Gitiles
Code Review Sign In
gerrit-public.fairphone.software / fp2-dev / platform / external / llvm / 03bbad64103133311920b6105339acd79058f6a7 / . / docs / LangRef.html
blob: f70b36360e290fd5e1f851a52509f81652a3c08c [file] [log] [blame]
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
2 "http://www.w3.org/TR/html4/strict.dtd">
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3<html>
4<head>
5 <title>LLVM Assembly Language Reference Manual</title>
Reid Spencer3921c742004-08-26 20:44:00 +0000[diff] [blame]6 <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
7 <meta name="author" content="Chris Lattner">
8 <meta name="description"
9 content="LLVM Assembly Language Reference Manual.">
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]10 <link rel="stylesheet" href="llvm.css" type="text/css">
11</head>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]12
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]13<body>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]14
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]15<div class="doc_title"> LLVM Language Reference Manual </div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]16<ol>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]17 <li><a href="#abstract">Abstract</a></li>
18 <li><a href="#introduction">Introduction</a></li>
19 <li><a href="#identifiers">Identifiers</a></li>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]20 <li><a href="#highlevel">High Level Structure</a>
21 <ol>
22 <li><a href="#modulestructure">Module Structure</a></li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]23 <li><a href="#linkage">Linkage Types</a></li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]24 <li><a href="#callingconv">Calling Conventions</a></li>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]25 <li><a href="#namedtypes">Named Types</a></li>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]26 <li><a href="#globalvars">Global Variables</a></li>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]27 <li><a href="#functionstructure">Functions</a></li>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]28 <li><a href="#aliasstructure">Aliases</a></li>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]29 <li><a href="#paramattrs">Parameter Attributes</a></li>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]30 <li><a href="#fnattrs">Function Attributes</a></li>
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame]31 <li><a href="#gc">Garbage Collector Names</a></li>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]32 <li><a href="#moduleasm">Module-Level Inline Assembly</a></li>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]33 <li><a href="#datalayout">Data Layout</a></li>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]34 </ol>
35 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]36 <li><a href="#typesystem">Type System</a>
37 <ol>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]38 <li><a href="#t_classifications">Type Classifications</a></li>
Robert Bocchino7b81c752006-02-17 21:18:08 +0000[diff] [blame]39 <li><a href="#t_primitive">Primitive Types</a>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]40 <ol>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]41 <li><a href="#t_floating">Floating Point Types</a></li>
42 <li><a href="#t_void">Void Type</a></li>
43 <li><a href="#t_label">Label Type</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]44 </ol>
45 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]46 <li><a href="#t_derived">Derived Types</a>
47 <ol>
Chris Lattnerb9488a62007-12-18 06:18:21 +0000[diff] [blame]48 <li><a href="#t_integer">Integer Type</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]49 <li><a href="#t_array">Array Type</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]50 <li><a href="#t_function">Function Type</a></li>
51 <li><a href="#t_pointer">Pointer Type</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]52 <li><a href="#t_struct">Structure Type</a></li>
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]53 <li><a href="#t_pstruct">Packed Structure Type</a></li>
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]54 <li><a href="#t_vector">Vector Type</a></li>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]55 <li><a href="#t_opaque">Opaque Type</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]56 </ol>
57 </li>
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]58 <li><a href="#t_uprefs">Type Up-references</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]59 </ol>
60 </li>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]61 <li><a href="#constants">Constants</a>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]62 <ol>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]63 <li><a href="#simpleconstants">Simple Constants</a></li>
64 <li><a href="#aggregateconstants">Aggregate Constants</a></li>
65 <li><a href="#globalconstants">Global Variable and Function Addresses</a></li>
