956 &lt;result&gt; = fsub float 4.0, %var <i>; yields {float}:result = 4.0 - %var</i>
3957 &lt;result&gt; = fsub float -0.0, %val <i>; yields {float}:result = -%var</i>
3958</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3959
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3960</div>
3961
3962<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3963<h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3964 <a name="i_mul">'<tt>mul</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]3965</h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3966
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]3967<div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3968
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3969<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3970<pre>
Dan Gohman08d012e2009-07-22 22:44:56 +0000[diff] [blame]3971 &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]3972 &lt;result&gt; = mul nuw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3973 &lt;result&gt; = mul nsw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3974 &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]3975</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3976
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3977<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3978<p>The '<tt>mul</tt>' instruction returns the product of its two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3979
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3980<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3981<p>The two arguments to the '<tt>mul</tt>' instruction must
3982 be <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of
3983 integer values. Both arguments must have identical types.</p>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]3984
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3985<h5>Semantics:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]3986<p>The value produced is the integer product of the two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]3987
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]3988<p>If the result of the multiplication has unsigned overflow, the result
3989 returned is the mathematical result modulo 2<sup>n</sup>, where n is the bit
3990 width of the result.</p>
3991
3992<p>Because LLVM integers use a two's complement representation, and the result
3993 is the same width as the operands, this instruction returns the correct
3994 result for both signed and unsigned integers. If a full product
3995 (e.g. <tt>i32</tt>x<tt>i32</tt>-><tt>i64</tt>) is needed, the operands should
3996 be sign-extended or zero-extended as appropriate to the width of the full
3997 product.</p>
3998
Dan Gohman08d012e2009-07-22 22:44:56 +0000[diff] [blame]3999<p><tt>nuw</tt> and <tt>nsw</tt> stand for &quot;No Unsigned Wrap&quot;
4000 and &quot;No Signed Wrap&quot;, respectively. If the <tt>nuw</tt> and/or
4001 <tt>nsw</tt> keywords are present, the result value of the <tt>mul</tt>
Dan Gohmanbfb056d2011-12-06 03:18:47 +0000[diff] [blame]4002 is a <a href="#poisonvalues">poison value</a> if unsigned and/or signed overflow,
Dan Gohmanfff6c532010-04-22 23:14:21 +0000[diff] [blame]4003 respectively, occurs.</p>
Dan Gohmancbb38f22009-07-20 22:41:19 +0000[diff] [blame]4004
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4005<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4006<pre>
4007 &lt;result&gt; = mul i32 4, %var <i>; yields {i32}:result = 4 * %var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4008</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4009
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4010</div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4011
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4012<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4013<h4>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]4014 <a name="i_fmul">'<tt>fmul</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4015</h4>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]4016
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4017<div>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]4018
4019<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4020<pre>
4021 &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]4022</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4023
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]4024<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4025<p>The '<tt>fmul</tt>' instruction returns the product of its two operands.</p>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]4026
4027<h5>Arguments:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]4028<p>The two arguments to the '<tt>fmul</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4029 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
4030 floating point values. Both arguments must have identical types.</p>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]4031
4032<h5>Semantics:</h5>
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]4033<p>The value produced is the floating point product of the two operands.</p>
4034
4035<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4036<pre>
4037 &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]4038</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4039
Dan Gohmanae3a0be2009-06-04 22:49:04 +0000[diff] [blame]4040</div>
4041
4042<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4043<h4>
4044 <a name="i_udiv">'<tt>udiv</tt>' Instruction</a>
4045</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4046
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4047<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4048
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]4049<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4050<pre>
Chris Lattner35bda892011-02-06 21:44:57 +0000[diff] [blame]4051 &lt;result&gt; = udiv &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
4052 &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]4053</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4054
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]4055<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4056<p>The '<tt>udiv</tt>' instruction returns the quotient of its two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4057
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]4058<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4059<p>The two arguments to the '<tt>udiv</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4060 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
4061 values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4062
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]4063<h5>Semantics:</h5>
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]4064<p>The value produced is the unsigned integer quotient of the two operands.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4065
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]4066<p>Note that unsigned integer division and signed integer division are distinct
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4067 operations; for signed integer division, use '<tt>sdiv</tt>'.</p>
4068
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]4069<p>Division by zero leads to undefined behavior.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4070
Chris Lattner35bda892011-02-06 21:44:57 +0000[diff] [blame]4071<p>If the <tt>exact</tt> keyword is present, the result value of the
Dan Gohmanbfb056d2011-12-06 03:18:47 +0000[diff] [blame]4072 <tt>udiv</tt> is a <a href="#poisonvalues">poison value</a> if %op1 is not a
Chris Lattner35bda892011-02-06 21:44:57 +0000[diff] [blame]4073 multiple of %op2 (as such, "((a udiv exact b) mul b) == a").</p>
4074
4075
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]4076<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4077<pre>
4078 &lt;result&gt; = udiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]4079</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4080
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]4081</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4082
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]4083<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4084<h4>
4085 <a name="i_sdiv">'<tt>sdiv</tt>' Instruction</a>
4086</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4087
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4088<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4089
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]4090<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4091<pre>
Dan Gohmancbb38f22009-07-20 22:41:19 +0000[diff] [blame]4092 &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]4093 &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]4094</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4095
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]4096<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4097<p>The '<tt>sdiv</tt>' instruction returns the quotient of its two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4098
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]4099<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4100<p>The two arguments to the '<tt>sdiv</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4101 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
4102 values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4103
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]4104<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4105<p>The value produced is the signed integer quotient of the two operands rounded
4106 towards zero.</p>
4107
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]4108<p>Note that signed integer division and unsigned integer division are distinct
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4109 operations; for unsigned integer division, use '<tt>udiv</tt>'.</p>
4110
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]4111<p>Division by zero leads to undefined behavior. Overflow also leads to
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4112 undefined behavior; this is a rare case, but can occur, for example, by doing
4113 a 32-bit division of -2147483648 by -1.</p>
4114
Dan Gohman9c5beed2009-07-22 00:04:19 +0000[diff] [blame]4115<p>If the <tt>exact</tt> keyword is present, the result value of the
Dan Gohmanbfb056d2011-12-06 03:18:47 +0000[diff] [blame]4116 <tt>sdiv</tt> is a <a href="#poisonvalues">poison value</a> if the result would
Dan Gohman38da9272010-07-11 00:08:34 +0000[diff] [blame]4117 be rounded.</p>
Dan Gohmancbb38f22009-07-20 22:41:19 +0000[diff] [blame]4118
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]4119<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4120<pre>
4121 &lt;result&gt; = sdiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]4122</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4123
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]4124</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4125
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]4126<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4127<h4>
4128 <a name="i_fdiv">'<tt>fdiv</tt>' Instruction</a>
4129</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4130
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4131<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4132
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4133<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4134<pre>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4135 &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]4136</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4137
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4138<h5>Overview:</h5>
4139<p>The '<tt>fdiv</tt>' instruction returns the quotient of its two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4140
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4141<h5>Arguments:</h5>
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]4142<p>The two arguments to the '<tt>fdiv</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4143 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
4144 floating point values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4145
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4146<h5>Semantics:</h5>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]4147<p>The value produced is the floating point quotient of the two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4148
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4149<h5>Example:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4150<pre>
4151 &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]4152</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4153
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4154</div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4155
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4156<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4157<h4>
4158 <a name="i_urem">'<tt>urem</tt>' Instruction</a>
4159</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4160
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4161<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4162
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]4163<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4164<pre>
4165 &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]4166</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4167
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]4168<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4169<p>The '<tt>urem</tt>' instruction returns the remainder from the unsigned
4170 division of its two arguments.</p>
4171
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]4172<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4173<p>The two arguments to the '<tt>urem</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4174 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
4175 values. Both arguments must have identical types.</p>
4176
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]4177<h5>Semantics:</h5>
4178<p>This instruction returns the unsigned integer <i>remainder</i> of a division.
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4179 This instruction always performs an unsigned division to get the
4180 remainder.</p>
4181
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]4182<p>Note that unsigned integer remainder and signed integer remainder are
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4183 distinct operations; for signed integer remainder, use '<tt>srem</tt>'.</p>
4184
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]4185<p>Taking the remainder of a division by zero leads to undefined behavior.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4186
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]4187<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4188<pre>
4189 &lt;result&gt; = urem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]4190</pre>
4191
4192</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4193
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]4194<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4195<h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4196 <a name="i_srem">'<tt>srem</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4197</h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4198
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4199<div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4200
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4201<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4202<pre>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4203 &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]4204</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4205
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4206<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4207<p>The '<tt>srem</tt>' instruction returns the remainder from the signed
4208 division of its two operands. This instruction can also take
4209 <a href="#t_vector">vector</a> versions of the values in which case the
4210 elements must be integers.</p>
Chris Lattnerc7d3ab32008-01-04 04:33:49 +0000[diff] [blame]4211
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4212<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4213<p>The two arguments to the '<tt>srem</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4214 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
4215 values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4216
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4217<h5>Semantics:</h5>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]4218<p>This instruction returns the <i>remainder</i> of a division (where the result
Duncan Sandsdea3a5e2011-03-07 09:12:24 +0000[diff] [blame]4219 is either zero or has the same sign as the dividend, <tt>op1</tt>), not the
4220 <i>modulo</i> operator (where the result is either zero or has the same sign
4221 as the divisor, <tt>op2</tt>) of a value.
4222 For more information about the difference,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4223 see <a href="http://mathforum.org/dr.math/problems/anne.4.28.99.html">The
4224 Math Forum</a>. For a table of how this is implemented in various languages,
4225 please see <a href="http://en.wikipedia.org/wiki/Modulo_operation">
4226 Wikipedia: modulo operation</a>.</p>
4227
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]4228<p>Note that signed integer remainder and unsigned integer remainder are
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4229 distinct operations; for unsigned integer remainder, use '<tt>urem</tt>'.</p>
4230
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]4231<p>Taking the remainder of a division by zero leads to undefined behavior.
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4232 Overflow also leads to undefined behavior; this is a rare case, but can
4233 occur, for example, by taking the remainder of a 32-bit division of
4234 -2147483648 by -1. (The remainder doesn't actually overflow, but this rule
4235 lets srem be implemented using instructions that return both the result of
4236 the division and the remainder.)</p>
4237
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4238<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4239<pre>
4240 &lt;result&gt; = srem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]4241</pre>
4242
4243</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4244
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]4245<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4246<h4>
4247 <a name="i_frem">'<tt>frem</tt>' Instruction</a>
4248</h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4249
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4250<div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4251
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]4252<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4253<pre>
4254 &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]4255</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4256
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]4257<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4258<p>The '<tt>frem</tt>' instruction returns the remainder from the division of
4259 its two operands.</p>
4260
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]4261<h5>Arguments:</h5>
4262<p>The two arguments to the '<tt>frem</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4263 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
4264 floating point values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4265
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]4266<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4267<p>This instruction returns the <i>remainder</i> of a division. The remainder
4268 has the same sign as the dividend.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4269
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]4270<h5>Example:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4271<pre>
4272 &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]4273</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4274
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4275</div>
Robert Bocchino7b81c752006-02-17 21:18:08 +0000[diff] [blame]4276
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4277</div>
4278
Reid Spencer8e11bf82007-02-02 13:57:07 +0000[diff] [blame]4279<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4280<h3>
4281 <a name="bitwiseops">Bitwise Binary Operations</a>
4282</h3>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4283
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4284<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4285
4286<p>Bitwise binary operators are used to do various forms of bit-twiddling in a
4287 program. They are generally very efficient instructions and can commonly be
4288 strength reduced from other instructions. They require two operands of the
4289 same type, execute an operation on them, and produce a single value. The
4290 resulting value is the same type as its operands.</p>
4291
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]4292<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4293<h4>
4294 <a name="i_shl">'<tt>shl</tt>' Instruction</a>
4295</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4296
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4297<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4298
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]4299<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4300<pre>
Chris Lattnerf067d582011-02-07 16:40:21 +0000[diff] [blame]4301 &lt;result&gt; = shl &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
4302 &lt;result&gt; = shl nuw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
4303 &lt;result&gt; = shl nsw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
4304 &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]4305</pre>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]4306
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]4307<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4308<p>The '<tt>shl</tt>' instruction returns the first operand shifted to the left
4309 a specified number of bits.</p>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]4310
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]4311<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4312<p>Both arguments to the '<tt>shl</tt>' instruction must be the
4313 same <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of
4314 integer type. '<tt>op2</tt>' is treated as an unsigned value.</p>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4315
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]4316<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4317<p>The value produced is <tt>op1</tt> * 2<sup><tt>op2</tt></sup> mod
4318 2<sup>n</sup>, where <tt>n</tt> is the width of the result. If <tt>op2</tt>
4319 is (statically or dynamically) negative or equal to or larger than the number
4320 of bits in <tt>op1</tt>, the result is undefined. If the arguments are
4321 vectors, each vector element of <tt>op1</tt> is shifted by the corresponding
4322 shift amount in <tt>op2</tt>.</p>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]4323
Chris Lattnerf067d582011-02-07 16:40:21 +0000[diff] [blame]4324<p>If the <tt>nuw</tt> keyword is present, then the shift produces a
Dan Gohmanbfb056d2011-12-06 03:18:47 +0000[diff] [blame]4325 <a href="#poisonvalues">poison value</a> if it shifts out any non-zero bits. If
Chris Lattner66298c12011-02-09 16:44:44 +0000[diff] [blame]4326 the <tt>nsw</tt> keyword is present, then the shift produces a
Dan Gohmanbfb056d2011-12-06 03:18:47 +0000[diff] [blame]4327 <a href="#poisonvalues">poison value</a> if it shifts out any bits that disagree
Chris Lattnerf067d582011-02-07 16:40:21 +0000[diff] [blame]4328 with the resultant sign bit. As such, NUW/NSW have the same semantics as
4329 they would if the shift were expressed as a mul instruction with the same
4330 nsw/nuw bits in (mul %op1, (shl 1, %op2)).</p>
4331
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4332<h5>Example:</h5>
4333<pre>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]4334 &lt;result&gt; = shl i32 4, %var <i>; yields {i32}: 4 &lt;&lt; %var</i>
4335 &lt;result&gt; = shl i32 4, 2 <i>; yields {i32}: 16</i>
4336 &lt;result&gt; = shl i32 1, 10 <i>; yields {i32}: 1024</i>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]4337 &lt;result&gt; = shl i32 1, 32 <i>; undefined</i>
Mon P Wange9f10152008-12-09 05:46:39 +0000[diff] [blame]4338 &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]4339</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4340
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]4341</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4342
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]4343<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4344<h4>
4345 <a name="i_lshr">'<tt>lshr</tt>' Instruction</a>
4346</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4347
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4348<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4349
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]4350<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4351<pre>
Chris Lattnerf067d582011-02-07 16:40:21 +0000[diff] [blame]4352 &lt;result&gt; = lshr &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
4353 &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]4354</pre>
4355
4356<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4357<p>The '<tt>lshr</tt>' instruction (logical shift right) returns the first
4358 operand shifted to the right a specified number of bits with zero fill.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]4359
4360<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4361<p>Both arguments to the '<tt>lshr</tt>' instruction must be the same
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4362 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
4363 type. '<tt>op2</tt>' is treated as an unsigned value.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]4364
4365<h5>Semantics:</h5>
4366<p>This instruction always performs a logical shift right operation. The most
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4367 significant bits of the result will be filled with zero bits after the shift.
4368 If <tt>op2</tt> is (statically or dynamically) equal to or larger than the
4369 number of bits in <tt>op1</tt>, the result is undefined. If the arguments are
4370 vectors, each vector element of <tt>op1</tt> is shifted by the corresponding
4371 shift amount in <tt>op2</tt>.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]4372
Chris Lattnerf067d582011-02-07 16:40:21 +0000[diff] [blame]4373<p>If the <tt>exact</tt> keyword is present, the result value of the
Dan Gohmanbfb056d2011-12-06 03:18:47 +0000[diff] [blame]4374 <tt>lshr</tt> is a <a href="#poisonvalues">poison value</a> if any of the bits
Chris Lattnerf067d582011-02-07 16:40:21 +0000[diff] [blame]4375 shifted out are non-zero.</p>
4376
4377
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]4378<h5>Example:</h5>
4379<pre>
4380 &lt;result&gt; = lshr i32 4, 1 <i>; yields {i32}:result = 2</i>
4381 &lt;result&gt; = lshr i32 4, 2 <i>; yields {i32}:result = 1</i>
4382 &lt;result&gt; = lshr i8 4, 3 <i>; yields {i8}:result = 0</i>
4383 &lt;result&gt; = lshr i8 -2, 1 <i>; yields {i8}:result = 0x7FFFFFFF </i>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]4384 &lt;result&gt; = lshr i32 1, 32 <i>; undefined</i>
Mon P Wange9f10152008-12-09 05:46:39 +0000[diff] [blame]4385 &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]4386</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4387
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]4388</div>
4389
Reid Spencer8e11bf82007-02-02 13:57:07 +0000[diff] [blame]4390<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4391<h4>
4392 <a name="i_ashr">'<tt>ashr</tt>' Instruction</a>
4393</h4>
4394
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4395<div>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]4396
4397<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4398<pre>
Chris Lattnerf067d582011-02-07 16:40:21 +0000[diff] [blame]4399 &lt;result&gt; = ashr &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
4400 &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]4401</pre>
4402
4403<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4404<p>The '<tt>ashr</tt>' instruction (arithmetic shift right) returns the first
4405 operand shifted to the right a specified number of bits with sign
4406 extension.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]4407
4408<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4409<p>Both arguments to the '<tt>ashr</tt>' instruction must be the same
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4410 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
4411 type. '<tt>op2</tt>' is treated as an unsigned value.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]4412
4413<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4414<p>This instruction always performs an arithmetic shift right operation, The
4415 most significant bits of the result will be filled with the sign bit
4416 of <tt>op1</tt>. If <tt>op2</tt> is (statically or dynamically) equal to or
4417 larger than the number of bits in <tt>op1</tt>, the result is undefined. If
4418 the arguments are vectors, each vector element of <tt>op1</tt> is shifted by
4419 the corresponding shift amount in <tt>op2</tt>.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]4420
Chris Lattnerf067d582011-02-07 16:40:21 +0000[diff] [blame]4421<p>If the <tt>exact</tt> keyword is present, the result value of the
Dan Gohmanbfb056d2011-12-06 03:18:47 +0000[diff] [blame]4422 <tt>ashr</tt> is a <a href="#poisonvalues">poison value</a> if any of the bits
Chris Lattnerf067d582011-02-07 16:40:21 +0000[diff] [blame]4423 shifted out are non-zero.</p>
4424
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]4425<h5>Example:</h5>
4426<pre>
4427 &lt;result&gt; = ashr i32 4, 1 <i>; yields {i32}:result = 2</i>
4428 &lt;result&gt; = ashr i32 4, 2 <i>; yields {i32}:result = 1</i>
4429 &lt;result&gt; = ashr i8 4, 3 <i>; yields {i8}:result = 0</i>
4430 &lt;result&gt; = ashr i8 -2, 1 <i>; yields {i8}:result = -1</i>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]4431 &lt;result&gt; = ashr i32 1, 32 <i>; undefined</i>
Mon P Wange9f10152008-12-09 05:46:39 +0000[diff] [blame]4432 &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]4433</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4434
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]4435</div>
4436
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4437<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4438<h4>
4439 <a name="i_and">'<tt>and</tt>' Instruction</a>
4440</h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4441
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4442<div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4443
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4444<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4445<pre>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4446 &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]4447</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4448
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4449<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4450<p>The '<tt>and</tt>' instruction returns the bitwise logical and of its two
4451 operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4452
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4453<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4454<p>The two arguments to the '<tt>and</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4455 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
4456 values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4457
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4458<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4459<p>The truth table used for the '<tt>and</tt>' instruction is:</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4460
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4461<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4462 <tbody>
4463 <tr>
Bill Wendling262396b2011-12-09 22:41:40 +0000[diff] [blame]4464 <th>In0</th>
4465 <th>In1</th>
4466 <th>Out</th>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4467 </tr>
4468 <tr>
4469 <td>0</td>
4470 <td>0</td>
4471 <td>0</td>
4472 </tr>
4473 <tr>
4474 <td>0</td>
4475 <td>1</td>
4476 <td>0</td>
4477 </tr>
4478 <tr>
4479 <td>1</td>
4480 <td>0</td>
4481 <td>0</td>
4482 </tr>
4483 <tr>
4484 <td>1</td>
4485 <td>1</td>
4486 <td>1</td>
4487 </tr>
4488 </tbody>
4489</table>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4490
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4491<h5>Example:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4492<pre>
4493 &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]4494 &lt;result&gt; = and i32 15, 40 <i>; yields {i32}:result = 8</i>
4495 &lt;result&gt; = and i32 4, 8 <i>; yields {i32}:result = 0</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4496</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4497</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4498<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4499<h4>
4500 <a name="i_or">'<tt>or</tt>' Instruction</a>
4501</h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4502
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4503<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4504
4505<h5>Syntax:</h5>
4506<pre>
4507 &lt;result&gt; = or &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
4508</pre>
4509
4510<h5>Overview:</h5>
4511<p>The '<tt>or</tt>' instruction returns the bitwise logical inclusive or of its
4512 two operands.</p>
4513
4514<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4515<p>The two arguments to the '<tt>or</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4516 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
4517 values. Both arguments must have identical types.</p>
4518
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4519<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4520<p>The truth table used for the '<tt>or</tt>' instruction is:</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4521
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4522<table border="1" cellspacing="0" cellpadding="4">
4523 <tbody>
4524 <tr>
Bill Wendling262396b2011-12-09 22:41:40 +0000[diff] [blame]4525 <th>In0</th>
4526 <th>In1</th>
4527 <th>Out</th>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4528 </tr>
4529 <tr>
4530 <td>0</td>
4531 <td>0</td>
4532 <td>0</td>
4533 </tr>
4534 <tr>
4535 <td>0</td>
4536 <td>1</td>
4537 <td>1</td>
4538 </tr>
4539 <tr>
4540 <td>1</td>
4541 <td>0</td>
4542 <td>1</td>
4543 </tr>
4544 <tr>
4545 <td>1</td>
4546 <td>1</td>
4547 <td>1</td>
4548 </tr>
4549 </tbody>
4550</table>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4551
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4552<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4553<pre>
4554 &lt;result&gt; = or i32 4, %var <i>; yields {i32}:result = 4 | %var</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4555 &lt;result&gt; = or i32 15, 40 <i>; yields {i32}:result = 47</i>
4556 &lt;result&gt; = or i32 4, 8 <i>; yields {i32}:result = 12</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4557</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4558
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4559</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4560
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4561<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4562<h4>
4563 <a name="i_xor">'<tt>xor</tt>' Instruction</a>
4564</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4565
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4566<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4567
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4568<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4569<pre>
4570 &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]4571</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4572
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4573<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4574<p>The '<tt>xor</tt>' instruction returns the bitwise logical exclusive or of
4575 its two operands. The <tt>xor</tt> is used to implement the "one's
4576 complement" operation, which is the "~" operator in C.</p>
4577
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4578<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4579<p>The two arguments to the '<tt>xor</tt>' instruction must be
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4580 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
4581 values. Both arguments must have identical types.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4582
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4583<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4584<p>The truth table used for the '<tt>xor</tt>' instruction is:</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4585
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4586<table border="1" cellspacing="0" cellpadding="4">
4587 <tbody>
4588 <tr>
Bill Wendling262396b2011-12-09 22:41:40 +0000[diff] [blame]4589 <th>In0</th>
4590 <th>In1</th>
4591 <th>Out</th>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4592 </tr>
4593 <tr>
4594 <td>0</td>
4595 <td>0</td>
4596 <td>0</td>
4597 </tr>
4598 <tr>
4599 <td>0</td>
4600 <td>1</td>
4601 <td>1</td>
4602 </tr>
4603 <tr>
4604 <td>1</td>
4605 <td>0</td>
4606 <td>1</td>
4607 </tr>
4608 <tr>
4609 <td>1</td>
4610 <td>1</td>
4611 <td>0</td>
4612 </tr>
4613 </tbody>
4614</table>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4615
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4616<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4617<pre>
4618 &lt;result&gt; = xor i32 4, %var <i>; yields {i32}:result = 4 ^ %var</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4619 &lt;result&gt; = xor i32 15, 40 <i>; yields {i32}:result = 39</i>
4620 &lt;result&gt; = xor i32 4, 8 <i>; yields {i32}:result = 12</i>
4621 &lt;result&gt; = xor i32 %V, -1 <i>; yields {i32}:result = ~%V</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4622</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4623
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4624</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4625
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4626</div>
4627
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4628<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4629<h3>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4630 <a name="vectorops">Vector Operations</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4631</h3>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4632
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4633<div>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4634
4635<p>LLVM supports several instructions to represent vector operations in a
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4636 target-independent manner. These instructions cover the element-access and
4637 vector-specific operations needed to process vectors effectively. While LLVM
4638 does directly support these vector operations, many sophisticated algorithms
4639 will want to use target-specific intrinsics to take full advantage of a
4640 specific target.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4641
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4642<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4643<h4>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4644 <a name="i_extractelement">'<tt>extractelement</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4645</h4>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4646
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4647<div>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4648
4649<h5>Syntax:</h5>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4650<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4651 &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]4652</pre>
4653
4654<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4655<p>The '<tt>extractelement</tt>' instruction extracts a single scalar element
4656 from a vector at a specified index.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4657
4658
4659<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4660<p>The first operand of an '<tt>extractelement</tt>' instruction is a value
4661 of <a href="#t_vector">vector</a> type. The second operand is an index
4662 indicating the position from which to extract the element. The index may be
4663 a variable.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4664
4665<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4666<p>The result is a scalar of the same type as the element type of
4667 <tt>val</tt>. Its value is the value at position <tt>idx</tt> of
4668 <tt>val</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
4669 results are undefined.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4670
4671<h5>Example:</h5>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4672<pre>
Gabor Greifa5b6f452009-10-28 13:14:50 +0000[diff] [blame]4673 &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]4674</pre>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4675
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4676</div>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4677
4678<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4679<h4>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4680 <a name="i_insertelement">'<tt>insertelement</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4681</h4>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4682
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4683<div>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4684
4685<h5>Syntax:</h5>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4686<pre>
Dan Gohmanf3480b92008-05-12 23:38:42 +0000[diff] [blame]4687 &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]4688</pre>
4689
4690<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4691<p>The '<tt>insertelement</tt>' instruction inserts a scalar element into a
4692 vector at a specified index.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4693
4694<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4695<p>The first operand of an '<tt>insertelement</tt>' instruction is a value
4696 of <a href="#t_vector">vector</a> type. The second operand is a scalar value
4697 whose type must equal the element type of the first operand. The third
4698 operand is an index indicating the position at which to insert the value.
