| Misha Brukman | daa4cb0 | 2004-03-01 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 Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 3 | <html> |
| 4 | <head> |
| 5 | <title>LLVM Assembly Language Reference Manual</title> |
| Reid Spencer | 3921c74 | 2004-08-26 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 Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 8 | <meta name="description" |
| Reid Spencer | 3921c74 | 2004-08-26 20:44:00 +0000 | [diff] [blame] | 9 | content="LLVM Assembly Language Reference Manual."> |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 10 | <link rel="stylesheet" href="llvm.css" type="text/css"> |
| 11 | </head> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 12 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 13 | <body> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 14 | |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 15 | <h1>LLVM Language Reference Manual</h1> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 16 | <ol> |
| Misha Brukman | 9d0919f | 2003-11-08 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 Lattner | fa73021 | 2004-12-09 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 Wendling | 3d10a5a | 2009-07-20 01:03:30 +0000 | [diff] [blame] | 23 | <li><a href="#linkage">Linkage Types</a> |
| 24 | <ol> |
| Bill Wendling | 987e7eb | 2009-07-20 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 Wendling | 5e721d7 | 2010-07-01 21:55:59 +0000 | [diff] [blame] | 27 | <li><a href="#linkage_linker_private_weak">'<tt>linker_private_weak</tt>' Linkage</a></li> |
| Bill Wendling | 55ae515 | 2010-08-20 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 Wendling | 987e7eb | 2009-07-20 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 Lattner | 5a2d875 | 2009-10-10 18:26:06 +0000 | [diff] [blame] | 36 | <li><a href="#linkage_linkonce_odr">'<tt>linkonce_odr</tt>' Linkage</a></li> |
| Bill Wendling | 987e7eb | 2009-07-20 02:41:50 +0000 | [diff] [blame] | 37 | <li><a href="#linkage_weak">'<tt>weak_odr</tt>' Linkage</a></li> |
| Bill Wendling | f7f0610 | 2011-10-11 06:41:28 +0000 | [diff] [blame] | 38 | <li><a href="#linkage_external">'<tt>external</tt>' Linkage</a></li> |
| Bill Wendling | 987e7eb | 2009-07-20 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 Wendling | 3d10a5a | 2009-07-20 01:03:30 +0000 | [diff] [blame] | 41 | </ol> |
| 42 | </li> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 43 | <li><a href="#callingconv">Calling Conventions</a></li> |
| Chris Lattner | e7886e4 | 2009-01-11 20:53:49 +0000 | [diff] [blame] | 44 | <li><a href="#namedtypes">Named Types</a></li> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 45 | <li><a href="#globalvars">Global Variables</a></li> |
| Chris Lattner | 4e9aba7 | 2006-01-23 23:23:47 +0000 | [diff] [blame] | 46 | <li><a href="#functionstructure">Functions</a></li> |
| Dan Gohman | 0e451ce | 2008-10-14 16:51:45 +0000 | [diff] [blame] | 47 | <li><a href="#aliasstructure">Aliases</a></li> |
| Devang Patel | cd1fd25 | 2010-01-11 19:35:55 +0000 | [diff] [blame] | 48 | <li><a href="#namedmetadatastructure">Named Metadata</a></li> |
| Reid Spencer | ca86e16 | 2006-12-31 07:07:53 +0000 | [diff] [blame] | 49 | <li><a href="#paramattrs">Parameter Attributes</a></li> |
| Devang Patel | 2c9c3e7 | 2008-09-26 23:51:19 +0000 | [diff] [blame] | 50 | <li><a href="#fnattrs">Function Attributes</a></li> |
| Gordon Henriksen | 80a75bf | 2007-12-10 03:18:06 +0000 | [diff] [blame] | 51 | <li><a href="#gc">Garbage Collector Names</a></li> |
| Chris Lattner | 4e9aba7 | 2006-01-23 23:23:47 +0000 | [diff] [blame] | 52 | <li><a href="#moduleasm">Module-Level Inline Assembly</a></li> |
| Reid Spencer | de15194 | 2007-02-19 23:54:10 +0000 | [diff] [blame] | 53 | <li><a href="#datalayout">Data Layout</a></li> |
| Dan Gohman | 556ca27 | 2009-07-27 18:07:55 +0000 | [diff] [blame] | 54 | <li><a href="#pointeraliasing">Pointer Aliasing Rules</a></li> |
| Jeffrey Yasskin | 93e066d | 2010-04-26 21:21:24 +0000 | [diff] [blame] | 55 | <li><a href="#volatile">Volatile Memory Accesses</a></li> |
| Eli Friedman | 5b60e1b | 2011-07-20 21:35:53 +0000 | [diff] [blame] | 56 | <li><a href="#memmodel">Memory Model for Concurrent Operations</a></li> |
| Eli Friedman | ff03048 | 2011-07-28 21:48:00 +0000 | [diff] [blame] | 57 | <li><a href="#ordering">Atomic Memory Ordering Constraints</a></li> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 58 | </ol> |
| 59 | </li> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 60 | <li><a href="#typesystem">Type System</a> |
| 61 | <ol> |
| Chris Lattner | 4f69f46 | 2008-01-04 04:32:38 +0000 | [diff] [blame] | 62 | <li><a href="#t_classifications">Type Classifications</a></li> |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 63 | <li><a href="#t_primitive">Primitive Types</a> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 64 | <ol> |
| Nick Lewycky | ec38da4 | 2009-09-27 00:45:11 +0000 | [diff] [blame] | 65 | <li><a href="#t_integer">Integer Type</a></li> |
| Chris Lattner | 4f69f46 | 2008-01-04 04:32:38 +0000 | [diff] [blame] | 66 | <li><a href="#t_floating">Floating Point Types</a></li> |
| Dale Johannesen | 21fe99b | 2010-10-01 00:48:59 +0000 | [diff] [blame] | 67 | <li><a href="#t_x86mmx">X86mmx Type</a></li> |
| Chris Lattner | 4f69f46 | 2008-01-04 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 Lewycky | 7a0370f | 2009-05-30 05:06:04 +0000 | [diff] [blame] | 70 | <li><a href="#t_metadata">Metadata Type</a></li> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 71 | </ol> |
| 72 | </li> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 73 | <li><a href="#t_derived">Derived Types</a> |
| 74 | <ol> |
| Chris Lattner | fdfeb69 | 2010-02-12 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 Lattner | 628ed39 | 2011-07-23 19:59:08 +0000 | [diff] [blame] | 79 | <li><a href="#t_opaque">Opaque Structure Types</a></li> |
| Chris Lattner | fdfeb69 | 2010-02-12 20:49:41 +0000 | [diff] [blame] | 80 | <li><a href="#t_vector">Vector Type</a></li> |
| 81 | </ol> |
| 82 | </li> |
| Misha Brukman | 9d0919f | 2003-11-08 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 Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 85 | </ol> |
| 86 | </li> |
| 87 | </ol> |
| 88 | </li> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 89 | <li><a href="#constants">Constants</a> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 90 | <ol> |
| Dan Gohman | 0e451ce | 2008-10-14 16:51:45 +0000 | [diff] [blame] | 91 | <li><a href="#simpleconstants">Simple Constants</a></li> |
| Chris Lattner | 7088279 | 2009-02-28 18:32:25 +0000 | [diff] [blame] | 92 | <li><a href="#complexconstants">Complex Constants</a></li> |
| Dan Gohman | 0e451ce | 2008-10-14 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 Gohman | bfb056d | 2011-12-06 03:18:47 +0000 | [diff] [blame] | 95 | <li><a href="#poisonvalues">Poison Values</a></li> |
| Chris Lattner | f9d078e | 2009-10-27 21:19:13 +0000 | [diff] [blame] | 96 | <li><a href="#blockaddress">Addresses of Basic Blocks</a></li> |
| Dan Gohman | 0e451ce | 2008-10-14 16:51:45 +0000 | [diff] [blame] | 97 | <li><a href="#constantexprs">Constant Expressions</a></li> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 98 | </ol> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 99 | </li> |
| Chris Lattner | e87d653 | 2006-01-25 23:47:57 +0000 | [diff] [blame] | 100 | <li><a href="#othervalues">Other Values</a> |
| 101 | <ol> |
| Dan Gohman | 0e451ce | 2008-10-14 16:51:45 +0000 | [diff] [blame] | 102 | <li><a href="#inlineasm">Inline Assembler Expressions</a></li> |
| Peter Collingbourne | 249d953 | 2011-10-27 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> |
| Peter Collingbourne | 999f90b | 2011-10-27 19:19:14 +0000 | [diff] [blame] | 106 | <li><a href="#fpaccuracy">'<tt>fpaccuracy</tt>' Metadata</a></li> |
| Peter Collingbourne | 249d953 | 2011-10-27 19:19:07 +0000 | [diff] [blame] | 107 | </ol> |
| 108 | </li> |
| Chris Lattner | e87d653 | 2006-01-25 23:47:57 +0000 | [diff] [blame] | 109 | </ol> |
| 110 | </li> |
| Bill Wendling | b9d75a9 | 2012-02-11 11:59:36 +0000 | [diff] [blame] | 111 | <li><a href="#module_flags">Module Flags Metadata</a> |
| 112 | <ol> |
| Bill Wendling | f7b367c | 2012-02-16 01:10:50 +0000 | [diff] [blame] | 113 | <li><a href="#objc_gc_flags">Objective-C Garbage Collection Module Flags Metadata</a></li> |
| Bill Wendling | b9d75a9 | 2012-02-11 11:59:36 +0000 | [diff] [blame] | 114 | </ol> |
| 115 | </li> |
| Chris Lattner | 857755c | 2009-07-20 05:55:19 +0000 | [diff] [blame] | 116 | <li><a href="#intrinsic_globals">Intrinsic Global Variables</a> |
| 117 | <ol> |
| 118 | <li><a href="#intg_used">The '<tt>llvm.used</tt>' Global Variable</a></li> |
| Chris Lattner | 401e10c | 2009-07-20 06:14:25 +0000 | [diff] [blame] | 119 | <li><a href="#intg_compiler_used">The '<tt>llvm.compiler.used</tt>' |
| 120 | Global Variable</a></li> |
| Chris Lattner | 857755c | 2009-07-20 05:55:19 +0000 | [diff] [blame] | 121 | <li><a href="#intg_global_ctors">The '<tt>llvm.global_ctors</tt>' |
| 122 | Global Variable</a></li> |
| 123 | <li><a href="#intg_global_dtors">The '<tt>llvm.global_dtors</tt>' |
| 124 | Global Variable</a></li> |
| 125 | </ol> |
| 126 | </li> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 127 | <li><a href="#instref">Instruction Reference</a> |
| 128 | <ol> |
| 129 | <li><a href="#terminators">Terminator Instructions</a> |
| 130 | <ol> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 131 | <li><a href="#i_ret">'<tt>ret</tt>' Instruction</a></li> |
| 132 | <li><a href="#i_br">'<tt>br</tt>' Instruction</a></li> |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 133 | <li><a href="#i_switch">'<tt>switch</tt>' Instruction</a></li> |
| Chris Lattner | ab21db7 | 2009-10-28 00:19:10 +0000 | [diff] [blame] | 134 | <li><a href="#i_indirectbr">'<tt>indirectbr</tt>' Instruction</a></li> |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 135 | <li><a href="#i_invoke">'<tt>invoke</tt>' Instruction</a></li> |
| Bill Wendling | dccc03b | 2011-07-31 06:30:59 +0000 | [diff] [blame] | 136 | <li><a href="#i_resume">'<tt>resume</tt>' Instruction</a></li> |
| Chris Lattner | 35eca58 | 2004-10-16 18:04:13 +0000 | [diff] [blame] | 137 | <li><a href="#i_unreachable">'<tt>unreachable</tt>' Instruction</a></li> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 138 | </ol> |
| 139 | </li> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 140 | <li><a href="#binaryops">Binary Operations</a> |
| 141 | <ol> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 142 | <li><a href="#i_add">'<tt>add</tt>' Instruction</a></li> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 143 | <li><a href="#i_fadd">'<tt>fadd</tt>' Instruction</a></li> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 144 | <li><a href="#i_sub">'<tt>sub</tt>' Instruction</a></li> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 145 | <li><a href="#i_fsub">'<tt>fsub</tt>' Instruction</a></li> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 146 | <li><a href="#i_mul">'<tt>mul</tt>' Instruction</a></li> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 147 | <li><a href="#i_fmul">'<tt>fmul</tt>' Instruction</a></li> |
| Reid Spencer | 1628cec | 2006-10-26 06:15:43 +0000 | [diff] [blame] | 148 | <li><a href="#i_udiv">'<tt>udiv</tt>' Instruction</a></li> |
| 149 | <li><a href="#i_sdiv">'<tt>sdiv</tt>' Instruction</a></li> |
| 150 | <li><a href="#i_fdiv">'<tt>fdiv</tt>' Instruction</a></li> |
| Reid Spencer | 0a783f7 | 2006-11-02 01:53:59 +0000 | [diff] [blame] | 151 | <li><a href="#i_urem">'<tt>urem</tt>' Instruction</a></li> |
| 152 | <li><a href="#i_srem">'<tt>srem</tt>' Instruction</a></li> |
| 153 | <li><a href="#i_frem">'<tt>frem</tt>' Instruction</a></li> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 154 | </ol> |
| 155 | </li> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 156 | <li><a href="#bitwiseops">Bitwise Binary Operations</a> |
| 157 | <ol> |
| Reid Spencer | 8e11bf8 | 2007-02-02 13:57:07 +0000 | [diff] [blame] | 158 | <li><a href="#i_shl">'<tt>shl</tt>' Instruction</a></li> |
| 159 | <li><a href="#i_lshr">'<tt>lshr</tt>' Instruction</a></li> |
| 160 | <li><a href="#i_ashr">'<tt>ashr</tt>' Instruction</a></li> |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 161 | <li><a href="#i_and">'<tt>and</tt>' Instruction</a></li> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 162 | <li><a href="#i_or">'<tt>or</tt>' Instruction</a></li> |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 163 | <li><a href="#i_xor">'<tt>xor</tt>' Instruction</a></li> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 164 | </ol> |
| 165 | </li> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 166 | <li><a href="#vectorops">Vector Operations</a> |
| 167 | <ol> |
| 168 | <li><a href="#i_extractelement">'<tt>extractelement</tt>' Instruction</a></li> |
| 169 | <li><a href="#i_insertelement">'<tt>insertelement</tt>' Instruction</a></li> |
| 170 | <li><a href="#i_shufflevector">'<tt>shufflevector</tt>' Instruction</a></li> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 171 | </ol> |
| 172 | </li> |
| Dan Gohman | a334d5f | 2008-05-12 23:51:09 +0000 | [diff] [blame] | 173 | <li><a href="#aggregateops">Aggregate Operations</a> |
| 174 | <ol> |
| 175 | <li><a href="#i_extractvalue">'<tt>extractvalue</tt>' Instruction</a></li> |
| 176 | <li><a href="#i_insertvalue">'<tt>insertvalue</tt>' Instruction</a></li> |
| 177 | </ol> |
| 178 | </li> |
| Chris Lattner | 884a970 | 2006-08-15 00:45:58 +0000 | [diff] [blame] | 179 | <li><a href="#memoryops">Memory Access and Addressing Operations</a> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 180 | <ol> |
| Eli Friedman | ff03048 | 2011-07-28 21:48:00 +0000 | [diff] [blame] | 181 | <li><a href="#i_alloca">'<tt>alloca</tt>' Instruction</a></li> |
| 182 | <li><a href="#i_load">'<tt>load</tt>' Instruction</a></li> |
| 183 | <li><a href="#i_store">'<tt>store</tt>' Instruction</a></li> |
| 184 | <li><a href="#i_fence">'<tt>fence</tt>' Instruction</a></li> |
| 185 | <li><a href="#i_cmpxchg">'<tt>cmpxchg</tt>' Instruction</a></li> |
| 186 | <li><a href="#i_atomicrmw">'<tt>atomicrmw</tt>' Instruction</a></li> |
| Robert Bocchino | 7b81c75 | 2006-02-17 21:18:08 +0000 | [diff] [blame] | 187 | <li><a href="#i_getelementptr">'<tt>getelementptr</tt>' Instruction</a></li> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 188 | </ol> |
| 189 | </li> |
| Reid Spencer | 2fd21e6 | 2006-11-08 01:18:52 +0000 | [diff] [blame] | 190 | <li><a href="#convertops">Conversion Operations</a> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 191 | <ol> |
| 192 | <li><a href="#i_trunc">'<tt>trunc .. to</tt>' Instruction</a></li> |
| 193 | <li><a href="#i_zext">'<tt>zext .. to</tt>' Instruction</a></li> |
| 194 | <li><a href="#i_sext">'<tt>sext .. to</tt>' Instruction</a></li> |
| 195 | <li><a href="#i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a></li> |
| 196 | <li><a href="#i_fpext">'<tt>fpext .. to</tt>' Instruction</a></li> |
| Reid Spencer | d444879 | 2006-11-09 23:03:26 +0000 | [diff] [blame] | 197 | <li><a href="#i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a></li> |
| 198 | <li><a href="#i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a></li> |
| 199 | <li><a href="#i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a></li> |
| 200 | <li><a href="#i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a></li> |
| Reid Spencer | 7267925 | 2006-11-11 21:00:47 +0000 | [diff] [blame] | 201 | <li><a href="#i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a></li> |
| 202 | <li><a href="#i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a></li> |
| Reid Spencer | 5c0ef47 | 2006-11-11 23:08:07 +0000 | [diff] [blame] | 203 | <li><a href="#i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a></li> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 204 | </ol> |
| Dan Gohman | 0e451ce | 2008-10-14 16:51:45 +0000 | [diff] [blame] | 205 | </li> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 206 | <li><a href="#otherops">Other Operations</a> |
| 207 | <ol> |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 208 | <li><a href="#i_icmp">'<tt>icmp</tt>' Instruction</a></li> |
| 209 | <li><a href="#i_fcmp">'<tt>fcmp</tt>' Instruction</a></li> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 210 | <li><a href="#i_phi">'<tt>phi</tt>' Instruction</a></li> |
| Chris Lattner | cc37aae | 2004-03-12 05:50:16 +0000 | [diff] [blame] | 211 | <li><a href="#i_select">'<tt>select</tt>' Instruction</a></li> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 212 | <li><a href="#i_call">'<tt>call</tt>' Instruction</a></li> |
| Chris Lattner | fb6977d | 2006-01-13 23:26:01 +0000 | [diff] [blame] | 213 | <li><a href="#i_va_arg">'<tt>va_arg</tt>' Instruction</a></li> |
| Bill Wendling | f78faf8 | 2011-08-02 21:52:38 +0000 | [diff] [blame] | 214 | <li><a href="#i_landingpad">'<tt>landingpad</tt>' Instruction</a></li> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 215 | </ol> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 216 | </li> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 217 | </ol> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 218 | </li> |
| Chris Lattner | d9ad5b3 | 2003-05-08 04:57:36 +0000 | [diff] [blame] | 219 | <li><a href="#intrinsics">Intrinsic Functions</a> |
| Chris Lattner | d9ad5b3 | 2003-05-08 04:57:36 +0000 | [diff] [blame] | 220 | <ol> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 221 | <li><a href="#int_varargs">Variable Argument Handling Intrinsics</a> |
| 222 | <ol> |
| Reid Spencer | a3e435f | 2007-04-04 02:42:35 +0000 | [diff] [blame] | 223 | <li><a href="#int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a></li> |
| 224 | <li><a href="#int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a></li> |
| 225 | <li><a href="#int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a></li> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 226 | </ol> |
| 227 | </li> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 228 | <li><a href="#int_gc">Accurate Garbage Collection Intrinsics</a> |
| 229 | <ol> |
| Reid Spencer | a3e435f | 2007-04-04 02:42:35 +0000 | [diff] [blame] | 230 | <li><a href="#int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a></li> |
| 231 | <li><a href="#int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a></li> |
| 232 | <li><a href="#int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a></li> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 233 | </ol> |
| 234 | </li> |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 235 | <li><a href="#int_codegen">Code Generator Intrinsics</a> |
| 236 | <ol> |
| Reid Spencer | a3e435f | 2007-04-04 02:42:35 +0000 | [diff] [blame] | 237 | <li><a href="#int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a></li> |
| 238 | <li><a href="#int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a></li> |
| 239 | <li><a href="#int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a></li> |
| 240 | <li><a href="#int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a></li> |
| 241 | <li><a href="#int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a></li> |
| 242 | <li><a href="#int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a></li> |
| Dan Gohman | 31f1af1 | 2010-05-26 21:56:15 +0000 | [diff] [blame] | 243 | <li><a href="#int_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a></li> |
| John Criswell | 7123e27 | 2004-04-09 16:43:20 +0000 | [diff] [blame] | 244 | </ol> |
| 245 | </li> |
| Chris Lattner | 33aec9e | 2004-02-12 17:01:32 +0000 | [diff] [blame] | 246 | <li><a href="#int_libc">Standard C Library Intrinsics</a> |
| 247 | <ol> |
| Reid Spencer | a3e435f | 2007-04-04 02:42:35 +0000 | [diff] [blame] | 248 | <li><a href="#int_memcpy">'<tt>llvm.memcpy.*</tt>' Intrinsic</a></li> |
| 249 | <li><a href="#int_memmove">'<tt>llvm.memmove.*</tt>' Intrinsic</a></li> |
| 250 | <li><a href="#int_memset">'<tt>llvm.memset.*</tt>' Intrinsic</a></li> |
| 251 | <li><a href="#int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a></li> |
| 252 | <li><a href="#int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a></li> |
| Dan Gohman | 91c284c | 2007-10-15 20:30:11 +0000 | [diff] [blame] | 253 | <li><a href="#int_sin">'<tt>llvm.sin.*</tt>' Intrinsic</a></li> |
| 254 | <li><a href="#int_cos">'<tt>llvm.cos.*</tt>' Intrinsic</a></li> |
| 255 | <li><a href="#int_pow">'<tt>llvm.pow.*</tt>' Intrinsic</a></li> |
| Dan Gohman | 08b280b | 2011-05-27 00:36:31 +0000 | [diff] [blame] | 256 | <li><a href="#int_exp">'<tt>llvm.exp.*</tt>' Intrinsic</a></li> |
| 257 | <li><a href="#int_log">'<tt>llvm.log.*</tt>' Intrinsic</a></li> |
| Cameron Zwarich | 3339084 | 2011-07-08 21:39:21 +0000 | [diff] [blame] | 258 | <li><a href="#int_fma">'<tt>llvm.fma.*</tt>' Intrinsic</a></li> |
| Chris Lattner | 33aec9e | 2004-02-12 17:01:32 +0000 | [diff] [blame] | 259 | </ol> |
| 260 | </li> |
| Nate Begeman | 7e36c47 | 2006-01-13 23:26:38 +0000 | [diff] [blame] | 261 | <li><a href="#int_manip">Bit Manipulation Intrinsics</a> |
| Andrew Lenharth | ec370fd | 2005-05-03 18:01:48 +0000 | [diff] [blame] | 262 | <ol> |
| Reid Spencer | a3e435f | 2007-04-04 02:42:35 +0000 | [diff] [blame] | 263 | <li><a href="#int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a></li> |
| Chris Lattner | 8a886be | 2006-01-16 22:34:14 +0000 | [diff] [blame] | 264 | <li><a href="#int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic </a></li> |
| 265 | <li><a href="#int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic </a></li> |
| 266 | <li><a href="#int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic </a></li> |
| Andrew Lenharth | ec370fd | 2005-05-03 18:01:48 +0000 | [diff] [blame] | 267 | </ol> |
| 268 | </li> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 269 | <li><a href="#int_overflow">Arithmetic with Overflow Intrinsics</a> |
| 270 | <ol> |
| Bill Wendling | da01af7 | 2009-02-08 04:04:40 +0000 | [diff] [blame] | 271 | <li><a href="#int_sadd_overflow">'<tt>llvm.sadd.with.overflow.*</tt> Intrinsics</a></li> |
| 272 | <li><a href="#int_uadd_overflow">'<tt>llvm.uadd.with.overflow.*</tt> Intrinsics</a></li> |
| 273 | <li><a href="#int_ssub_overflow">'<tt>llvm.ssub.with.overflow.*</tt> Intrinsics</a></li> |
| 274 | <li><a href="#int_usub_overflow">'<tt>llvm.usub.with.overflow.*</tt> Intrinsics</a></li> |
| 275 | <li><a href="#int_smul_overflow">'<tt>llvm.smul.with.overflow.*</tt> Intrinsics</a></li> |
| Bill Wendling | 41b485c | 2009-02-08 23:00:09 +0000 | [diff] [blame] | 276 | <li><a href="#int_umul_overflow">'<tt>llvm.umul.with.overflow.*</tt> Intrinsics</a></li> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 277 | </ol> |
| 278 | </li> |
| Anton Korobeynikov | f02e730 | 2010-03-14 18:42:47 +0000 | [diff] [blame] | 279 | <li><a href="#int_fp16">Half Precision Floating Point Intrinsics</a> |
| 280 | <ol> |
| Chris Lattner | 82c3dc6 | 2010-03-14 23:03:31 +0000 | [diff] [blame] | 281 | <li><a href="#int_convert_to_fp16">'<tt>llvm.convert.to.fp16</tt>' Intrinsic</a></li> |
| 282 | <li><a href="#int_convert_from_fp16">'<tt>llvm.convert.from.fp16</tt>' Intrinsic</a></li> |
| Anton Korobeynikov | f02e730 | 2010-03-14 18:42:47 +0000 | [diff] [blame] | 283 | </ol> |
| 284 | </li> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 285 | <li><a href="#int_debugger">Debugger intrinsics</a></li> |
| Jim Laskey | dd4ef1b | 2007-03-14 19:31:19 +0000 | [diff] [blame] | 286 | <li><a href="#int_eh">Exception Handling intrinsics</a></li> |
| Duncan Sands | 4a544a7 | 2011-09-06 13:37:06 +0000 | [diff] [blame] | 287 | <li><a href="#int_trampoline">Trampoline Intrinsics</a> |
| Duncan Sands | 36397f5 | 2007-07-27 12:58:54 +0000 | [diff] [blame] | 288 | <ol> |
| 289 | <li><a href="#int_it">'<tt>llvm.init.trampoline</tt>' Intrinsic</a></li> |
| Duncan Sands | 4a544a7 | 2011-09-06 13:37:06 +0000 | [diff] [blame] | 290 | <li><a href="#int_at">'<tt>llvm.adjust.trampoline</tt>' Intrinsic</a></li> |
| Duncan Sands | 36397f5 | 2007-07-27 12:58:54 +0000 | [diff] [blame] | 291 | </ol> |
| 292 | </li> |
| Nick Lewycky | cc27186 | 2009-10-13 07:03:23 +0000 | [diff] [blame] | 293 | <li><a href="#int_memorymarkers">Memory Use Markers</a> |
| 294 | <ol> |
| Jakub Staszak | 8e1b12a | 2011-12-04 20:44:25 +0000 | [diff] [blame] | 295 | <li><a href="#int_lifetime_start">'<tt>llvm.lifetime.start</tt>' Intrinsic</a></li> |
| 296 | <li><a href="#int_lifetime_end">'<tt>llvm.lifetime.end</tt>' Intrinsic</a></li> |
| 297 | <li><a href="#int_invariant_start">'<tt>llvm.invariant.start</tt>' Intrinsic</a></li> |
| 298 | <li><a href="#int_invariant_end">'<tt>llvm.invariant.end</tt>' Intrinsic</a></li> |
| Nick Lewycky | cc27186 | 2009-10-13 07:03:23 +0000 | [diff] [blame] | 299 | </ol> |
| 300 | </li> |
| Reid Spencer | 2067764 | 2007-07-20 19:59:11 +0000 | [diff] [blame] | 301 | <li><a href="#int_general">General intrinsics</a> |
| Tanya Lattner | 6d806e9 | 2007-06-15 20:50:54 +0000 | [diff] [blame] | 302 | <ol> |
| Reid Spencer | 2067764 | 2007-07-20 19:59:11 +0000 | [diff] [blame] | 303 | <li><a href="#int_var_annotation"> |
| Bill Wendling | 69e4adb | 2008-11-19 05:56:17 +0000 | [diff] [blame] | 304 | '<tt>llvm.var.annotation</tt>' Intrinsic</a></li> |
| Tanya Lattner | b636788 | 2007-09-21 22:59:12 +0000 | [diff] [blame] | 305 | <li><a href="#int_annotation"> |
| Bill Wendling | 69e4adb | 2008-11-19 05:56:17 +0000 | [diff] [blame] | 306 | '<tt>llvm.annotation.*</tt>' Intrinsic</a></li> |
| Anton Korobeynikov | 4cb8618 | 2008-01-15 22:31:34 +0000 | [diff] [blame] | 307 | <li><a href="#int_trap"> |
| Bill Wendling | 69e4adb | 2008-11-19 05:56:17 +0000 | [diff] [blame] | 308 | '<tt>llvm.trap</tt>' Intrinsic</a></li> |
| 309 | <li><a href="#int_stackprotector"> |
| 310 | '<tt>llvm.stackprotector</tt>' Intrinsic</a></li> |
| Eric Christopher | 0e67149 | 2009-11-30 08:03:53 +0000 | [diff] [blame] | 311 | <li><a href="#int_objectsize"> |
| 312 | '<tt>llvm.objectsize</tt>' Intrinsic</a></li> |
| Jakub Staszak | b170e2d | 2011-12-04 18:29:26 +0000 | [diff] [blame] | 313 | <li><a href="#int_expect"> |
| 314 | '<tt>llvm.expect</tt>' Intrinsic</a></li> |
| Tanya Lattner | b636788 | 2007-09-21 22:59:12 +0000 | [diff] [blame] | 315 | </ol> |
| Tanya Lattner | 6d806e9 | 2007-06-15 20:50:54 +0000 | [diff] [blame] | 316 | </li> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 317 | </ol> |
| 318 | </li> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 319 | </ol> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 320 | |
| 321 | <div class="doc_author"> |
| 322 | <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a> |
| 323 | and <a href="mailto:vadve@cs.uiuc.edu">Vikram Adve</a></p> |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 324 | </div> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 325 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 326 | <!-- *********************************************************************** --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 327 | <h2><a name="abstract">Abstract</a></h2> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 328 | <!-- *********************************************************************** --> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 329 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 330 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 331 | |
| 332 | <p>This document is a reference manual for the LLVM assembly language. LLVM is |
| 333 | a Static Single Assignment (SSA) based representation that provides type |
| 334 | safety, low-level operations, flexibility, and the capability of representing |
| 335 | 'all' high-level languages cleanly. It is the common code representation |
| 336 | used throughout all phases of the LLVM compilation strategy.</p> |
| 337 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 338 | </div> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 339 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 340 | <!-- *********************************************************************** --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 341 | <h2><a name="introduction">Introduction</a></h2> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 342 | <!-- *********************************************************************** --> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 343 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 344 | <div> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 345 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 346 | <p>The LLVM code representation is designed to be used in three different forms: |
| 347 | as an in-memory compiler IR, as an on-disk bitcode representation (suitable |
| 348 | for fast loading by a Just-In-Time compiler), and as a human readable |
| 349 | assembly language representation. This allows LLVM to provide a powerful |
| 350 | intermediate representation for efficient compiler transformations and |
| 351 | analysis, while providing a natural means to debug and visualize the |
| 352 | transformations. The three different forms of LLVM are all equivalent. This |
| 353 | document describes the human readable representation and notation.</p> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 354 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 355 | <p>The LLVM representation aims to be light-weight and low-level while being |
| 356 | expressive, typed, and extensible at the same time. It aims to be a |
| 357 | "universal IR" of sorts, by being at a low enough level that high-level ideas |
| 358 | may be cleanly mapped to it (similar to how microprocessors are "universal |
| 359 | IR's", allowing many source languages to be mapped to them). By providing |
| 360 | type information, LLVM can be used as the target of optimizations: for |
| 361 | example, through pointer analysis, it can be proven that a C automatic |
| Bill Wendling | f82d40a | 2009-11-02 00:24:16 +0000 | [diff] [blame] | 362 | variable is never accessed outside of the current function, allowing it to |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 363 | be promoted to a simple SSA value instead of a memory location.</p> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 364 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 365 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 366 | <h4> |
| 367 | <a name="wellformed">Well-Formedness</a> |
| 368 | </h4> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 369 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 370 | <div> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 371 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 372 | <p>It is important to note that this document describes 'well formed' LLVM |
| 373 | assembly language. There is a difference between what the parser accepts and |
| 374 | what is considered 'well formed'. For example, the following instruction is |
| 375 | syntactically okay, but not well formed:</p> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 376 | |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 377 | <pre class="doc_code"> |
| Bill Wendling | 2f7a8b0 | 2007-05-29 09:04:49 +0000 | [diff] [blame] | 378 | %x = <a href="#i_add">add</a> i32 1, %x |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 379 | </pre> |
| 380 | |
| Bill Wendling | f82d40a | 2009-11-02 00:24:16 +0000 | [diff] [blame] | 381 | <p>because the definition of <tt>%x</tt> does not dominate all of its uses. The |
| 382 | LLVM infrastructure provides a verification pass that may be used to verify |
| 383 | that an LLVM module is well formed. This pass is automatically run by the |
| 384 | parser after parsing input assembly and by the optimizer before it outputs |
| 385 | bitcode. The violations pointed out by the verifier pass indicate bugs in |
| 386 | transformation passes or input to the parser.</p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 387 | |
| Bill Wendling | 2f7a8b0 | 2007-05-29 09:04:49 +0000 | [diff] [blame] | 388 | </div> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 389 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 390 | </div> |
| 391 | |
| Chris Lattner | cc68939 | 2007-10-03 17:34:29 +0000 | [diff] [blame] | 392 | <!-- Describe the typesetting conventions here. --> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 393 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 394 | <!-- *********************************************************************** --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 395 | <h2><a name="identifiers">Identifiers</a></h2> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 396 | <!-- *********************************************************************** --> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 397 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 398 | <div> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 399 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 400 | <p>LLVM identifiers come in two basic types: global and local. Global |
| 401 | identifiers (functions, global variables) begin with the <tt>'@'</tt> |
| 402 | character. Local identifiers (register names, types) begin with |
| 403 | the <tt>'%'</tt> character. Additionally, there are three different formats |
| 404 | for identifiers, for different purposes:</p> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 405 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 406 | <ol> |
| Reid Spencer | 2c45228 | 2007-08-07 14:34:28 +0000 | [diff] [blame] | 407 | <li>Named values are represented as a string of characters with their prefix. |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 408 | For example, <tt>%foo</tt>, <tt>@DivisionByZero</tt>, |
| 409 | <tt>%a.really.long.identifier</tt>. The actual regular expression used is |
| 410 | '<tt>[%@][a-zA-Z$._][a-zA-Z$._0-9]*</tt>'. Identifiers which require |
| 411 | other characters in their names can be surrounded with quotes. Special |
| 412 | characters may be escaped using <tt>"\xx"</tt> where <tt>xx</tt> is the |
| 413 | ASCII code for the character in hexadecimal. In this way, any character |
| 414 | can be used in a name value, even quotes themselves.</li> |
| Chris Lattner | e5d947b | 2004-12-09 16:36:40 +0000 | [diff] [blame] | 415 | |
| Reid Spencer | 2c45228 | 2007-08-07 14:34:28 +0000 | [diff] [blame] | 416 | <li>Unnamed values are represented as an unsigned numeric value with their |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 417 | prefix. For example, <tt>%12</tt>, <tt>@2</tt>, <tt>%44</tt>.</li> |
| Chris Lattner | e5d947b | 2004-12-09 16:36:40 +0000 | [diff] [blame] | 418 | |
| Reid Spencer | cc16dc3 | 2004-12-09 18:02:53 +0000 | [diff] [blame] | 419 | <li>Constants, which are described in a <a href="#constants">section about |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 420 | constants</a>, below.</li> |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 421 | </ol> |
| Chris Lattner | e5d947b | 2004-12-09 16:36:40 +0000 | [diff] [blame] | 422 | |
| Reid Spencer | 2c45228 | 2007-08-07 14:34:28 +0000 | [diff] [blame] | 423 | <p>LLVM requires that values start with a prefix for two reasons: Compilers |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 424 | don't need to worry about name clashes with reserved words, and the set of |
| 425 | reserved words may be expanded in the future without penalty. Additionally, |
| 426 | unnamed identifiers allow a compiler to quickly come up with a temporary |
| 427 | variable without having to avoid symbol table conflicts.</p> |
| Chris Lattner | e5d947b | 2004-12-09 16:36:40 +0000 | [diff] [blame] | 428 | |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 429 | <p>Reserved words in LLVM are very similar to reserved words in other |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 430 | languages. There are keywords for different opcodes |
| 431 | ('<tt><a href="#i_add">add</a></tt>', |
| 432 | '<tt><a href="#i_bitcast">bitcast</a></tt>', |
| 433 | '<tt><a href="#i_ret">ret</a></tt>', etc...), for primitive type names |
| 434 | ('<tt><a href="#t_void">void</a></tt>', |
| 435 | '<tt><a href="#t_primitive">i32</a></tt>', etc...), and others. These |
| 436 | reserved words cannot conflict with variable names, because none of them |
| 437 | start with a prefix character (<tt>'%'</tt> or <tt>'@'</tt>).</p> |
| Chris Lattner | e5d947b | 2004-12-09 16:36:40 +0000 | [diff] [blame] | 438 | |
| 439 | <p>Here is an example of LLVM code to multiply the integer variable |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 440 | '<tt>%X</tt>' by 8:</p> |
| Chris Lattner | e5d947b | 2004-12-09 16:36:40 +0000 | [diff] [blame] | 441 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 442 | <p>The easy way:</p> |
| Chris Lattner | e5d947b | 2004-12-09 16:36:40 +0000 | [diff] [blame] | 443 | |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 444 | <pre class="doc_code"> |
| Bill Wendling | 2f7a8b0 | 2007-05-29 09:04:49 +0000 | [diff] [blame] | 445 | %result = <a href="#i_mul">mul</a> i32 %X, 8 |
| Chris Lattner | e5d947b | 2004-12-09 16:36:40 +0000 | [diff] [blame] | 446 | </pre> |
| 447 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 448 | <p>After strength reduction:</p> |
| Chris Lattner | e5d947b | 2004-12-09 16:36:40 +0000 | [diff] [blame] | 449 | |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 450 | <pre class="doc_code"> |
| Bill Wendling | 2f7a8b0 | 2007-05-29 09:04:49 +0000 | [diff] [blame] | 451 | %result = <a href="#i_shl">shl</a> i32 %X, i8 3 |
| Chris Lattner | e5d947b | 2004-12-09 16:36:40 +0000 | [diff] [blame] | 452 | </pre> |
| 453 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 454 | <p>And the hard way:</p> |
| Chris Lattner | e5d947b | 2004-12-09 16:36:40 +0000 | [diff] [blame] | 455 | |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 456 | <pre class="doc_code"> |
| Gabor Greif | ec58f75 | 2009-10-28 13:05:07 +0000 | [diff] [blame] | 457 | %0 = <a href="#i_add">add</a> i32 %X, %X <i>; yields {i32}:%0</i> |
| 458 | %1 = <a href="#i_add">add</a> i32 %0, %0 <i>; yields {i32}:%1</i> |
| Bill Wendling | 2f7a8b0 | 2007-05-29 09:04:49 +0000 | [diff] [blame] | 459 | %result = <a href="#i_add">add</a> i32 %1, %1 |
| Chris Lattner | e5d947b | 2004-12-09 16:36:40 +0000 | [diff] [blame] | 460 | </pre> |
| 461 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 462 | <p>This last way of multiplying <tt>%X</tt> by 8 illustrates several important |
| 463 | lexical features of LLVM:</p> |
| Chris Lattner | e5d947b | 2004-12-09 16:36:40 +0000 | [diff] [blame] | 464 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 465 | <ol> |
| Chris Lattner | e5d947b | 2004-12-09 16:36:40 +0000 | [diff] [blame] | 466 | <li>Comments are delimited with a '<tt>;</tt>' and go until the end of |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 467 | line.</li> |
| Chris Lattner | e5d947b | 2004-12-09 16:36:40 +0000 | [diff] [blame] | 468 | |
| 469 | <li>Unnamed temporaries are created when the result of a computation is not |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 470 | assigned to a named value.</li> |
| Chris Lattner | e5d947b | 2004-12-09 16:36:40 +0000 | [diff] [blame] | 471 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 472 | <li>Unnamed temporaries are numbered sequentially</li> |
| 473 | </ol> |
| Chris Lattner | e5d947b | 2004-12-09 16:36:40 +0000 | [diff] [blame] | 474 | |
| Bill Wendling | f82d40a | 2009-11-02 00:24:16 +0000 | [diff] [blame] | 475 | <p>It also shows a convention that we follow in this document. When |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 476 | demonstrating instructions, we will follow an instruction with a comment that |
| 477 | defines the type and name of value produced. Comments are shown in italic |
| 478 | text.</p> |
| Chris Lattner | e5d947b | 2004-12-09 16:36:40 +0000 | [diff] [blame] | 479 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 480 | </div> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 481 | |
| 482 | <!-- *********************************************************************** --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 483 | <h2><a name="highlevel">High Level Structure</a></h2> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 484 | <!-- *********************************************************************** --> |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 485 | <div> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 486 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 487 | <h3> |
| 488 | <a name="modulestructure">Module Structure</a> |
| 489 | </h3> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 490 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 491 | <div> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 492 | |
| Bill Wendling | 4cc2be6 | 2012-03-14 08:07:43 +0000 | [diff] [blame] | 493 | <p>LLVM programs are composed of <tt>Module</tt>s, each of which is a |
| 494 | translation unit of the input programs. Each module consists of functions, |
| 495 | global variables, and symbol table entries. Modules may be combined together |
| 496 | with the LLVM linker, which merges function (and global variable) |
| 497 | definitions, resolves forward declarations, and merges symbol table |
| 498 | entries. Here is an example of the "hello world" module:</p> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 499 | |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 500 | <pre class="doc_code"> |
| Chris Lattner | 63e4ccb | 2010-08-17 17:13:42 +0000 | [diff] [blame] | 501 | <i>; Declare the string constant as a global constant.</i> |
| Bill Wendling | 4cc2be6 | 2012-03-14 08:07:43 +0000 | [diff] [blame] | 502 | <a href="#identifiers">@.str</a> = <a href="#linkage_private">private</a> <a href="#globalvars">unnamed_addr</a> <a href="#globalvars">constant</a> <a href="#t_array">[13 x i8]</a> c"hello world\0A\00" |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 503 | |
| Chris Lattner | 63e4ccb | 2010-08-17 17:13:42 +0000 | [diff] [blame] | 504 | <i>; External declaration of the puts function</i> |
| Bill Wendling | 4cc2be6 | 2012-03-14 08:07:43 +0000 | [diff] [blame] | 505 | <a href="#functionstructure">declare</a> i32 @puts(i8* <a href="#nocapture">nocapture</a>) <a href="#fnattrs">nounwind</a> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 506 | |
| 507 | <i>; Definition of main function</i> |
| Chris Lattner | 63e4ccb | 2010-08-17 17:13:42 +0000 | [diff] [blame] | 508 | define i32 @main() { <i>; i32()* </i> |
| 509 | <i>; Convert [13 x i8]* to i8 *...</i> |
| Bill Wendling | 4cc2be6 | 2012-03-14 08:07:43 +0000 | [diff] [blame] | 510 | %cast210 = <a href="#i_getelementptr">getelementptr</a> [13 x i8]* @.str, i64 0, i64 0 |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 511 | |
| Chris Lattner | 63e4ccb | 2010-08-17 17:13:42 +0000 | [diff] [blame] | 512 | <i>; Call puts function to write out the string to stdout.</i> |
| Bill Wendling | 4cc2be6 | 2012-03-14 08:07:43 +0000 | [diff] [blame] | 513 | <a href="#i_call">call</a> i32 @puts(i8* %cast210) |
| Chris Lattner | 63e4ccb | 2010-08-17 17:13:42 +0000 | [diff] [blame] | 514 | <a href="#i_ret">ret</a> i32 0 |
| 515 | } |
| Devang Patel | cd1fd25 | 2010-01-11 19:35:55 +0000 | [diff] [blame] | 516 | |
| 517 | <i>; Named metadata</i> |
| Bill Wendling | 4cc2be6 | 2012-03-14 08:07:43 +0000 | [diff] [blame] | 518 | !1 = metadata !{i32 42} |
| Devang Patel | cd1fd25 | 2010-01-11 19:35:55 +0000 | [diff] [blame] | 519 | !foo = !{!1, null} |
| Bill Wendling | 2f7a8b0 | 2007-05-29 09:04:49 +0000 | [diff] [blame] | 520 | </pre> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 521 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 522 | <p>This example is made up of a <a href="#globalvars">global variable</a> named |
| Bill Wendling | 4cc2be6 | 2012-03-14 08:07:43 +0000 | [diff] [blame] | 523 | "<tt>.str</tt>", an external declaration of the "<tt>puts</tt>" function, |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 524 | a <a href="#functionstructure">function definition</a> for |
| Devang Patel | cd1fd25 | 2010-01-11 19:35:55 +0000 | [diff] [blame] | 525 | "<tt>main</tt>" and <a href="#namedmetadatastructure">named metadata</a> |
| Bill Wendling | 4cc2be6 | 2012-03-14 08:07:43 +0000 | [diff] [blame] | 526 | "<tt>foo</tt>".</p> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 527 | |
| Bill Wendling | 4cc2be6 | 2012-03-14 08:07:43 +0000 | [diff] [blame] | 528 | <p>In general, a module is made up of a list of global values (where both |
| 529 | functions and global variables are global values). Global values are |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 530 | represented by a pointer to a memory location (in this case, a pointer to an |
| 531 | array of char, and a pointer to a function), and have one of the |
| 532 | following <a href="#linkage">linkage types</a>.</p> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 533 | |
| Chris Lattner | e5d947b | 2004-12-09 16:36:40 +0000 | [diff] [blame] | 534 | </div> |
| 535 | |
| 536 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 537 | <h3> |
| Chris Lattner | e5d947b | 2004-12-09 16:36:40 +0000 | [diff] [blame] | 538 | <a name="linkage">Linkage Types</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 539 | </h3> |
| Chris Lattner | e5d947b | 2004-12-09 16:36:40 +0000 | [diff] [blame] | 540 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 541 | <div> |
| Chris Lattner | e5d947b | 2004-12-09 16:36:40 +0000 | [diff] [blame] | 542 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 543 | <p>All Global Variables and Functions have one of the following types of |
| 544 | linkage:</p> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 545 | |
| 546 | <dl> |
| Bill Wendling | f82d40a | 2009-11-02 00:24:16 +0000 | [diff] [blame] | 547 | <dt><tt><b><a name="linkage_private">private</a></b></tt></dt> |
| Bill Wendling | 5e721d7 | 2010-07-01 21:55:59 +0000 | [diff] [blame] | 548 | <dd>Global values with "<tt>private</tt>" linkage are only directly accessible |
| 549 | by objects in the current module. In particular, linking code into a |
| 550 | module with an private global value may cause the private to be renamed as |
| 551 | necessary to avoid collisions. Because the symbol is private to the |
| 552 | module, all references can be updated. This doesn't show up in any symbol |
| 553 | table in the object file.</dd> |
| Rafael Espindola | bb46f52 | 2009-01-15 20:18:42 +0000 | [diff] [blame] | 554 | |
| Bill Wendling | f82d40a | 2009-11-02 00:24:16 +0000 | [diff] [blame] | 555 | <dt><tt><b><a name="linkage_linker_private">linker_private</a></b></tt></dt> |
| Bill Wendling | 5e721d7 | 2010-07-01 21:55:59 +0000 | [diff] [blame] | 556 | <dd>Similar to <tt>private</tt>, but the symbol is passed through the |
| 557 | assembler and evaluated by the linker. Unlike normal strong symbols, they |
| 558 | are removed by the linker from the final linked image (executable or |
| 559 | dynamic library).</dd> |
| 560 | |
| 561 | <dt><tt><b><a name="linkage_linker_private_weak">linker_private_weak</a></b></tt></dt> |
| 562 | <dd>Similar to "<tt>linker_private</tt>", but the symbol is weak. Note that |
| 563 | <tt>linker_private_weak</tt> symbols are subject to coalescing by the |
| 564 | linker. The symbols are removed by the linker from the final linked image |
| 565 | (executable or dynamic library).</dd> |
| Bill Wendling | 3d10a5a | 2009-07-20 01:03:30 +0000 | [diff] [blame] | 566 | |
| Bill Wendling | 55ae515 | 2010-08-20 22:05:50 +0000 | [diff] [blame] | 567 | <dt><tt><b><a name="linkage_linker_private_weak_def_auto">linker_private_weak_def_auto</a></b></tt></dt> |
| 568 | <dd>Similar to "<tt>linker_private_weak</tt>", but it's known that the address |
| 569 | of the object is not taken. For instance, functions that had an inline |
| 570 | definition, but the compiler decided not to inline it. Note, |
| 571 | unlike <tt>linker_private</tt> and <tt>linker_private_weak</tt>, |
| 572 | <tt>linker_private_weak_def_auto</tt> may have only <tt>default</tt> |
| 573 | visibility. The symbols are removed by the linker from the final linked |
| 574 | image (executable or dynamic library).</dd> |
| 575 | |
| Bill Wendling | f82d40a | 2009-11-02 00:24:16 +0000 | [diff] [blame] | 576 | <dt><tt><b><a name="linkage_internal">internal</a></b></tt></dt> |
| Bill Wendling | 07d3177 | 2010-06-29 22:34:52 +0000 | [diff] [blame] | 577 | <dd>Similar to private, but the value shows as a local symbol |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 578 | (<tt>STB_LOCAL</tt> in the case of ELF) in the object file. This |
| 579 | corresponds to the notion of the '<tt>static</tt>' keyword in C.</dd> |
| Chris Lattner | e5d947b | 2004-12-09 16:36:40 +0000 | [diff] [blame] | 580 | |
| Bill Wendling | f82d40a | 2009-11-02 00:24:16 +0000 | [diff] [blame] | 581 | <dt><tt><b><a name="linkage_available_externally">available_externally</a></b></tt></dt> |
| Chris Lattner | 266c7bb | 2009-04-13 05:44:34 +0000 | [diff] [blame] | 582 | <dd>Globals with "<tt>available_externally</tt>" linkage are never emitted |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 583 | into the object file corresponding to the LLVM module. They exist to |
| 584 | allow inlining and other optimizations to take place given knowledge of |
| 585 | the definition of the global, which is known to be somewhere outside the |
| 586 | module. Globals with <tt>available_externally</tt> linkage are allowed to |
| 587 | be discarded at will, and are otherwise the same as <tt>linkonce_odr</tt>. |
| 588 | This linkage type is only allowed on definitions, not declarations.</dd> |
| Chris Lattner | 266c7bb | 2009-04-13 05:44:34 +0000 | [diff] [blame] | 589 | |
| Bill Wendling | f82d40a | 2009-11-02 00:24:16 +0000 | [diff] [blame] | 590 | <dt><tt><b><a name="linkage_linkonce">linkonce</a></b></tt></dt> |
| Chris Lattner | 4887bd8 | 2007-01-14 06:51:48 +0000 | [diff] [blame] | 591 | <dd>Globals with "<tt>linkonce</tt>" linkage are merged with other globals of |
| Chris Lattner | 873187c | 2010-01-09 19:15:14 +0000 | [diff] [blame] | 592 | the same name when linkage occurs. This can be used to implement |
| 593 | some forms of inline functions, templates, or other code which must be |
| 594 | generated in each translation unit that uses it, but where the body may |
| 595 | be overridden with a more definitive definition later. Unreferenced |
| 596 | <tt>linkonce</tt> globals are allowed to be discarded. Note that |
| 597 | <tt>linkonce</tt> linkage does not actually allow the optimizer to |
| 598 | inline the body of this function into callers because it doesn't know if |
| 599 | this definition of the function is the definitive definition within the |
| 600 | program or whether it will be overridden by a stronger definition. |
| 601 | To enable inlining and other optimizations, use "<tt>linkonce_odr</tt>" |
| 602 | linkage.</dd> |
| Chris Lattner | e5d947b | 2004-12-09 16:36:40 +0000 | [diff] [blame] | 603 | |
| Bill Wendling | f82d40a | 2009-11-02 00:24:16 +0000 | [diff] [blame] | 604 | <dt><tt><b><a name="linkage_weak">weak</a></b></tt></dt> |
| Chris Lattner | 26d054d | 2009-08-05 05:21:07 +0000 | [diff] [blame] | 605 | <dd>"<tt>weak</tt>" linkage has the same merging semantics as |
| 606 | <tt>linkonce</tt> linkage, except that unreferenced globals with |
| 607 | <tt>weak</tt> linkage may not be discarded. This is used for globals that |
| 608 | are declared "weak" in C source code.</dd> |
| 609 | |
| Bill Wendling | f82d40a | 2009-11-02 00:24:16 +0000 | [diff] [blame] | 610 | <dt><tt><b><a name="linkage_common">common</a></b></tt></dt> |
| Chris Lattner | 26d054d | 2009-08-05 05:21:07 +0000 | [diff] [blame] | 611 | <dd>"<tt>common</tt>" linkage is most similar to "<tt>weak</tt>" linkage, but |
| 612 | they are used for tentative definitions in C, such as "<tt>int X;</tt>" at |
| 613 | global scope. |
| 614 | Symbols with "<tt>common</tt>" linkage are merged in the same way as |
| 615 | <tt>weak symbols</tt>, and they may not be deleted if unreferenced. |
| Chris Lattner | cd81f5d | 2009-08-05 05:41:44 +0000 | [diff] [blame] | 616 | <tt>common</tt> symbols may not have an explicit section, |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 617 | must have a zero initializer, and may not be marked '<a |
| Chris Lattner | cd81f5d | 2009-08-05 05:41:44 +0000 | [diff] [blame] | 618 | href="#globalvars"><tt>constant</tt></a>'. Functions and aliases may not |
| 619 | have common linkage.</dd> |
| Chris Lattner | 26d054d | 2009-08-05 05:21:07 +0000 | [diff] [blame] | 620 | |
| Chris Lattner | e5d947b | 2004-12-09 16:36:40 +0000 | [diff] [blame] | 621 | |
| Bill Wendling | f82d40a | 2009-11-02 00:24:16 +0000 | [diff] [blame] | 622 | <dt><tt><b><a name="linkage_appending">appending</a></b></tt></dt> |
| Chris Lattner | e5d947b | 2004-12-09 16:36:40 +0000 | [diff] [blame] | 623 | <dd>"<tt>appending</tt>" linkage may only be applied to global variables of |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 624 | pointer to array type. When two global variables with appending linkage |
| 625 | are linked together, the two global arrays are appended together. This is |
| 626 | the LLVM, typesafe, equivalent of having the system linker append together |
| 627 | "sections" with identical names when .o files are linked.</dd> |
| Chris Lattner | e5d947b | 2004-12-09 16:36:40 +0000 | [diff] [blame] | 628 | |
| Bill Wendling | f82d40a | 2009-11-02 00:24:16 +0000 | [diff] [blame] | 629 | <dt><tt><b><a name="linkage_externweak">extern_weak</a></b></tt></dt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 630 | <dd>The semantics of this linkage follow the ELF object file model: the symbol |
| 631 | is weak until linked, if not linked, the symbol becomes null instead of |
| 632 | being an undefined reference.</dd> |
| Anton Korobeynikov | 7f70559 | 2007-01-12 19:20:47 +0000 | [diff] [blame] | 633 | |
| Bill Wendling | f82d40a | 2009-11-02 00:24:16 +0000 | [diff] [blame] | 634 | <dt><tt><b><a name="linkage_linkonce_odr">linkonce_odr</a></b></tt></dt> |
| 635 | <dt><tt><b><a name="linkage_weak_odr">weak_odr</a></b></tt></dt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 636 | <dd>Some languages allow differing globals to be merged, such as two functions |
| 637 | with different semantics. Other languages, such as <tt>C++</tt>, ensure |
| Bill Wendling | 5e721d7 | 2010-07-01 21:55:59 +0000 | [diff] [blame] | 638 | that only equivalent globals are ever merged (the "one definition rule" |
| 639 | — "ODR"). Such languages can use the <tt>linkonce_odr</tt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 640 | and <tt>weak_odr</tt> linkage types to indicate that the global will only |
| 641 | be merged with equivalent globals. These linkage types are otherwise the |
| 642 | same as their non-<tt>odr</tt> versions.</dd> |
| Duncan Sands | 667d4b8 | 2009-03-07 15:45:40 +0000 | [diff] [blame] | 643 | |
| Bill Wendling | 5c3a9f7 | 2011-11-04 20:40:41 +0000 | [diff] [blame] | 644 | <dt><tt><b><a name="linkage_external">external</a></b></tt></dt> |
| Chris Lattner | e5d947b | 2004-12-09 16:36:40 +0000 | [diff] [blame] | 645 | <dd>If none of the above identifiers are used, the global is externally |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 646 | visible, meaning that it participates in linkage and can be used to |
| 647 | resolve external symbol references.</dd> |
| Reid Spencer | c891084 | 2007-04-11 23:49:50 +0000 | [diff] [blame] | 648 | </dl> |
| Anton Korobeynikov | b74ed07 | 2006-09-14 18:23:27 +0000 | [diff] [blame] | 649 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 650 | <p>The next two types of linkage are targeted for Microsoft Windows platform |
| 651 | only. They are designed to support importing (exporting) symbols from (to) |
| 652 | DLLs (Dynamic Link Libraries).</p> |
| Anton Korobeynikov | b74ed07 | 2006-09-14 18:23:27 +0000 | [diff] [blame] | 653 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 654 | <dl> |
| Bill Wendling | f82d40a | 2009-11-02 00:24:16 +0000 | [diff] [blame] | 655 | <dt><tt><b><a name="linkage_dllimport">dllimport</a></b></tt></dt> |
| Anton Korobeynikov | b74ed07 | 2006-09-14 18:23:27 +0000 | [diff] [blame] | 656 | <dd>"<tt>dllimport</tt>" linkage causes the compiler to reference a function |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 657 | or variable via a global pointer to a pointer that is set up by the DLL |
| 658 | exporting the symbol. On Microsoft Windows targets, the pointer name is |
| 659 | formed by combining <code>__imp_</code> and the function or variable |
| 660 | name.</dd> |
| Anton Korobeynikov | b74ed07 | 2006-09-14 18:23:27 +0000 | [diff] [blame] | 661 | |
| Bill Wendling | f82d40a | 2009-11-02 00:24:16 +0000 | [diff] [blame] | 662 | <dt><tt><b><a name="linkage_dllexport">dllexport</a></b></tt></dt> |
| Anton Korobeynikov | b74ed07 | 2006-09-14 18:23:27 +0000 | [diff] [blame] | 663 | <dd>"<tt>dllexport</tt>" linkage causes the compiler to provide a global |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 664 | pointer to a pointer in a DLL, so that it can be referenced with the |
| 665 | <tt>dllimport</tt> attribute. On Microsoft Windows targets, the pointer |
| 666 | name is formed by combining <code>__imp_</code> and the function or |
| 667 | variable name.</dd> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 668 | </dl> |
| 669 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 670 | <p>For example, since the "<tt>.LC0</tt>" variable is defined to be internal, if |
| 671 | another module defined a "<tt>.LC0</tt>" variable and was linked with this |
| 672 | one, one of the two would be renamed, preventing a collision. Since |
| 673 | "<tt>main</tt>" and "<tt>puts</tt>" are external (i.e., lacking any linkage |
| 674 | declarations), they are accessible outside of the current module.</p> |
| 675 | |
| 676 | <p>It is illegal for a function <i>declaration</i> to have any linkage type |
| Bill Wendling | f7f0610 | 2011-10-11 06:41:28 +0000 | [diff] [blame] | 677 | other than <tt>external</tt>, <tt>dllimport</tt> |
| 678 | or <tt>extern_weak</tt>.</p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 679 | |
| Duncan Sands | 667d4b8 | 2009-03-07 15:45:40 +0000 | [diff] [blame] | 680 | <p>Aliases can have only <tt>external</tt>, <tt>internal</tt>, <tt>weak</tt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 681 | or <tt>weak_odr</tt> linkages.</p> |
| 682 | |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 683 | </div> |
| 684 | |
| 685 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 686 | <h3> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 687 | <a name="callingconv">Calling Conventions</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 688 | </h3> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 689 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 690 | <div> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 691 | |
| 692 | <p>LLVM <a href="#functionstructure">functions</a>, <a href="#i_call">calls</a> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 693 | and <a href="#i_invoke">invokes</a> can all have an optional calling |
| 694 | convention specified for the call. The calling convention of any pair of |
| 695 | dynamic caller/callee must match, or the behavior of the program is |
| 696 | undefined. The following calling conventions are supported by LLVM, and more |
| 697 | may be added in the future:</p> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 698 | |
| 699 | <dl> |
| 700 | <dt><b>"<tt>ccc</tt>" - The C calling convention</b>:</dt> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 701 | <dd>This calling convention (the default if no other calling convention is |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 702 | specified) matches the target C calling conventions. This calling |
| 703 | convention supports varargs function calls and tolerates some mismatch in |
| 704 | the declared prototype and implemented declaration of the function (as |
| 705 | does normal C).</dd> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 706 | |
| 707 | <dt><b>"<tt>fastcc</tt>" - The fast calling convention</b>:</dt> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 708 | <dd>This calling convention attempts to make calls as fast as possible |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 709 | (e.g. by passing things in registers). This calling convention allows the |
| 710 | target to use whatever tricks it wants to produce fast code for the |
| 711 | target, without having to conform to an externally specified ABI |
| Jeffrey Yasskin | 95fa80a | 2010-01-09 19:44:16 +0000 | [diff] [blame] | 712 | (Application Binary Interface). |
| 713 | <a href="CodeGenerator.html#tailcallopt">Tail calls can only be optimized |
| Chris Lattner | 2968943 | 2010-03-11 00:22:57 +0000 | [diff] [blame] | 714 | when this or the GHC convention is used.</a> This calling convention |
| 715 | does not support varargs and requires the prototype of all callees to |
| 716 | exactly match the prototype of the function definition.</dd> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 717 | |
| 718 | <dt><b>"<tt>coldcc</tt>" - The cold calling convention</b>:</dt> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 719 | <dd>This calling convention attempts to make code in the caller as efficient |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 720 | as possible under the assumption that the call is not commonly executed. |
| 721 | As such, these calls often preserve all registers so that the call does |
| 722 | not break any live ranges in the caller side. This calling convention |
| 723 | does not support varargs and requires the prototype of all callees to |
| 724 | exactly match the prototype of the function definition.</dd> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 725 | |
| Chris Lattner | 2968943 | 2010-03-11 00:22:57 +0000 | [diff] [blame] | 726 | <dt><b>"<tt>cc <em>10</em></tt>" - GHC convention</b>:</dt> |
| 727 | <dd>This calling convention has been implemented specifically for use by the |
| 728 | <a href="http://www.haskell.org/ghc">Glasgow Haskell Compiler (GHC)</a>. |
| 729 | It passes everything in registers, going to extremes to achieve this by |
| 730 | disabling callee save registers. This calling convention should not be |
| 731 | used lightly but only for specific situations such as an alternative to |
| 732 | the <em>register pinning</em> performance technique often used when |
| 733 | implementing functional programming languages.At the moment only X86 |
| 734 | supports this convention and it has the following limitations: |
| 735 | <ul> |
| 736 | <li>On <em>X86-32</em> only supports up to 4 bit type parameters. No |
| 737 | floating point types are supported.</li> |
| 738 | <li>On <em>X86-64</em> only supports up to 10 bit type parameters and |
| 739 | 6 floating point parameters.</li> |
| 740 | </ul> |
| 741 | This calling convention supports |
| 742 | <a href="CodeGenerator.html#tailcallopt">tail call optimization</a> but |
| 743 | requires both the caller and callee are using it. |
| 744 | </dd> |
| 745 | |
| Chris Lattner | cfe6b37 | 2005-05-07 01:46:40 +0000 | [diff] [blame] | 746 | <dt><b>"<tt>cc <<em>n</em>></tt>" - Numbered convention</b>:</dt> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 747 | <dd>Any calling convention may be specified by number, allowing |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 748 | target-specific calling conventions to be used. Target specific calling |
| 749 | conventions start at 64.</dd> |
| Chris Lattner | cfe6b37 | 2005-05-07 01:46:40 +0000 | [diff] [blame] | 750 | </dl> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 751 | |
| 752 | <p>More calling conventions can be added/defined on an as-needed basis, to |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 753 | support Pascal conventions or any other well-known target-independent |
| 754 | convention.</p> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 755 | |
| 756 | </div> |
| 757 | |
| 758 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 759 | <h3> |
| Anton Korobeynikov | 8cea37b | 2007-01-23 12:35:46 +0000 | [diff] [blame] | 760 | <a name="visibility">Visibility Styles</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 761 | </h3> |
| Anton Korobeynikov | 8cea37b | 2007-01-23 12:35:46 +0000 | [diff] [blame] | 762 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 763 | <div> |
| Anton Korobeynikov | 8cea37b | 2007-01-23 12:35:46 +0000 | [diff] [blame] | 764 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 765 | <p>All Global Variables and Functions have one of the following visibility |
| 766 | styles:</p> |
| Anton Korobeynikov | 8cea37b | 2007-01-23 12:35:46 +0000 | [diff] [blame] | 767 | |
| 768 | <dl> |
| 769 | <dt><b>"<tt>default</tt>" - Default style</b>:</dt> |
| Chris Lattner | d3eda89 | 2008-08-05 18:29:16 +0000 | [diff] [blame] | 770 | <dd>On targets that use the ELF object file format, default visibility means |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 771 | that the declaration is visible to other modules and, in shared libraries, |
| 772 | means that the declared entity may be overridden. On Darwin, default |
| 773 | visibility means that the declaration is visible to other modules. Default |
| 774 | visibility corresponds to "external linkage" in the language.</dd> |
| Anton Korobeynikov | 8cea37b | 2007-01-23 12:35:46 +0000 | [diff] [blame] | 775 | |
| 776 | <dt><b>"<tt>hidden</tt>" - Hidden style</b>:</dt> |
| Anton Korobeynikov | 8cea37b | 2007-01-23 12:35:46 +0000 | [diff] [blame] | 777 | <dd>Two declarations of an object with hidden visibility refer to the same |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 778 | object if they are in the same shared object. Usually, hidden visibility |
| 779 | indicates that the symbol will not be placed into the dynamic symbol |
| 780 | table, so no other module (executable or shared library) can reference it |
| 781 | directly.</dd> |
| Anton Korobeynikov | 8cea37b | 2007-01-23 12:35:46 +0000 | [diff] [blame] | 782 | |
| Anton Korobeynikov | 6f9896f | 2007-04-29 18:35:00 +0000 | [diff] [blame] | 783 | <dt><b>"<tt>protected</tt>" - Protected style</b>:</dt> |
| Anton Korobeynikov | 6f9896f | 2007-04-29 18:35:00 +0000 | [diff] [blame] | 784 | <dd>On ELF, protected visibility indicates that the symbol will be placed in |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 785 | the dynamic symbol table, but that references within the defining module |
| 786 | will bind to the local symbol. That is, the symbol cannot be overridden by |
| 787 | another module.</dd> |
| Anton Korobeynikov | 8cea37b | 2007-01-23 12:35:46 +0000 | [diff] [blame] | 788 | </dl> |
| 789 | |
| 790 | </div> |
| 791 | |
| 792 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 793 | <h3> |
| Chris Lattner | e7886e4 | 2009-01-11 20:53:49 +0000 | [diff] [blame] | 794 | <a name="namedtypes">Named Types</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 795 | </h3> |
| Chris Lattner | e7886e4 | 2009-01-11 20:53:49 +0000 | [diff] [blame] | 796 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 797 | <div> |
| Chris Lattner | e7886e4 | 2009-01-11 20:53:49 +0000 | [diff] [blame] | 798 | |
| 799 | <p>LLVM IR allows you to specify name aliases for certain types. This can make |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 800 | it easier to read the IR and make the IR more condensed (particularly when |
| 801 | recursive types are involved). An example of a name specification is:</p> |
| Chris Lattner | e7886e4 | 2009-01-11 20:53:49 +0000 | [diff] [blame] | 802 | |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 803 | <pre class="doc_code"> |
| Chris Lattner | e7886e4 | 2009-01-11 20:53:49 +0000 | [diff] [blame] | 804 | %mytype = type { %mytype*, i32 } |
| 805 | </pre> |
| Chris Lattner | e7886e4 | 2009-01-11 20:53:49 +0000 | [diff] [blame] | 806 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 807 | <p>You may give a name to any <a href="#typesystem">type</a> except |
| Chris Lattner | dc65f22 | 2010-08-17 23:26:04 +0000 | [diff] [blame] | 808 | "<a href="#t_void">void</a>". Type name aliases may be used anywhere a type |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 809 | is expected with the syntax "%mytype".</p> |
| Chris Lattner | e7886e4 | 2009-01-11 20:53:49 +0000 | [diff] [blame] | 810 | |
| 811 | <p>Note that type names are aliases for the structural type that they indicate, |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 812 | and that you can therefore specify multiple names for the same type. This |
| 813 | often leads to confusing behavior when dumping out a .ll file. Since LLVM IR |
| 814 | uses structural typing, the name is not part of the type. When printing out |
| 815 | LLVM IR, the printer will pick <em>one name</em> to render all types of a |
| 816 | particular shape. This means that if you have code where two different |
| 817 | source types end up having the same LLVM type, that the dumper will sometimes |
| 818 | print the "wrong" or unexpected type. This is an important design point and |
| 819 | isn't going to change.</p> |
| Chris Lattner | e7886e4 | 2009-01-11 20:53:49 +0000 | [diff] [blame] | 820 | |
| 821 | </div> |
| 822 | |
| Chris Lattner | e7886e4 | 2009-01-11 20:53:49 +0000 | [diff] [blame] | 823 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 824 | <h3> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 825 | <a name="globalvars">Global Variables</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 826 | </h3> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 827 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 828 | <div> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 829 | |
| Chris Lattner | 3689a34 | 2005-02-12 19:30:21 +0000 | [diff] [blame] | 830 | <p>Global variables define regions of memory allocated at compilation time |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 831 | instead of run-time. Global variables may optionally be initialized, may |
| 832 | have an explicit section to be placed in, and may have an optional explicit |
| 833 | alignment specified. A variable may be defined as "thread_local", which |
| 834 | means that it will not be shared by threads (each thread will have a |
| 835 | separated copy of the variable). A variable may be defined as a global |
| 836 | "constant," which indicates that the contents of the variable |
| 837 | will <b>never</b> be modified (enabling better optimization, allowing the |
| 838 | global data to be placed in the read-only section of an executable, etc). |
| 839 | Note that variables that need runtime initialization cannot be marked |
| 840 | "constant" as there is a store to the variable.</p> |
| Chris Lattner | 3689a34 | 2005-02-12 19:30:21 +0000 | [diff] [blame] | 841 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 842 | <p>LLVM explicitly allows <em>declarations</em> of global variables to be marked |
| 843 | constant, even if the final definition of the global is not. This capability |
| 844 | can be used to enable slightly better optimization of the program, but |
| 845 | requires the language definition to guarantee that optimizations based on the |
| 846 | 'constantness' are valid for the translation units that do not include the |
| 847 | definition.</p> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 848 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 849 | <p>As SSA values, global variables define pointer values that are in scope |
| 850 | (i.e. they dominate) all basic blocks in the program. Global variables |
| 851 | always define a pointer to their "content" type because they describe a |
| 852 | region of memory, and all memory objects in LLVM are accessed through |
| 853 | pointers.</p> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 854 | |
| Rafael Espindola | bea4626 | 2011-01-08 16:42:36 +0000 | [diff] [blame] | 855 | <p>Global variables can be marked with <tt>unnamed_addr</tt> which indicates |
| 856 | that the address is not significant, only the content. Constants marked |
| Rafael Espindola | a5eaa86 | 2011-01-15 08:20:57 +0000 | [diff] [blame] | 857 | like this can be merged with other constants if they have the same |
| 858 | initializer. Note that a constant with significant address <em>can</em> |
| 859 | be merged with a <tt>unnamed_addr</tt> constant, the result being a |
| 860 | constant whose address is significant.</p> |
| Rafael Espindola | bea4626 | 2011-01-08 16:42:36 +0000 | [diff] [blame] | 861 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 862 | <p>A global variable may be declared to reside in a target-specific numbered |
| 863 | address space. For targets that support them, address spaces may affect how |
| 864 | optimizations are performed and/or what target instructions are used to |
| 865 | access the variable. The default address space is zero. The address space |
| 866 | qualifier must precede any other attributes.</p> |
| Christopher Lamb | 284d992 | 2007-12-11 09:31:00 +0000 | [diff] [blame] | 867 | |
| Chris Lattner | 88f6c46 | 2005-11-12 00:45:07 +0000 | [diff] [blame] | 868 | <p>LLVM allows an explicit section to be specified for globals. If the target |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 869 | supports it, it will emit globals to the section specified.</p> |
| Chris Lattner | 88f6c46 | 2005-11-12 00:45:07 +0000 | [diff] [blame] | 870 | |
| Chris Lattner | ce99fa9 | 2010-04-28 00:13:42 +0000 | [diff] [blame] | 871 | <p>An explicit alignment may be specified for a global, which must be a power |
| 872 | of 2. If not present, or if the alignment is set to zero, the alignment of |
| 873 | the global is set by the target to whatever it feels convenient. If an |
| 874 | explicit alignment is specified, the global is forced to have exactly that |
| Chris Lattner | 2d4b8ee | 2010-04-28 00:31:12 +0000 | [diff] [blame] | 875 | alignment. Targets and optimizers are not allowed to over-align the global |
| 876 | if the global has an assigned section. In this case, the extra alignment |
| 877 | could be observable: for example, code could assume that the globals are |
| 878 | densely packed in their section and try to iterate over them as an array, |
| 879 | alignment padding would break this iteration.</p> |
| Chris Lattner | 2cbdc45 | 2005-11-06 08:02:57 +0000 | [diff] [blame] | 880 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 881 | <p>For example, the following defines a global in a numbered address space with |
| 882 | an initializer, section, and alignment:</p> |
| Chris Lattner | 68027ea | 2007-01-14 00:27:09 +0000 | [diff] [blame] | 883 | |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 884 | <pre class="doc_code"> |
| Dan Gohman | 398873c | 2009-01-11 00:40:00 +0000 | [diff] [blame] | 885 | @G = addrspace(5) constant float 1.0, section "foo", align 4 |
| Chris Lattner | 68027ea | 2007-01-14 00:27:09 +0000 | [diff] [blame] | 886 | </pre> |
| 887 | |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 888 | </div> |
| 889 | |
| 890 | |
| 891 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 892 | <h3> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 893 | <a name="functionstructure">Functions</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 894 | </h3> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 895 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 896 | <div> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 897 | |
| Dan Gohman | b55a1ee | 2010-03-01 17:41:39 +0000 | [diff] [blame] | 898 | <p>LLVM function definitions consist of the "<tt>define</tt>" keyword, an |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 899 | optional <a href="#linkage">linkage type</a>, an optional |
| 900 | <a href="#visibility">visibility style</a>, an optional |
| Rafael Espindola | bea4626 | 2011-01-08 16:42:36 +0000 | [diff] [blame] | 901 | <a href="#callingconv">calling convention</a>, |
| 902 | an optional <tt>unnamed_addr</tt> attribute, a return type, an optional |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 903 | <a href="#paramattrs">parameter attribute</a> for the return type, a function |
| 904 | name, a (possibly empty) argument list (each with optional |
| 905 | <a href="#paramattrs">parameter attributes</a>), optional |
| 906 | <a href="#fnattrs">function attributes</a>, an optional section, an optional |
| 907 | alignment, an optional <a href="#gc">garbage collector name</a>, an opening |
| 908 | curly brace, a list of basic blocks, and a closing curly brace.</p> |
| Anton Korobeynikov | 8cea37b | 2007-01-23 12:35:46 +0000 | [diff] [blame] | 909 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 910 | <p>LLVM function declarations consist of the "<tt>declare</tt>" keyword, an |
| 911 | optional <a href="#linkage">linkage type</a>, an optional |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 912 | <a href="#visibility">visibility style</a>, an optional |
| Rafael Espindola | bea4626 | 2011-01-08 16:42:36 +0000 | [diff] [blame] | 913 | <a href="#callingconv">calling convention</a>, |
| 914 | an optional <tt>unnamed_addr</tt> attribute, a return type, an optional |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 915 | <a href="#paramattrs">parameter attribute</a> for the return type, a function |
| 916 | name, a possibly empty list of arguments, an optional alignment, and an |