66 <li><a href="#undefvalues">Undefined Values</a></li>
67 <li><a href="#constantexprs">Constant Expressions</a></li>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]68 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]69 </li>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]70 <li><a href="#othervalues">Other Values</a>
71 <ol>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]72 <li><a href="#inlineasm">Inline Assembler Expressions</a></li>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]73 </ol>
74 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]75 <li><a href="#instref">Instruction Reference</a>
76 <ol>
77 <li><a href="#terminators">Terminator Instructions</a>
78 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]79 <li><a href="#i_ret">'<tt>ret</tt>' Instruction</a></li>
80 <li><a href="#i_br">'<tt>br</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]81 <li><a href="#i_switch">'<tt>switch</tt>' Instruction</a></li>
82 <li><a href="#i_invoke">'<tt>invoke</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]83 <li><a href="#i_unwind">'<tt>unwind</tt>' Instruction</a></li>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]84 <li><a href="#i_unreachable">'<tt>unreachable</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]85 </ol>
86 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]87 <li><a href="#binaryops">Binary Operations</a>
88 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]89 <li><a href="#i_add">'<tt>add</tt>' Instruction</a></li>
90 <li><a href="#i_sub">'<tt>sub</tt>' Instruction</a></li>
91 <li><a href="#i_mul">'<tt>mul</tt>' Instruction</a></li>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]92 <li><a href="#i_udiv">'<tt>udiv</tt>' Instruction</a></li>
93 <li><a href="#i_sdiv">'<tt>sdiv</tt>' Instruction</a></li>
94 <li><a href="#i_fdiv">'<tt>fdiv</tt>' Instruction</a></li>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]95 <li><a href="#i_urem">'<tt>urem</tt>' Instruction</a></li>
96 <li><a href="#i_srem">'<tt>srem</tt>' Instruction</a></li>
97 <li><a href="#i_frem">'<tt>frem</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]98 </ol>
99 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]100 <li><a href="#bitwiseops">Bitwise Binary Operations</a>
101 <ol>
Reid Spencer8e11bf82007-02-02 13:57:07 +0000[diff] [blame]102 <li><a href="#i_shl">'<tt>shl</tt>' Instruction</a></li>
103 <li><a href="#i_lshr">'<tt>lshr</tt>' Instruction</a></li>
104 <li><a href="#i_ashr">'<tt>ashr</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]105 <li><a href="#i_and">'<tt>and</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]106 <li><a href="#i_or">'<tt>or</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]107 <li><a href="#i_xor">'<tt>xor</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]108 </ol>
109 </li>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]110 <li><a href="#vectorops">Vector Operations</a>
111 <ol>
112 <li><a href="#i_extractelement">'<tt>extractelement</tt>' Instruction</a></li>
113 <li><a href="#i_insertelement">'<tt>insertelement</tt>' Instruction</a></li>
114 <li><a href="#i_shufflevector">'<tt>shufflevector</tt>' Instruction</a></li>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]115 </ol>
116 </li>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]117 <li><a href="#aggregateops">Aggregate Operations</a>
118 <ol>
119 <li><a href="#i_extractvalue">'<tt>extractvalue</tt>' Instruction</a></li>
120 <li><a href="#i_insertvalue">'<tt>insertvalue</tt>' Instruction</a></li>
121 </ol>
122 </li>
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]123 <li><a href="#memoryops">Memory Access and Addressing Operations</a>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]124 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]125 <li><a href="#i_malloc">'<tt>malloc</tt>' Instruction</a></li>
126 <li><a href="#i_free">'<tt>free</tt>' Instruction</a></li>
127 <li><a href="#i_alloca">'<tt>alloca</tt>' Instruction</a></li>
Robert Bocchino7b81c752006-02-17 21:18:08 +0000[diff] [blame]128 <li><a href="#i_load">'<tt>load</tt>' Instruction</a></li>
129 <li><a href="#i_store">'<tt>store</tt>' Instruction</a></li>
130 <li><a href="#i_getelementptr">'<tt>getelementptr</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]131 </ol>
132 </li>
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]133 <li><a href="#convertops">Conversion Operations</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]134 <ol>
135 <li><a href="#i_trunc">'<tt>trunc .. to</tt>' Instruction</a></li>
136 <li><a href="#i_zext">'<tt>zext .. to</tt>' Instruction</a></li>
137 <li><a href="#i_sext">'<tt>sext .. to</tt>' Instruction</a></li>
138 <li><a href="#i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a></li>
139 <li><a href="#i_fpext">'<tt>fpext .. to</tt>' Instruction</a></li>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]140 <li><a href="#i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a></li>