4699 The index may be a variable.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4700
4701<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4702<p>The result is a vector of the same type as <tt>val</tt>. Its element values
4703 are those of <tt>val</tt> except at position <tt>idx</tt>, where it gets the
4704 value <tt>elt</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
4705 results are undefined.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4706
4707<h5>Example:</h5>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4708<pre>
Gabor Greifa5b6f452009-10-28 13:14:50 +0000[diff] [blame]4709 &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]4710</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4711
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4712</div>
4713
4714<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4715<h4>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4716 <a name="i_shufflevector">'<tt>shufflevector</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4717</h4>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4718
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4719<div>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4720
4721<h5>Syntax:</h5>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4722<pre>
Mon P Wangaeb06d22008-11-10 04:46:22 +0000[diff] [blame]4723 &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]4724</pre>
4725
4726<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4727<p>The '<tt>shufflevector</tt>' instruction constructs a permutation of elements
4728 from two input vectors, returning a vector with the same element type as the
4729 input and length that is the same as the shuffle mask.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4730
4731<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4732<p>The first two operands of a '<tt>shufflevector</tt>' instruction are vectors
4733 with types that match each other. The third argument is a shuffle mask whose
4734 element type is always 'i32'. The result of the instruction is a vector
4735 whose length is the same as the shuffle mask and whose element type is the
4736 same as the element type of the first two operands.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4737
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4738<p>The shuffle mask operand is required to be a constant vector with either
4739 constant integer or undef values.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4740
4741<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4742<p>The elements of the two input vectors are numbered from left to right across
4743 both of the vectors. The shuffle mask operand specifies, for each element of
4744 the result vector, which element of the two input vectors the result element
4745 gets. The element selector may be undef (meaning "don't care") and the
4746 second operand may be undef if performing a shuffle from only one vector.</p>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4747
4748<h5>Example:</h5>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4749<pre>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]4750 &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]4751 &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]4752 &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]4753 &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]4754 &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]4755 &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]4756 &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]4757 &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]4758</pre>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4759
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4760</div>
Tanya Lattner09474292006-04-14 19:24:33 +0000[diff] [blame]4761
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4762</div>
4763
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]4764<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4765<h3>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4766 <a name="aggregateops">Aggregate Operations</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4767</h3>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4768
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4769<div>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4770
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4771<p>LLVM supports several instructions for working with
4772 <a href="#t_aggregate">aggregate</a> values.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4773
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4774<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4775<h4>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4776 <a name="i_extractvalue">'<tt>extractvalue</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4777</h4>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4778
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4779<div>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4780
4781<h5>Syntax:</h5>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4782<pre>
4783 &lt;result&gt; = extractvalue &lt;aggregate type&gt; &lt;val&gt;, &lt;idx&gt;{, &lt;idx&gt;}*
4784</pre>
4785
4786<h5>Overview:</h5>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4787<p>The '<tt>extractvalue</tt>' instruction extracts the value of a member field
4788 from an <a href="#t_aggregate">aggregate</a> value.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4789
4790<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4791<p>The first operand of an '<tt>extractvalue</tt>' instruction is a value
Chris Lattner61c70e92010-08-28 04:09:24 +0000[diff] [blame]4792 of <a href="#t_struct">struct</a> or
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4793 <a href="#t_array">array</a> type. The operands are constant indices to
4794 specify which value to extract in a similar manner as indices in a
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4795 '<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.</p>
Frits van Bommel13242892010-12-05 20:54:38 +0000[diff] [blame]4796 <p>The major differences to <tt>getelementptr</tt> indexing are:</p>
4797 <ul>
4798 <li>Since the value being indexed is not a pointer, the first index is
4799 omitted and assumed to be zero.</li>
4800 <li>At least one index must be specified.</li>
4801 <li>Not only struct indices but also array indices must be in
4802 bounds.</li>
4803 </ul>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4804
4805<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4806<p>The result is the value at the position in the aggregate specified by the
4807 index operands.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4808
4809<h5>Example:</h5>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4810<pre>
Gabor Greifa5b6f452009-10-28 13:14:50 +0000[diff] [blame]4811 &lt;result&gt; = extractvalue {i32, float} %agg, 0 <i>; yields i32</i>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4812</pre>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4813
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4814</div>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4815
4816<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4817<h4>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4818 <a name="i_insertvalue">'<tt>insertvalue</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4819</h4>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4820
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4821<div>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4822
4823<h5>Syntax:</h5>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4824<pre>
Bill Wendling194229e2011-07-26 20:42:28 +0000[diff] [blame]4825 &lt;result&gt; = insertvalue &lt;aggregate type&gt; &lt;val&gt;, &lt;ty&gt; &lt;elt&gt;, &lt;idx&gt;{, &lt;idx&gt;}* <i>; yields &lt;aggregate type&gt;</i>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4826</pre>
4827
4828<h5>Overview:</h5>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4829<p>The '<tt>insertvalue</tt>' instruction inserts a value into a member field
4830 in an <a href="#t_aggregate">aggregate</a> value.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4831
4832<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4833<p>The first operand of an '<tt>insertvalue</tt>' instruction is a value
Chris Lattner61c70e92010-08-28 04:09:24 +0000[diff] [blame]4834 of <a href="#t_struct">struct</a> or
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]4835 <a href="#t_array">array</a> type. The second operand is a first-class
4836 value to insert. The following operands are constant indices indicating
4837 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]4838 '<tt><a href="#i_extractvalue">extractvalue</a></tt>' instruction. The
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4839 value to insert must have the same type as the value identified by the
4840 indices.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4841
4842<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4843<p>The result is an aggregate of the same type as <tt>val</tt>. Its value is
4844 that of <tt>val</tt> except that the value at the position specified by the
4845 indices is that of <tt>elt</tt>.</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4846
4847<h5>Example:</h5>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4848<pre>
Chris Lattner8645d1a2011-05-22 07:18:08 +0000[diff] [blame]4849 %agg1 = insertvalue {i32, float} undef, i32 1, 0 <i>; yields {i32 1, float undef}</i>
4850 %agg2 = insertvalue {i32, float} %agg1, float %val, 1 <i>; yields {i32 1, float %val}</i>
4851 %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]4852</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4853
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4854</div>
4855
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4856</div>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]4857
4858<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4859<h3>
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]4860 <a name="memoryops">Memory Access and Addressing Operations</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4861</h3>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4862
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4863<div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4864
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4865<p>A key design point of an SSA-based representation is how it represents
4866 memory. In LLVM, no memory locations are in SSA form, which makes things
Victor Hernandez2fee2942009-10-26 23:44:29 +0000[diff] [blame]4867 very simple. This section describes how to read, write, and allocate
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4868 memory in LLVM.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4869
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4870<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4871<h4>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4872 <a name="i_alloca">'<tt>alloca</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4873</h4>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4874
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4875<div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4876
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4877<h5>Syntax:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4878<pre>
Dan Gohmanf75a7d32010-05-28 01:14:11 +0000[diff] [blame]4879 &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]4880</pre>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4881
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4882<h5>Overview:</h5>
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]4883<p>The '<tt>alloca</tt>' instruction allocates memory on the stack frame of the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4884 currently executing function, to be automatically released when this function
4885 returns to its caller. The object is always allocated in the generic address
4886 space (address space zero).</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4887
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4888<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4889<p>The '<tt>alloca</tt>' instruction
4890 allocates <tt>sizeof(&lt;type&gt;)*NumElements</tt> bytes of memory on the
4891 runtime stack, returning a pointer of the appropriate type to the program.
4892 If "NumElements" is specified, it is the number of elements allocated,
4893 otherwise "NumElements" is defaulted to be one. If a constant alignment is
4894 specified, the value result of the allocation is guaranteed to be aligned to
4895 at least that boundary. If not specified, or if zero, the target can choose
4896 to align the allocation on any convenient boundary compatible with the
4897 type.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4898
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4899<p>'<tt>type</tt>' may be any sized type.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4900
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4901<h5>Semantics:</h5>
Bill Wendling871eb0a2009-05-08 20:49:29 +0000[diff] [blame]4902<p>Memory is allocated; a pointer is returned. The operation is undefined if
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4903 there is insufficient stack space for the allocation. '<tt>alloca</tt>'d
4904 memory is automatically released when the function returns. The
4905 '<tt>alloca</tt>' instruction is commonly used to represent automatic
4906 variables that must have an address available. When the function returns
4907 (either with the <tt><a href="#i_ret">ret</a></tt>
Bill Wendling7b9e5392012-02-06 21:57:33 +0000[diff] [blame]4908 or <tt><a href="#i_resume">resume</a></tt> instructions), the memory is
Nick Lewycky84a1d232012-02-29 08:26:44 +0000[diff] [blame]4909 reclaimed. Allocating zero bytes is legal, but the result is undefined.
4910 The order in which memory is allocated (ie., which way the stack grows) is
Nick Lewycky75d05e62012-03-18 09:35:50 +0000[diff] [blame]4911 not specified.</p>
Nick Lewycky84a1d232012-02-29 08:26:44 +0000[diff] [blame]4912
4913<p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4914
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4915<h5>Example:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4916<pre>
Dan Gohman81e21672009-01-04 23:49:44 +0000[diff] [blame]4917 %ptr = alloca i32 <i>; yields {i32*}:ptr</i>
4918 %ptr = alloca i32, i32 4 <i>; yields {i32*}:ptr</i>
4919 %ptr = alloca i32, i32 4, align 1024 <i>; yields {i32*}:ptr</i>
4920 %ptr = alloca i32, align 1024 <i>; yields {i32*}:ptr</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4921</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4922
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4923</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]4924
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4925<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]4926<h4>
4927 <a name="i_load">'<tt>load</tt>' Instruction</a>
4928</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4929
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]4930<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4931
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4932<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4933<pre>
Pete Cooperf95acc62012-02-10 18:13:54 +0000[diff] [blame]4934 &lt;result&gt; = load [volatile] &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;][, !nontemporal !&lt;index&gt;][, !invariant.load !&lt;index&gt;]
Eli Friedmanf03bb262011-08-12 22:50:01 +0000[diff] [blame]4935 &lt;result&gt; = load atomic [volatile] &lt;ty&gt;* &lt;pointer&gt; [singlethread] &lt;ordering&gt;, align &lt;alignment&gt;
Bill Wendling7c78dbb2010-02-25 21:23:24 +0000[diff] [blame]4936 !&lt;index&gt; = !{ i32 1 }
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4937</pre>
4938
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4939<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4940<p>The '<tt>load</tt>' instruction is used to read from memory.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4941
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4942<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4943<p>The argument to the '<tt>load</tt>' instruction specifies the memory address
4944 from which to load. The pointer must point to
4945 a <a href="#t_firstclass">first class</a> type. If the <tt>load</tt> is
4946 marked as <tt>volatile</tt>, then the optimizer is not allowed to modify the
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]4947 number or order of execution of this <tt>load</tt> with other <a
4948 href="#volatile">volatile operations</a>.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4949
Eli Friedman21006d42011-08-09 23:02:53 +0000[diff] [blame]4950<p>If the <code>load</code> is marked as <code>atomic</code>, it takes an extra
4951 <a href="#ordering">ordering</a> and optional <code>singlethread</code>
4952 argument. The <code>release</code> and <code>acq_rel</code> orderings are
4953 not valid on <code>load</code> instructions. Atomic loads produce <a
4954 href="#memorymodel">defined</a> results when they may see multiple atomic
4955 stores. The type of the pointee must be an integer type whose bit width
4956 is a power of two greater than or equal to eight and less than or equal
4957 to a target-specific size limit. <code>align</code> must be explicitly
4958 specified on atomic loads, and the load has undefined behavior if the
4959 alignment is not set to a value which is at least the size in bytes of
4960 the pointee. <code>!nontemporal</code> does not have any defined semantics
4961 for atomic loads.</p>
4962
Bill Wendling7c78dbb2010-02-25 21:23:24 +0000[diff] [blame]4963<p>The optional constant <tt>align</tt> argument specifies the alignment of the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4964 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]4965 omitted <tt>align</tt> argument means that the operation has the preferential
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4966 alignment for the target. It is the responsibility of the code emitter to
4967 ensure that the alignment information is correct. Overestimating the
Bill Wendling7c78dbb2010-02-25 21:23:24 +0000[diff] [blame]4968 alignment results in undefined behavior. Underestimating the alignment may
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4969 produce less efficient code. An alignment of 1 is always safe.</p>
4970
Bill Wendling7c78dbb2010-02-25 21:23:24 +0000[diff] [blame]4971<p>The optional <tt>!nontemporal</tt> metadata must reference a single
4972 metatadata name &lt;index&gt; corresponding to a metadata node with
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]4973 one <tt>i32</tt> entry of value 1. The existence of
Bill Wendling7c78dbb2010-02-25 21:23:24 +0000[diff] [blame]4974 the <tt>!nontemporal</tt> metatadata on the instruction tells the optimizer
4975 and code generator that this load is not expected to be reused in the cache.
4976 The code generator may select special instructions to save cache bandwidth,
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]4977 such as the <tt>MOVNT</tt> instruction on x86.</p>
David Greene8939b0d2010-02-16 20:50:18 +0000[diff] [blame]4978
Pete Cooperf95acc62012-02-10 18:13:54 +0000[diff] [blame]4979<p>The optional <tt>!invariant.load</tt> metadata must reference a single
4980 metatadata name &lt;index&gt; corresponding to a metadata node with no
4981 entries. The existence of the <tt>!invariant.load</tt> metatadata on the
4982 instruction tells the optimizer and code generator that this load address
4983 points to memory which does not change value during program execution.
4984 The optimizer may then move this load around, for example, by hoisting it
4985 out of loops using loop invariant code motion.</p>
4986
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4987<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4988<p>The location of memory pointed to is loaded. If the value being loaded is of
4989 scalar type then the number of bytes read does not exceed the minimum number
4990 of bytes needed to hold all bits of the type. For example, loading an
4991 <tt>i24</tt> reads at most three bytes. When loading a value of a type like
4992 <tt>i20</tt> with a size that is not an integral number of bytes, the result
4993 is undefined if the value was not originally written using a store of the
4994 same type.</p>
4995
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]4996<h5>Examples:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]4997<pre>
4998 %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
4999 <a href="#i_store">store</a> i32 3, i32* %ptr <i>; yields {void}</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5000 %val = load i32* %ptr <i>; yields {i32}:val = i32 3</i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]5001</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5002
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5003</div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5004
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]5005<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5006<h4>
5007 <a name="i_store">'<tt>store</tt>' Instruction</a>
5008</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5009
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5010<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5011
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]5012<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5013<pre>
Bill Wendling262396b2011-12-09 22:41:40 +0000[diff] [blame]5014 store [volatile] &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;][, !nontemporal !&lt;index&gt;] <i>; yields {void}</i>
5015 store atomic [volatile] &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt; [singlethread] &lt;ordering&gt;, align &lt;alignment&gt; <i>; yields {void}</i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]5016</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5017
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]5018<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5019<p>The '<tt>store</tt>' instruction is used to write to memory.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5020
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]5021<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5022<p>There are two arguments to the '<tt>store</tt>' instruction: a value to store
5023 and an address at which to store it. The type of the
5024 '<tt>&lt;pointer&gt;</tt>' operand must be a pointer to
5025 the <a href="#t_firstclass">first class</a> type of the
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]5026 '<tt>&lt;value&gt;</tt>' operand. If the <tt>store</tt> is marked as
5027 <tt>volatile</tt>, then the optimizer is not allowed to modify the number or
5028 order of execution of this <tt>store</tt> with other <a
5029 href="#volatile">volatile operations</a>.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5030
Eli Friedman21006d42011-08-09 23:02:53 +0000[diff] [blame]5031<p>If the <code>store</code> is marked as <code>atomic</code>, it takes an extra
5032 <a href="#ordering">ordering</a> and optional <code>singlethread</code>
5033 argument. The <code>acquire</code> and <code>acq_rel</code> orderings aren't
5034 valid on <code>store</code> instructions. Atomic loads produce <a
5035 href="#memorymodel">defined</a> results when they may see multiple atomic
5036 stores. The type of the pointee must be an integer type whose bit width
5037 is a power of two greater than or equal to eight and less than or equal
5038 to a target-specific size limit. <code>align</code> must be explicitly
5039 specified on atomic stores, and the store has undefined behavior if the
5040 alignment is not set to a value which is at least the size in bytes of
5041 the pointee. <code>!nontemporal</code> does not have any defined semantics
5042 for atomic stores.</p>
5043
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5044<p>The optional constant "align" argument specifies the alignment of the
5045 operation (that is, the alignment of the memory address). A value of 0 or an
5046 omitted "align" argument means that the operation has the preferential
5047 alignment for the target. It is the responsibility of the code emitter to
5048 ensure that the alignment information is correct. Overestimating the
5049 alignment results in an undefined behavior. Underestimating the alignment may
5050 produce less efficient code. An alignment of 1 is always safe.</p>
5051
David Greene8939b0d2010-02-16 20:50:18 +0000[diff] [blame]5052<p>The optional !nontemporal metadata must reference a single metatadata
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]5053 name &lt;index&gt; corresponding to a metadata node with one i32 entry of
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]5054 value 1. The existence of the !nontemporal metatadata on the
David Greene8939b0d2010-02-16 20:50:18 +0000[diff] [blame]5055 instruction tells the optimizer and code generator that this load is
5056 not expected to be reused in the cache. The code generator may
5057 select special instructions to save cache bandwidth, such as the
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]5058 MOVNT instruction on x86.</p>
David Greene8939b0d2010-02-16 20:50:18 +0000[diff] [blame]5059
5060
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]5061<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5062<p>The contents of memory are updated to contain '<tt>&lt;value&gt;</tt>' at the
5063 location specified by the '<tt>&lt;pointer&gt;</tt>' operand. If
5064 '<tt>&lt;value&gt;</tt>' is of scalar type then the number of bytes written
5065 does not exceed the minimum number of bytes needed to hold all bits of the
5066 type. For example, storing an <tt>i24</tt> writes at most three bytes. When
5067 writing a value of a type like <tt>i20</tt> with a size that is not an
5068 integral number of bytes, it is unspecified what happens to the extra bits
5069 that do not belong to the type, but they will typically be overwritten.</p>
5070
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]5071<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5072<pre>
5073 %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
Bill Wendling8c6c72d2007-10-22 05:10:05 +0000[diff] [blame]5074 store i32 3, i32* %ptr <i>; yields {void}</i>
5075 %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]5076</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5077
Reid Spencer47ce1792006-11-09 21:15:49 +0000[diff] [blame]5078</div>
5079
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]5080<!-- _______________________________________________________________________ -->
NAKAMURA Takumib5bb29c2011-08-12 06:17:17 +0000[diff] [blame]5081<h4>
5082<a name="i_fence">'<tt>fence</tt>' Instruction</a>
5083</h4>
Eli Friedman47f35132011-07-25 23:16:38 +0000[diff] [blame]5084
NAKAMURA Takumib5bb29c2011-08-12 06:17:17 +0000[diff] [blame]5085<div>
Eli Friedman47f35132011-07-25 23:16:38 +0000[diff] [blame]5086
5087<h5>Syntax:</h5>
5088<pre>
5089 fence [singlethread] &lt;ordering&gt; <i>; yields {void}</i>
5090</pre>
5091
5092<h5>Overview:</h5>
5093<p>The '<tt>fence</tt>' instruction is used to introduce happens-before edges
5094between operations.</p>
5095
5096<h5>Arguments:</h5> <p>'<code>fence</code>' instructions take an <a
5097href="#ordering">ordering</a> argument which defines what
5098<i>synchronizes-with</i> edges they add. They can only be given
5099<code>acquire</code>, <code>release</code>, <code>acq_rel</code>, and
5100<code>seq_cst</code> orderings.</p>
5101
5102<h5>Semantics:</h5>
5103<p>A fence <var>A</var> which has (at least) <code>release</code> ordering
5104semantics <i>synchronizes with</i> a fence <var>B</var> with (at least)
5105<code>acquire</code> ordering semantics if and only if there exist atomic
5106operations <var>X</var> and <var>Y</var>, both operating on some atomic object
5107<var>M</var>, such that <var>A</var> is sequenced before <var>X</var>,
5108<var>X</var> modifies <var>M</var> (either directly or through some side effect
5109of a sequence headed by <var>X</var>), <var>Y</var> is sequenced before
5110<var>B</var>, and <var>Y</var> observes <var>M</var>. This provides a
5111<i>happens-before</i> dependency between <var>A</var> and <var>B</var>. Rather
5112than an explicit <code>fence</code>, one (but not both) of the atomic operations
5113<var>X</var> or <var>Y</var> might provide a <code>release</code> or
5114<code>acquire</code> (resp.) ordering constraint and still
5115<i>synchronize-with</i> the explicit <code>fence</code> and establish the
5116<i>happens-before</i> edge.</p>
5117
5118<p>A <code>fence</code> which has <code>seq_cst</code> ordering, in addition to
5119having both <code>acquire</code> and <code>release</code> semantics specified
5120above, participates in the global program order of other <code>seq_cst</code>
5121operations and/or fences.</p>
5122
5123<p>The optional "<a href="#singlethread"><code>singlethread</code></a>" argument
5124specifies that the fence only synchronizes with other fences in the same
5125thread. (This is useful for interacting with signal handlers.)</p>
5126
Eli Friedman47f35132011-07-25 23:16:38 +0000[diff] [blame]5127<h5>Example:</h5>
5128<pre>
5129 fence acquire <i>; yields {void}</i>
5130 fence singlethread seq_cst <i>; yields {void}</i>
5131</pre>
5132
5133</div>
5134
5135<!-- _______________________________________________________________________ -->
NAKAMURA Takumib5bb29c2011-08-12 06:17:17 +0000[diff] [blame]5136<h4>
5137<a name="i_cmpxchg">'<tt>cmpxchg</tt>' Instruction</a>
5138</h4>
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]5139
NAKAMURA Takumib5bb29c2011-08-12 06:17:17 +0000[diff] [blame]5140<div>
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]5141
5142<h5>Syntax:</h5>
5143<pre>
Bill Wendling262396b2011-12-09 22:41:40 +0000[diff] [blame]5144 cmpxchg [volatile] &lt;ty&gt;* &lt;pointer&gt;, &lt;ty&gt; &lt;cmp&gt;, &lt;ty&gt; &lt;new&gt; [singlethread] &lt;ordering&gt; <i>; yields {ty}</i>
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]5145</pre>
5146
5147<h5>Overview:</h5>
5148<p>The '<tt>cmpxchg</tt>' instruction is used to atomically modify memory.