| 917 | optional <a href="#gc">garbage collector name</a>.</p> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 918 | |
| Chris Lattner | d3eda89 | 2008-08-05 18:29:16 +0000 | [diff] [blame] | 919 | <p>A function definition contains a list of basic blocks, forming the CFG |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 920 | (Control Flow Graph) for the function. Each basic block may optionally start |
| 921 | with a label (giving the basic block a symbol table entry), contains a list |
| 922 | of instructions, and ends with a <a href="#terminators">terminator</a> |
| 923 | instruction (such as a branch or function return).</p> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 924 | |
| Chris Lattner | 4a3c901 | 2007-06-08 16:52:14 +0000 | [diff] [blame] | 925 | <p>The first basic block in a function is special in two ways: it is immediately |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 926 | executed on entrance to the function, and it is not allowed to have |
| 927 | predecessor basic blocks (i.e. there can not be any branches to the entry |
| 928 | block of a function). Because the block can have no predecessors, it also |
| 929 | cannot have any <a href="#i_phi">PHI nodes</a>.</p> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 930 | |
| Chris Lattner | 88f6c46 | 2005-11-12 00:45:07 +0000 | [diff] [blame] | 931 | <p>LLVM allows an explicit section to be specified for functions. If the target |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 932 | supports it, it will emit functions to the section specified.</p> |
| Chris Lattner | 88f6c46 | 2005-11-12 00:45:07 +0000 | [diff] [blame] | 933 | |
| Chris Lattner | 2cbdc45 | 2005-11-06 08:02:57 +0000 | [diff] [blame] | 934 | <p>An explicit alignment may be specified for a function. If not present, or if |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 935 | the alignment is set to zero, the alignment of the function is set by the |
| 936 | target to whatever it feels convenient. If an explicit alignment is |
| 937 | specified, the function is forced to have at least that much alignment. All |
| 938 | alignments must be a power of 2.</p> |
| Chris Lattner | 2cbdc45 | 2005-11-06 08:02:57 +0000 | [diff] [blame] | 939 | |
| Rafael Espindola | bea4626 | 2011-01-08 16:42:36 +0000 | [diff] [blame] | 940 | <p>If the <tt>unnamed_addr</tt> attribute is given, the address is know to not |
| Bill Wendling | 5c3a9f7 | 2011-11-04 20:40:41 +0000 | [diff] [blame] | 941 | be significant and two identical functions can be merged.</p> |
| Rafael Espindola | bea4626 | 2011-01-08 16:42:36 +0000 | [diff] [blame] | 942 | |
| Bill Wendling | c39e3e0 | 2009-07-20 02:39:26 +0000 | [diff] [blame] | 943 | <h5>Syntax:</h5> |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 944 | <pre class="doc_code"> |
| Chris Lattner | 50ad45c | 2008-10-13 16:55:18 +0000 | [diff] [blame] | 945 | define [<a href="#linkage">linkage</a>] [<a href="#visibility">visibility</a>] |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 946 | [<a href="#callingconv">cconv</a>] [<a href="#paramattrs">ret attrs</a>] |
| 947 | <ResultType> @<FunctionName> ([argument list]) |
| 948 | [<a href="#fnattrs">fn Attrs</a>] [section "name"] [align N] |
| 949 | [<a href="#gc">gc</a>] { ... } |
| 950 | </pre> |
| Devang Patel | 307e8ab | 2008-10-07 17:48:33 +0000 | [diff] [blame] | 951 | |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 952 | </div> |
| 953 | |
| Anton Korobeynikov | 8b0a8c8 | 2007-04-25 14:27:10 +0000 | [diff] [blame] | 954 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 955 | <h3> |
| Anton Korobeynikov | 8b0a8c8 | 2007-04-25 14:27:10 +0000 | [diff] [blame] | 956 | <a name="aliasstructure">Aliases</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 957 | </h3> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 958 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 959 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 960 | |
| 961 | <p>Aliases act as "second name" for the aliasee value (which can be either |
| 962 | function, global variable, another alias or bitcast of global value). Aliases |
| 963 | may have an optional <a href="#linkage">linkage type</a>, and an |
| 964 | optional <a href="#visibility">visibility style</a>.</p> |
| Anton Korobeynikov | 8b0a8c8 | 2007-04-25 14:27:10 +0000 | [diff] [blame] | 965 | |
| Bill Wendling | c39e3e0 | 2009-07-20 02:39:26 +0000 | [diff] [blame] | 966 | <h5>Syntax:</h5> |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 967 | <pre class="doc_code"> |
| Duncan Sands | 0b23ac1 | 2008-09-12 20:48:21 +0000 | [diff] [blame] | 968 | @<Name> = alias [Linkage] [Visibility] <AliaseeTy> @<Aliasee> |
| Bill Wendling | aac388b | 2007-05-29 09:42:13 +0000 | [diff] [blame] | 969 | </pre> |
| Anton Korobeynikov | 8b0a8c8 | 2007-04-25 14:27:10 +0000 | [diff] [blame] | 970 | |
| 971 | </div> |
| 972 | |
| Chris Lattner | 4e9aba7 | 2006-01-23 23:23:47 +0000 | [diff] [blame] | 973 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 974 | <h3> |
| Devang Patel | cd1fd25 | 2010-01-11 19:35:55 +0000 | [diff] [blame] | 975 | <a name="namedmetadatastructure">Named Metadata</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 976 | </h3> |
| Devang Patel | cd1fd25 | 2010-01-11 19:35:55 +0000 | [diff] [blame] | 977 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 978 | <div> |
| Devang Patel | cd1fd25 | 2010-01-11 19:35:55 +0000 | [diff] [blame] | 979 | |
| Chris Lattner | e6a5ddd | 2010-01-15 21:50:19 +0000 | [diff] [blame] | 980 | <p>Named metadata is a collection of metadata. <a href="#metadata">Metadata |
| Dan Gohman | 872814a | 2010-07-21 18:54:18 +0000 | [diff] [blame] | 981 | nodes</a> (but not metadata strings) are the only valid operands for |
| Chris Lattner | e6a5ddd | 2010-01-15 21:50:19 +0000 | [diff] [blame] | 982 | a named metadata.</p> |
| Devang Patel | cd1fd25 | 2010-01-11 19:35:55 +0000 | [diff] [blame] | 983 | |
| 984 | <h5>Syntax:</h5> |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 985 | <pre class="doc_code"> |
| Dan Gohman | 872814a | 2010-07-21 18:54:18 +0000 | [diff] [blame] | 986 | ; Some unnamed metadata nodes, which are referenced by the named metadata. |
| 987 | !0 = metadata !{metadata !"zero"} |
| Devang Patel | cd1fd25 | 2010-01-11 19:35:55 +0000 | [diff] [blame] | 988 | !1 = metadata !{metadata !"one"} |
| Dan Gohman | 872814a | 2010-07-21 18:54:18 +0000 | [diff] [blame] | 989 | !2 = metadata !{metadata !"two"} |
| Dan Gohman | 1005bc5 | 2010-07-13 19:48:13 +0000 | [diff] [blame] | 990 | ; A named metadata. |
| Dan Gohman | 872814a | 2010-07-21 18:54:18 +0000 | [diff] [blame] | 991 | !name = !{!0, !1, !2} |
| Devang Patel | cd1fd25 | 2010-01-11 19:35:55 +0000 | [diff] [blame] | 992 | </pre> |
| Devang Patel | cd1fd25 | 2010-01-11 19:35:55 +0000 | [diff] [blame] | 993 | |
| 994 | </div> |
| 995 | |
| 996 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 997 | <h3> |
| 998 | <a name="paramattrs">Parameter Attributes</a> |
| 999 | </h3> |
| Reid Spencer | ca86e16 | 2006-12-31 07:07:53 +0000 | [diff] [blame] | 1000 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 1001 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1002 | |
| 1003 | <p>The return type and each parameter of a function type may have a set of |
| 1004 | <i>parameter attributes</i> associated with them. Parameter attributes are |
| 1005 | used to communicate additional information about the result or parameters of |
| 1006 | a function. Parameter attributes are considered to be part of the function, |
| 1007 | not of the function type, so functions with different parameter attributes |
| 1008 | can have the same function type.</p> |
| 1009 | |
| 1010 | <p>Parameter attributes are simple keywords that follow the type specified. If |
| 1011 | multiple parameter attributes are needed, they are space separated. For |
| 1012 | example:</p> |
| Bill Wendling | 2f7a8b0 | 2007-05-29 09:04:49 +0000 | [diff] [blame] | 1013 | |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 1014 | <pre class="doc_code"> |
| Nick Lewycky | b6a7d25 | 2009-02-15 23:06:14 +0000 | [diff] [blame] | 1015 | declare i32 @printf(i8* noalias nocapture, ...) |
| Chris Lattner | 66d922c | 2008-10-04 18:33:34 +0000 | [diff] [blame] | 1016 | declare i32 @atoi(i8 zeroext) |
| 1017 | declare signext i8 @returns_signed_char() |
| Bill Wendling | 2f7a8b0 | 2007-05-29 09:04:49 +0000 | [diff] [blame] | 1018 | </pre> |
| Bill Wendling | 2f7a8b0 | 2007-05-29 09:04:49 +0000 | [diff] [blame] | 1019 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1020 | <p>Note that any attributes for the function result (<tt>nounwind</tt>, |
| 1021 | <tt>readonly</tt>) come immediately after the argument list.</p> |
| Reid Spencer | ca86e16 | 2006-12-31 07:07:53 +0000 | [diff] [blame] | 1022 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1023 | <p>Currently, only the following parameter attributes are defined:</p> |
| Chris Lattner | 47507de | 2008-01-11 06:20:47 +0000 | [diff] [blame] | 1024 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1025 | <dl> |
| Bill Wendling | f82d40a | 2009-11-02 00:24:16 +0000 | [diff] [blame] | 1026 | <dt><tt><b>zeroext</b></tt></dt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1027 | <dd>This indicates to the code generator that the parameter or return value |
| Cameron Zwarich | ebe8173 | 2011-03-16 22:20:18 +0000 | [diff] [blame] | 1028 | should be zero-extended to the extent required by the target's ABI (which |
| 1029 | is usually 32-bits, but is 8-bits for a i1 on x86-64) by the caller (for a |
| 1030 | parameter) or the callee (for a return value).</dd> |
| Chris Lattner | 47507de | 2008-01-11 06:20:47 +0000 | [diff] [blame] | 1031 | |
| Bill Wendling | f82d40a | 2009-11-02 00:24:16 +0000 | [diff] [blame] | 1032 | <dt><tt><b>signext</b></tt></dt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1033 | <dd>This indicates to the code generator that the parameter or return value |
| Cameron Zwarich | 9e69ff9 | 2011-03-17 14:21:58 +0000 | [diff] [blame] | 1034 | should be sign-extended to the extent required by the target's ABI (which |
| 1035 | is usually 32-bits) by the caller (for a parameter) or the callee (for a |
| 1036 | return value).</dd> |
| Chris Lattner | 47507de | 2008-01-11 06:20:47 +0000 | [diff] [blame] | 1037 | |
| Bill Wendling | f82d40a | 2009-11-02 00:24:16 +0000 | [diff] [blame] | 1038 | <dt><tt><b>inreg</b></tt></dt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1039 | <dd>This indicates that this parameter or return value should be treated in a |
| 1040 | special target-dependent fashion during while emitting code for a function |
| 1041 | call or return (usually, by putting it in a register as opposed to memory, |
| 1042 | though some targets use it to distinguish between two different kinds of |
| 1043 | registers). Use of this attribute is target-specific.</dd> |
| Chris Lattner | 47507de | 2008-01-11 06:20:47 +0000 | [diff] [blame] | 1044 | |
| Bill Wendling | f82d40a | 2009-11-02 00:24:16 +0000 | [diff] [blame] | 1045 | <dt><tt><b><a name="byval">byval</a></b></tt></dt> |
| Chris Lattner | a6fd81d | 2010-11-20 23:49:06 +0000 | [diff] [blame] | 1046 | <dd><p>This indicates that the pointer parameter should really be passed by |
| 1047 | value to the function. The attribute implies that a hidden copy of the |
| 1048 | pointee |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1049 | is made between the caller and the callee, so the callee is unable to |
| 1050 | modify the value in the callee. This attribute is only valid on LLVM |
| 1051 | pointer arguments. It is generally used to pass structs and arrays by |
| 1052 | value, but is also valid on pointers to scalars. The copy is considered |
| 1053 | to belong to the caller not the callee (for example, |
| 1054 | <tt><a href="#readonly">readonly</a></tt> functions should not write to |
| 1055 | <tt>byval</tt> parameters). This is not a valid attribute for return |
| Chris Lattner | a6fd81d | 2010-11-20 23:49:06 +0000 | [diff] [blame] | 1056 | values.</p> |
| 1057 | |
| 1058 | <p>The byval attribute also supports specifying an alignment with |
| 1059 | the align attribute. It indicates the alignment of the stack slot to |
| 1060 | form and the known alignment of the pointer specified to the call site. If |
| 1061 | the alignment is not specified, then the code generator makes a |
| 1062 | target-specific assumption.</p></dd> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1063 | |
| Dan Gohman | ff23535 | 2010-07-02 23:18:08 +0000 | [diff] [blame] | 1064 | <dt><tt><b><a name="sret">sret</a></b></tt></dt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1065 | <dd>This indicates that the pointer parameter specifies the address of a |
| 1066 | structure that is the return value of the function in the source program. |
| 1067 | This pointer must be guaranteed by the caller to be valid: loads and |
| 1068 | stores to the structure may be assumed by the callee to not to trap. This |
| 1069 | may only be applied to the first parameter. This is not a valid attribute |
| 1070 | for return values. </dd> |
| 1071 | |
| Dan Gohman | ff23535 | 2010-07-02 23:18:08 +0000 | [diff] [blame] | 1072 | <dt><tt><b><a name="noalias">noalias</a></b></tt></dt> |
| Dan Gohman | 1e10962 | 2010-07-02 18:41:32 +0000 | [diff] [blame] | 1073 | <dd>This indicates that pointer values |
| 1074 | <a href="#pointeraliasing"><i>based</i></a> on the argument or return |
| Dan Gohman | efca7f9 | 2010-07-02 23:46:54 +0000 | [diff] [blame] | 1075 | value do not alias pointer values which are not <i>based</i> on it, |
| 1076 | ignoring certain "irrelevant" dependencies. |
| 1077 | For a call to the parent function, dependencies between memory |
| 1078 | references from before or after the call and from those during the call |
| 1079 | are "irrelevant" to the <tt>noalias</tt> keyword for the arguments and |
| 1080 | return value used in that call. |
| Dan Gohman | 1e10962 | 2010-07-02 18:41:32 +0000 | [diff] [blame] | 1081 | The caller shares the responsibility with the callee for ensuring that |
| 1082 | these requirements are met. |
| 1083 | For further details, please see the discussion of the NoAlias response in |
| Dan Gohman | ff70fe4 | 2010-07-06 15:26:33 +0000 | [diff] [blame] | 1084 | <a href="AliasAnalysis.html#MustMayNo">alias analysis</a>.<br> |
| 1085 | <br> |
| John McCall | 191d4ee | 2010-07-06 21:07:14 +0000 | [diff] [blame] | 1086 | Note that this definition of <tt>noalias</tt> is intentionally |
| 1087 | similar to the definition of <tt>restrict</tt> in C99 for function |
| Chris Lattner | 211244a | 2010-07-06 20:51:35 +0000 | [diff] [blame] | 1088 | arguments, though it is slightly weaker. |
| Dan Gohman | ff70fe4 | 2010-07-06 15:26:33 +0000 | [diff] [blame] | 1089 | <br> |
| 1090 | For function return values, C99's <tt>restrict</tt> is not meaningful, |
| 1091 | while LLVM's <tt>noalias</tt> is. |
| 1092 | </dd> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1093 | |
| Dan Gohman | ff23535 | 2010-07-02 23:18:08 +0000 | [diff] [blame] | 1094 | <dt><tt><b><a name="nocapture">nocapture</a></b></tt></dt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1095 | <dd>This indicates that the callee does not make any copies of the pointer |
| 1096 | that outlive the callee itself. This is not a valid attribute for return |
| 1097 | values.</dd> |
| 1098 | |
| Dan Gohman | ff23535 | 2010-07-02 23:18:08 +0000 | [diff] [blame] | 1099 | <dt><tt><b><a name="nest">nest</a></b></tt></dt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1100 | <dd>This indicates that the pointer parameter can be excised using the |
| 1101 | <a href="#int_trampoline">trampoline intrinsics</a>. This is not a valid |
| 1102 | attribute for return values.</dd> |
| 1103 | </dl> |
| Reid Spencer | ca86e16 | 2006-12-31 07:07:53 +0000 | [diff] [blame] | 1104 | |
| Reid Spencer | ca86e16 | 2006-12-31 07:07:53 +0000 | [diff] [blame] | 1105 | </div> |
| 1106 | |
| 1107 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 1108 | <h3> |
| Gordon Henriksen | 80a75bf | 2007-12-10 03:18:06 +0000 | [diff] [blame] | 1109 | <a name="gc">Garbage Collector Names</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 1110 | </h3> |
| Gordon Henriksen | 80a75bf | 2007-12-10 03:18:06 +0000 | [diff] [blame] | 1111 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 1112 | <div> |
| Gordon Henriksen | 80a75bf | 2007-12-10 03:18:06 +0000 | [diff] [blame] | 1113 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1114 | <p>Each function may specify a garbage collector name, which is simply a |
| 1115 | string:</p> |
| 1116 | |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 1117 | <pre class="doc_code"> |
| Bill Wendling | f82d40a | 2009-11-02 00:24:16 +0000 | [diff] [blame] | 1118 | define void @f() gc "name" { ... } |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1119 | </pre> |
| Gordon Henriksen | 80a75bf | 2007-12-10 03:18:06 +0000 | [diff] [blame] | 1120 | |
| 1121 | <p>The compiler declares the supported values of <i>name</i>. Specifying a |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1122 | collector which will cause the compiler to alter its output in order to |
| 1123 | support the named garbage collection algorithm.</p> |
| 1124 | |
| Gordon Henriksen | 80a75bf | 2007-12-10 03:18:06 +0000 | [diff] [blame] | 1125 | </div> |
| 1126 | |
| 1127 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 1128 | <h3> |
| Devang Patel | 2c9c3e7 | 2008-09-26 23:51:19 +0000 | [diff] [blame] | 1129 | <a name="fnattrs">Function Attributes</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 1130 | </h3> |
| Devang Patel | f8b9481 | 2008-09-04 23:05:13 +0000 | [diff] [blame] | 1131 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 1132 | <div> |
| Devang Patel | 2c9c3e7 | 2008-09-26 23:51:19 +0000 | [diff] [blame] | 1133 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1134 | <p>Function attributes are set to communicate additional information about a |
| 1135 | function. Function attributes are considered to be part of the function, not |
| 1136 | of the function type, so functions with different parameter attributes can |
| 1137 | have the same function type.</p> |
| Devang Patel | 2c9c3e7 | 2008-09-26 23:51:19 +0000 | [diff] [blame] | 1138 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1139 | <p>Function attributes are simple keywords that follow the type specified. If |
| 1140 | multiple attributes are needed, they are space separated. For example:</p> |
| Devang Patel | f8b9481 | 2008-09-04 23:05:13 +0000 | [diff] [blame] | 1141 | |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 1142 | <pre class="doc_code"> |
| Devang Patel | 2c9c3e7 | 2008-09-26 23:51:19 +0000 | [diff] [blame] | 1143 | define void @f() noinline { ... } |
| 1144 | define void @f() alwaysinline { ... } |
| 1145 | define void @f() alwaysinline optsize { ... } |
| Bill Wendling | f82d40a | 2009-11-02 00:24:16 +0000 | [diff] [blame] | 1146 | define void @f() optsize { ... } |
| Bill Wendling | e36dccc | 2008-09-07 10:26:33 +0000 | [diff] [blame] | 1147 | </pre> |
| Devang Patel | f8b9481 | 2008-09-04 23:05:13 +0000 | [diff] [blame] | 1148 | |
| Bill Wendling | e36dccc | 2008-09-07 10:26:33 +0000 | [diff] [blame] | 1149 | <dl> |
| Kostya Serebryany | 164b86b | 2012-01-20 17:56:17 +0000 | [diff] [blame] | 1150 | <dt><tt><b>address_safety</b></tt></dt> |
| 1151 | <dd>This attribute indicates that the address safety analysis |
| 1152 | is enabled for this function. </dd> |
| 1153 | |
| Charles Davis | 1e063d1 | 2010-02-12 00:31:15 +0000 | [diff] [blame] | 1154 | <dt><tt><b>alignstack(<<em>n</em>>)</b></tt></dt> |
| 1155 | <dd>This attribute indicates that, when emitting the prologue and epilogue, |
| 1156 | the backend should forcibly align the stack pointer. Specify the |
| 1157 | desired alignment, which must be a power of two, in parentheses. |
| 1158 | |
| Bill Wendling | f82d40a | 2009-11-02 00:24:16 +0000 | [diff] [blame] | 1159 | <dt><tt><b>alwaysinline</b></tt></dt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1160 | <dd>This attribute indicates that the inliner should attempt to inline this |
| 1161 | function into callers whenever possible, ignoring any active inlining size |
| 1162 | threshold for this caller.</dd> |
| Bill Wendling | e36dccc | 2008-09-07 10:26:33 +0000 | [diff] [blame] | 1163 | |
| Dan Gohman | 129bd56 | 2011-06-16 16:03:13 +0000 | [diff] [blame] | 1164 | <dt><tt><b>nonlazybind</b></tt></dt> |
| 1165 | <dd>This attribute suppresses lazy symbol binding for the function. This |
| 1166 | may make calls to the function faster, at the cost of extra program |
| 1167 | startup time if the function is not called during program startup.</dd> |
| 1168 | |
| Jakob Stoklund Olesen | 570a4a5 | 2010-02-06 01:16:28 +0000 | [diff] [blame] | 1169 | <dt><tt><b>inlinehint</b></tt></dt> |
| 1170 | <dd>This attribute indicates that the source code contained a hint that inlining |
| 1171 | this function is desirable (such as the "inline" keyword in C/C++). It |
| 1172 | is just a hint; it imposes no requirements on the inliner.</dd> |
| 1173 | |
| Nick Lewycky | 76ec37a | 2010-07-06 18:24:09 +0000 | [diff] [blame] | 1174 | <dt><tt><b>naked</b></tt></dt> |
| 1175 | <dd>This attribute disables prologue / epilogue emission for the function. |
| 1176 | This can have very system-specific consequences.</dd> |
| 1177 | |
| 1178 | <dt><tt><b>noimplicitfloat</b></tt></dt> |
| 1179 | <dd>This attributes disables implicit floating point instructions.</dd> |
| 1180 | |
| Bill Wendling | f82d40a | 2009-11-02 00:24:16 +0000 | [diff] [blame] | 1181 | <dt><tt><b>noinline</b></tt></dt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1182 | <dd>This attribute indicates that the inliner should never inline this |
| 1183 | function in any situation. This attribute may not be used together with |
| 1184 | the <tt>alwaysinline</tt> attribute.</dd> |
| Devang Patel | 2c9c3e7 | 2008-09-26 23:51:19 +0000 | [diff] [blame] | 1185 | |
| Nick Lewycky | 76ec37a | 2010-07-06 18:24:09 +0000 | [diff] [blame] | 1186 | <dt><tt><b>noredzone</b></tt></dt> |
| 1187 | <dd>This attribute indicates that the code generator should not use a red |
| 1188 | zone, even if the target-specific ABI normally permits it.</dd> |
| Devang Patel | 2c9c3e7 | 2008-09-26 23:51:19 +0000 | [diff] [blame] | 1189 | |
| Bill Wendling | f82d40a | 2009-11-02 00:24:16 +0000 | [diff] [blame] | 1190 | <dt><tt><b>noreturn</b></tt></dt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1191 | <dd>This function attribute indicates that the function never returns |
| 1192 | normally. This produces undefined behavior at runtime if the function |
| 1193 | ever does dynamically return.</dd> |
| Bill Wendling | 31359ba | 2008-11-13 01:02:51 +0000 | [diff] [blame] | 1194 | |
| Bill Wendling | f82d40a | 2009-11-02 00:24:16 +0000 | [diff] [blame] | 1195 | <dt><tt><b>nounwind</b></tt></dt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1196 | <dd>This function attribute indicates that the function never returns with an |
| 1197 | unwind or exceptional control flow. If the function does unwind, its |
| 1198 | runtime behavior is undefined.</dd> |
| Bill Wendling | fbaa7ed | 2008-11-26 19:07:40 +0000 | [diff] [blame] | 1199 | |
| Nick Lewycky | 76ec37a | 2010-07-06 18:24:09 +0000 | [diff] [blame] | 1200 | <dt><tt><b>optsize</b></tt></dt> |
| 1201 | <dd>This attribute suggests that optimization passes and code generator passes |
| 1202 | make choices that keep the code size of this function low, and otherwise |
| 1203 | do optimizations specifically to reduce code size.</dd> |
| 1204 | |
| Bill Wendling | f82d40a | 2009-11-02 00:24:16 +0000 | [diff] [blame] | 1205 | <dt><tt><b>readnone</b></tt></dt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1206 | <dd>This attribute indicates that the function computes its result (or decides |
| 1207 | to unwind an exception) based strictly on its arguments, without |
| 1208 | dereferencing any pointer arguments or otherwise accessing any mutable |
| 1209 | state (e.g. memory, control registers, etc) visible to caller functions. |
| 1210 | It does not write through any pointer arguments |
| 1211 | (including <tt><a href="#byval">byval</a></tt> arguments) and never |
| 1212 | changes any state visible to callers. This means that it cannot unwind |
| Bill Wendling | 7b9e539 | 2012-02-06 21:57:33 +0000 | [diff] [blame] | 1213 | exceptions by calling the <tt>C++</tt> exception throwing methods.</dd> |
| Devang Patel | 5d96fda | 2009-06-12 19:45:19 +0000 | [diff] [blame] | 1214 | |
| Bill Wendling | f82d40a | 2009-11-02 00:24:16 +0000 | [diff] [blame] | 1215 | <dt><tt><b><a name="readonly">readonly</a></b></tt></dt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1216 | <dd>This attribute indicates that the function does not write through any |
| 1217 | pointer arguments (including <tt><a href="#byval">byval</a></tt> |
| 1218 | arguments) or otherwise modify any state (e.g. memory, control registers, |
| 1219 | etc) visible to caller functions. It may dereference pointer arguments |
| 1220 | and read state that may be set in the caller. A readonly function always |
| 1221 | returns the same value (or unwinds an exception identically) when called |
| 1222 | with the same set of arguments and global state. It cannot unwind an |
| Bill Wendling | 7b9e539 | 2012-02-06 21:57:33 +0000 | [diff] [blame] | 1223 | exception by calling the <tt>C++</tt> exception throwing methods.</dd> |
| Anton Korobeynikov | c5ec8a7 | 2009-07-17 18:07:26 +0000 | [diff] [blame] | 1224 | |
| Bill Wendling | 9bd5d04 | 2011-12-05 21:27:54 +0000 | [diff] [blame] | 1225 | <dt><tt><b><a name="returns_twice">returns_twice</a></b></tt></dt> |
| 1226 | <dd>This attribute indicates that this function can return twice. The |
| 1227 | C <code>setjmp</code> is an example of such a function. The compiler |
| 1228 | disables some optimizations (like tail calls) in the caller of these |
| 1229 | functions.</dd> |
| 1230 | |
| Bill Wendling | f82d40a | 2009-11-02 00:24:16 +0000 | [diff] [blame] | 1231 | <dt><tt><b><a name="ssp">ssp</a></b></tt></dt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1232 | <dd>This attribute indicates that the function should emit a stack smashing |
| 1233 | protector. It is in the form of a "canary"—a random value placed on |
| 1234 | the stack before the local variables that's checked upon return from the |
| 1235 | function to see if it has been overwritten. A heuristic is used to |
| 1236 | determine if a function needs stack protectors or not.<br> |
| 1237 | <br> |
| 1238 | If a function that has an <tt>ssp</tt> attribute is inlined into a |
| 1239 | function that doesn't have an <tt>ssp</tt> attribute, then the resulting |
| 1240 | function will have an <tt>ssp</tt> attribute.</dd> |
| 1241 | |
| Bill Wendling | f82d40a | 2009-11-02 00:24:16 +0000 | [diff] [blame] | 1242 | <dt><tt><b>sspreq</b></tt></dt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1243 | <dd>This attribute indicates that the function should <em>always</em> emit a |
| 1244 | stack smashing protector. This overrides |
| Bill Wendling | c39e3e0 | 2009-07-20 02:39:26 +0000 | [diff] [blame] | 1245 | the <tt><a href="#ssp">ssp</a></tt> function attribute.<br> |
| 1246 | <br> |
| 1247 | If a function that has an <tt>sspreq</tt> attribute is inlined into a |
| 1248 | function that doesn't have an <tt>sspreq</tt> attribute or which has |
| 1249 | an <tt>ssp</tt> attribute, then the resulting function will have |
| 1250 | an <tt>sspreq</tt> attribute.</dd> |
| Rafael Espindola | fbff0ec | 2011-07-25 15:27:59 +0000 | [diff] [blame] | 1251 | |
| 1252 | <dt><tt><b><a name="uwtable">uwtable</a></b></tt></dt> |
| 1253 | <dd>This attribute indicates that the ABI being targeted requires that |
| 1254 | an unwind table entry be produce for this function even if we can |
| 1255 | show that no exceptions passes by it. This is normally the case for |
| 1256 | the ELF x86-64 abi, but it can be disabled for some compilation |
| 1257 | units.</dd> |
| Bill Wendling | e36dccc | 2008-09-07 10:26:33 +0000 | [diff] [blame] | 1258 | </dl> |
| 1259 | |
| Devang Patel | f8b9481 | 2008-09-04 23:05:13 +0000 | [diff] [blame] | 1260 | </div> |
| 1261 | |
| 1262 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 1263 | <h3> |
| Chris Lattner | 1eeeb0c | 2006-04-08 04:40:53 +0000 | [diff] [blame] | 1264 | <a name="moduleasm">Module-Level Inline Assembly</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 1265 | </h3> |
| Chris Lattner | 4e9aba7 | 2006-01-23 23:23:47 +0000 | [diff] [blame] | 1266 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 1267 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1268 | |
| 1269 | <p>Modules may contain "module-level inline asm" blocks, which corresponds to |
| 1270 | the GCC "file scope inline asm" blocks. These blocks are internally |
| 1271 | concatenated by LLVM and treated as a single unit, but may be separated in |
| 1272 | the <tt>.ll</tt> file if desired. The syntax is very simple:</p> |
| Chris Lattner | 4e9aba7 | 2006-01-23 23:23:47 +0000 | [diff] [blame] | 1273 | |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 1274 | <pre class="doc_code"> |
| Bill Wendling | 2f7a8b0 | 2007-05-29 09:04:49 +0000 | [diff] [blame] | 1275 | module asm "inline asm code goes here" |
| 1276 | module asm "more can go here" |
| 1277 | </pre> |
| Chris Lattner | 4e9aba7 | 2006-01-23 23:23:47 +0000 | [diff] [blame] | 1278 | |
| 1279 | <p>The strings can contain any character by escaping non-printable characters. |
| 1280 | The escape sequence used is simply "\xx" where "xx" is the two digit hex code |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1281 | for the number.</p> |
| Chris Lattner | 4e9aba7 | 2006-01-23 23:23:47 +0000 | [diff] [blame] | 1282 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1283 | <p>The inline asm code is simply printed to the machine code .s file when |
| 1284 | assembly code is generated.</p> |
| 1285 | |
| Chris Lattner | 4e9aba7 | 2006-01-23 23:23:47 +0000 | [diff] [blame] | 1286 | </div> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 1287 | |
| Reid Spencer | de15194 | 2007-02-19 23:54:10 +0000 | [diff] [blame] | 1288 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 1289 | <h3> |
| Reid Spencer | de15194 | 2007-02-19 23:54:10 +0000 | [diff] [blame] | 1290 | <a name="datalayout">Data Layout</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 1291 | </h3> |
| Reid Spencer | de15194 | 2007-02-19 23:54:10 +0000 | [diff] [blame] | 1292 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 1293 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1294 | |
| Reid Spencer | de15194 | 2007-02-19 23:54:10 +0000 | [diff] [blame] | 1295 | <p>A module may specify a target specific data layout string that specifies how |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1296 | data is to be laid out in memory. The syntax for the data layout is |
| 1297 | simply:</p> |
| 1298 | |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 1299 | <pre class="doc_code"> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1300 | target datalayout = "<i>layout specification</i>" |
| 1301 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1302 | |
| 1303 | <p>The <i>layout specification</i> consists of a list of specifications |
| 1304 | separated by the minus sign character ('-'). Each specification starts with |
| 1305 | a letter and may include other information after the letter to define some |
| 1306 | aspect of the data layout. The specifications accepted are as follows:</p> |
| 1307 | |
| Reid Spencer | de15194 | 2007-02-19 23:54:10 +0000 | [diff] [blame] | 1308 | <dl> |
| 1309 | <dt><tt>E</tt></dt> |
| 1310 | <dd>Specifies that the target lays out data in big-endian form. That is, the |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1311 | bits with the most significance have the lowest address location.</dd> |
| 1312 | |
| Reid Spencer | de15194 | 2007-02-19 23:54:10 +0000 | [diff] [blame] | 1313 | <dt><tt>e</tt></dt> |
| Chris Lattner | d3eda89 | 2008-08-05 18:29:16 +0000 | [diff] [blame] | 1314 | <dd>Specifies that the target lays out data in little-endian form. That is, |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1315 | the bits with the least significance have the lowest address |
| 1316 | location.</dd> |
| 1317 | |
| Lang Hames | bb5b3f3 | 2011-10-10 23:42:08 +0000 | [diff] [blame] | 1318 | <dt><tt>S<i>size</i></tt></dt> |
| 1319 | <dd>Specifies the natural alignment of the stack in bits. Alignment promotion |
| 1320 | of stack variables is limited to the natural stack alignment to avoid |
| 1321 | dynamic stack realignment. The stack alignment must be a multiple of |
| Lang Hames | 5f119a6 | 2011-10-11 17:50:14 +0000 | [diff] [blame] | 1322 | 8-bits. If omitted, the natural stack alignment defaults to "unspecified", |
| 1323 | which does not prevent any alignment promotions.</dd> |
| Lang Hames | bb5b3f3 | 2011-10-10 23:42:08 +0000 | [diff] [blame] | 1324 | |
| Reid Spencer | de15194 | 2007-02-19 23:54:10 +0000 | [diff] [blame] | 1325 | <dt><tt>p:<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt> |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 1326 | <dd>This specifies the <i>size</i> of a pointer and its <i>abi</i> and |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1327 | <i>preferred</i> alignments. All sizes are in bits. Specifying |
| 1328 | the <i>pref</i> alignment is optional. If omitted, the |
| 1329 | preceding <tt>:</tt> should be omitted too.</dd> |
| 1330 | |
| Reid Spencer | de15194 | 2007-02-19 23:54:10 +0000 | [diff] [blame] | 1331 | <dt><tt>i<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt> |
| 1332 | <dd>This specifies the alignment for an integer type of a given bit |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1333 | <i>size</i>. The value of <i>size</i> must be in the range [1,2^23).</dd> |
| 1334 | |
| Reid Spencer | de15194 | 2007-02-19 23:54:10 +0000 | [diff] [blame] | 1335 | <dt><tt>v<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt> |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 1336 | <dd>This specifies the alignment for a vector type of a given bit |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1337 | <i>size</i>.</dd> |
| 1338 | |
| Reid Spencer | de15194 | 2007-02-19 23:54:10 +0000 | [diff] [blame] | 1339 | <dt><tt>f<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt> |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 1340 | <dd>This specifies the alignment for a floating point type of a given bit |
| Dale Johannesen | 9d8d221 | 2010-05-28 18:54:47 +0000 | [diff] [blame] | 1341 | <i>size</i>. Only values of <i>size</i> that are supported by the target |
| 1342 | will work. 32 (float) and 64 (double) are supported on all targets; |
| 1343 | 80 or 128 (different flavors of long double) are also supported on some |
| 1344 | targets. |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1345 | |
| Reid Spencer | de15194 | 2007-02-19 23:54:10 +0000 | [diff] [blame] | 1346 | <dt><tt>a<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt> |
| 1347 | <dd>This specifies the alignment for an aggregate type of a given bit |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1348 | <i>size</i>.</dd> |
| 1349 | |
| Daniel Dunbar | 87bde0b | 2009-06-08 22:17:53 +0000 | [diff] [blame] | 1350 | <dt><tt>s<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt> |
| 1351 | <dd>This specifies the alignment for a stack object of a given bit |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1352 | <i>size</i>.</dd> |
| Chris Lattner | e82bdc4 | 2009-11-07 09:35:34 +0000 | [diff] [blame] | 1353 | |
| 1354 | <dt><tt>n<i>size1</i>:<i>size2</i>:<i>size3</i>...</tt></dt> |
| 1355 | <dd>This specifies a set of native integer widths for the target CPU |
| 1356 | in bits. For example, it might contain "n32" for 32-bit PowerPC, |
| 1357 | "n32:64" for PowerPC 64, or "n8:16:32:64" for X86-64. Elements of |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 1358 | this set are considered to support most general arithmetic |
| Chris Lattner | e82bdc4 | 2009-11-07 09:35:34 +0000 | [diff] [blame] | 1359 | operations efficiently.</dd> |
| Reid Spencer | de15194 | 2007-02-19 23:54:10 +0000 | [diff] [blame] | 1360 | </dl> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1361 | |
| Reid Spencer | de15194 | 2007-02-19 23:54:10 +0000 | [diff] [blame] | 1362 | <p>When constructing the data layout for a given target, LLVM starts with a |
| Dan Gohman | 1c70c00 | 2010-04-28 00:36:01 +0000 | [diff] [blame] | 1363 | default set of specifications which are then (possibly) overridden by the |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1364 | specifications in the <tt>datalayout</tt> keyword. The default specifications |
| 1365 | are given in this list:</p> |
| 1366 | |
| Reid Spencer | de15194 | 2007-02-19 23:54:10 +0000 | [diff] [blame] | 1367 | <ul> |
| 1368 | <li><tt>E</tt> - big endian</li> |
| Dan Gohman | fdf2e8c | 2010-02-23 02:44:03 +0000 | [diff] [blame] | 1369 | <li><tt>p:64:64:64</tt> - 64-bit pointers with 64-bit alignment</li> |
| Reid Spencer | de15194 | 2007-02-19 23:54:10 +0000 | [diff] [blame] | 1370 | <li><tt>i1:8:8</tt> - i1 is 8-bit (byte) aligned</li> |
| 1371 | <li><tt>i8:8:8</tt> - i8 is 8-bit (byte) aligned</li> |
| 1372 | <li><tt>i16:16:16</tt> - i16 is 16-bit aligned</li> |
| 1373 | <li><tt>i32:32:32</tt> - i32 is 32-bit aligned</li> |
| Chris Lattner | d3eda89 | 2008-08-05 18:29:16 +0000 | [diff] [blame] | 1374 | <li><tt>i64:32:64</tt> - i64 has ABI alignment of 32-bits but preferred |
| Reid Spencer | de15194 | 2007-02-19 23:54:10 +0000 | [diff] [blame] | 1375 | alignment of 64-bits</li> |
| 1376 | <li><tt>f32:32:32</tt> - float is 32-bit aligned</li> |
| 1377 | <li><tt>f64:64:64</tt> - double is 64-bit aligned</li> |
| 1378 | <li><tt>v64:64:64</tt> - 64-bit vector is 64-bit aligned</li> |
| 1379 | <li><tt>v128:128:128</tt> - 128-bit vector is 128-bit aligned</li> |
| 1380 | <li><tt>a0:0:1</tt> - aggregates are 8-bit aligned</li> |
| Daniel Dunbar | 87bde0b | 2009-06-08 22:17:53 +0000 | [diff] [blame] | 1381 | <li><tt>s0:64:64</tt> - stack objects are 64-bit aligned</li> |
| Reid Spencer | de15194 | 2007-02-19 23:54:10 +0000 | [diff] [blame] | 1382 | </ul> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1383 | |
| 1384 | <p>When LLVM is determining the alignment for a given type, it uses the |
| 1385 | following rules:</p> |
| 1386 | |
| Reid Spencer | de15194 | 2007-02-19 23:54:10 +0000 | [diff] [blame] | 1387 | <ol> |
| 1388 | <li>If the type sought is an exact match for one of the specifications, that |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1389 | specification is used.</li> |
| 1390 | |
| Reid Spencer | de15194 | 2007-02-19 23:54:10 +0000 | [diff] [blame] | 1391 | <li>If no match is found, and the type sought is an integer type, then the |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1392 | smallest integer type that is larger than the bitwidth of the sought type |
| 1393 | is used. If none of the specifications are larger than the bitwidth then |
| 1394 | the the largest integer type is used. For example, given the default |
| 1395 | specifications above, the i7 type will use the alignment of i8 (next |
| 1396 | largest) while both i65 and i256 will use the alignment of i64 (largest |
| 1397 | specified).</li> |
| 1398 | |
| Reid Spencer | de15194 | 2007-02-19 23:54:10 +0000 | [diff] [blame] | 1399 | <li>If no match is found, and the type sought is a vector type, then the |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1400 | largest vector type that is smaller than the sought vector type will be |
| 1401 | used as a fall back. This happens because <128 x double> can be |
| 1402 | implemented in terms of 64 <2 x double>, for example.</li> |
| Reid Spencer | de15194 | 2007-02-19 23:54:10 +0000 | [diff] [blame] | 1403 | </ol> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1404 | |
| Chris Lattner | 6509f50 | 2011-10-11 23:01:39 +0000 | [diff] [blame] | 1405 | <p>The function of the data layout string may not be what you expect. Notably, |
| 1406 | this is not a specification from the frontend of what alignment the code |
| 1407 | generator should use.</p> |
| 1408 | |
| 1409 | <p>Instead, if specified, the target data layout is required to match what the |
| 1410 | ultimate <em>code generator</em> expects. This string is used by the |
| 1411 | mid-level optimizers to |
| 1412 | improve code, and this only works if it matches what the ultimate code |
| 1413 | generator uses. If you would like to generate IR that does not embed this |
| 1414 | target-specific detail into the IR, then you don't have to specify the |
| 1415 | string. This will disable some optimizations that require precise layout |
| 1416 | information, but this also prevents those optimizations from introducing |
| 1417 | target specificity into the IR.</p> |
| 1418 | |
| 1419 | |
| 1420 | |
| Reid Spencer | de15194 | 2007-02-19 23:54:10 +0000 | [diff] [blame] | 1421 | </div> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 1422 | |
| Dan Gohman | 556ca27 | 2009-07-27 18:07:55 +0000 | [diff] [blame] | 1423 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 1424 | <h3> |
| Dan Gohman | 556ca27 | 2009-07-27 18:07:55 +0000 | [diff] [blame] | 1425 | <a name="pointeraliasing">Pointer Aliasing Rules</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 1426 | </h3> |
| Dan Gohman | 556ca27 | 2009-07-27 18:07:55 +0000 | [diff] [blame] | 1427 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 1428 | <div> |
| Dan Gohman | 556ca27 | 2009-07-27 18:07:55 +0000 | [diff] [blame] | 1429 | |
| Andreas Bolka | 55e459a | 2009-07-29 00:02:05 +0000 | [diff] [blame] | 1430 | <p>Any memory access must be done through a pointer value associated |
| Andreas Bolka | 99a8205 | 2009-07-27 20:37:10 +0000 | [diff] [blame] | 1431 | with an address range of the memory access, otherwise the behavior |
| Dan Gohman | 556ca27 | 2009-07-27 18:07:55 +0000 | [diff] [blame] | 1432 | is undefined. Pointer values are associated with address ranges |
| 1433 | according to the following rules:</p> |
| 1434 | |
| 1435 | <ul> |
| Dan Gohman | 1e10962 | 2010-07-02 18:41:32 +0000 | [diff] [blame] | 1436 | <li>A pointer value is associated with the addresses associated with |
| 1437 | any value it is <i>based</i> on. |
| Andreas Bolka | 55e459a | 2009-07-29 00:02:05 +0000 | [diff] [blame] | 1438 | <li>An address of a global variable is associated with the address |
| Dan Gohman | 556ca27 | 2009-07-27 18:07:55 +0000 | [diff] [blame] | 1439 | range of the variable's storage.</li> |
| 1440 | <li>The result value of an allocation instruction is associated with |
| 1441 | the address range of the allocated storage.</li> |
| 1442 | <li>A null pointer in the default address-space is associated with |
| Andreas Bolka | 55e459a | 2009-07-29 00:02:05 +0000 | [diff] [blame] | 1443 | no address.</li> |
| Dan Gohman | 556ca27 | 2009-07-27 18:07:55 +0000 | [diff] [blame] | 1444 | <li>An integer constant other than zero or a pointer value returned |
| 1445 | from a function not defined within LLVM may be associated with address |
| 1446 | ranges allocated through mechanisms other than those provided by |
| Andreas Bolka | 55e459a | 2009-07-29 00:02:05 +0000 | [diff] [blame] | 1447 | LLVM. Such ranges shall not overlap with any ranges of addresses |
| Dan Gohman | 556ca27 | 2009-07-27 18:07:55 +0000 | [diff] [blame] | 1448 | allocated by mechanisms provided by LLVM.</li> |
| Dan Gohman | 1e10962 | 2010-07-02 18:41:32 +0000 | [diff] [blame] | 1449 | </ul> |
| 1450 | |
| 1451 | <p>A pointer value is <i>based</i> on another pointer value according |
| 1452 | to the following rules:</p> |
| 1453 | |
| 1454 | <ul> |
| 1455 | <li>A pointer value formed from a |
| 1456 | <tt><a href="#i_getelementptr">getelementptr</a></tt> operation |
| 1457 | is <i>based</i> on the first operand of the <tt>getelementptr</tt>.</li> |
| 1458 | <li>The result value of a |
| 1459 | <tt><a href="#i_bitcast">bitcast</a></tt> is <i>based</i> on the operand |
| 1460 | of the <tt>bitcast</tt>.</li> |
| 1461 | <li>A pointer value formed by an |
| 1462 | <tt><a href="#i_inttoptr">inttoptr</a></tt> is <i>based</i> on all |
| 1463 | pointer values that contribute (directly or indirectly) to the |
| 1464 | computation of the pointer's value.</li> |
| 1465 | <li>The "<i>based</i> on" relationship is transitive.</li> |
| 1466 | </ul> |
| 1467 | |
| 1468 | <p>Note that this definition of <i>"based"</i> is intentionally |
| 1469 | similar to the definition of <i>"based"</i> in C99, though it is |
| 1470 | slightly weaker.</p> |
| Dan Gohman | 556ca27 | 2009-07-27 18:07:55 +0000 | [diff] [blame] | 1471 | |
| 1472 | <p>LLVM IR does not associate types with memory. The result type of a |
| Andreas Bolka | 55e459a | 2009-07-29 00:02:05 +0000 | [diff] [blame] | 1473 | <tt><a href="#i_load">load</a></tt> merely indicates the size and |
| 1474 | alignment of the memory from which to load, as well as the |
| Dan Gohman | c22c0f3 | 2010-06-17 19:23:50 +0000 | [diff] [blame] | 1475 | interpretation of the value. The first operand type of a |
| Andreas Bolka | 55e459a | 2009-07-29 00:02:05 +0000 | [diff] [blame] | 1476 | <tt><a href="#i_store">store</a></tt> similarly only indicates the size |
| 1477 | and alignment of the store.</p> |
| Dan Gohman | 556ca27 | 2009-07-27 18:07:55 +0000 | [diff] [blame] | 1478 | |
| 1479 | <p>Consequently, type-based alias analysis, aka TBAA, aka |
| 1480 | <tt>-fstrict-aliasing</tt>, is not applicable to general unadorned |
| 1481 | LLVM IR. <a href="#metadata">Metadata</a> may be used to encode |
| 1482 | additional information which specialized optimization passes may use |
| 1483 | to implement type-based alias analysis.</p> |
| 1484 | |
| 1485 | </div> |
| 1486 | |
| Jeffrey Yasskin | 93e066d | 2010-04-26 21:21:24 +0000 | [diff] [blame] | 1487 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 1488 | <h3> |
| Jeffrey Yasskin | 93e066d | 2010-04-26 21:21:24 +0000 | [diff] [blame] | 1489 | <a name="volatile">Volatile Memory Accesses</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 1490 | </h3> |
| Jeffrey Yasskin | 93e066d | 2010-04-26 21:21:24 +0000 | [diff] [blame] | 1491 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 1492 | <div> |
| Jeffrey Yasskin | 93e066d | 2010-04-26 21:21:24 +0000 | [diff] [blame] | 1493 | |
| 1494 | <p>Certain memory accesses, such as <a href="#i_load"><tt>load</tt></a>s, <a |
| 1495 | href="#i_store"><tt>store</tt></a>s, and <a |
| 1496 | href="#int_memcpy"><tt>llvm.memcpy</tt></a>s may be marked <tt>volatile</tt>. |
| 1497 | The optimizers must not change the number of volatile operations or change their |
| 1498 | order of execution relative to other volatile operations. The optimizers |
| 1499 | <i>may</i> change the order of volatile operations relative to non-volatile |
| 1500 | operations. This is not Java's "volatile" and has no cross-thread |
| 1501 | synchronization behavior.</p> |
| 1502 | |
| 1503 | </div> |
| 1504 | |
| Eli Friedman | 5b60e1b | 2011-07-20 21:35:53 +0000 | [diff] [blame] | 1505 | <!-- ======================================================================= --> |
| 1506 | <h3> |
| 1507 | <a name="memmodel">Memory Model for Concurrent Operations</a> |
| 1508 | </h3> |
| 1509 | |
| 1510 | <div> |
| 1511 | |
| 1512 | <p>The LLVM IR does not define any way to start parallel threads of execution |
| 1513 | or to register signal handlers. Nonetheless, there are platform-specific |
| 1514 | ways to create them, and we define LLVM IR's behavior in their presence. This |
| 1515 | model is inspired by the C++0x memory model.</p> |
| 1516 | |
| Eli Friedman | 234bccd | 2011-08-22 21:35:27 +0000 | [diff] [blame] | 1517 | <p>For a more informal introduction to this model, see the |
| 1518 | <a href="Atomics.html">LLVM Atomic Instructions and Concurrency Guide</a>. |
| 1519 | |
| Eli Friedman | 5b60e1b | 2011-07-20 21:35:53 +0000 | [diff] [blame] | 1520 | <p>We define a <i>happens-before</i> partial order as the least partial order |
| 1521 | that</p> |
| 1522 | <ul> |
| 1523 | <li>Is a superset of single-thread program order, and</li> |
| 1524 | <li>When a <i>synchronizes-with</i> <tt>b</tt>, includes an edge from |
| 1525 | <tt>a</tt> to <tt>b</tt>. <i>Synchronizes-with</i> pairs are introduced |
| 1526 | by platform-specific techniques, like pthread locks, thread |
| Eli Friedman | ff03048 | 2011-07-28 21:48:00 +0000 | [diff] [blame] | 1527 | creation, thread joining, etc., and by atomic instructions. |
| 1528 | (See also <a href="#ordering">Atomic Memory Ordering Constraints</a>). |
| 1529 | </li> |
| Eli Friedman | 5b60e1b | 2011-07-20 21:35:53 +0000 | [diff] [blame] | 1530 | </ul> |
| 1531 | |
| 1532 | <p>Note that program order does not introduce <i>happens-before</i> edges |
| 1533 | between a thread and signals executing inside that thread.</p> |
| 1534 | |
| 1535 | <p>Every (defined) read operation (load instructions, memcpy, atomic |
| 1536 | loads/read-modify-writes, etc.) <var>R</var> reads a series of bytes written by |
| 1537 | (defined) write operations (store instructions, atomic |
| Eli Friedman | 118973a | 2011-07-22 03:04:45 +0000 | [diff] [blame] | 1538 | stores/read-modify-writes, memcpy, etc.). For the purposes of this section, |
| 1539 | initialized globals are considered to have a write of the initializer which is |
| 1540 | atomic and happens before any other read or write of the memory in question. |
| 1541 | For each byte of a read <var>R</var>, <var>R<sub>byte</sub></var> may see |
| 1542 | any write to the same byte, except:</p> |
| Eli Friedman | 5b60e1b | 2011-07-20 21:35:53 +0000 | [diff] [blame] | 1543 | |
| 1544 | <ul> |
| 1545 | <li>If <var>write<sub>1</sub></var> happens before |
| 1546 | <var>write<sub>2</sub></var>, and <var>write<sub>2</sub></var> happens |
| 1547 | before <var>R<sub>byte</sub></var>, then <var>R<sub>byte</sub></var> |
| Eli Friedman | 118973a | 2011-07-22 03:04:45 +0000 | [diff] [blame] | 1548 | does not see <var>write<sub>1</sub></var>. |
| Bill Wendling | 0246bb7 | 2011-07-31 06:45:03 +0000 | [diff] [blame] | 1549 | <li>If <var>R<sub>byte</sub></var> happens before |
| 1550 | <var>write<sub>3</sub></var>, then <var>R<sub>byte</sub></var> does not |
| 1551 | see <var>write<sub>3</sub></var>. |
| Eli Friedman | 5b60e1b | 2011-07-20 21:35:53 +0000 | [diff] [blame] | 1552 | </ul> |
| 1553 | |
| 1554 | <p>Given that definition, <var>R<sub>byte</sub></var> is defined as follows: |
| 1555 | <ul> |
| Eli Friedman | 234bccd | 2011-08-22 21:35:27 +0000 | [diff] [blame] | 1556 | <li>If <var>R</var> is volatile, the result is target-dependent. (Volatile |
| 1557 | is supposed to give guarantees which can support |
| 1558 | <code>sig_atomic_t</code> in C/C++, and may be used for accesses to |
| 1559 | addresses which do not behave like normal memory. It does not generally |
| 1560 | provide cross-thread synchronization.) |
| 1561 | <li>Otherwise, if there is no write to the same byte that happens before |
| Eli Friedman | 5b60e1b | 2011-07-20 21:35:53 +0000 | [diff] [blame] | 1562 | <var>R<sub>byte</sub></var>, <var>R<sub>byte</sub></var> returns |
| 1563 | <tt>undef</tt> for that byte. |
| Eli Friedman | 118973a | 2011-07-22 03:04:45 +0000 | [diff] [blame] | 1564 | <li>Otherwise, if <var>R<sub>byte</sub></var> may see exactly one write, |
| Eli Friedman | 5b60e1b | 2011-07-20 21:35:53 +0000 | [diff] [blame] | 1565 | <var>R<sub>byte</sub></var> returns the value written by that |
| 1566 | write.</li> |
| Eli Friedman | 118973a | 2011-07-22 03:04:45 +0000 | [diff] [blame] | 1567 | <li>Otherwise, if <var>R</var> is atomic, and all the writes |
| 1568 | <var>R<sub>byte</sub></var> may see are atomic, it chooses one of the |
| Eli Friedman | ff03048 | 2011-07-28 21:48:00 +0000 | [diff] [blame] | 1569 | values written. See the <a href="#ordering">Atomic Memory Ordering |
| 1570 | Constraints</a> section for additional constraints on how the choice |
| 1571 | is made. |
| Eli Friedman | 5b60e1b | 2011-07-20 21:35:53 +0000 | [diff] [blame] | 1572 | <li>Otherwise <var>R<sub>byte</sub></var> returns <tt>undef</tt>.</li> |
| 1573 | </ul> |
| 1574 | |
| 1575 | <p><var>R</var> returns the value composed of the series of bytes it read. |
| 1576 | This implies that some bytes within the value may be <tt>undef</tt> |
| 1577 | <b>without</b> the entire value being <tt>undef</tt>. Note that this only |
| 1578 | defines the semantics of the operation; it doesn't mean that targets will |
| 1579 | emit more than one instruction to read the series of bytes.</p> |
| 1580 | |
| 1581 | <p>Note that in cases where none of the atomic intrinsics are used, this model |
| 1582 | places only one restriction on IR transformations on top of what is required |
| 1583 | for single-threaded execution: introducing a store to a byte which might not |
| Eli Friedman | 101c81d | 2011-08-02 01:15:34 +0000 | [diff] [blame] | 1584 | otherwise be stored is not allowed in general. (Specifically, in the case |
| 1585 | where another thread might write to and read from an address, introducing a |
| 1586 | store can change a load that may see exactly one write into a load that may |
| 1587 | see multiple writes.)</p> |
| Eli Friedman | 5b60e1b | 2011-07-20 21:35:53 +0000 | [diff] [blame] | 1588 | |
| 1589 | <!-- FIXME: This model assumes all targets where concurrency is relevant have |
| 1590 | a byte-size store which doesn't affect adjacent bytes. As far as I can tell, |
| 1591 | none of the backends currently in the tree fall into this category; however, |
| 1592 | there might be targets which care. If there are, we want a paragraph |
| 1593 | like the following: |
| 1594 | |
| 1595 | Targets may specify that stores narrower than a certain width are not |
| 1596 | available; on such a target, for the purposes of this model, treat any |
| 1597 | non-atomic write with an alignment or width less than the minimum width |
| 1598 | as if it writes to the relevant surrounding bytes. |
| 1599 | --> |
| 1600 | |
| 1601 | </div> |
| 1602 | |
| Eli Friedman | ff03048 | 2011-07-28 21:48:00 +0000 | [diff] [blame] | 1603 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | b5bb29c | 2011-08-12 06:17:17 +0000 | [diff] [blame] | 1604 | <h3> |
| Eli Friedman | ff03048 | 2011-07-28 21:48:00 +0000 | [diff] [blame] | 1605 | <a name="ordering">Atomic Memory Ordering Constraints</a> |
| NAKAMURA Takumi | b5bb29c | 2011-08-12 06:17:17 +0000 | [diff] [blame] | 1606 | </h3> |
| Eli Friedman | ff03048 | 2011-07-28 21:48:00 +0000 | [diff] [blame] | 1607 | |
| NAKAMURA Takumi | b5bb29c | 2011-08-12 06:17:17 +0000 | [diff] [blame] | 1608 | <div> |
| Eli Friedman | ff03048 | 2011-07-28 21:48:00 +0000 | [diff] [blame] | 1609 | |
| 1610 | <p>Atomic instructions (<a href="#i_cmpxchg"><code>cmpxchg</code></a>, |
| Eli Friedman | 21006d4 | 2011-08-09 23:02:53 +0000 | [diff] [blame] | 1611 | <a href="#i_atomicrmw"><code>atomicrmw</code></a>, |
| 1612 | <a href="#i_fence"><code>fence</code></a>, |
| 1613 | <a href="#i_load"><code>atomic load</code></a>, and |
| Eli Friedman | 8fa281a | 2011-08-09 23:26:12 +0000 | [diff] [blame] | 1614 | <a href="#i_store"><code>atomic store</code></a>) take an ordering parameter |
| Eli Friedman | ff03048 | 2011-07-28 21:48:00 +0000 | [diff] [blame] | 1615 | that determines which other atomic instructions on the same address they |
| 1616 | <i>synchronize with</i>. These semantics are borrowed from Java and C++0x, |
| 1617 | but are somewhat more colloquial. If these descriptions aren't precise enough, |
| Eli Friedman | 234bccd | 2011-08-22 21:35:27 +0000 | [diff] [blame] | 1618 | check those specs (see spec references in the |
| Nick Lewycky | 300a263 | 2012-01-23 08:47:21 +0000 | [diff] [blame] | 1619 | <a href="Atomics.html#introduction">atomics guide</a>). |
| Eli Friedman | 234bccd | 2011-08-22 21:35:27 +0000 | [diff] [blame] | 1620 | <a href="#i_fence"><code>fence</code></a> instructions |
| Eli Friedman | ff03048 | 2011-07-28 21:48:00 +0000 | [diff] [blame] | 1621 | treat these orderings somewhat differently since they don't take an address. |
| 1622 | See that instruction's documentation for details.</p> |
| 1623 | |
| Eli Friedman | 234bccd | 2011-08-22 21:35:27 +0000 | [diff] [blame] | 1624 | <p>For a simpler introduction to the ordering constraints, see the |
| 1625 | <a href="Atomics.html">LLVM Atomic Instructions and Concurrency Guide</a>.</p> |
| 1626 | |
| Eli Friedman | ff03048 | 2011-07-28 21:48:00 +0000 | [diff] [blame] | 1627 | <dl> |
| Eli Friedman | ff03048 | 2011-07-28 21:48:00 +0000 | [diff] [blame] | 1628 | <dt><code>unordered</code></dt> |
| 1629 | <dd>The set of values that can be read is governed by the happens-before |
| 1630 | partial order. A value cannot be read unless some operation wrote it. |
| 1631 | This is intended to provide a guarantee strong enough to model Java's |
| 1632 | non-volatile shared variables. This ordering cannot be specified for |
| 1633 | read-modify-write operations; it is not strong enough to make them atomic |
| 1634 | in any interesting way.</dd> |
| 1635 | <dt><code>monotonic</code></dt> |
| 1636 | <dd>In addition to the guarantees of <code>unordered</code>, there is a single |
| 1637 | total order for modifications by <code>monotonic</code> operations on each |
| 1638 | address. All modification orders must be compatible with the happens-before |
| 1639 | order. There is no guarantee that the modification orders can be combined to |
| 1640 | a global total order for the whole program (and this often will not be |
| 1641 | possible). The read in an atomic read-modify-write operation |
| 1642 | (<a href="#i_cmpxchg"><code>cmpxchg</code></a> and |
| 1643 | <a href="#i_atomicrmw"><code>atomicrmw</code></a>) |
| 1644 | reads the value in the modification order immediately before the value it |
| 1645 | writes. If one atomic read happens before another atomic read of the same |
| 1646 | address, the later read must see the same value or a later value in the |
| 1647 | address's modification order. This disallows reordering of |
| 1648 | <code>monotonic</code> (or stronger) operations on the same address. If an |
| 1649 | address is written <code>monotonic</code>ally by one thread, and other threads |
| 1650 | <code>monotonic</code>ally read that address repeatedly, the other threads must |
| Eli Friedman | 234bccd | 2011-08-22 21:35:27 +0000 | [diff] [blame] | 1651 | eventually see the write. This corresponds to the C++0x/C1x |
| 1652 | <code>memory_order_relaxed</code>.</dd> |
| Eli Friedman | ff03048 | 2011-07-28 21:48:00 +0000 | [diff] [blame] | 1653 | <dt><code>acquire</code></dt> |
| Eli Friedman | ff03048 | 2011-07-28 21:48:00 +0000 | [diff] [blame] | 1654 | <dd>In addition to the guarantees of <code>monotonic</code>, |
| Eli Friedman | c264b2f | 2011-08-24 20:28:39 +0000 | [diff] [blame] | 1655 | a <i>synchronizes-with</i> edge may be formed with a <code>release</code> |
| 1656 | operation. This is intended to model C++'s <code>memory_order_acquire</code>.</dd> |
| 1657 | <dt><code>release</code></dt> |
| 1658 | <dd>In addition to the guarantees of <code>monotonic</code>, if this operation |
| 1659 | writes a value which is subsequently read by an <code>acquire</code> operation, |
| 1660 | it <i>synchronizes-with</i> that operation. (This isn't a complete |
| 1661 | description; see the C++0x definition of a release sequence.) This corresponds |
| 1662 | to the C++0x/C1x <code>memory_order_release</code>.</dd> |
| Eli Friedman | ff03048 | 2011-07-28 21:48:00 +0000 | [diff] [blame] | 1663 | <dt><code>acq_rel</code> (acquire+release)</dt><dd>Acts as both an |
| Eli Friedman | 234bccd | 2011-08-22 21:35:27 +0000 | [diff] [blame] | 1664 | <code>acquire</code> and <code>release</code> operation on its address. |
| 1665 | This corresponds to the C++0x/C1x <code>memory_order_acq_rel</code>.</dd> |
| Eli Friedman | ff03048 | 2011-07-28 21:48:00 +0000 | [diff] [blame] | 1666 | <dt><code>seq_cst</code> (sequentially consistent)</dt><dd> |
| 1667 | <dd>In addition to the guarantees of <code>acq_rel</code> |
| 1668 | (<code>acquire</code> for an operation which only reads, <code>release</code> |
| 1669 | for an operation which only writes), there is a global total order on all |
| 1670 | sequentially-consistent operations on all addresses, which is consistent with |
| 1671 | the <i>happens-before</i> partial order and with the modification orders of |
| 1672 | all the affected addresses. Each sequentially-consistent read sees the last |
| Eli Friedman | 234bccd | 2011-08-22 21:35:27 +0000 | [diff] [blame] | 1673 | preceding write to the same address in this global order. This corresponds |
| 1674 | to the C++0x/C1x <code>memory_order_seq_cst</code> and Java volatile.</dd> |
| Eli Friedman | ff03048 | 2011-07-28 21:48:00 +0000 | [diff] [blame] | 1675 | </dl> |
| 1676 | |
| 1677 | <p id="singlethread">If an atomic operation is marked <code>singlethread</code>, |
| 1678 | it only <i>synchronizes with</i> or participates in modification and seq_cst |
| 1679 | total orderings with other operations running in the same thread (for example, |
| 1680 | in signal handlers).</p> |
| 1681 | |
| 1682 | </div> |
| 1683 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 1684 | </div> |
| 1685 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 1686 | <!-- *********************************************************************** --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 1687 | <h2><a name="typesystem">Type System</a></h2> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 1688 | <!-- *********************************************************************** --> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 1689 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 1690 | <div> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 1691 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 1692 | <p>The LLVM type system is one of the most important features of the |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1693 | intermediate representation. Being typed enables a number of optimizations |
| 1694 | to be performed on the intermediate representation directly, without having |
| 1695 | to do extra analyses on the side before the transformation. A strong type |
| 1696 | system makes it easier to read the generated code and enables novel analyses |
| 1697 | and transformations that are not feasible to perform on normal three address |
| 1698 | code representations.</p> |
| Chris Lattner | fa73021 | 2004-12-09 16:11:40 +0000 | [diff] [blame] | 1699 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 1700 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 1701 | <h3> |
| 1702 | <a name="t_classifications">Type Classifications</a> |
| 1703 | </h3> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1704 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 1705 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1706 | |
| 1707 | <p>The types fall into a few useful classifications:</p> |
| Misha Brukman | daa4cb0 | 2004-03-01 17:47:27 +0000 | [diff] [blame] | 1708 | |
| 1709 | <table border="1" cellspacing="0" cellpadding="4"> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 1710 | <tbody> |
| Reid Spencer | d3f876c | 2004-11-01 08:19:36 +0000 | [diff] [blame] | 1711 | <tr><th>Classification</th><th>Types</th></tr> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 1712 | <tr> |
| Chris Lattner | 4f69f46 | 2008-01-04 04:32:38 +0000 | [diff] [blame] | 1713 | <td><a href="#t_integer">integer</a></td> |
| Reid Spencer | 2b91631 | 2007-05-16 18:44:01 +0000 | [diff] [blame] | 1714 | <td><tt>i1, i2, i3, ... i8, ... i16, ... i32, ... i64, ... </tt></td> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 1715 | </tr> |
| 1716 | <tr> |
| Chris Lattner | 4f69f46 | 2008-01-04 04:32:38 +0000 | [diff] [blame] | 1717 | <td><a href="#t_floating">floating point</a></td> |
| Dan Gohman | ce16339 | 2011-12-17 00:04:22 +0000 | [diff] [blame] | 1718 | <td><tt>half, float, double, x86_fp80, fp128, ppc_fp128</tt></td> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 1719 | </tr> |
| 1720 | <tr> |
| 1721 | <td><a name="t_firstclass">first class</a></td> |
| Chris Lattner | 4f69f46 | 2008-01-04 04:32:38 +0000 | [diff] [blame] | 1722 | <td><a href="#t_integer">integer</a>, |
| 1723 | <a href="#t_floating">floating point</a>, |
| 1724 | <a href="#t_pointer">pointer</a>, |
| Dan Gohman | 0066db6 | 2008-06-18 18:42:13 +0000 | [diff] [blame] | 1725 | <a href="#t_vector">vector</a>, |
| Dan Gohman | a334d5f | 2008-05-12 23:51:09 +0000 | [diff] [blame] | 1726 | <a href="#t_struct">structure</a>, |
| 1727 | <a href="#t_array">array</a>, |
| Nick Lewycky | 7a0370f | 2009-05-30 05:06:04 +0000 | [diff] [blame] | 1728 | <a href="#t_label">label</a>, |
| 1729 | <a href="#t_metadata">metadata</a>. |
| Reid Spencer | ca86e16 | 2006-12-31 07:07:53 +0000 | [diff] [blame] | 1730 | </td> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 1731 | </tr> |
| Chris Lattner | 4f69f46 | 2008-01-04 04:32:38 +0000 | [diff] [blame] | 1732 | <tr> |
| 1733 | <td><a href="#t_primitive">primitive</a></td> |
| 1734 | <td><a href="#t_label">label</a>, |
| 1735 | <a href="#t_void">void</a>, |
| Tobias Grosser | 0538729 | 2010-12-28 20:29:31 +0000 | [diff] [blame] | 1736 | <a href="#t_integer">integer</a>, |
| Nick Lewycky | 7a0370f | 2009-05-30 05:06:04 +0000 | [diff] [blame] | 1737 | <a href="#t_floating">floating point</a>, |
| Dale Johannesen | 21fe99b | 2010-10-01 00:48:59 +0000 | [diff] [blame] | 1738 | <a href="#t_x86mmx">x86mmx</a>, |
| Nick Lewycky | 7a0370f | 2009-05-30 05:06:04 +0000 | [diff] [blame] | 1739 | <a href="#t_metadata">metadata</a>.</td> |
| Chris Lattner | 4f69f46 | 2008-01-04 04:32:38 +0000 | [diff] [blame] | 1740 | </tr> |
| 1741 | <tr> |
| 1742 | <td><a href="#t_derived">derived</a></td> |
| Chris Lattner | fdfeb69 | 2010-02-12 20:49:41 +0000 | [diff] [blame] | 1743 | <td><a href="#t_array">array</a>, |
| Chris Lattner | 4f69f46 | 2008-01-04 04:32:38 +0000 | [diff] [blame] | 1744 | <a href="#t_function">function</a>, |
| 1745 | <a href="#t_pointer">pointer</a>, |
| 1746 | <a href="#t_struct">structure</a>, |
| Chris Lattner | 4f69f46 | 2008-01-04 04:32:38 +0000 | [diff] [blame] | 1747 | <a href="#t_vector">vector</a>, |
| 1748 | <a href="#t_opaque">opaque</a>. |
| Dan Gohman | 01ac101 | 2008-10-14 16:32:04 +0000 | [diff] [blame] | 1749 | </td> |
| Chris Lattner | 4f69f46 | 2008-01-04 04:32:38 +0000 | [diff] [blame] | 1750 | </tr> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 1751 | </tbody> |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 1752 | </table> |
| Misha Brukman | daa4cb0 | 2004-03-01 17:47:27 +0000 | [diff] [blame] | 1753 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1754 | <p>The <a href="#t_firstclass">first class</a> types are perhaps the most |
| 1755 | important. Values of these types are the only ones which can be produced by |
| Nick Lewycky | ec38da4 | 2009-09-27 00:45:11 +0000 | [diff] [blame] | 1756 | instructions.</p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1757 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 1758 | </div> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 1759 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 1760 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 1761 | <h3> |
| 1762 | <a name="t_primitive">Primitive Types</a> |
| 1763 | </h3> |
| Chris Lattner | 8f8c7b7 | 2008-01-04 04:34:14 +0000 | [diff] [blame] | 1764 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 1765 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1766 | |
| Chris Lattner | 4f69f46 | 2008-01-04 04:32:38 +0000 | [diff] [blame] | 1767 | <p>The primitive types are the fundamental building blocks of the LLVM |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1768 | system.</p> |
| Chris Lattner | 4f69f46 | 2008-01-04 04:32:38 +0000 | [diff] [blame] | 1769 | |
| 1770 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 1771 | <h4> |
| 1772 | <a name="t_integer">Integer Type</a> |
| 1773 | </h4> |
| Nick Lewycky | ec38da4 | 2009-09-27 00:45:11 +0000 | [diff] [blame] | 1774 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 1775 | <div> |
| Nick Lewycky | ec38da4 | 2009-09-27 00:45:11 +0000 | [diff] [blame] | 1776 | |
| 1777 | <h5>Overview:</h5> |
| 1778 | <p>The integer type is a very simple type that simply specifies an arbitrary |
| 1779 | bit width for the integer type desired. Any bit width from 1 bit to |
| 1780 | 2<sup>23</sup>-1 (about 8 million) can be specified.</p> |
| 1781 | |
| 1782 | <h5>Syntax:</h5> |
| 1783 | <pre> |
| 1784 | iN |
| 1785 | </pre> |
| 1786 | |
| 1787 | <p>The number of bits the integer will occupy is specified by the <tt>N</tt> |
| 1788 | value.</p> |
| 1789 | |
| 1790 | <h5>Examples:</h5> |
| 1791 | <table class="layout"> |
| 1792 | <tr class="layout"> |
| 1793 | <td class="left"><tt>i1</tt></td> |
| 1794 | <td class="left">a single-bit integer.</td> |
| 1795 | </tr> |
| 1796 | <tr class="layout"> |
| 1797 | <td class="left"><tt>i32</tt></td> |
| 1798 | <td class="left">a 32-bit integer.</td> |
| 1799 | </tr> |
| 1800 | <tr class="layout"> |
| 1801 | <td class="left"><tt>i1942652</tt></td> |
| 1802 | <td class="left">a really big integer of over 1 million bits.</td> |
| 1803 | </tr> |
| 1804 | </table> |
| 1805 | |
| Nick Lewycky | ec38da4 | 2009-09-27 00:45:11 +0000 | [diff] [blame] | 1806 | </div> |
| 1807 | |
| 1808 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 1809 | <h4> |
| 1810 | <a name="t_floating">Floating Point Types</a> |
| 1811 | </h4> |
| Chris Lattner | 4f69f46 | 2008-01-04 04:32:38 +0000 | [diff] [blame] | 1812 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 1813 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1814 | |
| 1815 | <table> |
| 1816 | <tbody> |
| 1817 | <tr><th>Type</th><th>Description</th></tr> |
| Dan Gohman | ce16339 | 2011-12-17 00:04:22 +0000 | [diff] [blame] | 1818 | <tr><td><tt>half</tt></td><td>16-bit floating point value</td></tr> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1819 | <tr><td><tt>float</tt></td><td>32-bit floating point value</td></tr> |
| 1820 | <tr><td><tt>double</tt></td><td>64-bit floating point value</td></tr> |
| 1821 | <tr><td><tt>fp128</tt></td><td>128-bit floating point value (112-bit mantissa)</td></tr> |
| 1822 | <tr><td><tt>x86_fp80</tt></td><td>80-bit floating point value (X87)</td></tr> |
| 1823 | <tr><td><tt>ppc_fp128</tt></td><td>128-bit floating point value (two 64-bits)</td></tr> |
| 1824 | </tbody> |
| 1825 | </table> |
| 1826 | |
| Chris Lattner | 4f69f46 | 2008-01-04 04:32:38 +0000 | [diff] [blame] | 1827 | </div> |
| 1828 | |
| 1829 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 1830 | <h4> |
| 1831 | <a name="t_x86mmx">X86mmx Type</a> |
| 1832 | </h4> |
| Dale Johannesen | 21fe99b | 2010-10-01 00:48:59 +0000 | [diff] [blame] | 1833 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 1834 | <div> |
| Dale Johannesen | 21fe99b | 2010-10-01 00:48:59 +0000 | [diff] [blame] | 1835 | |
| 1836 | <h5>Overview:</h5> |
| 1837 | <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> |
| 1838 | |
| 1839 | <h5>Syntax:</h5> |
| 1840 | <pre> |
| Dale Johannesen | 473a8c8 | 2010-10-01 01:07:02 +0000 | [diff] [blame] | 1841 | x86mmx |
| Dale Johannesen | 21fe99b | 2010-10-01 00:48:59 +0000 | [diff] [blame] | 1842 | </pre> |
| 1843 | |
| 1844 | </div> |
| 1845 | |
| 1846 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 1847 | <h4> |
| 1848 | <a name="t_void">Void Type</a> |
| 1849 | </h4> |
| Chris Lattner | 4f69f46 | 2008-01-04 04:32:38 +0000 | [diff] [blame] | 1850 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 1851 | <div> |
| Bill Wendling | c39e3e0 | 2009-07-20 02:39:26 +0000 | [diff] [blame] | 1852 | |
| Chris Lattner | 4f69f46 | 2008-01-04 04:32:38 +0000 | [diff] [blame] | 1853 | <h5>Overview:</h5> |
| 1854 | <p>The void type does not represent any value and has no size.</p> |
| 1855 | |
| 1856 | <h5>Syntax:</h5> |
| Chris Lattner | 4f69f46 | 2008-01-04 04:32:38 +0000 | [diff] [blame] | 1857 | <pre> |
| 1858 | void |
| 1859 | </pre> |
| Bill Wendling | c39e3e0 | 2009-07-20 02:39:26 +0000 | [diff] [blame] | 1860 | |
| Chris Lattner | 4f69f46 | 2008-01-04 04:32:38 +0000 | [diff] [blame] | 1861 | </div> |
| 1862 | |
| 1863 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 1864 | <h4> |
| 1865 | <a name="t_label">Label Type</a> |
| 1866 | </h4> |
| Chris Lattner | 4f69f46 | 2008-01-04 04:32:38 +0000 | [diff] [blame] | 1867 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 1868 | <div> |
| Bill Wendling | c39e3e0 | 2009-07-20 02:39:26 +0000 | [diff] [blame] | 1869 | |
| Chris Lattner | 4f69f46 | 2008-01-04 04:32:38 +0000 | [diff] [blame] | 1870 | <h5>Overview:</h5> |
| 1871 | <p>The label type represents code labels.</p> |
| 1872 | |
| 1873 | <h5>Syntax:</h5> |
| Chris Lattner | 4f69f46 | 2008-01-04 04:32:38 +0000 | [diff] [blame] | 1874 | <pre> |
| 1875 | label |
| 1876 | </pre> |
| Bill Wendling | c39e3e0 | 2009-07-20 02:39:26 +0000 | [diff] [blame] | 1877 | |
| Chris Lattner | 4f69f46 | 2008-01-04 04:32:38 +0000 | [diff] [blame] | 1878 | </div> |
| 1879 | |
| Nick Lewycky | 7a0370f | 2009-05-30 05:06:04 +0000 | [diff] [blame] | 1880 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 1881 | <h4> |
| 1882 | <a name="t_metadata">Metadata Type</a> |
| 1883 | </h4> |
| Nick Lewycky | 7a0370f | 2009-05-30 05:06:04 +0000 | [diff] [blame] | 1884 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 1885 | <div> |
| Bill Wendling | c39e3e0 | 2009-07-20 02:39:26 +0000 | [diff] [blame] | 1886 | |
| Nick Lewycky | 7a0370f | 2009-05-30 05:06:04 +0000 | [diff] [blame] | 1887 | <h5>Overview:</h5> |
| Nick Lewycky | c261df9 | 2009-09-27 23:27:42 +0000 | [diff] [blame] | 1888 | <p>The metadata type represents embedded metadata. No derived types may be |
| 1889 | created from metadata except for <a href="#t_function">function</a> |
| 1890 | arguments. |
| Nick Lewycky | 7a0370f | 2009-05-30 05:06:04 +0000 | [diff] [blame] | 1891 | |
| 1892 | <h5>Syntax:</h5> |
| Nick Lewycky | 7a0370f | 2009-05-30 05:06:04 +0000 | [diff] [blame] | 1893 | <pre> |
| 1894 | metadata |
| 1895 | </pre> |
| Bill Wendling | c39e3e0 | 2009-07-20 02:39:26 +0000 | [diff] [blame] | 1896 | |
| Nick Lewycky | 7a0370f | 2009-05-30 05:06:04 +0000 | [diff] [blame] | 1897 | </div> |
| 1898 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 1899 | </div> |
| Chris Lattner | 4f69f46 | 2008-01-04 04:32:38 +0000 | [diff] [blame] | 1900 | |
| 1901 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 1902 | <h3> |
| 1903 | <a name="t_derived">Derived Types</a> |
| 1904 | </h3> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 1905 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 1906 | <div> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 1907 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1908 | <p>The real power in LLVM comes from the derived types in the system. This is |
| 1909 | what allows a programmer to represent arrays, functions, pointers, and other |
| Nick Lewycky | ec38da4 | 2009-09-27 00:45:11 +0000 | [diff] [blame] | 1910 | useful types. Each of these types contain one or more element types which |
| 1911 | may be a primitive type, or another derived type. For example, it is |
| 1912 | possible to have a two dimensional array, using an array as the element type |
| 1913 | of another array.</p> |
| Dan Gohman | d8791e5 | 2009-01-24 15:58:40 +0000 | [diff] [blame] | 1914 | |
| Chris Lattner | fdfeb69 | 2010-02-12 20:49:41 +0000 | [diff] [blame] | 1915 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 1916 | <h4> |
| 1917 | <a name="t_aggregate">Aggregate Types</a> |
| 1918 | </h4> |
| Chris Lattner | fdfeb69 | 2010-02-12 20:49:41 +0000 | [diff] [blame] | 1919 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 1920 | <div> |
| Chris Lattner | fdfeb69 | 2010-02-12 20:49:41 +0000 | [diff] [blame] | 1921 | |
| 1922 | <p>Aggregate Types are a subset of derived types that can contain multiple |
| Duncan Sands | 20536b5 | 2011-12-14 15:44:20 +0000 | [diff] [blame] | 1923 | member types. <a href="#t_array">Arrays</a> and |
| 1924 | <a href="#t_struct">structs</a> are aggregate types. |
| 1925 | <a href="#t_vector">Vectors</a> are not considered to be aggregate types.</p> |
| Chris Lattner | fdfeb69 | 2010-02-12 20:49:41 +0000 | [diff] [blame] | 1926 | |
| 1927 | </div> |
| 1928 | |
| Reid Spencer | 2b91631 | 2007-05-16 18:44:01 +0000 | [diff] [blame] | 1929 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 1930 | <h4> |
| 1931 | <a name="t_array">Array Type</a> |
| 1932 | </h4> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 1933 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 1934 | <div> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 1935 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 1936 | <h5>Overview:</h5> |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 1937 | <p>The array type is a very simple derived type that arranges elements |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1938 | sequentially in memory. The array type requires a size (number of elements) |
| 1939 | and an underlying data type.</p> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 1940 | |