141 <li><a href="#i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a></li>
142 <li><a href="#i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a></li>
143 <li><a href="#i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a></li>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]144 <li><a href="#i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a></li>
145 <li><a href="#i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a></li>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]146 <li><a href="#i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a></li>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]147 </ol>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]148 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]149 <li><a href="#otherops">Other Operations</a>
150 <ol>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]151 <li><a href="#i_icmp">'<tt>icmp</tt>' Instruction</a></li>
152 <li><a href="#i_fcmp">'<tt>fcmp</tt>' Instruction</a></li>
Nate Begemanac80ade2008-05-12 19:01:56 +0000[diff] [blame]153 <li><a href="#i_vicmp">'<tt>vicmp</tt>' Instruction</a></li>
154 <li><a href="#i_vfcmp">'<tt>vfcmp</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]155 <li><a href="#i_phi">'<tt>phi</tt>' Instruction</a></li>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]156 <li><a href="#i_select">'<tt>select</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]157 <li><a href="#i_call">'<tt>call</tt>' Instruction</a></li>
Chris Lattnerfb6977d2006-01-13 23:26:01 +0000[diff] [blame]158 <li><a href="#i_va_arg">'<tt>va_arg</tt>' Instruction</a></li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]159 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]160 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]161 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]162 </li>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]163 <li><a href="#intrinsics">Intrinsic Functions</a>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]164 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]165 <li><a href="#int_varargs">Variable Argument Handling Intrinsics</a>
166 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]167 <li><a href="#int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a></li>
168 <li><a href="#int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a></li>
169 <li><a href="#int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]170 </ol>
171 </li>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]172 <li><a href="#int_gc">Accurate Garbage Collection Intrinsics</a>
173 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]174 <li><a href="#int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a></li>
175 <li><a href="#int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a></li>
176 <li><a href="#int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a></li>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]177 </ol>
178 </li>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]179 <li><a href="#int_codegen">Code Generator Intrinsics</a>
180 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]181 <li><a href="#int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a></li>
182 <li><a href="#int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a></li>
183 <li><a href="#int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a></li>
184 <li><a href="#int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a></li>
185 <li><a href="#int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a></li>
186 <li><a href="#int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a></li>
187 <li><a href="#int_readcyclecounter"><tt>llvm.readcyclecounter</tt>' Intrinsic</a></li>
John Criswell7123e272004-04-09 16:43:20 +0000[diff] [blame]188 </ol>
189 </li>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]190 <li><a href="#int_libc">Standard C Library Intrinsics</a>
191 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]192 <li><a href="#int_memcpy">'<tt>llvm.memcpy.*</tt>' Intrinsic</a></li>
193 <li><a href="#int_memmove">'<tt>llvm.memmove.*</tt>' Intrinsic</a></li>
194 <li><a href="#int_memset">'<tt>llvm.memset.*</tt>' Intrinsic</a></li>
195 <li><a href="#int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a></li>
196 <li><a href="#int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a></li>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]197 <li><a href="#int_sin">'<tt>llvm.sin.*</tt>' Intrinsic</a></li>
198 <li><a href="#int_cos">'<tt>llvm.cos.*</tt>' Intrinsic</a></li>
199 <li><a href="#int_pow">'<tt>llvm.pow.*</tt>' Intrinsic</a></li>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]200 </ol>
201 </li>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]202 <li><a href="#int_manip">Bit Manipulation Intrinsics</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]203 <ol>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]204 <li><a href="#int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a></li>