5149It loads a value in memory and compares it to a given value. If they are
5150equal, it stores a new value into the memory.</p>
5151
5152<h5>Arguments:</h5>
5153<p>There are three arguments to the '<code>cmpxchg</code>' instruction: an
5154address to operate on, a value to compare to the value currently be at that
5155address, and a new value to place at that address if the compared values are
5156equal. The type of '<var>&lt;cmp&gt;</var>' must be an integer type whose
5157bit width is a power of two greater than or equal to eight and less than
5158or equal to a target-specific size limit. '<var>&lt;cmp&gt;</var>' and
5159'<var>&lt;new&gt;</var>' must have the same type, and the type of
5160'<var>&lt;pointer&gt;</var>' must be a pointer to that type. If the
5161<code>cmpxchg</code> is marked as <code>volatile</code>, then the
5162optimizer is not allowed to modify the number or order of execution
5163of this <code>cmpxchg</code> with other <a href="#volatile">volatile
5164operations</a>.</p>
5165
5166<!-- FIXME: Extend allowed types. -->
5167
5168<p>The <a href="#ordering"><var>ordering</var></a> argument specifies how this
5169<code>cmpxchg</code> synchronizes with other atomic operations.</p>
5170
5171<p>The optional "<code>singlethread</code>" argument declares that the
5172<code>cmpxchg</code> is only atomic with respect to code (usually signal
5173handlers) running in the same thread as the <code>cmpxchg</code>. Otherwise the
5174cmpxchg is atomic with respect to all other code in the system.</p>
5175
5176<p>The pointer passed into cmpxchg must have alignment greater than or equal to
5177the size in memory of the operand.
5178
5179<h5>Semantics:</h5>
5180<p>The contents of memory at the location specified by the
5181'<tt>&lt;pointer&gt;</tt>' operand is read and compared to
5182'<tt>&lt;cmp&gt;</tt>'; if the read value is the equal,
5183'<tt>&lt;new&gt;</tt>' is written. The original value at the location
5184is returned.
5185
5186<p>A successful <code>cmpxchg</code> is a read-modify-write instruction for the
5187purpose of identifying <a href="#release_sequence">release sequences</a>. A
5188failed <code>cmpxchg</code> is equivalent to an atomic load with an ordering
5189parameter determined by dropping any <code>release</code> part of the
5190<code>cmpxchg</code>'s ordering.</p>
5191
5192<!--
5193FIXME: Is compare_exchange_weak() necessary? (Consider after we've done
5194optimization work on ARM.)
5195
5196FIXME: Is a weaker ordering constraint on failure helpful in practice?
5197-->
5198
5199<h5>Example:</h5>
5200<pre>
5201entry:
Bill Wendling262396b2011-12-09 22:41:40 +0000[diff] [blame]5202 %orig = atomic <a href="#i_load">load</a> i32* %ptr unordered <i>; yields {i32}</i>
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]5203 <a href="#i_br">br</a> label %loop
5204
5205loop:
5206 %cmp = <a href="#i_phi">phi</a> i32 [ %orig, %entry ], [%old, %loop]
5207 %squared = <a href="#i_mul">mul</a> i32 %cmp, %cmp
Bill Wendling262396b2011-12-09 22:41:40 +0000[diff] [blame]5208 %old = cmpxchg i32* %ptr, i32 %cmp, i32 %squared <i>; yields {i32}</i>
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]5209 %success = <a href="#i_icmp">icmp</a> eq i32 %cmp, %old
5210 <a href="#i_br">br</a> i1 %success, label %done, label %loop
5211
5212done:
5213 ...
5214</pre>
5215
5216</div>
5217
5218<!-- _______________________________________________________________________ -->
NAKAMURA Takumib5bb29c2011-08-12 06:17:17 +0000[diff] [blame]5219<h4>
5220<a name="i_atomicrmw">'<tt>atomicrmw</tt>' Instruction</a>
5221</h4>
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]5222
NAKAMURA Takumib5bb29c2011-08-12 06:17:17 +0000[diff] [blame]5223<div>
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]5224
5225<h5>Syntax:</h5>
5226<pre>
Eli Friedmanf03bb262011-08-12 22:50:01 +0000[diff] [blame]5227 atomicrmw [volatile] &lt;operation&gt; &lt;ty&gt;* &lt;pointer&gt;, &lt;ty&gt; &lt;value&gt; [singlethread] &lt;ordering&gt; <i>; yields {ty}</i>
Eli Friedmanff030482011-07-28 21:48:00 +0000[diff] [blame]5228</pre>
5229
5230<h5>Overview:</h5>
5231<p>The '<tt>atomicrmw</tt>' instruction is used to atomically modify memory.</p>
5232
5233<h5>Arguments:</h5>
5234<p>There are three arguments to the '<code>atomicrmw</code>' instruction: an
5235operation to apply, an address whose value to modify, an argument to the
5236operation. The operation must be one of the following keywords:</p>
5237<ul>
5238 <li>xchg</li>
5239 <li>add</li>
5240 <li>sub</li>
5241 <li>and</li>
5242 <li>nand</li>
5243 <li>or</li>
5244 <li>xor</li>
5245 <li>max</li>
5246 <li>min</li>
5247 <li>umax</li>
5248 <li>umin</li>
5249</ul>
5250
5251<p>The type of '<var>&lt;value&gt;</var>' must be an integer type whose
5252bit width is a power of two greater than or equal to eight and less than
5253or equal to a target-specific size limit. The type of the
5254'<code>&lt;pointer&gt;</code>' operand must be a pointer to that type.
5255If the <code>atomicrmw</code> is marked as <code>volatile</code>, then the
5256optimizer is not allowed to modify the number or order of execution of this
5257<code>atomicrmw</code> with other <a href="#volatile">volatile
5258 operations</a>.</p>
5259
5260<!-- FIXME: Extend allowed types. -->
5261
5262<h5>Semantics:</h5>
5263<p>The contents of memory at the location specified by the
5264'<tt>&lt;pointer&gt;</tt>' operand are atomically read, modified, and written
5265back. The original value at the location is returned. The modification is
5266specified by the <var>operation</var> argument:</p>
5267
5268<ul>
5269 <li>xchg: <code>*ptr = val</code></li>
5270 <li>add: <code>*ptr = *ptr + val</code></li>
5271 <li>sub: <code>*ptr = *ptr - val</code></li>
5272 <li>and: <code>*ptr = *ptr &amp; val</code></li>
5273 <li>nand: <code>*ptr = ~(*ptr &amp; val)</code></li>
5274 <li>or: <code>*ptr = *ptr | val</code></li>
5275 <li>xor: <code>*ptr = *ptr ^ val</code></li>
5276 <li>max: <code>*ptr = *ptr &gt; val ? *ptr : val</code> (using a signed comparison)</li>
5277 <li>min: <code>*ptr = *ptr &lt; val ? *ptr : val</code> (using a signed comparison)</li>
5278 <li>umax: <code>*ptr = *ptr &gt; val ? *ptr : val</code> (using an unsigned comparison)</li>
5279 <li>umin: <code>*ptr = *ptr &lt; val ? *ptr : val</code> (using an unsigned comparison)</li>
5280</ul>
5281
5282<h5>Example:</h5>
5283<pre>
5284 %old = atomicrmw add i32* %ptr, i32 1 acquire <i>; yields {i32}</i>
5285</pre>
5286
5287</div>
5288
5289<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5290<h4>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]5291 <a name="i_getelementptr">'<tt>getelementptr</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5292</h4>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]5293
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5294<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5295
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]5296<h5>Syntax:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]5297<pre>
Matthijs Kooijmane49d0bc2008-10-13 13:44:15 +0000[diff] [blame]5298 &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]5299 &lt;result&gt; = getelementptr inbounds &lt;pty&gt;* &lt;ptrval&gt;{, &lt;ty&gt; &lt;idx&gt;}*
Nadav Rotem16087692011-12-05 06:29:09 +0000[diff] [blame]5300 &lt;result&gt; = getelementptr &lt;ptr vector&gt; ptrval, &lt;vector index type&gt; idx
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]5301</pre>
5302
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]5303<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5304<p>The '<tt>getelementptr</tt>' instruction is used to get the address of a
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]5305 subelement of an <a href="#t_aggregate">aggregate</a> data structure.
5306 It performs address calculation only and does not access memory.</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]5307
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]5308<h5>Arguments:</h5>
Nadav Rotem16087692011-12-05 06:29:09 +0000[diff] [blame]5309<p>The first argument is always a pointer or a vector of pointers,
5310 and forms the basis of the
Chris Lattnerc8eef442009-07-29 06:44:13 +0000[diff] [blame]5311 calculation. The remaining arguments are indices that indicate which of the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5312 elements of the aggregate object are indexed. The interpretation of each
5313 index is dependent on the type being indexed into. The first index always
5314 indexes the pointer value given as the first argument, the second index
5315 indexes a value of the type pointed to (not necessarily the value directly
5316 pointed to, since the first index can be non-zero), etc. The first type
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]5317 indexed into must be a pointer value, subsequent types can be arrays,
Chris Lattner61c70e92010-08-28 04:09:24 +0000[diff] [blame]5318 vectors, and structs. Note that subsequent types being indexed into
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]5319 can never be pointers, since that would require loading the pointer before
5320 continuing calculation.</p>
Matthijs Kooijmane49d0bc2008-10-13 13:44:15 +0000[diff] [blame]5321
5322<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]5323 When indexing into a (optionally packed) structure, only <tt>i32</tt>
Chris Lattnerfdfeb692010-02-12 20:49:41 +0000[diff] [blame]5324 integer <b>constants</b> are allowed. When indexing into an array, pointer
5325 or vector, integers of any width are allowed, and they are not required to be
Eli Friedman266246c2011-08-12 23:37:55 +0000[diff] [blame]5326 constant. These integers are treated as signed values where relevant.</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]5327
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5328<p>For example, let's consider a C code fragment and how it gets compiled to
5329 LLVM:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]5330
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]5331<pre class="doc_code">
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]5332struct RT {
5333 char A;
Chris Lattnercabc8462007-05-29 15:43:56 +0000[diff] [blame]5334 int B[10][20];
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]5335 char C;
5336};
5337struct ST {
Chris Lattnercabc8462007-05-29 15:43:56 +0000[diff] [blame]5338 int X;
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]5339 double Y;
5340 struct RT Z;
5341};
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]5342
Chris Lattnercabc8462007-05-29 15:43:56 +0000[diff] [blame]5343int *foo(struct ST *s) {
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]5344 return &amp;s[1].Z.B[5][13];
5345}
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]5346</pre>
5347
Bill Wendlinga3495392011-12-13 01:07:07 +0000[diff] [blame]5348<p>The LLVM code generated by Clang is:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]5349
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]5350<pre class="doc_code">
Bill Wendlinga3495392011-12-13 01:07:07 +0000[diff] [blame]5351%struct.RT = <a href="#namedtypes">type</a> { i8, [10 x [20 x i32]], i8 }
5352%struct.ST = <a href="#namedtypes">type</a> { i32, double, %struct.RT }
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]5353
Bill Wendlinga3495392011-12-13 01:07:07 +0000[diff] [blame]5354define i32* @foo(%struct.ST* %s) nounwind uwtable readnone optsize ssp {
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]5355entry:
Bill Wendlinga3495392011-12-13 01:07:07 +0000[diff] [blame]5356 %arrayidx = getelementptr inbounds %struct.ST* %s, i64 1, i32 2, i32 1, i64 5, i64 13
5357 ret i32* %arrayidx
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]5358}
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]5359</pre>
5360
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]5361<h5>Semantics:</h5>
Bill Wendlinga3495392011-12-13 01:07:07 +0000[diff] [blame]5362<p>In the example above, the first index is indexing into the
5363 '<tt>%struct.ST*</tt>' type, which is a pointer, yielding a
5364 '<tt>%struct.ST</tt>' = '<tt>{ i32, double, %struct.RT }</tt>' type, a
5365 structure. The second index indexes into the third element of the structure,
5366 yielding a '<tt>%struct.RT</tt>' = '<tt>{ i8 , [10 x [20 x i32]], i8 }</tt>'
5367 type, another structure. The third index indexes into the second element of
5368 the structure, yielding a '<tt>[10 x [20 x i32]]</tt>' type, an array. The
5369 two dimensions of the array are subscripted into, yielding an '<tt>i32</tt>'
5370 type. The '<tt>getelementptr</tt>' instruction returns a pointer to this
5371 element, thus computing a value of '<tt>i32*</tt>' type.</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]5372
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5373<p>Note that it is perfectly legal to index partially through a structure,
5374 returning a pointer to an inner element. Because of this, the LLVM code for
5375 the given testcase is equivalent to:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]5376
Bill Wendlinga3495392011-12-13 01:07:07 +0000[diff] [blame]5377<pre class="doc_code">
5378define i32* @foo(%struct.ST* %s) {
5379 %t1 = getelementptr %struct.ST* %s, i32 1 <i>; yields %struct.ST*:%t1</i>
5380 %t2 = getelementptr %struct.ST* %t1, i32 0, i32 2 <i>; yields %struct.RT*:%t2</i>
5381 %t3 = getelementptr %struct.RT* %t2, i32 0, i32 1 <i>; yields [10 x [20 x i32]]*:%t3</i>
5382 %t4 = getelementptr [10 x [20 x i32]]* %t3, i32 0, i32 5 <i>; yields [20 x i32]*:%t4</i>
5383 %t5 = getelementptr [20 x i32]* %t4, i32 0, i32 13 <i>; yields i32*:%t5</i>
5384 ret i32* %t5
5385}
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]5386</pre>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]5387
Dan Gohmandd8004d2009-07-27 21:53:46 +0000[diff] [blame]5388<p>If the <tt>inbounds</tt> keyword is present, the result value of the
Dan Gohmanbfb056d2011-12-06 03:18:47 +0000[diff] [blame]5389 <tt>getelementptr</tt> is a <a href="#poisonvalues">poison value</a> if the
Dan Gohman27ef9972010-04-23 15:23:32 +0000[diff] [blame]5390 base pointer is not an <i>in bounds</i> address of an allocated object,
5391 or if any of the addresses that would be formed by successive addition of
5392 the offsets implied by the indices to the base address with infinitely
Eli Friedman266246c2011-08-12 23:37:55 +0000[diff] [blame]5393 precise signed arithmetic are not an <i>in bounds</i> address of that
5394 allocated object. The <i>in bounds</i> addresses for an allocated object
5395 are all the addresses that point into the object, plus the address one
Nadav Rotem16087692011-12-05 06:29:09 +0000[diff] [blame]5396 byte past the end.
5397 In cases where the base is a vector of pointers the <tt>inbounds</tt> keyword
5398 applies to each of the computations element-wise. </p>
Dan Gohmandd8004d2009-07-27 21:53:46 +0000[diff] [blame]5399
5400<p>If the <tt>inbounds</tt> keyword is not present, the offsets are added to
Eli Friedman266246c2011-08-12 23:37:55 +0000[diff] [blame]5401 the base address with silently-wrapping two's complement arithmetic. If the
5402 offsets have a different width from the pointer, they are sign-extended or
5403 truncated to the width of the pointer. The result value of the
5404 <tt>getelementptr</tt> may be outside the object pointed to by the base
5405 pointer. The result value may not necessarily be used to access memory
5406 though, even if it happens to point into allocated storage. See the
5407 <a href="#pointeraliasing">Pointer Aliasing Rules</a> section for more
5408 information.</p>
Dan Gohmandd8004d2009-07-27 21:53:46 +0000[diff] [blame]5409
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5410<p>The getelementptr instruction is often confusing. For some more insight into
5411 how it works, see <a href="GetElementPtr.html">the getelementptr FAQ</a>.</p>
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]5412
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]5413<h5>Example:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]5414<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5415 <i>; yields [12 x i8]*:aptr</i>
Matthijs Kooijmane49d0bc2008-10-13 13:44:15 +0000[diff] [blame]5416 %aptr = getelementptr {i32, [12 x i8]}* %saptr, i64 0, i32 1
5417 <i>; yields i8*:vptr</i>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5418 %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]5419 <i>; yields i8*:eptr</i>
5420 %eptr = getelementptr [12 x i8]* %aptr, i64 0, i32 1
Sanjiv Gupta9f805c22009-04-25 07:27:44 +0000[diff] [blame]5421 <i>; yields i32*:iptr</i>
Sanjiv Gupta16ffa802009-04-24 16:38:13 +0000[diff] [blame]5422 %iptr = getelementptr [10 x i32]* @arr, i16 0, i16 0
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]5423</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5424
Nadav Rotem16087692011-12-05 06:29:09 +0000[diff] [blame]5425<p>In cases where the pointer argument is a vector of pointers, only a
5426 single index may be used, and the number of vector elements has to be
5427 the same. For example: </p>
5428<pre class="doc_code">
5429 %A = getelementptr <4 x i8*> %ptrs, <4 x i64> %offsets,
5430</pre>
5431
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]5432</div>
Reid Spencer47ce1792006-11-09 21:15:49 +0000[diff] [blame]5433
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5434</div>
5435
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]5436<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5437<h3>
5438 <a name="convertops">Conversion Operations</a>
5439</h3>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5440
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5441<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5442
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5443<p>The instructions in this category are the conversion instructions (casting)
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5444 which all take a single operand and a type. They perform various bit
5445 conversions on the operand.</p>
5446
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]5447<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5448<h4>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5449 <a name="i_trunc">'<tt>trunc .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5450</h4>
5451
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5452<div>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5453
5454<h5>Syntax:</h5>
5455<pre>
5456 &lt;result&gt; = trunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
5457</pre>
5458
5459<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5460<p>The '<tt>trunc</tt>' instruction truncates its operand to the
5461 type <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5462
5463<h5>Arguments:</h5>
Nadav Rotem8c20ec52011-02-24 21:01:34 +0000[diff] [blame]5464<p>The '<tt>trunc</tt>' instruction takes a value to trunc, and a type to trunc it to.
5465 Both types must be of <a href="#t_integer">integer</a> types, or vectors
5466 of the same number of integers.
5467 The bit size of the <tt>value</tt> must be larger than
5468 the bit size of the destination type, <tt>ty2</tt>.
5469 Equal sized types are not allowed.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5470
5471<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5472<p>The '<tt>trunc</tt>' instruction truncates the high order bits
5473 in <tt>value</tt> and converts the remaining bits to <tt>ty2</tt>. Since the
5474 source size must be larger than the destination size, <tt>trunc</tt> cannot
5475 be a <i>no-op cast</i>. It will always truncate bits.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5476
5477<h5>Example:</h5>
5478<pre>
Nadav Rotem8c20ec52011-02-24 21:01:34 +0000[diff] [blame]5479 %X = trunc i32 257 to i8 <i>; yields i8:1</i>
5480 %Y = trunc i32 123 to i1 <i>; yields i1:true</i>
5481 %Z = trunc i32 122 to i1 <i>; yields i1:false</i>
5482 %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]5483</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5484
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5485</div>
5486
5487<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5488<h4>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5489 <a name="i_zext">'<tt>zext .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5490</h4>
5491
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5492<div>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5493
5494<h5>Syntax:</h5>
5495<pre>
5496 &lt;result&gt; = zext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
5497</pre>
5498
5499<h5>Overview:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5500<p>The '<tt>zext</tt>' instruction zero extends its operand to type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5501 <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5502
5503
5504<h5>Arguments:</h5>
Nadav Rotemed9b9342011-02-20 12:37:50 +0000[diff] [blame]5505<p>The '<tt>zext</tt>' instruction takes a value to cast, and a type to cast it to.
5506 Both types must be of <a href="#t_integer">integer</a> types, or vectors
5507 of the same number of integers.
5508 The bit size of the <tt>value</tt> must be smaller than
5509 the bit size of the destination type,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5510 <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5511
5512<h5>Semantics:</h5>
5513<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]5514 bits until it reaches the size of the destination type, <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5515
Reid Spencerb5929522007-01-12 15:46:11 +0000[diff] [blame]5516<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]5517
5518<h5>Example:</h5>
5519<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5520 %X = zext i32 257 to i64 <i>; yields i64:257</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]5521 %Y = zext i1 true to i32 <i>; yields i32:1</i>
Nadav Rotemed9b9342011-02-20 12:37:50 +0000[diff] [blame]5522 %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]5523</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5524
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5525</div>
5526
5527<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5528<h4>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5529 <a name="i_sext">'<tt>sext .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5530</h4>
5531
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5532<div>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5533
5534<h5>Syntax:</h5>
5535<pre>
5536 &lt;result&gt; = sext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
5537</pre>
5538
5539<h5>Overview:</h5>
5540<p>The '<tt>sext</tt>' sign extends <tt>value</tt> to the type <tt>ty2</tt>.</p>
5541
5542<h5>Arguments:</h5>
Nadav Rotem8c20ec52011-02-24 21:01:34 +0000[diff] [blame]5543<p>The '<tt>sext</tt>' instruction takes a value to cast, and a type to cast it to.
5544 Both types must be of <a href="#t_integer">integer</a> types, or vectors
5545 of the same number of integers.