| Chris Lattner | 7faa883 | 2002-04-14 06:13:44 +0000 | [diff] [blame] | 1941 | <h5>Syntax:</h5> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 1942 | <pre> |
| 1943 | [<# elements> x <elementtype>] |
| 1944 | </pre> |
| 1945 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1946 | <p>The number of elements is a constant integer value; <tt>elementtype</tt> may |
| 1947 | be any type with a size.</p> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 1948 | |
| Chris Lattner | 7faa883 | 2002-04-14 06:13:44 +0000 | [diff] [blame] | 1949 | <h5>Examples:</h5> |
| Reid Spencer | d3f876c | 2004-11-01 08:19:36 +0000 | [diff] [blame] | 1950 | <table class="layout"> |
| 1951 | <tr class="layout"> |
| Chris Lattner | 23ff1f9 | 2007-12-19 05:04:11 +0000 | [diff] [blame] | 1952 | <td class="left"><tt>[40 x i32]</tt></td> |
| 1953 | <td class="left">Array of 40 32-bit integer values.</td> |
| 1954 | </tr> |
| 1955 | <tr class="layout"> |
| 1956 | <td class="left"><tt>[41 x i32]</tt></td> |
| 1957 | <td class="left">Array of 41 32-bit integer values.</td> |
| 1958 | </tr> |
| 1959 | <tr class="layout"> |
| 1960 | <td class="left"><tt>[4 x i8]</tt></td> |
| 1961 | <td class="left">Array of 4 8-bit integer values.</td> |
| Reid Spencer | d3f876c | 2004-11-01 08:19:36 +0000 | [diff] [blame] | 1962 | </tr> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 1963 | </table> |
| Reid Spencer | d3f876c | 2004-11-01 08:19:36 +0000 | [diff] [blame] | 1964 | <p>Here are some examples of multidimensional arrays:</p> |
| 1965 | <table class="layout"> |
| 1966 | <tr class="layout"> |
| Chris Lattner | 23ff1f9 | 2007-12-19 05:04:11 +0000 | [diff] [blame] | 1967 | <td class="left"><tt>[3 x [4 x i32]]</tt></td> |
| 1968 | <td class="left">3x4 array of 32-bit integer values.</td> |
| 1969 | </tr> |
| 1970 | <tr class="layout"> |
| 1971 | <td class="left"><tt>[12 x [10 x float]]</tt></td> |
| 1972 | <td class="left">12x10 array of single precision floating point values.</td> |
| 1973 | </tr> |
| 1974 | <tr class="layout"> |
| 1975 | <td class="left"><tt>[2 x [3 x [4 x i16]]]</tt></td> |
| 1976 | <td class="left">2x3x4 array of 16-bit integer values.</td> |
| Reid Spencer | d3f876c | 2004-11-01 08:19:36 +0000 | [diff] [blame] | 1977 | </tr> |
| 1978 | </table> |
| Chris Lattner | e67a951 | 2005-06-24 17:22:57 +0000 | [diff] [blame] | 1979 | |
| Dan Gohman | 7657f6b | 2009-11-09 19:01:53 +0000 | [diff] [blame] | 1980 | <p>There is no restriction on indexing beyond the end of the array implied by |
| 1981 | a static type (though there are restrictions on indexing beyond the bounds |
| 1982 | of an allocated object in some cases). This means that single-dimension |
| 1983 | 'variable sized array' addressing can be implemented in LLVM with a zero |
| 1984 | length array type. An implementation of 'pascal style arrays' in LLVM could |
| 1985 | use the type "<tt>{ i32, [0 x float]}</tt>", for example.</p> |
| Chris Lattner | e67a951 | 2005-06-24 17:22:57 +0000 | [diff] [blame] | 1986 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 1987 | </div> |
| Reid Spencer | d3f876c | 2004-11-01 08:19:36 +0000 | [diff] [blame] | 1988 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 1989 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 1990 | <h4> |
| 1991 | <a name="t_function">Function Type</a> |
| 1992 | </h4> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1993 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 1994 | <div> |
| Chris Lattner | f4cde4e | 2008-04-23 04:59:35 +0000 | [diff] [blame] | 1995 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 1996 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 1997 | <p>The function type can be thought of as a function signature. It consists of |
| 1998 | a return type and a list of formal parameter types. The return type of a |
| Chris Lattner | 61c70e9 | 2010-08-28 04:09:24 +0000 | [diff] [blame] | 1999 | function type is a first class type or a void type.</p> |
| Devang Patel | c3fc6df | 2008-03-10 20:49:15 +0000 | [diff] [blame] | 2000 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 2001 | <h5>Syntax:</h5> |
| Chris Lattner | f4cde4e | 2008-04-23 04:59:35 +0000 | [diff] [blame] | 2002 | <pre> |
| Nick Lewycky | 5138694 | 2009-09-27 07:55:32 +0000 | [diff] [blame] | 2003 | <returntype> (<parameter list>) |
| Chris Lattner | f4cde4e | 2008-04-23 04:59:35 +0000 | [diff] [blame] | 2004 | </pre> |
| 2005 | |
| John Criswell | 0ec250c | 2005-10-24 16:17:18 +0000 | [diff] [blame] | 2006 | <p>...where '<tt><parameter list></tt>' is a comma-separated list of type |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2007 | specifiers. Optionally, the parameter list may include a type <tt>...</tt>, |
| 2008 | which indicates that the function takes a variable number of arguments. |
| 2009 | Variable argument functions can access their arguments with |
| 2010 | the <a href="#int_varargs">variable argument handling intrinsic</a> |
| Chris Lattner | 0724fbd | 2010-03-02 06:36:51 +0000 | [diff] [blame] | 2011 | functions. '<tt><returntype></tt>' is any type except |
| Nick Lewycky | c261df9 | 2009-09-27 23:27:42 +0000 | [diff] [blame] | 2012 | <a href="#t_label">label</a>.</p> |
| Chris Lattner | f4cde4e | 2008-04-23 04:59:35 +0000 | [diff] [blame] | 2013 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 2014 | <h5>Examples:</h5> |
| Reid Spencer | d3f876c | 2004-11-01 08:19:36 +0000 | [diff] [blame] | 2015 | <table class="layout"> |
| 2016 | <tr class="layout"> |
| Reid Spencer | 92f8230 | 2006-12-31 07:18:34 +0000 | [diff] [blame] | 2017 | <td class="left"><tt>i32 (i32)</tt></td> |
| 2018 | <td class="left">function taking an <tt>i32</tt>, returning an <tt>i32</tt> |
| Reid Spencer | d3f876c | 2004-11-01 08:19:36 +0000 | [diff] [blame] | 2019 | </td> |
| Reid Spencer | 92f8230 | 2006-12-31 07:18:34 +0000 | [diff] [blame] | 2020 | </tr><tr class="layout"> |
| Chris Lattner | 0724fbd | 2010-03-02 06:36:51 +0000 | [diff] [blame] | 2021 | <td class="left"><tt>float (i16, i32 *) * |
| Reid Spencer | f17a0b7 | 2006-12-31 07:20:23 +0000 | [diff] [blame] | 2022 | </tt></td> |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 2023 | <td class="left"><a href="#t_pointer">Pointer</a> to a function that takes |
| Chris Lattner | 0724fbd | 2010-03-02 06:36:51 +0000 | [diff] [blame] | 2024 | an <tt>i16</tt> and a <a href="#t_pointer">pointer</a> to <tt>i32</tt>, |
| 2025 | returning <tt>float</tt>. |
| Reid Spencer | 92f8230 | 2006-12-31 07:18:34 +0000 | [diff] [blame] | 2026 | </td> |
| 2027 | </tr><tr class="layout"> |
| 2028 | <td class="left"><tt>i32 (i8*, ...)</tt></td> |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 2029 | <td class="left">A vararg function that takes at least one |
| 2030 | <a href="#t_pointer">pointer</a> to <tt>i8 </tt> (char in C), |
| 2031 | which returns an integer. This is the signature for <tt>printf</tt> in |
| Reid Spencer | 92f8230 | 2006-12-31 07:18:34 +0000 | [diff] [blame] | 2032 | LLVM. |
| Reid Spencer | d3f876c | 2004-11-01 08:19:36 +0000 | [diff] [blame] | 2033 | </td> |
| Devang Patel | a582f40 | 2008-03-24 05:35:41 +0000 | [diff] [blame] | 2034 | </tr><tr class="layout"> |
| 2035 | <td class="left"><tt>{i32, i32} (i32)</tt></td> |
| Nick Lewycky | 5138694 | 2009-09-27 07:55:32 +0000 | [diff] [blame] | 2036 | <td class="left">A function taking an <tt>i32</tt>, returning a |
| 2037 | <a href="#t_struct">structure</a> containing two <tt>i32</tt> values |
| Devang Patel | a582f40 | 2008-03-24 05:35:41 +0000 | [diff] [blame] | 2038 | </td> |
| Reid Spencer | d3f876c | 2004-11-01 08:19:36 +0000 | [diff] [blame] | 2039 | </tr> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 2040 | </table> |
| Misha Brukman | daa4cb0 | 2004-03-01 17:47:27 +0000 | [diff] [blame] | 2041 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 2042 | </div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2043 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 2044 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 2045 | <h4> |
| 2046 | <a name="t_struct">Structure Type</a> |
| 2047 | </h4> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2048 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 2049 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2050 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 2051 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2052 | <p>The structure type is used to represent a collection of data members together |
| Chris Lattner | 1afcace | 2011-07-09 17:41:24 +0000 | [diff] [blame] | 2053 | in memory. The elements of a structure may be any type that has a size.</p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2054 | |
| Jeffrey Yasskin | 7a088cf | 2010-01-11 19:19:26 +0000 | [diff] [blame] | 2055 | <p>Structures in memory are accessed using '<tt><a href="#i_load">load</a></tt>' |
| 2056 | and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a field |
| 2057 | with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction. |
| 2058 | Structures in registers are accessed using the |
| 2059 | '<tt><a href="#i_extractvalue">extractvalue</a></tt>' and |
| 2060 | '<tt><a href="#i_insertvalue">insertvalue</a></tt>' instructions.</p> |
| Chris Lattner | 1afcace | 2011-07-09 17:41:24 +0000 | [diff] [blame] | 2061 | |
| 2062 | <p>Structures may optionally be "packed" structures, which indicate that the |
| 2063 | alignment of the struct is one byte, and that there is no padding between |
| Chris Lattner | 2c38d65 | 2011-08-12 17:31:02 +0000 | [diff] [blame] | 2064 | the elements. In non-packed structs, padding between field types is inserted |
| 2065 | as defined by the TargetData string in the module, which is required to match |
| Chris Lattner | e4617b0 | 2011-10-11 23:02:17 +0000 | [diff] [blame] | 2066 | what the underlying code generator expects.</p> |
| Chris Lattner | 1afcace | 2011-07-09 17:41:24 +0000 | [diff] [blame] | 2067 | |
| Chris Lattner | 2c38d65 | 2011-08-12 17:31:02 +0000 | [diff] [blame] | 2068 | <p>Structures can either be "literal" or "identified". A literal structure is |
| 2069 | defined inline with other types (e.g. <tt>{i32, i32}*</tt>) whereas identified |
| 2070 | types are always defined at the top level with a name. Literal types are |
| 2071 | uniqued by their contents and can never be recursive or opaque since there is |
| Chris Lattner | aa175c3 | 2011-08-12 18:12:40 +0000 | [diff] [blame] | 2072 | no way to write one. Identified types can be recursive, can be opaqued, and are |
| Chris Lattner | 2c38d65 | 2011-08-12 17:31:02 +0000 | [diff] [blame] | 2073 | never uniqued. |
| Chris Lattner | 1afcace | 2011-07-09 17:41:24 +0000 | [diff] [blame] | 2074 | </p> |
| 2075 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 2076 | <h5>Syntax:</h5> |
| Bill Wendling | c39e3e0 | 2009-07-20 02:39:26 +0000 | [diff] [blame] | 2077 | <pre> |
| Chris Lattner | 2c38d65 | 2011-08-12 17:31:02 +0000 | [diff] [blame] | 2078 | %T1 = type { <type list> } <i>; Identified normal struct type</i> |
| 2079 | %T2 = type <{ <type list> }> <i>; Identified packed struct type</i> |
| Bill Wendling | c39e3e0 | 2009-07-20 02:39:26 +0000 | [diff] [blame] | 2080 | </pre> |
| Chris Lattner | 1afcace | 2011-07-09 17:41:24 +0000 | [diff] [blame] | 2081 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 2082 | <h5>Examples:</h5> |
| Reid Spencer | d3f876c | 2004-11-01 08:19:36 +0000 | [diff] [blame] | 2083 | <table class="layout"> |
| 2084 | <tr class="layout"> |
| Jeff Cohen | 6f1cc77 | 2007-04-22 01:17:39 +0000 | [diff] [blame] | 2085 | <td class="left"><tt>{ i32, i32, i32 }</tt></td> |
| 2086 | <td class="left">A triple of three <tt>i32</tt> values</td> |
| Chris Lattner | 1afcace | 2011-07-09 17:41:24 +0000 | [diff] [blame] | 2087 | </tr> |
| 2088 | <tr class="layout"> |
| Jeff Cohen | 6f1cc77 | 2007-04-22 01:17:39 +0000 | [diff] [blame] | 2089 | <td class="left"><tt>{ float, i32 (i32) * }</tt></td> |
| 2090 | <td class="left">A pair, where the first element is a <tt>float</tt> and the |
| 2091 | second element is a <a href="#t_pointer">pointer</a> to a |
| 2092 | <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning |
| 2093 | an <tt>i32</tt>.</td> |
| Reid Spencer | d3f876c | 2004-11-01 08:19:36 +0000 | [diff] [blame] | 2094 | </tr> |
| Chris Lattner | 1afcace | 2011-07-09 17:41:24 +0000 | [diff] [blame] | 2095 | <tr class="layout"> |
| 2096 | <td class="left"><tt><{ i8, i32 }></tt></td> |
| 2097 | <td class="left">A packed struct known to be 5 bytes in size.</td> |
| 2098 | </tr> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 2099 | </table> |
| Dan Gohman | d8791e5 | 2009-01-24 15:58:40 +0000 | [diff] [blame] | 2100 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 2101 | </div> |
| Chris Lattner | 1afcace | 2011-07-09 17:41:24 +0000 | [diff] [blame] | 2102 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 2103 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 2104 | <h4> |
| Chris Lattner | 628ed39 | 2011-07-23 19:59:08 +0000 | [diff] [blame] | 2105 | <a name="t_opaque">Opaque Structure Types</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 2106 | </h4> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2107 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 2108 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2109 | |
| Andrew Lenharth | 75e1068 | 2006-12-08 17:13:00 +0000 | [diff] [blame] | 2110 | <h5>Overview:</h5> |
| Chris Lattner | 628ed39 | 2011-07-23 19:59:08 +0000 | [diff] [blame] | 2111 | <p>Opaque structure types are used to represent named structure types that do |
| 2112 | not have a body specified. This corresponds (for example) to the C notion of |
| 2113 | a forward declared structure.</p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2114 | |
| Andrew Lenharth | 75e1068 | 2006-12-08 17:13:00 +0000 | [diff] [blame] | 2115 | <h5>Syntax:</h5> |
| Bill Wendling | c39e3e0 | 2009-07-20 02:39:26 +0000 | [diff] [blame] | 2116 | <pre> |
| Chris Lattner | 1afcace | 2011-07-09 17:41:24 +0000 | [diff] [blame] | 2117 | %X = type opaque |
| 2118 | %52 = type opaque |
| Bill Wendling | c39e3e0 | 2009-07-20 02:39:26 +0000 | [diff] [blame] | 2119 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2120 | |
| Andrew Lenharth | 75e1068 | 2006-12-08 17:13:00 +0000 | [diff] [blame] | 2121 | <h5>Examples:</h5> |
| 2122 | <table class="layout"> |
| 2123 | <tr class="layout"> |
| Chris Lattner | 1afcace | 2011-07-09 17:41:24 +0000 | [diff] [blame] | 2124 | <td class="left"><tt>opaque</tt></td> |
| 2125 | <td class="left">An opaque type.</td> |
| Andrew Lenharth | 75e1068 | 2006-12-08 17:13:00 +0000 | [diff] [blame] | 2126 | </tr> |
| 2127 | </table> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2128 | |
| Andrew Lenharth | 75e1068 | 2006-12-08 17:13:00 +0000 | [diff] [blame] | 2129 | </div> |
| 2130 | |
| Chris Lattner | 1afcace | 2011-07-09 17:41:24 +0000 | [diff] [blame] | 2131 | |
| 2132 | |
| Andrew Lenharth | 75e1068 | 2006-12-08 17:13:00 +0000 | [diff] [blame] | 2133 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 2134 | <h4> |
| 2135 | <a name="t_pointer">Pointer Type</a> |
| 2136 | </h4> |
| Chris Lattner | 0fd4a27 | 2009-02-08 19:53:29 +0000 | [diff] [blame] | 2137 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 2138 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2139 | |
| 2140 | <h5>Overview:</h5> |
| Dan Gohman | ff3ef32 | 2010-02-25 16:50:07 +0000 | [diff] [blame] | 2141 | <p>The pointer type is used to specify memory locations. |
| 2142 | Pointers are commonly used to reference objects in memory.</p> |
| 2143 | |
| 2144 | <p>Pointer types may have an optional address space attribute defining the |
| 2145 | numbered address space where the pointed-to object resides. The default |
| 2146 | address space is number zero. The semantics of non-zero address |
| 2147 | spaces are target-specific.</p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2148 | |
| 2149 | <p>Note that LLVM does not permit pointers to void (<tt>void*</tt>) nor does it |
| 2150 | permit pointers to labels (<tt>label*</tt>). Use <tt>i8*</tt> instead.</p> |
| Chris Lattner | 0fd4a27 | 2009-02-08 19:53:29 +0000 | [diff] [blame] | 2151 | |
| Chris Lattner | 7faa883 | 2002-04-14 06:13:44 +0000 | [diff] [blame] | 2152 | <h5>Syntax:</h5> |
| Bill Wendling | c39e3e0 | 2009-07-20 02:39:26 +0000 | [diff] [blame] | 2153 | <pre> |
| 2154 | <type> * |
| 2155 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2156 | |
| Chris Lattner | 7faa883 | 2002-04-14 06:13:44 +0000 | [diff] [blame] | 2157 | <h5>Examples:</h5> |
| Reid Spencer | d3f876c | 2004-11-01 08:19:36 +0000 | [diff] [blame] | 2158 | <table class="layout"> |
| 2159 | <tr class="layout"> |
| Dan Gohman | 2a08c53 | 2009-01-04 23:44:43 +0000 | [diff] [blame] | 2160 | <td class="left"><tt>[4 x i32]*</tt></td> |
| Chris Lattner | 23ff1f9 | 2007-12-19 05:04:11 +0000 | [diff] [blame] | 2161 | <td class="left">A <a href="#t_pointer">pointer</a> to <a |
| 2162 | href="#t_array">array</a> of four <tt>i32</tt> values.</td> |
| 2163 | </tr> |
| 2164 | <tr class="layout"> |
| Dan Gohman | fe47aae | 2010-05-28 17:13:49 +0000 | [diff] [blame] | 2165 | <td class="left"><tt>i32 (i32*) *</tt></td> |
| Chris Lattner | 23ff1f9 | 2007-12-19 05:04:11 +0000 | [diff] [blame] | 2166 | <td class="left"> A <a href="#t_pointer">pointer</a> to a <a |
| Reid Spencer | ca86e16 | 2006-12-31 07:07:53 +0000 | [diff] [blame] | 2167 | href="#t_function">function</a> that takes an <tt>i32*</tt>, returning an |
| Chris Lattner | 23ff1f9 | 2007-12-19 05:04:11 +0000 | [diff] [blame] | 2168 | <tt>i32</tt>.</td> |
| 2169 | </tr> |
| 2170 | <tr class="layout"> |
| 2171 | <td class="left"><tt>i32 addrspace(5)*</tt></td> |
| 2172 | <td class="left">A <a href="#t_pointer">pointer</a> to an <tt>i32</tt> value |
| 2173 | that resides in address space #5.</td> |
| Reid Spencer | d3f876c | 2004-11-01 08:19:36 +0000 | [diff] [blame] | 2174 | </tr> |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 2175 | </table> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2176 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 2177 | </div> |
| Reid Spencer | d3f876c | 2004-11-01 08:19:36 +0000 | [diff] [blame] | 2178 | |
| Chris Lattner | a58561b | 2004-08-12 19:12:28 +0000 | [diff] [blame] | 2179 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 2180 | <h4> |
| 2181 | <a name="t_vector">Vector Type</a> |
| 2182 | </h4> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2183 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 2184 | <div> |
| Chris Lattner | 69c11bb | 2005-04-25 17:34:15 +0000 | [diff] [blame] | 2185 | |
| Chris Lattner | a58561b | 2004-08-12 19:12:28 +0000 | [diff] [blame] | 2186 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2187 | <p>A vector type is a simple derived type that represents a vector of elements. |
| 2188 | Vector types are used when multiple primitive data are operated in parallel |
| 2189 | using a single instruction (SIMD). A vector type requires a size (number of |
| Duncan Sands | d40d14e | 2009-11-27 13:38:03 +0000 | [diff] [blame] | 2190 | elements) and an underlying primitive data type. Vector types are considered |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2191 | <a href="#t_firstclass">first class</a>.</p> |
| Chris Lattner | 69c11bb | 2005-04-25 17:34:15 +0000 | [diff] [blame] | 2192 | |
| Chris Lattner | a58561b | 2004-08-12 19:12:28 +0000 | [diff] [blame] | 2193 | <h5>Syntax:</h5> |
| Chris Lattner | 69c11bb | 2005-04-25 17:34:15 +0000 | [diff] [blame] | 2194 | <pre> |
| 2195 | < <# elements> x <elementtype> > |
| 2196 | </pre> |
| 2197 | |
| Chris Lattner | 7d2e7be | 2010-10-10 18:20:35 +0000 | [diff] [blame] | 2198 | <p>The number of elements is a constant integer value larger than 0; elementtype |
| Nadav Rotem | 1608769 | 2011-12-05 06:29:09 +0000 | [diff] [blame] | 2199 | may be any integer or floating point type, or a pointer to these types. |
| 2200 | Vectors of size zero are not allowed. </p> |
| Chris Lattner | 69c11bb | 2005-04-25 17:34:15 +0000 | [diff] [blame] | 2201 | |
| Chris Lattner | a58561b | 2004-08-12 19:12:28 +0000 | [diff] [blame] | 2202 | <h5>Examples:</h5> |
| Reid Spencer | d3f876c | 2004-11-01 08:19:36 +0000 | [diff] [blame] | 2203 | <table class="layout"> |
| 2204 | <tr class="layout"> |
| Chris Lattner | 23ff1f9 | 2007-12-19 05:04:11 +0000 | [diff] [blame] | 2205 | <td class="left"><tt><4 x i32></tt></td> |
| 2206 | <td class="left">Vector of 4 32-bit integer values.</td> |
| 2207 | </tr> |
| 2208 | <tr class="layout"> |
| 2209 | <td class="left"><tt><8 x float></tt></td> |
| 2210 | <td class="left">Vector of 8 32-bit floating-point values.</td> |
| 2211 | </tr> |
| 2212 | <tr class="layout"> |
| 2213 | <td class="left"><tt><2 x i64></tt></td> |
| 2214 | <td class="left">Vector of 2 64-bit integer values.</td> |
| Reid Spencer | d3f876c | 2004-11-01 08:19:36 +0000 | [diff] [blame] | 2215 | </tr> |
| Nadav Rotem | 1608769 | 2011-12-05 06:29:09 +0000 | [diff] [blame] | 2216 | <tr class="layout"> |
| 2217 | <td class="left"><tt><4 x i64*></tt></td> |
| 2218 | <td class="left">Vector of 4 pointers to 64-bit integer values.</td> |
| 2219 | </tr> |
| Reid Spencer | d3f876c | 2004-11-01 08:19:36 +0000 | [diff] [blame] | 2220 | </table> |
| Dan Gohman | d8791e5 | 2009-01-24 15:58:40 +0000 | [diff] [blame] | 2221 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 2222 | </div> |
| 2223 | |
| Bill Wendling | af75f0c | 2011-07-31 06:47:33 +0000 | [diff] [blame] | 2224 | </div> |
| 2225 | |
| NAKAMURA Takumi | 4b2e07a | 2011-10-31 13:04:26 +0000 | [diff] [blame] | 2226 | </div> |
| 2227 | |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2228 | <!-- *********************************************************************** --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 2229 | <h2><a name="constants">Constants</a></h2> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2230 | <!-- *********************************************************************** --> |
| 2231 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 2232 | <div> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2233 | |
| 2234 | <p>LLVM has several different basic types of constants. This section describes |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2235 | them all and their syntax.</p> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2236 | |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2237 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 2238 | <h3> |
| 2239 | <a name="simpleconstants">Simple Constants</a> |
| 2240 | </h3> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2241 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 2242 | <div> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2243 | |
| 2244 | <dl> |
| 2245 | <dt><b>Boolean constants</b></dt> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2246 | <dd>The two strings '<tt>true</tt>' and '<tt>false</tt>' are both valid |
| Nick Lewycky | ec38da4 | 2009-09-27 00:45:11 +0000 | [diff] [blame] | 2247 | constants of the <tt><a href="#t_integer">i1</a></tt> type.</dd> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2248 | |
| 2249 | <dt><b>Integer constants</b></dt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2250 | <dd>Standard integers (such as '4') are constants of |
| 2251 | the <a href="#t_integer">integer</a> type. Negative numbers may be used |
| 2252 | with integer types.</dd> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2253 | |
| 2254 | <dt><b>Floating point constants</b></dt> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2255 | <dd>Floating point constants use standard decimal notation (e.g. 123.421), |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2256 | exponential notation (e.g. 1.23421e+2), or a more precise hexadecimal |
| 2257 | notation (see below). The assembler requires the exact decimal value of a |
| 2258 | floating-point constant. For example, the assembler accepts 1.25 but |
| 2259 | rejects 1.3 because 1.3 is a repeating decimal in binary. Floating point |
| 2260 | constants must have a <a href="#t_floating">floating point</a> type. </dd> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2261 | |
| 2262 | <dt><b>Null pointer constants</b></dt> |
| John Criswell | 9e2485c | 2004-12-10 15:51:16 +0000 | [diff] [blame] | 2263 | <dd>The identifier '<tt>null</tt>' is recognized as a null pointer constant |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2264 | and must be of <a href="#t_pointer">pointer type</a>.</dd> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2265 | </dl> |
| 2266 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2267 | <p>The one non-intuitive notation for constants is the hexadecimal form of |
| 2268 | floating point constants. For example, the form '<tt>double |
| 2269 | 0x432ff973cafa8000</tt>' is equivalent to (but harder to read than) |
| 2270 | '<tt>double 4.5e+15</tt>'. The only time hexadecimal floating point |
| 2271 | constants are required (and the only time that they are generated by the |
| 2272 | disassembler) is when a floating point constant must be emitted but it cannot |
| 2273 | be represented as a decimal floating point number in a reasonable number of |
| 2274 | digits. For example, NaN's, infinities, and other special values are |
| 2275 | represented in their IEEE hexadecimal format so that assembly and disassembly |
| 2276 | do not cause any bits to change in the constants.</p> |
| 2277 | |
| Dan Gohman | ce16339 | 2011-12-17 00:04:22 +0000 | [diff] [blame] | 2278 | <p>When using the hexadecimal form, constants of types half, float, and double are |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2279 | represented using the 16-digit form shown above (which matches the IEEE754 |
| Dan Gohman | ce16339 | 2011-12-17 00:04:22 +0000 | [diff] [blame] | 2280 | representation for double); half and float values must, however, be exactly |
| 2281 | representable as IEE754 half and single precision, respectively. |
| 2282 | Hexadecimal format is always used |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2283 | for long double, and there are three forms of long double. The 80-bit format |
| 2284 | used by x86 is represented as <tt>0xK</tt> followed by 20 hexadecimal digits. |
| 2285 | The 128-bit format used by PowerPC (two adjacent doubles) is represented |
| 2286 | by <tt>0xM</tt> followed by 32 hexadecimal digits. The IEEE 128-bit format |
| 2287 | is represented by <tt>0xL</tt> followed by 32 hexadecimal digits; no |
| 2288 | currently supported target uses this format. Long doubles will only work if |
| 2289 | they match the long double format on your target. All hexadecimal formats |
| 2290 | are big-endian (sign bit at the left).</p> |
| 2291 | |
| Dale Johannesen | 21fe99b | 2010-10-01 00:48:59 +0000 | [diff] [blame] | 2292 | <p>There are no constants of type x86mmx.</p> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2293 | </div> |
| 2294 | |
| 2295 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 2296 | <h3> |
| Bill Wendling | d9fe298 | 2009-07-20 02:32:41 +0000 | [diff] [blame] | 2297 | <a name="aggregateconstants"></a> <!-- old anchor --> |
| 2298 | <a name="complexconstants">Complex Constants</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 2299 | </h3> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2300 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 2301 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2302 | |
| Chris Lattner | 7088279 | 2009-02-28 18:32:25 +0000 | [diff] [blame] | 2303 | <p>Complex constants are a (potentially recursive) combination of simple |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2304 | constants and smaller complex constants.</p> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2305 | |
| 2306 | <dl> |
| 2307 | <dt><b>Structure constants</b></dt> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2308 | <dd>Structure constants are represented with notation similar to structure |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2309 | type definitions (a comma separated list of elements, surrounded by braces |
| 2310 | (<tt>{}</tt>)). For example: "<tt>{ i32 4, float 17.0, i32* @G }</tt>", |
| 2311 | where "<tt>@G</tt>" is declared as "<tt>@G = external global i32</tt>". |
| 2312 | Structure constants must have <a href="#t_struct">structure type</a>, and |
| 2313 | the number and types of elements must match those specified by the |
| 2314 | type.</dd> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2315 | |
| 2316 | <dt><b>Array constants</b></dt> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2317 | <dd>Array constants are represented with notation similar to array type |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2318 | definitions (a comma separated list of elements, surrounded by square |
| 2319 | brackets (<tt>[]</tt>)). For example: "<tt>[ i32 42, i32 11, i32 74 |
| 2320 | ]</tt>". Array constants must have <a href="#t_array">array type</a>, and |
| 2321 | the number and types of elements must match those specified by the |
| 2322 | type.</dd> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2323 | |
| Reid Spencer | 485bad1 | 2007-02-15 03:07:05 +0000 | [diff] [blame] | 2324 | <dt><b>Vector constants</b></dt> |
| Reid Spencer | 485bad1 | 2007-02-15 03:07:05 +0000 | [diff] [blame] | 2325 | <dd>Vector constants are represented with notation similar to vector type |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2326 | definitions (a comma separated list of elements, surrounded by |
| 2327 | less-than/greater-than's (<tt><></tt>)). For example: "<tt>< i32 |
| 2328 | 42, i32 11, i32 74, i32 100 ></tt>". Vector constants must |
| 2329 | have <a href="#t_vector">vector type</a>, and the number and types of |
| 2330 | elements must match those specified by the type.</dd> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2331 | |
| 2332 | <dt><b>Zero initialization</b></dt> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2333 | <dd>The string '<tt>zeroinitializer</tt>' can be used to zero initialize a |
| Chris Lattner | fdfeb69 | 2010-02-12 20:49:41 +0000 | [diff] [blame] | 2334 | value to zero of <em>any</em> type, including scalar and |
| 2335 | <a href="#t_aggregate">aggregate</a> types. |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2336 | This is often used to avoid having to print large zero initializers |
| 2337 | (e.g. for large arrays) and is always exactly equivalent to using explicit |
| 2338 | zero initializers.</dd> |
| Nick Lewycky | 21cc446 | 2009-04-04 07:22:01 +0000 | [diff] [blame] | 2339 | |
| 2340 | <dt><b>Metadata node</b></dt> |
| Nick Lewycky | 1e8c7a6 | 2009-05-30 16:08:30 +0000 | [diff] [blame] | 2341 | <dd>A metadata node is a structure-like constant with |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2342 | <a href="#t_metadata">metadata type</a>. For example: "<tt>metadata !{ |
| 2343 | i32 0, metadata !"test" }</tt>". Unlike other constants that are meant to |
| 2344 | be interpreted as part of the instruction stream, metadata is a place to |
| 2345 | attach additional information such as debug info.</dd> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2346 | </dl> |
| 2347 | |
| 2348 | </div> |
| 2349 | |
| 2350 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 2351 | <h3> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2352 | <a name="globalconstants">Global Variable and Function Addresses</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 2353 | </h3> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2354 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 2355 | <div> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2356 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2357 | <p>The addresses of <a href="#globalvars">global variables</a> |
| 2358 | and <a href="#functionstructure">functions</a> are always implicitly valid |
| 2359 | (link-time) constants. These constants are explicitly referenced when |
| 2360 | the <a href="#identifiers">identifier for the global</a> is used and always |
| 2361 | have <a href="#t_pointer">pointer</a> type. For example, the following is a |
| 2362 | legal LLVM file:</p> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2363 | |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 2364 | <pre class="doc_code"> |
| Chris Lattner | a18a424 | 2007-06-06 18:28:13 +0000 | [diff] [blame] | 2365 | @X = global i32 17 |
| 2366 | @Y = global i32 42 |
| 2367 | @Z = global [2 x i32*] [ i32* @X, i32* @Y ] |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2368 | </pre> |
| 2369 | |
| 2370 | </div> |
| 2371 | |
| 2372 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 2373 | <h3> |
| 2374 | <a name="undefvalues">Undefined Values</a> |
| 2375 | </h3> |
| 2376 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 2377 | <div> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2378 | |
| Chris Lattner | 48a109c | 2009-09-07 22:52:39 +0000 | [diff] [blame] | 2379 | <p>The string '<tt>undef</tt>' can be used anywhere a constant is expected, and |
| Benjamin Kramer | 8040cd3 | 2009-10-12 14:46:08 +0000 | [diff] [blame] | 2380 | indicates that the user of the value may receive an unspecified bit-pattern. |
| Bill Wendling | 1b383ba | 2010-10-27 01:07:41 +0000 | [diff] [blame] | 2381 | Undefined values may be of any type (other than '<tt>label</tt>' |
| 2382 | or '<tt>void</tt>') and be used anywhere a constant is permitted.</p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2383 | |
| Chris Lattner | c608cb1 | 2009-09-11 01:49:31 +0000 | [diff] [blame] | 2384 | <p>Undefined values are useful because they indicate to the compiler that the |
| Chris Lattner | 48a109c | 2009-09-07 22:52:39 +0000 | [diff] [blame] | 2385 | program is well defined no matter what value is used. This gives the |
| 2386 | compiler more freedom to optimize. Here are some examples of (potentially |
| 2387 | surprising) transformations that are valid (in pseudo IR):</p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2388 | |
| Chris Lattner | 48a109c | 2009-09-07 22:52:39 +0000 | [diff] [blame] | 2389 | |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 2390 | <pre class="doc_code"> |
| Chris Lattner | 48a109c | 2009-09-07 22:52:39 +0000 | [diff] [blame] | 2391 | %A = add %X, undef |
| 2392 | %B = sub %X, undef |
| 2393 | %C = xor %X, undef |
| 2394 | Safe: |
| 2395 | %A = undef |
| 2396 | %B = undef |
| 2397 | %C = undef |
| 2398 | </pre> |
| Chris Lattner | 48a109c | 2009-09-07 22:52:39 +0000 | [diff] [blame] | 2399 | |
| 2400 | <p>This is safe because all of the output bits are affected by the undef bits. |
| Bill Wendling | 1b383ba | 2010-10-27 01:07:41 +0000 | [diff] [blame] | 2401 | Any output bit can have a zero or one depending on the input bits.</p> |
| Chris Lattner | 48a109c | 2009-09-07 22:52:39 +0000 | [diff] [blame] | 2402 | |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 2403 | <pre class="doc_code"> |
| Chris Lattner | 48a109c | 2009-09-07 22:52:39 +0000 | [diff] [blame] | 2404 | %A = or %X, undef |
| 2405 | %B = and %X, undef |
| 2406 | Safe: |
| 2407 | %A = -1 |
| 2408 | %B = 0 |
| 2409 | Unsafe: |
| 2410 | %A = undef |
| 2411 | %B = undef |
| 2412 | </pre> |
| Chris Lattner | 48a109c | 2009-09-07 22:52:39 +0000 | [diff] [blame] | 2413 | |
| 2414 | <p>These logical operations have bits that are not always affected by the input. |
| Bill Wendling | 1b383ba | 2010-10-27 01:07:41 +0000 | [diff] [blame] | 2415 | For example, if <tt>%X</tt> has a zero bit, then the output of the |
| 2416 | '<tt>and</tt>' operation will always be a zero for that bit, no matter what |
| 2417 | the corresponding bit from the '<tt>undef</tt>' is. As such, it is unsafe to |
| 2418 | optimize or assume that the result of the '<tt>and</tt>' is '<tt>undef</tt>'. |
| 2419 | However, it is safe to assume that all bits of the '<tt>undef</tt>' could be |
| 2420 | 0, and optimize the '<tt>and</tt>' to 0. Likewise, it is safe to assume that |
| 2421 | all the bits of the '<tt>undef</tt>' operand to the '<tt>or</tt>' could be |
| 2422 | set, allowing the '<tt>or</tt>' to be folded to -1.</p> |
| Chris Lattner | 48a109c | 2009-09-07 22:52:39 +0000 | [diff] [blame] | 2423 | |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 2424 | <pre class="doc_code"> |
| Chris Lattner | 48a109c | 2009-09-07 22:52:39 +0000 | [diff] [blame] | 2425 | %A = select undef, %X, %Y |
| 2426 | %B = select undef, 42, %Y |
| 2427 | %C = select %X, %Y, undef |
| 2428 | Safe: |
| 2429 | %A = %X (or %Y) |
| 2430 | %B = 42 (or %Y) |
| 2431 | %C = %Y |
| 2432 | Unsafe: |
| 2433 | %A = undef |
| 2434 | %B = undef |
| 2435 | %C = undef |
| 2436 | </pre> |
| Chris Lattner | 48a109c | 2009-09-07 22:52:39 +0000 | [diff] [blame] | 2437 | |
| Bill Wendling | 1b383ba | 2010-10-27 01:07:41 +0000 | [diff] [blame] | 2438 | <p>This set of examples shows that undefined '<tt>select</tt>' (and conditional |
| 2439 | branch) conditions can go <em>either way</em>, but they have to come from one |
| 2440 | of the two operands. In the <tt>%A</tt> example, if <tt>%X</tt> and |
| 2441 | <tt>%Y</tt> were both known to have a clear low bit, then <tt>%A</tt> would |
| 2442 | have to have a cleared low bit. However, in the <tt>%C</tt> example, the |
| 2443 | optimizer is allowed to assume that the '<tt>undef</tt>' operand could be the |
| 2444 | same as <tt>%Y</tt>, allowing the whole '<tt>select</tt>' to be |
| 2445 | eliminated.</p> |
| Chris Lattner | 48a109c | 2009-09-07 22:52:39 +0000 | [diff] [blame] | 2446 | |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 2447 | <pre class="doc_code"> |
| Chris Lattner | 48a109c | 2009-09-07 22:52:39 +0000 | [diff] [blame] | 2448 | %A = xor undef, undef |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 2449 | |
| Chris Lattner | 48a109c | 2009-09-07 22:52:39 +0000 | [diff] [blame] | 2450 | %B = undef |
| 2451 | %C = xor %B, %B |
| 2452 | |
| 2453 | %D = undef |
| 2454 | %E = icmp lt %D, 4 |
| 2455 | %F = icmp gte %D, 4 |
| 2456 | |
| 2457 | Safe: |
| 2458 | %A = undef |
| 2459 | %B = undef |
| 2460 | %C = undef |
| 2461 | %D = undef |
| 2462 | %E = undef |
| 2463 | %F = undef |
| 2464 | </pre> |
| Chris Lattner | 48a109c | 2009-09-07 22:52:39 +0000 | [diff] [blame] | 2465 | |
| Bill Wendling | 1b383ba | 2010-10-27 01:07:41 +0000 | [diff] [blame] | 2466 | <p>This example points out that two '<tt>undef</tt>' operands are not |
| 2467 | necessarily the same. This can be surprising to people (and also matches C |
| 2468 | semantics) where they assume that "<tt>X^X</tt>" is always zero, even |
| 2469 | if <tt>X</tt> is undefined. This isn't true for a number of reasons, but the |
| 2470 | short answer is that an '<tt>undef</tt>' "variable" can arbitrarily change |
| 2471 | its value over its "live range". This is true because the variable doesn't |
| 2472 | actually <em>have a live range</em>. Instead, the value is logically read |
| 2473 | from arbitrary registers that happen to be around when needed, so the value |
| 2474 | is not necessarily consistent over time. In fact, <tt>%A</tt> and <tt>%C</tt> |
| 2475 | need to have the same semantics or the core LLVM "replace all uses with" |
| 2476 | concept would not hold.</p> |
| Chris Lattner | 6e9057b | 2009-09-07 23:33:52 +0000 | [diff] [blame] | 2477 | |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 2478 | <pre class="doc_code"> |
| Chris Lattner | 6e9057b | 2009-09-07 23:33:52 +0000 | [diff] [blame] | 2479 | %A = fdiv undef, %X |
| 2480 | %B = fdiv %X, undef |
| 2481 | Safe: |
| 2482 | %A = undef |
| 2483 | b: unreachable |
| 2484 | </pre> |
| Chris Lattner | 6e9057b | 2009-09-07 23:33:52 +0000 | [diff] [blame] | 2485 | |
| 2486 | <p>These examples show the crucial difference between an <em>undefined |
| Bill Wendling | 1b383ba | 2010-10-27 01:07:41 +0000 | [diff] [blame] | 2487 | value</em> and <em>undefined behavior</em>. An undefined value (like |
| 2488 | '<tt>undef</tt>') is allowed to have an arbitrary bit-pattern. This means that |
| 2489 | the <tt>%A</tt> operation can be constant folded to '<tt>undef</tt>', because |
| 2490 | the '<tt>undef</tt>' could be an SNaN, and <tt>fdiv</tt> is not (currently) |
| 2491 | defined on SNaN's. However, in the second example, we can make a more |
| 2492 | aggressive assumption: because the <tt>undef</tt> is allowed to be an |
| 2493 | arbitrary value, we are allowed to assume that it could be zero. Since a |
| 2494 | divide by zero has <em>undefined behavior</em>, we are allowed to assume that |
| 2495 | the operation does not execute at all. This allows us to delete the divide and |
| 2496 | all code after it. Because the undefined operation "can't happen", the |
| 2497 | optimizer can assume that it occurs in dead code.</p> |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 2498 | |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 2499 | <pre class="doc_code"> |
| Chris Lattner | 6e9057b | 2009-09-07 23:33:52 +0000 | [diff] [blame] | 2500 | a: store undef -> %X |
| 2501 | b: store %X -> undef |
| 2502 | Safe: |
| 2503 | a: <deleted> |
| 2504 | b: unreachable |
| 2505 | </pre> |
| Chris Lattner | 6e9057b | 2009-09-07 23:33:52 +0000 | [diff] [blame] | 2506 | |
| Bill Wendling | 1b383ba | 2010-10-27 01:07:41 +0000 | [diff] [blame] | 2507 | <p>These examples reiterate the <tt>fdiv</tt> example: a store <em>of</em> an |
| 2508 | undefined value can be assumed to not have any effect; we can assume that the |
| 2509 | value is overwritten with bits that happen to match what was already there. |
| 2510 | However, a store <em>to</em> an undefined location could clobber arbitrary |
| 2511 | memory, therefore, it has undefined behavior.</p> |
| Chris Lattner | 6e9057b | 2009-09-07 23:33:52 +0000 | [diff] [blame] | 2512 | |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2513 | </div> |
| 2514 | |
| 2515 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 2516 | <h3> |
| Dan Gohman | bfb056d | 2011-12-06 03:18:47 +0000 | [diff] [blame] | 2517 | <a name="poisonvalues">Poison Values</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 2518 | </h3> |
| 2519 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 2520 | <div> |
| Dan Gohman | fff6c53 | 2010-04-22 23:14:21 +0000 | [diff] [blame] | 2521 | |
| Dan Gohman | bfb056d | 2011-12-06 03:18:47 +0000 | [diff] [blame] | 2522 | <p>Poison values are similar to <a href="#undefvalues">undef values</a>, however |
| Dan Gohman | e1a2984 | 2011-12-06 03:35:58 +0000 | [diff] [blame] | 2523 | they also represent the fact that an instruction or constant expression which |
| 2524 | cannot evoke side effects has nevertheless detected a condition which results |
| 2525 | in undefined behavior.</p> |
| Dan Gohman | fff6c53 | 2010-04-22 23:14:21 +0000 | [diff] [blame] | 2526 | |
| Dan Gohman | bfb056d | 2011-12-06 03:18:47 +0000 | [diff] [blame] | 2527 | <p>There is currently no way of representing a poison value in the IR; they |
| Dan Gohman | 855abed | 2010-05-03 14:51:43 +0000 | [diff] [blame] | 2528 | only exist when produced by operations such as |
| Dan Gohman | 34b3d99 | 2010-04-28 00:49:41 +0000 | [diff] [blame] | 2529 | <a href="#i_add"><tt>add</tt></a> with the <tt>nsw</tt> flag.</p> |
| Dan Gohman | ae11c3f | 2010-04-26 23:36:52 +0000 | [diff] [blame] | 2530 | |
| Dan Gohman | bfb056d | 2011-12-06 03:18:47 +0000 | [diff] [blame] | 2531 | <p>Poison value behavior is defined in terms of value <i>dependence</i>:</p> |
| Dan Gohman | ae11c3f | 2010-04-26 23:36:52 +0000 | [diff] [blame] | 2532 | |
| Dan Gohman | 34b3d99 | 2010-04-28 00:49:41 +0000 | [diff] [blame] | 2533 | <ul> |
| 2534 | <li>Values other than <a href="#i_phi"><tt>phi</tt></a> nodes depend on |
| 2535 | their operands.</li> |
| 2536 | |
| 2537 | <li><a href="#i_phi"><tt>Phi</tt></a> nodes depend on the operand corresponding |
| 2538 | to their dynamic predecessor basic block.</li> |
| 2539 | |
| 2540 | <li>Function arguments depend on the corresponding actual argument values in |
| 2541 | the dynamic callers of their functions.</li> |
| 2542 | |
| 2543 | <li><a href="#i_call"><tt>Call</tt></a> instructions depend on the |
| 2544 | <a href="#i_ret"><tt>ret</tt></a> instructions that dynamically transfer |
| 2545 | control back to them.</li> |
| 2546 | |
| Dan Gohman | b532816 | 2010-05-03 14:55:22 +0000 | [diff] [blame] | 2547 | <li><a href="#i_invoke"><tt>Invoke</tt></a> instructions depend on the |
| Bill Wendling | 7b9e539 | 2012-02-06 21:57:33 +0000 | [diff] [blame] | 2548 | <a href="#i_ret"><tt>ret</tt></a>, <a href="#i_resume"><tt>resume</tt></a>, |
| Dan Gohman | b532816 | 2010-05-03 14:55:22 +0000 | [diff] [blame] | 2549 | or exception-throwing call instructions that dynamically transfer control |
| 2550 | back to them.</li> |
| 2551 | |
| Dan Gohman | 34b3d99 | 2010-04-28 00:49:41 +0000 | [diff] [blame] | 2552 | <li>Non-volatile loads and stores depend on the most recent stores to all of the |
| 2553 | referenced memory addresses, following the order in the IR |
| 2554 | (including loads and stores implied by intrinsics such as |
| 2555 | <a href="#int_memcpy"><tt>@llvm.memcpy</tt></a>.)</li> |
| 2556 | |
| Dan Gohman | 7c24ff1 | 2010-05-03 14:59:34 +0000 | [diff] [blame] | 2557 | <!-- TODO: In the case of multiple threads, this only applies if the store |
| 2558 | "happens-before" the load or store. --> |
| Dan Gohman | ae11c3f | 2010-04-26 23:36:52 +0000 | [diff] [blame] | 2559 | |
| Dan Gohman | 34b3d99 | 2010-04-28 00:49:41 +0000 | [diff] [blame] | 2560 | <!-- TODO: floating-point exception state --> |
| Dan Gohman | ae11c3f | 2010-04-26 23:36:52 +0000 | [diff] [blame] | 2561 | |
| Dan Gohman | 34b3d99 | 2010-04-28 00:49:41 +0000 | [diff] [blame] | 2562 | <li>An instruction with externally visible side effects depends on the most |
| 2563 | recent preceding instruction with externally visible side effects, following |
| Dan Gohman | ff70fe4 | 2010-07-06 15:26:33 +0000 | [diff] [blame] | 2564 | the order in the IR. (This includes |
| 2565 | <a href="#volatile">volatile operations</a>.)</li> |
| Dan Gohman | fff6c53 | 2010-04-22 23:14:21 +0000 | [diff] [blame] | 2566 | |
| Dan Gohman | b532816 | 2010-05-03 14:55:22 +0000 | [diff] [blame] | 2567 | <li>An instruction <i>control-depends</i> on a |
| 2568 | <a href="#terminators">terminator instruction</a> |
| 2569 | if the terminator instruction has multiple successors and the instruction |
| 2570 | is always executed when control transfers to one of the successors, and |
| Chris Lattner | 7a2bdde | 2011-04-15 05:18:47 +0000 | [diff] [blame] | 2571 | may not be executed when control is transferred to another.</li> |
| Dan Gohman | 34b3d99 | 2010-04-28 00:49:41 +0000 | [diff] [blame] | 2572 | |
| Dan Gohman | ca4cac4 | 2011-04-12 23:05:59 +0000 | [diff] [blame] | 2573 | <li>Additionally, an instruction also <i>control-depends</i> on a terminator |
| 2574 | instruction if the set of instructions it otherwise depends on would be |
| Chris Lattner | 7a2bdde | 2011-04-15 05:18:47 +0000 | [diff] [blame] | 2575 | different if the terminator had transferred control to a different |
| Dan Gohman | ca4cac4 | 2011-04-12 23:05:59 +0000 | [diff] [blame] | 2576 | successor.</li> |
| 2577 | |
| Dan Gohman | 34b3d99 | 2010-04-28 00:49:41 +0000 | [diff] [blame] | 2578 | <li>Dependence is transitive.</li> |
| 2579 | |
| 2580 | </ul> |
| Dan Gohman | 34b3d99 | 2010-04-28 00:49:41 +0000 | [diff] [blame] | 2581 | |
| Dan Gohman | e1a2984 | 2011-12-06 03:35:58 +0000 | [diff] [blame] | 2582 | <p>Poison Values have the same behavior as <a href="#undefvalues">undef values</a>, |
| 2583 | with the additional affect that any instruction which has a <i>dependence</i> |
| 2584 | on a poison value has undefined behavior.</p> |
| Dan Gohman | 34b3d99 | 2010-04-28 00:49:41 +0000 | [diff] [blame] | 2585 | |
| 2586 | <p>Here are some examples:</p> |
| Dan Gohman | c30f6e1 | 2010-04-26 20:54:53 +0000 | [diff] [blame] | 2587 | |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 2588 | <pre class="doc_code"> |
| Dan Gohman | ae11c3f | 2010-04-26 23:36:52 +0000 | [diff] [blame] | 2589 | entry: |
| Dan Gohman | bfb056d | 2011-12-06 03:18:47 +0000 | [diff] [blame] | 2590 | %poison = sub nuw i32 0, 1 ; Results in a poison value. |
| Dan Gohman | e1a2984 | 2011-12-06 03:35:58 +0000 | [diff] [blame] | 2591 | %still_poison = and i32 %poison, 0 ; 0, but also poison. |
| Dan Gohman | bfb056d | 2011-12-06 03:18:47 +0000 | [diff] [blame] | 2592 | %poison_yet_again = getelementptr i32* @h, i32 %still_poison |
| Dan Gohman | e1a2984 | 2011-12-06 03:35:58 +0000 | [diff] [blame] | 2593 | store i32 0, i32* %poison_yet_again ; memory at @h[0] is poisoned |
| Dan Gohman | 34b3d99 | 2010-04-28 00:49:41 +0000 | [diff] [blame] | 2594 | |
| Dan Gohman | e1a2984 | 2011-12-06 03:35:58 +0000 | [diff] [blame] | 2595 | store i32 %poison, i32* @g ; Poison value stored to memory. |
| 2596 | %poison2 = load i32* @g ; Poison value loaded back from memory. |
| Dan Gohman | 34b3d99 | 2010-04-28 00:49:41 +0000 | [diff] [blame] | 2597 | |
| Dan Gohman | bfb056d | 2011-12-06 03:18:47 +0000 | [diff] [blame] | 2598 | store volatile i32 %poison, i32* @g ; External observation; undefined behavior. |
| Dan Gohman | 34b3d99 | 2010-04-28 00:49:41 +0000 | [diff] [blame] | 2599 | |
| 2600 | %narrowaddr = bitcast i32* @g to i16* |
| 2601 | %wideaddr = bitcast i32* @g to i64* |
| Dan Gohman | bfb056d | 2011-12-06 03:18:47 +0000 | [diff] [blame] | 2602 | %poison3 = load i16* %narrowaddr ; Returns a poison value. |
| 2603 | %poison4 = load i64* %wideaddr ; Returns a poison value. |
| Dan Gohman | 34b3d99 | 2010-04-28 00:49:41 +0000 | [diff] [blame] | 2604 | |
| Dan Gohman | 5cdc51e | 2011-12-06 03:31:14 +0000 | [diff] [blame] | 2605 | %cmp = icmp slt i32 %poison, 0 ; Returns a poison value. |
| 2606 | br i1 %cmp, label %true, label %end ; Branch to either destination. |
| Dan Gohman | ae11c3f | 2010-04-26 23:36:52 +0000 | [diff] [blame] | 2607 | |
| 2608 | true: |
| Dan Gohman | 5cdc51e | 2011-12-06 03:31:14 +0000 | [diff] [blame] | 2609 | store volatile i32 0, i32* @g ; This is control-dependent on %cmp, so |
| 2610 | ; it has undefined behavior. |
| Dan Gohman | ae11c3f | 2010-04-26 23:36:52 +0000 | [diff] [blame] | 2611 | br label %end |
| 2612 | |
| 2613 | end: |
| 2614 | %p = phi i32 [ 0, %entry ], [ 1, %true ] |
| Dan Gohman | 5cdc51e | 2011-12-06 03:31:14 +0000 | [diff] [blame] | 2615 | ; Both edges into this PHI are |
| 2616 | ; control-dependent on %cmp, so this |
| 2617 | ; always results in a poison value. |
| Dan Gohman | ae11c3f | 2010-04-26 23:36:52 +0000 | [diff] [blame] | 2618 | |
| Dan Gohman | 5cdc51e | 2011-12-06 03:31:14 +0000 | [diff] [blame] | 2619 | store volatile i32 0, i32* @g ; This would depend on the store in %true |
| 2620 | ; if %cmp is true, or the store in %entry |
| 2621 | ; otherwise, so this is undefined behavior. |
| Dan Gohman | ca4cac4 | 2011-04-12 23:05:59 +0000 | [diff] [blame] | 2622 | |
| Nick Lewycky | 64f9fb1 | 2011-05-16 19:29:30 +0000 | [diff] [blame] | 2623 | br i1 %cmp, label %second_true, label %second_end |
| Dan Gohman | 5cdc51e | 2011-12-06 03:31:14 +0000 | [diff] [blame] | 2624 | ; The same branch again, but this time the |
| 2625 | ; true block doesn't have side effects. |
| Dan Gohman | ca4cac4 | 2011-04-12 23:05:59 +0000 | [diff] [blame] | 2626 | |
| 2627 | second_true: |
| 2628 | ; No side effects! |
| Nick Lewycky | 64f9fb1 | 2011-05-16 19:29:30 +0000 | [diff] [blame] | 2629 | ret void |
| Dan Gohman | ca4cac4 | 2011-04-12 23:05:59 +0000 | [diff] [blame] | 2630 | |
| 2631 | second_end: |
| Dan Gohman | 5cdc51e | 2011-12-06 03:31:14 +0000 | [diff] [blame] | 2632 | store volatile i32 0, i32* @g ; This time, the instruction always depends |
| 2633 | ; on the store in %end. Also, it is |
| 2634 | ; control-equivalent to %end, so this is |
| Dan Gohman | e1a2984 | 2011-12-06 03:35:58 +0000 | [diff] [blame] | 2635 | ; well-defined (ignoring earlier undefined |
| 2636 | ; behavior in this example). |
| Dan Gohman | ae11c3f | 2010-04-26 23:36:52 +0000 | [diff] [blame] | 2637 | </pre> |
| Dan Gohman | fff6c53 | 2010-04-22 23:14:21 +0000 | [diff] [blame] | 2638 | |
| Dan Gohman | fff6c53 | 2010-04-22 23:14:21 +0000 | [diff] [blame] | 2639 | </div> |
| 2640 | |
| 2641 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 2642 | <h3> |
| 2643 | <a name="blockaddress">Addresses of Basic Blocks</a> |
| 2644 | </h3> |
| 2645 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 2646 | <div> |
| Chris Lattner | c6f4436 | 2009-10-27 21:01:34 +0000 | [diff] [blame] | 2647 | |
| Chris Lattner | cdfc940 | 2009-11-01 01:27:45 +0000 | [diff] [blame] | 2648 | <p><b><tt>blockaddress(@function, %block)</tt></b></p> |
| Chris Lattner | c6f4436 | 2009-10-27 21:01:34 +0000 | [diff] [blame] | 2649 | |
| 2650 | <p>The '<tt>blockaddress</tt>' constant computes the address of the specified |
| Chris Lattner | 2dfdf2a | 2009-10-27 21:49:40 +0000 | [diff] [blame] | 2651 | basic block in the specified function, and always has an i8* type. Taking |
| Chris Lattner | cdfc940 | 2009-11-01 01:27:45 +0000 | [diff] [blame] | 2652 | the address of the entry block is illegal.</p> |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 2653 | |
| Chris Lattner | c6f4436 | 2009-10-27 21:01:34 +0000 | [diff] [blame] | 2654 | <p>This value only has defined behavior when used as an operand to the |
| Bill Wendling | 1b383ba | 2010-10-27 01:07:41 +0000 | [diff] [blame] | 2655 | '<a href="#i_indirectbr"><tt>indirectbr</tt></a>' instruction, or for |
| 2656 | comparisons against null. Pointer equality tests between labels addresses |
| 2657 | results in undefined behavior — though, again, comparison against null |
| 2658 | is ok, and no label is equal to the null pointer. This may be passed around |
| 2659 | as an opaque pointer sized value as long as the bits are not inspected. This |
| 2660 | allows <tt>ptrtoint</tt> and arithmetic to be performed on these values so |
| 2661 | long as the original value is reconstituted before the <tt>indirectbr</tt> |
| 2662 | instruction.</p> |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 2663 | |
| Bill Wendling | 1b383ba | 2010-10-27 01:07:41 +0000 | [diff] [blame] | 2664 | <p>Finally, some targets may provide defined semantics when using the value as |
| 2665 | the operand to an inline assembly, but that is target specific.</p> |
| Chris Lattner | c6f4436 | 2009-10-27 21:01:34 +0000 | [diff] [blame] | 2666 | |
| 2667 | </div> |
| 2668 | |
| 2669 | |
| 2670 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 2671 | <h3> |
| 2672 | <a name="constantexprs">Constant Expressions</a> |
| 2673 | </h3> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2674 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 2675 | <div> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2676 | |
| 2677 | <p>Constant expressions are used to allow expressions involving other constants |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2678 | to be used as constants. Constant expressions may be of |
| 2679 | any <a href="#t_firstclass">first class</a> type and may involve any LLVM |
| 2680 | operation that does not have side effects (e.g. load and call are not |
| Bill Wendling | 1b383ba | 2010-10-27 01:07:41 +0000 | [diff] [blame] | 2681 | supported). The following is the syntax for constant expressions:</p> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2682 | |
| 2683 | <dl> |
| Dan Gohman | 3dfb3cf | 2010-05-28 17:07:41 +0000 | [diff] [blame] | 2684 | <dt><b><tt>trunc (CST to TYPE)</tt></b></dt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2685 | <dd>Truncate a constant to another type. The bit size of CST must be larger |
| 2686 | than the bit size of TYPE. Both types must be integers.</dd> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2687 | |
| Dan Gohman | 3dfb3cf | 2010-05-28 17:07:41 +0000 | [diff] [blame] | 2688 | <dt><b><tt>zext (CST to TYPE)</tt></b></dt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2689 | <dd>Zero extend a constant to another type. The bit size of CST must be |
| Duncan Sands | 28afd43 | 2010-07-13 12:06:14 +0000 | [diff] [blame] | 2690 | smaller than the bit size of TYPE. Both types must be integers.</dd> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 2691 | |
| Dan Gohman | 3dfb3cf | 2010-05-28 17:07:41 +0000 | [diff] [blame] | 2692 | <dt><b><tt>sext (CST to TYPE)</tt></b></dt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2693 | <dd>Sign extend a constant to another type. The bit size of CST must be |
| Duncan Sands | 28afd43 | 2010-07-13 12:06:14 +0000 | [diff] [blame] | 2694 | smaller than the bit size of TYPE. Both types must be integers.</dd> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 2695 | |
| Dan Gohman | 3dfb3cf | 2010-05-28 17:07:41 +0000 | [diff] [blame] | 2696 | <dt><b><tt>fptrunc (CST to TYPE)</tt></b></dt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2697 | <dd>Truncate a floating point constant to another floating point type. The |
| 2698 | size of CST must be larger than the size of TYPE. Both types must be |
| 2699 | floating point.</dd> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 2700 | |
| Dan Gohman | 3dfb3cf | 2010-05-28 17:07:41 +0000 | [diff] [blame] | 2701 | <dt><b><tt>fpext (CST to TYPE)</tt></b></dt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2702 | <dd>Floating point extend a constant to another type. The size of CST must be |
| 2703 | smaller or equal to the size of TYPE. Both types must be floating |
| 2704 | point.</dd> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 2705 | |
| Dan Gohman | 3dfb3cf | 2010-05-28 17:07:41 +0000 | [diff] [blame] | 2706 | <dt><b><tt>fptoui (CST to TYPE)</tt></b></dt> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 2707 | <dd>Convert a floating point constant to the corresponding unsigned integer |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2708 | constant. TYPE must be a scalar or vector integer type. CST must be of |
| 2709 | scalar or vector floating point type. Both CST and TYPE must be scalars, |
| 2710 | or vectors of the same number of elements. If the value won't fit in the |
| 2711 | integer type, the results are undefined.</dd> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 2712 | |
| Dan Gohman | 3dfb3cf | 2010-05-28 17:07:41 +0000 | [diff] [blame] | 2713 | <dt><b><tt>fptosi (CST to TYPE)</tt></b></dt> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 2714 | <dd>Convert a floating point constant to the corresponding signed integer |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2715 | constant. TYPE must be a scalar or vector integer type. CST must be of |
| 2716 | scalar or vector floating point type. Both CST and TYPE must be scalars, |
| 2717 | or vectors of the same number of elements. If the value won't fit in the |
| 2718 | integer type, the results are undefined.</dd> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 2719 | |
| Dan Gohman | 3dfb3cf | 2010-05-28 17:07:41 +0000 | [diff] [blame] | 2720 | <dt><b><tt>uitofp (CST to TYPE)</tt></b></dt> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 2721 | <dd>Convert an unsigned integer constant to the corresponding floating point |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2722 | constant. TYPE must be a scalar or vector floating point type. CST must be |
| 2723 | of scalar or vector integer type. Both CST and TYPE must be scalars, or |
| 2724 | vectors of the same number of elements. If the value won't fit in the |
| 2725 | floating point type, the results are undefined.</dd> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 2726 | |
| Dan Gohman | 3dfb3cf | 2010-05-28 17:07:41 +0000 | [diff] [blame] | 2727 | <dt><b><tt>sitofp (CST to TYPE)</tt></b></dt> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 2728 | <dd>Convert a signed integer constant to the corresponding floating point |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2729 | constant. TYPE must be a scalar or vector floating point type. CST must be |
| 2730 | of scalar or vector integer type. Both CST and TYPE must be scalars, or |
| 2731 | vectors of the same number of elements. If the value won't fit in the |
| 2732 | floating point type, the results are undefined.</dd> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 2733 | |
| Dan Gohman | 3dfb3cf | 2010-05-28 17:07:41 +0000 | [diff] [blame] | 2734 | <dt><b><tt>ptrtoint (CST to TYPE)</tt></b></dt> |
| Reid Spencer | 5c0ef47 | 2006-11-11 23:08:07 +0000 | [diff] [blame] | 2735 | <dd>Convert a pointer typed constant to the corresponding integer constant |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2736 | <tt>TYPE</tt> must be an integer type. <tt>CST</tt> must be of pointer |
| 2737 | type. The <tt>CST</tt> value is zero extended, truncated, or unchanged to |
| 2738 | make it fit in <tt>TYPE</tt>.</dd> |
| Reid Spencer | 5c0ef47 | 2006-11-11 23:08:07 +0000 | [diff] [blame] | 2739 | |
| Dan Gohman | 3dfb3cf | 2010-05-28 17:07:41 +0000 | [diff] [blame] | 2740 | <dt><b><tt>inttoptr (CST to TYPE)</tt></b></dt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2741 | <dd>Convert a integer constant to a pointer constant. TYPE must be a pointer |
| 2742 | type. CST must be of integer type. The CST value is zero extended, |
| 2743 | truncated, or unchanged to make it fit in a pointer size. This one is |
| 2744 | <i>really</i> dangerous!</dd> |
| Reid Spencer | 5c0ef47 | 2006-11-11 23:08:07 +0000 | [diff] [blame] | 2745 | |
| Dan Gohman | 3dfb3cf | 2010-05-28 17:07:41 +0000 | [diff] [blame] | 2746 | <dt><b><tt>bitcast (CST to TYPE)</tt></b></dt> |
| Chris Lattner | 03bbad6 | 2009-02-28 18:27:03 +0000 | [diff] [blame] | 2747 | <dd>Convert a constant, CST, to another TYPE. The constraints of the operands |
| 2748 | are the same as those for the <a href="#i_bitcast">bitcast |
| 2749 | instruction</a>.</dd> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2750 | |
| Dan Gohman | 3dfb3cf | 2010-05-28 17:07:41 +0000 | [diff] [blame] | 2751 | <dt><b><tt>getelementptr (CSTPTR, IDX0, IDX1, ...)</tt></b></dt> |
| 2752 | <dt><b><tt>getelementptr inbounds (CSTPTR, IDX0, IDX1, ...)</tt></b></dt> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2753 | <dd>Perform the <a href="#i_getelementptr">getelementptr operation</a> on |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2754 | constants. As with the <a href="#i_getelementptr">getelementptr</a> |
| 2755 | instruction, the index list may have zero or more indexes, which are |
| 2756 | required to make sense for the type of "CSTPTR".</dd> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2757 | |
| Dan Gohman | 3dfb3cf | 2010-05-28 17:07:41 +0000 | [diff] [blame] | 2758 | <dt><b><tt>select (COND, VAL1, VAL2)</tt></b></dt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2759 | <dd>Perform the <a href="#i_select">select operation</a> on constants.</dd> |
| Reid Spencer | 01c4259 | 2006-12-04 19:23:19 +0000 | [diff] [blame] | 2760 | |
| Dan Gohman | 3dfb3cf | 2010-05-28 17:07:41 +0000 | [diff] [blame] | 2761 | <dt><b><tt>icmp COND (VAL1, VAL2)</tt></b></dt> |
| Reid Spencer | 01c4259 | 2006-12-04 19:23:19 +0000 | [diff] [blame] | 2762 | <dd>Performs the <a href="#i_icmp">icmp operation</a> on constants.</dd> |
| 2763 | |
| Dan Gohman | 3dfb3cf | 2010-05-28 17:07:41 +0000 | [diff] [blame] | 2764 | <dt><b><tt>fcmp COND (VAL1, VAL2)</tt></b></dt> |
| Reid Spencer | 01c4259 | 2006-12-04 19:23:19 +0000 | [diff] [blame] | 2765 | <dd>Performs the <a href="#i_fcmp">fcmp operation</a> on constants.</dd> |
| Robert Bocchino | 9fbe145 | 2006-01-10 19:31:34 +0000 | [diff] [blame] | 2766 | |
| Dan Gohman | 3dfb3cf | 2010-05-28 17:07:41 +0000 | [diff] [blame] | 2767 | <dt><b><tt>extractelement (VAL, IDX)</tt></b></dt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2768 | <dd>Perform the <a href="#i_extractelement">extractelement operation</a> on |
| 2769 | constants.</dd> |
| Robert Bocchino | 9fbe145 | 2006-01-10 19:31:34 +0000 | [diff] [blame] | 2770 | |
| Dan Gohman | 3dfb3cf | 2010-05-28 17:07:41 +0000 | [diff] [blame] | 2771 | <dt><b><tt>insertelement (VAL, ELT, IDX)</tt></b></dt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2772 | <dd>Perform the <a href="#i_insertelement">insertelement operation</a> on |
| 2773 | constants.</dd> |
| Chris Lattner | c198954 | 2006-04-08 00:13:41 +0000 | [diff] [blame] | 2774 | |
| Dan Gohman | 3dfb3cf | 2010-05-28 17:07:41 +0000 | [diff] [blame] | 2775 | <dt><b><tt>shufflevector (VEC1, VEC2, IDXMASK)</tt></b></dt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2776 | <dd>Perform the <a href="#i_shufflevector">shufflevector operation</a> on |
| 2777 | constants.</dd> |
| Chris Lattner | c198954 | 2006-04-08 00:13:41 +0000 | [diff] [blame] | 2778 | |
| Nick Lewycky | 9e130ce | 2010-05-29 06:44:15 +0000 | [diff] [blame] | 2779 | <dt><b><tt>extractvalue (VAL, IDX0, IDX1, ...)</tt></b></dt> |
| 2780 | <dd>Perform the <a href="#i_extractvalue">extractvalue operation</a> on |
| 2781 | constants. The index list is interpreted in a similar manner as indices in |
| 2782 | a '<a href="#i_getelementptr">getelementptr</a>' operation. At least one |
| 2783 | index value must be specified.</dd> |
| 2784 | |
| 2785 | <dt><b><tt>insertvalue (VAL, ELT, IDX0, IDX1, ...)</tt></b></dt> |
| 2786 | <dd>Perform the <a href="#i_insertvalue">insertvalue operation</a> on |
| 2787 | constants. The index list is interpreted in a similar manner as indices in |
| 2788 | a '<a href="#i_getelementptr">getelementptr</a>' operation. At least one |
| 2789 | index value must be specified.</dd> |
| 2790 | |
| Dan Gohman | 3dfb3cf | 2010-05-28 17:07:41 +0000 | [diff] [blame] | 2791 | <dt><b><tt>OPCODE (LHS, RHS)</tt></b></dt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2792 | <dd>Perform the specified operation of the LHS and RHS constants. OPCODE may |
| 2793 | be any of the <a href="#binaryops">binary</a> |
| 2794 | or <a href="#bitwiseops">bitwise binary</a> operations. The constraints |
| 2795 | on operands are the same as those for the corresponding instruction |
| 2796 | (e.g. no bitwise operations on floating point values are allowed).</dd> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2797 | </dl> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2798 | |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 2799 | </div> |
| Chris Lattner | 9ee5d22 | 2004-03-08 16:49:10 +0000 | [diff] [blame] | 2800 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 2801 | </div> |
| 2802 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 2803 | <!-- *********************************************************************** --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 2804 | <h2><a name="othervalues">Other Values</a></h2> |
| Chris Lattner | e87d653 | 2006-01-25 23:47:57 +0000 | [diff] [blame] | 2805 | <!-- *********************************************************************** --> |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 2806 | <div> |
| Chris Lattner | e87d653 | 2006-01-25 23:47:57 +0000 | [diff] [blame] | 2807 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 2808 | <h3> |
| Chris Lattner | e87d653 | 2006-01-25 23:47:57 +0000 | [diff] [blame] | 2809 | <a name="inlineasm">Inline Assembler Expressions</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 2810 | </h3> |
| Chris Lattner | e87d653 | 2006-01-25 23:47:57 +0000 | [diff] [blame] | 2811 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 2812 | <div> |
| Chris Lattner | e87d653 | 2006-01-25 23:47:57 +0000 | [diff] [blame] | 2813 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2814 | <p>LLVM supports inline assembler expressions (as opposed |
| Bill Wendling | aee0f45 | 2011-11-30 21:52:43 +0000 | [diff] [blame] | 2815 | to <a href="#moduleasm">Module-Level Inline Assembly</a>) through the use of |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2816 | a special value. This value represents the inline assembler as a string |
| 2817 | (containing the instructions to emit), a list of operand constraints (stored |
| Dale Johannesen | 09fed25 | 2009-10-13 21:56:55 +0000 | [diff] [blame] | 2818 | as a string), a flag that indicates whether or not the inline asm |
| Dale Johannesen | 8ba2d5b | 2009-10-21 23:28:00 +0000 | [diff] [blame] | 2819 | expression has side effects, and a flag indicating whether the function |
| 2820 | containing the asm needs to align its stack conservatively. An example |
| 2821 | inline assembler expression is:</p> |
| Chris Lattner | e87d653 | 2006-01-25 23:47:57 +0000 | [diff] [blame] | 2822 | |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 2823 | <pre class="doc_code"> |
| Bill Wendling | 2f7a8b0 | 2007-05-29 09:04:49 +0000 | [diff] [blame] | 2824 | i32 (i32) asm "bswap $0", "=r,r" |
| Chris Lattner | e87d653 | 2006-01-25 23:47:57 +0000 | [diff] [blame] | 2825 | </pre> |
| 2826 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2827 | <p>Inline assembler expressions may <b>only</b> be used as the callee operand of |
| 2828 | a <a href="#i_call"><tt>call</tt> instruction</a>. Thus, typically we |
| 2829 | have:</p> |
| Chris Lattner | e87d653 | 2006-01-25 23:47:57 +0000 | [diff] [blame] | 2830 | |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 2831 | <pre class="doc_code"> |
| Bill Wendling | 2f7a8b0 | 2007-05-29 09:04:49 +0000 | [diff] [blame] | 2832 | %X = call i32 asm "<a href="#int_bswap">bswap</a> $0", "=r,r"(i32 %Y) |
| Chris Lattner | e87d653 | 2006-01-25 23:47:57 +0000 | [diff] [blame] | 2833 | </pre> |
| 2834 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2835 | <p>Inline asms with side effects not visible in the constraint list must be |
| 2836 | marked as having side effects. This is done through the use of the |
| 2837 | '<tt>sideeffect</tt>' keyword, like so:</p> |
| Chris Lattner | e87d653 | 2006-01-25 23:47:57 +0000 | [diff] [blame] | 2838 | |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 2839 | <pre class="doc_code"> |
| Bill Wendling | 2f7a8b0 | 2007-05-29 09:04:49 +0000 | [diff] [blame] | 2840 | call void asm sideeffect "eieio", ""() |
| Chris Lattner | e87d653 | 2006-01-25 23:47:57 +0000 | [diff] [blame] | 2841 | </pre> |
| 2842 | |
| Dale Johannesen | 8ba2d5b | 2009-10-21 23:28:00 +0000 | [diff] [blame] | 2843 | <p>In some cases inline asms will contain code that will not work unless the |
| 2844 | stack is aligned in some way, such as calls or SSE instructions on x86, |
| 2845 | yet will not contain code that does that alignment within the asm. |
| 2846 | The compiler should make conservative assumptions about what the asm might |
| 2847 | contain and should generate its usual stack alignment code in the prologue |
| 2848 | if the '<tt>alignstack</tt>' keyword is present:</p> |
| Dale Johannesen | 09fed25 | 2009-10-13 21:56:55 +0000 | [diff] [blame] | 2849 | |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 2850 | <pre class="doc_code"> |
| Dale Johannesen | 8ba2d5b | 2009-10-21 23:28:00 +0000 | [diff] [blame] | 2851 | call void asm alignstack "eieio", ""() |
| Dale Johannesen | 09fed25 | 2009-10-13 21:56:55 +0000 | [diff] [blame] | 2852 | </pre> |
| Dale Johannesen | 09fed25 | 2009-10-13 21:56:55 +0000 | [diff] [blame] | 2853 | |
| 2854 | <p>If both keywords appear the '<tt>sideeffect</tt>' keyword must come |
| 2855 | first.</p> |
| 2856 | |
| Bill Wendling | aee0f45 | 2011-11-30 21:52:43 +0000 | [diff] [blame] | 2857 | <!-- |
| Chris Lattner | e87d653 | 2006-01-25 23:47:57 +0000 | [diff] [blame] | 2858 | <p>TODO: The format of the asm and constraints string still need to be |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 2859 | documented here. Constraints on what can be done (e.g. duplication, moving, |
| 2860 | etc need to be documented). This is probably best done by reference to |
| 2861 | another document that covers inline asm from a holistic perspective.</p> |
| Bill Wendling | aee0f45 | 2011-11-30 21:52:43 +0000 | [diff] [blame] | 2862 | --> |
| Chris Lattner | cf9a415 | 2010-04-07 05:38:05 +0000 | [diff] [blame] | 2863 | |
| Bill Wendling | aee0f45 | 2011-11-30 21:52:43 +0000 | [diff] [blame] | 2864 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 2865 | <h4> |
| Bill Wendling | aee0f45 | 2011-11-30 21:52:43 +0000 | [diff] [blame] | 2866 | <a name="inlineasm_md">Inline Asm Metadata</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 2867 | </h4> |
| Chris Lattner | cf9a415 | 2010-04-07 05:38:05 +0000 | [diff] [blame] | 2868 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 2869 | <div> |
| Chris Lattner | cf9a415 | 2010-04-07 05:38:05 +0000 | [diff] [blame] | 2870 | |
| Bill Wendling | aee0f45 | 2011-11-30 21:52:43 +0000 | [diff] [blame] | 2871 | <p>The call instructions that wrap inline asm nodes may have a |
| 2872 | "<tt>!srcloc</tt>" MDNode attached to it that contains a list of constant |
| 2873 | integers. If present, the code generator will use the integer as the |
| 2874 | location cookie value when report errors through the <tt>LLVMContext</tt> |
| 2875 | error reporting mechanisms. This allows a front-end to correlate backend |
| 2876 | errors that occur with inline asm back to the source code that produced it. |
| 2877 | For example:</p> |
| Chris Lattner | cf9a415 | 2010-04-07 05:38:05 +0000 | [diff] [blame] | 2878 | |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 2879 | <pre class="doc_code"> |
| Chris Lattner | cf9a415 | 2010-04-07 05:38:05 +0000 | [diff] [blame] | 2880 | call void asm sideeffect "something bad", ""()<b>, !srcloc !42</b> |
| 2881 | ... |
| 2882 | !42 = !{ i32 1234567 } |
| 2883 | </pre> |
| Chris Lattner | cf9a415 | 2010-04-07 05:38:05 +0000 | [diff] [blame] | 2884 | |
| 2885 | <p>It is up to the front-end to make sense of the magic numbers it places in the |
| Bill Wendling | aee0f45 | 2011-11-30 21:52:43 +0000 | [diff] [blame] | 2886 | IR. If the MDNode contains multiple constants, the code generator will use |
| Chris Lattner | ce1b9ad | 2010-11-17 08:20:42 +0000 | [diff] [blame] | 2887 | the one that corresponds to the line of the asm that the error occurs on.</p> |
| Chris Lattner | e87d653 | 2006-01-25 23:47:57 +0000 | [diff] [blame] | 2888 | |
| 2889 | </div> |
| 2890 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 2891 | </div> |
| 2892 | |
| Chris Lattner | e6a5ddd | 2010-01-15 21:50:19 +0000 | [diff] [blame] | 2893 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 2894 | <h3> |
| 2895 | <a name="metadata">Metadata Nodes and Metadata Strings</a> |
| 2896 | </h3> |
| Chris Lattner | e6a5ddd | 2010-01-15 21:50:19 +0000 | [diff] [blame] | 2897 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 2898 | <div> |
| Chris Lattner | e6a5ddd | 2010-01-15 21:50:19 +0000 | [diff] [blame] | 2899 | |
| 2900 | <p>LLVM IR allows metadata to be attached to instructions in the program that |
| 2901 | can convey extra information about the code to the optimizers and code |
| 2902 | generator. One example application of metadata is source-level debug |
| 2903 | information. There are two metadata primitives: strings and nodes. All |
| 2904 | metadata has the <tt>metadata</tt> type and is identified in syntax by a |
| 2905 | preceding exclamation point ('<tt>!</tt>').</p> |
| 2906 | |
| 2907 | <p>A metadata string is a string surrounded by double quotes. It can contain |
| Bill Wendling | f6cc4c2 | 2011-11-30 21:43:43 +0000 | [diff] [blame] | 2908 | any character by escaping non-printable characters with "<tt>\xx</tt>" where |
| 2909 | "<tt>xx</tt>" is the two digit hex code. For example: |
| 2910 | "<tt>!"test\00"</tt>".</p> |
| Chris Lattner | e6a5ddd | 2010-01-15 21:50:19 +0000 | [diff] [blame] | 2911 | |
| 2912 | <p>Metadata nodes are represented with notation similar to structure constants |
| 2913 | (a comma separated list of elements, surrounded by braces and preceded by an |
| Bill Wendling | f6cc4c2 | 2011-11-30 21:43:43 +0000 | [diff] [blame] | 2914 | exclamation point). Metadata nodes can have any values as their operand. For |
| 2915 | example:</p> |
| 2916 | |
| 2917 | <div class="doc_code"> |
| 2918 | <pre> |
| 2919 | !{ metadata !"test\00", i32 10} |
| 2920 | </pre> |
| 2921 | </div> |
| Chris Lattner | e6a5ddd | 2010-01-15 21:50:19 +0000 | [diff] [blame] | 2922 | |
| 2923 | <p>A <a href="#namedmetadatastructure">named metadata</a> is a collection of |
| 2924 | metadata nodes, which can be looked up in the module symbol table. For |
| Bill Wendling | f6cc4c2 | 2011-11-30 21:43:43 +0000 | [diff] [blame] | 2925 | example:</p> |
| 2926 | |
| 2927 | <div class="doc_code"> |
| 2928 | <pre> |
| 2929 | !foo = metadata !{!4, !3} |
| 2930 | </pre> |
| 2931 | </div> |
| Chris Lattner | e6a5ddd | 2010-01-15 21:50:19 +0000 | [diff] [blame] | 2932 | |
| Devang Patel | e1d50cd | 2010-03-04 23:44:48 +0000 | [diff] [blame] | 2933 | <p>Metadata can be used as function arguments. Here <tt>llvm.dbg.value</tt> |
| Bill Wendling | f6cc4c2 | 2011-11-30 21:43:43 +0000 | [diff] [blame] | 2934 | function is using two metadata arguments:</p> |
| Devang Patel | e1d50cd | 2010-03-04 23:44:48 +0000 | [diff] [blame] | 2935 | |
| Bill Wendling | 9ff5de9 | 2011-03-02 02:17:11 +0000 | [diff] [blame] | 2936 | <div class="doc_code"> |
| 2937 | <pre> |
| 2938 | call void @llvm.dbg.value(metadata !24, i64 0, metadata !25) |
| 2939 | </pre> |
| 2940 | </div> |
| Devang Patel | e1d50cd | 2010-03-04 23:44:48 +0000 | [diff] [blame] | 2941 | |
| 2942 | <p>Metadata can be attached with an instruction. Here metadata <tt>!21</tt> is |
| Bill Wendling | f6cc4c2 | 2011-11-30 21:43:43 +0000 | [diff] [blame] | 2943 | attached to the <tt>add</tt> instruction using the <tt>!dbg</tt> |
| 2944 | identifier:</p> |