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]205 <li><a href="#int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic </a></li>
206 <li><a href="#int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic </a></li>
207 <li><a href="#int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic </a></li>
Reid Spencerf86037f2007-04-11 23:23:49 +0000[diff] [blame]208 <li><a href="#int_part_select">'<tt>llvm.part.select.*</tt>' Intrinsic </a></li>
209 <li><a href="#int_part_set">'<tt>llvm.part.set.*</tt>' Intrinsic </a></li>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]210 </ol>
211 </li>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]212 <li><a href="#int_overflow">Arithmetic with Overflow Intrinsics</a>
213 <ol>
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]214 <li><a href="#int_sadd_overflow">'<tt>llvm.sadd.with.overflow.*</tt> Intrinsics</a></li>
215 <li><a href="#int_uadd_overflow">'<tt>llvm.uadd.with.overflow.*</tt> Intrinsics</a></li>
216 <li><a href="#int_ssub_overflow">'<tt>llvm.ssub.with.overflow.*</tt> Intrinsics</a></li>
217 <li><a href="#int_usub_overflow">'<tt>llvm.usub.with.overflow.*</tt> Intrinsics</a></li>
218 <li><a href="#int_smul_overflow">'<tt>llvm.smul.with.overflow.*</tt> Intrinsics</a></li>
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]219 <li><a href="#int_umul_overflow">'<tt>llvm.umul.with.overflow.*</tt> Intrinsics</a></li>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]220 </ol>
221 </li>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]222 <li><a href="#int_debugger">Debugger intrinsics</a></li>
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]223 <li><a href="#int_eh">Exception Handling intrinsics</a></li>
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]224 <li><a href="#int_trampoline">Trampoline Intrinsic</a>
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]225 <ol>
226 <li><a href="#int_it">'<tt>llvm.init.trampoline</tt>' Intrinsic</a></li>
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]227 </ol>
228 </li>
Bill Wendling3c44f5b2008-11-18 22:10:53 +0000[diff] [blame]229 <li><a href="#int_atomics">Atomic intrinsics</a>
230 <ol>
231 <li><a href="#int_memory_barrier"><tt>llvm.memory_barrier</tt></a></li>
232 <li><a href="#int_atomic_cmp_swap"><tt>llvm.atomic.cmp.swap</tt></a></li>
233 <li><a href="#int_atomic_swap"><tt>llvm.atomic.swap</tt></a></li>
234 <li><a href="#int_atomic_load_add"><tt>llvm.atomic.load.add</tt></a></li>
235 <li><a href="#int_atomic_load_sub"><tt>llvm.atomic.load.sub</tt></a></li>
236 <li><a href="#int_atomic_load_and"><tt>llvm.atomic.load.and</tt></a></li>
237 <li><a href="#int_atomic_load_nand"><tt>llvm.atomic.load.nand</tt></a></li>
238 <li><a href="#int_atomic_load_or"><tt>llvm.atomic.load.or</tt></a></li>
239 <li><a href="#int_atomic_load_xor"><tt>llvm.atomic.load.xor</tt></a></li>
240 <li><a href="#int_atomic_load_max"><tt>llvm.atomic.load.max</tt></a></li>
241 <li><a href="#int_atomic_load_min"><tt>llvm.atomic.load.min</tt></a></li>
242 <li><a href="#int_atomic_load_umax"><tt>llvm.atomic.load.umax</tt></a></li>
243 <li><a href="#int_atomic_load_umin"><tt>llvm.atomic.load.umin</tt></a></li>
244 </ol>
245 </li>
Reid Spencer20677642007-07-20 19:59:11 +0000[diff] [blame]246 <li><a href="#int_general">General intrinsics</a>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]247 <ol>
Reid Spencer20677642007-07-20 19:59:11 +0000[diff] [blame]248 <li><a href="#int_var_annotation">
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]249 '<tt>llvm.var.annotation</tt>' Intrinsic</a></li>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]250 <li><a href="#int_annotation">
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]251 '<tt>llvm.annotation.*</tt>' Intrinsic</a></li>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]252 <li><a href="#int_trap">
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]253 '<tt>llvm.trap</tt>' Intrinsic</a></li>
254 <li><a href="#int_stackprotector">
255 '<tt>llvm.stackprotector</tt>' Intrinsic</a></li>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]256 </ol>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]257 </li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]258 </ol>
259 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]260</ol>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]261
262<div class="doc_author">
263 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>
264 and <a href="mailto:vadve@cs.uiuc.edu">Vikram Adve</a></p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]265</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]266
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]267<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]268<div class="doc_section"> <a name="abstract">Abstract </a></div>
269<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]270
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]271<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]272<p>This document is a reference manual for the LLVM assembly language.