5546 The bit size of the <tt>value</tt> must be smaller than
5547 the bit size of the destination type,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5548 <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5549
5550<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5551<p>The '<tt>sext</tt>' instruction performs a sign extension by copying the sign
5552 bit (highest order bit) of the <tt>value</tt> until it reaches the bit size
5553 of the type <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5554
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]5555<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]5556
5557<h5>Example:</h5>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5558<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5559 %X = sext i8 -1 to i16 <i>; yields i16 :65535</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]5560 %Y = sext i1 true to i32 <i>; yields i32:-1</i>
Nadav Rotem8c20ec52011-02-24 21:01:34 +0000[diff] [blame]5561 %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]5562</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5563
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5564</div>
5565
5566<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5567<h4>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]5568 <a name="i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5569</h4>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]5570
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5571<div>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]5572
5573<h5>Syntax:</h5>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]5574<pre>
5575 &lt;result&gt; = fptrunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
5576</pre>
5577
5578<h5>Overview:</h5>
5579<p>The '<tt>fptrunc</tt>' instruction truncates <tt>value</tt> to type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5580 <tt>ty2</tt>.</p>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]5581
5582<h5>Arguments:</h5>
5583<p>The '<tt>fptrunc</tt>' instruction takes a <a href="#t_floating">floating
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5584 point</a> value to cast and a <a href="#t_floating">floating point</a> type
5585 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]5586 <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]5587 <i>no-op cast</i>.</p>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]5588
5589<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5590<p>The '<tt>fptrunc</tt>' instruction truncates a <tt>value</tt> from a larger
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5591 <a href="#t_floating">floating point</a> type to a smaller
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5592 <a href="#t_floating">floating point</a> type. If the value cannot fit
5593 within the destination type, <tt>ty2</tt>, then the results are
5594 undefined.</p>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]5595
5596<h5>Example:</h5>
5597<pre>
5598 %X = fptrunc double 123.0 to float <i>; yields float:123.0</i>
5599 %Y = fptrunc double 1.0E+300 to float <i>; yields undefined</i>
5600</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5601
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]5602</div>
5603
5604<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5605<h4>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5606 <a name="i_fpext">'<tt>fpext .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5607</h4>
5608
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5609<div>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5610
5611<h5>Syntax:</h5>
5612<pre>
5613 &lt;result&gt; = fpext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
5614</pre>
5615
5616<h5>Overview:</h5>
5617<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]5618 floating point value.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5619
5620<h5>Arguments:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5621<p>The '<tt>fpext</tt>' instruction takes a
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5622 <a href="#t_floating">floating point</a> <tt>value</tt> to cast, and
5623 a <a href="#t_floating">floating point</a> type to cast it to. The source
5624 type must be smaller than the destination type.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5625
5626<h5>Semantics:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]5627<p>The '<tt>fpext</tt>' instruction extends the <tt>value</tt> from a smaller
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5628 <a href="#t_floating">floating point</a> type to a larger
5629 <a href="#t_floating">floating point</a> type. The <tt>fpext</tt> cannot be
5630 used to make a <i>no-op cast</i> because it always changes bits. Use
5631 <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]5632
5633<h5>Example:</h5>
5634<pre>
Nick Lewycky5bb3ece2011-03-31 18:20:19 +0000[diff] [blame]5635 %X = fpext float 3.125 to double <i>; yields double:3.125000e+00</i>
5636 %Y = fpext double %X to fp128 <i>; yields fp128:0xL00000000000000004000900000000000</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5637</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5638
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5639</div>
5640
5641<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5642<h4>
Reid Spencer24d6da52007-01-21 00:29:26 +0000[diff] [blame]5643 <a name="i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5644</h4>
5645
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5646<div>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5647
5648<h5>Syntax:</h5>
5649<pre>
Reid Spencer1539a1c2007-07-31 14:40:14 +0000[diff] [blame]5650 &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]5651</pre>
5652
5653<h5>Overview:</h5>
Reid Spencer1539a1c2007-07-31 14:40:14 +0000[diff] [blame]5654<p>The '<tt>fptoui</tt>' converts a floating point <tt>value</tt> to its
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5655 unsigned integer equivalent of type <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5656
5657<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5658<p>The '<tt>fptoui</tt>' instruction takes a value to cast, which must be a
5659 scalar or vector <a href="#t_floating">floating point</a> value, and a type
5660 to cast it to <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a>
5661 type. If <tt>ty</tt> is a vector floating point type, <tt>ty2</tt> must be a
5662 vector integer type with the same number of elements as <tt>ty</tt></p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5663
5664<h5>Semantics:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5665<p>The '<tt>fptoui</tt>' instruction converts its
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5666 <a href="#t_floating">floating point</a> operand into the nearest (rounding
5667 towards zero) unsigned integer value. If the value cannot fit
5668 in <tt>ty2</tt>, the results are undefined.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5669
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5670<h5>Example:</h5>
5671<pre>
Reid Spencer1539a1c2007-07-31 14:40:14 +0000[diff] [blame]5672 %X = fptoui double 123.0 to i32 <i>; yields i32:123</i>
Chris Lattner88519042007-09-22 03:17:52 +0000[diff] [blame]5673 %Y = fptoui float 1.0E+300 to i1 <i>; yields undefined:1</i>
Gabor Greif6a292012009-10-28 09:21:30 +0000[diff] [blame]5674 %Z = fptoui float 1.04E+17 to i8 <i>; yields undefined:1</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5675</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5676
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5677</div>
5678
5679<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5680<h4>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]5681 <a name="i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5682</h4>
5683
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5684<div>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5685
5686<h5>Syntax:</h5>
5687<pre>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]5688 &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]5689</pre>
5690
5691<h5>Overview:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5692<p>The '<tt>fptosi</tt>' instruction converts
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5693 <a href="#t_floating">floating point</a> <tt>value</tt> to
5694 type <tt>ty2</tt>.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5695
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]5696<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5697<p>The '<tt>fptosi</tt>' instruction takes a value to cast, which must be a
5698 scalar or vector <a href="#t_floating">floating point</a> value, and a type
5699 to cast it to <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a>
5700 type. If <tt>ty</tt> is a vector floating point type, <tt>ty2</tt> must be a
5701 vector integer type with the same number of elements as <tt>ty</tt></p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5702
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]5703<h5>Semantics:</h5>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5704<p>The '<tt>fptosi</tt>' instruction converts its
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5705 <a href="#t_floating">floating point</a> operand into the nearest (rounding
5706 towards zero) signed integer value. If the value cannot fit in <tt>ty2</tt>,
5707 the results are undefined.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5708
Chris Lattner33ba0d92001-07-09 00:26:23 +0000[diff] [blame]5709<h5>Example:</h5>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5710<pre>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]5711 %X = fptosi double -123.0 to i32 <i>; yields i32:-123</i>
Chris Lattner88519042007-09-22 03:17:52 +0000[diff] [blame]5712 %Y = fptosi float 1.0E-247 to i1 <i>; yields undefined:1</i>
Gabor Greif6a292012009-10-28 09:21:30 +0000[diff] [blame]5713 %Z = fptosi float 1.04E+17 to i8 <i>; yields undefined:1</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5714</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5715
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5716</div>
5717
5718<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5719<h4>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]5720 <a name="i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5721</h4>
5722
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5723<div>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5724
5725<h5>Syntax:</h5>
5726<pre>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]5727 &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]5728</pre>
5729
5730<h5>Overview:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]5731<p>The '<tt>uitofp</tt>' instruction regards <tt>value</tt> as an unsigned
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5732 integer and converts that value to the <tt>ty2</tt> type.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5733
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5734<h5>Arguments:</h5>
Nate Begemanb348d182007-11-17 03:58:34 +0000[diff] [blame]5735<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]5736 scalar or vector <a href="#t_integer">integer</a> value, and a type to cast
5737 it to <tt>ty2</tt>, which must be an <a href="#t_floating">floating point</a>
5738 type. If <tt>ty</tt> is a vector integer type, <tt>ty2</tt> must be a vector
5739 floating point type with the same number of elements as <tt>ty</tt></p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5740
5741<h5>Semantics:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]5742<p>The '<tt>uitofp</tt>' instruction interprets its operand as an unsigned
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5743 integer quantity and converts it to the corresponding floating point
5744 value. If the value cannot fit in the floating point value, the results are
5745 undefined.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5746
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5747<h5>Example:</h5>
5748<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5749 %X = uitofp i32 257 to float <i>; yields float:257.0</i>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5750 %Y = uitofp i8 -1 to double <i>; yields double:255.0</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5751</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5752
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5753</div>
5754
5755<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5756<h4>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]5757 <a name="i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5758</h4>
5759
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5760<div>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5761
5762<h5>Syntax:</h5>
5763<pre>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]5764 &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]5765</pre>
5766
5767<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5768<p>The '<tt>sitofp</tt>' instruction regards <tt>value</tt> as a signed integer
5769 and converts that value to the <tt>ty2</tt> type.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5770
5771<h5>Arguments:</h5>
Nate Begemanb348d182007-11-17 03:58:34 +0000[diff] [blame]5772<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]5773 scalar or vector <a href="#t_integer">integer</a> value, and a type to cast
5774 it to <tt>ty2</tt>, which must be an <a href="#t_floating">floating point</a>
5775 type. If <tt>ty</tt> is a vector integer type, <tt>ty2</tt> must be a vector
5776 floating point type with the same number of elements as <tt>ty</tt></p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5777
5778<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5779<p>The '<tt>sitofp</tt>' instruction interprets its operand as a signed integer
5780 quantity and converts it to the corresponding floating point value. If the
5781 value cannot fit in the floating point value, the results are undefined.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5782
5783<h5>Example:</h5>
5784<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5785 %X = sitofp i32 257 to float <i>; yields float:257.0</i>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5786 %Y = sitofp i8 -1 to double <i>; yields double:-1.0</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5787</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5788
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5789</div>
5790
5791<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5792<h4>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5793 <a name="i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5794</h4>
5795
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5796<div>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5797
5798<h5>Syntax:</h5>
5799<pre>
5800 &lt;result&gt; = ptrtoint &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
5801</pre>
5802
5803<h5>Overview:</h5>
Nadav Rotem16087692011-12-05 06:29:09 +0000[diff] [blame]5804<p>The '<tt>ptrtoint</tt>' instruction converts the pointer or a vector of
5805 pointers <tt>value</tt> to
5806 the integer (or vector of integers) type <tt>ty2</tt>.</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5807
5808<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5809<p>The '<tt>ptrtoint</tt>' instruction takes a <tt>value</tt> to cast, which
Nadav Rotem16087692011-12-05 06:29:09 +0000[diff] [blame]5810 must be a a value of type <a href="#t_pointer">pointer</a> or a vector of
5811 pointers, and a type to cast it to
5812 <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a> or a vector
5813 of integers type.</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5814
5815<h5>Semantics:</h5>
5816<p>The '<tt>ptrtoint</tt>' instruction converts <tt>value</tt> to integer type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5817 <tt>ty2</tt> by interpreting the pointer value as an integer and either
5818 truncating or zero extending that value to the size of the integer type. If
5819 <tt>value</tt> is smaller than <tt>ty2</tt> then a zero extension is done. If
5820 <tt>value</tt> is larger than <tt>ty2</tt> then a truncation is done. If they
5821 are the same size, then nothing is done (<i>no-op cast</i>) other than a type
5822 change.</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5823
5824<h5>Example:</h5>
5825<pre>
Nadav Rotem16087692011-12-05 06:29:09 +0000[diff] [blame]5826 %X = ptrtoint i32* %P to i8 <i>; yields truncation on 32-bit architecture</i>
5827 %Y = ptrtoint i32* %P to i64 <i>; yields zero extension on 32-bit architecture</i>
5828 %Z = ptrtoint &lt;4 x i32*&gt; %P to &lt;4 x i64&gt;<i>; yields vector zero extension for a vector of addresses on 32-bit architecture</i>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5829</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5830
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5831</div>
5832
5833<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5834<h4>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5835 <a name="i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5836</h4>
5837
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5838<div>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5839
5840<h5>Syntax:</h5>
5841<pre>
5842 &lt;result&gt; = inttoptr &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
5843</pre>
5844
5845<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5846<p>The '<tt>inttoptr</tt>' instruction converts an integer <tt>value</tt> to a
5847 pointer type, <tt>ty2</tt>.</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5848
5849<h5>Arguments:</h5>
Duncan Sands8036ca42007-03-30 12:22:09 +0000[diff] [blame]5850<p>The '<tt>inttoptr</tt>' instruction takes an <a href="#t_integer">integer</a>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5851 value to cast, and a type to cast it to, which must be a
5852 <a href="#t_pointer">pointer</a> type.</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5853
5854<h5>Semantics:</h5>
5855<p>The '<tt>inttoptr</tt>' instruction converts <tt>value</tt> to type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5856 <tt>ty2</tt> by applying either a zero extension or a truncation depending on
5857 the size of the integer <tt>value</tt>. If <tt>value</tt> is larger than the
5858 size of a pointer then a truncation is done. If <tt>value</tt> is smaller
5859 than the size of a pointer then a zero extension is done. If they are the
5860 same size, nothing is done (<i>no-op cast</i>).</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5861
5862<h5>Example:</h5>
5863<pre>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5864 %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]5865 %Y = inttoptr i32 255 to i32* <i>; yields no-op on 32-bit architecture</i>
5866 %Z = inttoptr i64 0 to i32* <i>; yields truncation on 32-bit architecture</i>
Nadav Rotem16087692011-12-05 06:29:09 +0000[diff] [blame]5867 %Z = inttoptr &lt;4 x i32&gt; %G to &lt;4 x i8*&gt;<i>; yields truncation of vector G to four pointers</i>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5868</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5869
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]5870</div>
5871
5872<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5873<h4>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]5874 <a name="i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5875</h4>
5876
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5877<div>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5878
5879<h5>Syntax:</h5>
5880<pre>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]5881 &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]5882</pre>
5883
5884<h5>Overview:</h5>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]5885<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5886 <tt>ty2</tt> without changing any bits.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5887
5888<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5889<p>The '<tt>bitcast</tt>' instruction takes a value to cast, which must be a
5890 non-aggregate first class value, and a type to cast it to, which must also be
5891 a non-aggregate <a href="#t_firstclass">first class</a> type. The bit sizes
5892 of <tt>value</tt> and the destination type, <tt>ty2</tt>, must be
5893 identical. If the source type is a pointer, the destination type must also be
5894 a pointer. This instruction supports bitwise conversion of vectors to
5895 integers and to vectors of other types (as long as they have the same
5896 size).</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5897
5898<h5>Semantics:</h5>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]5899<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5900 <tt>ty2</tt>. It is always a <i>no-op cast</i> because no bits change with
5901 this conversion. The conversion is done as if the <tt>value</tt> had been
Nadav Rotem16087692011-12-05 06:29:09 +0000[diff] [blame]5902 stored to memory and read back as type <tt>ty2</tt>.
5903 Pointer (or vector of pointers) types may only be converted to other pointer
5904 (or vector of pointers) types with this instruction. To convert
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5905 pointers to other types, use the <a href="#i_inttoptr">inttoptr</a> or
5906 <a href="#i_ptrtoint">ptrtoint</a> instructions first.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]5907
5908<h5>Example:</h5>
5909<pre>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]5910 %X = bitcast i8 255 to i8 <i>; yields i8 :-1</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5911 %Y = bitcast i32* %x to sint* <i>; yields sint*:%x</i>
Nadav Rotem16087692011-12-05 06:29:09 +0000[diff] [blame]5912 %Z = bitcast &lt;2 x int&gt; %V to i64; <i>; yields i64: %V</i>
5913 %Z = bitcast &lt;2 x i32*&gt; %V to &lt;2 x i64*&gt; <i>; yields &lt;2 x i64*&gt;</i>
Chris Lattner33ba0d92001-07-09 00:26:23 +0000[diff] [blame]5914</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5915
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]5916</div>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]5917
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5918</div>
5919
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]5920<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5921<h3>
5922 <a name="otherops">Other Operations</a>
5923</h3>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5924
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]5925<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5926
5927<p>The instructions in this category are the "miscellaneous" instructions, which
5928 defy better classification.</p>
5929
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5930<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]5931<h4>
5932 <a name="i_icmp">'<tt>icmp</tt>' Instruction</a>
5933</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +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
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5937<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5938<pre>
5939 &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]5940</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5941
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5942<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5943<p>The '<tt>icmp</tt>' instruction returns a boolean value or a vector of
Nadav Rotem16087692011-12-05 06:29:09 +0000[diff] [blame]5944 boolean values based on comparison of its two integer, integer vector,
5945 pointer, or pointer vector operands.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5946
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5947<h5>Arguments:</h5>
5948<p>The '<tt>icmp</tt>' instruction takes three operands. The first operand is
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5949 the condition code indicating the kind of comparison to perform. It is not a
5950 value, just a keyword. The possible condition code are:</p>
5951
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5952<ol>
5953 <li><tt>eq</tt>: equal</li>
5954 <li><tt>ne</tt>: not equal </li>
5955 <li><tt>ugt</tt>: unsigned greater than</li>
5956 <li><tt>uge</tt>: unsigned greater or equal</li>
5957 <li><tt>ult</tt>: unsigned less than</li>
5958 <li><tt>ule</tt>: unsigned less or equal</li>
5959 <li><tt>sgt</tt>: signed greater than</li>
5960 <li><tt>sge</tt>: signed greater or equal</li>
5961 <li><tt>slt</tt>: signed less than</li>
5962 <li><tt>sle</tt>: signed less or equal</li>
5963</ol>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5964
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]5965<p>The remaining two arguments must be <a href="#t_integer">integer</a> or
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5966 <a href="#t_pointer">pointer</a> or integer <a href="#t_vector">vector</a>
5967 typed. They must also be identical types.</p>
5968
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5969<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5970<p>The '<tt>icmp</tt>' compares <tt>op1</tt> and <tt>op2</tt> according to the
5971 condition code given as <tt>cond</tt>. The comparison performed always yields
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]5972 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]5973 result, as follows:</p>
5974
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5975<ol>
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5976 <li><tt>eq</tt>: yields <tt>true</tt> if the operands are equal,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5977 <tt>false</tt> otherwise. No sign interpretation is necessary or
5978 performed.</li>
5979
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]5980 <li><tt>ne</tt>: yields <tt>true</tt> if the operands are unequal,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5981 <tt>false</tt> otherwise. No sign interpretation is necessary or
5982 performed.</li>
5983
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5984 <li><tt>ugt</tt>: interprets the operands as unsigned values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5985 <tt>true</tt> if <tt>op1</tt> is greater than <tt>op2</tt>.</li>
5986
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5987 <li><tt>uge</tt>: interprets the operands as unsigned values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5988 <tt>true</tt> if <tt>op1</tt> is greater than or equal
5989 to <tt>op2</tt>.</li>
5990
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5991 <li><tt>ult</tt>: interprets the operands as unsigned values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5992 <tt>true</tt> if <tt>op1</tt> is less than <tt>op2</tt>.</li>
5993
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5994 <li><tt>ule</tt>: interprets the operands as unsigned values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5995 <tt>true</tt> if <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
5996
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]5997 <li><tt>sgt</tt>: interprets the operands as signed values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]5998 <tt>true</tt> if <tt>op1</tt> is greater than <tt>op2</tt>.</li>
5999
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]6000 <li><tt>sge</tt>: interprets the operands as signed values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6001 <tt>true</tt> if <tt>op1</tt> is greater than or equal
6002 to <tt>op2</tt>.</li>
6003
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]6004 <li><tt>slt</tt>: interprets the operands as signed values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6005 <tt>true</tt> if <tt>op1</tt> is less than <tt>op2</tt>.</li>
6006
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]6007 <li><tt>sle</tt>: interprets the operands as signed values and yields
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6008 <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]6009</ol>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6010
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]6011<p>If the operands are <a href="#t_pointer">pointer</a> typed, the pointer
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6012 values are compared as if they were integers.</p>
6013
6014<p>If the operands are integer vectors, then they are compared element by
6015 element. The result is an <tt>i1</tt> vector with the same number of elements
6016 as the values being compared. Otherwise, the result is an <tt>i1</tt>.</p>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]6017
6018<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6019<pre>
6020 &lt;result&gt; = icmp eq i32 4, 5 <i>; yields: result=false</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]6021 &lt;result&gt; = icmp ne float* %X, %X <i>; yields: result=false</i>
6022 &lt;result&gt; = icmp ult i16 4, 5 <i>; yields: result=true</i>
6023 &lt;result&gt; = icmp sgt i16 4, 5 <i>; yields: result=false</i>
6024 &lt;result&gt; = icmp ule i16 -4, 5 <i>; yields: result=false</i>
6025 &lt;result&gt; = icmp sge i16 4, 5 <i>; yields: result=false</i>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]6026</pre>
Dan Gohmanc766f722009-01-22 01:39:38 +0000[diff] [blame]6027
6028<p>Note that the code generator does not yet support vector types with
6029 the <tt>icmp</tt> instruction.</p>
6030
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]6031</div>
6032
6033<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6034<h4>
6035 <a name="i_fcmp">'<tt>fcmp</tt>' Instruction</a>
6036</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6037
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6038<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6039
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]6040<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6041<pre>
6042 &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]6043</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6044
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]6045<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6046<p>The '<tt>fcmp</tt>' instruction returns a boolean value or vector of boolean
6047 values based on comparison of its operands.</p>
6048
6049<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]6050(<a href="#t_integer"><tt>i1</tt></a>).</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6051
6052<p>If the operands are floating point vectors, then the result type is a vector
6053 of boolean with the same number of elements as the operands being
6054 compared.</p>
6055
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]6056<h5>Arguments:</h5>
6057<p>The '<tt>fcmp</tt>' instruction takes three operands. The first operand is
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6058 the condition code indicating the kind of comparison to perform. It is not a
6059 value, just a keyword. The possible condition code are:</p>
6060
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]6061<ol>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]6062 <li><tt>false</tt>: no comparison, always returns false</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]6063 <li><tt>oeq</tt>: ordered and equal</li>
6064 <li><tt>ogt</tt>: ordered and greater than </li>
6065 <li><tt>oge</tt>: ordered and greater than or equal</li>
6066 <li><tt>olt</tt>: ordered and less than </li>
6067 <li><tt>ole</tt>: ordered and less than or equal</li>
6068 <li><tt>one</tt>: ordered and not equal</li>
6069 <li><tt>ord</tt>: ordered (no nans)</li>
6070 <li><tt>ueq</tt>: unordered or equal</li>
6071 <li><tt>ugt</tt>: unordered or greater than </li>
6072 <li><tt>uge</tt>: unordered or greater than or equal</li>
6073 <li><tt>ult</tt>: unordered or less than </li>
6074 <li><tt>ule</tt>: unordered or less than or equal</li>
6075 <li><tt>une</tt>: unordered or not equal</li>
6076 <li><tt>uno</tt>: unordered (either nans)</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]6077 <li><tt>true</tt>: no comparison, always returns true</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]6078</ol>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6079
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]6080<p><i>Ordered</i> means that neither operand is a QNAN while
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6081 <i>unordered</i> means that either operand may be a QNAN.</p>
6082
6083<p>Each of <tt>val1</tt> and <tt>val2</tt> arguments must be either
6084 a <a href="#t_floating">floating point</a> type or
6085 a <a href="#t_vector">vector</a> of floating point type. They must have
6086 identical types.</p>
6087
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]6088<h5>Semantics:</h5>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]6089<p>The '<tt>fcmp</tt>' instruction compares <tt>op1</tt> and <tt>op2</tt>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6090 according to the condition code given as <tt>cond</tt>. If the operands are
6091 vectors, then the vectors are compared element by element. Each comparison
Nick Lewyckyec38da42009-09-27 00:45:11 +0000[diff] [blame]6092 performed always yields an <a href="#t_integer">i1</a> result, as
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6093 follows:</p>
6094
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]6095<ol>
6096 <li><tt>false</tt>: always yields <tt>false</tt>, regardless of operands.</li>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6097
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]6098 <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]6099 <tt>op1</tt> is equal to <tt>op2</tt>.</li>
6100
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]6101 <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]6102 <tt>op1</tt> is greater than <tt>op2</tt>.</li>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6103
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]6104 <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]6105 <tt>op1</tt> is greater than or equal to <tt>op2</tt>.</li>
6106
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]6107 <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]6108 <tt>op1</tt> is less than <tt>op2</tt>.</li>
6109
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]6110 <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]6111 <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
6112
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]6113 <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]6114 <tt>op1</tt> is not equal to <tt>op2</tt>.</li>
6115
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]6116 <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]6117
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]6118 <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]6119 <tt>op1</tt> is equal to <tt>op2</tt>.</li>
6120
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]6121 <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]6122 <tt>op1</tt> is greater than <tt>op2</tt>.</li>
6123
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]6124 <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]6125 <tt>op1</tt> is greater than or equal to <tt>op2</tt>.</li>
6126
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]6127 <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]6128 <tt>op1</tt> is less than <tt>op2</tt>.</li>
6129
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]6130 <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]6131 <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
6132
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]6133 <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]6134 <tt>op1</tt> is not equal to <tt>op2</tt>.</li>
6135
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]6136 <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]6137
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]6138 <li><tt>true</tt>: always yields <tt>true</tt>, regardless of operands.</li>
6139</ol>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]6140
6141<h5>Example:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6142<pre>
6143 &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]6144 &lt;result&gt; = fcmp one float 4.0, 5.0 <i>; yields: result=true</i>
6145 &lt;result&gt; = fcmp olt float 4.0, 5.0 <i>; yields: result=true</i>
6146 &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]6147</pre>
Dan Gohmanc766f722009-01-22 01:39:38 +0000[diff] [blame]6148
6149<p>Note that the code generator does not yet support vector types with
6150 the <tt>fcmp</tt> instruction.</p>
6151
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]6152</div>
6153
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]6154<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6155<h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]6156 <a name="i_phi">'<tt>phi</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6157</h4>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]6158
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6159<div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]6160
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]6161<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6162<pre>
6163 &lt;result&gt; = phi &lt;ty&gt; [ &lt;val0&gt;, &lt;label0&gt;], ...
6164</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]6165
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]6166<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6167<p>The '<tt>phi</tt>' instruction is used to implement the &#966; node in the
6168 SSA graph representing the function.</p>
6169
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]6170<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6171<p>The type of the incoming values is specified with the first type field. After
6172 this, the '<tt>phi</tt>' instruction takes a list of pairs as arguments, with
6173 one pair for each predecessor basic block of the current block. Only values
6174 of <a href="#t_firstclass">first class</a> type may be used as the value
6175 arguments to the PHI node. Only labels may be used as the label
6176 arguments.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]6177
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6178<p>There must be no non-phi instructions between the start of a basic block and
6179 the PHI instructions: i.e. PHI instructions must be first in a basic
6180 block.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]6181
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6182<p>For the purposes of the SSA form, the use of each incoming value is deemed to
6183 occur on the edge from the corresponding predecessor block to the current
6184 block (but after any definition of an '<tt>invoke</tt>' instruction's return
6185 value on the same edge).</p>
Jay Foadd2449092009-06-03 10:20:10 +0000[diff] [blame]6186
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]6187<h5>Semantics:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]6188<p>At runtime, the '<tt>phi</tt>' instruction logically takes on the value
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6189 specified by the pair corresponding to the predecessor basic block that
6190 executed just prior to the current block.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]6191
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]6192<h5>Example:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]6193<pre>
6194Loop: ; Infinite loop that counts from 0 on up...