| Devang Patel | e1d50cd | 2010-03-04 23:44:48 +0000 | [diff] [blame] | 2945 | |
| Bill Wendling | 9ff5de9 | 2011-03-02 02:17:11 +0000 | [diff] [blame] | 2946 | <div class="doc_code"> |
| 2947 | <pre> |
| 2948 | %indvar.next = add i64 %indvar, 1, !dbg !21 |
| 2949 | </pre> |
| 2950 | </div> |
| 2951 | |
| Peter Collingbourne | 249d953 | 2011-10-27 19:19:07 +0000 | [diff] [blame] | 2952 | <p>More information about specific metadata nodes recognized by the optimizers |
| 2953 | and code generator is found below.</p> |
| 2954 | |
| Bill Wendling | f6cc4c2 | 2011-11-30 21:43:43 +0000 | [diff] [blame] | 2955 | <!-- _______________________________________________________________________ --> |
| Peter Collingbourne | 249d953 | 2011-10-27 19:19:07 +0000 | [diff] [blame] | 2956 | <h4> |
| 2957 | <a name="tbaa">'<tt>tbaa</tt>' Metadata</a> |
| 2958 | </h4> |
| 2959 | |
| 2960 | <div> |
| 2961 | |
| 2962 | <p>In LLVM IR, memory does not have types, so LLVM's own type system is not |
| 2963 | suitable for doing TBAA. Instead, metadata is added to the IR to describe |
| 2964 | a type system of a higher level language. This can be used to implement |
| 2965 | typical C/C++ TBAA, but it can also be used to implement custom alias |
| 2966 | analysis behavior for other languages.</p> |
| 2967 | |
| 2968 | <p>The current metadata format is very simple. TBAA metadata nodes have up to |
| 2969 | three fields, e.g.:</p> |
| 2970 | |
| 2971 | <div class="doc_code"> |
| 2972 | <pre> |
| 2973 | !0 = metadata !{ metadata !"an example type tree" } |
| 2974 | !1 = metadata !{ metadata !"int", metadata !0 } |
| 2975 | !2 = metadata !{ metadata !"float", metadata !0 } |
| 2976 | !3 = metadata !{ metadata !"const float", metadata !2, i64 1 } |
| 2977 | </pre> |
| 2978 | </div> |
| 2979 | |
| 2980 | <p>The first field is an identity field. It can be any value, usually |
| 2981 | a metadata string, which uniquely identifies the type. The most important |
| 2982 | name in the tree is the name of the root node. Two trees with |
| 2983 | different root node names are entirely disjoint, even if they |
| 2984 | have leaves with common names.</p> |
| 2985 | |
| 2986 | <p>The second field identifies the type's parent node in the tree, or |
| 2987 | is null or omitted for a root node. A type is considered to alias |
| 2988 | all of its descendants and all of its ancestors in the tree. Also, |
| 2989 | a type is considered to alias all types in other trees, so that |
| 2990 | bitcode produced from multiple front-ends is handled conservatively.</p> |
| 2991 | |
| 2992 | <p>If the third field is present, it's an integer which if equal to 1 |
| 2993 | indicates that the type is "constant" (meaning |
| 2994 | <tt>pointsToConstantMemory</tt> should return true; see |
| 2995 | <a href="AliasAnalysis.html#OtherItfs">other useful |
| 2996 | <tt>AliasAnalysis</tt> methods</a>).</p> |
| 2997 | |
| 2998 | </div> |
| 2999 | |
| Bill Wendling | f6cc4c2 | 2011-11-30 21:43:43 +0000 | [diff] [blame] | 3000 | <!-- _______________________________________________________________________ --> |
| Peter Collingbourne | 999f90b | 2011-10-27 19:19:14 +0000 | [diff] [blame] | 3001 | <h4> |
| 3002 | <a name="fpaccuracy">'<tt>fpaccuracy</tt>' Metadata</a> |
| 3003 | </h4> |
| 3004 | |
| 3005 | <div> |
| 3006 | |
| 3007 | <p><tt>fpaccuracy</tt> metadata may be attached to any instruction of floating |
| 3008 | point type. It expresses the maximum relative error of the result of |
| 3009 | that instruction, in ULPs. ULP is defined as follows:</p> |
| 3010 | |
| Bill Wendling | 0656e25 | 2011-11-09 19:33:56 +0000 | [diff] [blame] | 3011 | <blockquote> |
| 3012 | |
| 3013 | <p>If <tt>x</tt> is a real number that lies between two finite consecutive |
| 3014 | floating-point numbers <tt>a</tt> and <tt>b</tt>, without being equal to one |
| 3015 | of them, then <tt>ulp(x) = |b - a|</tt>, otherwise <tt>ulp(x)</tt> is the |
| 3016 | distance between the two non-equal finite floating-point numbers nearest |
| 3017 | <tt>x</tt>. Moreover, <tt>ulp(NaN)</tt> is <tt>NaN</tt>.</p> |
| 3018 | |
| 3019 | </blockquote> |
| Peter Collingbourne | 999f90b | 2011-10-27 19:19:14 +0000 | [diff] [blame] | 3020 | |
| 3021 | <p>The maximum relative error may be any rational number. The metadata node |
| 3022 | shall consist of a pair of unsigned integers respectively representing |
| 3023 | the numerator and denominator. For example, 2.5 ULP:</p> |
| 3024 | |
| 3025 | <div class="doc_code"> |
| 3026 | <pre> |
| 3027 | !0 = metadata !{ i32 5, i32 2 } |
| 3028 | </pre> |
| 3029 | </div> |
| 3030 | |
| 3031 | </div> |
| 3032 | |
| Chris Lattner | e6a5ddd | 2010-01-15 21:50:19 +0000 | [diff] [blame] | 3033 | </div> |
| 3034 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 3035 | </div> |
| Chris Lattner | 857755c | 2009-07-20 05:55:19 +0000 | [diff] [blame] | 3036 | |
| 3037 | <!-- *********************************************************************** --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3038 | <h2> |
| Bill Wendling | b9d75a9 | 2012-02-11 11:59:36 +0000 | [diff] [blame] | 3039 | <a name="module_flags">Module Flags Metadata</a> |
| 3040 | </h2> |
| 3041 | <!-- *********************************************************************** --> |
| 3042 | |
| 3043 | <div> |
| 3044 | |
| 3045 | <p>Information about the module as a whole is difficult to convey to LLVM's |
| 3046 | subsystems. The LLVM IR isn't sufficient to transmit this |
| 3047 | information. The <tt>llvm.module.flags</tt> named metadata exists in order to |
| 3048 | facilitate this. These flags are in the form of key / value pairs — |
| 3049 | much like a dictionary — making it easy for any subsystem who cares |
| 3050 | about a flag to look it up.</p> |
| 3051 | |
| 3052 | <p>The <tt>llvm.module.flags</tt> metadata contains a list of metadata |
| 3053 | triplets. Each triplet has the following form:</p> |
| 3054 | |
| 3055 | <ul> |
| 3056 | <li>The first element is a <i>behavior</i> flag, which specifies the behavior |
| 3057 | when two (or more) modules are merged together, and it encounters two (or |
| 3058 | more) metadata with the same ID. The supported behaviors are described |
| 3059 | below.</li> |
| 3060 | |
| 3061 | <li>The second element is a metadata string that is a unique ID for the |
| 3062 | metadata. How each ID is interpreted is documented below.</li> |
| 3063 | |
| 3064 | <li>The third element is the value of the flag.</li> |
| 3065 | </ul> |
| 3066 | |
| 3067 | <p>When two (or more) modules are merged together, the resulting |
| 3068 | <tt>llvm.module.flags</tt> metadata is the union of the |
| 3069 | modules' <tt>llvm.module.flags</tt> metadata. The only exception being a flag |
| 3070 | with the <i>Override</i> behavior, which may override another flag's value |
| 3071 | (see below).</p> |
| 3072 | |
| 3073 | <p>The following behaviors are supported:</p> |
| 3074 | |
| 3075 | <table border="1" cellspacing="0" cellpadding="4"> |
| 3076 | <tbody> |
| 3077 | <tr> |
| 3078 | <th>Value</th> |
| 3079 | <th>Behavior</th> |
| 3080 | </tr> |
| 3081 | <tr> |
| 3082 | <td>1</td> |
| 3083 | <td align="left"> |
| Bill Wendling | a0edecf | 2012-03-06 09:17:04 +0000 | [diff] [blame] | 3084 | <dl> |
| 3085 | <dt><b>Error</b></dt> |
| 3086 | <dd>Emits an error if two values disagree. It is an error to have an ID |
| 3087 | with both an Error and a Warning behavior.</dd> |
| 3088 | </dl> |
| Bill Wendling | b9d75a9 | 2012-02-11 11:59:36 +0000 | [diff] [blame] | 3089 | </td> |
| 3090 | </tr> |
| 3091 | <tr> |
| 3092 | <td>2</td> |
| 3093 | <td align="left"> |
| Bill Wendling | a0edecf | 2012-03-06 09:17:04 +0000 | [diff] [blame] | 3094 | <dl> |
| 3095 | <dt><b>Warning</b></dt> |
| 3096 | <dd>Emits a warning if two values disagree.</dd> |
| 3097 | </dl> |
| Bill Wendling | b9d75a9 | 2012-02-11 11:59:36 +0000 | [diff] [blame] | 3098 | </td> |
| 3099 | </tr> |
| 3100 | <tr> |
| 3101 | <td>3</td> |
| 3102 | <td align="left"> |
| Bill Wendling | a0edecf | 2012-03-06 09:17:04 +0000 | [diff] [blame] | 3103 | <dl> |
| 3104 | <dt><b>Require</b></dt> |
| 3105 | <dd>Emits an error when the specified value is not present or doesn't |
| 3106 | have the specified value. It is an error for two (or more) |
| 3107 | <tt>llvm.module.flags</tt> with the same ID to have the Require |
| 3108 | behavior but different values. There may be multiple Require flags |
| 3109 | per ID.</dd> |
| 3110 | </dl> |
| Bill Wendling | b9d75a9 | 2012-02-11 11:59:36 +0000 | [diff] [blame] | 3111 | </td> |
| 3112 | </tr> |
| 3113 | <tr> |
| 3114 | <td>4</td> |
| 3115 | <td align="left"> |
| Bill Wendling | a0edecf | 2012-03-06 09:17:04 +0000 | [diff] [blame] | 3116 | <dl> |
| 3117 | <dt><b>Override</b></dt> |
| 3118 | <dd>Uses the specified value if the two values disagree. It is an |
| 3119 | error for two (or more) <tt>llvm.module.flags</tt> with the same |
| 3120 | ID to have the Override behavior but different values.</dd> |
| 3121 | </dl> |
| Bill Wendling | b9d75a9 | 2012-02-11 11:59:36 +0000 | [diff] [blame] | 3122 | </td> |
| 3123 | </tr> |
| 3124 | </tbody> |
| 3125 | </table> |
| 3126 | |
| 3127 | <p>An example of module flags:</p> |
| 3128 | |
| 3129 | <pre class="doc_code"> |
| 3130 | !0 = metadata !{ i32 1, metadata !"foo", i32 1 } |
| 3131 | !1 = metadata !{ i32 4, metadata !"bar", i32 37 } |
| 3132 | !2 = metadata !{ i32 2, metadata !"qux", i32 42 } |
| 3133 | !3 = metadata !{ i32 3, metadata !"qux", |
| 3134 | metadata !{ |
| 3135 | metadata !"foo", i32 1 |
| 3136 | } |
| 3137 | } |
| 3138 | !llvm.module.flags = !{ !0, !1, !2, !3 } |
| 3139 | </pre> |
| 3140 | |
| 3141 | <ul> |
| 3142 | <li><p>Metadata <tt>!0</tt> has the ID <tt>!"foo"</tt> and the value '1'. The |
| 3143 | behavior if two or more <tt>!"foo"</tt> flags are seen is to emit an |
| 3144 | error if their values are not equal.</p></li> |
| 3145 | |
| 3146 | <li><p>Metadata <tt>!1</tt> has the ID <tt>!"bar"</tt> and the value '37'. The |
| 3147 | behavior if two or more <tt>!"bar"</tt> flags are seen is to use the |
| 3148 | value '37' if their values are not equal.</p></li> |
| 3149 | |
| 3150 | <li><p>Metadata <tt>!2</tt> has the ID <tt>!"qux"</tt> and the value '42'. The |
| 3151 | behavior if two or more <tt>!"qux"</tt> flags are seen is to emit a |
| 3152 | warning if their values are not equal.</p></li> |
| 3153 | |
| 3154 | <li><p>Metadata <tt>!3</tt> has the ID <tt>!"qux"</tt> and the value:</p> |
| 3155 | |
| 3156 | <pre class="doc_code"> |
| 3157 | metadata !{ metadata !"foo", i32 1 } |
| 3158 | </pre> |
| Bill Wendling | f7b367c | 2012-02-16 01:10:50 +0000 | [diff] [blame] | 3159 | |
| Bill Wendling | b9d75a9 | 2012-02-11 11:59:36 +0000 | [diff] [blame] | 3160 | <p>The behavior is to emit an error if the <tt>llvm.module.flags</tt> does |
| 3161 | not contain a flag with the ID <tt>!"foo"</tt> that has the value |
| 3162 | '1'. If two or more <tt>!"qux"</tt> flags exist, then they must have |
| 3163 | the same value or an error will be issued.</p></li> |
| 3164 | </ul> |
| 3165 | |
| Bill Wendling | f7b367c | 2012-02-16 01:10:50 +0000 | [diff] [blame] | 3166 | |
| 3167 | <!-- ======================================================================= --> |
| 3168 | <h3> |
| 3169 | <a name="objc_gc_flags">Objective-C Garbage Collection Module Flags Metadata</a> |
| 3170 | </h3> |
| 3171 | |
| 3172 | <div> |
| 3173 | |
| 3174 | <p>On the Mach-O platform, Objective-C stores metadata about garbage collection |
| 3175 | in a special section called "image info". The metadata consists of a version |
| 3176 | number and a bitmask specifying what types of garbage collection are |
| 3177 | supported (if any) by the file. If two or more modules are linked together |
| 3178 | their garbage collection metadata needs to be merged rather than appended |
| 3179 | together.</p> |
| 3180 | |
| 3181 | <p>The Objective-C garbage collection module flags metadata consists of the |
| 3182 | following key-value pairs:</p> |
| 3183 | |
| 3184 | <table border="1" cellspacing="0" cellpadding="4"> |
| Bill Wendling | b3ef223 | 2012-03-06 09:23:25 +0000 | [diff] [blame] | 3185 | <col width="30%"> |
| Bill Wendling | f7b367c | 2012-02-16 01:10:50 +0000 | [diff] [blame] | 3186 | <tbody> |
| 3187 | <tr> |
| Bill Wendling | a0edecf | 2012-03-06 09:17:04 +0000 | [diff] [blame] | 3188 | <th>Key</th> |
| Bill Wendling | f7b367c | 2012-02-16 01:10:50 +0000 | [diff] [blame] | 3189 | <th>Value</th> |
| 3190 | </tr> |
| 3191 | <tr> |
| 3192 | <td><tt>Objective-C Version</tt></td> |
| 3193 | <td align="left"><b>[Required]</b> — The Objective-C ABI |
| 3194 | version. Valid values are 1 and 2.</td> |
| 3195 | </tr> |
| 3196 | <tr> |
| 3197 | <td><tt>Objective-C Image Info Version</tt></td> |
| 3198 | <td align="left"><b>[Required]</b> — The version of the image info |
| 3199 | section. Currently always 0.</td> |
| 3200 | </tr> |
| 3201 | <tr> |
| 3202 | <td><tt>Objective-C Image Info Section</tt></td> |
| 3203 | <td align="left"><b>[Required]</b> — The section to place the |
| 3204 | metadata. Valid values are <tt>"__OBJC, __image_info, regular"</tt> for |
| 3205 | Objective-C ABI version 1, and <tt>"__DATA,__objc_imageinfo, regular, |
| 3206 | no_dead_strip"</tt> for Objective-C ABI version 2.</td> |
| 3207 | </tr> |
| 3208 | <tr> |
| 3209 | <td><tt>Objective-C Garbage Collection</tt></td> |
| 3210 | <td align="left"><b>[Required]</b> — Specifies whether garbage |
| 3211 | collection is supported or not. Valid values are 0, for no garbage |
| 3212 | collection, and 2, for garbage collection supported.</td> |
| 3213 | </tr> |
| 3214 | <tr> |
| 3215 | <td><tt>Objective-C GC Only</tt></td> |
| 3216 | <td align="left"><b>[Optional]</b> — Specifies that only garbage |
| 3217 | collection is supported. If present, its value must be 6. This flag |
| 3218 | requires that the <tt>Objective-C Garbage Collection</tt> flag have the |
| 3219 | value 2.</td> |
| 3220 | </tr> |
| 3221 | </tbody> |
| 3222 | </table> |
| 3223 | |
| 3224 | <p>Some important flag interactions:</p> |
| 3225 | |
| 3226 | <ul> |
| 3227 | <li>If a module with <tt>Objective-C Garbage Collection</tt> set to 0 is |
| 3228 | merged with a module with <tt>Objective-C Garbage Collection</tt> set to |
| 3229 | 2, then the resulting module has the <tt>Objective-C Garbage |
| 3230 | Collection</tt> flag set to 0.</li> |
| 3231 | |
| 3232 | <li>A module with <tt>Objective-C Garbage Collection</tt> set to 0 cannot be |
| 3233 | merged with a module with <tt>Objective-C GC Only</tt> set to 6.</li> |
| 3234 | </ul> |
| 3235 | |
| 3236 | </div> |
| 3237 | |
| Bill Wendling | b9d75a9 | 2012-02-11 11:59:36 +0000 | [diff] [blame] | 3238 | </div> |
| 3239 | |
| 3240 | <!-- *********************************************************************** --> |
| 3241 | <h2> |
| Chris Lattner | 857755c | 2009-07-20 05:55:19 +0000 | [diff] [blame] | 3242 | <a name="intrinsic_globals">Intrinsic Global Variables</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3243 | </h2> |
| Chris Lattner | 857755c | 2009-07-20 05:55:19 +0000 | [diff] [blame] | 3244 | <!-- *********************************************************************** --> |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 3245 | <div> |
| Chris Lattner | 857755c | 2009-07-20 05:55:19 +0000 | [diff] [blame] | 3246 | <p>LLVM has a number of "magic" global variables that contain data that affect |
| 3247 | code generation or other IR semantics. These are documented here. All globals |
| Chris Lattner | 401e10c | 2009-07-20 06:14:25 +0000 | [diff] [blame] | 3248 | of this sort should have a section specified as "<tt>llvm.metadata</tt>". This |
| 3249 | section and all globals that start with "<tt>llvm.</tt>" are reserved for use |
| 3250 | by LLVM.</p> |
| Chris Lattner | 857755c | 2009-07-20 05:55:19 +0000 | [diff] [blame] | 3251 | |
| 3252 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3253 | <h3> |
| Chris Lattner | 857755c | 2009-07-20 05:55:19 +0000 | [diff] [blame] | 3254 | <a name="intg_used">The '<tt>llvm.used</tt>' Global Variable</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3255 | </h3> |
| Chris Lattner | 857755c | 2009-07-20 05:55:19 +0000 | [diff] [blame] | 3256 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 3257 | <div> |
| Chris Lattner | 857755c | 2009-07-20 05:55:19 +0000 | [diff] [blame] | 3258 | |
| 3259 | <p>The <tt>@llvm.used</tt> global is an array with i8* element type which has <a |
| 3260 | href="#linkage_appending">appending linkage</a>. This array contains a list of |
| 3261 | pointers to global variables and functions which may optionally have a pointer |
| 3262 | cast formed of bitcast or getelementptr. For example, a legal use of it is:</p> |
| 3263 | |
| Bill Wendling | 9ae7563 | 2011-11-08 00:32:45 +0000 | [diff] [blame] | 3264 | <div class="doc_code"> |
| Chris Lattner | 857755c | 2009-07-20 05:55:19 +0000 | [diff] [blame] | 3265 | <pre> |
| Bill Wendling | 9ae7563 | 2011-11-08 00:32:45 +0000 | [diff] [blame] | 3266 | @X = global i8 4 |
| 3267 | @Y = global i32 123 |
| Chris Lattner | 857755c | 2009-07-20 05:55:19 +0000 | [diff] [blame] | 3268 | |
| Bill Wendling | 9ae7563 | 2011-11-08 00:32:45 +0000 | [diff] [blame] | 3269 | @llvm.used = appending global [2 x i8*] [ |
| 3270 | i8* @X, |
| 3271 | i8* bitcast (i32* @Y to i8*) |
| 3272 | ], section "llvm.metadata" |
| Chris Lattner | 857755c | 2009-07-20 05:55:19 +0000 | [diff] [blame] | 3273 | </pre> |
| Bill Wendling | 9ae7563 | 2011-11-08 00:32:45 +0000 | [diff] [blame] | 3274 | </div> |
| Chris Lattner | 857755c | 2009-07-20 05:55:19 +0000 | [diff] [blame] | 3275 | |
| 3276 | <p>If a global variable appears in the <tt>@llvm.used</tt> list, then the |
| Bill Wendling | 9ae7563 | 2011-11-08 00:32:45 +0000 | [diff] [blame] | 3277 | compiler, assembler, and linker are required to treat the symbol as if there |
| 3278 | is a reference to the global that it cannot see. For example, if a variable |
| 3279 | has internal linkage and no references other than that from |
| 3280 | the <tt>@llvm.used</tt> list, it cannot be deleted. This is commonly used to |
| 3281 | represent references from inline asms and other things the compiler cannot |
| 3282 | "see", and corresponds to "<tt>attribute((used))</tt>" in GNU C.</p> |
| Chris Lattner | 857755c | 2009-07-20 05:55:19 +0000 | [diff] [blame] | 3283 | |
| 3284 | <p>On some targets, the code generator must emit a directive to the assembler or |
| Bill Wendling | 9ae7563 | 2011-11-08 00:32:45 +0000 | [diff] [blame] | 3285 | object file to prevent the assembler and linker from molesting the |
| 3286 | symbol.</p> |
| Chris Lattner | 857755c | 2009-07-20 05:55:19 +0000 | [diff] [blame] | 3287 | |
| 3288 | </div> |
| 3289 | |
| 3290 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3291 | <h3> |
| 3292 | <a name="intg_compiler_used"> |
| 3293 | The '<tt>llvm.compiler.used</tt>' Global Variable |
| 3294 | </a> |
| 3295 | </h3> |
| Chris Lattner | 401e10c | 2009-07-20 06:14:25 +0000 | [diff] [blame] | 3296 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 3297 | <div> |
| Chris Lattner | 401e10c | 2009-07-20 06:14:25 +0000 | [diff] [blame] | 3298 | |
| 3299 | <p>The <tt>@llvm.compiler.used</tt> directive is the same as the |
| Bill Wendling | 9ae7563 | 2011-11-08 00:32:45 +0000 | [diff] [blame] | 3300 | <tt>@llvm.used</tt> directive, except that it only prevents the compiler from |
| 3301 | touching the symbol. On targets that support it, this allows an intelligent |
| 3302 | linker to optimize references to the symbol without being impeded as it would |
| 3303 | be by <tt>@llvm.used</tt>.</p> |
| Chris Lattner | 401e10c | 2009-07-20 06:14:25 +0000 | [diff] [blame] | 3304 | |
| 3305 | <p>This is a rare construct that should only be used in rare circumstances, and |
| Bill Wendling | 9ae7563 | 2011-11-08 00:32:45 +0000 | [diff] [blame] | 3306 | should not be exposed to source languages.</p> |
| Chris Lattner | 401e10c | 2009-07-20 06:14:25 +0000 | [diff] [blame] | 3307 | |
| 3308 | </div> |
| 3309 | |
| 3310 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3311 | <h3> |
| Chris Lattner | 857755c | 2009-07-20 05:55:19 +0000 | [diff] [blame] | 3312 | <a name="intg_global_ctors">The '<tt>llvm.global_ctors</tt>' Global Variable</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3313 | </h3> |
| Chris Lattner | 857755c | 2009-07-20 05:55:19 +0000 | [diff] [blame] | 3314 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 3315 | <div> |
| Bill Wendling | 9ae7563 | 2011-11-08 00:32:45 +0000 | [diff] [blame] | 3316 | |
| 3317 | <div class="doc_code"> |
| David Chisnall | e31e996 | 2010-04-30 19:23:49 +0000 | [diff] [blame] | 3318 | <pre> |
| 3319 | %0 = type { i32, void ()* } |
| David Chisnall | 27195a5 | 2010-04-30 19:27:35 +0000 | [diff] [blame] | 3320 | @llvm.global_ctors = appending global [1 x %0] [%0 { i32 65535, void ()* @ctor }] |
| David Chisnall | e31e996 | 2010-04-30 19:23:49 +0000 | [diff] [blame] | 3321 | </pre> |
| Bill Wendling | 9ae7563 | 2011-11-08 00:32:45 +0000 | [diff] [blame] | 3322 | </div> |
| 3323 | |
| 3324 | <p>The <tt>@llvm.global_ctors</tt> array contains a list of constructor |
| 3325 | functions and associated priorities. The functions referenced by this array |
| 3326 | will be called in ascending order of priority (i.e. lowest first) when the |
| 3327 | module is loaded. The order of functions with the same priority is not |
| 3328 | defined.</p> |
| Chris Lattner | 857755c | 2009-07-20 05:55:19 +0000 | [diff] [blame] | 3329 | |
| 3330 | </div> |
| 3331 | |
| 3332 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3333 | <h3> |
| Chris Lattner | 857755c | 2009-07-20 05:55:19 +0000 | [diff] [blame] | 3334 | <a name="intg_global_dtors">The '<tt>llvm.global_dtors</tt>' Global Variable</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3335 | </h3> |
| Chris Lattner | 857755c | 2009-07-20 05:55:19 +0000 | [diff] [blame] | 3336 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 3337 | <div> |
| Bill Wendling | 9ae7563 | 2011-11-08 00:32:45 +0000 | [diff] [blame] | 3338 | |
| 3339 | <div class="doc_code"> |
| David Chisnall | e31e996 | 2010-04-30 19:23:49 +0000 | [diff] [blame] | 3340 | <pre> |
| 3341 | %0 = type { i32, void ()* } |
| David Chisnall | 27195a5 | 2010-04-30 19:27:35 +0000 | [diff] [blame] | 3342 | @llvm.global_dtors = appending global [1 x %0] [%0 { i32 65535, void ()* @dtor }] |
| David Chisnall | e31e996 | 2010-04-30 19:23:49 +0000 | [diff] [blame] | 3343 | </pre> |
| Bill Wendling | 9ae7563 | 2011-11-08 00:32:45 +0000 | [diff] [blame] | 3344 | </div> |
| Chris Lattner | 857755c | 2009-07-20 05:55:19 +0000 | [diff] [blame] | 3345 | |
| Bill Wendling | 9ae7563 | 2011-11-08 00:32:45 +0000 | [diff] [blame] | 3346 | <p>The <tt>@llvm.global_dtors</tt> array contains a list of destructor functions |
| 3347 | and associated priorities. The functions referenced by this array will be |
| 3348 | called in descending order of priority (i.e. highest first) when the module |
| 3349 | is loaded. The order of functions with the same priority is not defined.</p> |
| Chris Lattner | 857755c | 2009-07-20 05:55:19 +0000 | [diff] [blame] | 3350 | |
| 3351 | </div> |
| 3352 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 3353 | </div> |
| Chris Lattner | 857755c | 2009-07-20 05:55:19 +0000 | [diff] [blame] | 3354 | |
| Chris Lattner | e87d653 | 2006-01-25 23:47:57 +0000 | [diff] [blame] | 3355 | <!-- *********************************************************************** --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3356 | <h2><a name="instref">Instruction Reference</a></h2> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 3357 | <!-- *********************************************************************** --> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 3358 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 3359 | <div> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 3360 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3361 | <p>The LLVM instruction set consists of several different classifications of |
| 3362 | instructions: <a href="#terminators">terminator |
| 3363 | instructions</a>, <a href="#binaryops">binary instructions</a>, |
| 3364 | <a href="#bitwiseops">bitwise binary instructions</a>, |
| 3365 | <a href="#memoryops">memory instructions</a>, and |
| 3366 | <a href="#otherops">other instructions</a>.</p> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 3367 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3368 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3369 | <h3> |
| 3370 | <a name="terminators">Terminator Instructions</a> |
| 3371 | </h3> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 3372 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 3373 | <div> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 3374 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3375 | <p>As mentioned <a href="#functionstructure">previously</a>, every basic block |
| 3376 | in a program ends with a "Terminator" instruction, which indicates which |
| 3377 | block should be executed after the current block is finished. These |
| 3378 | terminator instructions typically yield a '<tt>void</tt>' value: they produce |
| 3379 | control flow, not values (the one exception being the |
| 3380 | '<a href="#i_invoke"><tt>invoke</tt></a>' instruction).</p> |
| 3381 | |
| Chris Lattner | 6445ecb | 2011-08-02 20:29:13 +0000 | [diff] [blame] | 3382 | <p>The terminator instructions are: |
| 3383 | '<a href="#i_ret"><tt>ret</tt></a>', |
| 3384 | '<a href="#i_br"><tt>br</tt></a>', |
| 3385 | '<a href="#i_switch"><tt>switch</tt></a>', |
| 3386 | '<a href="#i_indirectbr"><tt>indirectbr</tt></a>', |
| 3387 | '<a href="#i_invoke"><tt>invoke</tt></a>', |
| Chris Lattner | 6445ecb | 2011-08-02 20:29:13 +0000 | [diff] [blame] | 3388 | '<a href="#i_resume"><tt>resume</tt></a>', and |
| 3389 | '<a href="#i_unreachable"><tt>unreachable</tt></a>'.</p> |
| Chris Lattner | c3f5976 | 2004-12-09 17:30:23 +0000 | [diff] [blame] | 3390 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3391 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3392 | <h4> |
| 3393 | <a name="i_ret">'<tt>ret</tt>' Instruction</a> |
| 3394 | </h4> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3395 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 3396 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3397 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3398 | <h5>Syntax:</h5> |
| Dan Gohman | b1e6b96 | 2008-10-04 19:00:07 +0000 | [diff] [blame] | 3399 | <pre> |
| 3400 | ret <type> <value> <i>; Return a value from a non-void function</i> |
| Chris Lattner | 7faa883 | 2002-04-14 06:13:44 +0000 | [diff] [blame] | 3401 | ret void <i>; Return from void function</i> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3402 | </pre> |
| Chris Lattner | f4cde4e | 2008-04-23 04:59:35 +0000 | [diff] [blame] | 3403 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3404 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3405 | <p>The '<tt>ret</tt>' instruction is used to return control flow (and optionally |
| 3406 | a value) from a function back to the caller.</p> |
| 3407 | |
| 3408 | <p>There are two forms of the '<tt>ret</tt>' instruction: one that returns a |
| 3409 | value and then causes control flow, and one that just causes control flow to |
| 3410 | occur.</p> |
| Chris Lattner | f4cde4e | 2008-04-23 04:59:35 +0000 | [diff] [blame] | 3411 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3412 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3413 | <p>The '<tt>ret</tt>' instruction optionally accepts a single argument, the |
| 3414 | return value. The type of the return value must be a |
| 3415 | '<a href="#t_firstclass">first class</a>' type.</p> |
| Dan Gohman | b1e6b96 | 2008-10-04 19:00:07 +0000 | [diff] [blame] | 3416 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3417 | <p>A function is not <a href="#wellformed">well formed</a> if it it has a |
| 3418 | non-void return type and contains a '<tt>ret</tt>' instruction with no return |
| 3419 | value or a return value with a type that does not match its type, or if it |
| 3420 | has a void return type and contains a '<tt>ret</tt>' instruction with a |
| 3421 | return value.</p> |
| Chris Lattner | f4cde4e | 2008-04-23 04:59:35 +0000 | [diff] [blame] | 3422 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3423 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3424 | <p>When the '<tt>ret</tt>' instruction is executed, control flow returns back to |
| 3425 | the calling function's context. If the caller is a |
| 3426 | "<a href="#i_call"><tt>call</tt></a>" instruction, execution continues at the |
| 3427 | instruction after the call. If the caller was an |
| 3428 | "<a href="#i_invoke"><tt>invoke</tt></a>" instruction, execution continues at |
| 3429 | the beginning of the "normal" destination block. If the instruction returns |
| 3430 | a value, that value shall set the call or invoke instruction's return |
| 3431 | value.</p> |
| Chris Lattner | f4cde4e | 2008-04-23 04:59:35 +0000 | [diff] [blame] | 3432 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3433 | <h5>Example:</h5> |
| Chris Lattner | f4cde4e | 2008-04-23 04:59:35 +0000 | [diff] [blame] | 3434 | <pre> |
| 3435 | ret i32 5 <i>; Return an integer value of 5</i> |
| Chris Lattner | 7faa883 | 2002-04-14 06:13:44 +0000 | [diff] [blame] | 3436 | ret void <i>; Return from a void function</i> |
| Bill Wendling | 0a4bbbf | 2009-02-28 22:12:54 +0000 | [diff] [blame] | 3437 | ret { i32, i8 } { i32 4, i8 2 } <i>; Return a struct of values 4 and 2</i> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3438 | </pre> |
| Dan Gohman | f3e60bd | 2009-01-12 23:12:39 +0000 | [diff] [blame] | 3439 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 3440 | </div> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3441 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3442 | <h4> |
| 3443 | <a name="i_br">'<tt>br</tt>' Instruction</a> |
| 3444 | </h4> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3445 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 3446 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3447 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3448 | <h5>Syntax:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3449 | <pre> |
| Bill Wendling | b3aa471 | 2011-07-26 10:41:15 +0000 | [diff] [blame] | 3450 | br i1 <cond>, label <iftrue>, label <iffalse> |
| 3451 | br label <dest> <i>; Unconditional branch</i> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3452 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3453 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3454 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3455 | <p>The '<tt>br</tt>' instruction is used to cause control flow to transfer to a |
| 3456 | different basic block in the current function. There are two forms of this |
| 3457 | instruction, corresponding to a conditional branch and an unconditional |
| 3458 | branch.</p> |
| 3459 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3460 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3461 | <p>The conditional branch form of the '<tt>br</tt>' instruction takes a single |
| 3462 | '<tt>i1</tt>' value and two '<tt>label</tt>' values. The unconditional form |
| 3463 | of the '<tt>br</tt>' instruction takes a single '<tt>label</tt>' value as a |
| 3464 | target.</p> |
| 3465 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3466 | <h5>Semantics:</h5> |
| Reid Spencer | c78f337 | 2007-01-12 03:35:51 +0000 | [diff] [blame] | 3467 | <p>Upon execution of a conditional '<tt>br</tt>' instruction, the '<tt>i1</tt>' |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3468 | argument is evaluated. If the value is <tt>true</tt>, control flows to the |
| 3469 | '<tt>iftrue</tt>' <tt>label</tt> argument. If "cond" is <tt>false</tt>, |
| 3470 | control flows to the '<tt>iffalse</tt>' <tt>label</tt> argument.</p> |
| 3471 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3472 | <h5>Example:</h5> |
| Bill Wendling | c39e3e0 | 2009-07-20 02:39:26 +0000 | [diff] [blame] | 3473 | <pre> |
| 3474 | Test: |
| 3475 | %cond = <a href="#i_icmp">icmp</a> eq i32 %a, %b |
| 3476 | br i1 %cond, label %IfEqual, label %IfUnequal |
| 3477 | IfEqual: |
| 3478 | <a href="#i_ret">ret</a> i32 1 |
| 3479 | IfUnequal: |
| 3480 | <a href="#i_ret">ret</a> i32 0 |
| 3481 | </pre> |
| 3482 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 3483 | </div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3484 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3485 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3486 | <h4> |
| Chris Lattner | c88c17b | 2004-02-24 04:54:45 +0000 | [diff] [blame] | 3487 | <a name="i_switch">'<tt>switch</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3488 | </h4> |
| Chris Lattner | c88c17b | 2004-02-24 04:54:45 +0000 | [diff] [blame] | 3489 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 3490 | <div> |
| Chris Lattner | c88c17b | 2004-02-24 04:54:45 +0000 | [diff] [blame] | 3491 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3492 | <h5>Syntax:</h5> |
| Chris Lattner | c88c17b | 2004-02-24 04:54:45 +0000 | [diff] [blame] | 3493 | <pre> |
| 3494 | switch <intty> <value>, label <defaultdest> [ <intty> <val>, label <dest> ... ] |
| 3495 | </pre> |
| 3496 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3497 | <h5>Overview:</h5> |
| Chris Lattner | c88c17b | 2004-02-24 04:54:45 +0000 | [diff] [blame] | 3498 | <p>The '<tt>switch</tt>' instruction is used to transfer control flow to one of |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3499 | several different places. It is a generalization of the '<tt>br</tt>' |
| 3500 | instruction, allowing a branch to occur to one of many possible |
| 3501 | destinations.</p> |
| Chris Lattner | c88c17b | 2004-02-24 04:54:45 +0000 | [diff] [blame] | 3502 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3503 | <h5>Arguments:</h5> |
| Chris Lattner | c88c17b | 2004-02-24 04:54:45 +0000 | [diff] [blame] | 3504 | <p>The '<tt>switch</tt>' instruction uses three parameters: an integer |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3505 | comparison value '<tt>value</tt>', a default '<tt>label</tt>' destination, |
| 3506 | and an array of pairs of comparison value constants and '<tt>label</tt>'s. |
| 3507 | The table is not allowed to contain duplicate constant entries.</p> |
| Chris Lattner | c88c17b | 2004-02-24 04:54:45 +0000 | [diff] [blame] | 3508 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3509 | <h5>Semantics:</h5> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 3510 | <p>The <tt>switch</tt> instruction specifies a table of values and |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3511 | destinations. When the '<tt>switch</tt>' instruction is executed, this table |
| 3512 | is searched for the given value. If the value is found, control flow is |
| Benjamin Kramer | 8040cd3 | 2009-10-12 14:46:08 +0000 | [diff] [blame] | 3513 | transferred to the corresponding destination; otherwise, control flow is |
| 3514 | transferred to the default destination.</p> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3515 | |
| Chris Lattner | c88c17b | 2004-02-24 04:54:45 +0000 | [diff] [blame] | 3516 | <h5>Implementation:</h5> |
| Chris Lattner | c88c17b | 2004-02-24 04:54:45 +0000 | [diff] [blame] | 3517 | <p>Depending on properties of the target machine and the particular |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3518 | <tt>switch</tt> instruction, this instruction may be code generated in |
| 3519 | different ways. For example, it could be generated as a series of chained |
| 3520 | conditional branches or with a lookup table.</p> |
| Chris Lattner | c88c17b | 2004-02-24 04:54:45 +0000 | [diff] [blame] | 3521 | |
| 3522 | <h5>Example:</h5> |
| Chris Lattner | c88c17b | 2004-02-24 04:54:45 +0000 | [diff] [blame] | 3523 | <pre> |
| 3524 | <i>; Emulate a conditional br instruction</i> |
| Reid Spencer | c78f337 | 2007-01-12 03:35:51 +0000 | [diff] [blame] | 3525 | %Val = <a href="#i_zext">zext</a> i1 %value to i32 |
| Dan Gohman | 2a08c53 | 2009-01-04 23:44:43 +0000 | [diff] [blame] | 3526 | switch i32 %Val, label %truedest [ i32 0, label %falsedest ] |
| Chris Lattner | c88c17b | 2004-02-24 04:54:45 +0000 | [diff] [blame] | 3527 | |
| 3528 | <i>; Emulate an unconditional br instruction</i> |
| Reid Spencer | ca86e16 | 2006-12-31 07:07:53 +0000 | [diff] [blame] | 3529 | switch i32 0, label %dest [ ] |
| Chris Lattner | c88c17b | 2004-02-24 04:54:45 +0000 | [diff] [blame] | 3530 | |
| 3531 | <i>; Implement a jump table:</i> |
| Dan Gohman | 2a08c53 | 2009-01-04 23:44:43 +0000 | [diff] [blame] | 3532 | switch i32 %val, label %otherwise [ i32 0, label %onzero |
| 3533 | i32 1, label %onone |
| 3534 | i32 2, label %ontwo ] |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3535 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3536 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 3537 | </div> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 3538 | |
| Chris Lattner | f9be95f | 2009-10-27 19:13:16 +0000 | [diff] [blame] | 3539 | |
| 3540 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3541 | <h4> |
| Chris Lattner | ab21db7 | 2009-10-28 00:19:10 +0000 | [diff] [blame] | 3542 | <a name="i_indirectbr">'<tt>indirectbr</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3543 | </h4> |
| Chris Lattner | f9be95f | 2009-10-27 19:13:16 +0000 | [diff] [blame] | 3544 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 3545 | <div> |
| Chris Lattner | f9be95f | 2009-10-27 19:13:16 +0000 | [diff] [blame] | 3546 | |
| 3547 | <h5>Syntax:</h5> |
| 3548 | <pre> |
| Chris Lattner | ab21db7 | 2009-10-28 00:19:10 +0000 | [diff] [blame] | 3549 | indirectbr <somety>* <address>, [ label <dest1>, label <dest2>, ... ] |
| Chris Lattner | f9be95f | 2009-10-27 19:13:16 +0000 | [diff] [blame] | 3550 | </pre> |
| 3551 | |
| 3552 | <h5>Overview:</h5> |
| 3553 | |
| Chris Lattner | ab21db7 | 2009-10-28 00:19:10 +0000 | [diff] [blame] | 3554 | <p>The '<tt>indirectbr</tt>' instruction implements an indirect branch to a label |
| Chris Lattner | f9be95f | 2009-10-27 19:13:16 +0000 | [diff] [blame] | 3555 | within the current function, whose address is specified by |
| Chris Lattner | c6f4436 | 2009-10-27 21:01:34 +0000 | [diff] [blame] | 3556 | "<tt>address</tt>". Address must be derived from a <a |
| 3557 | href="#blockaddress">blockaddress</a> constant.</p> |
| Chris Lattner | f9be95f | 2009-10-27 19:13:16 +0000 | [diff] [blame] | 3558 | |
| 3559 | <h5>Arguments:</h5> |
| 3560 | |
| 3561 | <p>The '<tt>address</tt>' argument is the address of the label to jump to. The |
| 3562 | rest of the arguments indicate the full set of possible destinations that the |
| 3563 | address may point to. Blocks are allowed to occur multiple times in the |
| 3564 | destination list, though this isn't particularly useful.</p> |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 3565 | |
| Chris Lattner | f9be95f | 2009-10-27 19:13:16 +0000 | [diff] [blame] | 3566 | <p>This destination list is required so that dataflow analysis has an accurate |
| 3567 | understanding of the CFG.</p> |
| 3568 | |
| 3569 | <h5>Semantics:</h5> |
| 3570 | |
| 3571 | <p>Control transfers to the block specified in the address argument. All |
| 3572 | possible destination blocks must be listed in the label list, otherwise this |
| 3573 | instruction has undefined behavior. This implies that jumps to labels |
| 3574 | defined in other functions have undefined behavior as well.</p> |
| 3575 | |
| 3576 | <h5>Implementation:</h5> |
| 3577 | |
| 3578 | <p>This is typically implemented with a jump through a register.</p> |
| 3579 | |
| 3580 | <h5>Example:</h5> |
| 3581 | <pre> |
| Chris Lattner | ab21db7 | 2009-10-28 00:19:10 +0000 | [diff] [blame] | 3582 | indirectbr i8* %Addr, [ label %bb1, label %bb2, label %bb3 ] |
| Chris Lattner | f9be95f | 2009-10-27 19:13:16 +0000 | [diff] [blame] | 3583 | </pre> |
| 3584 | |
| 3585 | </div> |
| 3586 | |
| 3587 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3588 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3589 | <h4> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 3590 | <a name="i_invoke">'<tt>invoke</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3591 | </h4> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 3592 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 3593 | <div> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 3594 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3595 | <h5>Syntax:</h5> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 3596 | <pre> |
| Devang Patel | 307e8ab | 2008-10-07 17:48:33 +0000 | [diff] [blame] | 3597 | <result> = invoke [<a href="#callingconv">cconv</a>] [<a href="#paramattrs">ret attrs</a>] <ptr to function ty> <function ptr val>(<function args>) [<a href="#fnattrs">fn attrs</a>] |
| Chris Lattner | 76b8a33 | 2006-05-14 18:23:06 +0000 | [diff] [blame] | 3598 | to label <normal label> unwind label <exception label> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 3599 | </pre> |
| 3600 | |
| Chris Lattner | 6536cfe | 2002-05-06 22:08:29 +0000 | [diff] [blame] | 3601 | <h5>Overview:</h5> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 3602 | <p>The '<tt>invoke</tt>' instruction causes control to transfer to a specified |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3603 | function, with the possibility of control flow transfer to either the |
| 3604 | '<tt>normal</tt>' label or the '<tt>exception</tt>' label. If the callee |
| 3605 | function returns with the "<tt><a href="#i_ret">ret</a></tt>" instruction, |
| 3606 | control flow will return to the "normal" label. If the callee (or any |
| Bill Wendling | 7b9e539 | 2012-02-06 21:57:33 +0000 | [diff] [blame] | 3607 | indirect callees) returns via the "<a href="#i_resume"><tt>resume</tt></a>" |
| 3608 | instruction or other exception handling mechanism, control is interrupted and |
| 3609 | continued at the dynamically nearest "exception" label.</p> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 3610 | |
| Bill Wendling | f78faf8 | 2011-08-02 21:52:38 +0000 | [diff] [blame] | 3611 | <p>The '<tt>exception</tt>' label is a |
| 3612 | <i><a href="ExceptionHandling.html#overview">landing pad</a></i> for the |
| 3613 | exception. As such, '<tt>exception</tt>' label is required to have the |
| 3614 | "<a href="#i_landingpad"><tt>landingpad</tt></a>" instruction, which contains |
| Chad Rosier | 85f5a1a | 2011-12-09 02:00:44 +0000 | [diff] [blame] | 3615 | the information about the behavior of the program after unwinding |
| Bill Wendling | f78faf8 | 2011-08-02 21:52:38 +0000 | [diff] [blame] | 3616 | happens, as its first non-PHI instruction. The restrictions on the |
| 3617 | "<tt>landingpad</tt>" instruction's tightly couples it to the |
| 3618 | "<tt>invoke</tt>" instruction, so that the important information contained |
| 3619 | within the "<tt>landingpad</tt>" instruction can't be lost through normal |
| 3620 | code motion.</p> |
| 3621 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3622 | <h5>Arguments:</h5> |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 3623 | <p>This instruction requires several arguments:</p> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 3624 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3625 | <ol> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3626 | <li>The optional "cconv" marker indicates which <a href="#callingconv">calling |
| 3627 | convention</a> the call should use. If none is specified, the call |
| 3628 | defaults to using C calling conventions.</li> |
| Devang Patel | f642f47 | 2008-10-06 18:50:38 +0000 | [diff] [blame] | 3629 | |
| 3630 | <li>The optional <a href="#paramattrs">Parameter Attributes</a> list for |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3631 | return values. Only '<tt>zeroext</tt>', '<tt>signext</tt>', and |
| 3632 | '<tt>inreg</tt>' attributes are valid here.</li> |
| Devang Patel | f642f47 | 2008-10-06 18:50:38 +0000 | [diff] [blame] | 3633 | |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 3634 | <li>'<tt>ptr to function ty</tt>': shall be the signature of the pointer to |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3635 | function value being invoked. In most cases, this is a direct function |
| 3636 | invocation, but indirect <tt>invoke</tt>s are just as possible, branching |
| 3637 | off an arbitrary pointer to function value.</li> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 3638 | |
| 3639 | <li>'<tt>function ptr val</tt>': An LLVM value containing a pointer to a |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3640 | function to be invoked. </li> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 3641 | |
| 3642 | <li>'<tt>function args</tt>': argument list whose types match the function |
| Chris Lattner | 0724fbd | 2010-03-02 06:36:51 +0000 | [diff] [blame] | 3643 | signature argument types and parameter attributes. All arguments must be |
| 3644 | of <a href="#t_firstclass">first class</a> type. If the function |
| 3645 | signature indicates the function accepts a variable number of arguments, |
| 3646 | the extra arguments can be specified.</li> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 3647 | |
| 3648 | <li>'<tt>normal label</tt>': the label reached when the called function |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3649 | executes a '<tt><a href="#i_ret">ret</a></tt>' instruction. </li> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 3650 | |
| Bill Wendling | 7b9e539 | 2012-02-06 21:57:33 +0000 | [diff] [blame] | 3651 | <li>'<tt>exception label</tt>': the label reached when a callee returns via |
| 3652 | the <a href="#i_resume"><tt>resume</tt></a> instruction or other exception |
| 3653 | handling mechanism.</li> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 3654 | |
| Devang Patel | 307e8ab | 2008-10-07 17:48:33 +0000 | [diff] [blame] | 3655 | <li>The optional <a href="#fnattrs">function attributes</a> list. Only |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3656 | '<tt>noreturn</tt>', '<tt>nounwind</tt>', '<tt>readonly</tt>' and |
| 3657 | '<tt>readnone</tt>' attributes are valid here.</li> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3658 | </ol> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 3659 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3660 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3661 | <p>This instruction is designed to operate as a standard |
| 3662 | '<tt><a href="#i_call">call</a></tt>' instruction in most regards. The |
| 3663 | primary difference is that it establishes an association with a label, which |
| 3664 | is used by the runtime library to unwind the stack.</p> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 3665 | |
| 3666 | <p>This instruction is used in languages with destructors to ensure that proper |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3667 | cleanup is performed in the case of either a <tt>longjmp</tt> or a thrown |
| 3668 | exception. Additionally, this is important for implementation of |
| 3669 | '<tt>catch</tt>' clauses in high-level languages that support them.</p> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 3670 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3671 | <p>For the purposes of the SSA form, the definition of the value returned by the |
| 3672 | '<tt>invoke</tt>' instruction is deemed to occur on the edge from the current |
| 3673 | block to the "normal" label. If the callee unwinds then no return value is |
| 3674 | available.</p> |
| Dan Gohman | f96a499 | 2009-05-22 21:47:08 +0000 | [diff] [blame] | 3675 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3676 | <h5>Example:</h5> |
| Chris Lattner | bad10ee | 2005-05-06 22:57:40 +0000 | [diff] [blame] | 3677 | <pre> |
| Nick Lewycky | d703f65 | 2008-03-16 07:18:12 +0000 | [diff] [blame] | 3678 | %retval = invoke i32 @Test(i32 15) to label %Continue |
| Jeff Cohen | 6f1cc77 | 2007-04-22 01:17:39 +0000 | [diff] [blame] | 3679 | unwind label %TestCleanup <i>; {i32}:retval set</i> |
| Nick Lewycky | d703f65 | 2008-03-16 07:18:12 +0000 | [diff] [blame] | 3680 | %retval = invoke <a href="#callingconv">coldcc</a> i32 %Testfnptr(i32 15) to label %Continue |
| Jeff Cohen | 6f1cc77 | 2007-04-22 01:17:39 +0000 | [diff] [blame] | 3681 | unwind label %TestCleanup <i>; {i32}:retval set</i> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3682 | </pre> |
| Chris Lattner | 35eca58 | 2004-10-16 18:04:13 +0000 | [diff] [blame] | 3683 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3684 | </div> |
| Chris Lattner | 35eca58 | 2004-10-16 18:04:13 +0000 | [diff] [blame] | 3685 | |
| Bill Wendling | dccc03b | 2011-07-31 06:30:59 +0000 | [diff] [blame] | 3686 | <!-- _______________________________________________________________________ --> |
| 3687 | |
| 3688 | <h4> |
| 3689 | <a name="i_resume">'<tt>resume</tt>' Instruction</a> |
| 3690 | </h4> |
| 3691 | |
| 3692 | <div> |
| 3693 | |
| 3694 | <h5>Syntax:</h5> |
| 3695 | <pre> |
| 3696 | resume <type> <value> |
| 3697 | </pre> |
| 3698 | |
| 3699 | <h5>Overview:</h5> |
| 3700 | <p>The '<tt>resume</tt>' instruction is a terminator instruction that has no |
| 3701 | successors.</p> |
| 3702 | |
| 3703 | <h5>Arguments:</h5> |
| Bill Wendling | f78faf8 | 2011-08-02 21:52:38 +0000 | [diff] [blame] | 3704 | <p>The '<tt>resume</tt>' instruction requires one argument, which must have the |
| Bill Wendling | e4ad50b | 2011-08-03 18:37:32 +0000 | [diff] [blame] | 3705 | same type as the result of any '<tt>landingpad</tt>' instruction in the same |
| 3706 | function.</p> |
| Bill Wendling | dccc03b | 2011-07-31 06:30:59 +0000 | [diff] [blame] | 3707 | |
| 3708 | <h5>Semantics:</h5> |
| 3709 | <p>The '<tt>resume</tt>' instruction resumes propagation of an existing |
| 3710 | (in-flight) exception whose unwinding was interrupted with |
| Bill Wendling | f78faf8 | 2011-08-02 21:52:38 +0000 | [diff] [blame] | 3711 | a <a href="#i_landingpad"><tt>landingpad</tt></a> instruction.</p> |
| Bill Wendling | dccc03b | 2011-07-31 06:30:59 +0000 | [diff] [blame] | 3712 | |
| 3713 | <h5>Example:</h5> |
| 3714 | <pre> |
| Bill Wendling | f78faf8 | 2011-08-02 21:52:38 +0000 | [diff] [blame] | 3715 | resume { i8*, i32 } %exn |
| Bill Wendling | dccc03b | 2011-07-31 06:30:59 +0000 | [diff] [blame] | 3716 | </pre> |
| 3717 | |
| 3718 | </div> |
| 3719 | |
| Chris Lattner | 35eca58 | 2004-10-16 18:04:13 +0000 | [diff] [blame] | 3720 | <!-- _______________________________________________________________________ --> |
| 3721 | |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3722 | <h4> |
| 3723 | <a name="i_unreachable">'<tt>unreachable</tt>' Instruction</a> |
| 3724 | </h4> |
| Chris Lattner | 35eca58 | 2004-10-16 18:04:13 +0000 | [diff] [blame] | 3725 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 3726 | <div> |
| Chris Lattner | 35eca58 | 2004-10-16 18:04:13 +0000 | [diff] [blame] | 3727 | |
| 3728 | <h5>Syntax:</h5> |
| 3729 | <pre> |
| 3730 | unreachable |
| 3731 | </pre> |
| 3732 | |
| 3733 | <h5>Overview:</h5> |
| Chris Lattner | 35eca58 | 2004-10-16 18:04:13 +0000 | [diff] [blame] | 3734 | <p>The '<tt>unreachable</tt>' instruction has no defined semantics. This |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3735 | instruction is used to inform the optimizer that a particular portion of the |
| 3736 | code is not reachable. This can be used to indicate that the code after a |
| 3737 | no-return function cannot be reached, and other facts.</p> |
| Chris Lattner | 35eca58 | 2004-10-16 18:04:13 +0000 | [diff] [blame] | 3738 | |
| 3739 | <h5>Semantics:</h5> |
| Chris Lattner | 35eca58 | 2004-10-16 18:04:13 +0000 | [diff] [blame] | 3740 | <p>The '<tt>unreachable</tt>' instruction has no defined semantics.</p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3741 | |
| Chris Lattner | 35eca58 | 2004-10-16 18:04:13 +0000 | [diff] [blame] | 3742 | </div> |
| 3743 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 3744 | </div> |
| 3745 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3746 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3747 | <h3> |
| 3748 | <a name="binaryops">Binary Operations</a> |
| 3749 | </h3> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3750 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 3751 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3752 | |
| 3753 | <p>Binary operators are used to do most of the computation in a program. They |
| 3754 | require two operands of the same type, execute an operation on them, and |
| 3755 | produce a single value. The operands might represent multiple data, as is |
| 3756 | the case with the <a href="#t_vector">vector</a> data type. The result value |
| 3757 | has the same type as its operands.</p> |
| 3758 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 3759 | <p>There are several different binary operators:</p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3760 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3761 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3762 | <h4> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 3763 | <a name="i_add">'<tt>add</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3764 | </h4> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 3765 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 3766 | <div> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 3767 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3768 | <h5>Syntax:</h5> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 3769 | <pre> |
| Dan Gohman | cbb38f2 | 2009-07-20 22:41:19 +0000 | [diff] [blame] | 3770 | <result> = add <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| Dan Gohman | fdfca79 | 2009-09-02 17:31:42 +0000 | [diff] [blame] | 3771 | <result> = add nuw <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| 3772 | <result> = add nsw <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| 3773 | <result> = add nuw nsw <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3774 | </pre> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 3775 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3776 | <h5>Overview:</h5> |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 3777 | <p>The '<tt>add</tt>' instruction returns the sum of its two operands.</p> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 3778 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3779 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3780 | <p>The two arguments to the '<tt>add</tt>' instruction must |
| 3781 | be <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of |
| 3782 | integer values. Both arguments must have identical types.</p> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 3783 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3784 | <h5>Semantics:</h5> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3785 | <p>The value produced is the integer sum of the two operands.</p> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 3786 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3787 | <p>If the sum has unsigned overflow, the result returned is the mathematical |
| 3788 | result modulo 2<sup>n</sup>, where n is the bit width of the result.</p> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 3789 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3790 | <p>Because LLVM integers use a two's complement representation, this instruction |
| 3791 | is appropriate for both signed and unsigned integers.</p> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 3792 | |
| Dan Gohman | 08d012e | 2009-07-22 22:44:56 +0000 | [diff] [blame] | 3793 | <p><tt>nuw</tt> and <tt>nsw</tt> stand for "No Unsigned Wrap" |
| 3794 | and "No Signed Wrap", respectively. If the <tt>nuw</tt> and/or |
| 3795 | <tt>nsw</tt> keywords are present, the result value of the <tt>add</tt> |
| Dan Gohman | bfb056d | 2011-12-06 03:18:47 +0000 | [diff] [blame] | 3796 | is a <a href="#poisonvalues">poison value</a> if unsigned and/or signed overflow, |
| Dan Gohman | fff6c53 | 2010-04-22 23:14:21 +0000 | [diff] [blame] | 3797 | respectively, occurs.</p> |
| Dan Gohman | cbb38f2 | 2009-07-20 22:41:19 +0000 | [diff] [blame] | 3798 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3799 | <h5>Example:</h5> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 3800 | <pre> |
| 3801 | <result> = add i32 4, %var <i>; yields {i32}:result = 4 + %var</i> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3802 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3803 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 3804 | </div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3805 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3806 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3807 | <h4> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3808 | <a name="i_fadd">'<tt>fadd</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3809 | </h4> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3810 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 3811 | <div> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3812 | |
| 3813 | <h5>Syntax:</h5> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3814 | <pre> |
| 3815 | <result> = fadd <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| 3816 | </pre> |
| 3817 | |
| 3818 | <h5>Overview:</h5> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3819 | <p>The '<tt>fadd</tt>' instruction returns the sum of its two operands.</p> |
| 3820 | |
| 3821 | <h5>Arguments:</h5> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3822 | <p>The two arguments to the '<tt>fadd</tt>' instruction must be |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3823 | <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of |
| 3824 | floating point values. Both arguments must have identical types.</p> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3825 | |
| 3826 | <h5>Semantics:</h5> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3827 | <p>The value produced is the floating point sum of the two operands.</p> |
| 3828 | |
| 3829 | <h5>Example:</h5> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3830 | <pre> |
| 3831 | <result> = fadd float 4.0, %var <i>; yields {float}:result = 4.0 + %var</i> |
| 3832 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3833 | |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3834 | </div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3835 | |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3836 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3837 | <h4> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 3838 | <a name="i_sub">'<tt>sub</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3839 | </h4> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 3840 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 3841 | <div> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 3842 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3843 | <h5>Syntax:</h5> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 3844 | <pre> |
| Dan Gohman | 08d012e | 2009-07-22 22:44:56 +0000 | [diff] [blame] | 3845 | <result> = sub <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| Dan Gohman | fdfca79 | 2009-09-02 17:31:42 +0000 | [diff] [blame] | 3846 | <result> = sub nuw <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| 3847 | <result> = sub nsw <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| 3848 | <result> = sub nuw nsw <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3849 | </pre> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 3850 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3851 | <h5>Overview:</h5> |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 3852 | <p>The '<tt>sub</tt>' instruction returns the difference of its two |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3853 | operands.</p> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 3854 | |
| 3855 | <p>Note that the '<tt>sub</tt>' instruction is used to represent the |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3856 | '<tt>neg</tt>' instruction present in most other intermediate |
| 3857 | representations.</p> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 3858 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3859 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3860 | <p>The two arguments to the '<tt>sub</tt>' instruction must |
| 3861 | be <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of |
| 3862 | integer values. Both arguments must have identical types.</p> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 3863 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3864 | <h5>Semantics:</h5> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3865 | <p>The value produced is the integer difference of the two operands.</p> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 3866 | |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3867 | <p>If the difference has unsigned overflow, the result returned is the |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3868 | mathematical result modulo 2<sup>n</sup>, where n is the bit width of the |
| 3869 | result.</p> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 3870 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3871 | <p>Because LLVM integers use a two's complement representation, this instruction |
| 3872 | is appropriate for both signed and unsigned integers.</p> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 3873 | |
| Dan Gohman | 08d012e | 2009-07-22 22:44:56 +0000 | [diff] [blame] | 3874 | <p><tt>nuw</tt> and <tt>nsw</tt> stand for "No Unsigned Wrap" |
| 3875 | and "No Signed Wrap", respectively. If the <tt>nuw</tt> and/or |
| 3876 | <tt>nsw</tt> keywords are present, the result value of the <tt>sub</tt> |
| Dan Gohman | bfb056d | 2011-12-06 03:18:47 +0000 | [diff] [blame] | 3877 | is a <a href="#poisonvalues">poison value</a> if unsigned and/or signed overflow, |
| Dan Gohman | fff6c53 | 2010-04-22 23:14:21 +0000 | [diff] [blame] | 3878 | respectively, occurs.</p> |
| Dan Gohman | cbb38f2 | 2009-07-20 22:41:19 +0000 | [diff] [blame] | 3879 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3880 | <h5>Example:</h5> |
| Bill Wendling | aac388b | 2007-05-29 09:42:13 +0000 | [diff] [blame] | 3881 | <pre> |
| 3882 | <result> = sub i32 4, %var <i>; yields {i32}:result = 4 - %var</i> |
| Reid Spencer | ca86e16 | 2006-12-31 07:07:53 +0000 | [diff] [blame] | 3883 | <result> = sub i32 0, %val <i>; yields {i32}:result = -%var</i> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3884 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3885 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 3886 | </div> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 3887 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3888 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3889 | <h4> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3890 | <a name="i_fsub">'<tt>fsub</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3891 | </h4> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3892 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 3893 | <div> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3894 | |
| 3895 | <h5>Syntax:</h5> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3896 | <pre> |
| 3897 | <result> = fsub <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| 3898 | </pre> |
| 3899 | |
| 3900 | <h5>Overview:</h5> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3901 | <p>The '<tt>fsub</tt>' instruction returns the difference of its two |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3902 | operands.</p> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3903 | |
| 3904 | <p>Note that the '<tt>fsub</tt>' instruction is used to represent the |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3905 | '<tt>fneg</tt>' instruction present in most other intermediate |
| 3906 | representations.</p> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3907 | |
| 3908 | <h5>Arguments:</h5> |
| Bill Wendling | d9fe298 | 2009-07-20 02:32:41 +0000 | [diff] [blame] | 3909 | <p>The two arguments to the '<tt>fsub</tt>' instruction must be |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3910 | <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of |
| 3911 | floating point values. Both arguments must have identical types.</p> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3912 | |
| 3913 | <h5>Semantics:</h5> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3914 | <p>The value produced is the floating point difference of the two operands.</p> |
| 3915 | |
| 3916 | <h5>Example:</h5> |
| 3917 | <pre> |
| 3918 | <result> = fsub float 4.0, %var <i>; yields {float}:result = 4.0 - %var</i> |
| 3919 | <result> = fsub float -0.0, %val <i>; yields {float}:result = -%var</i> |
| 3920 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3921 | |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3922 | </div> |
| 3923 | |
| 3924 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3925 | <h4> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 3926 | <a name="i_mul">'<tt>mul</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3927 | </h4> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 3928 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 3929 | <div> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 3930 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3931 | <h5>Syntax:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3932 | <pre> |
| Dan Gohman | 08d012e | 2009-07-22 22:44:56 +0000 | [diff] [blame] | 3933 | <result> = mul <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| Dan Gohman | fdfca79 | 2009-09-02 17:31:42 +0000 | [diff] [blame] | 3934 | <result> = mul nuw <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| 3935 | <result> = mul nsw <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| 3936 | <result> = mul nuw nsw <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3937 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3938 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3939 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3940 | <p>The '<tt>mul</tt>' instruction returns the product of its two operands.</p> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 3941 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3942 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3943 | <p>The two arguments to the '<tt>mul</tt>' instruction must |
| 3944 | be <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of |
| 3945 | integer values. Both arguments must have identical types.</p> |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 3946 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3947 | <h5>Semantics:</h5> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3948 | <p>The value produced is the integer product of the two operands.</p> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 3949 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3950 | <p>If the result of the multiplication has unsigned overflow, the result |
| 3951 | returned is the mathematical result modulo 2<sup>n</sup>, where n is the bit |
| 3952 | width of the result.</p> |
| 3953 | |
| 3954 | <p>Because LLVM integers use a two's complement representation, and the result |
| 3955 | is the same width as the operands, this instruction returns the correct |
| 3956 | result for both signed and unsigned integers. If a full product |
| 3957 | (e.g. <tt>i32</tt>x<tt>i32</tt>-><tt>i64</tt>) is needed, the operands should |
| 3958 | be sign-extended or zero-extended as appropriate to the width of the full |
| 3959 | product.</p> |
| 3960 | |
| Dan Gohman | 08d012e | 2009-07-22 22:44:56 +0000 | [diff] [blame] | 3961 | <p><tt>nuw</tt> and <tt>nsw</tt> stand for "No Unsigned Wrap" |
| 3962 | and "No Signed Wrap", respectively. If the <tt>nuw</tt> and/or |
| 3963 | <tt>nsw</tt> keywords are present, the result value of the <tt>mul</tt> |
| Dan Gohman | bfb056d | 2011-12-06 03:18:47 +0000 | [diff] [blame] | 3964 | is a <a href="#poisonvalues">poison value</a> if unsigned and/or signed overflow, |
| Dan Gohman | fff6c53 | 2010-04-22 23:14:21 +0000 | [diff] [blame] | 3965 | respectively, occurs.</p> |
| Dan Gohman | cbb38f2 | 2009-07-20 22:41:19 +0000 | [diff] [blame] | 3966 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3967 | <h5>Example:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3968 | <pre> |
| 3969 | <result> = mul i32 4, %var <i>; yields {i32}:result = 4 * %var</i> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3970 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3971 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 3972 | </div> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 3973 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 3974 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3975 | <h4> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3976 | <a name="i_fmul">'<tt>fmul</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 3977 | </h4> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3978 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 3979 | <div> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3980 | |
| 3981 | <h5>Syntax:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3982 | <pre> |
| 3983 | <result> = fmul <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3984 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3985 | |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3986 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3987 | <p>The '<tt>fmul</tt>' instruction returns the product of its two operands.</p> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3988 | |
| 3989 | <h5>Arguments:</h5> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3990 | <p>The two arguments to the '<tt>fmul</tt>' instruction must be |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3991 | <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of |
| 3992 | floating point values. Both arguments must have identical types.</p> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3993 | |
| 3994 | <h5>Semantics:</h5> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 3995 | <p>The value produced is the floating point product of the two operands.</p> |
| 3996 | |
| 3997 | <h5>Example:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 3998 | <pre> |
| 3999 | <result> = fmul float 4.0, %var <i>; yields {float}:result = 4.0 * %var</i> |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 4000 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4001 | |
| Dan Gohman | ae3a0be | 2009-06-04 22:49:04 +0000 | [diff] [blame] | 4002 | </div> |
| 4003 | |
| 4004 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4005 | <h4> |
| 4006 | <a name="i_udiv">'<tt>udiv</tt>' Instruction</a> |
| 4007 | </h4> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4008 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 4009 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4010 | |
| Reid Spencer | 1628cec | 2006-10-26 06:15:43 +0000 | [diff] [blame] | 4011 | <h5>Syntax:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4012 | <pre> |
| Chris Lattner | 35bda89 | 2011-02-06 21:44:57 +0000 | [diff] [blame] | 4013 | <result> = udiv <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| 4014 | <result> = udiv exact <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| Reid Spencer | 1628cec | 2006-10-26 06:15:43 +0000 | [diff] [blame] | 4015 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4016 | |
| Reid Spencer | 1628cec | 2006-10-26 06:15:43 +0000 | [diff] [blame] | 4017 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4018 | <p>The '<tt>udiv</tt>' instruction returns the quotient of its two operands.</p> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4019 | |
| Reid Spencer | 1628cec | 2006-10-26 06:15:43 +0000 | [diff] [blame] | 4020 | <h5>Arguments:</h5> |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 4021 | <p>The two arguments to the '<tt>udiv</tt>' instruction must be |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4022 | <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer |
| 4023 | values. Both arguments must have identical types.</p> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4024 | |
| Reid Spencer | 1628cec | 2006-10-26 06:15:43 +0000 | [diff] [blame] | 4025 | <h5>Semantics:</h5> |
| Chris Lattner | 5ec8983 | 2008-01-28 00:36:27 +0000 | [diff] [blame] | 4026 | <p>The value produced is the unsigned integer quotient of the two operands.</p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4027 | |
| Chris Lattner | 5ec8983 | 2008-01-28 00:36:27 +0000 | [diff] [blame] | 4028 | <p>Note that unsigned integer division and signed integer division are distinct |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4029 | operations; for signed integer division, use '<tt>sdiv</tt>'.</p> |
| 4030 | |
| Chris Lattner | 5ec8983 | 2008-01-28 00:36:27 +0000 | [diff] [blame] | 4031 | <p>Division by zero leads to undefined behavior.</p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4032 | |
| Chris Lattner | 35bda89 | 2011-02-06 21:44:57 +0000 | [diff] [blame] | 4033 | <p>If the <tt>exact</tt> keyword is present, the result value of the |
| Dan Gohman | bfb056d | 2011-12-06 03:18:47 +0000 | [diff] [blame] | 4034 | <tt>udiv</tt> is a <a href="#poisonvalues">poison value</a> if %op1 is not a |
| Chris Lattner | 35bda89 | 2011-02-06 21:44:57 +0000 | [diff] [blame] | 4035 | multiple of %op2 (as such, "((a udiv exact b) mul b) == a").</p> |
| 4036 | |
| 4037 | |
| Reid Spencer | 1628cec | 2006-10-26 06:15:43 +0000 | [diff] [blame] | 4038 | <h5>Example:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4039 | <pre> |
| 4040 | <result> = udiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i> |
| Reid Spencer | 1628cec | 2006-10-26 06:15:43 +0000 | [diff] [blame] | 4041 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4042 | |
| Reid Spencer | 1628cec | 2006-10-26 06:15:43 +0000 | [diff] [blame] | 4043 | </div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4044 | |
| Reid Spencer | 1628cec | 2006-10-26 06:15:43 +0000 | [diff] [blame] | 4045 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4046 | <h4> |
| 4047 | <a name="i_sdiv">'<tt>sdiv</tt>' Instruction</a> |
| 4048 | </h4> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4049 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 4050 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4051 | |
| Reid Spencer | 1628cec | 2006-10-26 06:15:43 +0000 | [diff] [blame] | 4052 | <h5>Syntax:</h5> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4053 | <pre> |
| Dan Gohman | cbb38f2 | 2009-07-20 22:41:19 +0000 | [diff] [blame] | 4054 | <result> = sdiv <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| Dan Gohman | fdfca79 | 2009-09-02 17:31:42 +0000 | [diff] [blame] | 4055 | <result> = sdiv exact <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| Reid Spencer | 1628cec | 2006-10-26 06:15:43 +0000 | [diff] [blame] | 4056 | </pre> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4057 | |
| Reid Spencer | 1628cec | 2006-10-26 06:15:43 +0000 | [diff] [blame] | 4058 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4059 | <p>The '<tt>sdiv</tt>' instruction returns the quotient of its two operands.</p> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4060 | |
| Reid Spencer | 1628cec | 2006-10-26 06:15:43 +0000 | [diff] [blame] | 4061 | <h5>Arguments:</h5> |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 4062 | <p>The two arguments to the '<tt>sdiv</tt>' instruction must be |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4063 | <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer |
| 4064 | values. Both arguments must have identical types.</p> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4065 | |
| Reid Spencer | 1628cec | 2006-10-26 06:15:43 +0000 | [diff] [blame] | 4066 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4067 | <p>The value produced is the signed integer quotient of the two operands rounded |
| 4068 | towards zero.</p> |
| 4069 | |
| Chris Lattner | 5ec8983 | 2008-01-28 00:36:27 +0000 | [diff] [blame] | 4070 | <p>Note that signed integer division and unsigned integer division are distinct |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4071 | operations; for unsigned integer division, use '<tt>udiv</tt>'.</p> |
| 4072 | |
| Chris Lattner | 5ec8983 | 2008-01-28 00:36:27 +0000 | [diff] [blame] | 4073 | <p>Division by zero leads to undefined behavior. Overflow also leads to |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4074 | undefined behavior; this is a rare case, but can occur, for example, by doing |
| 4075 | a 32-bit division of -2147483648 by -1.</p> |
| 4076 | |
| Dan Gohman | 9c5beed | 2009-07-22 00:04:19 +0000 | [diff] [blame] | 4077 | <p>If the <tt>exact</tt> keyword is present, the result value of the |
| Dan Gohman | bfb056d | 2011-12-06 03:18:47 +0000 | [diff] [blame] | 4078 | <tt>sdiv</tt> is a <a href="#poisonvalues">poison value</a> if the result would |
| Dan Gohman | 38da927 | 2010-07-11 00:08:34 +0000 | [diff] [blame] | 4079 | be rounded.</p> |
| Dan Gohman | cbb38f2 | 2009-07-20 22:41:19 +0000 | [diff] [blame] | 4080 | |
| Reid Spencer | 1628cec | 2006-10-26 06:15:43 +0000 | [diff] [blame] | 4081 | <h5>Example:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4082 | <pre> |