Bill Wendling837f39b2008-08-05 22:29:16 +0000[diff] [blame]273LLVM is a Static Single Assignment (SSA) based representation that provides
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]274type safety, low-level operations, flexibility, and the capability of
275representing 'all' high-level languages cleanly. It is the common code
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]276representation used throughout all phases of the LLVM compilation
277strategy.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]278</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]279
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]280<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]281<div class="doc_section"> <a name="introduction">Introduction</a> </div>
282<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]283
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]284<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]285
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]286<p>The LLVM code representation is designed to be used in three
Gabor Greif04367bf2007-07-06 22:07:22 +0000[diff] [blame]287different forms: as an in-memory compiler IR, as an on-disk bitcode
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]288representation (suitable for fast loading by a Just-In-Time compiler),
289and as a human readable assembly language representation. This allows
290LLVM to provide a powerful intermediate representation for efficient
291compiler transformations and analysis, while providing a natural means
292to debug and visualize the transformations. The three different forms
293of LLVM are all equivalent. This document describes the human readable
294representation and notation.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]295
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]296<p>The LLVM representation aims to be light-weight and low-level
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]297while being expressive, typed, and extensible at the same time. It
298aims to be a "universal IR" of sorts, by being at a low enough level
299that high-level ideas may be cleanly mapped to it (similar to how
300microprocessors are "universal IR's", allowing many source languages to
301be mapped to them). By providing type information, LLVM can be used as
302the target of optimizations: for example, through pointer analysis, it
303can be proven that a C automatic variable is never accessed outside of
304the current function... allowing it to be promoted to a simple SSA
305value instead of a memory location.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]306
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]307</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]308
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]309<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]310<div class="doc_subsubsection"> <a name="wellformed">Well-Formedness</a> </div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]311
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]312<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]313
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]314<p>It is important to note that this document describes 'well formed'
315LLVM assembly language. There is a difference between what the parser
316accepts and what is considered 'well formed'. For example, the
317following instruction is syntactically okay, but not well formed:</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]318
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]319<div class="doc_code">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]320<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]321%x = <a href="#i_add">add</a> i32 1, %x
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]322</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]323</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]324
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]325<p>...because the definition of <tt>%x</tt> does not dominate all of
326its uses. The LLVM infrastructure provides a verification pass that may
327be used to verify that an LLVM module is well formed. This pass is
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]328automatically run by the parser after parsing input assembly and by
Gabor Greif04367bf2007-07-06 22:07:22 +0000[diff] [blame]329the optimizer before it outputs bitcode. The violations pointed out
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]330by the verifier pass indicate bugs in transformation passes or input to
331the parser.</p>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]332</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]333
Chris Lattnercc689392007-10-03 17:34:29 +0000[diff] [blame]334<!-- Describe the typesetting conventions here. -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]335
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]336<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]337<div class="doc_section"> <a name="identifiers">Identifiers</a> </div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]338<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]339
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]340<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]341
Reid Spencer2c452282007-08-07 14:34:28 +0000[diff] [blame]342 <p>LLVM identifiers come in two basic types: global and local. Global
343 identifiers (functions, global variables) begin with the @ character. Local
344 identifiers (register names, types) begin with the % character. Additionally,
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]345 there are three different formats for identifiers, for different purposes:</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]346
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]347<ol>
Reid Spencer2c452282007-08-07 14:34:28 +0000[diff] [blame]348 <li>Named values are represented as a string of characters with their prefix.
349 For example, %foo, @DivisionByZero, %a.really.long.identifier. The actual
350 regular expression used is '<tt>[%@][a-zA-Z$._][a-zA-Z$._0-9]*</tt>'.