6195 %indvar = phi i32 [ 0, %LoopHeader ], [ %nextindvar, %Loop ]
6196 %nextindvar = add i32 %indvar, 1
6197 br label %Loop
6198</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6199
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]6200</div>
6201
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]6202<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6203<h4>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]6204 <a name="i_select">'<tt>select</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6205</h4>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]6206
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6207<div>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]6208
6209<h5>Syntax:</h5>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]6210<pre>
Dan Gohmanf72fb672008-09-09 01:02:47 +0000[diff] [blame]6211 &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>
6212
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]6213 <i>selty</i> is either i1 or {&lt;N x i1&gt;}
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]6214</pre>
6215
6216<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6217<p>The '<tt>select</tt>' instruction is used to choose one value based on a
6218 condition, without branching.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]6219
6220
6221<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6222<p>The '<tt>select</tt>' instruction requires an 'i1' value or a vector of 'i1'
6223 values indicating the condition, and two values of the
6224 same <a href="#t_firstclass">first class</a> type. If the val1/val2 are
6225 vectors and the condition is a scalar, then entire vectors are selected, not
6226 individual elements.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]6227
6228<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6229<p>If the condition is an i1 and it evaluates to 1, the instruction returns the
6230 first value argument; otherwise, it returns the second value argument.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]6231
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6232<p>If the condition is a vector of i1, then the value arguments must be vectors
6233 of the same size, and the selection is done element by element.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]6234
6235<h5>Example:</h5>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]6236<pre>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]6237 %X = select i1 true, i8 17, i8 42 <i>; yields i8:17</i>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]6238</pre>
Dan Gohmanc766f722009-01-22 01:39:38 +0000[diff] [blame]6239
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]6240</div>
6241
Robert Bocchino05ccd702006-01-15 20:48:27 +0000[diff] [blame]6242<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6243<h4>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]6244 <a name="i_call">'<tt>call</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6245</h4>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]6246
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6247<div>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]6248
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]6249<h5>Syntax:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]6250<pre>
Devang Patel307e8ab2008-10-07 17:48:33 +0000[diff] [blame]6251 &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]6252</pre>
6253
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]6254<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]6255<p>The '<tt>call</tt>' instruction represents a simple function call.</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]6256
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]6257<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]6258<p>This instruction requires several arguments:</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]6259
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]6260<ol>
Jeffrey Yasskin95fa80a2010-01-09 19:44:16 +0000[diff] [blame]6261 <li>The optional "tail" marker indicates that the callee function does not
6262 access any allocas or varargs in the caller. Note that calls may be
6263 marked "tail" even if they do not occur before
6264 a <a href="#i_ret"><tt>ret</tt></a> instruction. If the "tail" marker is
6265 present, the function call is eligible for tail call optimization,
6266 but <a href="CodeGenerator.html#tailcallopt">might not in fact be
Evan Chengdc444e92010-03-08 21:05:02 +0000[diff] [blame]6267 optimized into a jump</a>. The code generator may optimize calls marked
6268 "tail" with either 1) automatic <a href="CodeGenerator.html#sibcallopt">
6269 sibling call optimization</a> when the caller and callee have
6270 matching signatures, or 2) forced tail call optimization when the
6271 following extra requirements are met:
Jeffrey Yasskin95fa80a2010-01-09 19:44:16 +0000[diff] [blame]6272 <ul>
6273 <li>Caller and callee both have the calling
6274 convention <tt>fastcc</tt>.</li>
6275 <li>The call is in tail position (ret immediately follows call and ret
6276 uses value of call or is void).</li>
6277 <li>Option <tt>-tailcallopt</tt> is enabled,
Dan Gohmanfbbee8d2010-03-02 01:08:11 +0000[diff] [blame]6278 or <code>llvm::GuaranteedTailCallOpt</code> is <code>true</code>.</li>
Jeffrey Yasskin95fa80a2010-01-09 19:44:16 +0000[diff] [blame]6279 <li><a href="CodeGenerator.html#tailcallopt">Platform specific
6280 constraints are met.</a></li>
6281 </ul>
6282 </li>
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]6283
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6284 <li>The optional "cconv" marker indicates which <a href="#callingconv">calling
6285 convention</a> the call should use. If none is specified, the call
Jeffrey Yasskin95fa80a2010-01-09 19:44:16 +0000[diff] [blame]6286 defaults to using C calling conventions. The calling convention of the
6287 call must match the calling convention of the target function, or else the
6288 behavior is undefined.</li>
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]6289
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6290 <li>The optional <a href="#paramattrs">Parameter Attributes</a> list for
6291 return values. Only '<tt>zeroext</tt>', '<tt>signext</tt>', and
6292 '<tt>inreg</tt>' attributes are valid here.</li>
6293
6294 <li>'<tt>ty</tt>': the type of the call instruction itself which is also the
6295 type of the return value. Functions that return no value are marked
6296 <tt><a href="#t_void">void</a></tt>.</li>
6297
6298 <li>'<tt>fnty</tt>': shall be the signature of the pointer to function value
6299 being invoked. The argument types must match the types implied by this
6300 signature. This type can be omitted if the function is not varargs and if
6301 the function type does not return a pointer to a function.</li>
6302
6303 <li>'<tt>fnptrval</tt>': An LLVM value containing a pointer to a function to
6304 be invoked. In most cases, this is a direct function invocation, but
6305 indirect <tt>call</tt>s are just as possible, calling an arbitrary pointer
6306 to function value.</li>
6307
6308 <li>'<tt>function args</tt>': argument list whose types match the function
Chris Lattner0724fbd2010-03-02 06:36:51 +0000[diff] [blame]6309 signature argument types and parameter attributes. All arguments must be
6310 of <a href="#t_firstclass">first class</a> type. If the function
6311 signature indicates the function accepts a variable number of arguments,
6312 the extra arguments can be specified.</li>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6313
6314 <li>The optional <a href="#fnattrs">function attributes</a> list. Only
6315 '<tt>noreturn</tt>', '<tt>nounwind</tt>', '<tt>readonly</tt>' and
6316 '<tt>readnone</tt>' attributes are valid here.</li>
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]6317</ol>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]6318
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]6319<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6320<p>The '<tt>call</tt>' instruction is used to cause control flow to transfer to
6321 a specified function, with its incoming arguments bound to the specified
6322 values. Upon a '<tt><a href="#i_ret">ret</a></tt>' instruction in the called
6323 function, control flow continues with the instruction after the function
6324 call, and the return value of the function is bound to the result
6325 argument.</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]6326
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]6327<h5>Example:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]6328<pre>
Nick Lewyckydb7e3c92007-09-08 13:57:50 +0000[diff] [blame]6329 %retval = call i32 @test(i32 %argc)
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]6330 call i32 (i8*, ...)* @printf(i8* %msg, i32 12, i8 42) <i>; yields i32</i>
Chris Lattner772fccf2008-03-21 17:24:17 +0000[diff] [blame]6331 %X = tail call i32 @foo() <i>; yields i32</i>
6332 %Y = tail call <a href="#callingconv">fastcc</a> i32 @foo() <i>; yields i32</i>
6333 call void %foo(i8 97 signext)
Devang Patelc3fc6df2008-03-10 20:49:15 +0000[diff] [blame]6334
6335 %struct.A = type { i32, i8 }
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]6336 %r = call %struct.A @foo() <i>; yields { 32, i8 }</i>
Dan Gohmanb1e6b962008-10-04 19:00:07 +0000[diff] [blame]6337 %gr = extractvalue %struct.A %r, 0 <i>; yields i32</i>
6338 %gr1 = extractvalue %struct.A %r, 1 <i>; yields i8</i>
Chris Lattner85a350f2008-10-08 06:26:11 +0000[diff] [blame]6339 %Z = call void @foo() noreturn <i>; indicates that %foo never returns normally</i>
Matthijs Kooijmancb73d192008-10-07 10:03:45 +0000[diff] [blame]6340 %ZZ = call zeroext i32 @bar() <i>; Return value is %zero extended</i>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]6341</pre>
6342
Dale Johannesen07de8d12009-09-24 18:38:21 +0000[diff] [blame]6343<p>llvm treats calls to some functions with names and arguments that match the
Dale Johannesen9f8380b2009-09-25 17:04:42 +0000[diff] [blame]6344standard C99 library as being the C99 library functions, and may perform
6345optimizations or generate code for them under that assumption. This is
6346something we'd like to change in the future to provide better support for
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]6347freestanding environments and non-C-based languages.</p>
Dale Johannesen07de8d12009-09-24 18:38:21 +0000[diff] [blame]6348
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]6349</div>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]6350
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]6351<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6352<h4>
Chris Lattnerfb6977d2006-01-13 23:26:01 +0000[diff] [blame]6353 <a name="i_va_arg">'<tt>va_arg</tt>' Instruction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6354</h4>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]6355
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6356<div>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]6357
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]6358<h5>Syntax:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]6359<pre>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]6360 &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]6361</pre>
6362
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]6363<h5>Overview:</h5>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]6364<p>The '<tt>va_arg</tt>' instruction is used to access arguments passed through
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6365 the "variable argument" area of a function call. It is used to implement the
6366 <tt>va_arg</tt> macro in C.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]6367
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]6368<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6369<p>This instruction takes a <tt>va_list*</tt> value and the type of the
6370 argument. It returns a value of the specified argument type and increments
6371 the <tt>va_list</tt> to point to the next argument. The actual type
6372 of <tt>va_list</tt> is target specific.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]6373
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]6374<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6375<p>The '<tt>va_arg</tt>' instruction loads an argument of the specified type
6376 from the specified <tt>va_list</tt> and causes the <tt>va_list</tt> to point
6377 to the next argument. For more information, see the variable argument
6378 handling <a href="#int_varargs">Intrinsic Functions</a>.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]6379
6380<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]6381 take a variable number of arguments, for example, the <tt>vfprintf</tt>
6382 function.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]6383
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6384<p><tt>va_arg</tt> is an LLVM instruction instead of
6385 an <a href="#intrinsics">intrinsic function</a> because it takes a type as an
6386 argument.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]6387
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]6388<h5>Example:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]6389<p>See the <a href="#int_varargs">variable argument processing</a> section.</p>
6390
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6391<p>Note that the code generator does not yet fully support va_arg on many
6392 targets. Also, it does not currently support va_arg with aggregate types on
6393 any target.</p>
Dan Gohmanf3e60bd2009-01-12 23:12:39 +0000[diff] [blame]6394
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]6395</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]6396
Bill Wendlingf78faf82011-08-02 21:52:38 +0000[diff] [blame]6397<!-- _______________________________________________________________________ -->
6398<h4>
6399 <a name="i_landingpad">'<tt>landingpad</tt>' Instruction</a>
6400</h4>
6401
6402<div>
6403
6404<h5>Syntax:</h5>
6405<pre>
Duncan Sands8d6796b2012-01-13 19:59:16 +0000[diff] [blame]6406 &lt;resultval&gt; = landingpad &lt;resultty&gt; personality &lt;type&gt; &lt;pers_fn&gt; &lt;clause&gt;+
6407 &lt;resultval&gt; = landingpad &lt;resultty&gt; personality &lt;type&gt; &lt;pers_fn&gt; cleanup &lt;clause&gt;*
Bill Wendlingbf13ee12011-08-08 08:06:05 +0000[diff] [blame]6408
Bill Wendlingf78faf82011-08-02 21:52:38 +0000[diff] [blame]6409 &lt;clause&gt; := catch &lt;type&gt; &lt;value&gt;
Bill Wendlinge6e88262011-08-12 20:24:12 +0000[diff] [blame]6410 &lt;clause&gt; := filter &lt;array constant type&gt; &lt;array constant&gt;
Bill Wendlingf78faf82011-08-02 21:52:38 +0000[diff] [blame]6411</pre>
6412
6413<h5>Overview:</h5>
6414<p>The '<tt>landingpad</tt>' instruction is used by
6415 <a href="ExceptionHandling.html#overview">LLVM's exception handling
6416 system</a> to specify that a basic block is a landing pad &mdash; one where
6417 the exception lands, and corresponds to the code found in the
6418 <i><tt>catch</tt></i> portion of a <i><tt>try/catch</tt></i> sequence. It
6419 defines values supplied by the personality function (<tt>pers_fn</tt>) upon
6420 re-entry to the function. The <tt>resultval</tt> has the
Duncan Sands8d6796b2012-01-13 19:59:16 +0000[diff] [blame]6421 type <tt>resultty</tt>.</p>
Bill Wendlingf78faf82011-08-02 21:52:38 +0000[diff] [blame]6422
6423<h5>Arguments:</h5>
6424<p>This instruction takes a <tt>pers_fn</tt> value. This is the personality
6425 function associated with the unwinding mechanism. The optional
6426 <tt>cleanup</tt> flag indicates that the landing pad block is a cleanup.</p>
6427
6428<p>A <tt>clause</tt> begins with the clause type &mdash; <tt>catch</tt>
Bill Wendlinge6e88262011-08-12 20:24:12 +0000[diff] [blame]6429 or <tt>filter</tt> &mdash; and contains the global variable representing the
6430 "type" that may be caught or filtered respectively. Unlike the
6431 <tt>catch</tt> clause, the <tt>filter</tt> clause takes an array constant as
6432 its argument. Use "<tt>[0 x i8**] undef</tt>" for a filter which cannot
6433 throw. The '<tt>landingpad</tt>' instruction must contain <em>at least</em>
Bill Wendlingf78faf82011-08-02 21:52:38 +0000[diff] [blame]6434 one <tt>clause</tt> or the <tt>cleanup</tt> flag.</p>
6435
6436<h5>Semantics:</h5>
6437<p>The '<tt>landingpad</tt>' instruction defines the values which are set by the
6438 personality function (<tt>pers_fn</tt>) upon re-entry to the function, and
6439 therefore the "result type" of the <tt>landingpad</tt> instruction. As with
6440 calling conventions, how the personality function results are represented in
6441 LLVM IR is target specific.</p>
6442
Bill Wendlingb7a01352011-08-03 17:17:06 +0000[diff] [blame]6443<p>The clauses are applied in order from top to bottom. If two
6444 <tt>landingpad</tt> instructions are merged together through inlining, the
Duncan Sands8d6796b2012-01-13 19:59:16 +0000[diff] [blame]6445 clauses from the calling function are appended to the list of clauses.
6446 When the call stack is being unwound due to an exception being thrown, the
6447 exception is compared against each <tt>clause</tt> in turn. If it doesn't
6448 match any of the clauses, and the <tt>cleanup</tt> flag is not set, then
6449 unwinding continues further up the call stack.</p>
Bill Wendlingb7a01352011-08-03 17:17:06 +0000[diff] [blame]6450
Bill Wendlingf78faf82011-08-02 21:52:38 +0000[diff] [blame]6451<p>The <tt>landingpad</tt> instruction has several restrictions:</p>
6452
6453<ul>
6454 <li>A landing pad block is a basic block which is the unwind destination of an
6455 '<tt>invoke</tt>' instruction.</li>
6456 <li>A landing pad block must have a '<tt>landingpad</tt>' instruction as its
6457 first non-PHI instruction.</li>
6458 <li>There can be only one '<tt>landingpad</tt>' instruction within the landing
6459 pad block.</li>
6460 <li>A basic block that is not a landing pad block may not include a
6461 '<tt>landingpad</tt>' instruction.</li>
6462 <li>All '<tt>landingpad</tt>' instructions in a function must have the same
6463 personality function.</li>
6464</ul>
6465
6466<h5>Example:</h5>
6467<pre>
6468 ;; A landing pad which can catch an integer.
6469 %res = landingpad { i8*, i32 } personality i32 (...)* @__gxx_personality_v0
6470 catch i8** @_ZTIi
6471 ;; A landing pad that is a cleanup.
6472 %res = landingpad { i8*, i32 } personality i32 (...)* @__gxx_personality_v0
Bill Wendlinge6e88262011-08-12 20:24:12 +0000[diff] [blame]6473 cleanup
Bill Wendlingf78faf82011-08-02 21:52:38 +0000[diff] [blame]6474 ;; A landing pad which can catch an integer and can only throw a double.
6475 %res = landingpad { i8*, i32 } personality i32 (...)* @__gxx_personality_v0
6476 catch i8** @_ZTIi
Bill Wendlinge6e88262011-08-12 20:24:12 +0000[diff] [blame]6477 filter [1 x i8**] [@_ZTId]
Bill Wendlingf78faf82011-08-02 21:52:38 +0000[diff] [blame]6478</pre>
6479
6480</div>
6481
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6482</div>
6483
6484</div>
6485
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]6486<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6487<h2><a name="intrinsics">Intrinsic Functions</a></h2>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]6488<!-- *********************************************************************** -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]6489
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6490<div>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6491
6492<p>LLVM supports the notion of an "intrinsic function". These functions have
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6493 well known names and semantics and are required to follow certain
6494 restrictions. Overall, these intrinsics represent an extension mechanism for
6495 the LLVM language that does not require changing all of the transformations
6496 in LLVM when adding to the language (or the bitcode reader/writer, the
6497 parser, etc...).</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6498
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]6499<p>Intrinsic function names must all start with an "<tt>llvm.</tt>" prefix. This
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6500 prefix is reserved in LLVM for intrinsic names; thus, function names may not
6501 begin with this prefix. Intrinsic functions must always be external
6502 functions: you cannot define the body of intrinsic functions. Intrinsic
6503 functions may only be used in call or invoke instructions: it is illegal to
6504 take the address of an intrinsic function. Additionally, because intrinsic
6505 functions are part of the LLVM language, it is required if any are added that
6506 they be documented here.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6507
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6508<p>Some intrinsic functions can be overloaded, i.e., the intrinsic represents a
6509 family of functions that perform the same operation but on different data
6510 types. Because LLVM can represent over 8 million different integer types,
6511 overloading is used commonly to allow an intrinsic function to operate on any
6512 integer type. One or more of the argument types or the result type can be
6513 overloaded to accept any integer type. Argument types may also be defined as
6514 exactly matching a previous argument's type or the result type. This allows
6515 an intrinsic function which accepts multiple arguments, but needs all of them
6516 to be of the same type, to only be overloaded with respect to a single
6517 argument or the result.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6518
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6519<p>Overloaded intrinsics will have the names of its overloaded argument types
6520 encoded into its function name, each preceded by a period. Only those types
6521 which are overloaded result in a name suffix. Arguments whose type is matched
6522 against another type do not. For example, the <tt>llvm.ctpop</tt> function
6523 can take an integer of any width and returns an integer of exactly the same
6524 integer width. This leads to a family of functions such as
6525 <tt>i8 @llvm.ctpop.i8(i8 %val)</tt> and <tt>i29 @llvm.ctpop.i29(i29
6526 %val)</tt>. Only one type, the return type, is overloaded, and only one type
6527 suffix is required. Because the argument's type is matched against the return
6528 type, it does not require its own name suffix.</p>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]6529
Eric Christopher6c7e8a02009-12-05 02:46:03 +0000[diff] [blame]6530<p>To learn how to add an intrinsic function, please see the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6531 <a href="ExtendingLLVM.html">Extending LLVM Guide</a>.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6532
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]6533<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6534<h3>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]6535 <a name="int_varargs">Variable Argument Handling Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6536</h3>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]6537
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6538<div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6539
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6540<p>Variable argument support is defined in LLVM with
6541 the <a href="#i_va_arg"><tt>va_arg</tt></a> instruction and these three
6542 intrinsic functions. These functions are related to the similarly named
6543 macros defined in the <tt>&lt;stdarg.h&gt;</tt> header file.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6544
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6545<p>All of these functions operate on arguments that use a target-specific value
6546 type "<tt>va_list</tt>". The LLVM assembly language reference manual does
6547 not define what this type is, so all transformations should be prepared to
6548 handle these functions regardless of the type used.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6549
Chris Lattner374ab302006-05-15 17:26:46 +0000[diff] [blame]6550<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]6551 instruction and the variable argument handling intrinsic functions are
6552 used.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6553
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]6554<pre class="doc_code">
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]6555define i32 @test(i32 %X, ...) {
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6556 ; Initialize variable argument processing
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]6557 %ap = alloca i8*
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]6558 %ap2 = bitcast i8** %ap to i8*
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]6559 call void @llvm.va_start(i8* %ap2)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6560
6561 ; Read a single integer argument
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]6562 %tmp = va_arg i8** %ap, i32
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6563
6564 ; Demonstrate usage of llvm.va_copy and llvm.va_end
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]6565 %aq = alloca i8*
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]6566 %aq2 = bitcast i8** %aq to i8*
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]6567 call void @llvm.va_copy(i8* %aq2, i8* %ap2)
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]6568 call void @llvm.va_end(i8* %aq2)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6569
6570 ; Stop processing of arguments.
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]6571 call void @llvm.va_end(i8* %ap2)
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]6572 ret i32 %tmp
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6573}
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]6574
6575declare void @llvm.va_start(i8*)
6576declare void @llvm.va_copy(i8*, i8*)
6577declare void @llvm.va_end(i8*)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6578</pre>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]6579
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]6580<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6581<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6582 <a name="int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6583</h4>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]6584
6585
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6586<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6587
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]6588<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6589<pre>
6590 declare void %llvm.va_start(i8* &lt;arglist&gt;)
6591</pre>
6592
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]6593<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6594<p>The '<tt>llvm.va_start</tt>' intrinsic initializes <tt>*&lt;arglist&gt;</tt>
6595 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]6596
6597<h5>Arguments:</h5>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]6598<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]6599
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]6600<h5>Semantics:</h5>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]6601<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]6602 macro available in C. In a target-dependent way, it initializes
6603 the <tt>va_list</tt> element to which the argument points, so that the next
6604 call to <tt>va_arg</tt> will produce the first variable argument passed to
6605 the function. Unlike the C <tt>va_start</tt> macro, this intrinsic does not
6606 need to know the last argument of the function as the compiler can figure
6607 that out.</p>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]6608
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]6609</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]6610
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]6611<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6612<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6613 <a name="int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6614</h4>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]6615
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6616<div>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]6617
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6618<h5>Syntax:</h5>
6619<pre>
6620 declare void @llvm.va_end(i8* &lt;arglist&gt;)
6621</pre>
6622
6623<h5>Overview:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]6624<p>The '<tt>llvm.va_end</tt>' intrinsic destroys <tt>*&lt;arglist&gt;</tt>,
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6625 which has been initialized previously
6626 with <tt><a href="#int_va_start">llvm.va_start</a></tt>
6627 or <tt><a href="#i_va_copy">llvm.va_copy</a></tt>.</p>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]6628
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]6629<h5>Arguments:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]6630<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]6631
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]6632<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]6633<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]6634 macro available in C. In a target-dependent way, it destroys
6635 the <tt>va_list</tt> element to which the argument points. Calls
6636 to <a href="#int_va_start"><tt>llvm.va_start</tt></a>
6637 and <a href="#int_va_copy"> <tt>llvm.va_copy</tt></a> must be matched exactly
6638 with calls to <tt>llvm.va_end</tt>.</p>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]6639
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]6640</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]6641
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]6642<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6643<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6644 <a name="int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6645</h4>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]6646
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6647<div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6648
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]6649<h5>Syntax:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6650<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]6651 declare void @llvm.va_copy(i8* &lt;destarglist&gt;, i8* &lt;srcarglist&gt;)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6652</pre>
6653
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]6654<h5>Overview:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]6655<p>The '<tt>llvm.va_copy</tt>' intrinsic copies the current argument position
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6656 from the source argument list to the destination argument list.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6657
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]6658<h5>Arguments:</h5>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]6659<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]6660 The second argument is a pointer to a <tt>va_list</tt> element to copy
6661 from.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6662
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]6663<h5>Semantics:</h5>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]6664<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]6665 macro available in C. In a target-dependent way, it copies the
6666 source <tt>va_list</tt> element into the destination <tt>va_list</tt>
6667 element. This intrinsic is necessary because
6668 the <tt><a href="#int_va_start"> llvm.va_start</a></tt> intrinsic may be
6669 arbitrarily complex and require, for example, memory allocation.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6670
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]6671</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]6672
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6673</div>
6674
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]6675<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6676<h3>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6677 <a name="int_gc">Accurate Garbage Collection Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6678</h3>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6679
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6680<div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6681
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6682<p>LLVM support for <a href="GarbageCollection.html">Accurate Garbage
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]6683Collection</a> (GC) requires the implementation and generation of these
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6684intrinsics. These intrinsics allow identification of <a href="#int_gcroot">GC
6685roots on the stack</a>, as well as garbage collector implementations that
6686require <a href="#int_gcread">read</a> and <a href="#int_gcwrite">write</a>
6687barriers. Front-ends for type-safe garbage collected languages should generate
6688these intrinsics to make use of the LLVM garbage collectors. For more details,
6689see <a href="GarbageCollection.html">Accurate Garbage Collection with
6690LLVM</a>.</p>
Christopher Lamb303dae92007-12-17 01:00:21 +0000[diff] [blame]6691
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6692<p>The garbage collection intrinsics only operate on objects in the generic
6693 address space (address space zero).</p>
Christopher Lamb303dae92007-12-17 01:00:21 +0000[diff] [blame]6694
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6695<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6696<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6697 <a name="int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6698</h4>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6699
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6700<div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6701
6702<h5>Syntax:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6703<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]6704 declare void @llvm.gcroot(i8** %ptrloc, i8* %metadata)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6705</pre>
6706
6707<h5>Overview:</h5>
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]6708<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]6709 the code generator, and allows some metadata to be associated with it.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6710
6711<h5>Arguments:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6712<p>The first argument specifies the address of a stack object that contains the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6713 root pointer. The second pointer (which must be either a constant or a
6714 global value address) contains the meta-data to be associated with the
6715 root.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6716
6717<h5>Semantics:</h5>
Chris Lattner05d67092008-04-24 05:59:56 +0000[diff] [blame]6718<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]6719 location. At compile-time, the code generator generates information to allow
6720 the runtime to find the pointer at GC safe points. The '<tt>llvm.gcroot</tt>'
6721 intrinsic may only be used in a function which <a href="#gc">specifies a GC
6722 algorithm</a>.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6723
6724</div>
6725
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6726<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6727<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6728 <a name="int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6729</h4>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6730
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6731<div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6732
6733<h5>Syntax:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6734<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]6735 declare i8* @llvm.gcread(i8* %ObjPtr, i8** %Ptr)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6736</pre>
6737
6738<h5>Overview:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6739<p>The '<tt>llvm.gcread</tt>' intrinsic identifies reads of references from heap
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6740 locations, allowing garbage collector implementations that require read
6741 barriers.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6742
6743<h5>Arguments:</h5>
Chris Lattner80626e92006-03-14 20:02:51 +0000[diff] [blame]6744<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]6745 allocated from the garbage collector. The first object is a pointer to the
6746 start of the referenced object, if needed by the language runtime (otherwise
6747 null).</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6748
6749<h5>Semantics:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6750<p>The '<tt>llvm.gcread</tt>' intrinsic has the same semantics as a load
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6751 instruction, but may be replaced with substantially more complex code by the
6752 garbage collector runtime, as needed. The '<tt>llvm.gcread</tt>' intrinsic
6753 may only be used in a function which <a href="#gc">specifies a GC
6754 algorithm</a>.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6755
6756</div>
6757
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6758<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6759<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6760 <a name="int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6761</h4>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6762
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6763<div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6764
6765<h5>Syntax:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6766<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]6767 declare void @llvm.gcwrite(i8* %P1, i8* %Obj, i8** %P2)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6768</pre>
6769
6770<h5>Overview:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6771<p>The '<tt>llvm.gcwrite</tt>' intrinsic identifies writes of references to heap
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6772 locations, allowing garbage collector implementations that require write
6773 barriers (such as generational or reference counting collectors).</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6774
6775<h5>Arguments:</h5>
Chris Lattner80626e92006-03-14 20:02:51 +0000[diff] [blame]6776<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]6777 object to store it to, and the third is the address of the field of Obj to
6778 store to. If the runtime does not require a pointer to the object, Obj may
6779 be null.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6780
6781<h5>Semantics:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6782<p>The '<tt>llvm.gcwrite</tt>' intrinsic has the same semantics as a store
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6783 instruction, but may be replaced with substantially more complex code by the
6784 garbage collector runtime, as needed. The '<tt>llvm.gcwrite</tt>' intrinsic
6785 may only be used in a function which <a href="#gc">specifies a GC
6786 algorithm</a>.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6787
6788</div>
6789
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6790</div>
6791
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]6792<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6793<h3>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6794 <a name="int_codegen">Code Generator Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6795</h3>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6796
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6797<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6798
6799<p>These intrinsics are provided by LLVM to expose special features that may
6800 only be implemented with code generator support.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6801
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6802<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6803<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6804 <a name="int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6805</h4>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6806
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6807<div>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6808
6809<h5>Syntax:</h5>
6810<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]6811 declare i8 *@llvm.returnaddress(i32 &lt;level&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6812</pre>
6813
6814<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6815<p>The '<tt>llvm.returnaddress</tt>' intrinsic attempts to compute a
6816 target-specific value indicating the return address of the current function
6817 or one of its callers.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6818
6819<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6820<p>The argument to this intrinsic indicates which function to return the address
6821 for. Zero indicates the calling function, one indicates its caller, etc.