| 4083 | <result> = sdiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i> |
| Reid Spencer | 1628cec | 2006-10-26 06:15:43 +0000 | [diff] [blame] | 4084 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4085 | |
| Reid Spencer | 1628cec | 2006-10-26 06:15:43 +0000 | [diff] [blame] | 4086 | </div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4087 | |
| Reid Spencer | 1628cec | 2006-10-26 06:15:43 +0000 | [diff] [blame] | 4088 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4089 | <h4> |
| 4090 | <a name="i_fdiv">'<tt>fdiv</tt>' Instruction</a> |
| 4091 | </h4> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4092 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 4093 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4094 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4095 | <h5>Syntax:</h5> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4096 | <pre> |
| Gabor Greif | fb224a2 | 2008-08-07 21:46:00 +0000 | [diff] [blame] | 4097 | <result> = fdiv <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 4098 | </pre> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4099 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4100 | <h5>Overview:</h5> |
| 4101 | <p>The '<tt>fdiv</tt>' instruction returns the quotient of its two operands.</p> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4102 | |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 4103 | <h5>Arguments:</h5> |
| Jeff Cohen | 6f1cc77 | 2007-04-22 01:17:39 +0000 | [diff] [blame] | 4104 | <p>The two arguments to the '<tt>fdiv</tt>' instruction must be |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4105 | <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of |
| 4106 | floating point values. Both arguments must have identical types.</p> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4107 | |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 4108 | <h5>Semantics:</h5> |
| Reid Spencer | 1628cec | 2006-10-26 06:15:43 +0000 | [diff] [blame] | 4109 | <p>The value produced is the floating point quotient of the two operands.</p> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4110 | |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 4111 | <h5>Example:</h5> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4112 | <pre> |
| 4113 | <result> = fdiv float 4.0, %var <i>; yields {float}:result = 4.0 / %var</i> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 4114 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4115 | |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 4116 | </div> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4117 | |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 4118 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4119 | <h4> |
| 4120 | <a name="i_urem">'<tt>urem</tt>' Instruction</a> |
| 4121 | </h4> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4122 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 4123 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4124 | |
| Reid Spencer | 0a783f7 | 2006-11-02 01:53:59 +0000 | [diff] [blame] | 4125 | <h5>Syntax:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4126 | <pre> |
| 4127 | <result> = urem <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| Reid Spencer | 0a783f7 | 2006-11-02 01:53:59 +0000 | [diff] [blame] | 4128 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4129 | |
| Reid Spencer | 0a783f7 | 2006-11-02 01:53:59 +0000 | [diff] [blame] | 4130 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4131 | <p>The '<tt>urem</tt>' instruction returns the remainder from the unsigned |
| 4132 | division of its two arguments.</p> |
| 4133 | |
| Reid Spencer | 0a783f7 | 2006-11-02 01:53:59 +0000 | [diff] [blame] | 4134 | <h5>Arguments:</h5> |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 4135 | <p>The two arguments to the '<tt>urem</tt>' instruction must be |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4136 | <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer |
| 4137 | values. Both arguments must have identical types.</p> |
| 4138 | |
| Reid Spencer | 0a783f7 | 2006-11-02 01:53:59 +0000 | [diff] [blame] | 4139 | <h5>Semantics:</h5> |
| 4140 | <p>This instruction returns the unsigned integer <i>remainder</i> of a division. |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4141 | This instruction always performs an unsigned division to get the |
| 4142 | remainder.</p> |
| 4143 | |
| Chris Lattner | 5ec8983 | 2008-01-28 00:36:27 +0000 | [diff] [blame] | 4144 | <p>Note that unsigned integer remainder and signed integer remainder are |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4145 | distinct operations; for signed integer remainder, use '<tt>srem</tt>'.</p> |
| 4146 | |
| Chris Lattner | 5ec8983 | 2008-01-28 00:36:27 +0000 | [diff] [blame] | 4147 | <p>Taking the remainder of a division by zero leads to undefined behavior.</p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4148 | |
| Reid Spencer | 0a783f7 | 2006-11-02 01:53:59 +0000 | [diff] [blame] | 4149 | <h5>Example:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4150 | <pre> |
| 4151 | <result> = urem i32 4, %var <i>; yields {i32}:result = 4 % %var</i> |
| Reid Spencer | 0a783f7 | 2006-11-02 01:53:59 +0000 | [diff] [blame] | 4152 | </pre> |
| 4153 | |
| 4154 | </div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4155 | |
| Reid Spencer | 0a783f7 | 2006-11-02 01:53:59 +0000 | [diff] [blame] | 4156 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4157 | <h4> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4158 | <a name="i_srem">'<tt>srem</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4159 | </h4> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4160 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 4161 | <div> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4162 | |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 4163 | <h5>Syntax:</h5> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4164 | <pre> |
| Gabor Greif | fb224a2 | 2008-08-07 21:46:00 +0000 | [diff] [blame] | 4165 | <result> = srem <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 4166 | </pre> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4167 | |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 4168 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4169 | <p>The '<tt>srem</tt>' instruction returns the remainder from the signed |
| 4170 | division of its two operands. This instruction can also take |
| 4171 | <a href="#t_vector">vector</a> versions of the values in which case the |
| 4172 | elements must be integers.</p> |
| Chris Lattner | c7d3ab3 | 2008-01-04 04:33:49 +0000 | [diff] [blame] | 4173 | |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 4174 | <h5>Arguments:</h5> |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 4175 | <p>The two arguments to the '<tt>srem</tt>' instruction must be |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4176 | <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer |
| 4177 | values. Both arguments must have identical types.</p> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4178 | |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 4179 | <h5>Semantics:</h5> |
| Reid Spencer | 0a783f7 | 2006-11-02 01:53:59 +0000 | [diff] [blame] | 4180 | <p>This instruction returns the <i>remainder</i> of a division (where the result |
| Duncan Sands | dea3a5e | 2011-03-07 09:12:24 +0000 | [diff] [blame] | 4181 | is either zero or has the same sign as the dividend, <tt>op1</tt>), not the |
| 4182 | <i>modulo</i> operator (where the result is either zero or has the same sign |
| 4183 | as the divisor, <tt>op2</tt>) of a value. |
| 4184 | For more information about the difference, |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4185 | see <a href="http://mathforum.org/dr.math/problems/anne.4.28.99.html">The |
| 4186 | Math Forum</a>. For a table of how this is implemented in various languages, |
| 4187 | please see <a href="http://en.wikipedia.org/wiki/Modulo_operation"> |
| 4188 | Wikipedia: modulo operation</a>.</p> |
| 4189 | |
| Chris Lattner | 5ec8983 | 2008-01-28 00:36:27 +0000 | [diff] [blame] | 4190 | <p>Note that signed integer remainder and unsigned integer remainder are |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4191 | distinct operations; for unsigned integer remainder, use '<tt>urem</tt>'.</p> |
| 4192 | |
| Chris Lattner | 5ec8983 | 2008-01-28 00:36:27 +0000 | [diff] [blame] | 4193 | <p>Taking the remainder of a division by zero leads to undefined behavior. |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4194 | Overflow also leads to undefined behavior; this is a rare case, but can |
| 4195 | occur, for example, by taking the remainder of a 32-bit division of |
| 4196 | -2147483648 by -1. (The remainder doesn't actually overflow, but this rule |
| 4197 | lets srem be implemented using instructions that return both the result of |
| 4198 | the division and the remainder.)</p> |
| 4199 | |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 4200 | <h5>Example:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4201 | <pre> |
| 4202 | <result> = srem i32 4, %var <i>; yields {i32}:result = 4 % %var</i> |
| Reid Spencer | 0a783f7 | 2006-11-02 01:53:59 +0000 | [diff] [blame] | 4203 | </pre> |
| 4204 | |
| 4205 | </div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4206 | |
| Reid Spencer | 0a783f7 | 2006-11-02 01:53:59 +0000 | [diff] [blame] | 4207 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4208 | <h4> |
| 4209 | <a name="i_frem">'<tt>frem</tt>' Instruction</a> |
| 4210 | </h4> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4211 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 4212 | <div> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4213 | |
| Reid Spencer | 0a783f7 | 2006-11-02 01:53:59 +0000 | [diff] [blame] | 4214 | <h5>Syntax:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4215 | <pre> |
| 4216 | <result> = frem <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| Reid Spencer | 0a783f7 | 2006-11-02 01:53:59 +0000 | [diff] [blame] | 4217 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4218 | |
| Reid Spencer | 0a783f7 | 2006-11-02 01:53:59 +0000 | [diff] [blame] | 4219 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4220 | <p>The '<tt>frem</tt>' instruction returns the remainder from the division of |
| 4221 | its two operands.</p> |
| 4222 | |
| Reid Spencer | 0a783f7 | 2006-11-02 01:53:59 +0000 | [diff] [blame] | 4223 | <h5>Arguments:</h5> |
| 4224 | <p>The two arguments to the '<tt>frem</tt>' instruction must be |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4225 | <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of |
| 4226 | floating point values. Both arguments must have identical types.</p> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4227 | |
| Reid Spencer | 0a783f7 | 2006-11-02 01:53:59 +0000 | [diff] [blame] | 4228 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4229 | <p>This instruction returns the <i>remainder</i> of a division. The remainder |
| 4230 | has the same sign as the dividend.</p> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4231 | |
| Reid Spencer | 0a783f7 | 2006-11-02 01:53:59 +0000 | [diff] [blame] | 4232 | <h5>Example:</h5> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4233 | <pre> |
| 4234 | <result> = frem float 4.0, %var <i>; yields {float}:result = 4.0 % %var</i> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 4235 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4236 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 4237 | </div> |
| Robert Bocchino | 7b81c75 | 2006-02-17 21:18:08 +0000 | [diff] [blame] | 4238 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 4239 | </div> |
| 4240 | |
| Reid Spencer | 8e11bf8 | 2007-02-02 13:57:07 +0000 | [diff] [blame] | 4241 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4242 | <h3> |
| 4243 | <a name="bitwiseops">Bitwise Binary Operations</a> |
| 4244 | </h3> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4245 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 4246 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4247 | |
| 4248 | <p>Bitwise binary operators are used to do various forms of bit-twiddling in a |
| 4249 | program. They are generally very efficient instructions and can commonly be |
| 4250 | strength reduced from other instructions. They require two operands of the |
| 4251 | same type, execute an operation on them, and produce a single value. The |
| 4252 | resulting value is the same type as its operands.</p> |
| 4253 | |
| Reid Spencer | 569f2fa | 2007-01-31 21:39:12 +0000 | [diff] [blame] | 4254 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4255 | <h4> |
| 4256 | <a name="i_shl">'<tt>shl</tt>' Instruction</a> |
| 4257 | </h4> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4258 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 4259 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4260 | |
| Reid Spencer | 569f2fa | 2007-01-31 21:39:12 +0000 | [diff] [blame] | 4261 | <h5>Syntax:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4262 | <pre> |
| Chris Lattner | f067d58 | 2011-02-07 16:40:21 +0000 | [diff] [blame] | 4263 | <result> = shl <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| 4264 | <result> = shl nuw <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| 4265 | <result> = shl nsw <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| 4266 | <result> = shl nuw nsw <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| Reid Spencer | 569f2fa | 2007-01-31 21:39:12 +0000 | [diff] [blame] | 4267 | </pre> |
| Chris Lattner | 6ccc2d5 | 2007-10-03 21:01:14 +0000 | [diff] [blame] | 4268 | |
| Reid Spencer | 569f2fa | 2007-01-31 21:39:12 +0000 | [diff] [blame] | 4269 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4270 | <p>The '<tt>shl</tt>' instruction returns the first operand shifted to the left |
| 4271 | a specified number of bits.</p> |
| Chris Lattner | 6ccc2d5 | 2007-10-03 21:01:14 +0000 | [diff] [blame] | 4272 | |
| Reid Spencer | 569f2fa | 2007-01-31 21:39:12 +0000 | [diff] [blame] | 4273 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4274 | <p>Both arguments to the '<tt>shl</tt>' instruction must be the |
| 4275 | same <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of |
| 4276 | integer type. '<tt>op2</tt>' is treated as an unsigned value.</p> |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 4277 | |
| Reid Spencer | 569f2fa | 2007-01-31 21:39:12 +0000 | [diff] [blame] | 4278 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4279 | <p>The value produced is <tt>op1</tt> * 2<sup><tt>op2</tt></sup> mod |
| 4280 | 2<sup>n</sup>, where <tt>n</tt> is the width of the result. If <tt>op2</tt> |
| 4281 | is (statically or dynamically) negative or equal to or larger than the number |
| 4282 | of bits in <tt>op1</tt>, the result is undefined. If the arguments are |
| 4283 | vectors, each vector element of <tt>op1</tt> is shifted by the corresponding |
| 4284 | shift amount in <tt>op2</tt>.</p> |
| Chris Lattner | 6ccc2d5 | 2007-10-03 21:01:14 +0000 | [diff] [blame] | 4285 | |
| Chris Lattner | f067d58 | 2011-02-07 16:40:21 +0000 | [diff] [blame] | 4286 | <p>If the <tt>nuw</tt> keyword is present, then the shift produces a |
| Dan Gohman | bfb056d | 2011-12-06 03:18:47 +0000 | [diff] [blame] | 4287 | <a href="#poisonvalues">poison value</a> if it shifts out any non-zero bits. If |
| Chris Lattner | 66298c1 | 2011-02-09 16:44:44 +0000 | [diff] [blame] | 4288 | the <tt>nsw</tt> keyword is present, then the shift produces a |
| Dan Gohman | bfb056d | 2011-12-06 03:18:47 +0000 | [diff] [blame] | 4289 | <a href="#poisonvalues">poison value</a> if it shifts out any bits that disagree |
| Chris Lattner | f067d58 | 2011-02-07 16:40:21 +0000 | [diff] [blame] | 4290 | with the resultant sign bit. As such, NUW/NSW have the same semantics as |
| 4291 | they would if the shift were expressed as a mul instruction with the same |
| 4292 | nsw/nuw bits in (mul %op1, (shl 1, %op2)).</p> |
| 4293 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4294 | <h5>Example:</h5> |
| 4295 | <pre> |
| Reid Spencer | 569f2fa | 2007-01-31 21:39:12 +0000 | [diff] [blame] | 4296 | <result> = shl i32 4, %var <i>; yields {i32}: 4 << %var</i> |
| 4297 | <result> = shl i32 4, 2 <i>; yields {i32}: 16</i> |
| 4298 | <result> = shl i32 1, 10 <i>; yields {i32}: 1024</i> |
| Chris Lattner | 6ccc2d5 | 2007-10-03 21:01:14 +0000 | [diff] [blame] | 4299 | <result> = shl i32 1, 32 <i>; undefined</i> |
| Mon P Wang | e9f1015 | 2008-12-09 05:46:39 +0000 | [diff] [blame] | 4300 | <result> = shl <2 x i32> < i32 1, i32 1>, < i32 1, i32 2> <i>; yields: result=<2 x i32> < i32 2, i32 4></i> |
| Reid Spencer | 569f2fa | 2007-01-31 21:39:12 +0000 | [diff] [blame] | 4301 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4302 | |
| Reid Spencer | 569f2fa | 2007-01-31 21:39:12 +0000 | [diff] [blame] | 4303 | </div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4304 | |
| Reid Spencer | 569f2fa | 2007-01-31 21:39:12 +0000 | [diff] [blame] | 4305 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4306 | <h4> |
| 4307 | <a name="i_lshr">'<tt>lshr</tt>' Instruction</a> |
| 4308 | </h4> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4309 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 4310 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4311 | |
| Reid Spencer | 569f2fa | 2007-01-31 21:39:12 +0000 | [diff] [blame] | 4312 | <h5>Syntax:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4313 | <pre> |
| Chris Lattner | f067d58 | 2011-02-07 16:40:21 +0000 | [diff] [blame] | 4314 | <result> = lshr <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| 4315 | <result> = lshr exact <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| Reid Spencer | 569f2fa | 2007-01-31 21:39:12 +0000 | [diff] [blame] | 4316 | </pre> |
| 4317 | |
| 4318 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4319 | <p>The '<tt>lshr</tt>' instruction (logical shift right) returns the first |
| 4320 | operand shifted to the right a specified number of bits with zero fill.</p> |
| Reid Spencer | 569f2fa | 2007-01-31 21:39:12 +0000 | [diff] [blame] | 4321 | |
| 4322 | <h5>Arguments:</h5> |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 4323 | <p>Both arguments to the '<tt>lshr</tt>' instruction must be the same |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4324 | <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer |
| 4325 | type. '<tt>op2</tt>' is treated as an unsigned value.</p> |
| Reid Spencer | 569f2fa | 2007-01-31 21:39:12 +0000 | [diff] [blame] | 4326 | |
| 4327 | <h5>Semantics:</h5> |
| 4328 | <p>This instruction always performs a logical shift right operation. The most |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4329 | significant bits of the result will be filled with zero bits after the shift. |
| 4330 | If <tt>op2</tt> is (statically or dynamically) equal to or larger than the |
| 4331 | number of bits in <tt>op1</tt>, the result is undefined. If the arguments are |
| 4332 | vectors, each vector element of <tt>op1</tt> is shifted by the corresponding |
| 4333 | shift amount in <tt>op2</tt>.</p> |
| Reid Spencer | 569f2fa | 2007-01-31 21:39:12 +0000 | [diff] [blame] | 4334 | |
| Chris Lattner | f067d58 | 2011-02-07 16:40:21 +0000 | [diff] [blame] | 4335 | <p>If the <tt>exact</tt> keyword is present, the result value of the |
| Dan Gohman | bfb056d | 2011-12-06 03:18:47 +0000 | [diff] [blame] | 4336 | <tt>lshr</tt> is a <a href="#poisonvalues">poison value</a> if any of the bits |
| Chris Lattner | f067d58 | 2011-02-07 16:40:21 +0000 | [diff] [blame] | 4337 | shifted out are non-zero.</p> |
| 4338 | |
| 4339 | |
| Reid Spencer | 569f2fa | 2007-01-31 21:39:12 +0000 | [diff] [blame] | 4340 | <h5>Example:</h5> |
| 4341 | <pre> |
| 4342 | <result> = lshr i32 4, 1 <i>; yields {i32}:result = 2</i> |
| 4343 | <result> = lshr i32 4, 2 <i>; yields {i32}:result = 1</i> |
| 4344 | <result> = lshr i8 4, 3 <i>; yields {i8}:result = 0</i> |
| 4345 | <result> = lshr i8 -2, 1 <i>; yields {i8}:result = 0x7FFFFFFF </i> |
| Chris Lattner | 6ccc2d5 | 2007-10-03 21:01:14 +0000 | [diff] [blame] | 4346 | <result> = lshr i32 1, 32 <i>; undefined</i> |
| Mon P Wang | e9f1015 | 2008-12-09 05:46:39 +0000 | [diff] [blame] | 4347 | <result> = lshr <2 x i32> < i32 -2, i32 4>, < i32 1, i32 2> <i>; yields: result=<2 x i32> < i32 0x7FFFFFFF, i32 1></i> |
| Reid Spencer | 569f2fa | 2007-01-31 21:39:12 +0000 | [diff] [blame] | 4348 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4349 | |
| Reid Spencer | 569f2fa | 2007-01-31 21:39:12 +0000 | [diff] [blame] | 4350 | </div> |
| 4351 | |
| Reid Spencer | 8e11bf8 | 2007-02-02 13:57:07 +0000 | [diff] [blame] | 4352 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4353 | <h4> |
| 4354 | <a name="i_ashr">'<tt>ashr</tt>' Instruction</a> |
| 4355 | </h4> |
| 4356 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 4357 | <div> |
| Reid Spencer | 569f2fa | 2007-01-31 21:39:12 +0000 | [diff] [blame] | 4358 | |
| 4359 | <h5>Syntax:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4360 | <pre> |
| Chris Lattner | f067d58 | 2011-02-07 16:40:21 +0000 | [diff] [blame] | 4361 | <result> = ashr <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| 4362 | <result> = ashr exact <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| Reid Spencer | 569f2fa | 2007-01-31 21:39:12 +0000 | [diff] [blame] | 4363 | </pre> |
| 4364 | |
| 4365 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4366 | <p>The '<tt>ashr</tt>' instruction (arithmetic shift right) returns the first |
| 4367 | operand shifted to the right a specified number of bits with sign |
| 4368 | extension.</p> |
| Reid Spencer | 569f2fa | 2007-01-31 21:39:12 +0000 | [diff] [blame] | 4369 | |
| 4370 | <h5>Arguments:</h5> |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 4371 | <p>Both arguments to the '<tt>ashr</tt>' instruction must be the same |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4372 | <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer |
| 4373 | type. '<tt>op2</tt>' is treated as an unsigned value.</p> |
| Reid Spencer | 569f2fa | 2007-01-31 21:39:12 +0000 | [diff] [blame] | 4374 | |
| 4375 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4376 | <p>This instruction always performs an arithmetic shift right operation, The |
| 4377 | most significant bits of the result will be filled with the sign bit |
| 4378 | of <tt>op1</tt>. If <tt>op2</tt> is (statically or dynamically) equal to or |
| 4379 | larger than the number of bits in <tt>op1</tt>, the result is undefined. If |
| 4380 | the arguments are vectors, each vector element of <tt>op1</tt> is shifted by |
| 4381 | the corresponding shift amount in <tt>op2</tt>.</p> |
| Reid Spencer | 569f2fa | 2007-01-31 21:39:12 +0000 | [diff] [blame] | 4382 | |
| Chris Lattner | f067d58 | 2011-02-07 16:40:21 +0000 | [diff] [blame] | 4383 | <p>If the <tt>exact</tt> keyword is present, the result value of the |
| Dan Gohman | bfb056d | 2011-12-06 03:18:47 +0000 | [diff] [blame] | 4384 | <tt>ashr</tt> is a <a href="#poisonvalues">poison value</a> if any of the bits |
| Chris Lattner | f067d58 | 2011-02-07 16:40:21 +0000 | [diff] [blame] | 4385 | shifted out are non-zero.</p> |
| 4386 | |
| Reid Spencer | 569f2fa | 2007-01-31 21:39:12 +0000 | [diff] [blame] | 4387 | <h5>Example:</h5> |
| 4388 | <pre> |
| 4389 | <result> = ashr i32 4, 1 <i>; yields {i32}:result = 2</i> |
| 4390 | <result> = ashr i32 4, 2 <i>; yields {i32}:result = 1</i> |
| 4391 | <result> = ashr i8 4, 3 <i>; yields {i8}:result = 0</i> |
| 4392 | <result> = ashr i8 -2, 1 <i>; yields {i8}:result = -1</i> |
| Chris Lattner | 6ccc2d5 | 2007-10-03 21:01:14 +0000 | [diff] [blame] | 4393 | <result> = ashr i32 1, 32 <i>; undefined</i> |
| Mon P Wang | e9f1015 | 2008-12-09 05:46:39 +0000 | [diff] [blame] | 4394 | <result> = ashr <2 x i32> < i32 -2, i32 4>, < i32 1, i32 3> <i>; yields: result=<2 x i32> < i32 -1, i32 0></i> |
| Reid Spencer | 569f2fa | 2007-01-31 21:39:12 +0000 | [diff] [blame] | 4395 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4396 | |
| Reid Spencer | 569f2fa | 2007-01-31 21:39:12 +0000 | [diff] [blame] | 4397 | </div> |
| 4398 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4399 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4400 | <h4> |
| 4401 | <a name="i_and">'<tt>and</tt>' Instruction</a> |
| 4402 | </h4> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4403 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 4404 | <div> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4405 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4406 | <h5>Syntax:</h5> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4407 | <pre> |
| Gabor Greif | fb224a2 | 2008-08-07 21:46:00 +0000 | [diff] [blame] | 4408 | <result> = and <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4409 | </pre> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4410 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4411 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4412 | <p>The '<tt>and</tt>' instruction returns the bitwise logical and of its two |
| 4413 | operands.</p> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4414 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4415 | <h5>Arguments:</h5> |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 4416 | <p>The two arguments to the '<tt>and</tt>' instruction must be |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4417 | <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer |
| 4418 | values. Both arguments must have identical types.</p> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4419 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4420 | <h5>Semantics:</h5> |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 4421 | <p>The truth table used for the '<tt>and</tt>' instruction is:</p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4422 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 4423 | <table border="1" cellspacing="0" cellpadding="4"> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 4424 | <tbody> |
| 4425 | <tr> |
| Bill Wendling | 262396b | 2011-12-09 22:41:40 +0000 | [diff] [blame] | 4426 | <th>In0</th> |
| 4427 | <th>In1</th> |
| 4428 | <th>Out</th> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 4429 | </tr> |
| 4430 | <tr> |
| 4431 | <td>0</td> |
| 4432 | <td>0</td> |
| 4433 | <td>0</td> |
| 4434 | </tr> |
| 4435 | <tr> |
| 4436 | <td>0</td> |
| 4437 | <td>1</td> |
| 4438 | <td>0</td> |
| 4439 | </tr> |
| 4440 | <tr> |
| 4441 | <td>1</td> |
| 4442 | <td>0</td> |
| 4443 | <td>0</td> |
| 4444 | </tr> |
| 4445 | <tr> |
| 4446 | <td>1</td> |
| 4447 | <td>1</td> |
| 4448 | <td>1</td> |
| 4449 | </tr> |
| 4450 | </tbody> |
| 4451 | </table> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4452 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4453 | <h5>Example:</h5> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4454 | <pre> |
| 4455 | <result> = and i32 4, %var <i>; yields {i32}:result = 4 & %var</i> |
| Reid Spencer | ca86e16 | 2006-12-31 07:07:53 +0000 | [diff] [blame] | 4456 | <result> = and i32 15, 40 <i>; yields {i32}:result = 8</i> |
| 4457 | <result> = and i32 4, 8 <i>; yields {i32}:result = 0</i> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4458 | </pre> |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 4459 | </div> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4460 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4461 | <h4> |
| 4462 | <a name="i_or">'<tt>or</tt>' Instruction</a> |
| 4463 | </h4> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4464 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 4465 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4466 | |
| 4467 | <h5>Syntax:</h5> |
| 4468 | <pre> |
| 4469 | <result> = or <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| 4470 | </pre> |
| 4471 | |
| 4472 | <h5>Overview:</h5> |
| 4473 | <p>The '<tt>or</tt>' instruction returns the bitwise logical inclusive or of its |
| 4474 | two operands.</p> |
| 4475 | |
| 4476 | <h5>Arguments:</h5> |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 4477 | <p>The two arguments to the '<tt>or</tt>' instruction must be |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4478 | <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer |
| 4479 | values. Both arguments must have identical types.</p> |
| 4480 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4481 | <h5>Semantics:</h5> |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 4482 | <p>The truth table used for the '<tt>or</tt>' instruction is:</p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4483 | |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 4484 | <table border="1" cellspacing="0" cellpadding="4"> |
| 4485 | <tbody> |
| 4486 | <tr> |
| Bill Wendling | 262396b | 2011-12-09 22:41:40 +0000 | [diff] [blame] | 4487 | <th>In0</th> |
| 4488 | <th>In1</th> |
| 4489 | <th>Out</th> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 4490 | </tr> |
| 4491 | <tr> |
| 4492 | <td>0</td> |
| 4493 | <td>0</td> |
| 4494 | <td>0</td> |
| 4495 | </tr> |
| 4496 | <tr> |
| 4497 | <td>0</td> |
| 4498 | <td>1</td> |
| 4499 | <td>1</td> |
| 4500 | </tr> |
| 4501 | <tr> |
| 4502 | <td>1</td> |
| 4503 | <td>0</td> |
| 4504 | <td>1</td> |
| 4505 | </tr> |
| 4506 | <tr> |
| 4507 | <td>1</td> |
| 4508 | <td>1</td> |
| 4509 | <td>1</td> |
| 4510 | </tr> |
| 4511 | </tbody> |
| 4512 | </table> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4513 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4514 | <h5>Example:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4515 | <pre> |
| 4516 | <result> = or i32 4, %var <i>; yields {i32}:result = 4 | %var</i> |
| Reid Spencer | ca86e16 | 2006-12-31 07:07:53 +0000 | [diff] [blame] | 4517 | <result> = or i32 15, 40 <i>; yields {i32}:result = 47</i> |
| 4518 | <result> = or i32 4, 8 <i>; yields {i32}:result = 12</i> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4519 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4520 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 4521 | </div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4522 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4523 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4524 | <h4> |
| 4525 | <a name="i_xor">'<tt>xor</tt>' Instruction</a> |
| 4526 | </h4> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4527 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 4528 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4529 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4530 | <h5>Syntax:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4531 | <pre> |
| 4532 | <result> = xor <ty> <op1>, <op2> <i>; yields {ty}:result</i> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4533 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4534 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4535 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4536 | <p>The '<tt>xor</tt>' instruction returns the bitwise logical exclusive or of |
| 4537 | its two operands. The <tt>xor</tt> is used to implement the "one's |
| 4538 | complement" operation, which is the "~" operator in C.</p> |
| 4539 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4540 | <h5>Arguments:</h5> |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 4541 | <p>The two arguments to the '<tt>xor</tt>' instruction must be |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4542 | <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer |
| 4543 | values. Both arguments must have identical types.</p> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 4544 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4545 | <h5>Semantics:</h5> |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 4546 | <p>The truth table used for the '<tt>xor</tt>' instruction is:</p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4547 | |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 4548 | <table border="1" cellspacing="0" cellpadding="4"> |
| 4549 | <tbody> |
| 4550 | <tr> |
| Bill Wendling | 262396b | 2011-12-09 22:41:40 +0000 | [diff] [blame] | 4551 | <th>In0</th> |
| 4552 | <th>In1</th> |
| 4553 | <th>Out</th> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 4554 | </tr> |
| 4555 | <tr> |
| 4556 | <td>0</td> |
| 4557 | <td>0</td> |
| 4558 | <td>0</td> |
| 4559 | </tr> |
| 4560 | <tr> |
| 4561 | <td>0</td> |
| 4562 | <td>1</td> |
| 4563 | <td>1</td> |
| 4564 | </tr> |
| 4565 | <tr> |
| 4566 | <td>1</td> |
| 4567 | <td>0</td> |
| 4568 | <td>1</td> |
| 4569 | </tr> |
| 4570 | <tr> |
| 4571 | <td>1</td> |
| 4572 | <td>1</td> |
| 4573 | <td>0</td> |
| 4574 | </tr> |
| 4575 | </tbody> |
| 4576 | </table> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4577 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4578 | <h5>Example:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4579 | <pre> |
| 4580 | <result> = xor i32 4, %var <i>; yields {i32}:result = 4 ^ %var</i> |
| Reid Spencer | ca86e16 | 2006-12-31 07:07:53 +0000 | [diff] [blame] | 4581 | <result> = xor i32 15, 40 <i>; yields {i32}:result = 39</i> |
| 4582 | <result> = xor i32 4, 8 <i>; yields {i32}:result = 12</i> |
| 4583 | <result> = xor i32 %V, -1 <i>; yields {i32}:result = ~%V</i> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4584 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4585 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 4586 | </div> |
| Chris Lattner | 2cbdc45 | 2005-11-06 08:02:57 +0000 | [diff] [blame] | 4587 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 4588 | </div> |
| 4589 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4590 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4591 | <h3> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4592 | <a name="vectorops">Vector Operations</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4593 | </h3> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4594 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 4595 | <div> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4596 | |
| 4597 | <p>LLVM supports several instructions to represent vector operations in a |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4598 | target-independent manner. These instructions cover the element-access and |
| 4599 | vector-specific operations needed to process vectors effectively. While LLVM |
| 4600 | does directly support these vector operations, many sophisticated algorithms |
| 4601 | will want to use target-specific intrinsics to take full advantage of a |
| 4602 | specific target.</p> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4603 | |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4604 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4605 | <h4> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4606 | <a name="i_extractelement">'<tt>extractelement</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4607 | </h4> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4608 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 4609 | <div> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4610 | |
| 4611 | <h5>Syntax:</h5> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4612 | <pre> |
| Reid Spencer | ca86e16 | 2006-12-31 07:07:53 +0000 | [diff] [blame] | 4613 | <result> = extractelement <n x <ty>> <val>, i32 <idx> <i>; yields <ty></i> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4614 | </pre> |
| 4615 | |
| 4616 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4617 | <p>The '<tt>extractelement</tt>' instruction extracts a single scalar element |
| 4618 | from a vector at a specified index.</p> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4619 | |
| 4620 | |
| 4621 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4622 | <p>The first operand of an '<tt>extractelement</tt>' instruction is a value |
| 4623 | of <a href="#t_vector">vector</a> type. The second operand is an index |
| 4624 | indicating the position from which to extract the element. The index may be |
| 4625 | a variable.</p> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4626 | |
| 4627 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4628 | <p>The result is a scalar of the same type as the element type of |
| 4629 | <tt>val</tt>. Its value is the value at position <tt>idx</tt> of |
| 4630 | <tt>val</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the |
| 4631 | results are undefined.</p> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4632 | |
| 4633 | <h5>Example:</h5> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4634 | <pre> |
| Gabor Greif | a5b6f45 | 2009-10-28 13:14:50 +0000 | [diff] [blame] | 4635 | <result> = extractelement <4 x i32> %vec, i32 0 <i>; yields i32</i> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4636 | </pre> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4637 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4638 | </div> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4639 | |
| 4640 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4641 | <h4> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4642 | <a name="i_insertelement">'<tt>insertelement</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4643 | </h4> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4644 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 4645 | <div> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4646 | |
| 4647 | <h5>Syntax:</h5> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4648 | <pre> |
| Dan Gohman | f3480b9 | 2008-05-12 23:38:42 +0000 | [diff] [blame] | 4649 | <result> = insertelement <n x <ty>> <val>, <ty> <elt>, i32 <idx> <i>; yields <n x <ty>></i> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4650 | </pre> |
| 4651 | |
| 4652 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4653 | <p>The '<tt>insertelement</tt>' instruction inserts a scalar element into a |
| 4654 | vector at a specified index.</p> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4655 | |
| 4656 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4657 | <p>The first operand of an '<tt>insertelement</tt>' instruction is a value |
| 4658 | of <a href="#t_vector">vector</a> type. The second operand is a scalar value |
| 4659 | whose type must equal the element type of the first operand. The third |
| 4660 | operand is an index indicating the position at which to insert the value. |
| 4661 | The index may be a variable.</p> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4662 | |
| 4663 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4664 | <p>The result is a vector of the same type as <tt>val</tt>. Its element values |
| 4665 | are those of <tt>val</tt> except at position <tt>idx</tt>, where it gets the |
| 4666 | value <tt>elt</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the |
| 4667 | results are undefined.</p> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4668 | |
| 4669 | <h5>Example:</h5> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4670 | <pre> |
| Gabor Greif | a5b6f45 | 2009-10-28 13:14:50 +0000 | [diff] [blame] | 4671 | <result> = insertelement <4 x i32> %vec, i32 1, i32 0 <i>; yields <4 x i32></i> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4672 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4673 | |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4674 | </div> |
| 4675 | |
| 4676 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4677 | <h4> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4678 | <a name="i_shufflevector">'<tt>shufflevector</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4679 | </h4> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4680 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 4681 | <div> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4682 | |
| 4683 | <h5>Syntax:</h5> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4684 | <pre> |
| Mon P Wang | aeb06d2 | 2008-11-10 04:46:22 +0000 | [diff] [blame] | 4685 | <result> = shufflevector <n x <ty>> <v1>, <n x <ty>> <v2>, <m x i32> <mask> <i>; yields <m x <ty>></i> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4686 | </pre> |
| 4687 | |
| 4688 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4689 | <p>The '<tt>shufflevector</tt>' instruction constructs a permutation of elements |
| 4690 | from two input vectors, returning a vector with the same element type as the |
| 4691 | input and length that is the same as the shuffle mask.</p> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4692 | |
| 4693 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4694 | <p>The first two operands of a '<tt>shufflevector</tt>' instruction are vectors |
| 4695 | with types that match each other. The third argument is a shuffle mask whose |
| 4696 | element type is always 'i32'. The result of the instruction is a vector |
| 4697 | whose length is the same as the shuffle mask and whose element type is the |
| 4698 | same as the element type of the first two operands.</p> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4699 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4700 | <p>The shuffle mask operand is required to be a constant vector with either |
| 4701 | constant integer or undef values.</p> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4702 | |
| 4703 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4704 | <p>The elements of the two input vectors are numbered from left to right across |
| 4705 | both of the vectors. The shuffle mask operand specifies, for each element of |
| 4706 | the result vector, which element of the two input vectors the result element |
| 4707 | gets. The element selector may be undef (meaning "don't care") and the |
| 4708 | second operand may be undef if performing a shuffle from only one vector.</p> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4709 | |
| 4710 | <h5>Example:</h5> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4711 | <pre> |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 4712 | <result> = shufflevector <4 x i32> %v1, <4 x i32> %v2, |
| Jeff Cohen | 6f1cc77 | 2007-04-22 01:17:39 +0000 | [diff] [blame] | 4713 | <4 x i32> <i32 0, i32 4, i32 1, i32 5> <i>; yields <4 x i32></i> |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 4714 | <result> = shufflevector <4 x i32> %v1, <4 x i32> undef, |
| Reid Spencer | ca86e16 | 2006-12-31 07:07:53 +0000 | [diff] [blame] | 4715 | <4 x i32> <i32 0, i32 1, i32 2, i32 3> <i>; yields <4 x i32></i> - Identity shuffle. |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 4716 | <result> = shufflevector <8 x i32> %v1, <8 x i32> undef, |
| Mon P Wang | aeb06d2 | 2008-11-10 04:46:22 +0000 | [diff] [blame] | 4717 | <4 x i32> <i32 0, i32 1, i32 2, i32 3> <i>; yields <4 x i32></i> |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 4718 | <result> = shufflevector <4 x i32> %v1, <4 x i32> %v2, |
| Mon P Wang | aeb06d2 | 2008-11-10 04:46:22 +0000 | [diff] [blame] | 4719 | <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7 > <i>; yields <8 x i32></i> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4720 | </pre> |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4721 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4722 | </div> |
| Tanya Lattner | 0947429 | 2006-04-14 19:24:33 +0000 | [diff] [blame] | 4723 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 4724 | </div> |
| 4725 | |
| Chris Lattner | 3df241e | 2006-04-08 23:07:04 +0000 | [diff] [blame] | 4726 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4727 | <h3> |
| Dan Gohman | a334d5f | 2008-05-12 23:51:09 +0000 | [diff] [blame] | 4728 | <a name="aggregateops">Aggregate Operations</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4729 | </h3> |
| Dan Gohman | a334d5f | 2008-05-12 23:51:09 +0000 | [diff] [blame] | 4730 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 4731 | <div> |
| Dan Gohman | a334d5f | 2008-05-12 23:51:09 +0000 | [diff] [blame] | 4732 | |
| Chris Lattner | fdfeb69 | 2010-02-12 20:49:41 +0000 | [diff] [blame] | 4733 | <p>LLVM supports several instructions for working with |
| 4734 | <a href="#t_aggregate">aggregate</a> values.</p> |
| Dan Gohman | a334d5f | 2008-05-12 23:51:09 +0000 | [diff] [blame] | 4735 | |
| Dan Gohman | a334d5f | 2008-05-12 23:51:09 +0000 | [diff] [blame] | 4736 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4737 | <h4> |
| Dan Gohman | a334d5f | 2008-05-12 23:51:09 +0000 | [diff] [blame] | 4738 | <a name="i_extractvalue">'<tt>extractvalue</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4739 | </h4> |
| Dan Gohman | a334d5f | 2008-05-12 23:51:09 +0000 | [diff] [blame] | 4740 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 4741 | <div> |
| Dan Gohman | a334d5f | 2008-05-12 23:51:09 +0000 | [diff] [blame] | 4742 | |
| 4743 | <h5>Syntax:</h5> |
| Dan Gohman | a334d5f | 2008-05-12 23:51:09 +0000 | [diff] [blame] | 4744 | <pre> |
| 4745 | <result> = extractvalue <aggregate type> <val>, <idx>{, <idx>}* |
| 4746 | </pre> |
| 4747 | |
| 4748 | <h5>Overview:</h5> |
| Chris Lattner | fdfeb69 | 2010-02-12 20:49:41 +0000 | [diff] [blame] | 4749 | <p>The '<tt>extractvalue</tt>' instruction extracts the value of a member field |
| 4750 | from an <a href="#t_aggregate">aggregate</a> value.</p> |
| Dan Gohman | a334d5f | 2008-05-12 23:51:09 +0000 | [diff] [blame] | 4751 | |
| 4752 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4753 | <p>The first operand of an '<tt>extractvalue</tt>' instruction is a value |
| Chris Lattner | 61c70e9 | 2010-08-28 04:09:24 +0000 | [diff] [blame] | 4754 | of <a href="#t_struct">struct</a> or |
| Chris Lattner | fdfeb69 | 2010-02-12 20:49:41 +0000 | [diff] [blame] | 4755 | <a href="#t_array">array</a> type. The operands are constant indices to |
| 4756 | specify which value to extract in a similar manner as indices in a |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4757 | '<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.</p> |
| Frits van Bommel | 1324289 | 2010-12-05 20:54:38 +0000 | [diff] [blame] | 4758 | <p>The major differences to <tt>getelementptr</tt> indexing are:</p> |
| 4759 | <ul> |
| 4760 | <li>Since the value being indexed is not a pointer, the first index is |
| 4761 | omitted and assumed to be zero.</li> |
| 4762 | <li>At least one index must be specified.</li> |
| 4763 | <li>Not only struct indices but also array indices must be in |
| 4764 | bounds.</li> |
| 4765 | </ul> |
| Dan Gohman | a334d5f | 2008-05-12 23:51:09 +0000 | [diff] [blame] | 4766 | |
| 4767 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4768 | <p>The result is the value at the position in the aggregate specified by the |
| 4769 | index operands.</p> |
| Dan Gohman | a334d5f | 2008-05-12 23:51:09 +0000 | [diff] [blame] | 4770 | |
| 4771 | <h5>Example:</h5> |
| Dan Gohman | a334d5f | 2008-05-12 23:51:09 +0000 | [diff] [blame] | 4772 | <pre> |
| Gabor Greif | a5b6f45 | 2009-10-28 13:14:50 +0000 | [diff] [blame] | 4773 | <result> = extractvalue {i32, float} %agg, 0 <i>; yields i32</i> |
| Dan Gohman | a334d5f | 2008-05-12 23:51:09 +0000 | [diff] [blame] | 4774 | </pre> |
| Dan Gohman | a334d5f | 2008-05-12 23:51:09 +0000 | [diff] [blame] | 4775 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4776 | </div> |
| Dan Gohman | a334d5f | 2008-05-12 23:51:09 +0000 | [diff] [blame] | 4777 | |
| 4778 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4779 | <h4> |
| Dan Gohman | a334d5f | 2008-05-12 23:51:09 +0000 | [diff] [blame] | 4780 | <a name="i_insertvalue">'<tt>insertvalue</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4781 | </h4> |
| Dan Gohman | a334d5f | 2008-05-12 23:51:09 +0000 | [diff] [blame] | 4782 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 4783 | <div> |
| Dan Gohman | a334d5f | 2008-05-12 23:51:09 +0000 | [diff] [blame] | 4784 | |
| 4785 | <h5>Syntax:</h5> |
| Dan Gohman | a334d5f | 2008-05-12 23:51:09 +0000 | [diff] [blame] | 4786 | <pre> |
| Bill Wendling | 194229e | 2011-07-26 20:42:28 +0000 | [diff] [blame] | 4787 | <result> = insertvalue <aggregate type> <val>, <ty> <elt>, <idx>{, <idx>}* <i>; yields <aggregate type></i> |
| Dan Gohman | a334d5f | 2008-05-12 23:51:09 +0000 | [diff] [blame] | 4788 | </pre> |
| 4789 | |
| 4790 | <h5>Overview:</h5> |
| Chris Lattner | fdfeb69 | 2010-02-12 20:49:41 +0000 | [diff] [blame] | 4791 | <p>The '<tt>insertvalue</tt>' instruction inserts a value into a member field |
| 4792 | in an <a href="#t_aggregate">aggregate</a> value.</p> |
| Dan Gohman | a334d5f | 2008-05-12 23:51:09 +0000 | [diff] [blame] | 4793 | |
| 4794 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4795 | <p>The first operand of an '<tt>insertvalue</tt>' instruction is a value |
| Chris Lattner | 61c70e9 | 2010-08-28 04:09:24 +0000 | [diff] [blame] | 4796 | of <a href="#t_struct">struct</a> or |
| Chris Lattner | fdfeb69 | 2010-02-12 20:49:41 +0000 | [diff] [blame] | 4797 | <a href="#t_array">array</a> type. The second operand is a first-class |
| 4798 | value to insert. The following operands are constant indices indicating |
| 4799 | the position at which to insert the value in a similar manner as indices in a |
| Frits van Bommel | 1324289 | 2010-12-05 20:54:38 +0000 | [diff] [blame] | 4800 | '<tt><a href="#i_extractvalue">extractvalue</a></tt>' instruction. The |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4801 | value to insert must have the same type as the value identified by the |
| 4802 | indices.</p> |
| Dan Gohman | a334d5f | 2008-05-12 23:51:09 +0000 | [diff] [blame] | 4803 | |
| 4804 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4805 | <p>The result is an aggregate of the same type as <tt>val</tt>. Its value is |
| 4806 | that of <tt>val</tt> except that the value at the position specified by the |
| 4807 | indices is that of <tt>elt</tt>.</p> |
| Dan Gohman | a334d5f | 2008-05-12 23:51:09 +0000 | [diff] [blame] | 4808 | |
| 4809 | <h5>Example:</h5> |
| Dan Gohman | a334d5f | 2008-05-12 23:51:09 +0000 | [diff] [blame] | 4810 | <pre> |
| Chris Lattner | 8645d1a | 2011-05-22 07:18:08 +0000 | [diff] [blame] | 4811 | %agg1 = insertvalue {i32, float} undef, i32 1, 0 <i>; yields {i32 1, float undef}</i> |
| 4812 | %agg2 = insertvalue {i32, float} %agg1, float %val, 1 <i>; yields {i32 1, float %val}</i> |
| 4813 | %agg3 = insertvalue {i32, {float}} %agg1, float %val, 1, 0 <i>; yields {i32 1, float %val}</i> |
| Dan Gohman | a334d5f | 2008-05-12 23:51:09 +0000 | [diff] [blame] | 4814 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4815 | |
| Dan Gohman | a334d5f | 2008-05-12 23:51:09 +0000 | [diff] [blame] | 4816 | </div> |
| 4817 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 4818 | </div> |
| Dan Gohman | a334d5f | 2008-05-12 23:51:09 +0000 | [diff] [blame] | 4819 | |
| 4820 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4821 | <h3> |
| Chris Lattner | 884a970 | 2006-08-15 00:45:58 +0000 | [diff] [blame] | 4822 | <a name="memoryops">Memory Access and Addressing Operations</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4823 | </h3> |
| Chris Lattner | 2cbdc45 | 2005-11-06 08:02:57 +0000 | [diff] [blame] | 4824 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 4825 | <div> |
| Chris Lattner | 2cbdc45 | 2005-11-06 08:02:57 +0000 | [diff] [blame] | 4826 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4827 | <p>A key design point of an SSA-based representation is how it represents |
| 4828 | memory. In LLVM, no memory locations are in SSA form, which makes things |
| Victor Hernandez | 2fee294 | 2009-10-26 23:44:29 +0000 | [diff] [blame] | 4829 | very simple. This section describes how to read, write, and allocate |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4830 | memory in LLVM.</p> |
| Chris Lattner | 2cbdc45 | 2005-11-06 08:02:57 +0000 | [diff] [blame] | 4831 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4832 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4833 | <h4> |
| Chris Lattner | 2cbdc45 | 2005-11-06 08:02:57 +0000 | [diff] [blame] | 4834 | <a name="i_alloca">'<tt>alloca</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4835 | </h4> |
| Chris Lattner | 2cbdc45 | 2005-11-06 08:02:57 +0000 | [diff] [blame] | 4836 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 4837 | <div> |
| Chris Lattner | 2cbdc45 | 2005-11-06 08:02:57 +0000 | [diff] [blame] | 4838 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4839 | <h5>Syntax:</h5> |
| Chris Lattner | 2cbdc45 | 2005-11-06 08:02:57 +0000 | [diff] [blame] | 4840 | <pre> |
| Dan Gohman | f75a7d3 | 2010-05-28 01:14:11 +0000 | [diff] [blame] | 4841 | <result> = alloca <type>[, <ty> <NumElements>][, align <alignment>] <i>; yields {type*}:result</i> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4842 | </pre> |
| Chris Lattner | 2cbdc45 | 2005-11-06 08:02:57 +0000 | [diff] [blame] | 4843 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4844 | <h5>Overview:</h5> |
| Jeff Cohen | 6f1cc77 | 2007-04-22 01:17:39 +0000 | [diff] [blame] | 4845 | <p>The '<tt>alloca</tt>' instruction allocates memory on the stack frame of the |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4846 | currently executing function, to be automatically released when this function |
| 4847 | returns to its caller. The object is always allocated in the generic address |
| 4848 | space (address space zero).</p> |
| Chris Lattner | 2cbdc45 | 2005-11-06 08:02:57 +0000 | [diff] [blame] | 4849 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4850 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4851 | <p>The '<tt>alloca</tt>' instruction |
| 4852 | allocates <tt>sizeof(<type>)*NumElements</tt> bytes of memory on the |
| 4853 | runtime stack, returning a pointer of the appropriate type to the program. |
| 4854 | If "NumElements" is specified, it is the number of elements allocated, |
| 4855 | otherwise "NumElements" is defaulted to be one. If a constant alignment is |
| 4856 | specified, the value result of the allocation is guaranteed to be aligned to |
| 4857 | at least that boundary. If not specified, or if zero, the target can choose |
| 4858 | to align the allocation on any convenient boundary compatible with the |
| 4859 | type.</p> |
| Chris Lattner | 2cbdc45 | 2005-11-06 08:02:57 +0000 | [diff] [blame] | 4860 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 4861 | <p>'<tt>type</tt>' may be any sized type.</p> |
| Chris Lattner | 2cbdc45 | 2005-11-06 08:02:57 +0000 | [diff] [blame] | 4862 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4863 | <h5>Semantics:</h5> |
| Bill Wendling | 871eb0a | 2009-05-08 20:49:29 +0000 | [diff] [blame] | 4864 | <p>Memory is allocated; a pointer is returned. The operation is undefined if |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4865 | there is insufficient stack space for the allocation. '<tt>alloca</tt>'d |
| 4866 | memory is automatically released when the function returns. The |
| 4867 | '<tt>alloca</tt>' instruction is commonly used to represent automatic |
| 4868 | variables that must have an address available. When the function returns |
| 4869 | (either with the <tt><a href="#i_ret">ret</a></tt> |
| Bill Wendling | 7b9e539 | 2012-02-06 21:57:33 +0000 | [diff] [blame] | 4870 | or <tt><a href="#i_resume">resume</a></tt> instructions), the memory is |
| Nick Lewycky | 84a1d23 | 2012-02-29 08:26:44 +0000 | [diff] [blame] | 4871 | reclaimed. Allocating zero bytes is legal, but the result is undefined. |
| 4872 | The order in which memory is allocated (ie., which way the stack grows) is |
| Nick Lewycky | 75d05e6 | 2012-03-18 09:35:50 +0000 | [diff] [blame^] | 4873 | not specified.</p> |
| Nick Lewycky | 84a1d23 | 2012-02-29 08:26:44 +0000 | [diff] [blame] | 4874 | |
| 4875 | <p> |
| Chris Lattner | 2cbdc45 | 2005-11-06 08:02:57 +0000 | [diff] [blame] | 4876 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4877 | <h5>Example:</h5> |
| Chris Lattner | 2cbdc45 | 2005-11-06 08:02:57 +0000 | [diff] [blame] | 4878 | <pre> |
| Dan Gohman | 81e2167 | 2009-01-04 23:49:44 +0000 | [diff] [blame] | 4879 | %ptr = alloca i32 <i>; yields {i32*}:ptr</i> |
| 4880 | %ptr = alloca i32, i32 4 <i>; yields {i32*}:ptr</i> |
| 4881 | %ptr = alloca i32, i32 4, align 1024 <i>; yields {i32*}:ptr</i> |
| 4882 | %ptr = alloca i32, align 1024 <i>; yields {i32*}:ptr</i> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4883 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4884 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 4885 | </div> |
| Chris Lattner | 2cbdc45 | 2005-11-06 08:02:57 +0000 | [diff] [blame] | 4886 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 4887 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4888 | <h4> |
| 4889 | <a name="i_load">'<tt>load</tt>' Instruction</a> |
| 4890 | </h4> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4891 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 4892 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4893 | |
| Chris Lattner | 2b7d320 | 2002-05-06 03:03:22 +0000 | [diff] [blame] | 4894 | <h5>Syntax:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4895 | <pre> |
| Pete Cooper | f95acc6 | 2012-02-10 18:13:54 +0000 | [diff] [blame] | 4896 | <result> = load [volatile] <ty>* <pointer>[, align <alignment>][, !nontemporal !<index>][, !invariant.load !<index>] |
| Eli Friedman | f03bb26 | 2011-08-12 22:50:01 +0000 | [diff] [blame] | 4897 | <result> = load atomic [volatile] <ty>* <pointer> [singlethread] <ordering>, align <alignment> |
| Bill Wendling | 7c78dbb | 2010-02-25 21:23:24 +0000 | [diff] [blame] | 4898 | !<index> = !{ i32 1 } |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4899 | </pre> |
| 4900 | |
| Chris Lattner | 2b7d320 | 2002-05-06 03:03:22 +0000 | [diff] [blame] | 4901 | <h5>Overview:</h5> |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 4902 | <p>The '<tt>load</tt>' instruction is used to read from memory.</p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4903 | |
| Chris Lattner | 2b7d320 | 2002-05-06 03:03:22 +0000 | [diff] [blame] | 4904 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4905 | <p>The argument to the '<tt>load</tt>' instruction specifies the memory address |
| 4906 | from which to load. The pointer must point to |
| 4907 | a <a href="#t_firstclass">first class</a> type. If the <tt>load</tt> is |
| 4908 | marked as <tt>volatile</tt>, then the optimizer is not allowed to modify the |
| Jeffrey Yasskin | 93e066d | 2010-04-26 21:21:24 +0000 | [diff] [blame] | 4909 | number or order of execution of this <tt>load</tt> with other <a |
| 4910 | href="#volatile">volatile operations</a>.</p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4911 | |
| Eli Friedman | 21006d4 | 2011-08-09 23:02:53 +0000 | [diff] [blame] | 4912 | <p>If the <code>load</code> is marked as <code>atomic</code>, it takes an extra |
| 4913 | <a href="#ordering">ordering</a> and optional <code>singlethread</code> |
| 4914 | argument. The <code>release</code> and <code>acq_rel</code> orderings are |
| 4915 | not valid on <code>load</code> instructions. Atomic loads produce <a |
| 4916 | href="#memorymodel">defined</a> results when they may see multiple atomic |
| 4917 | stores. The type of the pointee must be an integer type whose bit width |
| 4918 | is a power of two greater than or equal to eight and less than or equal |
| 4919 | to a target-specific size limit. <code>align</code> must be explicitly |
| 4920 | specified on atomic loads, and the load has undefined behavior if the |
| 4921 | alignment is not set to a value which is at least the size in bytes of |
| 4922 | the pointee. <code>!nontemporal</code> does not have any defined semantics |
| 4923 | for atomic loads.</p> |
| 4924 | |
| Bill Wendling | 7c78dbb | 2010-02-25 21:23:24 +0000 | [diff] [blame] | 4925 | <p>The optional constant <tt>align</tt> argument specifies the alignment of the |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4926 | operation (that is, the alignment of the memory address). A value of 0 or an |
| Bill Wendling | 7c78dbb | 2010-02-25 21:23:24 +0000 | [diff] [blame] | 4927 | omitted <tt>align</tt> argument means that the operation has the preferential |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4928 | alignment for the target. It is the responsibility of the code emitter to |
| 4929 | ensure that the alignment information is correct. Overestimating the |
| Bill Wendling | 7c78dbb | 2010-02-25 21:23:24 +0000 | [diff] [blame] | 4930 | alignment results in undefined behavior. Underestimating the alignment may |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4931 | produce less efficient code. An alignment of 1 is always safe.</p> |
| 4932 | |
| Bill Wendling | 7c78dbb | 2010-02-25 21:23:24 +0000 | [diff] [blame] | 4933 | <p>The optional <tt>!nontemporal</tt> metadata must reference a single |
| 4934 | metatadata name <index> corresponding to a metadata node with |
| Dan Gohman | b55a1ee | 2010-03-01 17:41:39 +0000 | [diff] [blame] | 4935 | one <tt>i32</tt> entry of value 1. The existence of |
| Bill Wendling | 7c78dbb | 2010-02-25 21:23:24 +0000 | [diff] [blame] | 4936 | the <tt>!nontemporal</tt> metatadata on the instruction tells the optimizer |
| 4937 | and code generator that this load is not expected to be reused in the cache. |
| 4938 | The code generator may select special instructions to save cache bandwidth, |
| Dan Gohman | b55a1ee | 2010-03-01 17:41:39 +0000 | [diff] [blame] | 4939 | such as the <tt>MOVNT</tt> instruction on x86.</p> |
| David Greene | 8939b0d | 2010-02-16 20:50:18 +0000 | [diff] [blame] | 4940 | |
| Pete Cooper | f95acc6 | 2012-02-10 18:13:54 +0000 | [diff] [blame] | 4941 | <p>The optional <tt>!invariant.load</tt> metadata must reference a single |
| 4942 | metatadata name <index> corresponding to a metadata node with no |
| 4943 | entries. The existence of the <tt>!invariant.load</tt> metatadata on the |
| 4944 | instruction tells the optimizer and code generator that this load address |
| 4945 | points to memory which does not change value during program execution. |
| 4946 | The optimizer may then move this load around, for example, by hoisting it |
| 4947 | out of loops using loop invariant code motion.</p> |
| 4948 | |
| Chris Lattner | 2b7d320 | 2002-05-06 03:03:22 +0000 | [diff] [blame] | 4949 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4950 | <p>The location of memory pointed to is loaded. If the value being loaded is of |
| 4951 | scalar type then the number of bytes read does not exceed the minimum number |
| 4952 | of bytes needed to hold all bits of the type. For example, loading an |
| 4953 | <tt>i24</tt> reads at most three bytes. When loading a value of a type like |
| 4954 | <tt>i20</tt> with a size that is not an integral number of bytes, the result |
| 4955 | is undefined if the value was not originally written using a store of the |
| 4956 | same type.</p> |
| 4957 | |
| Chris Lattner | 2b7d320 | 2002-05-06 03:03:22 +0000 | [diff] [blame] | 4958 | <h5>Examples:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4959 | <pre> |
| 4960 | %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i> |
| 4961 | <a href="#i_store">store</a> i32 3, i32* %ptr <i>; yields {void}</i> |
| Reid Spencer | ca86e16 | 2006-12-31 07:07:53 +0000 | [diff] [blame] | 4962 | %val = load i32* %ptr <i>; yields {i32}:val = i32 3</i> |
| Chris Lattner | 2b7d320 | 2002-05-06 03:03:22 +0000 | [diff] [blame] | 4963 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4964 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 4965 | </div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4966 | |
| Chris Lattner | 2b7d320 | 2002-05-06 03:03:22 +0000 | [diff] [blame] | 4967 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 4968 | <h4> |
| 4969 | <a name="i_store">'<tt>store</tt>' Instruction</a> |
| 4970 | </h4> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4971 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 4972 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4973 | |
| Chris Lattner | 2b7d320 | 2002-05-06 03:03:22 +0000 | [diff] [blame] | 4974 | <h5>Syntax:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4975 | <pre> |
| Bill Wendling | 262396b | 2011-12-09 22:41:40 +0000 | [diff] [blame] | 4976 | store [volatile] <ty> <value>, <ty>* <pointer>[, align <alignment>][, !nontemporal !<index>] <i>; yields {void}</i> |
| 4977 | store atomic [volatile] <ty> <value>, <ty>* <pointer> [singlethread] <ordering>, align <alignment> <i>; yields {void}</i> |
| Chris Lattner | 2b7d320 | 2002-05-06 03:03:22 +0000 | [diff] [blame] | 4978 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4979 | |
| Chris Lattner | 2b7d320 | 2002-05-06 03:03:22 +0000 | [diff] [blame] | 4980 | <h5>Overview:</h5> |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 4981 | <p>The '<tt>store</tt>' instruction is used to write to memory.</p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4982 | |
| Chris Lattner | 2b7d320 | 2002-05-06 03:03:22 +0000 | [diff] [blame] | 4983 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4984 | <p>There are two arguments to the '<tt>store</tt>' instruction: a value to store |
| 4985 | and an address at which to store it. The type of the |
| 4986 | '<tt><pointer></tt>' operand must be a pointer to |
| 4987 | the <a href="#t_firstclass">first class</a> type of the |
| Jeffrey Yasskin | 93e066d | 2010-04-26 21:21:24 +0000 | [diff] [blame] | 4988 | '<tt><value></tt>' operand. If the <tt>store</tt> is marked as |
| 4989 | <tt>volatile</tt>, then the optimizer is not allowed to modify the number or |
| 4990 | order of execution of this <tt>store</tt> with other <a |
| 4991 | href="#volatile">volatile operations</a>.</p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 4992 | |
| Eli Friedman | 21006d4 | 2011-08-09 23:02:53 +0000 | [diff] [blame] | 4993 | <p>If the <code>store</code> is marked as <code>atomic</code>, it takes an extra |
| 4994 | <a href="#ordering">ordering</a> and optional <code>singlethread</code> |
| 4995 | argument. The <code>acquire</code> and <code>acq_rel</code> orderings aren't |
| 4996 | valid on <code>store</code> instructions. Atomic loads produce <a |
| 4997 | href="#memorymodel">defined</a> results when they may see multiple atomic |
| 4998 | stores. The type of the pointee must be an integer type whose bit width |
| 4999 | is a power of two greater than or equal to eight and less than or equal |
| 5000 | to a target-specific size limit. <code>align</code> must be explicitly |
| 5001 | specified on atomic stores, and the store has undefined behavior if the |
| 5002 | alignment is not set to a value which is at least the size in bytes of |
| 5003 | the pointee. <code>!nontemporal</code> does not have any defined semantics |
| 5004 | for atomic stores.</p> |
| 5005 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5006 | <p>The optional constant "align" argument specifies the alignment of the |
| 5007 | operation (that is, the alignment of the memory address). A value of 0 or an |
| 5008 | omitted "align" argument means that the operation has the preferential |
| 5009 | alignment for the target. It is the responsibility of the code emitter to |
| 5010 | ensure that the alignment information is correct. Overestimating the |
| 5011 | alignment results in an undefined behavior. Underestimating the alignment may |
| 5012 | produce less efficient code. An alignment of 1 is always safe.</p> |
| 5013 | |
| David Greene | 8939b0d | 2010-02-16 20:50:18 +0000 | [diff] [blame] | 5014 | <p>The optional !nontemporal metadata must reference a single metatadata |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 5015 | name <index> corresponding to a metadata node with one i32 entry of |
| Dan Gohman | b55a1ee | 2010-03-01 17:41:39 +0000 | [diff] [blame] | 5016 | value 1. The existence of the !nontemporal metatadata on the |
| David Greene | 8939b0d | 2010-02-16 20:50:18 +0000 | [diff] [blame] | 5017 | instruction tells the optimizer and code generator that this load is |
| 5018 | not expected to be reused in the cache. The code generator may |
| 5019 | select special instructions to save cache bandwidth, such as the |
| Dan Gohman | b55a1ee | 2010-03-01 17:41:39 +0000 | [diff] [blame] | 5020 | MOVNT instruction on x86.</p> |
| David Greene | 8939b0d | 2010-02-16 20:50:18 +0000 | [diff] [blame] | 5021 | |
| 5022 | |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 5023 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5024 | <p>The contents of memory are updated to contain '<tt><value></tt>' at the |
| 5025 | location specified by the '<tt><pointer></tt>' operand. If |
| 5026 | '<tt><value></tt>' is of scalar type then the number of bytes written |
| 5027 | does not exceed the minimum number of bytes needed to hold all bits of the |
| 5028 | type. For example, storing an <tt>i24</tt> writes at most three bytes. When |
| 5029 | writing a value of a type like <tt>i20</tt> with a size that is not an |
| 5030 | integral number of bytes, it is unspecified what happens to the extra bits |
| 5031 | that do not belong to the type, but they will typically be overwritten.</p> |
| 5032 | |
| Chris Lattner | 2b7d320 | 2002-05-06 03:03:22 +0000 | [diff] [blame] | 5033 | <h5>Example:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5034 | <pre> |
| 5035 | %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i> |
| Bill Wendling | 8c6c72d | 2007-10-22 05:10:05 +0000 | [diff] [blame] | 5036 | store i32 3, i32* %ptr <i>; yields {void}</i> |
| 5037 | %val = <a href="#i_load">load</a> i32* %ptr <i>; yields {i32}:val = i32 3</i> |
| Chris Lattner | 2b7d320 | 2002-05-06 03:03:22 +0000 | [diff] [blame] | 5038 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5039 | |
| Reid Spencer | 47ce179 | 2006-11-09 21:15:49 +0000 | [diff] [blame] | 5040 | </div> |
| 5041 | |
| Chris Lattner | 2b7d320 | 2002-05-06 03:03:22 +0000 | [diff] [blame] | 5042 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | b5bb29c | 2011-08-12 06:17:17 +0000 | [diff] [blame] | 5043 | <h4> |
| 5044 | <a name="i_fence">'<tt>fence</tt>' Instruction</a> |
| 5045 | </h4> |
| Eli Friedman | 47f3513 | 2011-07-25 23:16:38 +0000 | [diff] [blame] | 5046 | |
| NAKAMURA Takumi | b5bb29c | 2011-08-12 06:17:17 +0000 | [diff] [blame] | 5047 | <div> |
| Eli Friedman | 47f3513 | 2011-07-25 23:16:38 +0000 | [diff] [blame] | 5048 | |
| 5049 | <h5>Syntax:</h5> |
| 5050 | <pre> |
| 5051 | fence [singlethread] <ordering> <i>; yields {void}</i> |
| 5052 | </pre> |
| 5053 | |
| 5054 | <h5>Overview:</h5> |
| 5055 | <p>The '<tt>fence</tt>' instruction is used to introduce happens-before edges |
| 5056 | between operations.</p> |
| 5057 | |
| 5058 | <h5>Arguments:</h5> <p>'<code>fence</code>' instructions take an <a |
| 5059 | href="#ordering">ordering</a> argument which defines what |
| 5060 | <i>synchronizes-with</i> edges they add. They can only be given |
| 5061 | <code>acquire</code>, <code>release</code>, <code>acq_rel</code>, and |
| 5062 | <code>seq_cst</code> orderings.</p> |
| 5063 | |
| 5064 | <h5>Semantics:</h5> |
| 5065 | <p>A fence <var>A</var> which has (at least) <code>release</code> ordering |
| 5066 | semantics <i>synchronizes with</i> a fence <var>B</var> with (at least) |
| 5067 | <code>acquire</code> ordering semantics if and only if there exist atomic |
| 5068 | operations <var>X</var> and <var>Y</var>, both operating on some atomic object |
| 5069 | <var>M</var>, such that <var>A</var> is sequenced before <var>X</var>, |
| 5070 | <var>X</var> modifies <var>M</var> (either directly or through some side effect |
| 5071 | of a sequence headed by <var>X</var>), <var>Y</var> is sequenced before |
| 5072 | <var>B</var>, and <var>Y</var> observes <var>M</var>. This provides a |
| 5073 | <i>happens-before</i> dependency between <var>A</var> and <var>B</var>. Rather |
| 5074 | than an explicit <code>fence</code>, one (but not both) of the atomic operations |
| 5075 | <var>X</var> or <var>Y</var> might provide a <code>release</code> or |
| 5076 | <code>acquire</code> (resp.) ordering constraint and still |
| 5077 | <i>synchronize-with</i> the explicit <code>fence</code> and establish the |
| 5078 | <i>happens-before</i> edge.</p> |
| 5079 | |
| 5080 | <p>A <code>fence</code> which has <code>seq_cst</code> ordering, in addition to |
| 5081 | having both <code>acquire</code> and <code>release</code> semantics specified |
| 5082 | above, participates in the global program order of other <code>seq_cst</code> |
| 5083 | operations and/or fences.</p> |
| 5084 | |
| 5085 | <p>The optional "<a href="#singlethread"><code>singlethread</code></a>" argument |
| 5086 | specifies that the fence only synchronizes with other fences in the same |
| 5087 | thread. (This is useful for interacting with signal handlers.)</p> |
| 5088 | |
| Eli Friedman | 47f3513 | 2011-07-25 23:16:38 +0000 | [diff] [blame] | 5089 | <h5>Example:</h5> |
| 5090 | <pre> |
| 5091 | fence acquire <i>; yields {void}</i> |
| 5092 | fence singlethread seq_cst <i>; yields {void}</i> |
| 5093 | </pre> |
| 5094 | |
| 5095 | </div> |
| 5096 | |
| 5097 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | b5bb29c | 2011-08-12 06:17:17 +0000 | [diff] [blame] | 5098 | <h4> |
| 5099 | <a name="i_cmpxchg">'<tt>cmpxchg</tt>' Instruction</a> |
| 5100 | </h4> |
| Eli Friedman | ff03048 | 2011-07-28 21:48:00 +0000 | [diff] [blame] | 5101 | |
| NAKAMURA Takumi | b5bb29c | 2011-08-12 06:17:17 +0000 | [diff] [blame] | 5102 | <div> |
| Eli Friedman | ff03048 | 2011-07-28 21:48:00 +0000 | [diff] [blame] | 5103 | |
| 5104 | <h5>Syntax:</h5> |
| 5105 | <pre> |
| Bill Wendling | 262396b | 2011-12-09 22:41:40 +0000 | [diff] [blame] | 5106 | cmpxchg [volatile] <ty>* <pointer>, <ty> <cmp>, <ty> <new> [singlethread] <ordering> <i>; yields {ty}</i> |
| Eli Friedman | ff03048 | 2011-07-28 21:48:00 +0000 | [diff] [blame] | 5107 | </pre> |
| 5108 | |
| 5109 | <h5>Overview:</h5> |
| 5110 | <p>The '<tt>cmpxchg</tt>' instruction is used to atomically modify memory. |
| 5111 | It loads a value in memory and compares it to a given value. If they are |
| 5112 | equal, it stores a new value into the memory.</p> |
| 5113 | |
| 5114 | <h5>Arguments:</h5> |
| 5115 | <p>There are three arguments to the '<code>cmpxchg</code>' instruction: an |
| 5116 | address to operate on, a value to compare to the value currently be at that |
| 5117 | address, and a new value to place at that address if the compared values are |
| 5118 | equal. The type of '<var><cmp></var>' must be an integer type whose |
| 5119 | bit width is a power of two greater than or equal to eight and less than |
| 5120 | or equal to a target-specific size limit. '<var><cmp></var>' and |
| 5121 | '<var><new></var>' must have the same type, and the type of |
| 5122 | '<var><pointer></var>' must be a pointer to that type. If the |
| 5123 | <code>cmpxchg</code> is marked as <code>volatile</code>, then the |
| 5124 | optimizer is not allowed to modify the number or order of execution |
| 5125 | of this <code>cmpxchg</code> with other <a href="#volatile">volatile |
| 5126 | operations</a>.</p> |
| 5127 | |
| 5128 | <!-- FIXME: Extend allowed types. --> |
| 5129 | |
| 5130 | <p>The <a href="#ordering"><var>ordering</var></a> argument specifies how this |
| 5131 | <code>cmpxchg</code> synchronizes with other atomic operations.</p> |
| 5132 | |
| 5133 | <p>The optional "<code>singlethread</code>" argument declares that the |
| 5134 | <code>cmpxchg</code> is only atomic with respect to code (usually signal |
| 5135 | handlers) running in the same thread as the <code>cmpxchg</code>. Otherwise the |
| 5136 | cmpxchg is atomic with respect to all other code in the system.</p> |
| 5137 | |
| 5138 | <p>The pointer passed into cmpxchg must have alignment greater than or equal to |
| 5139 | the size in memory of the operand. |
| 5140 | |
| 5141 | <h5>Semantics:</h5> |
| 5142 | <p>The contents of memory at the location specified by the |
| 5143 | '<tt><pointer></tt>' operand is read and compared to |
| 5144 | '<tt><cmp></tt>'; if the read value is the equal, |
| 5145 | '<tt><new></tt>' is written. The original value at the location |
| 5146 | is returned. |
| 5147 | |
| 5148 | <p>A successful <code>cmpxchg</code> is a read-modify-write instruction for the |
| 5149 | purpose of identifying <a href="#release_sequence">release sequences</a>. A |
| 5150 | failed <code>cmpxchg</code> is equivalent to an atomic load with an ordering |
| 5151 | parameter determined by dropping any <code>release</code> part of the |
| 5152 | <code>cmpxchg</code>'s ordering.</p> |
| 5153 | |
| 5154 | <!-- |
| 5155 | FIXME: Is compare_exchange_weak() necessary? (Consider after we've done |
| 5156 | optimization work on ARM.) |
| 5157 | |
| 5158 | FIXME: Is a weaker ordering constraint on failure helpful in practice? |
| 5159 | --> |
| 5160 | |
| 5161 | <h5>Example:</h5> |
| 5162 | <pre> |
| 5163 | entry: |
| Bill Wendling | 262396b | 2011-12-09 22:41:40 +0000 | [diff] [blame] | 5164 | %orig = atomic <a href="#i_load">load</a> i32* %ptr unordered <i>; yields {i32}</i> |
| Eli Friedman | ff03048 | 2011-07-28 21:48:00 +0000 | [diff] [blame] | 5165 | <a href="#i_br">br</a> label %loop |
| 5166 | |
| 5167 | loop: |
| 5168 | %cmp = <a href="#i_phi">phi</a> i32 [ %orig, %entry ], [%old, %loop] |
| 5169 | %squared = <a href="#i_mul">mul</a> i32 %cmp, %cmp |
| Bill Wendling | 262396b | 2011-12-09 22:41:40 +0000 | [diff] [blame] | 5170 | %old = cmpxchg i32* %ptr, i32 %cmp, i32 %squared <i>; yields {i32}</i> |