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]351 Identifiers which require other characters in their names can be surrounded
Daniel Dunbar76dea952008-10-14 23:51:43 +0000[diff] [blame]352 with quotes. Special characters may be escaped using "\xx" where xx is the
353 ASCII code for the character in hexadecimal. In this way, any character can
354 be used in a name value, even quotes themselves.
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]355
Reid Spencer2c452282007-08-07 14:34:28 +0000[diff] [blame]356 <li>Unnamed values are represented as an unsigned numeric value with their
357 prefix. For example, %12, @2, %44.</li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]358
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]359 <li>Constants, which are described in a <a href="#constants">section about
360 constants</a>, below.</li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]361</ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]362
Reid Spencer2c452282007-08-07 14:34:28 +0000[diff] [blame]363<p>LLVM requires that values start with a prefix for two reasons: Compilers
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]364don't need to worry about name clashes with reserved words, and the set of
365reserved words may be expanded in the future without penalty. Additionally,
366unnamed identifiers allow a compiler to quickly come up with a temporary
367variable without having to avoid symbol table conflicts.</p>
368
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]369<p>Reserved words in LLVM are very similar to reserved words in other
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]370languages. There are keywords for different opcodes
371('<tt><a href="#i_add">add</a></tt>',
372 '<tt><a href="#i_bitcast">bitcast</a></tt>',
373 '<tt><a href="#i_ret">ret</a></tt>', etc...), for primitive type names ('<tt><a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]374href="#t_void">void</a></tt>', '<tt><a href="#t_primitive">i32</a></tt>', etc...),
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]375and others. These reserved words cannot conflict with variable names, because
Reid Spencer2c452282007-08-07 14:34:28 +0000[diff] [blame]376none of them start with a prefix character ('%' or '@').</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]377
378<p>Here is an example of LLVM code to multiply the integer variable
379'<tt>%X</tt>' by 8:</p>
380
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]381<p>The easy way:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]382
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]383<div class="doc_code">
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]384<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]385%result = <a href="#i_mul">mul</a> i32 %X, 8
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]386</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]387</div>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]388
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]389<p>After strength reduction:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]390
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]391<div class="doc_code">
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]392<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]393%result = <a href="#i_shl">shl</a> i32 %X, i8 3
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]394</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]395</div>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]396
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]397<p>And the hard way:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]398
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]399<div class="doc_code">
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]400<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]401<a href="#i_add">add</a> i32 %X, %X <i>; yields {i32}:%0</i>
402<a href="#i_add">add</a> i32 %0, %0 <i>; yields {i32}:%1</i>
403%result = <a href="#i_add">add</a> i32 %1, %1
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]404</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]405</div>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]406
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]407<p>This last way of multiplying <tt>%X</tt> by 8 illustrates several
408important lexical features of LLVM:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]409
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]410<ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]411
412 <li>Comments are delimited with a '<tt>;</tt>' and go until the end of
413 line.</li>
414
415 <li>Unnamed temporaries are created when the result of a computation is not
416 assigned to a named value.</li>
417
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]418 <li>Unnamed temporaries are numbered sequentially</li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]419
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]420</ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]421
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]422<p>...and it also shows a convention that we follow in this document. When
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]423demonstrating instructions, we will follow an instruction with a comment that
424defines the type and name of value produced. Comments are shown in italic
425text.</p>
426
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]427</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]428
429<!-- *********************************************************************** -->
430<div class="doc_section"> <a name="highlevel">High Level Structure</a> </div>
431<!-- *********************************************************************** -->
432
433<!-- ======================================================================= -->