6822 The argument is <b>required</b> to be a constant integer value.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6823
6824<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6825<p>The '<tt>llvm.returnaddress</tt>' intrinsic either returns a pointer
6826 indicating the return address of the specified call frame, or zero if it
6827 cannot be identified. The value returned by this intrinsic is likely to be
6828 incorrect or 0 for arguments other than zero, so it should only be used for
6829 debugging purposes.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6830
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6831<p>Note that calling this intrinsic does not prevent function inlining or other
6832 aggressive transformations, so the value returned may not be that of the
6833 obvious source-language caller.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6834
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6835</div>
6836
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6837<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6838<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6839 <a name="int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6840</h4>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6841
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6842<div>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6843
6844<h5>Syntax:</h5>
6845<pre>
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]6846 declare i8* @llvm.frameaddress(i32 &lt;level&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6847</pre>
6848
6849<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6850<p>The '<tt>llvm.frameaddress</tt>' intrinsic attempts to return the
6851 target-specific frame pointer value for the specified stack frame.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6852
6853<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6854<p>The argument to this intrinsic indicates which function to return the frame
6855 pointer for. Zero indicates the calling function, one indicates its caller,
6856 etc. The argument is <b>required</b> to be a constant integer value.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6857
6858<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6859<p>The '<tt>llvm.frameaddress</tt>' intrinsic either returns a pointer
6860 indicating the frame address of the specified call frame, or zero if it
6861 cannot be identified. The value returned by this intrinsic is likely to be
6862 incorrect or 0 for arguments other than zero, so it should only be used for
6863 debugging purposes.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6864
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6865<p>Note that calling this intrinsic does not prevent function inlining or other
6866 aggressive transformations, so the value returned may not be that of the
6867 obvious source-language caller.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6868
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]6869</div>
6870
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]6871<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6872<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6873 <a name="int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6874</h4>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6875
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6876<div>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6877
6878<h5>Syntax:</h5>
6879<pre>
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]6880 declare i8* @llvm.stacksave()
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6881</pre>
6882
6883<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6884<p>The '<tt>llvm.stacksave</tt>' intrinsic is used to remember the current state
6885 of the function stack, for use
6886 with <a href="#int_stackrestore"> <tt>llvm.stackrestore</tt></a>. This is
6887 useful for implementing language features like scoped automatic variable
6888 sized arrays in C99.</p>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6889
6890<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6891<p>This intrinsic returns a opaque pointer value that can be passed
6892 to <a href="#int_stackrestore"><tt>llvm.stackrestore</tt></a>. When
6893 an <tt>llvm.stackrestore</tt> intrinsic is executed with a value saved
6894 from <tt>llvm.stacksave</tt>, it effectively restores the state of the stack
6895 to the state it was in when the <tt>llvm.stacksave</tt> intrinsic executed.
6896 In practice, this pops any <a href="#i_alloca">alloca</a> blocks from the
6897 stack that were allocated after the <tt>llvm.stacksave</tt> was executed.</p>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6898
6899</div>
6900
6901<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6902<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6903 <a name="int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6904</h4>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6905
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6906<div>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6907
6908<h5>Syntax:</h5>
6909<pre>
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]6910 declare void @llvm.stackrestore(i8* %ptr)
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6911</pre>
6912
6913<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6914<p>The '<tt>llvm.stackrestore</tt>' intrinsic is used to restore the state of
6915 the function stack to the state it was in when the
6916 corresponding <a href="#int_stacksave"><tt>llvm.stacksave</tt></a> intrinsic
6917 executed. This is useful for implementing language features like scoped
6918 automatic variable sized arrays in C99.</p>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6919
6920<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6921<p>See the description
6922 for <a href="#int_stacksave"><tt>llvm.stacksave</tt></a>.</p>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6923
6924</div>
6925
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]6926<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6927<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6928 <a name="int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6929</h4>
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]6930
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6931<div>
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]6932
6933<h5>Syntax:</h5>
6934<pre>
Bruno Cardoso Lopes9a767332011-06-14 04:58:37 +0000[diff] [blame]6935 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]6936</pre>
6937
6938<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6939<p>The '<tt>llvm.prefetch</tt>' intrinsic is a hint to the code generator to
6940 insert a prefetch instruction if supported; otherwise, it is a noop.
6941 Prefetches have no effect on the behavior of the program but can change its
6942 performance characteristics.</p>
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]6943
6944<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6945<p><tt>address</tt> is the address to be prefetched, <tt>rw</tt> is the
6946 specifier determining if the fetch should be for a read (0) or write (1),
6947 and <tt>locality</tt> is a temporal locality specifier ranging from (0) - no
Bruno Cardoso Lopes9a767332011-06-14 04:58:37 +0000[diff] [blame]6948 locality, to (3) - extremely local keep in cache. The <tt>cache type</tt>
6949 specifies whether the prefetch is performed on the data (1) or instruction (0)
6950 cache. The <tt>rw</tt>, <tt>locality</tt> and <tt>cache type</tt> arguments
6951 must be constant integers.</p>
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]6952
6953<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6954<p>This intrinsic does not modify the behavior of the program. In particular,
6955 prefetches cannot trap and do not produce a value. On targets that support
6956 this intrinsic, the prefetch can provide hints to the processor cache for
6957 better performance.</p>
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]6958
6959</div>
6960
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]6961<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6962<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6963 <a name="int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6964</h4>
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]6965
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6966<div>
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]6967
6968<h5>Syntax:</h5>
6969<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]6970 declare void @llvm.pcmarker(i32 &lt;id&gt;)
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]6971</pre>
6972
6973<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6974<p>The '<tt>llvm.pcmarker</tt>' intrinsic is a method to export a Program
6975 Counter (PC) in a region of code to simulators and other tools. The method
6976 is target specific, but it is expected that the marker will use exported
6977 symbols to transmit the PC of the marker. The marker makes no guarantees
6978 that it will remain with any specific instruction after optimizations. It is
6979 possible that the presence of a marker will inhibit optimizations. The
6980 intended use is to be inserted after optimizations to allow correlations of
6981 simulation runs.</p>
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]6982
6983<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6984<p><tt>id</tt> is a numerical id identifying the marker.</p>
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]6985
6986<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]6987<p>This intrinsic does not modify the behavior of the program. Backends that do
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]6988 not support this intrinsic may ignore it.</p>
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]6989
6990</div>
6991
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]6992<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6993<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]6994 <a name="int_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]6995</h4>
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]6996
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]6997<div>
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]6998
6999<h5>Syntax:</h5>
7000<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]7001 declare i64 @llvm.readcyclecounter()
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]7002</pre>
7003
7004<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7005<p>The '<tt>llvm.readcyclecounter</tt>' intrinsic provides access to the cycle
7006 counter register (or similar low latency, high accuracy clocks) on those
7007 targets that support it. On X86, it should map to RDTSC. On Alpha, it
7008 should map to RPCC. As the backing counters overflow quickly (on the order
7009 of 9 seconds on alpha), this should only be used for small timings.</p>
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]7010
7011<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7012<p>When directly supported, reading the cycle counter should not modify any
7013 memory. Implementations are allowed to either return a application specific
7014 value or a system wide value. On backends without support, this is lowered
7015 to a constant 0.</p>
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]7016
7017</div>
7018
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7019</div>
7020
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]7021<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7022<h3>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]7023 <a name="int_libc">Standard C Library Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7024</h3>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]7025
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7026<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7027
7028<p>LLVM provides intrinsics for a few important standard C library functions.
7029 These intrinsics allow source-language front-ends to pass information about
7030 the alignment of the pointer arguments to the code generator, providing
7031 opportunity for more efficient code generation.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]7032
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]7033<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7034<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]7035 <a name="int_memcpy">'<tt>llvm.memcpy</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7036</h4>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]7037
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7038<div>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]7039
7040<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7041<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]7042 integer bit width and for different address spaces. Not all targets support
7043 all bit widths however.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7044
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]7045<pre>
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]7046 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]7047 i32 &lt;len&gt;, i32 &lt;align&gt;, i1 &lt;isvolatile&gt;)
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]7048 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]7049 i64 &lt;len&gt;, i32 &lt;align&gt;, i1 &lt;isvolatile&gt;)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]7050</pre>
7051
7052<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7053<p>The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the
7054 source location to the destination location.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]7055
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7056<p>Note that, unlike the standard libc function, the <tt>llvm.memcpy.*</tt>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]7057 intrinsics do not return a value, takes extra alignment/isvolatile arguments
7058 and the pointers can be in specified address spaces.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]7059
7060<h5>Arguments:</h5>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]7061
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7062<p>The first argument is a pointer to the destination, the second is a pointer
7063 to the source. The third argument is an integer argument specifying the
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]7064 number of bytes to copy, the fourth argument is the alignment of the
7065 source and destination locations, and the fifth is a boolean indicating a
7066 volatile access.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]7067
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]7068<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]7069 then the caller guarantees that both the source and destination pointers are
7070 aligned to that boundary.</p>
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]7071
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]7072<p>If the <tt>isvolatile</tt> parameter is <tt>true</tt>, the
7073 <tt>llvm.memcpy</tt> call is a <a href="#volatile">volatile operation</a>.
7074 The detailed access behavior is not very cleanly specified and it is unwise
7075 to depend on it.</p>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]7076
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]7077<h5>Semantics:</h5>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]7078
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7079<p>The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the
7080 source location to the destination location, which are not allowed to
7081 overlap. It copies "len" bytes of memory over. If the argument is known to
7082 be aligned to some boundary, this can be specified as the fourth argument,
7083 otherwise it should be set to 0 or 1.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]7084
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]7085</div>
7086
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]7087<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7088<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]7089 <a name="int_memmove">'<tt>llvm.memmove</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7090</h4>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]7091
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7092<div>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]7093
7094<h5>Syntax:</h5>
Chris Lattner824b9582008-11-21 16:42:48 +0000[diff] [blame]7095<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]7096 width and for different address space. Not all targets support all bit
7097 widths however.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7098
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]7099<pre>
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]7100 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]7101 i32 &lt;len&gt;, i32 &lt;align&gt;, i1 &lt;isvolatile&gt;)
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]7102 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]7103 i64 &lt;len&gt;, i32 &lt;align&gt;, i1 &lt;isvolatile&gt;)
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]7104</pre>
7105
7106<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7107<p>The '<tt>llvm.memmove.*</tt>' intrinsics move a block of memory from the
7108 source location to the destination location. It is similar to the
7109 '<tt>llvm.memcpy</tt>' intrinsic but allows the two memory locations to
7110 overlap.</p>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]7111
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7112<p>Note that, unlike the standard libc function, the <tt>llvm.memmove.*</tt>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]7113 intrinsics do not return a value, takes extra alignment/isvolatile arguments
7114 and the pointers can be in specified address spaces.</p>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]7115
7116<h5>Arguments:</h5>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]7117
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7118<p>The first argument is a pointer to the destination, the second is a pointer
7119 to the source. The third argument is an integer argument specifying the
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]7120 number of bytes to copy, the fourth argument is the alignment of the
7121 source and destination locations, and the fifth is a boolean indicating a
7122 volatile access.</p>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]7123
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]7124<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]7125 then the caller guarantees that the source and destination pointers are
7126 aligned to that boundary.</p>
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]7127
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]7128<p>If the <tt>isvolatile</tt> parameter is <tt>true</tt>, the
7129 <tt>llvm.memmove</tt> call is a <a href="#volatile">volatile operation</a>.
7130 The detailed access behavior is not very cleanly specified and it is unwise
7131 to depend on it.</p>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]7132
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]7133<h5>Semantics:</h5>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]7134
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7135<p>The '<tt>llvm.memmove.*</tt>' intrinsics copy a block of memory from the
7136 source location to the destination location, which may overlap. It copies
7137 "len" bytes of memory over. If the argument is known to be aligned to some
7138 boundary, this can be specified as the fourth argument, otherwise it should
7139 be set to 0 or 1.</p>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]7140
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]7141</div>
7142
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]7143<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7144<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]7145 <a name="int_memset">'<tt>llvm.memset.*</tt>' Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7146</h4>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]7147
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7148<div>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]7149
7150<h5>Syntax:</h5>
Chris Lattner824b9582008-11-21 16:42:48 +0000[diff] [blame]7151<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]7152 width and for different address spaces. However, not all targets support all
7153 bit widths.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7154
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]7155<pre>
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]7156 declare void @llvm.memset.p0i8.i32(i8* &lt;dest&gt;, i8 &lt;val&gt;,
Chris Lattnerff35c3f2010-04-08 00:54:34 +0000[diff] [blame]7157 i32 &lt;len&gt;, i32 &lt;align&gt;, i1 &lt;isvolatile&gt;)
Dan Gohmanfe47aae2010-05-28 17:13:49 +0000[diff] [blame]7158 declare void @llvm.memset.p0i8.i64(i8* &lt;dest&gt;, i8 &lt;val&gt;,
Chris Lattnerff35c3f2010-04-08 00:54:34 +0000[diff] [blame]7159 i64 &lt;len&gt;, i32 &lt;align&gt;, i1 &lt;isvolatile&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]7160</pre>
7161
7162<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7163<p>The '<tt>llvm.memset.*</tt>' intrinsics fill a block of memory with a
7164 particular byte value.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]7165
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7166<p>Note that, unlike the standard libc function, the <tt>llvm.memset</tt>
John Criswellcdcbbfc2010-07-30 16:30:28 +0000[diff] [blame]7167 intrinsic does not return a value and takes extra alignment/volatile
7168 arguments. Also, the destination can be in an arbitrary address space.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]7169
7170<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7171<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]7172 byte value with which to fill it, the third argument is an integer argument
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7173 specifying the number of bytes to fill, and the fourth argument is the known
John Criswellcdcbbfc2010-07-30 16:30:28 +0000[diff] [blame]7174 alignment of the destination location.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]7175
Dan Gohmanb55a1ee2010-03-01 17:41:39 +0000[diff] [blame]7176<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]7177 then the caller guarantees that the destination pointer is aligned to that
7178 boundary.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]7179
Jeffrey Yasskin93e066d2010-04-26 21:21:24 +0000[diff] [blame]7180<p>If the <tt>isvolatile</tt> parameter is <tt>true</tt>, the
7181 <tt>llvm.memset</tt> call is a <a href="#volatile">volatile operation</a>.
7182 The detailed access behavior is not very cleanly specified and it is unwise
7183 to depend on it.</p>
Chris Lattner9f636de2010-04-08 00:53:57 +0000[diff] [blame]7184
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]7185<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7186<p>The '<tt>llvm.memset.*</tt>' intrinsics fill "len" bytes of memory starting
7187 at the destination location. If the argument is known to be aligned to some
7188 boundary, this can be specified as the fourth argument, otherwise it should
7189 be set to 0 or 1.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]7190
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]7191</div>
7192
Chris Lattner32006282004-06-11 02:28:03 +0000[diff] [blame]7193<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7194<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]7195 <a name="int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7196</h4>
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]7197
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7198<div>
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]7199
7200<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7201<p>This is an overloaded intrinsic. You can use <tt>llvm.sqrt</tt> on any
7202 floating point or vector of floating point type. Not all targets support all
7203 types however.</p>
7204
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]7205<pre>
Dale Johannesen408f9c12007-10-02 17:47:38 +0000[diff] [blame]7206 declare float @llvm.sqrt.f32(float %Val)
7207 declare double @llvm.sqrt.f64(double %Val)
7208 declare x86_fp80 @llvm.sqrt.f80(x86_fp80 %Val)
7209 declare fp128 @llvm.sqrt.f128(fp128 %Val)
7210 declare ppc_fp128 @llvm.sqrt.ppcf128(ppc_fp128 %Val)
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]7211</pre>
7212
7213<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7214<p>The '<tt>llvm.sqrt</tt>' intrinsics return the sqrt of the specified operand,
7215 returning the same value as the libm '<tt>sqrt</tt>' functions would.
7216 Unlike <tt>sqrt</tt> in libm, however, <tt>llvm.sqrt</tt> has undefined
7217 behavior for negative numbers other than -0.0 (which allows for better
7218 optimization, because there is no need to worry about errno being
7219 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]7220
7221<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7222<p>The argument and return value are floating point numbers of the same
7223 type.</p>
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]7224
7225<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7226<p>This function returns the sqrt of the specified operand if it is a
7227 nonnegative floating point number.</p>
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]7228
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]7229</div>
7230
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]7231<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7232<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]7233 <a name="int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7234</h4>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]7235
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7236<div>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]7237
7238<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7239<p>This is an overloaded intrinsic. You can use <tt>llvm.powi</tt> on any
7240 floating point or vector of floating point type. Not all targets support all
7241 types however.</p>
7242
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]7243<pre>
Dale Johannesen408f9c12007-10-02 17:47:38 +0000[diff] [blame]7244 declare float @llvm.powi.f32(float %Val, i32 %power)
7245 declare double @llvm.powi.f64(double %Val, i32 %power)
7246 declare x86_fp80 @llvm.powi.f80(x86_fp80 %Val, i32 %power)
7247 declare fp128 @llvm.powi.f128(fp128 %Val, i32 %power)
7248 declare ppc_fp128 @llvm.powi.ppcf128(ppc_fp128 %Val, i32 %power)
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]7249</pre>
7250
7251<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7252<p>The '<tt>llvm.powi.*</tt>' intrinsics return the first operand raised to the
7253 specified (positive or negative) power. The order of evaluation of
7254 multiplications is not defined. When a vector of floating point type is
7255 used, the second argument remains a scalar integer value.</p>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]7256
7257<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7258<p>The second argument is an integer power, and the first is a value to raise to
7259 that power.</p>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]7260
7261<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7262<p>This function returns the first value raised to the second power with an
7263 unspecified sequence of rounding operations.</p>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]7264
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]7265</div>
7266
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]7267<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7268<h4>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]7269 <a name="int_sin">'<tt>llvm.sin.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7270</h4>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]7271
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7272<div>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]7273
7274<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7275<p>This is an overloaded intrinsic. You can use <tt>llvm.sin</tt> on any
7276 floating point or vector of floating point type. Not all targets support all
7277 types however.</p>
7278
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]7279<pre>
7280 declare float @llvm.sin.f32(float %Val)
7281 declare double @llvm.sin.f64(double %Val)
7282 declare x86_fp80 @llvm.sin.f80(x86_fp80 %Val)
7283 declare fp128 @llvm.sin.f128(fp128 %Val)
7284 declare ppc_fp128 @llvm.sin.ppcf128(ppc_fp128 %Val)
7285</pre>
7286
7287<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7288<p>The '<tt>llvm.sin.*</tt>' intrinsics return the sine of the operand.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]7289
7290<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7291<p>The argument and return value are floating point numbers of the same
7292 type.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]7293
7294<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7295<p>This function returns the sine of the specified operand, returning the same
7296 values as the libm <tt>sin</tt> functions would, and handles error conditions
7297 in the same way.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]7298
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]7299</div>
7300
7301<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7302<h4>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]7303 <a name="int_cos">'<tt>llvm.cos.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7304</h4>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]7305
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7306<div>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]7307
7308<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7309<p>This is an overloaded intrinsic. You can use <tt>llvm.cos</tt> on any
7310 floating point or vector of floating point type. Not all targets support all
7311 types however.</p>
7312
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]7313<pre>
7314 declare float @llvm.cos.f32(float %Val)
7315 declare double @llvm.cos.f64(double %Val)
7316 declare x86_fp80 @llvm.cos.f80(x86_fp80 %Val)
7317 declare fp128 @llvm.cos.f128(fp128 %Val)
7318 declare ppc_fp128 @llvm.cos.ppcf128(ppc_fp128 %Val)
7319</pre>
7320
7321<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7322<p>The '<tt>llvm.cos.*</tt>' intrinsics return the cosine of the operand.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]7323
7324<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7325<p>The argument and return value are floating point numbers of the same
7326 type.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]7327
7328<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7329<p>This function returns the cosine of the specified operand, returning the same
7330 values as the libm <tt>cos</tt> functions would, and handles error conditions
7331 in the same way.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]7332
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]7333</div>
7334
7335<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7336<h4>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]7337 <a name="int_pow">'<tt>llvm.pow.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7338</h4>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]7339
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7340<div>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]7341
7342<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7343<p>This is an overloaded intrinsic. You can use <tt>llvm.pow</tt> on any
7344 floating point or vector of floating point type. Not all targets support all
7345 types however.</p>
7346
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]7347<pre>
7348 declare float @llvm.pow.f32(float %Val, float %Power)
7349 declare double @llvm.pow.f64(double %Val, double %Power)
7350 declare x86_fp80 @llvm.pow.f80(x86_fp80 %Val, x86_fp80 %Power)
7351 declare fp128 @llvm.pow.f128(fp128 %Val, fp128 %Power)
7352 declare ppc_fp128 @llvm.pow.ppcf128(ppc_fp128 %Val, ppc_fp128 Power)
7353</pre>
7354
7355<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7356<p>The '<tt>llvm.pow.*</tt>' intrinsics return the first operand raised to the
7357 specified (positive or negative) power.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]7358
7359<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7360<p>The second argument is a floating point power, and the first is a value to
7361 raise to that power.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]7362
7363<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7364<p>This function returns the first value raised to the second power, returning
7365 the same values as the libm <tt>pow</tt> functions would, and handles error
7366 conditions in the same way.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]7367
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]7368</div>
7369
Dan Gohman4e9011c2011-05-23 21:13:03 +0000[diff] [blame]7370<!-- _______________________________________________________________________ -->
7371<h4>
7372 <a name="int_exp">'<tt>llvm.exp.*</tt>' Intrinsic</a>
7373</h4>
7374
7375<div>
7376
7377<h5>Syntax:</h5>
7378<p>This is an overloaded intrinsic. You can use <tt>llvm.exp</tt> on any
7379 floating point or vector of floating point type. Not all targets support all
7380 types however.</p>
7381
7382<pre>
7383 declare float @llvm.exp.f32(float %Val)
7384 declare double @llvm.exp.f64(double %Val)
7385 declare x86_fp80 @llvm.exp.f80(x86_fp80 %Val)
7386 declare fp128 @llvm.exp.f128(fp128 %Val)
7387 declare ppc_fp128 @llvm.exp.ppcf128(ppc_fp128 %Val)
7388</pre>
7389
7390<h5>Overview:</h5>
7391<p>The '<tt>llvm.exp.*</tt>' intrinsics perform the exp function.</p>
7392
7393<h5>Arguments:</h5>
7394<p>The argument and return value are floating point numbers of the same
7395 type.</p>
7396
7397<h5>Semantics:</h5>
7398<p>This function returns the same values as the libm <tt>exp</tt> functions
7399 would, and handles error conditions in the same way.</p>
7400
7401</div>
7402
7403<!-- _______________________________________________________________________ -->
7404<h4>
7405 <a name="int_log">'<tt>llvm.log.*</tt>' Intrinsic</a>
7406</h4>
7407
7408<div>
7409
7410<h5>Syntax:</h5>
7411<p>This is an overloaded intrinsic. You can use <tt>llvm.log</tt> on any
7412 floating point or vector of floating point type. Not all targets support all
7413 types however.</p>
7414
7415<pre>
7416 declare float @llvm.log.f32(float %Val)
7417 declare double @llvm.log.f64(double %Val)
7418 declare x86_fp80 @llvm.log.f80(x86_fp80 %Val)
7419 declare fp128 @llvm.log.f128(fp128 %Val)
7420 declare ppc_fp128 @llvm.log.ppcf128(ppc_fp128 %Val)
7421</pre>
7422
7423<h5>Overview:</h5>
7424<p>The '<tt>llvm.log.*</tt>' intrinsics perform the log function.</p>
7425
7426<h5>Arguments:</h5>
7427<p>The argument and return value are floating point numbers of the same
7428 type.</p>
7429
7430<h5>Semantics:</h5>
7431<p>This function returns the same values as the libm <tt>log</tt> functions
7432 would, and handles error conditions in the same way.</p>
7433
Nick Lewycky1c929be2011-10-31 01:32:21 +0000[diff] [blame]7434</div>
7435
7436<!-- _______________________________________________________________________ -->
Cameron Zwarich33390842011-07-08 21:39:21 +0000[diff] [blame]7437<h4>
7438 <a name="int_fma">'<tt>llvm.fma.*</tt>' Intrinsic</a>
7439</h4>
7440
7441<div>
7442
7443<h5>Syntax:</h5>
7444<p>This is an overloaded intrinsic. You can use <tt>llvm.fma</tt> on any
7445 floating point or vector of floating point type. Not all targets support all
7446 types however.</p>
7447
7448<pre>
7449 declare float @llvm.fma.f32(float %a, float %b, float %c)
7450 declare double @llvm.fma.f64(double %a, double %b, double %c)
7451 declare x86_fp80 @llvm.fma.f80(x86_fp80 %a, x86_fp80 %b, x86_fp80 %c)
7452 declare fp128 @llvm.fma.f128(fp128 %a, fp128 %b, fp128 %c)
7453 declare ppc_fp128 @llvm.fma.ppcf128(ppc_fp128 %a, ppc_fp128 %b, ppc_fp128 %c)
7454</pre>
7455
7456<h5>Overview:</h5>
Cameron Zwarichabc43e62011-07-08 22:13:55 +0000[diff] [blame]7457<p>The '<tt>llvm.fma.*</tt>' intrinsics perform the fused multiply-add
Cameron Zwarich33390842011-07-08 21:39:21 +0000[diff] [blame]7458 operation.</p>
7459
7460<h5>Arguments:</h5>
7461<p>The argument and return value are floating point numbers of the same
7462 type.</p>
7463
7464<h5>Semantics:</h5>
7465<p>This function returns the same values as the libm <tt>fma</tt> functions
7466 would.</p>
7467
Dan Gohman4e9011c2011-05-23 21:13:03 +0000[diff] [blame]7468</div>
7469
NAKAMURA Takumi4b2e07a2011-10-31 13:04:26 +0000[diff] [blame]7470</div>
7471
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7472<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7473<h3>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]7474 <a name="int_manip">Bit Manipulation Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7475</h3>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7476
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7477<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7478
7479<p>LLVM provides intrinsics for a few important bit manipulation operations.