| Eli Friedman | ff03048 | 2011-07-28 21:48:00 +0000 | [diff] [blame] | 5171 | %success = <a href="#i_icmp">icmp</a> eq i32 %cmp, %old |
| 5172 | <a href="#i_br">br</a> i1 %success, label %done, label %loop |
| 5173 | |
| 5174 | done: |
| 5175 | ... |
| 5176 | </pre> |
| 5177 | |
| 5178 | </div> |
| 5179 | |
| 5180 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | b5bb29c | 2011-08-12 06:17:17 +0000 | [diff] [blame] | 5181 | <h4> |
| 5182 | <a name="i_atomicrmw">'<tt>atomicrmw</tt>' Instruction</a> |
| 5183 | </h4> |
| Eli Friedman | ff03048 | 2011-07-28 21:48:00 +0000 | [diff] [blame] | 5184 | |
| NAKAMURA Takumi | b5bb29c | 2011-08-12 06:17:17 +0000 | [diff] [blame] | 5185 | <div> |
| Eli Friedman | ff03048 | 2011-07-28 21:48:00 +0000 | [diff] [blame] | 5186 | |
| 5187 | <h5>Syntax:</h5> |
| 5188 | <pre> |
| Eli Friedman | f03bb26 | 2011-08-12 22:50:01 +0000 | [diff] [blame] | 5189 | atomicrmw [volatile] <operation> <ty>* <pointer>, <ty> <value> [singlethread] <ordering> <i>; yields {ty}</i> |
| Eli Friedman | ff03048 | 2011-07-28 21:48:00 +0000 | [diff] [blame] | 5190 | </pre> |
| 5191 | |
| 5192 | <h5>Overview:</h5> |
| 5193 | <p>The '<tt>atomicrmw</tt>' instruction is used to atomically modify memory.</p> |
| 5194 | |
| 5195 | <h5>Arguments:</h5> |
| 5196 | <p>There are three arguments to the '<code>atomicrmw</code>' instruction: an |
| 5197 | operation to apply, an address whose value to modify, an argument to the |
| 5198 | operation. The operation must be one of the following keywords:</p> |
| 5199 | <ul> |
| 5200 | <li>xchg</li> |
| 5201 | <li>add</li> |
| 5202 | <li>sub</li> |
| 5203 | <li>and</li> |
| 5204 | <li>nand</li> |
| 5205 | <li>or</li> |
| 5206 | <li>xor</li> |
| 5207 | <li>max</li> |
| 5208 | <li>min</li> |
| 5209 | <li>umax</li> |
| 5210 | <li>umin</li> |
| 5211 | </ul> |
| 5212 | |
| 5213 | <p>The type of '<var><value></var>' must be an integer type whose |
| 5214 | bit width is a power of two greater than or equal to eight and less than |
| 5215 | or equal to a target-specific size limit. The type of the |
| 5216 | '<code><pointer></code>' operand must be a pointer to that type. |
| 5217 | If the <code>atomicrmw</code> is marked as <code>volatile</code>, then the |
| 5218 | optimizer is not allowed to modify the number or order of execution of this |
| 5219 | <code>atomicrmw</code> with other <a href="#volatile">volatile |
| 5220 | operations</a>.</p> |
| 5221 | |
| 5222 | <!-- FIXME: Extend allowed types. --> |
| 5223 | |
| 5224 | <h5>Semantics:</h5> |
| 5225 | <p>The contents of memory at the location specified by the |
| 5226 | '<tt><pointer></tt>' operand are atomically read, modified, and written |
| 5227 | back. The original value at the location is returned. The modification is |
| 5228 | specified by the <var>operation</var> argument:</p> |
| 5229 | |
| 5230 | <ul> |
| 5231 | <li>xchg: <code>*ptr = val</code></li> |
| 5232 | <li>add: <code>*ptr = *ptr + val</code></li> |
| 5233 | <li>sub: <code>*ptr = *ptr - val</code></li> |
| 5234 | <li>and: <code>*ptr = *ptr & val</code></li> |
| 5235 | <li>nand: <code>*ptr = ~(*ptr & val)</code></li> |
| 5236 | <li>or: <code>*ptr = *ptr | val</code></li> |
| 5237 | <li>xor: <code>*ptr = *ptr ^ val</code></li> |
| 5238 | <li>max: <code>*ptr = *ptr > val ? *ptr : val</code> (using a signed comparison)</li> |
| 5239 | <li>min: <code>*ptr = *ptr < val ? *ptr : val</code> (using a signed comparison)</li> |
| 5240 | <li>umax: <code>*ptr = *ptr > val ? *ptr : val</code> (using an unsigned comparison)</li> |
| 5241 | <li>umin: <code>*ptr = *ptr < val ? *ptr : val</code> (using an unsigned comparison)</li> |
| 5242 | </ul> |
| 5243 | |
| 5244 | <h5>Example:</h5> |
| 5245 | <pre> |
| 5246 | %old = atomicrmw add i32* %ptr, i32 1 acquire <i>; yields {i32}</i> |
| 5247 | </pre> |
| 5248 | |
| 5249 | </div> |
| 5250 | |
| 5251 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 5252 | <h4> |
| Chris Lattner | f74d5c7 | 2004-04-05 01:30:49 +0000 | [diff] [blame] | 5253 | <a name="i_getelementptr">'<tt>getelementptr</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 5254 | </h4> |
| Chris Lattner | f74d5c7 | 2004-04-05 01:30:49 +0000 | [diff] [blame] | 5255 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 5256 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5257 | |
| Chris Lattner | 7faa883 | 2002-04-14 06:13:44 +0000 | [diff] [blame] | 5258 | <h5>Syntax:</h5> |
| Chris Lattner | f74d5c7 | 2004-04-05 01:30:49 +0000 | [diff] [blame] | 5259 | <pre> |
| Matthijs Kooijman | e49d0bc | 2008-10-13 13:44:15 +0000 | [diff] [blame] | 5260 | <result> = getelementptr <pty>* <ptrval>{, <ty> <idx>}* |
| Dan Gohman | dd8004d | 2009-07-27 21:53:46 +0000 | [diff] [blame] | 5261 | <result> = getelementptr inbounds <pty>* <ptrval>{, <ty> <idx>}* |
| Nadav Rotem | 1608769 | 2011-12-05 06:29:09 +0000 | [diff] [blame] | 5262 | <result> = getelementptr <ptr vector> ptrval, <vector index type> idx |
| Chris Lattner | f74d5c7 | 2004-04-05 01:30:49 +0000 | [diff] [blame] | 5263 | </pre> |
| 5264 | |
| Chris Lattner | 7faa883 | 2002-04-14 06:13:44 +0000 | [diff] [blame] | 5265 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5266 | <p>The '<tt>getelementptr</tt>' instruction is used to get the address of a |
| Chris Lattner | fdfeb69 | 2010-02-12 20:49:41 +0000 | [diff] [blame] | 5267 | subelement of an <a href="#t_aggregate">aggregate</a> data structure. |
| 5268 | It performs address calculation only and does not access memory.</p> |
| Chris Lattner | f74d5c7 | 2004-04-05 01:30:49 +0000 | [diff] [blame] | 5269 | |
| Chris Lattner | 7faa883 | 2002-04-14 06:13:44 +0000 | [diff] [blame] | 5270 | <h5>Arguments:</h5> |
| Nadav Rotem | 1608769 | 2011-12-05 06:29:09 +0000 | [diff] [blame] | 5271 | <p>The first argument is always a pointer or a vector of pointers, |
| 5272 | and forms the basis of the |
| Chris Lattner | c8eef44 | 2009-07-29 06:44:13 +0000 | [diff] [blame] | 5273 | calculation. The remaining arguments are indices that indicate which of the |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5274 | elements of the aggregate object are indexed. The interpretation of each |
| 5275 | index is dependent on the type being indexed into. The first index always |
| 5276 | indexes the pointer value given as the first argument, the second index |
| 5277 | indexes a value of the type pointed to (not necessarily the value directly |
| 5278 | pointed to, since the first index can be non-zero), etc. The first type |
| Chris Lattner | fdfeb69 | 2010-02-12 20:49:41 +0000 | [diff] [blame] | 5279 | indexed into must be a pointer value, subsequent types can be arrays, |
| Chris Lattner | 61c70e9 | 2010-08-28 04:09:24 +0000 | [diff] [blame] | 5280 | vectors, and structs. Note that subsequent types being indexed into |
| Chris Lattner | fdfeb69 | 2010-02-12 20:49:41 +0000 | [diff] [blame] | 5281 | can never be pointers, since that would require loading the pointer before |
| 5282 | continuing calculation.</p> |
| Matthijs Kooijman | e49d0bc | 2008-10-13 13:44:15 +0000 | [diff] [blame] | 5283 | |
| 5284 | <p>The type of each index argument depends on the type it is indexing into. |
| Chris Lattner | 61c70e9 | 2010-08-28 04:09:24 +0000 | [diff] [blame] | 5285 | When indexing into a (optionally packed) structure, only <tt>i32</tt> |
| Chris Lattner | fdfeb69 | 2010-02-12 20:49:41 +0000 | [diff] [blame] | 5286 | integer <b>constants</b> are allowed. When indexing into an array, pointer |
| 5287 | or vector, integers of any width are allowed, and they are not required to be |
| Eli Friedman | 266246c | 2011-08-12 23:37:55 +0000 | [diff] [blame] | 5288 | constant. These integers are treated as signed values where relevant.</p> |
| Chris Lattner | f74d5c7 | 2004-04-05 01:30:49 +0000 | [diff] [blame] | 5289 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5290 | <p>For example, let's consider a C code fragment and how it gets compiled to |
| 5291 | LLVM:</p> |
| Chris Lattner | f74d5c7 | 2004-04-05 01:30:49 +0000 | [diff] [blame] | 5292 | |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 5293 | <pre class="doc_code"> |
| Bill Wendling | 2f7a8b0 | 2007-05-29 09:04:49 +0000 | [diff] [blame] | 5294 | struct RT { |
| 5295 | char A; |
| Chris Lattner | cabc846 | 2007-05-29 15:43:56 +0000 | [diff] [blame] | 5296 | int B[10][20]; |
| Bill Wendling | 2f7a8b0 | 2007-05-29 09:04:49 +0000 | [diff] [blame] | 5297 | char C; |
| 5298 | }; |
| 5299 | struct ST { |
| Chris Lattner | cabc846 | 2007-05-29 15:43:56 +0000 | [diff] [blame] | 5300 | int X; |
| Bill Wendling | 2f7a8b0 | 2007-05-29 09:04:49 +0000 | [diff] [blame] | 5301 | double Y; |
| 5302 | struct RT Z; |
| 5303 | }; |
| Chris Lattner | f74d5c7 | 2004-04-05 01:30:49 +0000 | [diff] [blame] | 5304 | |
| Chris Lattner | cabc846 | 2007-05-29 15:43:56 +0000 | [diff] [blame] | 5305 | int *foo(struct ST *s) { |
| Bill Wendling | 2f7a8b0 | 2007-05-29 09:04:49 +0000 | [diff] [blame] | 5306 | return &s[1].Z.B[5][13]; |
| 5307 | } |
| Chris Lattner | f74d5c7 | 2004-04-05 01:30:49 +0000 | [diff] [blame] | 5308 | </pre> |
| 5309 | |
| Bill Wendling | a349539 | 2011-12-13 01:07:07 +0000 | [diff] [blame] | 5310 | <p>The LLVM code generated by Clang is:</p> |
| Chris Lattner | f74d5c7 | 2004-04-05 01:30:49 +0000 | [diff] [blame] | 5311 | |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 5312 | <pre class="doc_code"> |
| Bill Wendling | a349539 | 2011-12-13 01:07:07 +0000 | [diff] [blame] | 5313 | %struct.RT = <a href="#namedtypes">type</a> { i8, [10 x [20 x i32]], i8 } |
| 5314 | %struct.ST = <a href="#namedtypes">type</a> { i32, double, %struct.RT } |
| Chris Lattner | f74d5c7 | 2004-04-05 01:30:49 +0000 | [diff] [blame] | 5315 | |
| Bill Wendling | a349539 | 2011-12-13 01:07:07 +0000 | [diff] [blame] | 5316 | define i32* @foo(%struct.ST* %s) nounwind uwtable readnone optsize ssp { |
| Bill Wendling | 2f7a8b0 | 2007-05-29 09:04:49 +0000 | [diff] [blame] | 5317 | entry: |
| Bill Wendling | a349539 | 2011-12-13 01:07:07 +0000 | [diff] [blame] | 5318 | %arrayidx = getelementptr inbounds %struct.ST* %s, i64 1, i32 2, i32 1, i64 5, i64 13 |
| 5319 | ret i32* %arrayidx |
| Bill Wendling | 2f7a8b0 | 2007-05-29 09:04:49 +0000 | [diff] [blame] | 5320 | } |
| Chris Lattner | f74d5c7 | 2004-04-05 01:30:49 +0000 | [diff] [blame] | 5321 | </pre> |
| 5322 | |
| Chris Lattner | 7faa883 | 2002-04-14 06:13:44 +0000 | [diff] [blame] | 5323 | <h5>Semantics:</h5> |
| Bill Wendling | a349539 | 2011-12-13 01:07:07 +0000 | [diff] [blame] | 5324 | <p>In the example above, the first index is indexing into the |
| 5325 | '<tt>%struct.ST*</tt>' type, which is a pointer, yielding a |
| 5326 | '<tt>%struct.ST</tt>' = '<tt>{ i32, double, %struct.RT }</tt>' type, a |
| 5327 | structure. The second index indexes into the third element of the structure, |
| 5328 | yielding a '<tt>%struct.RT</tt>' = '<tt>{ i8 , [10 x [20 x i32]], i8 }</tt>' |
| 5329 | type, another structure. The third index indexes into the second element of |
| 5330 | the structure, yielding a '<tt>[10 x [20 x i32]]</tt>' type, an array. The |
| 5331 | two dimensions of the array are subscripted into, yielding an '<tt>i32</tt>' |
| 5332 | type. The '<tt>getelementptr</tt>' instruction returns a pointer to this |
| 5333 | element, thus computing a value of '<tt>i32*</tt>' type.</p> |
| Chris Lattner | f74d5c7 | 2004-04-05 01:30:49 +0000 | [diff] [blame] | 5334 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5335 | <p>Note that it is perfectly legal to index partially through a structure, |
| 5336 | returning a pointer to an inner element. Because of this, the LLVM code for |
| 5337 | the given testcase is equivalent to:</p> |
| Chris Lattner | f74d5c7 | 2004-04-05 01:30:49 +0000 | [diff] [blame] | 5338 | |
| Bill Wendling | a349539 | 2011-12-13 01:07:07 +0000 | [diff] [blame] | 5339 | <pre class="doc_code"> |
| 5340 | define i32* @foo(%struct.ST* %s) { |
| 5341 | %t1 = getelementptr %struct.ST* %s, i32 1 <i>; yields %struct.ST*:%t1</i> |
| 5342 | %t2 = getelementptr %struct.ST* %t1, i32 0, i32 2 <i>; yields %struct.RT*:%t2</i> |
| 5343 | %t3 = getelementptr %struct.RT* %t2, i32 0, i32 1 <i>; yields [10 x [20 x i32]]*:%t3</i> |
| 5344 | %t4 = getelementptr [10 x [20 x i32]]* %t3, i32 0, i32 5 <i>; yields [20 x i32]*:%t4</i> |
| 5345 | %t5 = getelementptr [20 x i32]* %t4, i32 0, i32 13 <i>; yields i32*:%t5</i> |
| 5346 | ret i32* %t5 |
| 5347 | } |
| Chris Lattner | 6536cfe | 2002-05-06 22:08:29 +0000 | [diff] [blame] | 5348 | </pre> |
| Chris Lattner | e67a951 | 2005-06-24 17:22:57 +0000 | [diff] [blame] | 5349 | |
| Dan Gohman | dd8004d | 2009-07-27 21:53:46 +0000 | [diff] [blame] | 5350 | <p>If the <tt>inbounds</tt> keyword is present, the result value of the |
| Dan Gohman | bfb056d | 2011-12-06 03:18:47 +0000 | [diff] [blame] | 5351 | <tt>getelementptr</tt> is a <a href="#poisonvalues">poison value</a> if the |
| Dan Gohman | 27ef997 | 2010-04-23 15:23:32 +0000 | [diff] [blame] | 5352 | base pointer is not an <i>in bounds</i> address of an allocated object, |
| 5353 | or if any of the addresses that would be formed by successive addition of |
| 5354 | the offsets implied by the indices to the base address with infinitely |
| Eli Friedman | 266246c | 2011-08-12 23:37:55 +0000 | [diff] [blame] | 5355 | precise signed arithmetic are not an <i>in bounds</i> address of that |
| 5356 | allocated object. The <i>in bounds</i> addresses for an allocated object |
| 5357 | are all the addresses that point into the object, plus the address one |
| Nadav Rotem | 1608769 | 2011-12-05 06:29:09 +0000 | [diff] [blame] | 5358 | byte past the end. |
| 5359 | In cases where the base is a vector of pointers the <tt>inbounds</tt> keyword |
| 5360 | applies to each of the computations element-wise. </p> |
| Dan Gohman | dd8004d | 2009-07-27 21:53:46 +0000 | [diff] [blame] | 5361 | |
| 5362 | <p>If the <tt>inbounds</tt> keyword is not present, the offsets are added to |
| Eli Friedman | 266246c | 2011-08-12 23:37:55 +0000 | [diff] [blame] | 5363 | the base address with silently-wrapping two's complement arithmetic. If the |
| 5364 | offsets have a different width from the pointer, they are sign-extended or |
| 5365 | truncated to the width of the pointer. The result value of the |
| 5366 | <tt>getelementptr</tt> may be outside the object pointed to by the base |
| 5367 | pointer. The result value may not necessarily be used to access memory |
| 5368 | though, even if it happens to point into allocated storage. See the |
| 5369 | <a href="#pointeraliasing">Pointer Aliasing Rules</a> section for more |
| 5370 | information.</p> |
| Dan Gohman | dd8004d | 2009-07-27 21:53:46 +0000 | [diff] [blame] | 5371 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5372 | <p>The getelementptr instruction is often confusing. For some more insight into |
| 5373 | how it works, see <a href="GetElementPtr.html">the getelementptr FAQ</a>.</p> |
| Chris Lattner | 884a970 | 2006-08-15 00:45:58 +0000 | [diff] [blame] | 5374 | |
| Chris Lattner | 7faa883 | 2002-04-14 06:13:44 +0000 | [diff] [blame] | 5375 | <h5>Example:</h5> |
| Chris Lattner | f74d5c7 | 2004-04-05 01:30:49 +0000 | [diff] [blame] | 5376 | <pre> |
| Reid Spencer | ca86e16 | 2006-12-31 07:07:53 +0000 | [diff] [blame] | 5377 | <i>; yields [12 x i8]*:aptr</i> |
| Matthijs Kooijman | e49d0bc | 2008-10-13 13:44:15 +0000 | [diff] [blame] | 5378 | %aptr = getelementptr {i32, [12 x i8]}* %saptr, i64 0, i32 1 |
| 5379 | <i>; yields i8*:vptr</i> |
| Dan Gohman | 0e451ce | 2008-10-14 16:51:45 +0000 | [diff] [blame] | 5380 | %vptr = getelementptr {i32, <2 x i8>}* %svptr, i64 0, i32 1, i32 1 |
| Matthijs Kooijman | e49d0bc | 2008-10-13 13:44:15 +0000 | [diff] [blame] | 5381 | <i>; yields i8*:eptr</i> |
| 5382 | %eptr = getelementptr [12 x i8]* %aptr, i64 0, i32 1 |
| Sanjiv Gupta | 9f805c2 | 2009-04-25 07:27:44 +0000 | [diff] [blame] | 5383 | <i>; yields i32*:iptr</i> |
| Sanjiv Gupta | 16ffa80 | 2009-04-24 16:38:13 +0000 | [diff] [blame] | 5384 | %iptr = getelementptr [10 x i32]* @arr, i16 0, i16 0 |
| Chris Lattner | f74d5c7 | 2004-04-05 01:30:49 +0000 | [diff] [blame] | 5385 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5386 | |
| Nadav Rotem | 1608769 | 2011-12-05 06:29:09 +0000 | [diff] [blame] | 5387 | <p>In cases where the pointer argument is a vector of pointers, only a |
| 5388 | single index may be used, and the number of vector elements has to be |
| 5389 | the same. For example: </p> |
| 5390 | <pre class="doc_code"> |
| 5391 | %A = getelementptr <4 x i8*> %ptrs, <4 x i64> %offsets, |
| 5392 | </pre> |
| 5393 | |
| Chris Lattner | f74d5c7 | 2004-04-05 01:30:49 +0000 | [diff] [blame] | 5394 | </div> |
| Reid Spencer | 47ce179 | 2006-11-09 21:15:49 +0000 | [diff] [blame] | 5395 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 5396 | </div> |
| 5397 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 5398 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 5399 | <h3> |
| 5400 | <a name="convertops">Conversion Operations</a> |
| 5401 | </h3> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5402 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 5403 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5404 | |
| Reid Spencer | 2fd21e6 | 2006-11-08 01:18:52 +0000 | [diff] [blame] | 5405 | <p>The instructions in this category are the conversion instructions (casting) |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5406 | which all take a single operand and a type. They perform various bit |
| 5407 | conversions on the operand.</p> |
| 5408 | |
| Chris Lattner | 6536cfe | 2002-05-06 22:08:29 +0000 | [diff] [blame] | 5409 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 5410 | <h4> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5411 | <a name="i_trunc">'<tt>trunc .. to</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 5412 | </h4> |
| 5413 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 5414 | <div> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5415 | |
| 5416 | <h5>Syntax:</h5> |
| 5417 | <pre> |
| 5418 | <result> = trunc <ty> <value> to <ty2> <i>; yields ty2</i> |
| 5419 | </pre> |
| 5420 | |
| 5421 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5422 | <p>The '<tt>trunc</tt>' instruction truncates its operand to the |
| 5423 | type <tt>ty2</tt>.</p> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5424 | |
| 5425 | <h5>Arguments:</h5> |
| Nadav Rotem | 8c20ec5 | 2011-02-24 21:01:34 +0000 | [diff] [blame] | 5426 | <p>The '<tt>trunc</tt>' instruction takes a value to trunc, and a type to trunc it to. |
| 5427 | Both types must be of <a href="#t_integer">integer</a> types, or vectors |
| 5428 | of the same number of integers. |
| 5429 | The bit size of the <tt>value</tt> must be larger than |
| 5430 | the bit size of the destination type, <tt>ty2</tt>. |
| 5431 | Equal sized types are not allowed.</p> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5432 | |
| 5433 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5434 | <p>The '<tt>trunc</tt>' instruction truncates the high order bits |
| 5435 | in <tt>value</tt> and converts the remaining bits to <tt>ty2</tt>. Since the |
| 5436 | source size must be larger than the destination size, <tt>trunc</tt> cannot |
| 5437 | be a <i>no-op cast</i>. It will always truncate bits.</p> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5438 | |
| 5439 | <h5>Example:</h5> |
| 5440 | <pre> |
| Nadav Rotem | 8c20ec5 | 2011-02-24 21:01:34 +0000 | [diff] [blame] | 5441 | %X = trunc i32 257 to i8 <i>; yields i8:1</i> |
| 5442 | %Y = trunc i32 123 to i1 <i>; yields i1:true</i> |
| 5443 | %Z = trunc i32 122 to i1 <i>; yields i1:false</i> |
| 5444 | %W = trunc <2 x i16> <i16 8, i16 7> to <2 x i8> <i>; yields <i8 8, i8 7></i> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5445 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5446 | |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5447 | </div> |
| 5448 | |
| 5449 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 5450 | <h4> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5451 | <a name="i_zext">'<tt>zext .. to</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 5452 | </h4> |
| 5453 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 5454 | <div> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5455 | |
| 5456 | <h5>Syntax:</h5> |
| 5457 | <pre> |
| 5458 | <result> = zext <ty> <value> to <ty2> <i>; yields ty2</i> |
| 5459 | </pre> |
| 5460 | |
| 5461 | <h5>Overview:</h5> |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 5462 | <p>The '<tt>zext</tt>' instruction zero extends its operand to type |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5463 | <tt>ty2</tt>.</p> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5464 | |
| 5465 | |
| 5466 | <h5>Arguments:</h5> |
| Nadav Rotem | ed9b934 | 2011-02-20 12:37:50 +0000 | [diff] [blame] | 5467 | <p>The '<tt>zext</tt>' instruction takes a value to cast, and a type to cast it to. |
| 5468 | Both types must be of <a href="#t_integer">integer</a> types, or vectors |
| 5469 | of the same number of integers. |
| 5470 | The bit size of the <tt>value</tt> must be smaller than |
| 5471 | the bit size of the destination type, |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5472 | <tt>ty2</tt>.</p> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5473 | |
| 5474 | <h5>Semantics:</h5> |
| 5475 | <p>The <tt>zext</tt> fills the high order bits of the <tt>value</tt> with zero |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5476 | bits until it reaches the size of the destination type, <tt>ty2</tt>.</p> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5477 | |
| Reid Spencer | b592952 | 2007-01-12 15:46:11 +0000 | [diff] [blame] | 5478 | <p>When zero extending from i1, the result will always be either 0 or 1.</p> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5479 | |
| 5480 | <h5>Example:</h5> |
| 5481 | <pre> |
| Reid Spencer | ca86e16 | 2006-12-31 07:07:53 +0000 | [diff] [blame] | 5482 | %X = zext i32 257 to i64 <i>; yields i64:257</i> |
| Reid Spencer | c78f337 | 2007-01-12 03:35:51 +0000 | [diff] [blame] | 5483 | %Y = zext i1 true to i32 <i>; yields i32:1</i> |
| Nadav Rotem | ed9b934 | 2011-02-20 12:37:50 +0000 | [diff] [blame] | 5484 | %Z = zext <2 x i16> <i16 8, i16 7> to <2 x i32> <i>; yields <i32 8, i32 7></i> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5485 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5486 | |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5487 | </div> |
| 5488 | |
| 5489 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 5490 | <h4> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5491 | <a name="i_sext">'<tt>sext .. to</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 5492 | </h4> |
| 5493 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 5494 | <div> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5495 | |
| 5496 | <h5>Syntax:</h5> |
| 5497 | <pre> |
| 5498 | <result> = sext <ty> <value> to <ty2> <i>; yields ty2</i> |
| 5499 | </pre> |
| 5500 | |
| 5501 | <h5>Overview:</h5> |
| 5502 | <p>The '<tt>sext</tt>' sign extends <tt>value</tt> to the type <tt>ty2</tt>.</p> |
| 5503 | |
| 5504 | <h5>Arguments:</h5> |
| Nadav Rotem | 8c20ec5 | 2011-02-24 21:01:34 +0000 | [diff] [blame] | 5505 | <p>The '<tt>sext</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 Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5510 | <tt>ty2</tt>.</p> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5511 | |
| 5512 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5513 | <p>The '<tt>sext</tt>' instruction performs a sign extension by copying the sign |
| 5514 | bit (highest order bit) of the <tt>value</tt> until it reaches the bit size |
| 5515 | of the type <tt>ty2</tt>.</p> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5516 | |
| Reid Spencer | c78f337 | 2007-01-12 03:35:51 +0000 | [diff] [blame] | 5517 | <p>When sign extending from i1, the extension always results in -1 or 0.</p> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5518 | |
| 5519 | <h5>Example:</h5> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5520 | <pre> |
| Reid Spencer | ca86e16 | 2006-12-31 07:07:53 +0000 | [diff] [blame] | 5521 | %X = sext i8 -1 to i16 <i>; yields i16 :65535</i> |
| Reid Spencer | c78f337 | 2007-01-12 03:35:51 +0000 | [diff] [blame] | 5522 | %Y = sext i1 true to i32 <i>; yields i32:-1</i> |
| Nadav Rotem | 8c20ec5 | 2011-02-24 21:01:34 +0000 | [diff] [blame] | 5523 | %Z = sext <2 x i16> <i16 8, i16 7> to <2 x i32> <i>; yields <i32 8, i32 7></i> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5524 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5525 | |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5526 | </div> |
| 5527 | |
| 5528 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 5529 | <h4> |
| Reid Spencer | 3fa91b0 | 2006-11-09 21:48:10 +0000 | [diff] [blame] | 5530 | <a name="i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 5531 | </h4> |
| Reid Spencer | 3fa91b0 | 2006-11-09 21:48:10 +0000 | [diff] [blame] | 5532 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 5533 | <div> |
| Reid Spencer | 3fa91b0 | 2006-11-09 21:48:10 +0000 | [diff] [blame] | 5534 | |
| 5535 | <h5>Syntax:</h5> |
| Reid Spencer | 3fa91b0 | 2006-11-09 21:48:10 +0000 | [diff] [blame] | 5536 | <pre> |
| 5537 | <result> = fptrunc <ty> <value> to <ty2> <i>; yields ty2</i> |
| 5538 | </pre> |
| 5539 | |
| 5540 | <h5>Overview:</h5> |
| 5541 | <p>The '<tt>fptrunc</tt>' instruction truncates <tt>value</tt> to type |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5542 | <tt>ty2</tt>.</p> |
| Reid Spencer | 3fa91b0 | 2006-11-09 21:48:10 +0000 | [diff] [blame] | 5543 | |
| 5544 | <h5>Arguments:</h5> |
| 5545 | <p>The '<tt>fptrunc</tt>' instruction takes a <a href="#t_floating">floating |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5546 | point</a> value to cast and a <a href="#t_floating">floating point</a> type |
| 5547 | to cast it to. The size of <tt>value</tt> must be larger than the size of |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 5548 | <tt>ty2</tt>. This implies that <tt>fptrunc</tt> cannot be used to make a |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5549 | <i>no-op cast</i>.</p> |
| Reid Spencer | 3fa91b0 | 2006-11-09 21:48:10 +0000 | [diff] [blame] | 5550 | |
| 5551 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5552 | <p>The '<tt>fptrunc</tt>' instruction truncates a <tt>value</tt> from a larger |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 5553 | <a href="#t_floating">floating point</a> type to a smaller |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5554 | <a href="#t_floating">floating point</a> type. If the value cannot fit |
| 5555 | within the destination type, <tt>ty2</tt>, then the results are |
| 5556 | undefined.</p> |
| Reid Spencer | 3fa91b0 | 2006-11-09 21:48:10 +0000 | [diff] [blame] | 5557 | |
| 5558 | <h5>Example:</h5> |
| 5559 | <pre> |
| 5560 | %X = fptrunc double 123.0 to float <i>; yields float:123.0</i> |
| 5561 | %Y = fptrunc double 1.0E+300 to float <i>; yields undefined</i> |
| 5562 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5563 | |
| Reid Spencer | 3fa91b0 | 2006-11-09 21:48:10 +0000 | [diff] [blame] | 5564 | </div> |
| 5565 | |
| 5566 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 5567 | <h4> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5568 | <a name="i_fpext">'<tt>fpext .. to</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 5569 | </h4> |
| 5570 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 5571 | <div> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5572 | |
| 5573 | <h5>Syntax:</h5> |
| 5574 | <pre> |
| 5575 | <result> = fpext <ty> <value> to <ty2> <i>; yields ty2</i> |
| 5576 | </pre> |
| 5577 | |
| 5578 | <h5>Overview:</h5> |
| 5579 | <p>The '<tt>fpext</tt>' extends a floating point <tt>value</tt> to a larger |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5580 | floating point value.</p> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5581 | |
| 5582 | <h5>Arguments:</h5> |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 5583 | <p>The '<tt>fpext</tt>' instruction takes a |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5584 | <a href="#t_floating">floating point</a> <tt>value</tt> to cast, and |
| 5585 | a <a href="#t_floating">floating point</a> type to cast it to. The source |
| 5586 | type must be smaller than the destination type.</p> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5587 | |
| 5588 | <h5>Semantics:</h5> |
| Reid Spencer | d444879 | 2006-11-09 23:03:26 +0000 | [diff] [blame] | 5589 | <p>The '<tt>fpext</tt>' instruction extends the <tt>value</tt> from a smaller |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5590 | <a href="#t_floating">floating point</a> type to a larger |
| 5591 | <a href="#t_floating">floating point</a> type. The <tt>fpext</tt> cannot be |
| 5592 | used to make a <i>no-op cast</i> because it always changes bits. Use |
| 5593 | <tt>bitcast</tt> to make a <i>no-op cast</i> for a floating point cast.</p> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5594 | |
| 5595 | <h5>Example:</h5> |
| 5596 | <pre> |
| Nick Lewycky | 5bb3ece | 2011-03-31 18:20:19 +0000 | [diff] [blame] | 5597 | %X = fpext float 3.125 to double <i>; yields double:3.125000e+00</i> |
| 5598 | %Y = fpext double %X to fp128 <i>; yields fp128:0xL00000000000000004000900000000000</i> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5599 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5600 | |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5601 | </div> |
| 5602 | |
| 5603 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 5604 | <h4> |
| Reid Spencer | 24d6da5 | 2007-01-21 00:29:26 +0000 | [diff] [blame] | 5605 | <a name="i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 5606 | </h4> |
| 5607 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 5608 | <div> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5609 | |
| 5610 | <h5>Syntax:</h5> |
| 5611 | <pre> |
| Reid Spencer | 1539a1c | 2007-07-31 14:40:14 +0000 | [diff] [blame] | 5612 | <result> = fptoui <ty> <value> to <ty2> <i>; yields ty2</i> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5613 | </pre> |
| 5614 | |
| 5615 | <h5>Overview:</h5> |
| Reid Spencer | 1539a1c | 2007-07-31 14:40:14 +0000 | [diff] [blame] | 5616 | <p>The '<tt>fptoui</tt>' converts a floating point <tt>value</tt> to its |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5617 | unsigned integer equivalent of type <tt>ty2</tt>.</p> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5618 | |
| 5619 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5620 | <p>The '<tt>fptoui</tt>' instruction takes a value to cast, which must be a |
| 5621 | scalar or vector <a href="#t_floating">floating point</a> value, and a type |
| 5622 | to cast it to <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a> |
| 5623 | type. If <tt>ty</tt> is a vector floating point type, <tt>ty2</tt> must be a |
| 5624 | vector integer type with the same number of elements as <tt>ty</tt></p> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5625 | |
| 5626 | <h5>Semantics:</h5> |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 5627 | <p>The '<tt>fptoui</tt>' instruction converts its |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5628 | <a href="#t_floating">floating point</a> operand into the nearest (rounding |
| 5629 | towards zero) unsigned integer value. If the value cannot fit |
| 5630 | in <tt>ty2</tt>, the results are undefined.</p> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5631 | |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5632 | <h5>Example:</h5> |
| 5633 | <pre> |
| Reid Spencer | 1539a1c | 2007-07-31 14:40:14 +0000 | [diff] [blame] | 5634 | %X = fptoui double 123.0 to i32 <i>; yields i32:123</i> |
| Chris Lattner | 8851904 | 2007-09-22 03:17:52 +0000 | [diff] [blame] | 5635 | %Y = fptoui float 1.0E+300 to i1 <i>; yields undefined:1</i> |
| Gabor Greif | 6a29201 | 2009-10-28 09:21:30 +0000 | [diff] [blame] | 5636 | %Z = fptoui float 1.04E+17 to i8 <i>; yields undefined:1</i> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5637 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5638 | |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5639 | </div> |
| 5640 | |
| 5641 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 5642 | <h4> |
| Reid Spencer | d444879 | 2006-11-09 23:03:26 +0000 | [diff] [blame] | 5643 | <a name="i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 5644 | </h4> |
| 5645 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 5646 | <div> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5647 | |
| 5648 | <h5>Syntax:</h5> |
| 5649 | <pre> |
| Reid Spencer | d444879 | 2006-11-09 23:03:26 +0000 | [diff] [blame] | 5650 | <result> = fptosi <ty> <value> to <ty2> <i>; yields ty2</i> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5651 | </pre> |
| 5652 | |
| 5653 | <h5>Overview:</h5> |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 5654 | <p>The '<tt>fptosi</tt>' instruction converts |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5655 | <a href="#t_floating">floating point</a> <tt>value</tt> to |
| 5656 | type <tt>ty2</tt>.</p> |
| Chris Lattner | cc37aae | 2004-03-12 05:50:16 +0000 | [diff] [blame] | 5657 | |
| Chris Lattner | 6536cfe | 2002-05-06 22:08:29 +0000 | [diff] [blame] | 5658 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5659 | <p>The '<tt>fptosi</tt>' instruction takes a value to cast, which must be a |
| 5660 | scalar or vector <a href="#t_floating">floating point</a> value, and a type |
| 5661 | to cast it to <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a> |
| 5662 | type. If <tt>ty</tt> is a vector floating point type, <tt>ty2</tt> must be a |
| 5663 | vector integer type with the same number of elements as <tt>ty</tt></p> |
| Chris Lattner | cc37aae | 2004-03-12 05:50:16 +0000 | [diff] [blame] | 5664 | |
| Chris Lattner | 6536cfe | 2002-05-06 22:08:29 +0000 | [diff] [blame] | 5665 | <h5>Semantics:</h5> |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 5666 | <p>The '<tt>fptosi</tt>' instruction converts its |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5667 | <a href="#t_floating">floating point</a> operand into the nearest (rounding |
| 5668 | towards zero) signed integer value. If the value cannot fit in <tt>ty2</tt>, |
| 5669 | the results are undefined.</p> |
| Chris Lattner | cc37aae | 2004-03-12 05:50:16 +0000 | [diff] [blame] | 5670 | |
| Chris Lattner | 33ba0d9 | 2001-07-09 00:26:23 +0000 | [diff] [blame] | 5671 | <h5>Example:</h5> |
| Chris Lattner | cc37aae | 2004-03-12 05:50:16 +0000 | [diff] [blame] | 5672 | <pre> |
| Reid Spencer | c78f337 | 2007-01-12 03:35:51 +0000 | [diff] [blame] | 5673 | %X = fptosi double -123.0 to i32 <i>; yields i32:-123</i> |
| Chris Lattner | 8851904 | 2007-09-22 03:17:52 +0000 | [diff] [blame] | 5674 | %Y = fptosi float 1.0E-247 to i1 <i>; yields undefined:1</i> |
| Gabor Greif | 6a29201 | 2009-10-28 09:21:30 +0000 | [diff] [blame] | 5675 | %Z = fptosi float 1.04E+17 to i8 <i>; yields undefined:1</i> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5676 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5677 | |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5678 | </div> |
| 5679 | |
| 5680 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 5681 | <h4> |
| Reid Spencer | d444879 | 2006-11-09 23:03:26 +0000 | [diff] [blame] | 5682 | <a name="i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 5683 | </h4> |
| 5684 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 5685 | <div> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5686 | |
| 5687 | <h5>Syntax:</h5> |
| 5688 | <pre> |
| Reid Spencer | d444879 | 2006-11-09 23:03:26 +0000 | [diff] [blame] | 5689 | <result> = uitofp <ty> <value> to <ty2> <i>; yields ty2</i> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5690 | </pre> |
| 5691 | |
| 5692 | <h5>Overview:</h5> |
| Reid Spencer | d444879 | 2006-11-09 23:03:26 +0000 | [diff] [blame] | 5693 | <p>The '<tt>uitofp</tt>' instruction regards <tt>value</tt> as an unsigned |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5694 | integer and converts that value to the <tt>ty2</tt> type.</p> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5695 | |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5696 | <h5>Arguments:</h5> |
| Nate Begeman | b348d18 | 2007-11-17 03:58:34 +0000 | [diff] [blame] | 5697 | <p>The '<tt>uitofp</tt>' instruction takes a value to cast, which must be a |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5698 | scalar or vector <a href="#t_integer">integer</a> value, and a type to cast |
| 5699 | it to <tt>ty2</tt>, which must be an <a href="#t_floating">floating point</a> |
| 5700 | type. If <tt>ty</tt> is a vector integer type, <tt>ty2</tt> must be a vector |
| 5701 | floating point type with the same number of elements as <tt>ty</tt></p> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5702 | |
| 5703 | <h5>Semantics:</h5> |
| Reid Spencer | d444879 | 2006-11-09 23:03:26 +0000 | [diff] [blame] | 5704 | <p>The '<tt>uitofp</tt>' instruction interprets its operand as an unsigned |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5705 | integer quantity and converts it to the corresponding floating point |
| 5706 | value. If the value cannot fit in the floating point value, the results are |
| 5707 | undefined.</p> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5708 | |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5709 | <h5>Example:</h5> |
| 5710 | <pre> |
| Reid Spencer | ca86e16 | 2006-12-31 07:07:53 +0000 | [diff] [blame] | 5711 | %X = uitofp i32 257 to float <i>; yields float:257.0</i> |
| Dan Gohman | 0e451ce | 2008-10-14 16:51:45 +0000 | [diff] [blame] | 5712 | %Y = uitofp i8 -1 to double <i>; yields double:255.0</i> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5713 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5714 | |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5715 | </div> |
| 5716 | |
| 5717 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 5718 | <h4> |
| Reid Spencer | d444879 | 2006-11-09 23:03:26 +0000 | [diff] [blame] | 5719 | <a name="i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 5720 | </h4> |
| 5721 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 5722 | <div> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5723 | |
| 5724 | <h5>Syntax:</h5> |
| 5725 | <pre> |
| Reid Spencer | d444879 | 2006-11-09 23:03:26 +0000 | [diff] [blame] | 5726 | <result> = sitofp <ty> <value> to <ty2> <i>; yields ty2</i> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5727 | </pre> |
| 5728 | |
| 5729 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5730 | <p>The '<tt>sitofp</tt>' instruction regards <tt>value</tt> as a signed integer |
| 5731 | and converts that value to the <tt>ty2</tt> type.</p> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5732 | |
| 5733 | <h5>Arguments:</h5> |
| Nate Begeman | b348d18 | 2007-11-17 03:58:34 +0000 | [diff] [blame] | 5734 | <p>The '<tt>sitofp</tt>' instruction takes a value to cast, which must be a |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5735 | scalar or vector <a href="#t_integer">integer</a> value, and a type to cast |
| 5736 | it to <tt>ty2</tt>, which must be an <a href="#t_floating">floating point</a> |
| 5737 | type. If <tt>ty</tt> is a vector integer type, <tt>ty2</tt> must be a vector |
| 5738 | floating point type with the same number of elements as <tt>ty</tt></p> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5739 | |
| 5740 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5741 | <p>The '<tt>sitofp</tt>' instruction interprets its operand as a signed integer |
| 5742 | quantity and converts it to the corresponding floating point value. If the |
| 5743 | value cannot fit in the floating point value, the results are undefined.</p> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5744 | |
| 5745 | <h5>Example:</h5> |
| 5746 | <pre> |
| Reid Spencer | ca86e16 | 2006-12-31 07:07:53 +0000 | [diff] [blame] | 5747 | %X = sitofp i32 257 to float <i>; yields float:257.0</i> |
| Dan Gohman | 0e451ce | 2008-10-14 16:51:45 +0000 | [diff] [blame] | 5748 | %Y = sitofp i8 -1 to double <i>; yields double:-1.0</i> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5749 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5750 | |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5751 | </div> |
| 5752 | |
| 5753 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 5754 | <h4> |
| Reid Spencer | 7267925 | 2006-11-11 21:00:47 +0000 | [diff] [blame] | 5755 | <a name="i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 5756 | </h4> |
| 5757 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 5758 | <div> |
| Reid Spencer | 7267925 | 2006-11-11 21:00:47 +0000 | [diff] [blame] | 5759 | |
| 5760 | <h5>Syntax:</h5> |
| 5761 | <pre> |
| 5762 | <result> = ptrtoint <ty> <value> to <ty2> <i>; yields ty2</i> |
| 5763 | </pre> |
| 5764 | |
| 5765 | <h5>Overview:</h5> |
| Nadav Rotem | 1608769 | 2011-12-05 06:29:09 +0000 | [diff] [blame] | 5766 | <p>The '<tt>ptrtoint</tt>' instruction converts the pointer or a vector of |
| 5767 | pointers <tt>value</tt> to |
| 5768 | the integer (or vector of integers) type <tt>ty2</tt>.</p> |
| Reid Spencer | 7267925 | 2006-11-11 21:00:47 +0000 | [diff] [blame] | 5769 | |
| 5770 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5771 | <p>The '<tt>ptrtoint</tt>' instruction takes a <tt>value</tt> to cast, which |
| Nadav Rotem | 1608769 | 2011-12-05 06:29:09 +0000 | [diff] [blame] | 5772 | must be a a value of type <a href="#t_pointer">pointer</a> or a vector of |
| 5773 | pointers, and a type to cast it to |
| 5774 | <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a> or a vector |
| 5775 | of integers type.</p> |
| Reid Spencer | 7267925 | 2006-11-11 21:00:47 +0000 | [diff] [blame] | 5776 | |
| 5777 | <h5>Semantics:</h5> |
| 5778 | <p>The '<tt>ptrtoint</tt>' instruction converts <tt>value</tt> to integer type |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5779 | <tt>ty2</tt> by interpreting the pointer value as an integer and either |
| 5780 | truncating or zero extending that value to the size of the integer type. If |
| 5781 | <tt>value</tt> is smaller than <tt>ty2</tt> then a zero extension is done. If |
| 5782 | <tt>value</tt> is larger than <tt>ty2</tt> then a truncation is done. If they |
| 5783 | are the same size, then nothing is done (<i>no-op cast</i>) other than a type |
| 5784 | change.</p> |
| Reid Spencer | 7267925 | 2006-11-11 21:00:47 +0000 | [diff] [blame] | 5785 | |
| 5786 | <h5>Example:</h5> |
| 5787 | <pre> |
| Nadav Rotem | 1608769 | 2011-12-05 06:29:09 +0000 | [diff] [blame] | 5788 | %X = ptrtoint i32* %P to i8 <i>; yields truncation on 32-bit architecture</i> |
| 5789 | %Y = ptrtoint i32* %P to i64 <i>; yields zero extension on 32-bit architecture</i> |
| 5790 | %Z = ptrtoint <4 x i32*> %P to <4 x i64><i>; yields vector zero extension for a vector of addresses on 32-bit architecture</i> |
| Reid Spencer | 7267925 | 2006-11-11 21:00:47 +0000 | [diff] [blame] | 5791 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5792 | |
| Reid Spencer | 7267925 | 2006-11-11 21:00:47 +0000 | [diff] [blame] | 5793 | </div> |
| 5794 | |
| 5795 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 5796 | <h4> |
| Reid Spencer | 7267925 | 2006-11-11 21:00:47 +0000 | [diff] [blame] | 5797 | <a name="i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 5798 | </h4> |
| 5799 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 5800 | <div> |
| Reid Spencer | 7267925 | 2006-11-11 21:00:47 +0000 | [diff] [blame] | 5801 | |
| 5802 | <h5>Syntax:</h5> |
| 5803 | <pre> |
| 5804 | <result> = inttoptr <ty> <value> to <ty2> <i>; yields ty2</i> |
| 5805 | </pre> |
| 5806 | |
| 5807 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5808 | <p>The '<tt>inttoptr</tt>' instruction converts an integer <tt>value</tt> to a |
| 5809 | pointer type, <tt>ty2</tt>.</p> |
| Reid Spencer | 7267925 | 2006-11-11 21:00:47 +0000 | [diff] [blame] | 5810 | |
| 5811 | <h5>Arguments:</h5> |
| Duncan Sands | 8036ca4 | 2007-03-30 12:22:09 +0000 | [diff] [blame] | 5812 | <p>The '<tt>inttoptr</tt>' instruction takes an <a href="#t_integer">integer</a> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5813 | value to cast, and a type to cast it to, which must be a |
| 5814 | <a href="#t_pointer">pointer</a> type.</p> |
| Reid Spencer | 7267925 | 2006-11-11 21:00:47 +0000 | [diff] [blame] | 5815 | |
| 5816 | <h5>Semantics:</h5> |
| 5817 | <p>The '<tt>inttoptr</tt>' instruction converts <tt>value</tt> to type |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5818 | <tt>ty2</tt> by applying either a zero extension or a truncation depending on |
| 5819 | the size of the integer <tt>value</tt>. If <tt>value</tt> is larger than the |
| 5820 | size of a pointer then a truncation is done. If <tt>value</tt> is smaller |
| 5821 | than the size of a pointer then a zero extension is done. If they are the |
| 5822 | same size, nothing is done (<i>no-op cast</i>).</p> |
| Reid Spencer | 7267925 | 2006-11-11 21:00:47 +0000 | [diff] [blame] | 5823 | |
| 5824 | <h5>Example:</h5> |
| 5825 | <pre> |
| Jeff Cohen | b627eab | 2007-04-29 01:07:00 +0000 | [diff] [blame] | 5826 | %X = inttoptr i32 255 to i32* <i>; yields zero extension on 64-bit architecture</i> |
| Gabor Greif | 6a29201 | 2009-10-28 09:21:30 +0000 | [diff] [blame] | 5827 | %Y = inttoptr i32 255 to i32* <i>; yields no-op on 32-bit architecture</i> |
| 5828 | %Z = inttoptr i64 0 to i32* <i>; yields truncation on 32-bit architecture</i> |
| Nadav Rotem | 1608769 | 2011-12-05 06:29:09 +0000 | [diff] [blame] | 5829 | %Z = inttoptr <4 x i32> %G to <4 x i8*><i>; yields truncation of vector G to four pointers</i> |
| Reid Spencer | 7267925 | 2006-11-11 21:00:47 +0000 | [diff] [blame] | 5830 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5831 | |
| Reid Spencer | 7267925 | 2006-11-11 21:00:47 +0000 | [diff] [blame] | 5832 | </div> |
| 5833 | |
| 5834 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 5835 | <h4> |
| Reid Spencer | 5c0ef47 | 2006-11-11 23:08:07 +0000 | [diff] [blame] | 5836 | <a name="i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 5837 | </h4> |
| 5838 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 5839 | <div> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5840 | |
| 5841 | <h5>Syntax:</h5> |
| 5842 | <pre> |
| Reid Spencer | 5c0ef47 | 2006-11-11 23:08:07 +0000 | [diff] [blame] | 5843 | <result> = bitcast <ty> <value> to <ty2> <i>; yields ty2</i> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5844 | </pre> |
| 5845 | |
| 5846 | <h5>Overview:</h5> |
| Reid Spencer | 5c0ef47 | 2006-11-11 23:08:07 +0000 | [diff] [blame] | 5847 | <p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5848 | <tt>ty2</tt> without changing any bits.</p> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5849 | |
| 5850 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5851 | <p>The '<tt>bitcast</tt>' instruction takes a value to cast, which must be a |
| 5852 | non-aggregate first class value, and a type to cast it to, which must also be |
| 5853 | a non-aggregate <a href="#t_firstclass">first class</a> type. The bit sizes |
| 5854 | of <tt>value</tt> and the destination type, <tt>ty2</tt>, must be |
| 5855 | identical. If the source type is a pointer, the destination type must also be |
| 5856 | a pointer. This instruction supports bitwise conversion of vectors to |
| 5857 | integers and to vectors of other types (as long as they have the same |
| 5858 | size).</p> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5859 | |
| 5860 | <h5>Semantics:</h5> |
| Reid Spencer | 5c0ef47 | 2006-11-11 23:08:07 +0000 | [diff] [blame] | 5861 | <p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5862 | <tt>ty2</tt>. It is always a <i>no-op cast</i> because no bits change with |
| 5863 | this conversion. The conversion is done as if the <tt>value</tt> had been |
| Nadav Rotem | 1608769 | 2011-12-05 06:29:09 +0000 | [diff] [blame] | 5864 | stored to memory and read back as type <tt>ty2</tt>. |
| 5865 | Pointer (or vector of pointers) types may only be converted to other pointer |
| 5866 | (or vector of pointers) types with this instruction. To convert |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5867 | pointers to other types, use the <a href="#i_inttoptr">inttoptr</a> or |
| 5868 | <a href="#i_ptrtoint">ptrtoint</a> instructions first.</p> |
| Reid Spencer | 9dee3ac | 2006-11-08 01:11:31 +0000 | [diff] [blame] | 5869 | |
| 5870 | <h5>Example:</h5> |
| 5871 | <pre> |
| Jeff Cohen | b627eab | 2007-04-29 01:07:00 +0000 | [diff] [blame] | 5872 | %X = bitcast i8 255 to i8 <i>; yields i8 :-1</i> |
| Reid Spencer | ca86e16 | 2006-12-31 07:07:53 +0000 | [diff] [blame] | 5873 | %Y = bitcast i32* %x to sint* <i>; yields sint*:%x</i> |
| Nadav Rotem | 1608769 | 2011-12-05 06:29:09 +0000 | [diff] [blame] | 5874 | %Z = bitcast <2 x int> %V to i64; <i>; yields i64: %V</i> |
| 5875 | %Z = bitcast <2 x i32*> %V to <2 x i64*> <i>; yields <2 x i64*></i> |
| Chris Lattner | 33ba0d9 | 2001-07-09 00:26:23 +0000 | [diff] [blame] | 5876 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5877 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 5878 | </div> |
| Chris Lattner | cc37aae | 2004-03-12 05:50:16 +0000 | [diff] [blame] | 5879 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 5880 | </div> |
| 5881 | |
| Reid Spencer | 2fd21e6 | 2006-11-08 01:18:52 +0000 | [diff] [blame] | 5882 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 5883 | <h3> |
| 5884 | <a name="otherops">Other Operations</a> |
| 5885 | </h3> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5886 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 5887 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5888 | |
| 5889 | <p>The instructions in this category are the "miscellaneous" instructions, which |
| 5890 | defy better classification.</p> |
| 5891 | |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 5892 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 5893 | <h4> |
| 5894 | <a name="i_icmp">'<tt>icmp</tt>' Instruction</a> |
| 5895 | </h4> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5896 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 5897 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5898 | |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 5899 | <h5>Syntax:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5900 | <pre> |
| 5901 | <result> = icmp <cond> <ty> <op1>, <op2> <i>; yields {i1} or {<N x i1>}:result</i> |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 5902 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5903 | |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 5904 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5905 | <p>The '<tt>icmp</tt>' instruction returns a boolean value or a vector of |
| Nadav Rotem | 1608769 | 2011-12-05 06:29:09 +0000 | [diff] [blame] | 5906 | boolean values based on comparison of its two integer, integer vector, |
| 5907 | pointer, or pointer vector operands.</p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5908 | |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 5909 | <h5>Arguments:</h5> |
| 5910 | <p>The '<tt>icmp</tt>' instruction takes three operands. The first operand is |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5911 | the condition code indicating the kind of comparison to perform. It is not a |
| 5912 | value, just a keyword. The possible condition code are:</p> |
| 5913 | |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 5914 | <ol> |
| 5915 | <li><tt>eq</tt>: equal</li> |
| 5916 | <li><tt>ne</tt>: not equal </li> |
| 5917 | <li><tt>ugt</tt>: unsigned greater than</li> |
| 5918 | <li><tt>uge</tt>: unsigned greater or equal</li> |
| 5919 | <li><tt>ult</tt>: unsigned less than</li> |
| 5920 | <li><tt>ule</tt>: unsigned less or equal</li> |
| 5921 | <li><tt>sgt</tt>: signed greater than</li> |
| 5922 | <li><tt>sge</tt>: signed greater or equal</li> |
| 5923 | <li><tt>slt</tt>: signed less than</li> |
| 5924 | <li><tt>sle</tt>: signed less or equal</li> |
| 5925 | </ol> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5926 | |
| Chris Lattner | 3b19d65 | 2007-01-15 01:54:13 +0000 | [diff] [blame] | 5927 | <p>The remaining two arguments must be <a href="#t_integer">integer</a> or |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5928 | <a href="#t_pointer">pointer</a> or integer <a href="#t_vector">vector</a> |
| 5929 | typed. They must also be identical types.</p> |
| 5930 | |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 5931 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5932 | <p>The '<tt>icmp</tt>' compares <tt>op1</tt> and <tt>op2</tt> according to the |
| 5933 | condition code given as <tt>cond</tt>. The comparison performed always yields |
| Nick Lewycky | ec38da4 | 2009-09-27 00:45:11 +0000 | [diff] [blame] | 5934 | either an <a href="#t_integer"><tt>i1</tt></a> or vector of <tt>i1</tt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5935 | result, as follows:</p> |
| 5936 | |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 5937 | <ol> |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 5938 | <li><tt>eq</tt>: yields <tt>true</tt> if the operands are equal, |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5939 | <tt>false</tt> otherwise. No sign interpretation is necessary or |
| 5940 | performed.</li> |
| 5941 | |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 5942 | <li><tt>ne</tt>: yields <tt>true</tt> if the operands are unequal, |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5943 | <tt>false</tt> otherwise. No sign interpretation is necessary or |
| 5944 | performed.</li> |
| 5945 | |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 5946 | <li><tt>ugt</tt>: interprets the operands as unsigned values and yields |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5947 | <tt>true</tt> if <tt>op1</tt> is greater than <tt>op2</tt>.</li> |
| 5948 | |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 5949 | <li><tt>uge</tt>: interprets the operands as unsigned values and yields |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5950 | <tt>true</tt> if <tt>op1</tt> is greater than or equal |
| 5951 | to <tt>op2</tt>.</li> |
| 5952 | |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 5953 | <li><tt>ult</tt>: interprets the operands as unsigned values and yields |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5954 | <tt>true</tt> if <tt>op1</tt> is less than <tt>op2</tt>.</li> |
| 5955 | |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 5956 | <li><tt>ule</tt>: interprets the operands as unsigned values and yields |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5957 | <tt>true</tt> if <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li> |
| 5958 | |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 5959 | <li><tt>sgt</tt>: interprets the operands as signed values and yields |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5960 | <tt>true</tt> if <tt>op1</tt> is greater than <tt>op2</tt>.</li> |
| 5961 | |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 5962 | <li><tt>sge</tt>: interprets the operands as signed values and yields |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5963 | <tt>true</tt> if <tt>op1</tt> is greater than or equal |
| 5964 | to <tt>op2</tt>.</li> |
| 5965 | |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 5966 | <li><tt>slt</tt>: interprets the operands as signed values and yields |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5967 | <tt>true</tt> if <tt>op1</tt> is less than <tt>op2</tt>.</li> |
| 5968 | |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 5969 | <li><tt>sle</tt>: interprets the operands as signed values and yields |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5970 | <tt>true</tt> if <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li> |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 5971 | </ol> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5972 | |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 5973 | <p>If the operands are <a href="#t_pointer">pointer</a> typed, the pointer |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5974 | values are compared as if they were integers.</p> |
| 5975 | |
| 5976 | <p>If the operands are integer vectors, then they are compared element by |
| 5977 | element. The result is an <tt>i1</tt> vector with the same number of elements |
| 5978 | as the values being compared. Otherwise, the result is an <tt>i1</tt>.</p> |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 5979 | |
| 5980 | <h5>Example:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5981 | <pre> |
| 5982 | <result> = icmp eq i32 4, 5 <i>; yields: result=false</i> |
| Reid Spencer | ca86e16 | 2006-12-31 07:07:53 +0000 | [diff] [blame] | 5983 | <result> = icmp ne float* %X, %X <i>; yields: result=false</i> |
| 5984 | <result> = icmp ult i16 4, 5 <i>; yields: result=true</i> |
| 5985 | <result> = icmp sgt i16 4, 5 <i>; yields: result=false</i> |
| 5986 | <result> = icmp ule i16 -4, 5 <i>; yields: result=false</i> |
| 5987 | <result> = icmp sge i16 4, 5 <i>; yields: result=false</i> |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 5988 | </pre> |
| Dan Gohman | c766f72 | 2009-01-22 01:39:38 +0000 | [diff] [blame] | 5989 | |
| 5990 | <p>Note that the code generator does not yet support vector types with |
| 5991 | the <tt>icmp</tt> instruction.</p> |
| 5992 | |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 5993 | </div> |
| 5994 | |
| 5995 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 5996 | <h4> |
| 5997 | <a name="i_fcmp">'<tt>fcmp</tt>' Instruction</a> |
| 5998 | </h4> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 5999 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 6000 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6001 | |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 6002 | <h5>Syntax:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6003 | <pre> |
| 6004 | <result> = fcmp <cond> <ty> <op1>, <op2> <i>; yields {i1} or {<N x i1>}:result</i> |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 6005 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6006 | |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 6007 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6008 | <p>The '<tt>fcmp</tt>' instruction returns a boolean value or vector of boolean |
| 6009 | values based on comparison of its operands.</p> |
| 6010 | |
| 6011 | <p>If the operands are floating point scalars, then the result type is a boolean |
| Nick Lewycky | ec38da4 | 2009-09-27 00:45:11 +0000 | [diff] [blame] | 6012 | (<a href="#t_integer"><tt>i1</tt></a>).</p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6013 | |
| 6014 | <p>If the operands are floating point vectors, then the result type is a vector |
| 6015 | of boolean with the same number of elements as the operands being |
| 6016 | compared.</p> |
| 6017 | |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 6018 | <h5>Arguments:</h5> |
| 6019 | <p>The '<tt>fcmp</tt>' instruction takes three operands. The first operand is |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6020 | the condition code indicating the kind of comparison to perform. It is not a |
| 6021 | value, just a keyword. The possible condition code are:</p> |
| 6022 | |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 6023 | <ol> |
| Reid Spencer | b7f2628 | 2006-11-19 03:00:14 +0000 | [diff] [blame] | 6024 | <li><tt>false</tt>: no comparison, always returns false</li> |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 6025 | <li><tt>oeq</tt>: ordered and equal</li> |
| 6026 | <li><tt>ogt</tt>: ordered and greater than </li> |
| 6027 | <li><tt>oge</tt>: ordered and greater than or equal</li> |
| 6028 | <li><tt>olt</tt>: ordered and less than </li> |
| 6029 | <li><tt>ole</tt>: ordered and less than or equal</li> |
| 6030 | <li><tt>one</tt>: ordered and not equal</li> |
| 6031 | <li><tt>ord</tt>: ordered (no nans)</li> |
| 6032 | <li><tt>ueq</tt>: unordered or equal</li> |
| 6033 | <li><tt>ugt</tt>: unordered or greater than </li> |
| 6034 | <li><tt>uge</tt>: unordered or greater than or equal</li> |
| 6035 | <li><tt>ult</tt>: unordered or less than </li> |
| 6036 | <li><tt>ule</tt>: unordered or less than or equal</li> |
| 6037 | <li><tt>une</tt>: unordered or not equal</li> |
| 6038 | <li><tt>uno</tt>: unordered (either nans)</li> |
| Reid Spencer | b7f2628 | 2006-11-19 03:00:14 +0000 | [diff] [blame] | 6039 | <li><tt>true</tt>: no comparison, always returns true</li> |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 6040 | </ol> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6041 | |
| Jeff Cohen | b627eab | 2007-04-29 01:07:00 +0000 | [diff] [blame] | 6042 | <p><i>Ordered</i> means that neither operand is a QNAN while |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6043 | <i>unordered</i> means that either operand may be a QNAN.</p> |
| 6044 | |
| 6045 | <p>Each of <tt>val1</tt> and <tt>val2</tt> arguments must be either |
| 6046 | a <a href="#t_floating">floating point</a> type or |
| 6047 | a <a href="#t_vector">vector</a> of floating point type. They must have |
| 6048 | identical types.</p> |
| 6049 | |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 6050 | <h5>Semantics:</h5> |
| Gabor Greif | fb224a2 | 2008-08-07 21:46:00 +0000 | [diff] [blame] | 6051 | <p>The '<tt>fcmp</tt>' instruction compares <tt>op1</tt> and <tt>op2</tt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6052 | according to the condition code given as <tt>cond</tt>. If the operands are |
| 6053 | vectors, then the vectors are compared element by element. Each comparison |
| Nick Lewycky | ec38da4 | 2009-09-27 00:45:11 +0000 | [diff] [blame] | 6054 | performed always yields an <a href="#t_integer">i1</a> result, as |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6055 | follows:</p> |
| 6056 | |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 6057 | <ol> |
| 6058 | <li><tt>false</tt>: always yields <tt>false</tt>, regardless of operands.</li> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6059 | |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 6060 | <li><tt>oeq</tt>: yields <tt>true</tt> if both operands are not a QNAN and |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6061 | <tt>op1</tt> is equal to <tt>op2</tt>.</li> |
| 6062 | |
| Reid Spencer | b7f2628 | 2006-11-19 03:00:14 +0000 | [diff] [blame] | 6063 | <li><tt>ogt</tt>: yields <tt>true</tt> if both operands are not a QNAN and |
| Dan Gohman | b55a1ee | 2010-03-01 17:41:39 +0000 | [diff] [blame] | 6064 | <tt>op1</tt> is greater than <tt>op2</tt>.</li> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6065 | |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 6066 | <li><tt>oge</tt>: yields <tt>true</tt> if both operands are not a QNAN and |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6067 | <tt>op1</tt> is greater than or equal to <tt>op2</tt>.</li> |
| 6068 | |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 6069 | <li><tt>olt</tt>: yields <tt>true</tt> if both operands are not a QNAN and |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6070 | <tt>op1</tt> is less than <tt>op2</tt>.</li> |
| 6071 | |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 6072 | <li><tt>ole</tt>: yields <tt>true</tt> if both operands are not a QNAN and |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6073 | <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li> |
| 6074 | |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 6075 | <li><tt>one</tt>: yields <tt>true</tt> if both operands are not a QNAN and |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6076 | <tt>op1</tt> is not equal to <tt>op2</tt>.</li> |
| 6077 | |
| Reid Spencer | b7f2628 | 2006-11-19 03:00:14 +0000 | [diff] [blame] | 6078 | <li><tt>ord</tt>: yields <tt>true</tt> if both operands are not a QNAN.</li> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6079 | |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 6080 | <li><tt>ueq</tt>: yields <tt>true</tt> if either operand is a QNAN or |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6081 | <tt>op1</tt> is equal to <tt>op2</tt>.</li> |
| 6082 | |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 6083 | <li><tt>ugt</tt>: yields <tt>true</tt> if either operand is a QNAN or |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6084 | <tt>op1</tt> is greater than <tt>op2</tt>.</li> |
| 6085 | |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 6086 | <li><tt>uge</tt>: yields <tt>true</tt> if either operand is a QNAN or |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6087 | <tt>op1</tt> is greater than or equal to <tt>op2</tt>.</li> |
| 6088 | |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 6089 | <li><tt>ult</tt>: yields <tt>true</tt> if either operand is a QNAN or |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6090 | <tt>op1</tt> is less than <tt>op2</tt>.</li> |
| 6091 | |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 6092 | <li><tt>ule</tt>: yields <tt>true</tt> if either operand is a QNAN or |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6093 | <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li> |
| 6094 | |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 6095 | <li><tt>une</tt>: yields <tt>true</tt> if either operand is a QNAN or |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6096 | <tt>op1</tt> is not equal to <tt>op2</tt>.</li> |
| 6097 | |
| Reid Spencer | b7f2628 | 2006-11-19 03:00:14 +0000 | [diff] [blame] | 6098 | <li><tt>uno</tt>: yields <tt>true</tt> if either operand is a QNAN.</li> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6099 | |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 6100 | <li><tt>true</tt>: always yields <tt>true</tt>, regardless of operands.</li> |
| 6101 | </ol> |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 6102 | |
| 6103 | <h5>Example:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6104 | <pre> |
| 6105 | <result> = fcmp oeq float 4.0, 5.0 <i>; yields: result=false</i> |
| Dan Gohman | f72fb67 | 2008-09-09 01:02:47 +0000 | [diff] [blame] | 6106 | <result> = fcmp one float 4.0, 5.0 <i>; yields: result=true</i> |
| 6107 | <result> = fcmp olt float 4.0, 5.0 <i>; yields: result=true</i> |
| 6108 | <result> = fcmp ueq double 1.0, 2.0 <i>; yields: result=false</i> |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 6109 | </pre> |
| Dan Gohman | c766f72 | 2009-01-22 01:39:38 +0000 | [diff] [blame] | 6110 | |
| 6111 | <p>Note that the code generator does not yet support vector types with |
| 6112 | the <tt>fcmp</tt> instruction.</p> |
| 6113 | |
| Reid Spencer | f3a70a6 | 2006-11-18 21:50:54 +0000 | [diff] [blame] | 6114 | </div> |
| 6115 | |
| Reid Spencer | 2fd21e6 | 2006-11-08 01:18:52 +0000 | [diff] [blame] | 6116 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6117 | <h4> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 6118 | <a name="i_phi">'<tt>phi</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6119 | </h4> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 6120 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 6121 | <div> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 6122 | |
| Reid Spencer | 2fd21e6 | 2006-11-08 01:18:52 +0000 | [diff] [blame] | 6123 | <h5>Syntax:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6124 | <pre> |
| 6125 | <result> = phi <ty> [ <val0>, <label0>], ... |
| 6126 | </pre> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 6127 | |
| Reid Spencer | 2fd21e6 | 2006-11-08 01:18:52 +0000 | [diff] [blame] | 6128 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6129 | <p>The '<tt>phi</tt>' instruction is used to implement the φ node in the |
| 6130 | SSA graph representing the function.</p> |
| 6131 | |
| Reid Spencer | 2fd21e6 | 2006-11-08 01:18:52 +0000 | [diff] [blame] | 6132 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6133 | <p>The type of the incoming values is specified with the first type field. After |
| 6134 | this, the '<tt>phi</tt>' instruction takes a list of pairs as arguments, with |
| 6135 | one pair for each predecessor basic block of the current block. Only values |
| 6136 | of <a href="#t_firstclass">first class</a> type may be used as the value |
| 6137 | arguments to the PHI node. Only labels may be used as the label |
| 6138 | arguments.</p> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 6139 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6140 | <p>There must be no non-phi instructions between the start of a basic block and |
| 6141 | the PHI instructions: i.e. PHI instructions must be first in a basic |
| 6142 | block.</p> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 6143 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6144 | <p>For the purposes of the SSA form, the use of each incoming value is deemed to |
| 6145 | occur on the edge from the corresponding predecessor block to the current |
| 6146 | block (but after any definition of an '<tt>invoke</tt>' instruction's return |
| 6147 | value on the same edge).</p> |
| Jay Foad | d244909 | 2009-06-03 10:20:10 +0000 | [diff] [blame] | 6148 | |
| Reid Spencer | 2fd21e6 | 2006-11-08 01:18:52 +0000 | [diff] [blame] | 6149 | <h5>Semantics:</h5> |
| Jeff Cohen | b627eab | 2007-04-29 01:07:00 +0000 | [diff] [blame] | 6150 | <p>At runtime, the '<tt>phi</tt>' instruction logically takes on the value |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6151 | specified by the pair corresponding to the predecessor basic block that |
| 6152 | executed just prior to the current block.</p> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 6153 | |
| Reid Spencer | 2fd21e6 | 2006-11-08 01:18:52 +0000 | [diff] [blame] | 6154 | <h5>Example:</h5> |
| Chris Lattner | 5568e94 | 2008-05-20 20:48:21 +0000 | [diff] [blame] | 6155 | <pre> |
| 6156 | Loop: ; Infinite loop that counts from 0 on up... |
| 6157 | %indvar = phi i32 [ 0, %LoopHeader ], [ %nextindvar, %Loop ] |
| 6158 | %nextindvar = add i32 %indvar, 1 |
| 6159 | br label %Loop |
| 6160 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6161 | |
| Reid Spencer | 2fd21e6 | 2006-11-08 01:18:52 +0000 | [diff] [blame] | 6162 | </div> |
| 6163 | |
| Chris Lattner | cc37aae | 2004-03-12 05:50:16 +0000 | [diff] [blame] | 6164 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6165 | <h4> |
| Chris Lattner | cc37aae | 2004-03-12 05:50:16 +0000 | [diff] [blame] | 6166 | <a name="i_select">'<tt>select</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6167 | </h4> |
| Chris Lattner | cc37aae | 2004-03-12 05:50:16 +0000 | [diff] [blame] | 6168 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 6169 | <div> |
| Chris Lattner | cc37aae | 2004-03-12 05:50:16 +0000 | [diff] [blame] | 6170 | |
| 6171 | <h5>Syntax:</h5> |
| Chris Lattner | cc37aae | 2004-03-12 05:50:16 +0000 | [diff] [blame] | 6172 | <pre> |
| Dan Gohman | f72fb67 | 2008-09-09 01:02:47 +0000 | [diff] [blame] | 6173 | <result> = select <i>selty</i> <cond>, <ty> <val1>, <ty> <val2> <i>; yields ty</i> |
| 6174 | |
| Dan Gohman | 0e451ce | 2008-10-14 16:51:45 +0000 | [diff] [blame] | 6175 | <i>selty</i> is either i1 or {<N x i1>} |
| Chris Lattner | cc37aae | 2004-03-12 05:50:16 +0000 | [diff] [blame] | 6176 | </pre> |
| 6177 | |
| 6178 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6179 | <p>The '<tt>select</tt>' instruction is used to choose one value based on a |
| 6180 | condition, without branching.</p> |
| Chris Lattner | cc37aae | 2004-03-12 05:50:16 +0000 | [diff] [blame] | 6181 | |
| 6182 | |
| 6183 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6184 | <p>The '<tt>select</tt>' instruction requires an 'i1' value or a vector of 'i1' |
| 6185 | values indicating the condition, and two values of the |
| 6186 | same <a href="#t_firstclass">first class</a> type. If the val1/val2 are |
| 6187 | vectors and the condition is a scalar, then entire vectors are selected, not |
| 6188 | individual elements.</p> |
| Chris Lattner | cc37aae | 2004-03-12 05:50:16 +0000 | [diff] [blame] | 6189 | |
| 6190 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6191 | <p>If the condition is an i1 and it evaluates to 1, the instruction returns the |
| 6192 | first value argument; otherwise, it returns the second value argument.</p> |
| Chris Lattner | cc37aae | 2004-03-12 05:50:16 +0000 | [diff] [blame] | 6193 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6194 | <p>If the condition is a vector of i1, then the value arguments must be vectors |
| 6195 | of the same size, and the selection is done element by element.</p> |
| Chris Lattner | cc37aae | 2004-03-12 05:50:16 +0000 | [diff] [blame] | 6196 | |
| 6197 | <h5>Example:</h5> |
| Chris Lattner | cc37aae | 2004-03-12 05:50:16 +0000 | [diff] [blame] | 6198 | <pre> |
| Reid Spencer | c78f337 | 2007-01-12 03:35:51 +0000 | [diff] [blame] | 6199 | %X = select i1 true, i8 17, i8 42 <i>; yields i8:17</i> |
| Chris Lattner | cc37aae | 2004-03-12 05:50:16 +0000 | [diff] [blame] | 6200 | </pre> |
| Dan Gohman | c766f72 | 2009-01-22 01:39:38 +0000 | [diff] [blame] | 6201 | |
| Chris Lattner | cc37aae | 2004-03-12 05:50:16 +0000 | [diff] [blame] | 6202 | </div> |
| 6203 | |
| Robert Bocchino | 05ccd70 | 2006-01-15 20:48:27 +0000 | [diff] [blame] | 6204 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6205 | <h4> |
| Chris Lattner | 2bff524 | 2005-05-06 05:47:36 +0000 | [diff] [blame] | 6206 | <a name="i_call">'<tt>call</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6207 | </h4> |
| Chris Lattner | 2bff524 | 2005-05-06 05:47:36 +0000 | [diff] [blame] | 6208 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 6209 | <div> |
| Chris Lattner | 2bff524 | 2005-05-06 05:47:36 +0000 | [diff] [blame] | 6210 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 6211 | <h5>Syntax:</h5> |
| Chris Lattner | 2bff524 | 2005-05-06 05:47:36 +0000 | [diff] [blame] | 6212 | <pre> |
| Devang Patel | 307e8ab | 2008-10-07 17:48:33 +0000 | [diff] [blame] | 6213 | <result> = [tail] call [<a href="#callingconv">cconv</a>] [<a href="#paramattrs">ret attrs</a>] <ty> [<fnty>*] <fnptrval>(<function args>) [<a href="#fnattrs">fn attrs</a>] |
| Chris Lattner | 2bff524 | 2005-05-06 05:47:36 +0000 | [diff] [blame] | 6214 | </pre> |
| 6215 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 6216 | <h5>Overview:</h5> |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 6217 | <p>The '<tt>call</tt>' instruction represents a simple function call.</p> |
| Chris Lattner | 2bff524 | 2005-05-06 05:47:36 +0000 | [diff] [blame] | 6218 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 6219 | <h5>Arguments:</h5> |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 6220 | <p>This instruction requires several arguments:</p> |
| Chris Lattner | 2bff524 | 2005-05-06 05:47:36 +0000 | [diff] [blame] | 6221 | |
| Chris Lattner | 6536cfe | 2002-05-06 22:08:29 +0000 | [diff] [blame] | 6222 | <ol> |
| Jeffrey Yasskin | 95fa80a | 2010-01-09 19:44:16 +0000 | [diff] [blame] | 6223 | <li>The optional "tail" marker indicates that the callee function does not |
| 6224 | access any allocas or varargs in the caller. Note that calls may be |
| 6225 | marked "tail" even if they do not occur before |
| 6226 | a <a href="#i_ret"><tt>ret</tt></a> instruction. If the "tail" marker is |
| 6227 | present, the function call is eligible for tail call optimization, |
| 6228 | but <a href="CodeGenerator.html#tailcallopt">might not in fact be |
| Evan Cheng | dc444e9 | 2010-03-08 21:05:02 +0000 | [diff] [blame] | 6229 | optimized into a jump</a>. The code generator may optimize calls marked |
| 6230 | "tail" with either 1) automatic <a href="CodeGenerator.html#sibcallopt"> |
| 6231 | sibling call optimization</a> when the caller and callee have |
| 6232 | matching signatures, or 2) forced tail call optimization when the |
| 6233 | following extra requirements are met: |
| Jeffrey Yasskin | 95fa80a | 2010-01-09 19:44:16 +0000 | [diff] [blame] | 6234 | <ul> |
| 6235 | <li>Caller and callee both have the calling |
| 6236 | convention <tt>fastcc</tt>.</li> |
| 6237 | <li>The call is in tail position (ret immediately follows call and ret |
| 6238 | uses value of call or is void).</li> |
| 6239 | <li>Option <tt>-tailcallopt</tt> is enabled, |
| Dan Gohman | fbbee8d | 2010-03-02 01:08:11 +0000 | [diff] [blame] | 6240 | or <code>llvm::GuaranteedTailCallOpt</code> is <code>true</code>.</li> |
| Jeffrey Yasskin | 95fa80a | 2010-01-09 19:44:16 +0000 | [diff] [blame] | 6241 | <li><a href="CodeGenerator.html#tailcallopt">Platform specific |
| 6242 | constraints are met.</a></li> |
| 6243 | </ul> |
| 6244 | </li> |
| Devang Patel | f642f47 | 2008-10-06 18:50:38 +0000 | [diff] [blame] | 6245 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6246 | <li>The optional "cconv" marker indicates which <a href="#callingconv">calling |
| 6247 | convention</a> the call should use. If none is specified, the call |
| Jeffrey Yasskin | 95fa80a | 2010-01-09 19:44:16 +0000 | [diff] [blame] | 6248 | defaults to using C calling conventions. The calling convention of the |
| 6249 | call must match the calling convention of the target function, or else the |
| 6250 | behavior is undefined.</li> |