434<div class="doc_subsection"> <a name="modulestructure">Module Structure</a>
435</div>
436
437<div class="doc_text">
438
439<p>LLVM programs are composed of "Module"s, each of which is a
440translation unit of the input programs. Each module consists of
441functions, global variables, and symbol table entries. Modules may be
442combined together with the LLVM linker, which merges function (and
443global variable) definitions, resolves forward declarations, and merges
444symbol table entries. Here is an example of the "hello world" module:</p>
445
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]446<div class="doc_code">
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]447<pre><i>; Declare the string constant as a global constant...</i>
Chris Lattnera89e5f12007-06-12 17:00:26 +0000[diff] [blame]448<a href="#identifiers">@.LC0</a> = <a href="#linkage_internal">internal</a> <a
449 href="#globalvars">constant</a> <a href="#t_array">[13 x i8]</a> c"hello world\0A\00" <i>; [13 x i8]*</i>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]450
451<i>; External declaration of the puts function</i>
Chris Lattnera89e5f12007-06-12 17:00:26 +0000[diff] [blame]452<a href="#functionstructure">declare</a> i32 @puts(i8 *) <i>; i32(i8 *)* </i>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]453
454<i>; Definition of main function</i>
Chris Lattnera89e5f12007-06-12 17:00:26 +0000[diff] [blame]455define i32 @main() { <i>; i32()* </i>
Dan Gohman2a08c532009-01-04 23:44:43 +0000[diff] [blame]456 <i>; Convert [13 x i8]* to i8 *...</i>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]457 %cast210 = <a
Dan Gohman2a08c532009-01-04 23:44:43 +0000[diff] [blame]458 href="#i_getelementptr">getelementptr</a> [13 x i8]* @.LC0, i64 0, i64 0 <i>; i8 *</i>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]459
460 <i>; Call puts function to write out the string to stdout...</i>
461 <a
Chris Lattnera89e5f12007-06-12 17:00:26 +0000[diff] [blame]462 href="#i_call">call</a> i32 @puts(i8 * %cast210) <i>; i32</i>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]463 <a
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]464 href="#i_ret">ret</a> i32 0<br>}<br>
465</pre>
466</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]467
468<p>This example is made up of a <a href="#globalvars">global variable</a>
469named "<tt>.LC0</tt>", an external declaration of the "<tt>puts</tt>"
470function, and a <a href="#functionstructure">function definition</a>
471for "<tt>main</tt>".</p>
472
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]473<p>In general, a module is made up of a list of global values,
474where both functions and global variables are global values. Global values are
475represented by a pointer to a memory location (in this case, a pointer to an
476array of char, and a pointer to a function), and have one of the following <a
477href="#linkage">linkage types</a>.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]478
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]479</div>
480
481<!-- ======================================================================= -->
482<div class="doc_subsection">
483 <a name="linkage">Linkage Types</a>
484</div>
485
486<div class="doc_text">
487
488<p>
489All Global Variables and Functions have one of the following types of linkage:
490</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]491
492<dl>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]493
Rafael Espindolabb46f522009-01-15 20:18:42 +0000[diff] [blame]494 <dt><tt><b><a name="linkage_private">private</a></b></tt>: </dt>
495
496 <dd>Global values with private linkage are only directly accessible by
497 objects in the current module. In particular, linking code into a module with
498 an private global value may cause the private to be renamed as necessary to
499 avoid collisions. Because the symbol is private to the module, all
500 references can be updated. This doesn't show up in any symbol table in the
501 object file.
502 </dd>
503
Dale Johannesen2307a7f2008-05-23 23:13:41 +0000[diff] [blame]504 <dt><tt><b><a name="linkage_internal">internal</a></b></tt>: </dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]505
Duncan Sands81d05c22009-01-16 09:29:46 +0000[diff] [blame]506 <dd> Similar to private, but the value shows as a local symbol (STB_LOCAL in
Rafael Espindolabb46f522009-01-15 20:18:42 +0000[diff] [blame]507 the case of ELF) in the object file. This corresponds to the notion of the
Chris Lattner4887bd82007-01-14 06:51:48 +0000[diff] [blame]508 '<tt>static</tt>' keyword in C.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]509 </dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]510
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]511 <dt><tt><b><a name="linkage_linkonce">linkonce</a></b></tt>: </dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]512
Chris Lattner4887bd82007-01-14 06:51:48 +0000[diff] [blame]513 <dd>Globals with "<tt>linkonce</tt>" linkage are merged with other globals of
514 the same name when linkage occurs. This is typically used to implement
515 inline functions, templates, or other code which must be generated in each
516 translation unit that uses it. Unreferenced <tt>linkonce</tt> globals are
517 allowed to be discarded.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]518 </dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]519
Dale Johannesen2307a7f2008-05-23 23:13:41 +0000[diff] [blame]520 <dt><tt><b><a name="linkage_common">common</a></b></tt>: </dt>
521
522 <dd>"<tt>common</tt>" linkage is exactly the same as <tt>linkonce</tt>
523 linkage, except that unreferenced <tt>common</tt> globals may not be
524 discarded. This is used for globals that may be emitted in multiple
525 translation units, but that are not guaranteed to be emitted into every
526 translation unit that uses them. One example of this is tentative
527 definitions in C, such as "<tt>int X;</tt>" at global scope.