7480 These allow efficient code generation for some algorithms.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7481
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7482<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7483<h4>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]7484 <a name="int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7485</h4>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]7486
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7487<div>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]7488
7489<h5>Syntax:</h5>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]7490<p>This is an overloaded intrinsic function. You can use bswap on any integer
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7491 type that is an even number of bytes (i.e. BitWidth % 16 == 0).</p>
7492
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]7493<pre>
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]7494 declare i16 @llvm.bswap.i16(i16 &lt;id&gt;)
7495 declare i32 @llvm.bswap.i32(i32 &lt;id&gt;)
7496 declare i64 @llvm.bswap.i64(i64 &lt;id&gt;)
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]7497</pre>
7498
7499<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7500<p>The '<tt>llvm.bswap</tt>' family of intrinsics is used to byte swap integer
7501 values with an even number of bytes (positive multiple of 16 bits). These
7502 are useful for performing operations on data that is not in the target's
7503 native byte order.</p>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]7504
7505<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7506<p>The <tt>llvm.bswap.i16</tt> intrinsic returns an i16 value that has the high
7507 and low byte of the input i16 swapped. Similarly,
7508 the <tt>llvm.bswap.i32</tt> intrinsic returns an i32 value that has the four
7509 bytes of the input i32 swapped, so that if the input bytes are numbered 0, 1,
7510 2, 3 then the returned i32 will have its bytes in 3, 2, 1, 0 order.
7511 The <tt>llvm.bswap.i48</tt>, <tt>llvm.bswap.i64</tt> and other intrinsics
7512 extend this concept to additional even-byte lengths (6 bytes, 8 bytes and
7513 more, respectively).</p>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]7514
7515</div>
7516
7517<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7518<h4>
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]7519 <a name="int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7520</h4>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7521
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7522<div>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7523
7524<h5>Syntax:</h5>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]7525<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]7526 width, or on any vector with integer elements. Not all targets support all
7527 bit widths or vector types, however.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7528
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7529<pre>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7530 declare i8 @llvm.ctpop.i8(i8 &lt;src&gt;)
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]7531 declare i16 @llvm.ctpop.i16(i16 &lt;src&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]7532 declare i32 @llvm.ctpop.i32(i32 &lt;src&gt;)
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]7533 declare i64 @llvm.ctpop.i64(i64 &lt;src&gt;)
7534 declare i256 @llvm.ctpop.i256(i256 &lt;src&gt;)
Owen Andersonf1ac4652011-07-01 21:52:38 +0000[diff] [blame]7535 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]7536</pre>
7537
7538<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7539<p>The '<tt>llvm.ctpop</tt>' family of intrinsics counts the number of bits set
7540 in a value.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7541
7542<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7543<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]7544 integer type, or a vector with integer elements.
7545 The return type must match the argument type.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7546
7547<h5>Semantics:</h5>
Owen Andersonf1ac4652011-07-01 21:52:38 +0000[diff] [blame]7548<p>The '<tt>llvm.ctpop</tt>' intrinsic counts the 1's in a variable, or within each
7549 element of a vector.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7550
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7551</div>
7552
7553<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7554<h4>
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]7555 <a name="int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7556</h4>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7557
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7558<div>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7559
7560<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7561<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]7562 integer bit width, or any vector whose elements are integers. Not all
7563 targets support all bit widths or vector types, however.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7564
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7565<pre>
Chandler Carruth48b0bbf2011-12-12 04:36:04 +0000[diff] [blame]7566 declare i8 @llvm.ctlz.i8 (i8 &lt;src&gt;, i1 &lt;is_zero_undef&gt;)
7567 declare i16 @llvm.ctlz.i16 (i16 &lt;src&gt;, i1 &lt;is_zero_undef&gt;)
7568 declare i32 @llvm.ctlz.i32 (i32 &lt;src&gt;, i1 &lt;is_zero_undef&gt;)
7569 declare i64 @llvm.ctlz.i64 (i64 &lt;src&gt;, i1 &lt;is_zero_undef&gt;)
7570 declare i256 @llvm.ctlz.i256(i256 &lt;src&gt;, i1 &lt;is_zero_undef&gt;)
7571 declase &lt;2 x i32&gt; @llvm.ctlz.v2i32(&lt;2 x i32&gt; &lt;src&gt;, i1 &lt;is_zero_undef&gt;)
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7572</pre>
7573
7574<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7575<p>The '<tt>llvm.ctlz</tt>' family of intrinsic functions counts the number of
7576 leading zeros in a variable.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7577
7578<h5>Arguments:</h5>
Chandler Carruth48b0bbf2011-12-12 04:36:04 +0000[diff] [blame]7579<p>The first argument is the value to be counted. This argument may be of any
7580 integer type, or a vectory with integer element type. The return type
7581 must match the first argument type.</p>
7582
7583<p>The second argument must be a constant and is a flag to indicate whether the
7584 intrinsic should ensure that a zero as the first argument produces a defined
7585 result. Historically some architectures did not provide a defined result for
7586 zero values as efficiently, and many algorithms are now predicated on
7587 avoiding zero-value inputs.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7588
7589<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7590<p>The '<tt>llvm.ctlz</tt>' intrinsic counts the leading (most significant)
Chandler Carruth48b0bbf2011-12-12 04:36:04 +0000[diff] [blame]7591 zeros in a variable, or within each element of the vector.
7592 If <tt>src == 0</tt> then the result is the size in bits of the type of
7593 <tt>src</tt> if <tt>is_zero_undef == 0</tt> and <tt>undef</tt> otherwise.
7594 For example, <tt>llvm.ctlz(i32 2) = 30</tt>.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7595
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]7596</div>
Chris Lattner32006282004-06-11 02:28:03 +0000[diff] [blame]7597
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]7598<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7599<h4>
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]7600 <a name="int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7601</h4>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]7602
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7603<div>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]7604
7605<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7606<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]7607 integer bit width, or any vector of integer elements. Not all targets
7608 support all bit widths or vector types, however.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7609
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]7610<pre>
Chandler Carruth48b0bbf2011-12-12 04:36:04 +0000[diff] [blame]7611 declare i8 @llvm.cttz.i8 (i8 &lt;src&gt;, i1 &lt;is_zero_undef&gt;)
7612 declare i16 @llvm.cttz.i16 (i16 &lt;src&gt;, i1 &lt;is_zero_undef&gt;)
7613 declare i32 @llvm.cttz.i32 (i32 &lt;src&gt;, i1 &lt;is_zero_undef&gt;)
7614 declare i64 @llvm.cttz.i64 (i64 &lt;src&gt;, i1 &lt;is_zero_undef&gt;)
7615 declare i256 @llvm.cttz.i256(i256 &lt;src&gt;, i1 &lt;is_zero_undef&gt;)
7616 declase &lt;2 x i32&gt; @llvm.cttz.v2i32(&lt;2 x i32&gt; &lt;src&gt;, i1 &lt;is_zero_undef&gt;)
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]7617</pre>
7618
7619<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7620<p>The '<tt>llvm.cttz</tt>' family of intrinsic functions counts the number of
7621 trailing zeros.</p>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]7622
7623<h5>Arguments:</h5>
Chandler Carruth48b0bbf2011-12-12 04:36:04 +0000[diff] [blame]7624<p>The first argument is the value to be counted. This argument may be of any
7625 integer type, or a vectory with integer element type. The return type
7626 must match the first argument type.</p>
7627
7628<p>The second argument must be a constant and is a flag to indicate whether the
7629 intrinsic should ensure that a zero as the first argument produces a defined
7630 result. Historically some architectures did not provide a defined result for
7631 zero values as efficiently, and many algorithms are now predicated on
7632 avoiding zero-value inputs.</p>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]7633
7634<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7635<p>The '<tt>llvm.cttz</tt>' intrinsic counts the trailing (least significant)
Owen Andersonf1ac4652011-07-01 21:52:38 +0000[diff] [blame]7636 zeros in a variable, or within each element of a vector.
Chandler Carruth48b0bbf2011-12-12 04:36:04 +0000[diff] [blame]7637 If <tt>src == 0</tt> then the result is the size in bits of the type of
7638 <tt>src</tt> if <tt>is_zero_undef == 0</tt> and <tt>undef</tt> otherwise.
7639 For example, <tt>llvm.cttz(2) = 1</tt>.</p>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]7640
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]7641</div>
7642
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7643</div>
7644
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]7645<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7646<h3>
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]7647 <a name="int_overflow">Arithmetic with Overflow Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7648</h3>
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]7649
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7650<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7651
7652<p>LLVM provides intrinsics for some arithmetic with overflow operations.</p>
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]7653
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7654<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7655<h4>
7656 <a name="int_sadd_overflow">
7657 '<tt>llvm.sadd.with.overflow.*</tt>' Intrinsics
7658 </a>
7659</h4>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7660
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7661<div>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7662
7663<h5>Syntax:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7664<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]7665 on any integer bit width.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7666
7667<pre>
7668 declare {i16, i1} @llvm.sadd.with.overflow.i16(i16 %a, i16 %b)
7669 declare {i32, i1} @llvm.sadd.with.overflow.i32(i32 %a, i32 %b)
7670 declare {i64, i1} @llvm.sadd.with.overflow.i64(i64 %a, i64 %b)
7671</pre>
7672
7673<h5>Overview:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7674<p>The '<tt>llvm.sadd.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7675 a signed addition of the two arguments, and indicate whether an overflow
7676 occurred during the signed summation.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7677
7678<h5>Arguments:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7679<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]7680 be of integer types of any bit width, but they must have the same bit
7681 width. The second element of the result structure must be of
7682 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
7683 undergo signed addition.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7684
7685<h5>Semantics:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7686<p>The '<tt>llvm.sadd.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7687 a signed addition of the two variables. They return a structure &mdash; the
7688 first element of which is the signed summation, and the second element of
7689 which is a bit specifying if the signed summation resulted in an
7690 overflow.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7691
7692<h5>Examples:</h5>
7693<pre>
7694 %res = call {i32, i1} @llvm.sadd.with.overflow.i32(i32 %a, i32 %b)
7695 %sum = extractvalue {i32, i1} %res, 0
7696 %obit = extractvalue {i32, i1} %res, 1
7697 br i1 %obit, label %overflow, label %normal
7698</pre>
7699
7700</div>
7701
7702<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7703<h4>
7704 <a name="int_uadd_overflow">
7705 '<tt>llvm.uadd.with.overflow.*</tt>' Intrinsics
7706 </a>
7707</h4>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7708
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7709<div>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7710
7711<h5>Syntax:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7712<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]7713 on any integer bit width.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7714
7715<pre>
7716 declare {i16, i1} @llvm.uadd.with.overflow.i16(i16 %a, i16 %b)
7717 declare {i32, i1} @llvm.uadd.with.overflow.i32(i32 %a, i32 %b)
7718 declare {i64, i1} @llvm.uadd.with.overflow.i64(i64 %a, i64 %b)
7719</pre>
7720
7721<h5>Overview:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7722<p>The '<tt>llvm.uadd.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7723 an unsigned addition of the two arguments, and indicate whether a carry
7724 occurred during the unsigned summation.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7725
7726<h5>Arguments:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7727<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]7728 be of integer types of any bit width, but they must have the same bit
7729 width. The second element of the result structure must be of
7730 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
7731 undergo unsigned addition.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7732
7733<h5>Semantics:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7734<p>The '<tt>llvm.uadd.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7735 an unsigned addition of the two arguments. They return a structure &mdash;
7736 the first element of which is the sum, and the second element of which is a
7737 bit specifying if the unsigned summation resulted in a carry.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7738
7739<h5>Examples:</h5>
7740<pre>
7741 %res = call {i32, i1} @llvm.uadd.with.overflow.i32(i32 %a, i32 %b)
7742 %sum = extractvalue {i32, i1} %res, 0
7743 %obit = extractvalue {i32, i1} %res, 1
7744 br i1 %obit, label %carry, label %normal
7745</pre>
7746
7747</div>
7748
7749<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7750<h4>
7751 <a name="int_ssub_overflow">
7752 '<tt>llvm.ssub.with.overflow.*</tt>' Intrinsics
7753 </a>
7754</h4>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7755
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7756<div>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7757
7758<h5>Syntax:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7759<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]7760 on any integer bit width.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7761
7762<pre>
7763 declare {i16, i1} @llvm.ssub.with.overflow.i16(i16 %a, i16 %b)
7764 declare {i32, i1} @llvm.ssub.with.overflow.i32(i32 %a, i32 %b)
7765 declare {i64, i1} @llvm.ssub.with.overflow.i64(i64 %a, i64 %b)
7766</pre>
7767
7768<h5>Overview:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7769<p>The '<tt>llvm.ssub.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7770 a signed subtraction of the two arguments, and indicate whether an overflow
7771 occurred during the signed subtraction.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7772
7773<h5>Arguments:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7774<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]7775 be of integer types of any bit width, but they must have the same bit
7776 width. The second element of the result structure must be of
7777 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
7778 undergo signed subtraction.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7779
7780<h5>Semantics:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7781<p>The '<tt>llvm.ssub.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7782 a signed subtraction of the two arguments. They return a structure &mdash;
7783 the first element of which is the subtraction, and the second element of
7784 which is a bit specifying if the signed subtraction resulted in an
7785 overflow.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7786
7787<h5>Examples:</h5>
7788<pre>
7789 %res = call {i32, i1} @llvm.ssub.with.overflow.i32(i32 %a, i32 %b)
7790 %sum = extractvalue {i32, i1} %res, 0
7791 %obit = extractvalue {i32, i1} %res, 1
7792 br i1 %obit, label %overflow, label %normal
7793</pre>
7794
7795</div>
7796
7797<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7798<h4>
7799 <a name="int_usub_overflow">
7800 '<tt>llvm.usub.with.overflow.*</tt>' Intrinsics
7801 </a>
7802</h4>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7803
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7804<div>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7805
7806<h5>Syntax:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7807<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]7808 on any integer bit width.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7809
7810<pre>
7811 declare {i16, i1} @llvm.usub.with.overflow.i16(i16 %a, i16 %b)
7812 declare {i32, i1} @llvm.usub.with.overflow.i32(i32 %a, i32 %b)
7813 declare {i64, i1} @llvm.usub.with.overflow.i64(i64 %a, i64 %b)
7814</pre>
7815
7816<h5>Overview:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7817<p>The '<tt>llvm.usub.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7818 an unsigned subtraction of the two arguments, and indicate whether an
7819 overflow occurred during the unsigned subtraction.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7820
7821<h5>Arguments:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7822<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]7823 be of integer types of any bit width, but they must have the same bit
7824 width. The second element of the result structure must be of
7825 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
7826 undergo unsigned subtraction.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7827
7828<h5>Semantics:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7829<p>The '<tt>llvm.usub.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7830 an unsigned subtraction of the two arguments. They return a structure &mdash;
7831 the first element of which is the subtraction, and the second element of
7832 which is a bit specifying if the unsigned subtraction resulted in an
7833 overflow.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7834
7835<h5>Examples:</h5>
7836<pre>
7837 %res = call {i32, i1} @llvm.usub.with.overflow.i32(i32 %a, i32 %b)
7838 %sum = extractvalue {i32, i1} %res, 0
7839 %obit = extractvalue {i32, i1} %res, 1
7840 br i1 %obit, label %overflow, label %normal
7841</pre>
7842
7843</div>
7844
7845<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7846<h4>
7847 <a name="int_smul_overflow">
7848 '<tt>llvm.smul.with.overflow.*</tt>' Intrinsics
7849 </a>
7850</h4>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7851
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7852<div>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7853
7854<h5>Syntax:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7855<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]7856 on any integer bit width.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7857
7858<pre>
7859 declare {i16, i1} @llvm.smul.with.overflow.i16(i16 %a, i16 %b)
7860 declare {i32, i1} @llvm.smul.with.overflow.i32(i32 %a, i32 %b)
7861 declare {i64, i1} @llvm.smul.with.overflow.i64(i64 %a, i64 %b)
7862</pre>
7863
7864<h5>Overview:</h5>
7865
7866<p>The '<tt>llvm.smul.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7867 a signed multiplication of the two arguments, and indicate whether an
7868 overflow occurred during the signed multiplication.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7869
7870<h5>Arguments:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7871<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]7872 be of integer types of any bit width, but they must have the same bit
7873 width. The second element of the result structure must be of
7874 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
7875 undergo signed multiplication.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7876
7877<h5>Semantics:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7878<p>The '<tt>llvm.smul.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7879 a signed multiplication of the two arguments. They return a structure &mdash;
7880 the first element of which is the multiplication, and the second element of
7881 which is a bit specifying if the signed multiplication resulted in an
7882 overflow.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]7883
7884<h5>Examples:</h5>
7885<pre>
7886 %res = call {i32, i1} @llvm.smul.with.overflow.i32(i32 %a, i32 %b)
7887 %sum = extractvalue {i32, i1} %res, 0
7888 %obit = extractvalue {i32, i1} %res, 1
7889 br i1 %obit, label %overflow, label %normal
7890</pre>
7891
Reid Spencerf86037f2007-04-11 23:23:49 +0000[diff] [blame]7892</div>
7893
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7894<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7895<h4>
7896 <a name="int_umul_overflow">
7897 '<tt>llvm.umul.with.overflow.*</tt>' Intrinsics
7898 </a>
7899</h4>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7900
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7901<div>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7902
7903<h5>Syntax:</h5>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7904<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]7905 on any integer bit width.</p>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7906
7907<pre>
7908 declare {i16, i1} @llvm.umul.with.overflow.i16(i16 %a, i16 %b)
7909 declare {i32, i1} @llvm.umul.with.overflow.i32(i32 %a, i32 %b)
7910 declare {i64, i1} @llvm.umul.with.overflow.i64(i64 %a, i64 %b)
7911</pre>
7912
7913<h5>Overview:</h5>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7914<p>The '<tt>llvm.umul.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7915 a unsigned multiplication of the two arguments, and indicate whether an
7916 overflow occurred during the unsigned multiplication.</p>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7917
7918<h5>Arguments:</h5>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7919<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]7920 be of integer types of any bit width, but they must have the same bit
7921 width. The second element of the result structure must be of
7922 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
7923 undergo unsigned multiplication.</p>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7924
7925<h5>Semantics:</h5>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7926<p>The '<tt>llvm.umul.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]7927 an unsigned multiplication of the two arguments. They return a structure
7928 &mdash; the first element of which is the multiplication, and the second
7929 element of which is a bit specifying if the unsigned multiplication resulted
7930 in an overflow.</p>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]7931
7932<h5>Examples:</h5>
7933<pre>
7934 %res = call {i32, i1} @llvm.umul.with.overflow.i32(i32 %a, i32 %b)
7935 %sum = extractvalue {i32, i1} %res, 0
7936 %obit = extractvalue {i32, i1} %res, 1
7937 br i1 %obit, label %overflow, label %normal
7938</pre>
7939
7940</div>
7941
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7942</div>
7943
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]7944<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7945<h3>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]7946 <a name="int_fp16">Half Precision Floating Point Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7947</h3>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]7948
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7949<div>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]7950
Tobias Grosser057beb82012-05-24 15:59:06 +0000[diff] [blame]7951<p>For most target platforms, half precision floating point is a storage-only
7952 format. This means that it is
Chris Lattner0cec9c82010-03-15 04:12:21 +0000[diff] [blame]7953 a dense encoding (in memory) but does not support computation in the
7954 format.</p>
Chris Lattner82c3dc62010-03-14 23:03:31 +0000[diff] [blame]7955
Chris Lattner0cec9c82010-03-15 04:12:21 +0000[diff] [blame]7956<p>This means that code must first load the half-precision floating point
Chris Lattner82c3dc62010-03-14 23:03:31 +0000[diff] [blame]7957 value as an i16, then convert it to float with <a
7958 href="#int_convert_from_fp16"><tt>llvm.convert.from.fp16</tt></a>.