| Devang Patel | f642f47 | 2008-10-06 18:50:38 +0000 | [diff] [blame] | 6251 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6252 | <li>The optional <a href="#paramattrs">Parameter Attributes</a> list for |
| 6253 | return values. Only '<tt>zeroext</tt>', '<tt>signext</tt>', and |
| 6254 | '<tt>inreg</tt>' attributes are valid here.</li> |
| 6255 | |
| 6256 | <li>'<tt>ty</tt>': the type of the call instruction itself which is also the |
| 6257 | type of the return value. Functions that return no value are marked |
| 6258 | <tt><a href="#t_void">void</a></tt>.</li> |
| 6259 | |
| 6260 | <li>'<tt>fnty</tt>': shall be the signature of the pointer to function value |
| 6261 | being invoked. The argument types must match the types implied by this |
| 6262 | signature. This type can be omitted if the function is not varargs and if |
| 6263 | the function type does not return a pointer to a function.</li> |
| 6264 | |
| 6265 | <li>'<tt>fnptrval</tt>': An LLVM value containing a pointer to a function to |
| 6266 | be invoked. In most cases, this is a direct function invocation, but |
| 6267 | indirect <tt>call</tt>s are just as possible, calling an arbitrary pointer |
| 6268 | to function value.</li> |
| 6269 | |
| 6270 | <li>'<tt>function args</tt>': argument list whose types match the function |
| Chris Lattner | 0724fbd | 2010-03-02 06:36:51 +0000 | [diff] [blame] | 6271 | signature argument types and parameter attributes. All arguments must be |
| 6272 | of <a href="#t_firstclass">first class</a> type. If the function |
| 6273 | signature indicates the function accepts a variable number of arguments, |
| 6274 | the extra arguments can be specified.</li> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6275 | |
| 6276 | <li>The optional <a href="#fnattrs">function attributes</a> list. Only |
| 6277 | '<tt>noreturn</tt>', '<tt>nounwind</tt>', '<tt>readonly</tt>' and |
| 6278 | '<tt>readnone</tt>' attributes are valid here.</li> |
| Chris Lattner | 6536cfe | 2002-05-06 22:08:29 +0000 | [diff] [blame] | 6279 | </ol> |
| Chris Lattner | 2bff524 | 2005-05-06 05:47:36 +0000 | [diff] [blame] | 6280 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 6281 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6282 | <p>The '<tt>call</tt>' instruction is used to cause control flow to transfer to |
| 6283 | a specified function, with its incoming arguments bound to the specified |
| 6284 | values. Upon a '<tt><a href="#i_ret">ret</a></tt>' instruction in the called |
| 6285 | function, control flow continues with the instruction after the function |
| 6286 | call, and the return value of the function is bound to the result |
| 6287 | argument.</p> |
| Chris Lattner | 2bff524 | 2005-05-06 05:47:36 +0000 | [diff] [blame] | 6288 | |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 6289 | <h5>Example:</h5> |
| Chris Lattner | 2bff524 | 2005-05-06 05:47:36 +0000 | [diff] [blame] | 6290 | <pre> |
| Nick Lewycky | db7e3c9 | 2007-09-08 13:57:50 +0000 | [diff] [blame] | 6291 | %retval = call i32 @test(i32 %argc) |
| Dan Gohman | fe47aae | 2010-05-28 17:13:49 +0000 | [diff] [blame] | 6292 | call i32 (i8*, ...)* @printf(i8* %msg, i32 12, i8 42) <i>; yields i32</i> |
| Chris Lattner | 772fccf | 2008-03-21 17:24:17 +0000 | [diff] [blame] | 6293 | %X = tail call i32 @foo() <i>; yields i32</i> |
| 6294 | %Y = tail call <a href="#callingconv">fastcc</a> i32 @foo() <i>; yields i32</i> |
| 6295 | call void %foo(i8 97 signext) |
| Devang Patel | c3fc6df | 2008-03-10 20:49:15 +0000 | [diff] [blame] | 6296 | |
| 6297 | %struct.A = type { i32, i8 } |
| Devang Patel | f642f47 | 2008-10-06 18:50:38 +0000 | [diff] [blame] | 6298 | %r = call %struct.A @foo() <i>; yields { 32, i8 }</i> |
| Dan Gohman | b1e6b96 | 2008-10-04 19:00:07 +0000 | [diff] [blame] | 6299 | %gr = extractvalue %struct.A %r, 0 <i>; yields i32</i> |
| 6300 | %gr1 = extractvalue %struct.A %r, 1 <i>; yields i8</i> |
| Chris Lattner | 85a350f | 2008-10-08 06:26:11 +0000 | [diff] [blame] | 6301 | %Z = call void @foo() noreturn <i>; indicates that %foo never returns normally</i> |
| Matthijs Kooijman | cb73d19 | 2008-10-07 10:03:45 +0000 | [diff] [blame] | 6302 | %ZZ = call zeroext i32 @bar() <i>; Return value is %zero extended</i> |
| Chris Lattner | 2bff524 | 2005-05-06 05:47:36 +0000 | [diff] [blame] | 6303 | </pre> |
| 6304 | |
| Dale Johannesen | 07de8d1 | 2009-09-24 18:38:21 +0000 | [diff] [blame] | 6305 | <p>llvm treats calls to some functions with names and arguments that match the |
| Dale Johannesen | 9f8380b | 2009-09-25 17:04:42 +0000 | [diff] [blame] | 6306 | standard C99 library as being the C99 library functions, and may perform |
| 6307 | optimizations or generate code for them under that assumption. This is |
| 6308 | something we'd like to change in the future to provide better support for |
| Dan Gohman | b55a1ee | 2010-03-01 17:41:39 +0000 | [diff] [blame] | 6309 | freestanding environments and non-C-based languages.</p> |
| Dale Johannesen | 07de8d1 | 2009-09-24 18:38:21 +0000 | [diff] [blame] | 6310 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 6311 | </div> |
| Chris Lattner | e19d7a7 | 2004-09-27 21:51:25 +0000 | [diff] [blame] | 6312 | |
| Chris Lattner | d9ad5b3 | 2003-05-08 04:57:36 +0000 | [diff] [blame] | 6313 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6314 | <h4> |
| Chris Lattner | fb6977d | 2006-01-13 23:26:01 +0000 | [diff] [blame] | 6315 | <a name="i_va_arg">'<tt>va_arg</tt>' Instruction</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6316 | </h4> |
| Chris Lattner | e19d7a7 | 2004-09-27 21:51:25 +0000 | [diff] [blame] | 6317 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 6318 | <div> |
| Chris Lattner | e19d7a7 | 2004-09-27 21:51:25 +0000 | [diff] [blame] | 6319 | |
| Chris Lattner | 8d1a81d | 2003-10-18 05:51:36 +0000 | [diff] [blame] | 6320 | <h5>Syntax:</h5> |
| Chris Lattner | e19d7a7 | 2004-09-27 21:51:25 +0000 | [diff] [blame] | 6321 | <pre> |
| Andrew Lenharth | 8bf607a | 2005-06-18 18:28:17 +0000 | [diff] [blame] | 6322 | <resultval> = va_arg <va_list*> <arglist>, <argty> |
| Chris Lattner | e19d7a7 | 2004-09-27 21:51:25 +0000 | [diff] [blame] | 6323 | </pre> |
| 6324 | |
| Chris Lattner | 8d1a81d | 2003-10-18 05:51:36 +0000 | [diff] [blame] | 6325 | <h5>Overview:</h5> |
| Andrew Lenharth | 8bf607a | 2005-06-18 18:28:17 +0000 | [diff] [blame] | 6326 | <p>The '<tt>va_arg</tt>' instruction is used to access arguments passed through |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6327 | the "variable argument" area of a function call. It is used to implement the |
| 6328 | <tt>va_arg</tt> macro in C.</p> |
| Chris Lattner | e19d7a7 | 2004-09-27 21:51:25 +0000 | [diff] [blame] | 6329 | |
| Chris Lattner | 8d1a81d | 2003-10-18 05:51:36 +0000 | [diff] [blame] | 6330 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6331 | <p>This instruction takes a <tt>va_list*</tt> value and the type of the |
| 6332 | argument. It returns a value of the specified argument type and increments |
| 6333 | the <tt>va_list</tt> to point to the next argument. The actual type |
| 6334 | of <tt>va_list</tt> is target specific.</p> |
| Chris Lattner | e19d7a7 | 2004-09-27 21:51:25 +0000 | [diff] [blame] | 6335 | |
| Chris Lattner | 8d1a81d | 2003-10-18 05:51:36 +0000 | [diff] [blame] | 6336 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6337 | <p>The '<tt>va_arg</tt>' instruction loads an argument of the specified type |
| 6338 | from the specified <tt>va_list</tt> and causes the <tt>va_list</tt> to point |
| 6339 | to the next argument. For more information, see the variable argument |
| 6340 | handling <a href="#int_varargs">Intrinsic Functions</a>.</p> |
| Chris Lattner | e19d7a7 | 2004-09-27 21:51:25 +0000 | [diff] [blame] | 6341 | |
| 6342 | <p>It is legal for this instruction to be called in a function which does not |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6343 | take a variable number of arguments, for example, the <tt>vfprintf</tt> |
| 6344 | function.</p> |
| Chris Lattner | e19d7a7 | 2004-09-27 21:51:25 +0000 | [diff] [blame] | 6345 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6346 | <p><tt>va_arg</tt> is an LLVM instruction instead of |
| 6347 | an <a href="#intrinsics">intrinsic function</a> because it takes a type as an |
| 6348 | argument.</p> |
| Chris Lattner | e19d7a7 | 2004-09-27 21:51:25 +0000 | [diff] [blame] | 6349 | |
| Chris Lattner | 8d1a81d | 2003-10-18 05:51:36 +0000 | [diff] [blame] | 6350 | <h5>Example:</h5> |
| Chris Lattner | e19d7a7 | 2004-09-27 21:51:25 +0000 | [diff] [blame] | 6351 | <p>See the <a href="#int_varargs">variable argument processing</a> section.</p> |
| 6352 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6353 | <p>Note that the code generator does not yet fully support va_arg on many |
| 6354 | targets. Also, it does not currently support va_arg with aggregate types on |
| 6355 | any target.</p> |
| Dan Gohman | f3e60bd | 2009-01-12 23:12:39 +0000 | [diff] [blame] | 6356 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 6357 | </div> |
| Chris Lattner | 8ff7590 | 2004-01-06 05:31:32 +0000 | [diff] [blame] | 6358 | |
| Bill Wendling | f78faf8 | 2011-08-02 21:52:38 +0000 | [diff] [blame] | 6359 | <!-- _______________________________________________________________________ --> |
| 6360 | <h4> |
| 6361 | <a name="i_landingpad">'<tt>landingpad</tt>' Instruction</a> |
| 6362 | </h4> |
| 6363 | |
| 6364 | <div> |
| 6365 | |
| 6366 | <h5>Syntax:</h5> |
| 6367 | <pre> |
| Duncan Sands | 8d6796b | 2012-01-13 19:59:16 +0000 | [diff] [blame] | 6368 | <resultval> = landingpad <resultty> personality <type> <pers_fn> <clause>+ |
| 6369 | <resultval> = landingpad <resultty> personality <type> <pers_fn> cleanup <clause>* |
| Bill Wendling | bf13ee1 | 2011-08-08 08:06:05 +0000 | [diff] [blame] | 6370 | |
| Bill Wendling | f78faf8 | 2011-08-02 21:52:38 +0000 | [diff] [blame] | 6371 | <clause> := catch <type> <value> |
| Bill Wendling | e6e8826 | 2011-08-12 20:24:12 +0000 | [diff] [blame] | 6372 | <clause> := filter <array constant type> <array constant> |
| Bill Wendling | f78faf8 | 2011-08-02 21:52:38 +0000 | [diff] [blame] | 6373 | </pre> |
| 6374 | |
| 6375 | <h5>Overview:</h5> |
| 6376 | <p>The '<tt>landingpad</tt>' instruction is used by |
| 6377 | <a href="ExceptionHandling.html#overview">LLVM's exception handling |
| 6378 | system</a> to specify that a basic block is a landing pad — one where |
| 6379 | the exception lands, and corresponds to the code found in the |
| 6380 | <i><tt>catch</tt></i> portion of a <i><tt>try/catch</tt></i> sequence. It |
| 6381 | defines values supplied by the personality function (<tt>pers_fn</tt>) upon |
| 6382 | re-entry to the function. The <tt>resultval</tt> has the |
| Duncan Sands | 8d6796b | 2012-01-13 19:59:16 +0000 | [diff] [blame] | 6383 | type <tt>resultty</tt>.</p> |
| Bill Wendling | f78faf8 | 2011-08-02 21:52:38 +0000 | [diff] [blame] | 6384 | |
| 6385 | <h5>Arguments:</h5> |
| 6386 | <p>This instruction takes a <tt>pers_fn</tt> value. This is the personality |
| 6387 | function associated with the unwinding mechanism. The optional |
| 6388 | <tt>cleanup</tt> flag indicates that the landing pad block is a cleanup.</p> |
| 6389 | |
| 6390 | <p>A <tt>clause</tt> begins with the clause type — <tt>catch</tt> |
| Bill Wendling | e6e8826 | 2011-08-12 20:24:12 +0000 | [diff] [blame] | 6391 | or <tt>filter</tt> — and contains the global variable representing the |
| 6392 | "type" that may be caught or filtered respectively. Unlike the |
| 6393 | <tt>catch</tt> clause, the <tt>filter</tt> clause takes an array constant as |
| 6394 | its argument. Use "<tt>[0 x i8**] undef</tt>" for a filter which cannot |
| 6395 | throw. The '<tt>landingpad</tt>' instruction must contain <em>at least</em> |
| Bill Wendling | f78faf8 | 2011-08-02 21:52:38 +0000 | [diff] [blame] | 6396 | one <tt>clause</tt> or the <tt>cleanup</tt> flag.</p> |
| 6397 | |
| 6398 | <h5>Semantics:</h5> |
| 6399 | <p>The '<tt>landingpad</tt>' instruction defines the values which are set by the |
| 6400 | personality function (<tt>pers_fn</tt>) upon re-entry to the function, and |
| 6401 | therefore the "result type" of the <tt>landingpad</tt> instruction. As with |
| 6402 | calling conventions, how the personality function results are represented in |
| 6403 | LLVM IR is target specific.</p> |
| 6404 | |
| Bill Wendling | b7a0135 | 2011-08-03 17:17:06 +0000 | [diff] [blame] | 6405 | <p>The clauses are applied in order from top to bottom. If two |
| 6406 | <tt>landingpad</tt> instructions are merged together through inlining, the |
| Duncan Sands | 8d6796b | 2012-01-13 19:59:16 +0000 | [diff] [blame] | 6407 | clauses from the calling function are appended to the list of clauses. |
| 6408 | When the call stack is being unwound due to an exception being thrown, the |
| 6409 | exception is compared against each <tt>clause</tt> in turn. If it doesn't |
| 6410 | match any of the clauses, and the <tt>cleanup</tt> flag is not set, then |
| 6411 | unwinding continues further up the call stack.</p> |
| Bill Wendling | b7a0135 | 2011-08-03 17:17:06 +0000 | [diff] [blame] | 6412 | |
| Bill Wendling | f78faf8 | 2011-08-02 21:52:38 +0000 | [diff] [blame] | 6413 | <p>The <tt>landingpad</tt> instruction has several restrictions:</p> |
| 6414 | |
| 6415 | <ul> |
| 6416 | <li>A landing pad block is a basic block which is the unwind destination of an |
| 6417 | '<tt>invoke</tt>' instruction.</li> |
| 6418 | <li>A landing pad block must have a '<tt>landingpad</tt>' instruction as its |
| 6419 | first non-PHI instruction.</li> |
| 6420 | <li>There can be only one '<tt>landingpad</tt>' instruction within the landing |
| 6421 | pad block.</li> |
| 6422 | <li>A basic block that is not a landing pad block may not include a |
| 6423 | '<tt>landingpad</tt>' instruction.</li> |
| 6424 | <li>All '<tt>landingpad</tt>' instructions in a function must have the same |
| 6425 | personality function.</li> |
| 6426 | </ul> |
| 6427 | |
| 6428 | <h5>Example:</h5> |
| 6429 | <pre> |
| 6430 | ;; A landing pad which can catch an integer. |
| 6431 | %res = landingpad { i8*, i32 } personality i32 (...)* @__gxx_personality_v0 |
| 6432 | catch i8** @_ZTIi |
| 6433 | ;; A landing pad that is a cleanup. |
| 6434 | %res = landingpad { i8*, i32 } personality i32 (...)* @__gxx_personality_v0 |
| Bill Wendling | e6e8826 | 2011-08-12 20:24:12 +0000 | [diff] [blame] | 6435 | cleanup |
| Bill Wendling | f78faf8 | 2011-08-02 21:52:38 +0000 | [diff] [blame] | 6436 | ;; A landing pad which can catch an integer and can only throw a double. |
| 6437 | %res = landingpad { i8*, i32 } personality i32 (...)* @__gxx_personality_v0 |
| 6438 | catch i8** @_ZTIi |
| Bill Wendling | e6e8826 | 2011-08-12 20:24:12 +0000 | [diff] [blame] | 6439 | filter [1 x i8**] [@_ZTId] |
| Bill Wendling | f78faf8 | 2011-08-02 21:52:38 +0000 | [diff] [blame] | 6440 | </pre> |
| 6441 | |
| 6442 | </div> |
| 6443 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 6444 | </div> |
| 6445 | |
| 6446 | </div> |
| 6447 | |
| Chris Lattner | d9ad5b3 | 2003-05-08 04:57:36 +0000 | [diff] [blame] | 6448 | <!-- *********************************************************************** --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6449 | <h2><a name="intrinsics">Intrinsic Functions</a></h2> |
| Chris Lattner | 261efe9 | 2003-11-25 01:02:51 +0000 | [diff] [blame] | 6450 | <!-- *********************************************************************** --> |
| Chris Lattner | 8ff7590 | 2004-01-06 05:31:32 +0000 | [diff] [blame] | 6451 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 6452 | <div> |
| Chris Lattner | 33aec9e | 2004-02-12 17:01:32 +0000 | [diff] [blame] | 6453 | |
| 6454 | <p>LLVM supports the notion of an "intrinsic function". These functions have |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6455 | well known names and semantics and are required to follow certain |
| 6456 | restrictions. Overall, these intrinsics represent an extension mechanism for |
| 6457 | the LLVM language that does not require changing all of the transformations |
| 6458 | in LLVM when adding to the language (or the bitcode reader/writer, the |
| 6459 | parser, etc...).</p> |
| Chris Lattner | 33aec9e | 2004-02-12 17:01:32 +0000 | [diff] [blame] | 6460 | |
| John Criswell | fc6b895 | 2005-05-16 16:17:45 +0000 | [diff] [blame] | 6461 | <p>Intrinsic function names must all start with an "<tt>llvm.</tt>" prefix. This |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6462 | prefix is reserved in LLVM for intrinsic names; thus, function names may not |
| 6463 | begin with this prefix. Intrinsic functions must always be external |
| 6464 | functions: you cannot define the body of intrinsic functions. Intrinsic |
| 6465 | functions may only be used in call or invoke instructions: it is illegal to |
| 6466 | take the address of an intrinsic function. Additionally, because intrinsic |
| 6467 | functions are part of the LLVM language, it is required if any are added that |
| 6468 | they be documented here.</p> |
| Chris Lattner | 33aec9e | 2004-02-12 17:01:32 +0000 | [diff] [blame] | 6469 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6470 | <p>Some intrinsic functions can be overloaded, i.e., the intrinsic represents a |
| 6471 | family of functions that perform the same operation but on different data |
| 6472 | types. Because LLVM can represent over 8 million different integer types, |
| 6473 | overloading is used commonly to allow an intrinsic function to operate on any |
| 6474 | integer type. One or more of the argument types or the result type can be |
| 6475 | overloaded to accept any integer type. Argument types may also be defined as |
| 6476 | exactly matching a previous argument's type or the result type. This allows |
| 6477 | an intrinsic function which accepts multiple arguments, but needs all of them |
| 6478 | to be of the same type, to only be overloaded with respect to a single |
| 6479 | argument or the result.</p> |
| Chris Lattner | 33aec9e | 2004-02-12 17:01:32 +0000 | [diff] [blame] | 6480 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6481 | <p>Overloaded intrinsics will have the names of its overloaded argument types |
| 6482 | encoded into its function name, each preceded by a period. Only those types |
| 6483 | which are overloaded result in a name suffix. Arguments whose type is matched |
| 6484 | against another type do not. For example, the <tt>llvm.ctpop</tt> function |
| 6485 | can take an integer of any width and returns an integer of exactly the same |
| 6486 | integer width. This leads to a family of functions such as |
| 6487 | <tt>i8 @llvm.ctpop.i8(i8 %val)</tt> and <tt>i29 @llvm.ctpop.i29(i29 |
| 6488 | %val)</tt>. Only one type, the return type, is overloaded, and only one type |
| 6489 | suffix is required. Because the argument's type is matched against the return |
| 6490 | type, it does not require its own name suffix.</p> |
| Reid Spencer | 409e28f | 2007-04-01 08:04:23 +0000 | [diff] [blame] | 6491 | |
| Eric Christopher | 6c7e8a0 | 2009-12-05 02:46:03 +0000 | [diff] [blame] | 6492 | <p>To learn how to add an intrinsic function, please see the |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6493 | <a href="ExtendingLLVM.html">Extending LLVM Guide</a>.</p> |
| Chris Lattner | 33aec9e | 2004-02-12 17:01:32 +0000 | [diff] [blame] | 6494 | |
| Chris Lattner | d9ad5b3 | 2003-05-08 04:57:36 +0000 | [diff] [blame] | 6495 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6496 | <h3> |
| Chris Lattner | 8ff7590 | 2004-01-06 05:31:32 +0000 | [diff] [blame] | 6497 | <a name="int_varargs">Variable Argument Handling Intrinsics</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6498 | </h3> |
| Chris Lattner | 8ff7590 | 2004-01-06 05:31:32 +0000 | [diff] [blame] | 6499 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 6500 | <div> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6501 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6502 | <p>Variable argument support is defined in LLVM with |
| 6503 | the <a href="#i_va_arg"><tt>va_arg</tt></a> instruction and these three |
| 6504 | intrinsic functions. These functions are related to the similarly named |
| 6505 | macros defined in the <tt><stdarg.h></tt> header file.</p> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6506 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6507 | <p>All of these functions operate on arguments that use a target-specific value |
| 6508 | type "<tt>va_list</tt>". The LLVM assembly language reference manual does |
| 6509 | not define what this type is, so all transformations should be prepared to |
| 6510 | handle these functions regardless of the type used.</p> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6511 | |
| Chris Lattner | 374ab30 | 2006-05-15 17:26:46 +0000 | [diff] [blame] | 6512 | <p>This example shows how the <a href="#i_va_arg"><tt>va_arg</tt></a> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6513 | instruction and the variable argument handling intrinsic functions are |
| 6514 | used.</p> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6515 | |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 6516 | <pre class="doc_code"> |
| Anton Korobeynikov | 5d522f3 | 2007-03-21 23:58:04 +0000 | [diff] [blame] | 6517 | define i32 @test(i32 %X, ...) { |
| Chris Lattner | 33aec9e | 2004-02-12 17:01:32 +0000 | [diff] [blame] | 6518 | ; Initialize variable argument processing |
| Jeff Cohen | b627eab | 2007-04-29 01:07:00 +0000 | [diff] [blame] | 6519 | %ap = alloca i8* |
| Chris Lattner | b75137d | 2007-01-08 07:55:15 +0000 | [diff] [blame] | 6520 | %ap2 = bitcast i8** %ap to i8* |
| Anton Korobeynikov | 5d522f3 | 2007-03-21 23:58:04 +0000 | [diff] [blame] | 6521 | call void @llvm.va_start(i8* %ap2) |
| Chris Lattner | 33aec9e | 2004-02-12 17:01:32 +0000 | [diff] [blame] | 6522 | |
| 6523 | ; Read a single integer argument |
| Jeff Cohen | b627eab | 2007-04-29 01:07:00 +0000 | [diff] [blame] | 6524 | %tmp = va_arg i8** %ap, i32 |
| Chris Lattner | 33aec9e | 2004-02-12 17:01:32 +0000 | [diff] [blame] | 6525 | |
| 6526 | ; Demonstrate usage of llvm.va_copy and llvm.va_end |
| Jeff Cohen | b627eab | 2007-04-29 01:07:00 +0000 | [diff] [blame] | 6527 | %aq = alloca i8* |
| Chris Lattner | b75137d | 2007-01-08 07:55:15 +0000 | [diff] [blame] | 6528 | %aq2 = bitcast i8** %aq to i8* |
| Jeff Cohen | b627eab | 2007-04-29 01:07:00 +0000 | [diff] [blame] | 6529 | call void @llvm.va_copy(i8* %aq2, i8* %ap2) |
| Anton Korobeynikov | 5d522f3 | 2007-03-21 23:58:04 +0000 | [diff] [blame] | 6530 | call void @llvm.va_end(i8* %aq2) |
| Chris Lattner | 33aec9e | 2004-02-12 17:01:32 +0000 | [diff] [blame] | 6531 | |
| 6532 | ; Stop processing of arguments. |
| Anton Korobeynikov | 5d522f3 | 2007-03-21 23:58:04 +0000 | [diff] [blame] | 6533 | call void @llvm.va_end(i8* %ap2) |
| Reid Spencer | ca86e16 | 2006-12-31 07:07:53 +0000 | [diff] [blame] | 6534 | ret i32 %tmp |
| Chris Lattner | 33aec9e | 2004-02-12 17:01:32 +0000 | [diff] [blame] | 6535 | } |
| Anton Korobeynikov | 5d522f3 | 2007-03-21 23:58:04 +0000 | [diff] [blame] | 6536 | |
| 6537 | declare void @llvm.va_start(i8*) |
| 6538 | declare void @llvm.va_copy(i8*, i8*) |
| 6539 | declare void @llvm.va_end(i8*) |
| Chris Lattner | 33aec9e | 2004-02-12 17:01:32 +0000 | [diff] [blame] | 6540 | </pre> |
| Chris Lattner | 8ff7590 | 2004-01-06 05:31:32 +0000 | [diff] [blame] | 6541 | |
| Chris Lattner | d9ad5b3 | 2003-05-08 04:57:36 +0000 | [diff] [blame] | 6542 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6543 | <h4> |
| Reid Spencer | a3e435f | 2007-04-04 02:42:35 +0000 | [diff] [blame] | 6544 | <a name="int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6545 | </h4> |
| Chris Lattner | 8ff7590 | 2004-01-06 05:31:32 +0000 | [diff] [blame] | 6546 | |
| 6547 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 6548 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6549 | |
| Chris Lattner | d9ad5b3 | 2003-05-08 04:57:36 +0000 | [diff] [blame] | 6550 | <h5>Syntax:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6551 | <pre> |
| 6552 | declare void %llvm.va_start(i8* <arglist>) |
| 6553 | </pre> |
| 6554 | |
| Chris Lattner | d9ad5b3 | 2003-05-08 04:57:36 +0000 | [diff] [blame] | 6555 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6556 | <p>The '<tt>llvm.va_start</tt>' intrinsic initializes <tt>*<arglist></tt> |
| 6557 | for subsequent use by <tt><a href="#i_va_arg">va_arg</a></tt>.</p> |
| Andrew Lenharth | 8bf607a | 2005-06-18 18:28:17 +0000 | [diff] [blame] | 6558 | |
| 6559 | <h5>Arguments:</h5> |
| Dan Gohman | 0e451ce | 2008-10-14 16:51:45 +0000 | [diff] [blame] | 6560 | <p>The argument is a pointer to a <tt>va_list</tt> element to initialize.</p> |
| Andrew Lenharth | 8bf607a | 2005-06-18 18:28:17 +0000 | [diff] [blame] | 6561 | |
| Chris Lattner | d9ad5b3 | 2003-05-08 04:57:36 +0000 | [diff] [blame] | 6562 | <h5>Semantics:</h5> |
| Dan Gohman | 0e451ce | 2008-10-14 16:51:45 +0000 | [diff] [blame] | 6563 | <p>The '<tt>llvm.va_start</tt>' intrinsic works just like the <tt>va_start</tt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6564 | macro available in C. In a target-dependent way, it initializes |
| 6565 | the <tt>va_list</tt> element to which the argument points, so that the next |
| 6566 | call to <tt>va_arg</tt> will produce the first variable argument passed to |
| 6567 | the function. Unlike the C <tt>va_start</tt> macro, this intrinsic does not |
| 6568 | need to know the last argument of the function as the compiler can figure |
| 6569 | that out.</p> |
| Andrew Lenharth | 8bf607a | 2005-06-18 18:28:17 +0000 | [diff] [blame] | 6570 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 6571 | </div> |
| Chris Lattner | 8ff7590 | 2004-01-06 05:31:32 +0000 | [diff] [blame] | 6572 | |
| Chris Lattner | d9ad5b3 | 2003-05-08 04:57:36 +0000 | [diff] [blame] | 6573 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6574 | <h4> |
| Reid Spencer | a3e435f | 2007-04-04 02:42:35 +0000 | [diff] [blame] | 6575 | <a name="int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6576 | </h4> |
| Chris Lattner | 8ff7590 | 2004-01-06 05:31:32 +0000 | [diff] [blame] | 6577 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 6578 | <div> |
| Chris Lattner | b75137d | 2007-01-08 07:55:15 +0000 | [diff] [blame] | 6579 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6580 | <h5>Syntax:</h5> |
| 6581 | <pre> |
| 6582 | declare void @llvm.va_end(i8* <arglist>) |
| 6583 | </pre> |
| 6584 | |
| 6585 | <h5>Overview:</h5> |
| Jeff Cohen | b627eab | 2007-04-29 01:07:00 +0000 | [diff] [blame] | 6586 | <p>The '<tt>llvm.va_end</tt>' intrinsic destroys <tt>*<arglist></tt>, |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6587 | which has been initialized previously |
| 6588 | with <tt><a href="#int_va_start">llvm.va_start</a></tt> |
| 6589 | or <tt><a href="#i_va_copy">llvm.va_copy</a></tt>.</p> |
| Chris Lattner | b75137d | 2007-01-08 07:55:15 +0000 | [diff] [blame] | 6590 | |
| Chris Lattner | d9ad5b3 | 2003-05-08 04:57:36 +0000 | [diff] [blame] | 6591 | <h5>Arguments:</h5> |
| Jeff Cohen | b627eab | 2007-04-29 01:07:00 +0000 | [diff] [blame] | 6592 | <p>The argument is a pointer to a <tt>va_list</tt> to destroy.</p> |
| Chris Lattner | b75137d | 2007-01-08 07:55:15 +0000 | [diff] [blame] | 6593 | |
| Chris Lattner | d9ad5b3 | 2003-05-08 04:57:36 +0000 | [diff] [blame] | 6594 | <h5>Semantics:</h5> |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 6595 | <p>The '<tt>llvm.va_end</tt>' intrinsic works just like the <tt>va_end</tt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6596 | macro available in C. In a target-dependent way, it destroys |
| 6597 | the <tt>va_list</tt> element to which the argument points. Calls |
| 6598 | to <a href="#int_va_start"><tt>llvm.va_start</tt></a> |
| 6599 | and <a href="#int_va_copy"> <tt>llvm.va_copy</tt></a> must be matched exactly |
| 6600 | with calls to <tt>llvm.va_end</tt>.</p> |
| Chris Lattner | b75137d | 2007-01-08 07:55:15 +0000 | [diff] [blame] | 6601 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 6602 | </div> |
| Chris Lattner | 8ff7590 | 2004-01-06 05:31:32 +0000 | [diff] [blame] | 6603 | |
| Chris Lattner | d9ad5b3 | 2003-05-08 04:57:36 +0000 | [diff] [blame] | 6604 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6605 | <h4> |
| Reid Spencer | a3e435f | 2007-04-04 02:42:35 +0000 | [diff] [blame] | 6606 | <a name="int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6607 | </h4> |
| Chris Lattner | 8ff7590 | 2004-01-06 05:31:32 +0000 | [diff] [blame] | 6608 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 6609 | <div> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6610 | |
| Chris Lattner | d9ad5b3 | 2003-05-08 04:57:36 +0000 | [diff] [blame] | 6611 | <h5>Syntax:</h5> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6612 | <pre> |
| Anton Korobeynikov | ec43a06 | 2007-03-22 00:02:17 +0000 | [diff] [blame] | 6613 | declare void @llvm.va_copy(i8* <destarglist>, i8* <srcarglist>) |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6614 | </pre> |
| 6615 | |
| Chris Lattner | d9ad5b3 | 2003-05-08 04:57:36 +0000 | [diff] [blame] | 6616 | <h5>Overview:</h5> |
| Jeff Cohen | b627eab | 2007-04-29 01:07:00 +0000 | [diff] [blame] | 6617 | <p>The '<tt>llvm.va_copy</tt>' intrinsic copies the current argument position |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6618 | from the source argument list to the destination argument list.</p> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6619 | |
| Chris Lattner | d9ad5b3 | 2003-05-08 04:57:36 +0000 | [diff] [blame] | 6620 | <h5>Arguments:</h5> |
| Andrew Lenharth | 8bf607a | 2005-06-18 18:28:17 +0000 | [diff] [blame] | 6621 | <p>The first argument is a pointer to a <tt>va_list</tt> element to initialize. |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6622 | The second argument is a pointer to a <tt>va_list</tt> element to copy |
| 6623 | from.</p> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6624 | |
| Chris Lattner | d9ad5b3 | 2003-05-08 04:57:36 +0000 | [diff] [blame] | 6625 | <h5>Semantics:</h5> |
| Jeff Cohen | b627eab | 2007-04-29 01:07:00 +0000 | [diff] [blame] | 6626 | <p>The '<tt>llvm.va_copy</tt>' intrinsic works just like the <tt>va_copy</tt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6627 | macro available in C. In a target-dependent way, it copies the |
| 6628 | source <tt>va_list</tt> element into the destination <tt>va_list</tt> |
| 6629 | element. This intrinsic is necessary because |
| 6630 | the <tt><a href="#int_va_start"> llvm.va_start</a></tt> intrinsic may be |
| 6631 | arbitrarily complex and require, for example, memory allocation.</p> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6632 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 6633 | </div> |
| Chris Lattner | 8ff7590 | 2004-01-06 05:31:32 +0000 | [diff] [blame] | 6634 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 6635 | </div> |
| 6636 | |
| Chris Lattner | 33aec9e | 2004-02-12 17:01:32 +0000 | [diff] [blame] | 6637 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6638 | <h3> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6639 | <a name="int_gc">Accurate Garbage Collection Intrinsics</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6640 | </h3> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6641 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 6642 | <div> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6643 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6644 | <p>LLVM support for <a href="GarbageCollection.html">Accurate Garbage |
| Chris Lattner | d3eda89 | 2008-08-05 18:29:16 +0000 | [diff] [blame] | 6645 | Collection</a> (GC) requires the implementation and generation of these |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6646 | intrinsics. These intrinsics allow identification of <a href="#int_gcroot">GC |
| 6647 | roots on the stack</a>, as well as garbage collector implementations that |
| 6648 | require <a href="#int_gcread">read</a> and <a href="#int_gcwrite">write</a> |
| 6649 | barriers. Front-ends for type-safe garbage collected languages should generate |
| 6650 | these intrinsics to make use of the LLVM garbage collectors. For more details, |
| 6651 | see <a href="GarbageCollection.html">Accurate Garbage Collection with |
| 6652 | LLVM</a>.</p> |
| Christopher Lamb | 303dae9 | 2007-12-17 01:00:21 +0000 | [diff] [blame] | 6653 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6654 | <p>The garbage collection intrinsics only operate on objects in the generic |
| 6655 | address space (address space zero).</p> |
| Christopher Lamb | 303dae9 | 2007-12-17 01:00:21 +0000 | [diff] [blame] | 6656 | |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6657 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6658 | <h4> |
| Reid Spencer | a3e435f | 2007-04-04 02:42:35 +0000 | [diff] [blame] | 6659 | <a name="int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6660 | </h4> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6661 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 6662 | <div> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6663 | |
| 6664 | <h5>Syntax:</h5> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6665 | <pre> |
| Chris Lattner | 1df4f75 | 2007-09-21 17:30:40 +0000 | [diff] [blame] | 6666 | declare void @llvm.gcroot(i8** %ptrloc, i8* %metadata) |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6667 | </pre> |
| 6668 | |
| 6669 | <h5>Overview:</h5> |
| John Criswell | 9e2485c | 2004-12-10 15:51:16 +0000 | [diff] [blame] | 6670 | <p>The '<tt>llvm.gcroot</tt>' intrinsic declares the existence of a GC root to |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6671 | the code generator, and allows some metadata to be associated with it.</p> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6672 | |
| 6673 | <h5>Arguments:</h5> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6674 | <p>The first argument specifies the address of a stack object that contains the |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6675 | root pointer. The second pointer (which must be either a constant or a |
| 6676 | global value address) contains the meta-data to be associated with the |
| 6677 | root.</p> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6678 | |
| 6679 | <h5>Semantics:</h5> |
| Chris Lattner | 05d6709 | 2008-04-24 05:59:56 +0000 | [diff] [blame] | 6680 | <p>At runtime, a call to this intrinsic stores a null pointer into the "ptrloc" |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6681 | location. At compile-time, the code generator generates information to allow |
| 6682 | the runtime to find the pointer at GC safe points. The '<tt>llvm.gcroot</tt>' |
| 6683 | intrinsic may only be used in a function which <a href="#gc">specifies a GC |
| 6684 | algorithm</a>.</p> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6685 | |
| 6686 | </div> |
| 6687 | |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6688 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6689 | <h4> |
| Reid Spencer | a3e435f | 2007-04-04 02:42:35 +0000 | [diff] [blame] | 6690 | <a name="int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6691 | </h4> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6692 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 6693 | <div> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6694 | |
| 6695 | <h5>Syntax:</h5> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6696 | <pre> |
| Chris Lattner | 1df4f75 | 2007-09-21 17:30:40 +0000 | [diff] [blame] | 6697 | declare i8* @llvm.gcread(i8* %ObjPtr, i8** %Ptr) |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6698 | </pre> |
| 6699 | |
| 6700 | <h5>Overview:</h5> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6701 | <p>The '<tt>llvm.gcread</tt>' intrinsic identifies reads of references from heap |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6702 | locations, allowing garbage collector implementations that require read |
| 6703 | barriers.</p> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6704 | |
| 6705 | <h5>Arguments:</h5> |
| Chris Lattner | 80626e9 | 2006-03-14 20:02:51 +0000 | [diff] [blame] | 6706 | <p>The second argument is the address to read from, which should be an address |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6707 | allocated from the garbage collector. The first object is a pointer to the |
| 6708 | start of the referenced object, if needed by the language runtime (otherwise |
| 6709 | null).</p> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6710 | |
| 6711 | <h5>Semantics:</h5> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6712 | <p>The '<tt>llvm.gcread</tt>' intrinsic has the same semantics as a load |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6713 | instruction, but may be replaced with substantially more complex code by the |
| 6714 | garbage collector runtime, as needed. The '<tt>llvm.gcread</tt>' intrinsic |
| 6715 | may only be used in a function which <a href="#gc">specifies a GC |
| 6716 | algorithm</a>.</p> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6717 | |
| 6718 | </div> |
| 6719 | |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6720 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6721 | <h4> |
| Reid Spencer | a3e435f | 2007-04-04 02:42:35 +0000 | [diff] [blame] | 6722 | <a name="int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6723 | </h4> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6724 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 6725 | <div> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6726 | |
| 6727 | <h5>Syntax:</h5> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6728 | <pre> |
| Chris Lattner | 1df4f75 | 2007-09-21 17:30:40 +0000 | [diff] [blame] | 6729 | declare void @llvm.gcwrite(i8* %P1, i8* %Obj, i8** %P2) |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6730 | </pre> |
| 6731 | |
| 6732 | <h5>Overview:</h5> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6733 | <p>The '<tt>llvm.gcwrite</tt>' intrinsic identifies writes of references to heap |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6734 | locations, allowing garbage collector implementations that require write |
| 6735 | barriers (such as generational or reference counting collectors).</p> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6736 | |
| 6737 | <h5>Arguments:</h5> |
| Chris Lattner | 80626e9 | 2006-03-14 20:02:51 +0000 | [diff] [blame] | 6738 | <p>The first argument is the reference to store, the second is the start of the |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6739 | object to store it to, and the third is the address of the field of Obj to |
| 6740 | store to. If the runtime does not require a pointer to the object, Obj may |
| 6741 | be null.</p> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6742 | |
| 6743 | <h5>Semantics:</h5> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6744 | <p>The '<tt>llvm.gcwrite</tt>' intrinsic has the same semantics as a store |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6745 | instruction, but may be replaced with substantially more complex code by the |
| 6746 | garbage collector runtime, as needed. The '<tt>llvm.gcwrite</tt>' intrinsic |
| 6747 | may only be used in a function which <a href="#gc">specifies a GC |
| 6748 | algorithm</a>.</p> |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6749 | |
| 6750 | </div> |
| 6751 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 6752 | </div> |
| 6753 | |
| Chris Lattner | d792391 | 2004-05-23 21:06:01 +0000 | [diff] [blame] | 6754 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6755 | <h3> |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 6756 | <a name="int_codegen">Code Generator Intrinsics</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6757 | </h3> |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 6758 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 6759 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6760 | |
| 6761 | <p>These intrinsics are provided by LLVM to expose special features that may |
| 6762 | only be implemented with code generator support.</p> |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 6763 | |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 6764 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6765 | <h4> |
| Reid Spencer | a3e435f | 2007-04-04 02:42:35 +0000 | [diff] [blame] | 6766 | <a name="int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6767 | </h4> |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 6768 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 6769 | <div> |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 6770 | |
| 6771 | <h5>Syntax:</h5> |
| 6772 | <pre> |
| Anton Korobeynikov | ec43a06 | 2007-03-22 00:02:17 +0000 | [diff] [blame] | 6773 | declare i8 *@llvm.returnaddress(i32 <level>) |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 6774 | </pre> |
| 6775 | |
| 6776 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6777 | <p>The '<tt>llvm.returnaddress</tt>' intrinsic attempts to compute a |
| 6778 | target-specific value indicating the return address of the current function |
| 6779 | or one of its callers.</p> |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 6780 | |
| 6781 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6782 | <p>The argument to this intrinsic indicates which function to return the address |
| 6783 | for. Zero indicates the calling function, one indicates its caller, etc. |
| 6784 | The argument is <b>required</b> to be a constant integer value.</p> |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 6785 | |
| 6786 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6787 | <p>The '<tt>llvm.returnaddress</tt>' intrinsic either returns a pointer |
| 6788 | indicating the return address of the specified call frame, or zero if it |
| 6789 | cannot be identified. The value returned by this intrinsic is likely to be |
| 6790 | incorrect or 0 for arguments other than zero, so it should only be used for |
| 6791 | debugging purposes.</p> |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 6792 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6793 | <p>Note that calling this intrinsic does not prevent function inlining or other |
| 6794 | aggressive transformations, so the value returned may not be that of the |
| 6795 | obvious source-language caller.</p> |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 6796 | |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 6797 | </div> |
| 6798 | |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 6799 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6800 | <h4> |
| Reid Spencer | a3e435f | 2007-04-04 02:42:35 +0000 | [diff] [blame] | 6801 | <a name="int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6802 | </h4> |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 6803 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 6804 | <div> |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 6805 | |
| 6806 | <h5>Syntax:</h5> |
| 6807 | <pre> |
| Dan Gohman | fe47aae | 2010-05-28 17:13:49 +0000 | [diff] [blame] | 6808 | declare i8* @llvm.frameaddress(i32 <level>) |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 6809 | </pre> |
| 6810 | |
| 6811 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6812 | <p>The '<tt>llvm.frameaddress</tt>' intrinsic attempts to return the |
| 6813 | target-specific frame pointer value for the specified stack frame.</p> |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 6814 | |
| 6815 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6816 | <p>The argument to this intrinsic indicates which function to return the frame |
| 6817 | pointer for. Zero indicates the calling function, one indicates its caller, |
| 6818 | etc. The argument is <b>required</b> to be a constant integer value.</p> |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 6819 | |
| 6820 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6821 | <p>The '<tt>llvm.frameaddress</tt>' intrinsic either returns a pointer |
| 6822 | indicating the frame address of the specified call frame, or zero if it |
| 6823 | cannot be identified. The value returned by this intrinsic is likely to be |
| 6824 | incorrect or 0 for arguments other than zero, so it should only be used for |
| 6825 | debugging purposes.</p> |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 6826 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6827 | <p>Note that calling this intrinsic does not prevent function inlining or other |
| 6828 | aggressive transformations, so the value returned may not be that of the |
| 6829 | obvious source-language caller.</p> |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 6830 | |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 6831 | </div> |
| 6832 | |
| Chris Lattner | 9a9d7ac | 2005-02-28 19:24:19 +0000 | [diff] [blame] | 6833 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6834 | <h4> |
| Reid Spencer | a3e435f | 2007-04-04 02:42:35 +0000 | [diff] [blame] | 6835 | <a name="int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6836 | </h4> |
| Chris Lattner | 57e1f39 | 2006-01-13 02:03:13 +0000 | [diff] [blame] | 6837 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 6838 | <div> |
| Chris Lattner | 57e1f39 | 2006-01-13 02:03:13 +0000 | [diff] [blame] | 6839 | |
| 6840 | <h5>Syntax:</h5> |
| 6841 | <pre> |
| Dan Gohman | fe47aae | 2010-05-28 17:13:49 +0000 | [diff] [blame] | 6842 | declare i8* @llvm.stacksave() |
| Chris Lattner | 57e1f39 | 2006-01-13 02:03:13 +0000 | [diff] [blame] | 6843 | </pre> |
| 6844 | |
| 6845 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6846 | <p>The '<tt>llvm.stacksave</tt>' intrinsic is used to remember the current state |
| 6847 | of the function stack, for use |
| 6848 | with <a href="#int_stackrestore"> <tt>llvm.stackrestore</tt></a>. This is |
| 6849 | useful for implementing language features like scoped automatic variable |
| 6850 | sized arrays in C99.</p> |
| Chris Lattner | 57e1f39 | 2006-01-13 02:03:13 +0000 | [diff] [blame] | 6851 | |
| 6852 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6853 | <p>This intrinsic returns a opaque pointer value that can be passed |
| 6854 | to <a href="#int_stackrestore"><tt>llvm.stackrestore</tt></a>. When |
| 6855 | an <tt>llvm.stackrestore</tt> intrinsic is executed with a value saved |
| 6856 | from <tt>llvm.stacksave</tt>, it effectively restores the state of the stack |
| 6857 | to the state it was in when the <tt>llvm.stacksave</tt> intrinsic executed. |
| 6858 | In practice, this pops any <a href="#i_alloca">alloca</a> blocks from the |
| 6859 | stack that were allocated after the <tt>llvm.stacksave</tt> was executed.</p> |
| Chris Lattner | 57e1f39 | 2006-01-13 02:03:13 +0000 | [diff] [blame] | 6860 | |
| 6861 | </div> |
| 6862 | |
| 6863 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6864 | <h4> |
| Reid Spencer | a3e435f | 2007-04-04 02:42:35 +0000 | [diff] [blame] | 6865 | <a name="int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6866 | </h4> |
| Chris Lattner | 57e1f39 | 2006-01-13 02:03:13 +0000 | [diff] [blame] | 6867 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 6868 | <div> |
| Chris Lattner | 57e1f39 | 2006-01-13 02:03:13 +0000 | [diff] [blame] | 6869 | |
| 6870 | <h5>Syntax:</h5> |
| 6871 | <pre> |
| Dan Gohman | fe47aae | 2010-05-28 17:13:49 +0000 | [diff] [blame] | 6872 | declare void @llvm.stackrestore(i8* %ptr) |
| Chris Lattner | 57e1f39 | 2006-01-13 02:03:13 +0000 | [diff] [blame] | 6873 | </pre> |
| 6874 | |
| 6875 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6876 | <p>The '<tt>llvm.stackrestore</tt>' intrinsic is used to restore the state of |
| 6877 | the function stack to the state it was in when the |
| 6878 | corresponding <a href="#int_stacksave"><tt>llvm.stacksave</tt></a> intrinsic |
| 6879 | executed. This is useful for implementing language features like scoped |
| 6880 | automatic variable sized arrays in C99.</p> |
| Chris Lattner | 57e1f39 | 2006-01-13 02:03:13 +0000 | [diff] [blame] | 6881 | |
| 6882 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6883 | <p>See the description |
| 6884 | for <a href="#int_stacksave"><tt>llvm.stacksave</tt></a>.</p> |
| Chris Lattner | 57e1f39 | 2006-01-13 02:03:13 +0000 | [diff] [blame] | 6885 | |
| 6886 | </div> |
| 6887 | |
| Chris Lattner | 57e1f39 | 2006-01-13 02:03:13 +0000 | [diff] [blame] | 6888 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6889 | <h4> |
| Reid Spencer | a3e435f | 2007-04-04 02:42:35 +0000 | [diff] [blame] | 6890 | <a name="int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6891 | </h4> |
| Chris Lattner | 9a9d7ac | 2005-02-28 19:24:19 +0000 | [diff] [blame] | 6892 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 6893 | <div> |
| Chris Lattner | 9a9d7ac | 2005-02-28 19:24:19 +0000 | [diff] [blame] | 6894 | |
| 6895 | <h5>Syntax:</h5> |
| 6896 | <pre> |
| Bruno Cardoso Lopes | 9a76733 | 2011-06-14 04:58:37 +0000 | [diff] [blame] | 6897 | declare void @llvm.prefetch(i8* <address>, i32 <rw>, i32 <locality>, i32 <cache type>) |
| Chris Lattner | 9a9d7ac | 2005-02-28 19:24:19 +0000 | [diff] [blame] | 6898 | </pre> |
| 6899 | |
| 6900 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6901 | <p>The '<tt>llvm.prefetch</tt>' intrinsic is a hint to the code generator to |
| 6902 | insert a prefetch instruction if supported; otherwise, it is a noop. |
| 6903 | Prefetches have no effect on the behavior of the program but can change its |
| 6904 | performance characteristics.</p> |
| Chris Lattner | 9a9d7ac | 2005-02-28 19:24:19 +0000 | [diff] [blame] | 6905 | |
| 6906 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6907 | <p><tt>address</tt> is the address to be prefetched, <tt>rw</tt> is the |
| 6908 | specifier determining if the fetch should be for a read (0) or write (1), |
| 6909 | and <tt>locality</tt> is a temporal locality specifier ranging from (0) - no |
| Bruno Cardoso Lopes | 9a76733 | 2011-06-14 04:58:37 +0000 | [diff] [blame] | 6910 | locality, to (3) - extremely local keep in cache. The <tt>cache type</tt> |
| 6911 | specifies whether the prefetch is performed on the data (1) or instruction (0) |
| 6912 | cache. The <tt>rw</tt>, <tt>locality</tt> and <tt>cache type</tt> arguments |
| 6913 | must be constant integers.</p> |
| Chris Lattner | 9a9d7ac | 2005-02-28 19:24:19 +0000 | [diff] [blame] | 6914 | |
| 6915 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6916 | <p>This intrinsic does not modify the behavior of the program. In particular, |
| 6917 | prefetches cannot trap and do not produce a value. On targets that support |
| 6918 | this intrinsic, the prefetch can provide hints to the processor cache for |
| 6919 | better performance.</p> |
| Chris Lattner | 9a9d7ac | 2005-02-28 19:24:19 +0000 | [diff] [blame] | 6920 | |
| 6921 | </div> |
| 6922 | |
| Andrew Lenharth | 7f4ec3b | 2005-03-28 20:05:49 +0000 | [diff] [blame] | 6923 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6924 | <h4> |
| Reid Spencer | a3e435f | 2007-04-04 02:42:35 +0000 | [diff] [blame] | 6925 | <a name="int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6926 | </h4> |
| Andrew Lenharth | 7f4ec3b | 2005-03-28 20:05:49 +0000 | [diff] [blame] | 6927 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 6928 | <div> |
| Andrew Lenharth | 7f4ec3b | 2005-03-28 20:05:49 +0000 | [diff] [blame] | 6929 | |
| 6930 | <h5>Syntax:</h5> |
| 6931 | <pre> |
| Chris Lattner | 1df4f75 | 2007-09-21 17:30:40 +0000 | [diff] [blame] | 6932 | declare void @llvm.pcmarker(i32 <id>) |
| Andrew Lenharth | 7f4ec3b | 2005-03-28 20:05:49 +0000 | [diff] [blame] | 6933 | </pre> |
| 6934 | |
| 6935 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6936 | <p>The '<tt>llvm.pcmarker</tt>' intrinsic is a method to export a Program |
| 6937 | Counter (PC) in a region of code to simulators and other tools. The method |
| 6938 | is target specific, but it is expected that the marker will use exported |
| 6939 | symbols to transmit the PC of the marker. The marker makes no guarantees |
| 6940 | that it will remain with any specific instruction after optimizations. It is |
| 6941 | possible that the presence of a marker will inhibit optimizations. The |
| 6942 | intended use is to be inserted after optimizations to allow correlations of |
| 6943 | simulation runs.</p> |
| Andrew Lenharth | 7f4ec3b | 2005-03-28 20:05:49 +0000 | [diff] [blame] | 6944 | |
| 6945 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6946 | <p><tt>id</tt> is a numerical id identifying the marker.</p> |
| Andrew Lenharth | 7f4ec3b | 2005-03-28 20:05:49 +0000 | [diff] [blame] | 6947 | |
| 6948 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6949 | <p>This intrinsic does not modify the behavior of the program. Backends that do |
| Dan Gohman | b55a1ee | 2010-03-01 17:41:39 +0000 | [diff] [blame] | 6950 | not support this intrinsic may ignore it.</p> |
| Andrew Lenharth | 7f4ec3b | 2005-03-28 20:05:49 +0000 | [diff] [blame] | 6951 | |
| 6952 | </div> |
| 6953 | |
| Andrew Lenharth | 51b8d54 | 2005-11-11 16:47:30 +0000 | [diff] [blame] | 6954 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6955 | <h4> |
| Reid Spencer | a3e435f | 2007-04-04 02:42:35 +0000 | [diff] [blame] | 6956 | <a name="int_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6957 | </h4> |
| Andrew Lenharth | 51b8d54 | 2005-11-11 16:47:30 +0000 | [diff] [blame] | 6958 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 6959 | <div> |
| Andrew Lenharth | 51b8d54 | 2005-11-11 16:47:30 +0000 | [diff] [blame] | 6960 | |
| 6961 | <h5>Syntax:</h5> |
| 6962 | <pre> |
| Dan Gohman | 3dfb3cf | 2010-05-28 17:07:41 +0000 | [diff] [blame] | 6963 | declare i64 @llvm.readcyclecounter() |
| Andrew Lenharth | 51b8d54 | 2005-11-11 16:47:30 +0000 | [diff] [blame] | 6964 | </pre> |
| 6965 | |
| 6966 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6967 | <p>The '<tt>llvm.readcyclecounter</tt>' intrinsic provides access to the cycle |
| 6968 | counter register (or similar low latency, high accuracy clocks) on those |
| 6969 | targets that support it. On X86, it should map to RDTSC. On Alpha, it |
| 6970 | should map to RPCC. As the backing counters overflow quickly (on the order |
| 6971 | of 9 seconds on alpha), this should only be used for small timings.</p> |
| Andrew Lenharth | 51b8d54 | 2005-11-11 16:47:30 +0000 | [diff] [blame] | 6972 | |
| 6973 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6974 | <p>When directly supported, reading the cycle counter should not modify any |
| 6975 | memory. Implementations are allowed to either return a application specific |
| 6976 | value or a system wide value. On backends without support, this is lowered |
| 6977 | to a constant 0.</p> |
| Andrew Lenharth | 51b8d54 | 2005-11-11 16:47:30 +0000 | [diff] [blame] | 6978 | |
| 6979 | </div> |
| 6980 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 6981 | </div> |
| 6982 | |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 6983 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6984 | <h3> |
| Chris Lattner | 33aec9e | 2004-02-12 17:01:32 +0000 | [diff] [blame] | 6985 | <a name="int_libc">Standard C Library Intrinsics</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6986 | </h3> |
| Chris Lattner | 33aec9e | 2004-02-12 17:01:32 +0000 | [diff] [blame] | 6987 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 6988 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 6989 | |
| 6990 | <p>LLVM provides intrinsics for a few important standard C library functions. |
| 6991 | These intrinsics allow source-language front-ends to pass information about |
| 6992 | the alignment of the pointer arguments to the code generator, providing |
| 6993 | opportunity for more efficient code generation.</p> |
| Chris Lattner | 33aec9e | 2004-02-12 17:01:32 +0000 | [diff] [blame] | 6994 | |
| Chris Lattner | 33aec9e | 2004-02-12 17:01:32 +0000 | [diff] [blame] | 6995 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6996 | <h4> |
| Reid Spencer | a3e435f | 2007-04-04 02:42:35 +0000 | [diff] [blame] | 6997 | <a name="int_memcpy">'<tt>llvm.memcpy</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 6998 | </h4> |
| Chris Lattner | 33aec9e | 2004-02-12 17:01:32 +0000 | [diff] [blame] | 6999 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 7000 | <div> |
| Chris Lattner | 33aec9e | 2004-02-12 17:01:32 +0000 | [diff] [blame] | 7001 | |
| 7002 | <h5>Syntax:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7003 | <p>This is an overloaded intrinsic. You can use <tt>llvm.memcpy</tt> on any |
| Mon P Wang | e88909b | 2010-04-07 06:35:53 +0000 | [diff] [blame] | 7004 | integer bit width and for different address spaces. Not all targets support |
| 7005 | all bit widths however.</p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7006 | |
| Chris Lattner | 33aec9e | 2004-02-12 17:01:32 +0000 | [diff] [blame] | 7007 | <pre> |
| Dan Gohman | fe47aae | 2010-05-28 17:13:49 +0000 | [diff] [blame] | 7008 | declare void @llvm.memcpy.p0i8.p0i8.i32(i8* <dest>, i8* <src>, |
| Chris Lattner | 9f636de | 2010-04-08 00:53:57 +0000 | [diff] [blame] | 7009 | i32 <len>, i32 <align>, i1 <isvolatile>) |
| Dan Gohman | fe47aae | 2010-05-28 17:13:49 +0000 | [diff] [blame] | 7010 | declare void @llvm.memcpy.p0i8.p0i8.i64(i8* <dest>, i8* <src>, |
| Chris Lattner | 9f636de | 2010-04-08 00:53:57 +0000 | [diff] [blame] | 7011 | i64 <len>, i32 <align>, i1 <isvolatile>) |
| Chris Lattner | 33aec9e | 2004-02-12 17:01:32 +0000 | [diff] [blame] | 7012 | </pre> |
| 7013 | |
| 7014 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7015 | <p>The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the |
| 7016 | source location to the destination location.</p> |
| Chris Lattner | 33aec9e | 2004-02-12 17:01:32 +0000 | [diff] [blame] | 7017 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7018 | <p>Note that, unlike the standard libc function, the <tt>llvm.memcpy.*</tt> |
| Chris Lattner | 9f636de | 2010-04-08 00:53:57 +0000 | [diff] [blame] | 7019 | intrinsics do not return a value, takes extra alignment/isvolatile arguments |
| 7020 | and the pointers can be in specified address spaces.</p> |
| Chris Lattner | 33aec9e | 2004-02-12 17:01:32 +0000 | [diff] [blame] | 7021 | |
| 7022 | <h5>Arguments:</h5> |
| Chris Lattner | 9f636de | 2010-04-08 00:53:57 +0000 | [diff] [blame] | 7023 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7024 | <p>The first argument is a pointer to the destination, the second is a pointer |
| 7025 | to the source. The third argument is an integer argument specifying the |
| Chris Lattner | 9f636de | 2010-04-08 00:53:57 +0000 | [diff] [blame] | 7026 | number of bytes to copy, the fourth argument is the alignment of the |
| 7027 | source and destination locations, and the fifth is a boolean indicating a |
| 7028 | volatile access.</p> |
| Chris Lattner | 33aec9e | 2004-02-12 17:01:32 +0000 | [diff] [blame] | 7029 | |
| Dan Gohman | b55a1ee | 2010-03-01 17:41:39 +0000 | [diff] [blame] | 7030 | <p>If the call to this intrinsic has an alignment value that is not 0 or 1, |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7031 | then the caller guarantees that both the source and destination pointers are |
| 7032 | aligned to that boundary.</p> |
| Chris Lattner | 3301ced | 2004-02-12 21:18:15 +0000 | [diff] [blame] | 7033 | |
| Jeffrey Yasskin | 93e066d | 2010-04-26 21:21:24 +0000 | [diff] [blame] | 7034 | <p>If the <tt>isvolatile</tt> parameter is <tt>true</tt>, the |
| 7035 | <tt>llvm.memcpy</tt> call is a <a href="#volatile">volatile operation</a>. |
| 7036 | The detailed access behavior is not very cleanly specified and it is unwise |
| 7037 | to depend on it.</p> |
| Chris Lattner | 9f636de | 2010-04-08 00:53:57 +0000 | [diff] [blame] | 7038 | |
| Chris Lattner | 33aec9e | 2004-02-12 17:01:32 +0000 | [diff] [blame] | 7039 | <h5>Semantics:</h5> |
| Chris Lattner | 9f636de | 2010-04-08 00:53:57 +0000 | [diff] [blame] | 7040 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7041 | <p>The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the |
| 7042 | source location to the destination location, which are not allowed to |
| 7043 | overlap. It copies "len" bytes of memory over. If the argument is known to |
| 7044 | be aligned to some boundary, this can be specified as the fourth argument, |
| 7045 | otherwise it should be set to 0 or 1.</p> |
| Chris Lattner | 33aec9e | 2004-02-12 17:01:32 +0000 | [diff] [blame] | 7046 | |
| Chris Lattner | 33aec9e | 2004-02-12 17:01:32 +0000 | [diff] [blame] | 7047 | </div> |
| 7048 | |
| Chris Lattner | 0eb51b4 | 2004-02-12 18:10:10 +0000 | [diff] [blame] | 7049 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7050 | <h4> |
| Reid Spencer | a3e435f | 2007-04-04 02:42:35 +0000 | [diff] [blame] | 7051 | <a name="int_memmove">'<tt>llvm.memmove</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7052 | </h4> |
| Chris Lattner | 0eb51b4 | 2004-02-12 18:10:10 +0000 | [diff] [blame] | 7053 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 7054 | <div> |
| Chris Lattner | 0eb51b4 | 2004-02-12 18:10:10 +0000 | [diff] [blame] | 7055 | |
| 7056 | <h5>Syntax:</h5> |
| Chris Lattner | 824b958 | 2008-11-21 16:42:48 +0000 | [diff] [blame] | 7057 | <p>This is an overloaded intrinsic. You can use llvm.memmove on any integer bit |
| Mon P Wang | e88909b | 2010-04-07 06:35:53 +0000 | [diff] [blame] | 7058 | width and for different address space. Not all targets support all bit |
| 7059 | widths however.</p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7060 | |
| Chris Lattner | 0eb51b4 | 2004-02-12 18:10:10 +0000 | [diff] [blame] | 7061 | <pre> |
| Dan Gohman | fe47aae | 2010-05-28 17:13:49 +0000 | [diff] [blame] | 7062 | declare void @llvm.memmove.p0i8.p0i8.i32(i8* <dest>, i8* <src>, |
| Chris Lattner | 9f636de | 2010-04-08 00:53:57 +0000 | [diff] [blame] | 7063 | i32 <len>, i32 <align>, i1 <isvolatile>) |
| Dan Gohman | fe47aae | 2010-05-28 17:13:49 +0000 | [diff] [blame] | 7064 | declare void @llvm.memmove.p0i8.p0i8.i64(i8* <dest>, i8* <src>, |
| Chris Lattner | 9f636de | 2010-04-08 00:53:57 +0000 | [diff] [blame] | 7065 | i64 <len>, i32 <align>, i1 <isvolatile>) |
| Chris Lattner | 0eb51b4 | 2004-02-12 18:10:10 +0000 | [diff] [blame] | 7066 | </pre> |
| 7067 | |
| 7068 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7069 | <p>The '<tt>llvm.memmove.*</tt>' intrinsics move a block of memory from the |
| 7070 | source location to the destination location. It is similar to the |
| 7071 | '<tt>llvm.memcpy</tt>' intrinsic but allows the two memory locations to |
| 7072 | overlap.</p> |
| Chris Lattner | 0eb51b4 | 2004-02-12 18:10:10 +0000 | [diff] [blame] | 7073 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7074 | <p>Note that, unlike the standard libc function, the <tt>llvm.memmove.*</tt> |
| Chris Lattner | 9f636de | 2010-04-08 00:53:57 +0000 | [diff] [blame] | 7075 | intrinsics do not return a value, takes extra alignment/isvolatile arguments |
| 7076 | and the pointers can be in specified address spaces.</p> |
| Chris Lattner | 0eb51b4 | 2004-02-12 18:10:10 +0000 | [diff] [blame] | 7077 | |
| 7078 | <h5>Arguments:</h5> |
| Chris Lattner | 9f636de | 2010-04-08 00:53:57 +0000 | [diff] [blame] | 7079 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7080 | <p>The first argument is a pointer to the destination, the second is a pointer |
| 7081 | to the source. The third argument is an integer argument specifying the |
| Chris Lattner | 9f636de | 2010-04-08 00:53:57 +0000 | [diff] [blame] | 7082 | number of bytes to copy, the fourth argument is the alignment of the |
| 7083 | source and destination locations, and the fifth is a boolean indicating a |
| 7084 | volatile access.</p> |
| Chris Lattner | 0eb51b4 | 2004-02-12 18:10:10 +0000 | [diff] [blame] | 7085 | |
| Dan Gohman | b55a1ee | 2010-03-01 17:41:39 +0000 | [diff] [blame] | 7086 | <p>If the call to this intrinsic has an alignment value that is not 0 or 1, |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7087 | then the caller guarantees that the source and destination pointers are |
| 7088 | aligned to that boundary.</p> |
| Chris Lattner | 3301ced | 2004-02-12 21:18:15 +0000 | [diff] [blame] | 7089 | |
| Jeffrey Yasskin | 93e066d | 2010-04-26 21:21:24 +0000 | [diff] [blame] | 7090 | <p>If the <tt>isvolatile</tt> parameter is <tt>true</tt>, the |
| 7091 | <tt>llvm.memmove</tt> call is a <a href="#volatile">volatile operation</a>. |
| 7092 | The detailed access behavior is not very cleanly specified and it is unwise |
| 7093 | to depend on it.</p> |
| Chris Lattner | 9f636de | 2010-04-08 00:53:57 +0000 | [diff] [blame] | 7094 | |
| Chris Lattner | 0eb51b4 | 2004-02-12 18:10:10 +0000 | [diff] [blame] | 7095 | <h5>Semantics:</h5> |
| Chris Lattner | 9f636de | 2010-04-08 00:53:57 +0000 | [diff] [blame] | 7096 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7097 | <p>The '<tt>llvm.memmove.*</tt>' intrinsics copy a block of memory from the |
| 7098 | source location to the destination location, which may overlap. It copies |
| 7099 | "len" bytes of memory over. If the argument is known to be aligned to some |
| 7100 | boundary, this can be specified as the fourth argument, otherwise it should |
| 7101 | be set to 0 or 1.</p> |
| Chris Lattner | 0eb51b4 | 2004-02-12 18:10:10 +0000 | [diff] [blame] | 7102 | |
| Chris Lattner | 0eb51b4 | 2004-02-12 18:10:10 +0000 | [diff] [blame] | 7103 | </div> |
| 7104 | |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 7105 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7106 | <h4> |
| Reid Spencer | a3e435f | 2007-04-04 02:42:35 +0000 | [diff] [blame] | 7107 | <a name="int_memset">'<tt>llvm.memset.*</tt>' Intrinsics</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7108 | </h4> |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 7109 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 7110 | <div> |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 7111 | |
| 7112 | <h5>Syntax:</h5> |
| Chris Lattner | 824b958 | 2008-11-21 16:42:48 +0000 | [diff] [blame] | 7113 | <p>This is an overloaded intrinsic. You can use llvm.memset on any integer bit |
| John Criswell | cdcbbfc | 2010-07-30 16:30:28 +0000 | [diff] [blame] | 7114 | width and for different address spaces. However, not all targets support all |
| 7115 | bit widths.</p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7116 | |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 7117 | <pre> |
| Dan Gohman | fe47aae | 2010-05-28 17:13:49 +0000 | [diff] [blame] | 7118 | declare void @llvm.memset.p0i8.i32(i8* <dest>, i8 <val>, |
| Chris Lattner | ff35c3f | 2010-04-08 00:54:34 +0000 | [diff] [blame] | 7119 | i32 <len>, i32 <align>, i1 <isvolatile>) |
| Dan Gohman | fe47aae | 2010-05-28 17:13:49 +0000 | [diff] [blame] | 7120 | declare void @llvm.memset.p0i8.i64(i8* <dest>, i8 <val>, |
| Chris Lattner | ff35c3f | 2010-04-08 00:54:34 +0000 | [diff] [blame] | 7121 | i64 <len>, i32 <align>, i1 <isvolatile>) |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 7122 | </pre> |
| 7123 | |
| 7124 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7125 | <p>The '<tt>llvm.memset.*</tt>' intrinsics fill a block of memory with a |
| 7126 | particular byte value.</p> |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 7127 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7128 | <p>Note that, unlike the standard libc function, the <tt>llvm.memset</tt> |
| John Criswell | cdcbbfc | 2010-07-30 16:30:28 +0000 | [diff] [blame] | 7129 | intrinsic does not return a value and takes extra alignment/volatile |
| 7130 | arguments. Also, the destination can be in an arbitrary address space.</p> |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 7131 | |
| 7132 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7133 | <p>The first argument is a pointer to the destination to fill, the second is the |
| John Criswell | cdcbbfc | 2010-07-30 16:30:28 +0000 | [diff] [blame] | 7134 | byte value with which to fill it, the third argument is an integer argument |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7135 | specifying the number of bytes to fill, and the fourth argument is the known |
| John Criswell | cdcbbfc | 2010-07-30 16:30:28 +0000 | [diff] [blame] | 7136 | alignment of the destination location.</p> |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 7137 | |
| Dan Gohman | b55a1ee | 2010-03-01 17:41:39 +0000 | [diff] [blame] | 7138 | <p>If the call to this intrinsic has an alignment value that is not 0 or 1, |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7139 | then the caller guarantees that the destination pointer is aligned to that |
| 7140 | boundary.</p> |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 7141 | |
| Jeffrey Yasskin | 93e066d | 2010-04-26 21:21:24 +0000 | [diff] [blame] | 7142 | <p>If the <tt>isvolatile</tt> parameter is <tt>true</tt>, the |
| 7143 | <tt>llvm.memset</tt> call is a <a href="#volatile">volatile operation</a>. |
| 7144 | The detailed access behavior is not very cleanly specified and it is unwise |
| 7145 | to depend on it.</p> |
| Chris Lattner | 9f636de | 2010-04-08 00:53:57 +0000 | [diff] [blame] | 7146 | |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 7147 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7148 | <p>The '<tt>llvm.memset.*</tt>' intrinsics fill "len" bytes of memory starting |
| 7149 | at the destination location. If the argument is known to be aligned to some |
| 7150 | boundary, this can be specified as the fourth argument, otherwise it should |
| 7151 | be set to 0 or 1.</p> |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 7152 | |
| Chris Lattner | 1061064 | 2004-02-14 04:08:35 +0000 | [diff] [blame] | 7153 | </div> |
| 7154 | |
| Chris Lattner | 3200628 | 2004-06-11 02:28:03 +0000 | [diff] [blame] | 7155 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7156 | <h4> |
| Reid Spencer | a3e435f | 2007-04-04 02:42:35 +0000 | [diff] [blame] | 7157 | <a name="int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7158 | </h4> |
| Chris Lattner | a4d7414 | 2005-07-21 01:29:16 +0000 | [diff] [blame] | 7159 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 7160 | <div> |
| Chris Lattner | a4d7414 | 2005-07-21 01:29:16 +0000 | [diff] [blame] | 7161 | |
| 7162 | <h5>Syntax:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7163 | <p>This is an overloaded intrinsic. You can use <tt>llvm.sqrt</tt> on any |
| 7164 | floating point or vector of floating point type. Not all targets support all |
| 7165 | types however.</p> |
| 7166 | |
| Chris Lattner | a4d7414 | 2005-07-21 01:29:16 +0000 | [diff] [blame] | 7167 | <pre> |
| Dale Johannesen | 408f9c1 | 2007-10-02 17:47:38 +0000 | [diff] [blame] | 7168 | declare float @llvm.sqrt.f32(float %Val) |
| 7169 | declare double @llvm.sqrt.f64(double %Val) |
| 7170 | declare x86_fp80 @llvm.sqrt.f80(x86_fp80 %Val) |
| 7171 | declare fp128 @llvm.sqrt.f128(fp128 %Val) |
| 7172 | declare ppc_fp128 @llvm.sqrt.ppcf128(ppc_fp128 %Val) |
| Chris Lattner | a4d7414 | 2005-07-21 01:29:16 +0000 | [diff] [blame] | 7173 | </pre> |
| 7174 | |
| 7175 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7176 | <p>The '<tt>llvm.sqrt</tt>' intrinsics return the sqrt of the specified operand, |
| 7177 | returning the same value as the libm '<tt>sqrt</tt>' functions would. |
| 7178 | Unlike <tt>sqrt</tt> in libm, however, <tt>llvm.sqrt</tt> has undefined |
| 7179 | behavior for negative numbers other than -0.0 (which allows for better |
| 7180 | optimization, because there is no need to worry about errno being |
| 7181 | set). <tt>llvm.sqrt(-0.0)</tt> is defined to return -0.0 like IEEE sqrt.</p> |
| Chris Lattner | a4d7414 | 2005-07-21 01:29:16 +0000 | [diff] [blame] | 7182 | |
| 7183 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7184 | <p>The argument and return value are floating point numbers of the same |
| 7185 | type.</p> |
| Chris Lattner | a4d7414 | 2005-07-21 01:29:16 +0000 | [diff] [blame] | 7186 | |
| 7187 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7188 | <p>This function returns the sqrt of the specified operand if it is a |
| 7189 | nonnegative floating point number.</p> |
| Chris Lattner | a4d7414 | 2005-07-21 01:29:16 +0000 | [diff] [blame] | 7190 | |
| Chris Lattner | a4d7414 | 2005-07-21 01:29:16 +0000 | [diff] [blame] | 7191 | </div> |
| 7192 | |
| Chris Lattner | f4d252d | 2006-09-08 06:34:02 +0000 | [diff] [blame] | 7193 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7194 | <h4> |
| Reid Spencer | a3e435f | 2007-04-04 02:42:35 +0000 | [diff] [blame] | 7195 | <a name="int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7196 | </h4> |
| Chris Lattner | f4d252d | 2006-09-08 06:34:02 +0000 | [diff] [blame] | 7197 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 7198 | <div> |
| Chris Lattner | f4d252d | 2006-09-08 06:34:02 +0000 | [diff] [blame] | 7199 | |
| 7200 | <h5>Syntax:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7201 | <p>This is an overloaded intrinsic. You can use <tt>llvm.powi</tt> on any |
| 7202 | floating point or vector of floating point type. Not all targets support all |
| 7203 | types however.</p> |
| 7204 | |
| Chris Lattner | f4d252d | 2006-09-08 06:34:02 +0000 | [diff] [blame] | 7205 | <pre> |
| Dale Johannesen | 408f9c1 | 2007-10-02 17:47:38 +0000 | [diff] [blame] | 7206 | declare float @llvm.powi.f32(float %Val, i32 %power) |
| 7207 | declare double @llvm.powi.f64(double %Val, i32 %power) |
| 7208 | declare x86_fp80 @llvm.powi.f80(x86_fp80 %Val, i32 %power) |
| 7209 | declare fp128 @llvm.powi.f128(fp128 %Val, i32 %power) |
| 7210 | declare ppc_fp128 @llvm.powi.ppcf128(ppc_fp128 %Val, i32 %power) |
| Chris Lattner | f4d252d | 2006-09-08 06:34:02 +0000 | [diff] [blame] | 7211 | </pre> |
| 7212 | |
| 7213 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7214 | <p>The '<tt>llvm.powi.*</tt>' intrinsics return the first operand raised to the |
| 7215 | specified (positive or negative) power. The order of evaluation of |
| 7216 | multiplications is not defined. When a vector of floating point type is |
| 7217 | used, the second argument remains a scalar integer value.</p> |
| Chris Lattner | f4d252d | 2006-09-08 06:34:02 +0000 | [diff] [blame] | 7218 | |
| 7219 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7220 | <p>The second argument is an integer power, and the first is a value to raise to |
| 7221 | that power.</p> |
| Chris Lattner | f4d252d | 2006-09-08 06:34:02 +0000 | [diff] [blame] | 7222 | |
| 7223 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7224 | <p>This function returns the first value raised to the second power with an |
| 7225 | unspecified sequence of rounding operations.</p> |
| Chris Lattner | f4d252d | 2006-09-08 06:34:02 +0000 | [diff] [blame] | 7226 | |
| Chris Lattner | f4d252d | 2006-09-08 06:34:02 +0000 | [diff] [blame] | 7227 | </div> |
| 7228 | |
| Dan Gohman | 91c284c | 2007-10-15 20:30:11 +0000 | [diff] [blame] | 7229 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7230 | <h4> |
| Dan Gohman | 91c284c | 2007-10-15 20:30:11 +0000 | [diff] [blame] | 7231 | <a name="int_sin">'<tt>llvm.sin.*</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7232 | </h4> |
| Dan Gohman | 91c284c | 2007-10-15 20:30:11 +0000 | [diff] [blame] | 7233 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 7234 | <div> |
| Dan Gohman | 91c284c | 2007-10-15 20:30:11 +0000 | [diff] [blame] | 7235 | |
| 7236 | <h5>Syntax:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7237 | <p>This is an overloaded intrinsic. You can use <tt>llvm.sin</tt> on any |
| 7238 | floating point or vector of floating point type. Not all targets support all |
| 7239 | types however.</p> |
| 7240 | |
| Dan Gohman | 91c284c | 2007-10-15 20:30:11 +0000 | [diff] [blame] | 7241 | <pre> |
| 7242 | declare float @llvm.sin.f32(float %Val) |
| 7243 | declare double @llvm.sin.f64(double %Val) |
| 7244 | declare x86_fp80 @llvm.sin.f80(x86_fp80 %Val) |
| 7245 | declare fp128 @llvm.sin.f128(fp128 %Val) |
| 7246 | declare ppc_fp128 @llvm.sin.ppcf128(ppc_fp128 %Val) |
| 7247 | </pre> |
| 7248 | |
| 7249 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7250 | <p>The '<tt>llvm.sin.*</tt>' intrinsics return the sine of the operand.</p> |
| Dan Gohman | 91c284c | 2007-10-15 20:30:11 +0000 | [diff] [blame] | 7251 | |
| 7252 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7253 | <p>The argument and return value are floating point numbers of the same |
| 7254 | type.</p> |
| Dan Gohman | 91c284c | 2007-10-15 20:30:11 +0000 | [diff] [blame] | 7255 | |
| 7256 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7257 | <p>This function returns the sine of the specified operand, returning the same |
| 7258 | values as the libm <tt>sin</tt> functions would, and handles error conditions |
| 7259 | in the same way.</p> |
| Dan Gohman | 91c284c | 2007-10-15 20:30:11 +0000 | [diff] [blame] | 7260 | |
| Dan Gohman | 91c284c | 2007-10-15 20:30:11 +0000 | [diff] [blame] | 7261 | </div> |
| 7262 | |
| 7263 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7264 | <h4> |
| Dan Gohman | 91c284c | 2007-10-15 20:30:11 +0000 | [diff] [blame] | 7265 | <a name="int_cos">'<tt>llvm.cos.*</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7266 | </h4> |
| Dan Gohman | 91c284c | 2007-10-15 20:30:11 +0000 | [diff] [blame] | 7267 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 7268 | <div> |
| Dan Gohman | 91c284c | 2007-10-15 20:30:11 +0000 | [diff] [blame] | 7269 | |
| 7270 | <h5>Syntax:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7271 | <p>This is an overloaded intrinsic. You can use <tt>llvm.cos</tt> on any |
| 7272 | floating point or vector of floating point type. Not all targets support all |
| 7273 | types however.</p> |
| 7274 | |
| Dan Gohman | 91c284c | 2007-10-15 20:30:11 +0000 | [diff] [blame] | 7275 | <pre> |
| 7276 | declare float @llvm.cos.f32(float %Val) |
| 7277 | declare double @llvm.cos.f64(double %Val) |
| 7278 | declare x86_fp80 @llvm.cos.f80(x86_fp80 %Val) |
| 7279 | declare fp128 @llvm.cos.f128(fp128 %Val) |
| 7280 | declare ppc_fp128 @llvm.cos.ppcf128(ppc_fp128 %Val) |
| 7281 | </pre> |
| 7282 | |
| 7283 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7284 | <p>The '<tt>llvm.cos.*</tt>' intrinsics return the cosine of the operand.</p> |
| Dan Gohman | 91c284c | 2007-10-15 20:30:11 +0000 | [diff] [blame] | 7285 | |
| 7286 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7287 | <p>The argument and return value are floating point numbers of the same |
| 7288 | type.</p> |
| Dan Gohman | 91c284c | 2007-10-15 20:30:11 +0000 | [diff] [blame] | 7289 | |
| 7290 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7291 | <p>This function returns the cosine of the specified operand, returning the same |
| 7292 | values as the libm <tt>cos</tt> functions would, and handles error conditions |
| 7293 | in the same way.</p> |
| Dan Gohman | 91c284c | 2007-10-15 20:30:11 +0000 | [diff] [blame] | 7294 | |
| Dan Gohman | 91c284c | 2007-10-15 20:30:11 +0000 | [diff] [blame] | 7295 | </div> |
| 7296 | |
| 7297 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7298 | <h4> |
| Dan Gohman | 91c284c | 2007-10-15 20:30:11 +0000 | [diff] [blame] | 7299 | <a name="int_pow">'<tt>llvm.pow.*</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7300 | </h4> |
| Dan Gohman | 91c284c | 2007-10-15 20:30:11 +0000 | [diff] [blame] | 7301 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 7302 | <div> |
| Dan Gohman | 91c284c | 2007-10-15 20:30:11 +0000 | [diff] [blame] | 7303 | |
| 7304 | <h5>Syntax:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7305 | <p>This is an overloaded intrinsic. You can use <tt>llvm.pow</tt> on any |
| 7306 | floating point or vector of floating point type. Not all targets support all |
| 7307 | types however.</p> |
| 7308 | |
| Dan Gohman | 91c284c | 2007-10-15 20:30:11 +0000 | [diff] [blame] | 7309 | <pre> |
| 7310 | declare float @llvm.pow.f32(float %Val, float %Power) |
| 7311 | declare double @llvm.pow.f64(double %Val, double %Power) |
| 7312 | declare x86_fp80 @llvm.pow.f80(x86_fp80 %Val, x86_fp80 %Power) |
| 7313 | declare fp128 @llvm.pow.f128(fp128 %Val, fp128 %Power) |
| 7314 | declare ppc_fp128 @llvm.pow.ppcf128(ppc_fp128 %Val, ppc_fp128 Power) |
| 7315 | </pre> |
| 7316 | |
| 7317 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7318 | <p>The '<tt>llvm.pow.*</tt>' intrinsics return the first operand raised to the |
| 7319 | specified (positive or negative) power.</p> |
| Dan Gohman | 91c284c | 2007-10-15 20:30:11 +0000 | [diff] [blame] | 7320 | |
| 7321 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7322 | <p>The second argument is a floating point power, and the first is a value to |
| 7323 | raise to that power.</p> |
| Dan Gohman | 91c284c | 2007-10-15 20:30:11 +0000 | [diff] [blame] | 7324 | |
| 7325 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7326 | <p>This function returns the first value raised to the second power, returning |
| 7327 | the same values as the libm <tt>pow</tt> functions would, and handles error |
| 7328 | conditions in the same way.</p> |
| Dan Gohman | 91c284c | 2007-10-15 20:30:11 +0000 | [diff] [blame] | 7329 | |
| Dan Gohman | 91c284c | 2007-10-15 20:30:11 +0000 | [diff] [blame] | 7330 | </div> |
| 7331 | |
| Dan Gohman | 4e9011c | 2011-05-23 21:13:03 +0000 | [diff] [blame] | 7332 | <!-- _______________________________________________________________________ --> |
| 7333 | <h4> |
| 7334 | <a name="int_exp">'<tt>llvm.exp.*</tt>' Intrinsic</a> |
| 7335 | </h4> |
| 7336 | |
| 7337 | <div> |
| 7338 | |
| 7339 | <h5>Syntax:</h5> |
| 7340 | <p>This is an overloaded intrinsic. You can use <tt>llvm.exp</tt> on any |
| 7341 | floating point or vector of floating point type. Not all targets support all |
| 7342 | types however.</p> |
| 7343 | |
| 7344 | <pre> |
| 7345 | declare float @llvm.exp.f32(float %Val) |
| 7346 | declare double @llvm.exp.f64(double %Val) |
| 7347 | declare x86_fp80 @llvm.exp.f80(x86_fp80 %Val) |
| 7348 | declare fp128 @llvm.exp.f128(fp128 %Val) |
| 7349 | declare ppc_fp128 @llvm.exp.ppcf128(ppc_fp128 %Val) |
| 7350 | </pre> |
| 7351 | |
| 7352 | <h5>Overview:</h5> |
| 7353 | <p>The '<tt>llvm.exp.*</tt>' intrinsics perform the exp function.</p> |
| 7354 | |
| 7355 | <h5>Arguments:</h5> |
| 7356 | <p>The argument and return value are floating point numbers of the same |
| 7357 | type.</p> |
| 7358 | |
| 7359 | <h5>Semantics:</h5> |
| 7360 | <p>This function returns the same values as the libm <tt>exp</tt> functions |
| 7361 | would, and handles error conditions in the same way.</p> |
| 7362 | |
| 7363 | </div> |
| 7364 | |
| 7365 | <!-- _______________________________________________________________________ --> |
| 7366 | <h4> |
| 7367 | <a name="int_log">'<tt>llvm.log.*</tt>' Intrinsic</a> |
| 7368 | </h4> |
| 7369 | |
| 7370 | <div> |
| 7371 | |
| 7372 | <h5>Syntax:</h5> |
| 7373 | <p>This is an overloaded intrinsic. You can use <tt>llvm.log</tt> on any |
| 7374 | floating point or vector of floating point type. Not all targets support all |
| 7375 | types however.</p> |
| 7376 | |
| 7377 | <pre> |
| 7378 | declare float @llvm.log.f32(float %Val) |
| 7379 | declare double @llvm.log.f64(double %Val) |
| 7380 | declare x86_fp80 @llvm.log.f80(x86_fp80 %Val) |
| 7381 | declare fp128 @llvm.log.f128(fp128 %Val) |
| 7382 | declare ppc_fp128 @llvm.log.ppcf128(ppc_fp128 %Val) |
| 7383 | </pre> |
| 7384 | |
| 7385 | <h5>Overview:</h5> |
| 7386 | <p>The '<tt>llvm.log.*</tt>' intrinsics perform the log function.</p> |
| 7387 | |
| 7388 | <h5>Arguments:</h5> |
| 7389 | <p>The argument and return value are floating point numbers of the same |
| 7390 | type.</p> |
| 7391 | |
| 7392 | <h5>Semantics:</h5> |
| 7393 | <p>This function returns the same values as the libm <tt>log</tt> functions |
| 7394 | would, and handles error conditions in the same way.</p> |
| 7395 | |
| Nick Lewycky | 1c929be | 2011-10-31 01:32:21 +0000 | [diff] [blame] | 7396 | </div> |
| 7397 | |
| 7398 | <!-- _______________________________________________________________________ --> |
| Cameron Zwarich | 3339084 | 2011-07-08 21:39:21 +0000 | [diff] [blame] | 7399 | <h4> |
| 7400 | <a name="int_fma">'<tt>llvm.fma.*</tt>' Intrinsic</a> |
| 7401 | </h4> |
| 7402 | |
| 7403 | <div> |
| 7404 | |
| 7405 | <h5>Syntax:</h5> |
| 7406 | <p>This is an overloaded intrinsic. You can use <tt>llvm.fma</tt> on any |
| 7407 | floating point or vector of floating point type. Not all targets support all |
| 7408 | types however.</p> |
| 7409 | |
| 7410 | <pre> |
| 7411 | declare float @llvm.fma.f32(float %a, float %b, float %c) |
| 7412 | declare double @llvm.fma.f64(double %a, double %b, double %c) |
| 7413 | declare x86_fp80 @llvm.fma.f80(x86_fp80 %a, x86_fp80 %b, x86_fp80 %c) |
| 7414 | declare fp128 @llvm.fma.f128(fp128 %a, fp128 %b, fp128 %c) |
| 7415 | declare ppc_fp128 @llvm.fma.ppcf128(ppc_fp128 %a, ppc_fp128 %b, ppc_fp128 %c) |
| 7416 | </pre> |
| 7417 | |
| 7418 | <h5>Overview:</h5> |
| Cameron Zwarich | abc43e6 | 2011-07-08 22:13:55 +0000 | [diff] [blame] | 7419 | <p>The '<tt>llvm.fma.*</tt>' intrinsics perform the fused multiply-add |
| Cameron Zwarich | 3339084 | 2011-07-08 21:39:21 +0000 | [diff] [blame] | 7420 | operation.</p> |
| 7421 | |
| 7422 | <h5>Arguments:</h5> |
| 7423 | <p>The argument and return value are floating point numbers of the same |
| 7424 | type.</p> |
| 7425 | |
| 7426 | <h5>Semantics:</h5> |
| 7427 | <p>This function returns the same values as the libm <tt>fma</tt> functions |
| 7428 | would.</p> |
| 7429 | |
| Dan Gohman | 4e9011c | 2011-05-23 21:13:03 +0000 | [diff] [blame] | 7430 | </div> |
| 7431 | |
| NAKAMURA Takumi | 4b2e07a | 2011-10-31 13:04:26 +0000 | [diff] [blame] | 7432 | </div> |
| 7433 | |
| Andrew Lenharth | ec370fd | 2005-05-03 18:01:48 +0000 | [diff] [blame] | 7434 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7435 | <h3> |
| Nate Begeman | 7e36c47 | 2006-01-13 23:26:38 +0000 | [diff] [blame] | 7436 | <a name="int_manip">Bit Manipulation Intrinsics</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7437 | </h3> |
| Andrew Lenharth | ec370fd | 2005-05-03 18:01:48 +0000 | [diff] [blame] | 7438 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 7439 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7440 | |
| 7441 | <p>LLVM provides intrinsics for a few important bit manipulation operations. |
| 7442 | These allow efficient code generation for some algorithms.</p> |
| Andrew Lenharth | ec370fd | 2005-05-03 18:01:48 +0000 | [diff] [blame] | 7443 | |
| Andrew Lenharth | ec370fd | 2005-05-03 18:01:48 +0000 | [diff] [blame] | 7444 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7445 | <h4> |
| Reid Spencer | a3e435f | 2007-04-04 02:42:35 +0000 | [diff] [blame] | 7446 | <a name="int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7447 | </h4> |
| Nate Begeman | 7e36c47 | 2006-01-13 23:26:38 +0000 | [diff] [blame] | 7448 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 7449 | <div> |
| Nate Begeman | 7e36c47 | 2006-01-13 23:26:38 +0000 | [diff] [blame] | 7450 | |
| 7451 | <h5>Syntax:</h5> |
| Reid Spencer | 409e28f | 2007-04-01 08:04:23 +0000 | [diff] [blame] | 7452 | <p>This is an overloaded intrinsic function. You can use bswap on any integer |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7453 | type that is an even number of bytes (i.e. BitWidth % 16 == 0).</p> |
| 7454 | |
| Nate Begeman | 7e36c47 | 2006-01-13 23:26:38 +0000 | [diff] [blame] | 7455 | <pre> |
| Chandler Carruth | 6994040 | 2007-08-04 01:51:18 +0000 | [diff] [blame] | 7456 | declare i16 @llvm.bswap.i16(i16 <id>) |
| 7457 | declare i32 @llvm.bswap.i32(i32 <id>) |
| 7458 | declare i64 @llvm.bswap.i64(i64 <id>) |
| Nate Begeman | 7e36c47 | 2006-01-13 23:26:38 +0000 | [diff] [blame] | 7459 | </pre> |
| 7460 | |
| 7461 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7462 | <p>The '<tt>llvm.bswap</tt>' family of intrinsics is used to byte swap integer |
| 7463 | values with an even number of bytes (positive multiple of 16 bits). These |
| 7464 | are useful for performing operations on data that is not in the target's |
| 7465 | native byte order.</p> |
| Nate Begeman | 7e36c47 | 2006-01-13 23:26:38 +0000 | [diff] [blame] | 7466 | |
| 7467 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7468 | <p>The <tt>llvm.bswap.i16</tt> intrinsic returns an i16 value that has the high |
| 7469 | and low byte of the input i16 swapped. Similarly, |
| 7470 | the <tt>llvm.bswap.i32</tt> intrinsic returns an i32 value that has the four |
| 7471 | bytes of the input i32 swapped, so that if the input bytes are numbered 0, 1, |
| 7472 | 2, 3 then the returned i32 will have its bytes in 3, 2, 1, 0 order. |
| 7473 | The <tt>llvm.bswap.i48</tt>, <tt>llvm.bswap.i64</tt> and other intrinsics |
| 7474 | extend this concept to additional even-byte lengths (6 bytes, 8 bytes and |
| 7475 | more, respectively).</p> |
| Nate Begeman | 7e36c47 | 2006-01-13 23:26:38 +0000 | [diff] [blame] | 7476 | |
| 7477 | </div> |
| 7478 | |
| 7479 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7480 | <h4> |
| Reid Spencer | 0b11820 | 2006-01-16 21:12:35 +0000 | [diff] [blame] | 7481 | <a name="int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7482 | </h4> |
| Andrew Lenharth | ec370fd | 2005-05-03 18:01:48 +0000 | [diff] [blame] | 7483 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 7484 | <div> |
| Andrew Lenharth | ec370fd | 2005-05-03 18:01:48 +0000 | [diff] [blame] | 7485 | |
| 7486 | <h5>Syntax:</h5> |
| Reid Spencer | 409e28f | 2007-04-01 08:04:23 +0000 | [diff] [blame] | 7487 | <p>This is an overloaded intrinsic. You can use llvm.ctpop on any integer bit |
| Owen Anderson | f1ac465 | 2011-07-01 21:52:38 +0000 | [diff] [blame] | 7488 | width, or on any vector with integer elements. Not all targets support all |
| 7489 | bit widths or vector types, however.</p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7490 | |
| Andrew Lenharth | ec370fd | 2005-05-03 18:01:48 +0000 | [diff] [blame] | 7491 | <pre> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7492 | declare i8 @llvm.ctpop.i8(i8 <src>) |
| Chandler Carruth | 6994040 | 2007-08-04 01:51:18 +0000 | [diff] [blame] | 7493 | declare i16 @llvm.ctpop.i16(i16 <src>) |
| Anton Korobeynikov | ec43a06 | 2007-03-22 00:02:17 +0000 | [diff] [blame] | 7494 | declare i32 @llvm.ctpop.i32(i32 <src>) |
| Chandler Carruth | 6994040 | 2007-08-04 01:51:18 +0000 | [diff] [blame] | 7495 | declare i64 @llvm.ctpop.i64(i64 <src>) |
| 7496 | declare i256 @llvm.ctpop.i256(i256 <src>) |
| Owen Anderson | f1ac465 | 2011-07-01 21:52:38 +0000 | [diff] [blame] | 7497 | declare <2 x i32> @llvm.ctpop.v2i32(<2 x i32> <src>) |
| Andrew Lenharth | ec370fd | 2005-05-03 18:01:48 +0000 | [diff] [blame] | 7498 | </pre> |
| 7499 | |
| 7500 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7501 | <p>The '<tt>llvm.ctpop</tt>' family of intrinsics counts the number of bits set |
| 7502 | in a value.</p> |
| Andrew Lenharth | ec370fd | 2005-05-03 18:01:48 +0000 | [diff] [blame] | 7503 | |
| 7504 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7505 | <p>The only argument is the value to be counted. The argument may be of any |
| Owen Anderson | f1ac465 | 2011-07-01 21:52:38 +0000 | [diff] [blame] | 7506 | integer type, or a vector with integer elements. |
| 7507 | The return type must match the argument type.</p> |
| Andrew Lenharth | ec370fd | 2005-05-03 18:01:48 +0000 | [diff] [blame] | 7508 | |
| 7509 | <h5>Semantics:</h5> |
| Owen Anderson | f1ac465 | 2011-07-01 21:52:38 +0000 | [diff] [blame] | 7510 | <p>The '<tt>llvm.ctpop</tt>' intrinsic counts the 1's in a variable, or within each |
| 7511 | element of a vector.</p> |
| Andrew Lenharth | ec370fd | 2005-05-03 18:01:48 +0000 | [diff] [blame] | 7512 | |
| Andrew Lenharth | ec370fd | 2005-05-03 18:01:48 +0000 | [diff] [blame] | 7513 | </div> |
| 7514 | |
| 7515 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7516 | <h4> |
| Chris Lattner | 8a886be | 2006-01-16 22:34:14 +0000 | [diff] [blame] | 7517 | <a name="int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7518 | </h4> |
| Andrew Lenharth | ec370fd | 2005-05-03 18:01:48 +0000 | [diff] [blame] | 7519 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 7520 | <div> |
| Andrew Lenharth | ec370fd | 2005-05-03 18:01:48 +0000 | [diff] [blame] | 7521 | |
| 7522 | <h5>Syntax:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7523 | <p>This is an overloaded intrinsic. You can use <tt>llvm.ctlz</tt> on any |
| Owen Anderson | f1ac465 | 2011-07-01 21:52:38 +0000 | [diff] [blame] | 7524 | integer bit width, or any vector whose elements are integers. Not all |
| 7525 | targets support all bit widths or vector types, however.</p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7526 | |
| Andrew Lenharth | ec370fd | 2005-05-03 18:01:48 +0000 | [diff] [blame] | 7527 | <pre> |
| Chandler Carruth | 48b0bbf | 2011-12-12 04:36:04 +0000 | [diff] [blame] | 7528 | declare i8 @llvm.ctlz.i8 (i8 <src>, i1 <is_zero_undef>) |
| 7529 | declare i16 @llvm.ctlz.i16 (i16 <src>, i1 <is_zero_undef>) |
| 7530 | declare i32 @llvm.ctlz.i32 (i32 <src>, i1 <is_zero_undef>) |
| 7531 | declare i64 @llvm.ctlz.i64 (i64 <src>, i1 <is_zero_undef>) |
| 7532 | declare i256 @llvm.ctlz.i256(i256 <src>, i1 <is_zero_undef>) |
| 7533 | declase <2 x i32> @llvm.ctlz.v2i32(<2 x i32> <src>, i1 <is_zero_undef>) |
| Andrew Lenharth | ec370fd | 2005-05-03 18:01:48 +0000 | [diff] [blame] | 7534 | </pre> |
| 7535 | |
| 7536 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7537 | <p>The '<tt>llvm.ctlz</tt>' family of intrinsic functions counts the number of |
| 7538 | leading zeros in a variable.</p> |
| Andrew Lenharth | ec370fd | 2005-05-03 18:01:48 +0000 | [diff] [blame] | 7539 | |
| 7540 | <h5>Arguments:</h5> |
| Chandler Carruth | 48b0bbf | 2011-12-12 04:36:04 +0000 | [diff] [blame] | 7541 | <p>The first argument is the value to be counted. This argument may be of any |
| 7542 | integer type, or a vectory with integer element type. The return type |
| 7543 | must match the first argument type.</p> |
| 7544 | |
| 7545 | <p>The second argument must be a constant and is a flag to indicate whether the |
| 7546 | intrinsic should ensure that a zero as the first argument produces a defined |
| 7547 | result. Historically some architectures did not provide a defined result for |
| 7548 | zero values as efficiently, and many algorithms are now predicated on |
| 7549 | avoiding zero-value inputs.</p> |
| Andrew Lenharth | ec370fd | 2005-05-03 18:01:48 +0000 | [diff] [blame] | 7550 | |
| 7551 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7552 | <p>The '<tt>llvm.ctlz</tt>' intrinsic counts the leading (most significant) |
| Chandler Carruth | 48b0bbf | 2011-12-12 04:36:04 +0000 | [diff] [blame] | 7553 | zeros in a variable, or within each element of the vector. |
| 7554 | If <tt>src == 0</tt> then the result is the size in bits of the type of |
| 7555 | <tt>src</tt> if <tt>is_zero_undef == 0</tt> and <tt>undef</tt> otherwise. |
| 7556 | For example, <tt>llvm.ctlz(i32 2) = 30</tt>.</p> |
| Andrew Lenharth | ec370fd | 2005-05-03 18:01:48 +0000 | [diff] [blame] | 7557 | |
| Andrew Lenharth | ec370fd | 2005-05-03 18:01:48 +0000 | [diff] [blame] | 7558 | </div> |
| Chris Lattner | 3200628 | 2004-06-11 02:28:03 +0000 | [diff] [blame] | 7559 | |
| Chris Lattner | eff29ab | 2005-05-15 19:39:26 +0000 | [diff] [blame] | 7560 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7561 | <h4> |
| Chris Lattner | 8a886be | 2006-01-16 22:34:14 +0000 | [diff] [blame] | 7562 | <a name="int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7563 | </h4> |
| Chris Lattner | eff29ab | 2005-05-15 19:39:26 +0000 | [diff] [blame] | 7564 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 7565 | <div> |
| Chris Lattner | eff29ab | 2005-05-15 19:39:26 +0000 | [diff] [blame] | 7566 | |
| 7567 | <h5>Syntax:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7568 | <p>This is an overloaded intrinsic. You can use <tt>llvm.cttz</tt> on any |
| Owen Anderson | f1ac465 | 2011-07-01 21:52:38 +0000 | [diff] [blame] | 7569 | integer bit width, or any vector of integer elements. Not all targets |
| 7570 | support all bit widths or vector types, however.</p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7571 | |
| Chris Lattner | eff29ab | 2005-05-15 19:39:26 +0000 | [diff] [blame] | 7572 | <pre> |
| Chandler Carruth | 48b0bbf | 2011-12-12 04:36:04 +0000 | [diff] [blame] | 7573 | declare i8 @llvm.cttz.i8 (i8 <src>, i1 <is_zero_undef>) |
| 7574 | declare i16 @llvm.cttz.i16 (i16 <src>, i1 <is_zero_undef>) |
| 7575 | declare i32 @llvm.cttz.i32 (i32 <src>, i1 <is_zero_undef>) |
| 7576 | declare i64 @llvm.cttz.i64 (i64 <src>, i1 <is_zero_undef>) |
| 7577 | declare i256 @llvm.cttz.i256(i256 <src>, i1 <is_zero_undef>) |
| 7578 | declase <2 x i32> @llvm.cttz.v2i32(<2 x i32> <src>, i1 <is_zero_undef>) |
| Chris Lattner | eff29ab | 2005-05-15 19:39:26 +0000 | [diff] [blame] | 7579 | </pre> |
| 7580 | |
| 7581 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7582 | <p>The '<tt>llvm.cttz</tt>' family of intrinsic functions counts the number of |
| 7583 | trailing zeros.</p> |
| Chris Lattner | eff29ab | 2005-05-15 19:39:26 +0000 | [diff] [blame] | 7584 | |
| 7585 | <h5>Arguments:</h5> |
| Chandler Carruth | 48b0bbf | 2011-12-12 04:36:04 +0000 | [diff] [blame] | 7586 | <p>The first argument is the value to be counted. This argument may be of any |
| 7587 | integer type, or a vectory with integer element type. The return type |
| 7588 | must match the first argument type.</p> |
| 7589 | |
| 7590 | <p>The second argument must be a constant and is a flag to indicate whether the |
| 7591 | intrinsic should ensure that a zero as the first argument produces a defined |
| 7592 | result. Historically some architectures did not provide a defined result for |
| 7593 | zero values as efficiently, and many algorithms are now predicated on |
| 7594 | avoiding zero-value inputs.</p> |
| Chris Lattner | eff29ab | 2005-05-15 19:39:26 +0000 | [diff] [blame] | 7595 | |
| 7596 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7597 | <p>The '<tt>llvm.cttz</tt>' intrinsic counts the trailing (least significant) |
| Owen Anderson | f1ac465 | 2011-07-01 21:52:38 +0000 | [diff] [blame] | 7598 | zeros in a variable, or within each element of a vector. |
| Chandler Carruth | 48b0bbf | 2011-12-12 04:36:04 +0000 | [diff] [blame] | 7599 | If <tt>src == 0</tt> then the result is the size in bits of the type of |
| 7600 | <tt>src</tt> if <tt>is_zero_undef == 0</tt> and <tt>undef</tt> otherwise. |
| 7601 | For example, <tt>llvm.cttz(2) = 1</tt>.</p> |
| Chris Lattner | eff29ab | 2005-05-15 19:39:26 +0000 | [diff] [blame] | 7602 | |
| Chris Lattner | eff29ab | 2005-05-15 19:39:26 +0000 | [diff] [blame] | 7603 | </div> |
| 7604 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 7605 | </div> |
| 7606 | |
| Bill Wendling | da01af7 | 2009-02-08 04:04:40 +0000 | [diff] [blame] | 7607 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7608 | <h3> |
| Bill Wendling | da01af7 | 2009-02-08 04:04:40 +0000 | [diff] [blame] | 7609 | <a name="int_overflow">Arithmetic with Overflow Intrinsics</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7610 | </h3> |
| Bill Wendling | da01af7 | 2009-02-08 04:04:40 +0000 | [diff] [blame] | 7611 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 7612 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7613 | |
| 7614 | <p>LLVM provides intrinsics for some arithmetic with overflow operations.</p> |
| Bill Wendling | da01af7 | 2009-02-08 04:04:40 +0000 | [diff] [blame] | 7615 | |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7616 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7617 | <h4> |
| 7618 | <a name="int_sadd_overflow"> |
| 7619 | '<tt>llvm.sadd.with.overflow.*</tt>' Intrinsics |
| 7620 | </a> |
| 7621 | </h4> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7622 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 7623 | <div> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7624 | |
| 7625 | <h5>Syntax:</h5> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7626 | <p>This is an overloaded intrinsic. You can use <tt>llvm.sadd.with.overflow</tt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7627 | on any integer bit width.</p> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7628 | |
| 7629 | <pre> |
| 7630 | declare {i16, i1} @llvm.sadd.with.overflow.i16(i16 %a, i16 %b) |
| 7631 | declare {i32, i1} @llvm.sadd.with.overflow.i32(i32 %a, i32 %b) |
| 7632 | declare {i64, i1} @llvm.sadd.with.overflow.i64(i64 %a, i64 %b) |
| 7633 | </pre> |
| 7634 | |
| 7635 | <h5>Overview:</h5> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7636 | <p>The '<tt>llvm.sadd.with.overflow</tt>' family of intrinsic functions perform |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7637 | a signed addition of the two arguments, and indicate whether an overflow |
| 7638 | occurred during the signed summation.</p> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7639 | |
| 7640 | <h5>Arguments:</h5> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7641 | <p>The arguments (%a and %b) and the first element of the result structure may |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7642 | be of integer types of any bit width, but they must have the same bit |
| 7643 | width. The second element of the result structure must be of |
| 7644 | type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will |
| 7645 | undergo signed addition.</p> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7646 | |
| 7647 | <h5>Semantics:</h5> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7648 | <p>The '<tt>llvm.sadd.with.overflow</tt>' family of intrinsic functions perform |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7649 | a signed addition of the two variables. They return a structure — the |
| 7650 | first element of which is the signed summation, and the second element of |
| 7651 | which is a bit specifying if the signed summation resulted in an |
| 7652 | overflow.</p> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7653 | |
| 7654 | <h5>Examples:</h5> |
| 7655 | <pre> |
| 7656 | %res = call {i32, i1} @llvm.sadd.with.overflow.i32(i32 %a, i32 %b) |
| 7657 | %sum = extractvalue {i32, i1} %res, 0 |
| 7658 | %obit = extractvalue {i32, i1} %res, 1 |
| 7659 | br i1 %obit, label %overflow, label %normal |
| 7660 | </pre> |
| 7661 | |
| 7662 | </div> |
| 7663 | |
| 7664 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7665 | <h4> |
| 7666 | <a name="int_uadd_overflow"> |
| 7667 | '<tt>llvm.uadd.with.overflow.*</tt>' Intrinsics |
| 7668 | </a> |
| 7669 | </h4> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7670 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 7671 | <div> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7672 | |
| 7673 | <h5>Syntax:</h5> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7674 | <p>This is an overloaded intrinsic. You can use <tt>llvm.uadd.with.overflow</tt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7675 | on any integer bit width.</p> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7676 | |
| 7677 | <pre> |
| 7678 | declare {i16, i1} @llvm.uadd.with.overflow.i16(i16 %a, i16 %b) |
| 7679 | declare {i32, i1} @llvm.uadd.with.overflow.i32(i32 %a, i32 %b) |
| 7680 | declare {i64, i1} @llvm.uadd.with.overflow.i64(i64 %a, i64 %b) |
| 7681 | </pre> |
| 7682 | |
| 7683 | <h5>Overview:</h5> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7684 | <p>The '<tt>llvm.uadd.with.overflow</tt>' family of intrinsic functions perform |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7685 | an unsigned addition of the two arguments, and indicate whether a carry |
| 7686 | occurred during the unsigned summation.</p> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7687 | |
| 7688 | <h5>Arguments:</h5> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7689 | <p>The arguments (%a and %b) and the first element of the result structure may |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7690 | be of integer types of any bit width, but they must have the same bit |
| 7691 | width. The second element of the result structure must be of |
| 7692 | type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will |
| 7693 | undergo unsigned addition.</p> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7694 | |
| 7695 | <h5>Semantics:</h5> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7696 | <p>The '<tt>llvm.uadd.with.overflow</tt>' family of intrinsic functions perform |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7697 | an unsigned addition of the two arguments. They return a structure — |
| 7698 | the first element of which is the sum, and the second element of which is a |
| 7699 | bit specifying if the unsigned summation resulted in a carry.</p> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7700 | |
| 7701 | <h5>Examples:</h5> |
| 7702 | <pre> |
| 7703 | %res = call {i32, i1} @llvm.uadd.with.overflow.i32(i32 %a, i32 %b) |
| 7704 | %sum = extractvalue {i32, i1} %res, 0 |
| 7705 | %obit = extractvalue {i32, i1} %res, 1 |
| 7706 | br i1 %obit, label %carry, label %normal |
| 7707 | </pre> |
| 7708 | |
| 7709 | </div> |
| 7710 | |
| 7711 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7712 | <h4> |
| 7713 | <a name="int_ssub_overflow"> |
| 7714 | '<tt>llvm.ssub.with.overflow.*</tt>' Intrinsics |
| 7715 | </a> |
| 7716 | </h4> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7717 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 7718 | <div> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7719 | |
| 7720 | <h5>Syntax:</h5> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7721 | <p>This is an overloaded intrinsic. You can use <tt>llvm.ssub.with.overflow</tt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7722 | on any integer bit width.</p> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7723 | |
| 7724 | <pre> |
| 7725 | declare {i16, i1} @llvm.ssub.with.overflow.i16(i16 %a, i16 %b) |
| 7726 | declare {i32, i1} @llvm.ssub.with.overflow.i32(i32 %a, i32 %b) |
| 7727 | declare {i64, i1} @llvm.ssub.with.overflow.i64(i64 %a, i64 %b) |
| 7728 | </pre> |
| 7729 | |
| 7730 | <h5>Overview:</h5> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7731 | <p>The '<tt>llvm.ssub.with.overflow</tt>' family of intrinsic functions perform |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7732 | a signed subtraction of the two arguments, and indicate whether an overflow |
| 7733 | occurred during the signed subtraction.</p> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7734 | |
| 7735 | <h5>Arguments:</h5> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7736 | <p>The arguments (%a and %b) and the first element of the result structure may |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7737 | be of integer types of any bit width, but they must have the same bit |
| 7738 | width. The second element of the result structure must be of |
| 7739 | type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will |
| 7740 | undergo signed subtraction.</p> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7741 | |
| 7742 | <h5>Semantics:</h5> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7743 | <p>The '<tt>llvm.ssub.with.overflow</tt>' family of intrinsic functions perform |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7744 | a signed subtraction of the two arguments. They return a structure — |
| 7745 | the first element of which is the subtraction, and the second element of |
| 7746 | which is a bit specifying if the signed subtraction resulted in an |
| 7747 | overflow.</p> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7748 | |
| 7749 | <h5>Examples:</h5> |
| 7750 | <pre> |
| 7751 | %res = call {i32, i1} @llvm.ssub.with.overflow.i32(i32 %a, i32 %b) |
| 7752 | %sum = extractvalue {i32, i1} %res, 0 |
| 7753 | %obit = extractvalue {i32, i1} %res, 1 |
| 7754 | br i1 %obit, label %overflow, label %normal |
| 7755 | </pre> |
| 7756 | |
| 7757 | </div> |
| 7758 | |
| 7759 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7760 | <h4> |
| 7761 | <a name="int_usub_overflow"> |
| 7762 | '<tt>llvm.usub.with.overflow.*</tt>' Intrinsics |
| 7763 | </a> |
| 7764 | </h4> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7765 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 7766 | <div> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7767 | |
| 7768 | <h5>Syntax:</h5> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7769 | <p>This is an overloaded intrinsic. You can use <tt>llvm.usub.with.overflow</tt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7770 | on any integer bit width.</p> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7771 | |
| 7772 | <pre> |
| 7773 | declare {i16, i1} @llvm.usub.with.overflow.i16(i16 %a, i16 %b) |
| 7774 | declare {i32, i1} @llvm.usub.with.overflow.i32(i32 %a, i32 %b) |
| 7775 | declare {i64, i1} @llvm.usub.with.overflow.i64(i64 %a, i64 %b) |
| 7776 | </pre> |
| 7777 | |
| 7778 | <h5>Overview:</h5> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7779 | <p>The '<tt>llvm.usub.with.overflow</tt>' family of intrinsic functions perform |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7780 | an unsigned subtraction of the two arguments, and indicate whether an |
| 7781 | overflow occurred during the unsigned subtraction.</p> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7782 | |
| 7783 | <h5>Arguments:</h5> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7784 | <p>The arguments (%a and %b) and the first element of the result structure may |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7785 | be of integer types of any bit width, but they must have the same bit |
| 7786 | width. The second element of the result structure must be of |
| 7787 | type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will |
| 7788 | undergo unsigned subtraction.</p> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7789 | |
| 7790 | <h5>Semantics:</h5> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7791 | <p>The '<tt>llvm.usub.with.overflow</tt>' family of intrinsic functions perform |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7792 | an unsigned subtraction of the two arguments. They return a structure — |
| 7793 | the first element of which is the subtraction, and the second element of |
| 7794 | which is a bit specifying if the unsigned subtraction resulted in an |
| 7795 | overflow.</p> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7796 | |
| 7797 | <h5>Examples:</h5> |
| 7798 | <pre> |
| 7799 | %res = call {i32, i1} @llvm.usub.with.overflow.i32(i32 %a, i32 %b) |
| 7800 | %sum = extractvalue {i32, i1} %res, 0 |
| 7801 | %obit = extractvalue {i32, i1} %res, 1 |
| 7802 | br i1 %obit, label %overflow, label %normal |
| 7803 | </pre> |
| 7804 | |
| 7805 | </div> |
| 7806 | |
| 7807 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7808 | <h4> |
| 7809 | <a name="int_smul_overflow"> |
| 7810 | '<tt>llvm.smul.with.overflow.*</tt>' Intrinsics |
| 7811 | </a> |
| 7812 | </h4> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7813 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 7814 | <div> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7815 | |
| 7816 | <h5>Syntax:</h5> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7817 | <p>This is an overloaded intrinsic. You can use <tt>llvm.smul.with.overflow</tt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7818 | on any integer bit width.</p> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7819 | |
| 7820 | <pre> |
| 7821 | declare {i16, i1} @llvm.smul.with.overflow.i16(i16 %a, i16 %b) |
| 7822 | declare {i32, i1} @llvm.smul.with.overflow.i32(i32 %a, i32 %b) |
| 7823 | declare {i64, i1} @llvm.smul.with.overflow.i64(i64 %a, i64 %b) |
| 7824 | </pre> |
| 7825 | |
| 7826 | <h5>Overview:</h5> |
| 7827 | |
| 7828 | <p>The '<tt>llvm.smul.with.overflow</tt>' family of intrinsic functions perform |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7829 | a signed multiplication of the two arguments, and indicate whether an |
| 7830 | overflow occurred during the signed multiplication.</p> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7831 | |
| 7832 | <h5>Arguments:</h5> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7833 | <p>The arguments (%a and %b) and the first element of the result structure may |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7834 | be of integer types of any bit width, but they must have the same bit |
| 7835 | width. The second element of the result structure must be of |
| 7836 | type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will |
| 7837 | undergo signed multiplication.</p> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7838 | |
| 7839 | <h5>Semantics:</h5> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7840 | <p>The '<tt>llvm.smul.with.overflow</tt>' family of intrinsic functions perform |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7841 | a signed multiplication of the two arguments. They return a structure — |
| 7842 | the first element of which is the multiplication, and the second element of |
| 7843 | which is a bit specifying if the signed multiplication resulted in an |
| 7844 | overflow.</p> |
| Bill Wendling | ac1df8e | 2009-02-08 01:40:31 +0000 | [diff] [blame] | 7845 | |
| 7846 | <h5>Examples:</h5> |
| 7847 | <pre> |
| 7848 | %res = call {i32, i1} @llvm.smul.with.overflow.i32(i32 %a, i32 %b) |
| 7849 | %sum = extractvalue {i32, i1} %res, 0 |
| 7850 | %obit = extractvalue {i32, i1} %res, 1 |
| 7851 | br i1 %obit, label %overflow, label %normal |
| 7852 | </pre> |
| 7853 | |
| Reid Spencer | f86037f | 2007-04-11 23:23:49 +0000 | [diff] [blame] | 7854 | </div> |
| 7855 | |
| Bill Wendling | 41b485c | 2009-02-08 23:00:09 +0000 | [diff] [blame] | 7856 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7857 | <h4> |
| 7858 | <a name="int_umul_overflow"> |
| 7859 | '<tt>llvm.umul.with.overflow.*</tt>' Intrinsics |
| 7860 | </a> |
| 7861 | </h4> |
| Bill Wendling | 41b485c | 2009-02-08 23:00:09 +0000 | [diff] [blame] | 7862 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 7863 | <div> |
| Bill Wendling | 41b485c | 2009-02-08 23:00:09 +0000 | [diff] [blame] | 7864 | |
| 7865 | <h5>Syntax:</h5> |
| Bill Wendling | 41b485c | 2009-02-08 23:00:09 +0000 | [diff] [blame] | 7866 | <p>This is an overloaded intrinsic. You can use <tt>llvm.umul.with.overflow</tt> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7867 | on any integer bit width.</p> |
| Bill Wendling | 41b485c | 2009-02-08 23:00:09 +0000 | [diff] [blame] | 7868 | |
| 7869 | <pre> |
| 7870 | declare {i16, i1} @llvm.umul.with.overflow.i16(i16 %a, i16 %b) |
| 7871 | declare {i32, i1} @llvm.umul.with.overflow.i32(i32 %a, i32 %b) |
| 7872 | declare {i64, i1} @llvm.umul.with.overflow.i64(i64 %a, i64 %b) |
| 7873 | </pre> |
| 7874 | |
| 7875 | <h5>Overview:</h5> |
| Bill Wendling | 41b485c | 2009-02-08 23:00:09 +0000 | [diff] [blame] | 7876 | <p>The '<tt>llvm.umul.with.overflow</tt>' family of intrinsic functions perform |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7877 | a unsigned multiplication of the two arguments, and indicate whether an |
| 7878 | overflow occurred during the unsigned multiplication.</p> |
| Bill Wendling | 41b485c | 2009-02-08 23:00:09 +0000 | [diff] [blame] | 7879 | |
| 7880 | <h5>Arguments:</h5> |
| Bill Wendling | 41b485c | 2009-02-08 23:00:09 +0000 | [diff] [blame] | 7881 | <p>The arguments (%a and %b) and the first element of the result structure may |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7882 | be of integer types of any bit width, but they must have the same bit |
| 7883 | width. The second element of the result structure must be of |
| 7884 | type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will |
| 7885 | undergo unsigned multiplication.</p> |
| Bill Wendling | 41b485c | 2009-02-08 23:00:09 +0000 | [diff] [blame] | 7886 | |
| 7887 | <h5>Semantics:</h5> |
| Bill Wendling | 41b485c | 2009-02-08 23:00:09 +0000 | [diff] [blame] | 7888 | <p>The '<tt>llvm.umul.with.overflow</tt>' family of intrinsic functions perform |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 7889 | an unsigned multiplication of the two arguments. They return a structure |
| 7890 | — the first element of which is the multiplication, and the second |
| 7891 | element of which is a bit specifying if the unsigned multiplication resulted |
| 7892 | in an overflow.</p> |
| Bill Wendling | 41b485c | 2009-02-08 23:00:09 +0000 | [diff] [blame] | 7893 | |
| 7894 | <h5>Examples:</h5> |
| 7895 | <pre> |
| 7896 | %res = call {i32, i1} @llvm.umul.with.overflow.i32(i32 %a, i32 %b) |
| 7897 | %sum = extractvalue {i32, i1} %res, 0 |
| 7898 | %obit = extractvalue {i32, i1} %res, 1 |
| 7899 | br i1 %obit, label %overflow, label %normal |
| 7900 | </pre> |
| 7901 | |
| 7902 | </div> |
| 7903 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 7904 | </div> |
| 7905 | |
| Chris Lattner | 8ff7590 | 2004-01-06 05:31:32 +0000 | [diff] [blame] | 7906 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7907 | <h3> |
| Anton Korobeynikov | f02e730 | 2010-03-14 18:42:47 +0000 | [diff] [blame] | 7908 | <a name="int_fp16">Half Precision Floating Point Intrinsics</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7909 | </h3> |
| Anton Korobeynikov | f02e730 | 2010-03-14 18:42:47 +0000 | [diff] [blame] | 7910 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 7911 | <div> |
| Anton Korobeynikov | f02e730 | 2010-03-14 18:42:47 +0000 | [diff] [blame] | 7912 | |
| Chris Lattner | 0cec9c8 | 2010-03-15 04:12:21 +0000 | [diff] [blame] | 7913 | <p>Half precision floating point is a storage-only format. This means that it is |
| 7914 | a dense encoding (in memory) but does not support computation in the |
| 7915 | format.</p> |
| Chris Lattner | 82c3dc6 | 2010-03-14 23:03:31 +0000 | [diff] [blame] | 7916 | |
| Chris Lattner | 0cec9c8 | 2010-03-15 04:12:21 +0000 | [diff] [blame] | 7917 | <p>This means that code must first load the half-precision floating point |
| Chris Lattner | 82c3dc6 | 2010-03-14 23:03:31 +0000 | [diff] [blame] | 7918 | value as an i16, then convert it to float with <a |
| 7919 | href="#int_convert_from_fp16"><tt>llvm.convert.from.fp16</tt></a>. |
| 7920 | Computation can then be performed on the float value (including extending to |
| Chris Lattner | 0cec9c8 | 2010-03-15 04:12:21 +0000 | [diff] [blame] | 7921 | double etc). To store the value back to memory, it is first converted to |
| 7922 | float if needed, then converted to i16 with |
| Chris Lattner | 82c3dc6 | 2010-03-14 23:03:31 +0000 | [diff] [blame] | 7923 | <a href="#int_convert_to_fp16"><tt>llvm.convert.to.fp16</tt></a>, then |
| 7924 | storing as an i16 value.</p> |
| Anton Korobeynikov | f02e730 | 2010-03-14 18:42:47 +0000 | [diff] [blame] | 7925 | |
| 7926 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7927 | <h4> |
| 7928 | <a name="int_convert_to_fp16"> |
| 7929 | '<tt>llvm.convert.to.fp16</tt>' Intrinsic |
| 7930 | </a> |
| 7931 | </h4> |
| Anton Korobeynikov | f02e730 | 2010-03-14 18:42:47 +0000 | [diff] [blame] | 7932 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 7933 | <div> |
| Anton Korobeynikov | f02e730 | 2010-03-14 18:42:47 +0000 | [diff] [blame] | 7934 | |
| 7935 | <h5>Syntax:</h5> |
| 7936 | <pre> |
| 7937 | declare i16 @llvm.convert.to.fp16(f32 %a) |
| 7938 | </pre> |
| 7939 | |
| 7940 | <h5>Overview:</h5> |
| 7941 | <p>The '<tt>llvm.convert.to.fp16</tt>' intrinsic function performs |
| 7942 | a conversion from single precision floating point format to half precision |
| 7943 | floating point format.</p> |
| 7944 | |
| 7945 | <h5>Arguments:</h5> |
| 7946 | <p>The intrinsic function contains single argument - the value to be |
| 7947 | converted.</p> |
| 7948 | |
| 7949 | <h5>Semantics:</h5> |
| 7950 | <p>The '<tt>llvm.convert.to.fp16</tt>' intrinsic function performs |
| 7951 | a conversion from single precision floating point format to half precision |
| Chris Lattner | 0cec9c8 | 2010-03-15 04:12:21 +0000 | [diff] [blame] | 7952 | floating point format. The return value is an <tt>i16</tt> which |
| Chris Lattner | 82c3dc6 | 2010-03-14 23:03:31 +0000 | [diff] [blame] | 7953 | contains the converted number.</p> |
| Anton Korobeynikov | f02e730 | 2010-03-14 18:42:47 +0000 | [diff] [blame] | 7954 | |
| 7955 | <h5>Examples:</h5> |
| 7956 | <pre> |
| 7957 | %res = call i16 @llvm.convert.to.fp16(f32 %a) |
| 7958 | store i16 %res, i16* @x, align 2 |
| 7959 | </pre> |
| 7960 | |
| 7961 | </div> |
| 7962 | |
| 7963 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 7964 | <h4> |
| 7965 | <a name="int_convert_from_fp16"> |
| 7966 | '<tt>llvm.convert.from.fp16</tt>' Intrinsic |
| 7967 | </a> |
| 7968 | </h4> |
| Anton Korobeynikov | f02e730 | 2010-03-14 18:42:47 +0000 | [diff] [blame] | 7969 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 7970 | <div> |
| Anton Korobeynikov | f02e730 | 2010-03-14 18:42:47 +0000 | [diff] [blame] | 7971 | |
| 7972 | <h5>Syntax:</h5> |
| 7973 | <pre> |
| 7974 | declare f32 @llvm.convert.from.fp16(i16 %a) |
| 7975 | </pre> |
| 7976 | |
| 7977 | <h5>Overview:</h5> |
| 7978 | <p>The '<tt>llvm.convert.from.fp16</tt>' intrinsic function performs |
| 7979 | a conversion from half precision floating point format to single precision |
| 7980 | floating point format.</p> |
| 7981 | |
| 7982 | <h5>Arguments:</h5> |
| 7983 | <p>The intrinsic function contains single argument - the value to be |
| 7984 | converted.</p> |
| 7985 | |
| 7986 | <h5>Semantics:</h5> |
| 7987 | <p>The '<tt>llvm.convert.from.fp16</tt>' intrinsic function performs a |
| Chris Lattner | 0cec9c8 | 2010-03-15 04:12:21 +0000 | [diff] [blame] | 7988 | conversion from half single precision floating point format to single |
| Chris Lattner | 82c3dc6 | 2010-03-14 23:03:31 +0000 | [diff] [blame] | 7989 | precision floating point format. The input half-float value is represented by |
| 7990 | an <tt>i16</tt> value.</p> |
| Anton Korobeynikov | f02e730 | 2010-03-14 18:42:47 +0000 | [diff] [blame] | 7991 | |
| 7992 | <h5>Examples:</h5> |
| 7993 | <pre> |
| 7994 | %a = load i16* @x, align 2 |
| 7995 | %res = call f32 @llvm.convert.from.fp16(i16 %a) |
| 7996 | </pre> |
| 7997 | |
| 7998 | </div> |
| 7999 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 8000 | </div> |
| 8001 | |
| Anton Korobeynikov | f02e730 | 2010-03-14 18:42:47 +0000 | [diff] [blame] | 8002 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 8003 | <h3> |
| Chris Lattner | 8ff7590 | 2004-01-06 05:31:32 +0000 | [diff] [blame] | 8004 | <a name="int_debugger">Debugger Intrinsics</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 8005 | </h3> |
| Chris Lattner | 8ff7590 | 2004-01-06 05:31:32 +0000 | [diff] [blame] | 8006 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 8007 | <div> |
| Chris Lattner | 8ff7590 | 2004-01-06 05:31:32 +0000 | [diff] [blame] | 8008 | |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8009 | <p>The LLVM debugger intrinsics (which all start with <tt>llvm.dbg.</tt> |
| 8010 | prefix), are described in |
| 8011 | the <a href="SourceLevelDebugging.html#format_common_intrinsics">LLVM Source |
| 8012 | Level Debugging</a> document.</p> |
| 8013 | |
| 8014 | </div> |
| Chris Lattner | 8ff7590 | 2004-01-06 05:31:32 +0000 | [diff] [blame] | 8015 | |
| Jim Laskey | dd4ef1b | 2007-03-14 19:31:19 +0000 | [diff] [blame] | 8016 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 8017 | <h3> |
| Jim Laskey | dd4ef1b | 2007-03-14 19:31:19 +0000 | [diff] [blame] | 8018 | <a name="int_eh">Exception Handling Intrinsics</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 8019 | </h3> |
| Jim Laskey | dd4ef1b | 2007-03-14 19:31:19 +0000 | [diff] [blame] | 8020 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 8021 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8022 | |
| 8023 | <p>The LLVM exception handling intrinsics (which all start with |
| 8024 | <tt>llvm.eh.</tt> prefix), are described in |
| 8025 | the <a href="ExceptionHandling.html#format_common_intrinsics">LLVM Exception |
| 8026 | Handling</a> document.</p> |
| 8027 | |
| Jim Laskey | dd4ef1b | 2007-03-14 19:31:19 +0000 | [diff] [blame] | 8028 | </div> |
| 8029 | |
| Tanya Lattner | 6d806e9 | 2007-06-15 20:50:54 +0000 | [diff] [blame] | 8030 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 8031 | <h3> |
| Duncan Sands | 4a544a7 | 2011-09-06 13:37:06 +0000 | [diff] [blame] | 8032 | <a name="int_trampoline">Trampoline Intrinsics</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 8033 | </h3> |
| Duncan Sands | 36397f5 | 2007-07-27 12:58:54 +0000 | [diff] [blame] | 8034 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 8035 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8036 | |
| Duncan Sands | 4a544a7 | 2011-09-06 13:37:06 +0000 | [diff] [blame] | 8037 | <p>These intrinsics make it possible to excise one parameter, marked with |
| Dan Gohman | ff23535 | 2010-07-02 23:18:08 +0000 | [diff] [blame] | 8038 | the <a href="#nest"><tt>nest</tt></a> attribute, from a function. |
| 8039 | The result is a callable |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8040 | function pointer lacking the nest parameter - the caller does not need to |
| 8041 | provide a value for it. Instead, the value to use is stored in advance in a |
| 8042 | "trampoline", a block of memory usually allocated on the stack, which also |
| 8043 | contains code to splice the nest value into the argument list. This is used |
| 8044 | to implement the GCC nested function address extension.</p> |
| 8045 | |
| 8046 | <p>For example, if the function is |
| 8047 | <tt>i32 f(i8* nest %c, i32 %x, i32 %y)</tt> then the resulting function |
| 8048 | pointer has signature <tt>i32 (i32, i32)*</tt>. It can be created as |
| 8049 | follows:</p> |
| 8050 | |
| Benjamin Kramer | 26fe25f | 2010-07-13 12:26:09 +0000 | [diff] [blame] | 8051 | <pre class="doc_code"> |
| Duncan Sands | f7331b3 | 2007-09-11 14:10:23 +0000 | [diff] [blame] | 8052 | %tramp = alloca [10 x i8], align 4 ; size and alignment only correct for X86 |
| 8053 | %tramp1 = getelementptr [10 x i8]* %tramp, i32 0, i32 0 |
| Duncan Sands | 4a544a7 | 2011-09-06 13:37:06 +0000 | [diff] [blame] | 8054 | call i8* @llvm.init.trampoline(i8* %tramp1, i8* bitcast (i32 (i8*, i32, i32)* @f to i8*), i8* %nval) |
| 8055 | %p = call i8* @llvm.adjust.trampoline(i8* %tramp1) |
| Duncan Sands | f7331b3 | 2007-09-11 14:10:23 +0000 | [diff] [blame] | 8056 | %fp = bitcast i8* %p to i32 (i32, i32)* |
| Duncan Sands | 36397f5 | 2007-07-27 12:58:54 +0000 | [diff] [blame] | 8057 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8058 | |
| Dan Gohman | 3dfb3cf | 2010-05-28 17:07:41 +0000 | [diff] [blame] | 8059 | <p>The call <tt>%val = call i32 %fp(i32 %x, i32 %y)</tt> is then equivalent |
| 8060 | to <tt>%val = call i32 %f(i8* %nval, i32 %x, i32 %y)</tt>.</p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8061 | |
| Duncan Sands | 36397f5 | 2007-07-27 12:58:54 +0000 | [diff] [blame] | 8062 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 8063 | <h4> |
| 8064 | <a name="int_it"> |
| 8065 | '<tt>llvm.init.trampoline</tt>' Intrinsic |
| 8066 | </a> |
| 8067 | </h4> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8068 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 8069 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8070 | |
| Duncan Sands | 36397f5 | 2007-07-27 12:58:54 +0000 | [diff] [blame] | 8071 | <h5>Syntax:</h5> |
| 8072 | <pre> |
| Duncan Sands | 4a544a7 | 2011-09-06 13:37:06 +0000 | [diff] [blame] | 8073 | declare void @llvm.init.trampoline(i8* <tramp>, i8* <func>, i8* <nval>) |
| Duncan Sands | 36397f5 | 2007-07-27 12:58:54 +0000 | [diff] [blame] | 8074 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8075 | |
| Duncan Sands | 36397f5 | 2007-07-27 12:58:54 +0000 | [diff] [blame] | 8076 | <h5>Overview:</h5> |
| Duncan Sands | 4a544a7 | 2011-09-06 13:37:06 +0000 | [diff] [blame] | 8077 | <p>This fills the memory pointed to by <tt>tramp</tt> with executable code, |
| 8078 | turning it into a trampoline.</p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8079 | |
| Duncan Sands | 36397f5 | 2007-07-27 12:58:54 +0000 | [diff] [blame] | 8080 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8081 | <p>The <tt>llvm.init.trampoline</tt> intrinsic takes three arguments, all |
| 8082 | pointers. The <tt>tramp</tt> argument must point to a sufficiently large and |
| 8083 | sufficiently aligned block of memory; this memory is written to by the |
| 8084 | intrinsic. Note that the size and the alignment are target-specific - LLVM |
| 8085 | currently provides no portable way of determining them, so a front-end that |
| 8086 | generates this intrinsic needs to have some target-specific knowledge. |
| 8087 | The <tt>func</tt> argument must hold a function bitcast to |
| 8088 | an <tt>i8*</tt>.</p> |
| 8089 | |
| Duncan Sands | 36397f5 | 2007-07-27 12:58:54 +0000 | [diff] [blame] | 8090 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8091 | <p>The block of memory pointed to by <tt>tramp</tt> is filled with target |
| Duncan Sands | 4a544a7 | 2011-09-06 13:37:06 +0000 | [diff] [blame] | 8092 | dependent code, turning it into a function. Then <tt>tramp</tt> needs to be |
| 8093 | passed to <a href="#int_at">llvm.adjust.trampoline</a> to get a pointer |
| 8094 | which can be <a href="#int_trampoline">bitcast (to a new function) and |
| 8095 | called</a>. The new function's signature is the same as that of |
| 8096 | <tt>func</tt> with any arguments marked with the <tt>nest</tt> attribute |
| 8097 | removed. At most one such <tt>nest</tt> argument is allowed, and it must be of |
| 8098 | pointer type. Calling the new function is equivalent to calling <tt>func</tt> |
| 8099 | with the same argument list, but with <tt>nval</tt> used for the missing |
| 8100 | <tt>nest</tt> argument. If, after calling <tt>llvm.init.trampoline</tt>, the |
| 8101 | memory pointed to by <tt>tramp</tt> is modified, then the effect of any later call |
| 8102 | to the returned function pointer is undefined.</p> |
| 8103 | </div> |
| 8104 | |
| 8105 | <!-- _______________________________________________________________________ --> |
| 8106 | <h4> |
| 8107 | <a name="int_at"> |
| 8108 | '<tt>llvm.adjust.trampoline</tt>' Intrinsic |
| 8109 | </a> |
| 8110 | </h4> |
| 8111 | |
| 8112 | <div> |
| 8113 | |
| 8114 | <h5>Syntax:</h5> |
| 8115 | <pre> |
| 8116 | declare i8* @llvm.adjust.trampoline(i8* <tramp>) |
| 8117 | </pre> |
| 8118 | |
| 8119 | <h5>Overview:</h5> |
| 8120 | <p>This performs any required machine-specific adjustment to the address of a |
| 8121 | trampoline (passed as <tt>tramp</tt>).</p> |
| 8122 | |
| 8123 | <h5>Arguments:</h5> |
| 8124 | <p><tt>tramp</tt> must point to a block of memory which already has trampoline code |
| 8125 | filled in by a previous call to <a href="#int_it"><tt>llvm.init.trampoline</tt> |
| 8126 | </a>.</p> |
| 8127 | |
| 8128 | <h5>Semantics:</h5> |
| 8129 | <p>On some architectures the address of the code to be executed needs to be |
| 8130 | different to the address where the trampoline is actually stored. This |
| 8131 | intrinsic returns the executable address corresponding to <tt>tramp</tt> |
| 8132 | after performing the required machine specific adjustments. |
| 8133 | The pointer returned can then be <a href="#int_trampoline"> bitcast and |
| 8134 | executed</a>. |
| 8135 | </p> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8136 | |
| Duncan Sands | 36397f5 | 2007-07-27 12:58:54 +0000 | [diff] [blame] | 8137 | </div> |
| 8138 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 8139 | </div> |
| 8140 | |
| Duncan Sands | 36397f5 | 2007-07-27 12:58:54 +0000 | [diff] [blame] | 8141 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 8142 | <h3> |
| Nick Lewycky | cc27186 | 2009-10-13 07:03:23 +0000 | [diff] [blame] | 8143 | <a name="int_memorymarkers">Memory Use Markers</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 8144 | </h3> |
| Nick Lewycky | cc27186 | 2009-10-13 07:03:23 +0000 | [diff] [blame] | 8145 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 8146 | <div> |
| Nick Lewycky | cc27186 | 2009-10-13 07:03:23 +0000 | [diff] [blame] | 8147 | |
| 8148 | <p>This class of intrinsics exists to information about the lifetime of memory |
| 8149 | objects and ranges where variables are immutable.</p> |
| 8150 | |
| Nick Lewycky | cc27186 | 2009-10-13 07:03:23 +0000 | [diff] [blame] | 8151 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 8152 | <h4> |
| Nick Lewycky | cc27186 | 2009-10-13 07:03:23 +0000 | [diff] [blame] | 8153 | <a name="int_lifetime_start">'<tt>llvm.lifetime.start</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 8154 | </h4> |
| Nick Lewycky | cc27186 | 2009-10-13 07:03:23 +0000 | [diff] [blame] | 8155 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 8156 | <div> |
| Nick Lewycky | cc27186 | 2009-10-13 07:03:23 +0000 | [diff] [blame] | 8157 | |
| 8158 | <h5>Syntax:</h5> |
| 8159 | <pre> |
| 8160 | declare void @llvm.lifetime.start(i64 <size>, i8* nocapture <ptr>) |
| 8161 | </pre> |
| 8162 | |
| 8163 | <h5>Overview:</h5> |
| 8164 | <p>The '<tt>llvm.lifetime.start</tt>' intrinsic specifies the start of a memory |
| 8165 | object's lifetime.</p> |
| 8166 | |
| 8167 | <h5>Arguments:</h5> |
| Nick Lewycky | 321333e | 2009-10-13 07:57:33 +0000 | [diff] [blame] | 8168 | <p>The first argument is a constant integer representing the size of the |
| 8169 | object, or -1 if it is variable sized. The second argument is a pointer to |
| 8170 | the object.</p> |
| Nick Lewycky | cc27186 | 2009-10-13 07:03:23 +0000 | [diff] [blame] | 8171 | |
| 8172 | <h5>Semantics:</h5> |
| 8173 | <p>This intrinsic indicates that before this point in the code, the value of the |
| 8174 | memory pointed to by <tt>ptr</tt> is dead. This means that it is known to |
| Nick Lewycky | 8d33659 | 2009-10-27 16:56:58 +0000 | [diff] [blame] | 8175 | never be used and has an undefined value. A load from the pointer that |
| 8176 | precedes this intrinsic can be replaced with |
| Nick Lewycky | cc27186 | 2009-10-13 07:03:23 +0000 | [diff] [blame] | 8177 | <tt>'<a href="#undefvalues">undef</a>'</tt>.</p> |
| 8178 | |
| 8179 | </div> |
| 8180 | |
| 8181 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 8182 | <h4> |
| Nick Lewycky | cc27186 | 2009-10-13 07:03:23 +0000 | [diff] [blame] | 8183 | <a name="int_lifetime_end">'<tt>llvm.lifetime.end</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 8184 | </h4> |
| Nick Lewycky | cc27186 | 2009-10-13 07:03:23 +0000 | [diff] [blame] | 8185 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 8186 | <div> |
| Nick Lewycky | cc27186 | 2009-10-13 07:03:23 +0000 | [diff] [blame] | 8187 | |
| 8188 | <h5>Syntax:</h5> |
| 8189 | <pre> |
| 8190 | declare void @llvm.lifetime.end(i64 <size>, i8* nocapture <ptr>) |
| 8191 | </pre> |
| 8192 | |
| 8193 | <h5>Overview:</h5> |
| 8194 | <p>The '<tt>llvm.lifetime.end</tt>' intrinsic specifies the end of a memory |
| 8195 | object's lifetime.</p> |
| 8196 | |
| 8197 | <h5>Arguments:</h5> |
| Nick Lewycky | 321333e | 2009-10-13 07:57:33 +0000 | [diff] [blame] | 8198 | <p>The first argument is a constant integer representing the size of the |
| 8199 | object, or -1 if it is variable sized. The second argument is a pointer to |
| 8200 | the object.</p> |
| Nick Lewycky | cc27186 | 2009-10-13 07:03:23 +0000 | [diff] [blame] | 8201 | |
| 8202 | <h5>Semantics:</h5> |
| 8203 | <p>This intrinsic indicates that after this point in the code, the value of the |
| 8204 | memory pointed to by <tt>ptr</tt> is dead. This means that it is known to |
| 8205 | never be used and has an undefined value. Any stores into the memory object |
| 8206 | following this intrinsic may be removed as dead. |
| 8207 | |
| 8208 | </div> |
| 8209 | |
| 8210 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 8211 | <h4> |
| Nick Lewycky | cc27186 | 2009-10-13 07:03:23 +0000 | [diff] [blame] | 8212 | <a name="int_invariant_start">'<tt>llvm.invariant.start</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 8213 | </h4> |
| Nick Lewycky | cc27186 | 2009-10-13 07:03:23 +0000 | [diff] [blame] | 8214 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 8215 | <div> |
| Nick Lewycky | cc27186 | 2009-10-13 07:03:23 +0000 | [diff] [blame] | 8216 | |
| 8217 | <h5>Syntax:</h5> |
| 8218 | <pre> |
| Nick Lewycky | 29b6cb4 | 2010-11-30 04:13:41 +0000 | [diff] [blame] | 8219 | declare {}* @llvm.invariant.start(i64 <size>, i8* nocapture <ptr>) |
| Nick Lewycky | cc27186 | 2009-10-13 07:03:23 +0000 | [diff] [blame] | 8220 | </pre> |
| 8221 | |
| 8222 | <h5>Overview:</h5> |
| 8223 | <p>The '<tt>llvm.invariant.start</tt>' intrinsic specifies that the contents of |
| 8224 | a memory object will not change.</p> |
| 8225 | |
| 8226 | <h5>Arguments:</h5> |
| Nick Lewycky | 321333e | 2009-10-13 07:57:33 +0000 | [diff] [blame] | 8227 | <p>The first argument is a constant integer representing the size of the |
| 8228 | object, or -1 if it is variable sized. The second argument is a pointer to |
| 8229 | the object.</p> |
| Nick Lewycky | cc27186 | 2009-10-13 07:03:23 +0000 | [diff] [blame] | 8230 | |
| 8231 | <h5>Semantics:</h5> |
| 8232 | <p>This intrinsic indicates that until an <tt>llvm.invariant.end</tt> that uses |
| 8233 | the return value, the referenced memory location is constant and |
| 8234 | unchanging.</p> |
| 8235 | |
| 8236 | </div> |
| 8237 | |
| 8238 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 8239 | <h4> |
| Nick Lewycky | cc27186 | 2009-10-13 07:03:23 +0000 | [diff] [blame] | 8240 | <a name="int_invariant_end">'<tt>llvm.invariant.end</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 8241 | </h4> |
| Nick Lewycky | cc27186 | 2009-10-13 07:03:23 +0000 | [diff] [blame] | 8242 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 8243 | <div> |
| Nick Lewycky | cc27186 | 2009-10-13 07:03:23 +0000 | [diff] [blame] | 8244 | |
| 8245 | <h5>Syntax:</h5> |
| 8246 | <pre> |
| 8247 | declare void @llvm.invariant.end({}* <start>, i64 <size>, i8* nocapture <ptr>) |
| 8248 | </pre> |
| 8249 | |
| 8250 | <h5>Overview:</h5> |
| 8251 | <p>The '<tt>llvm.invariant.end</tt>' intrinsic specifies that the contents of |
| 8252 | a memory object are mutable.</p> |
| 8253 | |
| 8254 | <h5>Arguments:</h5> |
| 8255 | <p>The first argument is the matching <tt>llvm.invariant.start</tt> intrinsic. |
| Nick Lewycky | 321333e | 2009-10-13 07:57:33 +0000 | [diff] [blame] | 8256 | The second argument is a constant integer representing the size of the |
| 8257 | object, or -1 if it is variable sized and the third argument is a pointer |
| 8258 | to the object.</p> |
| Nick Lewycky | cc27186 | 2009-10-13 07:03:23 +0000 | [diff] [blame] | 8259 | |
| 8260 | <h5>Semantics:</h5> |
| 8261 | <p>This intrinsic indicates that the memory is mutable again.</p> |
| 8262 | |
| 8263 | </div> |
| 8264 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 8265 | </div> |
| 8266 | |
| Andrew Lenharth | 22c5c1b | 2008-02-16 01:24:58 +0000 | [diff] [blame] | 8267 | <!-- ======================================================================= --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 8268 | <h3> |
| Tanya Lattner | 6d806e9 | 2007-06-15 20:50:54 +0000 | [diff] [blame] | 8269 | <a name="int_general">General Intrinsics</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 8270 | </h3> |
| Tanya Lattner | 6d806e9 | 2007-06-15 20:50:54 +0000 | [diff] [blame] | 8271 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 8272 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8273 | |
| 8274 | <p>This class of intrinsics is designed to be generic and has no specific |
| 8275 | purpose.</p> |
| 8276 | |
| Tanya Lattner | 6d806e9 | 2007-06-15 20:50:54 +0000 | [diff] [blame] | 8277 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 8278 | <h4> |
| Tanya Lattner | 6d806e9 | 2007-06-15 20:50:54 +0000 | [diff] [blame] | 8279 | <a name="int_var_annotation">'<tt>llvm.var.annotation</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 8280 | </h4> |
| Tanya Lattner | 6d806e9 | 2007-06-15 20:50:54 +0000 | [diff] [blame] | 8281 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 8282 | <div> |
| Tanya Lattner | 6d806e9 | 2007-06-15 20:50:54 +0000 | [diff] [blame] | 8283 | |
| 8284 | <h5>Syntax:</h5> |
| 8285 | <pre> |
| Dan Gohman | 3dfb3cf | 2010-05-28 17:07:41 +0000 | [diff] [blame] | 8286 | declare void @llvm.var.annotation(i8* <val>, i8* <str>, i8* <str>, i32 <int>) |
| Tanya Lattner | 6d806e9 | 2007-06-15 20:50:54 +0000 | [diff] [blame] | 8287 | </pre> |
| 8288 | |
| 8289 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8290 | <p>The '<tt>llvm.var.annotation</tt>' intrinsic.</p> |
| Tanya Lattner | 6d806e9 | 2007-06-15 20:50:54 +0000 | [diff] [blame] | 8291 | |
| 8292 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8293 | <p>The first argument is a pointer to a value, the second is a pointer to a |
| 8294 | global string, the third is a pointer to a global string which is the source |
| 8295 | file name, and the last argument is the line number.</p> |
| Tanya Lattner | 6d806e9 | 2007-06-15 20:50:54 +0000 | [diff] [blame] | 8296 | |
| 8297 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8298 | <p>This intrinsic allows annotation of local variables with arbitrary strings. |
| 8299 | This can be useful for special purpose optimizations that want to look for |
| John Criswell | e865c03 | 2011-08-19 16:57:55 +0000 | [diff] [blame] | 8300 | these annotations. These have no other defined use; they are ignored by code |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8301 | generation and optimization.</p> |
| Tanya Lattner | 6d806e9 | 2007-06-15 20:50:54 +0000 | [diff] [blame] | 8302 | |
| Tanya Lattner | 6d806e9 | 2007-06-15 20:50:54 +0000 | [diff] [blame] | 8303 | </div> |
| 8304 | |
| Tanya Lattner | b636788 | 2007-09-21 22:59:12 +0000 | [diff] [blame] | 8305 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 8306 | <h4> |
| Tanya Lattner | e1a8da0 | 2007-09-21 23:57:59 +0000 | [diff] [blame] | 8307 | <a name="int_annotation">'<tt>llvm.annotation.*</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 8308 | </h4> |
| Tanya Lattner | b636788 | 2007-09-21 22:59:12 +0000 | [diff] [blame] | 8309 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 8310 | <div> |
| Tanya Lattner | b636788 | 2007-09-21 22:59:12 +0000 | [diff] [blame] | 8311 | |
| 8312 | <h5>Syntax:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8313 | <p>This is an overloaded intrinsic. You can use '<tt>llvm.annotation</tt>' on |
| 8314 | any integer bit width.</p> |
| 8315 | |
| Tanya Lattner | b636788 | 2007-09-21 22:59:12 +0000 | [diff] [blame] | 8316 | <pre> |
| Dan Gohman | 3dfb3cf | 2010-05-28 17:07:41 +0000 | [diff] [blame] | 8317 | declare i8 @llvm.annotation.i8(i8 <val>, i8* <str>, i8* <str>, i32 <int>) |
| 8318 | declare i16 @llvm.annotation.i16(i16 <val>, i8* <str>, i8* <str>, i32 <int>) |
| 8319 | declare i32 @llvm.annotation.i32(i32 <val>, i8* <str>, i8* <str>, i32 <int>) |
| 8320 | declare i64 @llvm.annotation.i64(i64 <val>, i8* <str>, i8* <str>, i32 <int>) |
| 8321 | declare i256 @llvm.annotation.i256(i256 <val>, i8* <str>, i8* <str>, i32 <int>) |
| Tanya Lattner | b636788 | 2007-09-21 22:59:12 +0000 | [diff] [blame] | 8322 | </pre> |
| 8323 | |
| 8324 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8325 | <p>The '<tt>llvm.annotation</tt>' intrinsic.</p> |
| Tanya Lattner | b636788 | 2007-09-21 22:59:12 +0000 | [diff] [blame] | 8326 | |
| 8327 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8328 | <p>The first argument is an integer value (result of some expression), the |
| 8329 | second is a pointer to a global string, the third is a pointer to a global |
| 8330 | string which is the source file name, and the last argument is the line |
| 8331 | number. It returns the value of the first argument.</p> |
| Tanya Lattner | b636788 | 2007-09-21 22:59:12 +0000 | [diff] [blame] | 8332 | |
| 8333 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8334 | <p>This intrinsic allows annotations to be put on arbitrary expressions with |
| 8335 | arbitrary strings. This can be useful for special purpose optimizations that |
| John Criswell | e865c03 | 2011-08-19 16:57:55 +0000 | [diff] [blame] | 8336 | want to look for these annotations. These have no other defined use; they |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8337 | are ignored by code generation and optimization.</p> |
| Tanya Lattner | b636788 | 2007-09-21 22:59:12 +0000 | [diff] [blame] | 8338 | |
| Tanya Lattner | b636788 | 2007-09-21 22:59:12 +0000 | [diff] [blame] | 8339 | </div> |
| Jim Laskey | dd4ef1b | 2007-03-14 19:31:19 +0000 | [diff] [blame] | 8340 | |
| Anton Korobeynikov | 4cb8618 | 2008-01-15 22:31:34 +0000 | [diff] [blame] | 8341 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 8342 | <h4> |
| Anton Korobeynikov | 4cb8618 | 2008-01-15 22:31:34 +0000 | [diff] [blame] | 8343 | <a name="int_trap">'<tt>llvm.trap</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 8344 | </h4> |
| Anton Korobeynikov | 4cb8618 | 2008-01-15 22:31:34 +0000 | [diff] [blame] | 8345 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 8346 | <div> |
| Anton Korobeynikov | 4cb8618 | 2008-01-15 22:31:34 +0000 | [diff] [blame] | 8347 | |
| 8348 | <h5>Syntax:</h5> |
| 8349 | <pre> |
| 8350 | declare void @llvm.trap() |
| 8351 | </pre> |
| 8352 | |
| 8353 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8354 | <p>The '<tt>llvm.trap</tt>' intrinsic.</p> |
| Anton Korobeynikov | 4cb8618 | 2008-01-15 22:31:34 +0000 | [diff] [blame] | 8355 | |
| 8356 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8357 | <p>None.</p> |
| Anton Korobeynikov | 4cb8618 | 2008-01-15 22:31:34 +0000 | [diff] [blame] | 8358 | |
| 8359 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8360 | <p>This intrinsics is lowered to the target dependent trap instruction. If the |
| 8361 | target does not have a trap instruction, this intrinsic will be lowered to |
| 8362 | the call of the <tt>abort()</tt> function.</p> |
| Anton Korobeynikov | 4cb8618 | 2008-01-15 22:31:34 +0000 | [diff] [blame] | 8363 | |
| Anton Korobeynikov | 4cb8618 | 2008-01-15 22:31:34 +0000 | [diff] [blame] | 8364 | </div> |
| 8365 | |
| Bill Wendling | 69e4adb | 2008-11-19 05:56:17 +0000 | [diff] [blame] | 8366 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 8367 | <h4> |
| Misha Brukman | dccb025 | 2008-11-22 23:55:29 +0000 | [diff] [blame] | 8368 | <a name="int_stackprotector">'<tt>llvm.stackprotector</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 8369 | </h4> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8370 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 8371 | <div> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8372 | |
| Bill Wendling | 69e4adb | 2008-11-19 05:56:17 +0000 | [diff] [blame] | 8373 | <h5>Syntax:</h5> |
| 8374 | <pre> |
| Dan Gohman | 3dfb3cf | 2010-05-28 17:07:41 +0000 | [diff] [blame] | 8375 | declare void @llvm.stackprotector(i8* <guard>, i8** <slot>) |
| Bill Wendling | 69e4adb | 2008-11-19 05:56:17 +0000 | [diff] [blame] | 8376 | </pre> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8377 | |
| Bill Wendling | 69e4adb | 2008-11-19 05:56:17 +0000 | [diff] [blame] | 8378 | <h5>Overview:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8379 | <p>The <tt>llvm.stackprotector</tt> intrinsic takes the <tt>guard</tt> and |
| 8380 | stores it onto the stack at <tt>slot</tt>. The stack slot is adjusted to |
| 8381 | ensure that it is placed on the stack before local variables.</p> |
| 8382 | |
| Bill Wendling | 69e4adb | 2008-11-19 05:56:17 +0000 | [diff] [blame] | 8383 | <h5>Arguments:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8384 | <p>The <tt>llvm.stackprotector</tt> intrinsic requires two pointer |
| 8385 | arguments. The first argument is the value loaded from the stack |
| 8386 | guard <tt>@__stack_chk_guard</tt>. The second variable is an <tt>alloca</tt> |
| 8387 | that has enough space to hold the value of the guard.</p> |
| 8388 | |
| Bill Wendling | 69e4adb | 2008-11-19 05:56:17 +0000 | [diff] [blame] | 8389 | <h5>Semantics:</h5> |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8390 | <p>This intrinsic causes the prologue/epilogue inserter to force the position of |
| 8391 | the <tt>AllocaInst</tt> stack slot to be before local variables on the |
| 8392 | stack. This is to ensure that if a local variable on the stack is |
| 8393 | overwritten, it will destroy the value of the guard. When the function exits, |
| Bill Wendling | 1b383ba | 2010-10-27 01:07:41 +0000 | [diff] [blame] | 8394 | the guard on the stack is checked against the original guard. If they are |
| Bill Wendling | e910b4c | 2009-07-20 02:29:24 +0000 | [diff] [blame] | 8395 | different, then the program aborts by calling the <tt>__stack_chk_fail()</tt> |
| 8396 | function.</p> |
| 8397 | |
| Bill Wendling | 69e4adb | 2008-11-19 05:56:17 +0000 | [diff] [blame] | 8398 | </div> |
| 8399 | |
| Eric Christopher | 0e67149 | 2009-11-30 08:03:53 +0000 | [diff] [blame] | 8400 | <!-- _______________________________________________________________________ --> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 8401 | <h4> |
| Eric Christopher | 0e67149 | 2009-11-30 08:03:53 +0000 | [diff] [blame] | 8402 | <a name="int_objectsize">'<tt>llvm.objectsize</tt>' Intrinsic</a> |
| NAKAMURA Takumi | 05d0265 | 2011-04-18 23:59:50 +0000 | [diff] [blame] | 8403 | </h4> |
| Eric Christopher | 0e67149 | 2009-11-30 08:03:53 +0000 | [diff] [blame] | 8404 | |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 8405 | <div> |
| Eric Christopher | 0e67149 | 2009-11-30 08:03:53 +0000 | [diff] [blame] | 8406 | |
| 8407 | <h5>Syntax:</h5> |
| 8408 | <pre> |
| Dan Gohman | 3dfb3cf | 2010-05-28 17:07:41 +0000 | [diff] [blame] | 8409 | declare i32 @llvm.objectsize.i32(i8* <object>, i1 <type>) |
| 8410 | declare i64 @llvm.objectsize.i64(i8* <object>, i1 <type>) |
| Eric Christopher | 0e67149 | 2009-11-30 08:03:53 +0000 | [diff] [blame] | 8411 | </pre> |
| 8412 | |
| 8413 | <h5>Overview:</h5> |
| Bill Wendling | 1b383ba | 2010-10-27 01:07:41 +0000 | [diff] [blame] | 8414 | <p>The <tt>llvm.objectsize</tt> intrinsic is designed to provide information to |
| 8415 | the optimizers to determine at compile time whether a) an operation (like |
| 8416 | memcpy) will overflow a buffer that corresponds to an object, or b) that a |
| 8417 | runtime check for overflow isn't necessary. An object in this context means |
| 8418 | an allocation of a specific class, structure, array, or other object.</p> |
| Eric Christopher | 0e67149 | 2009-11-30 08:03:53 +0000 | [diff] [blame] | 8419 | |
| 8420 | <h5>Arguments:</h5> |
| Bill Wendling | 1b383ba | 2010-10-27 01:07:41 +0000 | [diff] [blame] | 8421 | <p>The <tt>llvm.objectsize</tt> intrinsic takes two arguments. The first |
| Eric Christopher | 8295a0a | 2009-12-23 00:29:49 +0000 | [diff] [blame] | 8422 | argument is a pointer to or into the <tt>object</tt>. The second argument |
| Bill Wendling | 1b383ba | 2010-10-27 01:07:41 +0000 | [diff] [blame] | 8423 | is a boolean 0 or 1. This argument determines whether you want the |
| 8424 | maximum (0) or minimum (1) bytes remaining. This needs to be a literal 0 or |
| Eric Christopher | 8295a0a | 2009-12-23 00:29:49 +0000 | [diff] [blame] | 8425 | 1, variables are not allowed.</p> |
| 8426 | |
| Eric Christopher | 0e67149 | 2009-11-30 08:03:53 +0000 | [diff] [blame] | 8427 | <h5>Semantics:</h5> |
| 8428 | <p>The <tt>llvm.objectsize</tt> intrinsic is lowered to either a constant |
| Bill Wendling | 1b383ba | 2010-10-27 01:07:41 +0000 | [diff] [blame] | 8429 | representing the size of the object concerned, or <tt>i32/i64 -1 or 0</tt>, |
| 8430 | depending on the <tt>type</tt> argument, if the size cannot be determined at |
| 8431 | compile time.</p> |
| Eric Christopher | 0e67149 | 2009-11-30 08:03:53 +0000 | [diff] [blame] | 8432 | |
| 8433 | </div> |
| Jakub Staszak | b170e2d | 2011-12-04 18:29:26 +0000 | [diff] [blame] | 8434 | <!-- _______________________________________________________________________ --> |
| 8435 | <h4> |
| 8436 | <a name="int_expect">'<tt>llvm.expect</tt>' Intrinsic</a> |
| 8437 | </h4> |
| Eric Christopher | 0e67149 | 2009-11-30 08:03:53 +0000 | [diff] [blame] | 8438 | |
| Jakub Staszak | b170e2d | 2011-12-04 18:29:26 +0000 | [diff] [blame] | 8439 | <div> |
| 8440 | |
| 8441 | <h5>Syntax:</h5> |
| 8442 | <pre> |
| 8443 | declare i32 @llvm.expect.i32(i32 <val>, i32 <expected_val>) |
| 8444 | declare i64 @llvm.expect.i64(i64 <val>, i64 <expected_val>) |
| 8445 | </pre> |
| 8446 | |
| 8447 | <h5>Overview:</h5> |
| 8448 | <p>The <tt>llvm.expect</tt> intrinsic provides information about expected (the |
| 8449 | most probable) value of <tt>val</tt>, which can be used by optimizers.</p> |
| 8450 | |
| 8451 | <h5>Arguments:</h5> |
| 8452 | <p>The <tt>llvm.expect</tt> intrinsic takes two arguments. The first |
| 8453 | argument is a value. The second argument is an expected value, this needs to |
| 8454 | be a constant value, variables are not allowed.</p> |
| 8455 | |
| 8456 | <h5>Semantics:</h5> |
| 8457 | <p>This intrinsic is lowered to the <tt>val</tt>.</p> |
| NAKAMURA Takumi | f5af6ad | 2011-04-23 00:30:22 +0000 | [diff] [blame] | 8458 | </div> |
| 8459 | |
| 8460 | </div> |
| 8461 | |
| Jakub Staszak | b170e2d | 2011-12-04 18:29:26 +0000 | [diff] [blame] | 8462 | </div> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 8463 | <!-- *********************************************************************** --> |
| Chris Lattner | 0095054 | 2001-06-06 20:29:01 +0000 | [diff] [blame] | 8464 | <hr> |
| Misha Brukman | daa4cb0 | 2004-03-01 17:47:27 +0000 | [diff] [blame] | 8465 | <address> |
| 8466 | <a href="http://jigsaw.w3.org/css-validator/check/referer"><img |
| Misha Brukman | 4440870 | 2008-12-11 17:34:48 +0000 | [diff] [blame] | 8467 | src="http://jigsaw.w3.org/css-validator/images/vcss-blue" alt="Valid CSS"></a> |
| Misha Brukman | daa4cb0 | 2004-03-01 17:47:27 +0000 | [diff] [blame] | 8468 | <a href="http://validator.w3.org/check/referer"><img |
| Misha Brukman | 4440870 | 2008-12-11 17:34:48 +0000 | [diff] [blame] | 8469 | src="http://www.w3.org/Icons/valid-html401-blue" alt="Valid HTML 4.01"></a> |
| Misha Brukman | daa4cb0 | 2004-03-01 17:47:27 +0000 | [diff] [blame] | 8470 | |
| 8471 | <a href="mailto:sabre@nondot.org">Chris Lattner</a><br> |
| NAKAMURA Takumi | b9a3363 | 2011-04-09 02:13:37 +0000 | [diff] [blame] | 8472 | <a href="http://llvm.org/">The LLVM Compiler Infrastructure</a><br> |
| Misha Brukman | daa4cb0 | 2004-03-01 17:47:27 +0000 | [diff] [blame] | 8473 | Last modified: $Date$ |
| 8474 | </address> |
| Chris Lattner | c7d3ab3 | 2008-01-04 04:33:49 +0000 | [diff] [blame] | 8475 | |
| Misha Brukman | 9d0919f | 2003-11-08 01:05:38 +0000 | [diff] [blame] | 8476 | </body> |
| 8477 | </html> |