7959 Computation can then be performed on the float value (including extending to
Chris Lattner0cec9c82010-03-15 04:12:21 +0000[diff] [blame]7960 double etc). To store the value back to memory, it is first converted to
7961 float if needed, then converted to i16 with
Chris Lattner82c3dc62010-03-14 23:03:31 +0000[diff] [blame]7962 <a href="#int_convert_to_fp16"><tt>llvm.convert.to.fp16</tt></a>, then
7963 storing as an i16 value.</p>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]7964
7965<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]7966<h4>
7967 <a name="int_convert_to_fp16">
7968 '<tt>llvm.convert.to.fp16</tt>' Intrinsic
7969 </a>
7970</h4>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]7971
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]7972<div>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]7973
7974<h5>Syntax:</h5>
7975<pre>
7976 declare i16 @llvm.convert.to.fp16(f32 %a)
7977</pre>
7978
7979<h5>Overview:</h5>
7980<p>The '<tt>llvm.convert.to.fp16</tt>' intrinsic function performs
7981 a conversion from single precision floating point format to half precision
7982 floating point format.</p>
7983
7984<h5>Arguments:</h5>
7985<p>The intrinsic function contains single argument - the value to be
7986 converted.</p>
7987
7988<h5>Semantics:</h5>
7989<p>The '<tt>llvm.convert.to.fp16</tt>' intrinsic function performs
7990 a conversion from single precision floating point format to half precision
Chris Lattner0cec9c82010-03-15 04:12:21 +0000[diff] [blame]7991 floating point format. The return value is an <tt>i16</tt> which
Chris Lattner82c3dc62010-03-14 23:03:31 +0000[diff] [blame]7992 contains the converted number.</p>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]7993
7994<h5>Examples:</h5>
7995<pre>
7996 %res = call i16 @llvm.convert.to.fp16(f32 %a)
7997 store i16 %res, i16* @x, align 2
7998</pre>
7999
8000</div>
8001
8002<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8003<h4>
8004 <a name="int_convert_from_fp16">
8005 '<tt>llvm.convert.from.fp16</tt>' Intrinsic
8006 </a>
8007</h4>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]8008
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8009<div>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]8010
8011<h5>Syntax:</h5>
8012<pre>
8013 declare f32 @llvm.convert.from.fp16(i16 %a)
8014</pre>
8015
8016<h5>Overview:</h5>
8017<p>The '<tt>llvm.convert.from.fp16</tt>' intrinsic function performs
8018 a conversion from half precision floating point format to single precision
8019 floating point format.</p>
8020
8021<h5>Arguments:</h5>
8022<p>The intrinsic function contains single argument - the value to be
8023 converted.</p>
8024
8025<h5>Semantics:</h5>
8026<p>The '<tt>llvm.convert.from.fp16</tt>' intrinsic function performs a
Chris Lattner0cec9c82010-03-15 04:12:21 +0000[diff] [blame]8027 conversion from half single precision floating point format to single
Chris Lattner82c3dc62010-03-14 23:03:31 +0000[diff] [blame]8028 precision floating point format. The input half-float value is represented by
8029 an <tt>i16</tt> value.</p>
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]8030
8031<h5>Examples:</h5>
8032<pre>
8033 %a = load i16* @x, align 2
8034 %res = call f32 @llvm.convert.from.fp16(i16 %a)
8035</pre>
8036
8037</div>
8038
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8039</div>
8040
Anton Korobeynikovf02e7302010-03-14 18:42:47 +0000[diff] [blame]8041<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8042<h3>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]8043 <a name="int_debugger">Debugger Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8044</h3>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]8045
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8046<div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]8047
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8048<p>The LLVM debugger intrinsics (which all start with <tt>llvm.dbg.</tt>
8049 prefix), are described in
8050 the <a href="SourceLevelDebugging.html#format_common_intrinsics">LLVM Source
8051 Level Debugging</a> document.</p>
8052
8053</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]8054
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]8055<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8056<h3>
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]8057 <a name="int_eh">Exception Handling Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8058</h3>
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]8059
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8060<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8061
8062<p>The LLVM exception handling intrinsics (which all start with
8063 <tt>llvm.eh.</tt> prefix), are described in
8064 the <a href="ExceptionHandling.html#format_common_intrinsics">LLVM Exception
8065 Handling</a> document.</p>
8066
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]8067</div>
8068
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]8069<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8070<h3>
Duncan Sands4a544a72011-09-06 13:37:06 +0000[diff] [blame]8071 <a name="int_trampoline">Trampoline Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8072</h3>
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]8073
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8074<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8075
Duncan Sands4a544a72011-09-06 13:37:06 +0000[diff] [blame]8076<p>These intrinsics make it possible to excise one parameter, marked with
Dan Gohmanff235352010-07-02 23:18:08 +0000[diff] [blame]8077 the <a href="#nest"><tt>nest</tt></a> attribute, from a function.
8078 The result is a callable
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8079 function pointer lacking the nest parameter - the caller does not need to
8080 provide a value for it. Instead, the value to use is stored in advance in a
8081 "trampoline", a block of memory usually allocated on the stack, which also
8082 contains code to splice the nest value into the argument list. This is used
8083 to implement the GCC nested function address extension.</p>
8084
8085<p>For example, if the function is
8086 <tt>i32 f(i8* nest %c, i32 %x, i32 %y)</tt> then the resulting function
8087 pointer has signature <tt>i32 (i32, i32)*</tt>. It can be created as
8088 follows:</p>
8089
Benjamin Kramer26fe25f2010-07-13 12:26:09 +0000[diff] [blame]8090<pre class="doc_code">
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]8091 %tramp = alloca [10 x i8], align 4 ; size and alignment only correct for X86
8092 %tramp1 = getelementptr [10 x i8]* %tramp, i32 0, i32 0
Duncan Sands4a544a72011-09-06 13:37:06 +0000[diff] [blame]8093 call i8* @llvm.init.trampoline(i8* %tramp1, i8* bitcast (i32 (i8*, i32, i32)* @f to i8*), i8* %nval)
8094 %p = call i8* @llvm.adjust.trampoline(i8* %tramp1)
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]8095 %fp = bitcast i8* %p to i32 (i32, i32)*
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]8096</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8097
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]8098<p>The call <tt>%val = call i32 %fp(i32 %x, i32 %y)</tt> is then equivalent
8099 to <tt>%val = call i32 %f(i8* %nval, i32 %x, i32 %y)</tt>.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8100
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]8101<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8102<h4>
8103 <a name="int_it">
8104 '<tt>llvm.init.trampoline</tt>' Intrinsic
8105 </a>
8106</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8107
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8108<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8109
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]8110<h5>Syntax:</h5>
8111<pre>
Duncan Sands4a544a72011-09-06 13:37:06 +0000[diff] [blame]8112 declare void @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]8113</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8114
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]8115<h5>Overview:</h5>
Duncan Sands4a544a72011-09-06 13:37:06 +0000[diff] [blame]8116<p>This fills the memory pointed to by <tt>tramp</tt> with executable code,
8117 turning it into a trampoline.</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8118
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]8119<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8120<p>The <tt>llvm.init.trampoline</tt> intrinsic takes three arguments, all
8121 pointers. The <tt>tramp</tt> argument must point to a sufficiently large and
8122 sufficiently aligned block of memory; this memory is written to by the
8123 intrinsic. Note that the size and the alignment are target-specific - LLVM
8124 currently provides no portable way of determining them, so a front-end that
8125 generates this intrinsic needs to have some target-specific knowledge.
8126 The <tt>func</tt> argument must hold a function bitcast to
8127 an <tt>i8*</tt>.</p>
8128
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]8129<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8130<p>The block of memory pointed to by <tt>tramp</tt> is filled with target
Duncan Sands4a544a72011-09-06 13:37:06 +0000[diff] [blame]8131 dependent code, turning it into a function. Then <tt>tramp</tt> needs to be
8132 passed to <a href="#int_at">llvm.adjust.trampoline</a> to get a pointer
8133 which can be <a href="#int_trampoline">bitcast (to a new function) and
8134 called</a>. The new function's signature is the same as that of
8135 <tt>func</tt> with any arguments marked with the <tt>nest</tt> attribute
8136 removed. At most one such <tt>nest</tt> argument is allowed, and it must be of
8137 pointer type. Calling the new function is equivalent to calling <tt>func</tt>
8138 with the same argument list, but with <tt>nval</tt> used for the missing
8139 <tt>nest</tt> argument. If, after calling <tt>llvm.init.trampoline</tt>, the
8140 memory pointed to by <tt>tramp</tt> is modified, then the effect of any later call
8141 to the returned function pointer is undefined.</p>
8142</div>
8143
8144<!-- _______________________________________________________________________ -->
8145<h4>
8146 <a name="int_at">
8147 '<tt>llvm.adjust.trampoline</tt>' Intrinsic
8148 </a>
8149</h4>
8150
8151<div>
8152
8153<h5>Syntax:</h5>
8154<pre>
8155 declare i8* @llvm.adjust.trampoline(i8* &lt;tramp&gt;)
8156</pre>
8157
8158<h5>Overview:</h5>
8159<p>This performs any required machine-specific adjustment to the address of a
8160 trampoline (passed as <tt>tramp</tt>).</p>
8161
8162<h5>Arguments:</h5>
8163<p><tt>tramp</tt> must point to a block of memory which already has trampoline code
8164 filled in by a previous call to <a href="#int_it"><tt>llvm.init.trampoline</tt>
8165 </a>.</p>
8166
8167<h5>Semantics:</h5>
8168<p>On some architectures the address of the code to be executed needs to be
8169 different to the address where the trampoline is actually stored. This
8170 intrinsic returns the executable address corresponding to <tt>tramp</tt>
8171 after performing the required machine specific adjustments.
8172 The pointer returned can then be <a href="#int_trampoline"> bitcast and
8173 executed</a>.
8174</p>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8175
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]8176</div>
8177
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8178</div>
8179
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]8180<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8181<h3>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8182 <a name="int_memorymarkers">Memory Use Markers</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8183</h3>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8184
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8185<div>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8186
8187<p>This class of intrinsics exists to information about the lifetime of memory
8188 objects and ranges where variables are immutable.</p>
8189
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8190<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8191<h4>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8192 <a name="int_lifetime_start">'<tt>llvm.lifetime.start</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8193</h4>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8194
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8195<div>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8196
8197<h5>Syntax:</h5>
8198<pre>
8199 declare void @llvm.lifetime.start(i64 &lt;size&gt;, i8* nocapture &lt;ptr&gt;)
8200</pre>
8201
8202<h5>Overview:</h5>
8203<p>The '<tt>llvm.lifetime.start</tt>' intrinsic specifies the start of a memory
8204 object's lifetime.</p>
8205
8206<h5>Arguments:</h5>
Nick Lewycky321333e2009-10-13 07:57:33 +0000[diff] [blame]8207<p>The first argument is a constant integer representing the size of the
8208 object, or -1 if it is variable sized. The second argument is a pointer to
8209 the object.</p>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8210
8211<h5>Semantics:</h5>
8212<p>This intrinsic indicates that before this point in the code, the value of the
8213 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]8214 never be used and has an undefined value. A load from the pointer that
8215 precedes this intrinsic can be replaced with
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8216 <tt>'<a href="#undefvalues">undef</a>'</tt>.</p>
8217
8218</div>
8219
8220<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8221<h4>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8222 <a name="int_lifetime_end">'<tt>llvm.lifetime.end</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8223</h4>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8224
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8225<div>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8226
8227<h5>Syntax:</h5>
8228<pre>
8229 declare void @llvm.lifetime.end(i64 &lt;size&gt;, i8* nocapture &lt;ptr&gt;)
8230</pre>
8231
8232<h5>Overview:</h5>
8233<p>The '<tt>llvm.lifetime.end</tt>' intrinsic specifies the end of a memory
8234 object's lifetime.</p>
8235
8236<h5>Arguments:</h5>
Nick Lewycky321333e2009-10-13 07:57:33 +0000[diff] [blame]8237<p>The first argument is a constant integer representing the size of the
8238 object, or -1 if it is variable sized. The second argument is a pointer to
8239 the object.</p>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8240
8241<h5>Semantics:</h5>
8242<p>This intrinsic indicates that after this point in the code, the value of the
8243 memory pointed to by <tt>ptr</tt> is dead. This means that it is known to
8244 never be used and has an undefined value. Any stores into the memory object
8245 following this intrinsic may be removed as dead.
8246
8247</div>
8248
8249<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8250<h4>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8251 <a name="int_invariant_start">'<tt>llvm.invariant.start</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8252</h4>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8253
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8254<div>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8255
8256<h5>Syntax:</h5>
8257<pre>
Nick Lewycky29b6cb42010-11-30 04:13:41 +0000[diff] [blame]8258 declare {}* @llvm.invariant.start(i64 &lt;size&gt;, i8* nocapture &lt;ptr&gt;)
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8259</pre>
8260
8261<h5>Overview:</h5>
8262<p>The '<tt>llvm.invariant.start</tt>' intrinsic specifies that the contents of
8263 a memory object will not change.</p>
8264
8265<h5>Arguments:</h5>
Nick Lewycky321333e2009-10-13 07:57:33 +0000[diff] [blame]8266<p>The first argument is a constant integer representing the size of the
8267 object, or -1 if it is variable sized. The second argument is a pointer to
8268 the object.</p>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8269
8270<h5>Semantics:</h5>
8271<p>This intrinsic indicates that until an <tt>llvm.invariant.end</tt> that uses
8272 the return value, the referenced memory location is constant and
8273 unchanging.</p>
8274
8275</div>
8276
8277<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8278<h4>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8279 <a name="int_invariant_end">'<tt>llvm.invariant.end</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8280</h4>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8281
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8282<div>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8283
8284<h5>Syntax:</h5>
8285<pre>
8286 declare void @llvm.invariant.end({}* &lt;start&gt;, i64 &lt;size&gt;, i8* nocapture &lt;ptr&gt;)
8287</pre>
8288
8289<h5>Overview:</h5>
8290<p>The '<tt>llvm.invariant.end</tt>' intrinsic specifies that the contents of
8291 a memory object are mutable.</p>
8292
8293<h5>Arguments:</h5>
8294<p>The first argument is the matching <tt>llvm.invariant.start</tt> intrinsic.
Nick Lewycky321333e2009-10-13 07:57:33 +0000[diff] [blame]8295 The second argument is a constant integer representing the size of the
8296 object, or -1 if it is variable sized and the third argument is a pointer
8297 to the object.</p>
Nick Lewyckycc271862009-10-13 07:03:23 +0000[diff] [blame]8298
8299<h5>Semantics:</h5>
8300<p>This intrinsic indicates that the memory is mutable again.</p>
8301
8302</div>
8303
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8304</div>
8305
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]8306<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8307<h3>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]8308 <a name="int_general">General Intrinsics</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8309</h3>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]8310
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8311<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8312
8313<p>This class of intrinsics is designed to be generic and has no specific
8314 purpose.</p>
8315
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]8316<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8317<h4>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]8318 <a name="int_var_annotation">'<tt>llvm.var.annotation</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8319</h4>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]8320
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8321<div>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]8322
8323<h5>Syntax:</h5>
8324<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]8325 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]8326</pre>
8327
8328<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8329<p>The '<tt>llvm.var.annotation</tt>' intrinsic.</p>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]8330
8331<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8332<p>The first argument is a pointer to a value, the second is a pointer to a
8333 global string, the third is a pointer to a global string which is the source
8334 file name, and the last argument is the line number.</p>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]8335
8336<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8337<p>This intrinsic allows annotation of local variables with arbitrary strings.
8338 This can be useful for special purpose optimizations that want to look for
John Criswelle865c032011-08-19 16:57:55 +0000[diff] [blame]8339 these annotations. These have no other defined use; they are ignored by code
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8340 generation and optimization.</p>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]8341
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]8342</div>
8343
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]8344<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8345<h4>
Tanya Lattnere1a8da02007-09-21 23:57:59 +0000[diff] [blame]8346 <a name="int_annotation">'<tt>llvm.annotation.*</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8347</h4>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]8348
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8349<div>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]8350
8351<h5>Syntax:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8352<p>This is an overloaded intrinsic. You can use '<tt>llvm.annotation</tt>' on
8353 any integer bit width.</p>
8354
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]8355<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]8356 declare i8 @llvm.annotation.i8(i8 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt;)
8357 declare i16 @llvm.annotation.i16(i16 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt;)
8358 declare i32 @llvm.annotation.i32(i32 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt;)
8359 declare i64 @llvm.annotation.i64(i64 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt;)
8360 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]8361</pre>
8362
8363<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8364<p>The '<tt>llvm.annotation</tt>' intrinsic.</p>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]8365
8366<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8367<p>The first argument is an integer value (result of some expression), the
8368 second is a pointer to a global string, the third is a pointer to a global
8369 string which is the source file name, and the last argument is the line
8370 number. It returns the value of the first argument.</p>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]8371
8372<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8373<p>This intrinsic allows annotations to be put on arbitrary expressions with
8374 arbitrary strings. This can be useful for special purpose optimizations that
John Criswelle865c032011-08-19 16:57:55 +0000[diff] [blame]8375 want to look for these annotations. These have no other defined use; they
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8376 are ignored by code generation and optimization.</p>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]8377
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]8378</div>
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]8379
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]8380<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8381<h4>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]8382 <a name="int_trap">'<tt>llvm.trap</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8383</h4>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]8384
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8385<div>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]8386
8387<h5>Syntax:</h5>
8388<pre>
8389 declare void @llvm.trap()
8390</pre>
8391
8392<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8393<p>The '<tt>llvm.trap</tt>' intrinsic.</p>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]8394
8395<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8396<p>None.</p>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]8397
8398<h5>Semantics:</h5>
John Criswell99534f32012-05-16 00:26:51 +0000[diff] [blame]8399<p>This intrinsic is lowered to the target dependent trap instruction. If the
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8400 target does not have a trap instruction, this intrinsic will be lowered to
John Criswell99534f32012-05-16 00:26:51 +0000[diff] [blame]8401 a call of the <tt>abort()</tt> function.</p>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]8402
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]8403</div>
8404
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]8405<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8406<h4>
Dan Gohmana6063c62012-05-14 18:58:10 +0000[diff] [blame]8407 <a name="int_debugtrap">'<tt>llvm.debugtrap</tt>' Intrinsic</a>
Dan Gohmand4347e12012-05-11 00:19:32 +0000[diff] [blame]8408</h4>
8409
8410<div>
8411
8412<h5>Syntax:</h5>
8413<pre>
Dan Gohmana6063c62012-05-14 18:58:10 +0000[diff] [blame]8414 declare void @llvm.debugtrap()
Dan Gohmand4347e12012-05-11 00:19:32 +0000[diff] [blame]8415</pre>
8416
8417<h5>Overview:</h5>
Dan Gohmana6063c62012-05-14 18:58:10 +0000[diff] [blame]8418<p>The '<tt>llvm.debugtrap</tt>' intrinsic.</p>
Dan Gohmand4347e12012-05-11 00:19:32 +0000[diff] [blame]8419
8420<h5>Arguments:</h5>
8421<p>None.</p>
8422
8423<h5>Semantics:</h5>
8424<p>This intrinsic is lowered to code which is intended to cause an execution
8425 trap with the intention of requesting the attention of a debugger.</p>
8426
8427</div>
8428
8429<!-- _______________________________________________________________________ -->
8430<h4>
Misha Brukmandccb0252008-11-22 23:55:29 +0000[diff] [blame]8431 <a name="int_stackprotector">'<tt>llvm.stackprotector</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8432</h4>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8433
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8434<div>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8435
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]8436<h5>Syntax:</h5>
8437<pre>
Dan Gohman3dfb3cf2010-05-28 17:07:41 +0000[diff] [blame]8438 declare void @llvm.stackprotector(i8* &lt;guard&gt;, i8** &lt;slot&gt;)
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]8439</pre>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8440
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]8441<h5>Overview:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8442<p>The <tt>llvm.stackprotector</tt> intrinsic takes the <tt>guard</tt> and
8443 stores it onto the stack at <tt>slot</tt>. The stack slot is adjusted to
8444 ensure that it is placed on the stack before local variables.</p>
8445
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]8446<h5>Arguments:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8447<p>The <tt>llvm.stackprotector</tt> intrinsic requires two pointer
8448 arguments. The first argument is the value loaded from the stack
8449 guard <tt>@__stack_chk_guard</tt>. The second variable is an <tt>alloca</tt>
8450 that has enough space to hold the value of the guard.</p>
8451
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]8452<h5>Semantics:</h5>
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8453<p>This intrinsic causes the prologue/epilogue inserter to force the position of
8454 the <tt>AllocaInst</tt> stack slot to be before local variables on the
8455 stack. This is to ensure that if a local variable on the stack is
8456 overwritten, it will destroy the value of the guard. When the function exits,
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]8457 the guard on the stack is checked against the original guard. If they are
Bill Wendlinge910b4c2009-07-20 02:29:24 +0000[diff] [blame]8458 different, then the program aborts by calling the <tt>__stack_chk_fail()</tt>
8459 function.</p>
8460
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]8461</div>
8462
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]8463<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8464<h4>
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]8465 <a name="int_objectsize">'<tt>llvm.objectsize</tt>' Intrinsic</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]8466</h4>
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]8467
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8468<div>
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]8469
8470<h5>Syntax:</h5>
8471<pre>
Nuno Lopes23e75da2012-05-22 15:25:31 +0000[diff] [blame]8472 declare i32 @llvm.objectsize.i32(i8* &lt;object&gt;, i1 &lt;min&gt;)
8473 declare i64 @llvm.objectsize.i64(i8* &lt;object&gt;, i1 &lt;min&gt;)
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]8474</pre>
8475
8476<h5>Overview:</h5>
Bill Wendling1b383ba2010-10-27 01:07:41 +0000[diff] [blame]8477<p>The <tt>llvm.objectsize</tt> intrinsic is designed to provide information to
8478 the optimizers to determine at compile time whether a) an operation (like
8479 memcpy) will overflow a buffer that corresponds to an object, or b) that a
8480 runtime check for overflow isn't necessary. An object in this context means
8481 an allocation of a specific class, structure, array, or other object.</p>
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]8482
8483<h5>Arguments:</h5>
Nuno Lopes23e75da2012-05-22 15:25:31 +0000[diff] [blame]8484<p>The <tt>llvm.objectsize</tt> intrinsic takes two arguments. The first
Eric Christopher8295a0a2009-12-23 00:29:49 +0000[diff] [blame]8485 argument is a pointer to or into the <tt>object</tt>. The second argument
Nuno Lopes23e75da2012-05-22 15:25:31 +0000[diff] [blame]8486 is a boolean and determines whether <tt>llvm.objectsize</tt> returns 0 (if
8487 true) or -1 (if false) when the object size is unknown.
8488 The second argument only accepts constants.</p>
Eric Christopher8295a0a2009-12-23 00:29:49 +0000[diff] [blame]8489
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]8490<h5>Semantics:</h5>
Nuno Lopes30759542012-05-09 15:52:43 +0000[diff] [blame]8491<p>The <tt>llvm.objectsize</tt> intrinsic is lowered to a constant representing
8492 the size of the object concerned. If the size cannot be determined at compile
Nuno Lopes23e75da2012-05-22 15:25:31 +0000[diff] [blame]8493 time, <tt>llvm.objectsize</tt> returns <tt>i32/i64 -1 or 0</tt>
8494 (depending on the <tt>min</tt> argument).</p>
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]8495
8496</div>
Jakub Staszakb170e2d2011-12-04 18:29:26 +0000[diff] [blame]8497<!-- _______________________________________________________________________ -->
8498<h4>
8499 <a name="int_expect">'<tt>llvm.expect</tt>' Intrinsic</a>
8500</h4>
Eric Christopher0e671492009-11-30 08:03:53 +0000[diff] [blame]8501
Jakub Staszakb170e2d2011-12-04 18:29:26 +0000[diff] [blame]8502<div>
8503
8504<h5>Syntax:</h5>
8505<pre>
8506 declare i32 @llvm.expect.i32(i32 &lt;val&gt;, i32 &lt;expected_val&gt;)
8507 declare i64 @llvm.expect.i64(i64 &lt;val&gt;, i64 &lt;expected_val&gt;)
8508</pre>
8509
8510<h5>Overview:</h5>
8511<p>The <tt>llvm.expect</tt> intrinsic provides information about expected (the
8512 most probable) value of <tt>val</tt>, which can be used by optimizers.</p>
8513
8514<h5>Arguments:</h5>
8515<p>The <tt>llvm.expect</tt> intrinsic takes two arguments. The first
8516 argument is a value. The second argument is an expected value, this needs to
8517 be a constant value, variables are not allowed.</p>
8518
8519<h5>Semantics:</h5>
8520<p>This intrinsic is lowered to the <tt>val</tt>.</p>
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]8521</div>
8522
8523</div>
8524
Jakub Staszakb170e2d2011-12-04 18:29:26 +0000[diff] [blame]8525</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]8526<!-- *********************************************************************** -->
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]8527<hr>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]8528<address>
8529 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
Misha Brukman44408702008-12-11 17:34:48 +0000[diff] [blame]8530 src="http://jigsaw.w3.org/css-validator/images/vcss-blue" alt="Valid CSS"></a>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]8531 <a href="http://validator.w3.org/check/referer"><img
Misha Brukman44408702008-12-11 17:34:48 +0000[diff] [blame]8532 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]8533
8534 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
NAKAMURA Takumib9a33632011-04-09 02:13:37 +0000[diff] [blame]8535 <a href="http://llvm.org/">The LLVM Compiler Infrastructure</a><br>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]8536 Last modified: $Date$
8537</address>
Chris Lattnerc7d3ab32008-01-04 04:33:49 +0000[diff] [blame]8538
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]8539</body>
8540</html>