528 </dd>
529
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]530 <dt><tt><b><a name="linkage_weak">weak</a></b></tt>: </dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]531
Dale Johannesen2307a7f2008-05-23 23:13:41 +0000[diff] [blame]532 <dd>"<tt>weak</tt>" linkage is the same as <tt>common</tt> linkage, except
533 that some targets may choose to emit different assembly sequences for them
534 for target-dependent reasons. This is used for globals that are declared
535 "weak" in C source code.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]536 </dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]537
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]538 <dt><tt><b><a name="linkage_appending">appending</a></b></tt>: </dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]539
540 <dd>"<tt>appending</tt>" linkage may only be applied to global variables of
541 pointer to array type. When two global variables with appending linkage are
542 linked together, the two global arrays are appended together. This is the
543 LLVM, typesafe, equivalent of having the system linker append together
544 "sections" with identical names when .o files are linked.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]545 </dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]546
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame]547 <dt><tt><b><a name="linkage_externweak">extern_weak</a></b></tt>: </dt>
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]548 <dd>The semantics of this linkage follow the ELF object file model: the
549 symbol is weak until linked, if not linked, the symbol becomes null instead
550 of being an undefined reference.
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame]551 </dd>
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame]552
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]553 <dt><tt><b><a name="linkage_external">externally visible</a></b></tt>:</dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]554
555 <dd>If none of the above identifiers are used, the global is externally
556 visible, meaning that it participates in linkage and can be used to resolve
557 external symbol references.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]558 </dd>
Reid Spencerc8910842007-04-11 23:49:50 +0000[diff] [blame]559</dl>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]560
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]561 <p>
562 The next two types of linkage are targeted for Microsoft Windows platform
563 only. They are designed to support importing (exporting) symbols from (to)
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]564 DLLs (Dynamic Link Libraries).
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]565 </p>
566
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame]567 <dl>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]568 <dt><tt><b><a name="linkage_dllimport">dllimport</a></b></tt>: </dt>
569
570 <dd>"<tt>dllimport</tt>" linkage causes the compiler to reference a function
571 or variable via a global pointer to a pointer that is set up by the DLL
572 exporting the symbol. On Microsoft Windows targets, the pointer name is
Dan Gohman79564122009-01-12 21:35:55 +0000[diff] [blame]573 formed by combining <code>__imp_</code> and the function or variable name.
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]574 </dd>
575
576 <dt><tt><b><a name="linkage_dllexport">dllexport</a></b></tt>: </dt>
577
578 <dd>"<tt>dllexport</tt>" linkage causes the compiler to provide a global
579 pointer to a pointer in a DLL, so that it can be referenced with the
580 <tt>dllimport</tt> attribute. On Microsoft Windows targets, the pointer
Dan Gohman79564122009-01-12 21:35:55 +0000[diff] [blame]581 name is formed by combining <code>__imp_</code> and the function or variable
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]582 name.
583 </dd>
584
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]585</dl>
586
Dan Gohmanf0032762008-11-24 17:18:39 +0000[diff] [blame]587<p>For example, since the "<tt>.LC0</tt>"
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]588variable is defined to be internal, if another module defined a "<tt>.LC0</tt>"
589variable and was linked with this one, one of the two would be renamed,
590preventing a collision. Since "<tt>main</tt>" and "<tt>puts</tt>" are
591external (i.e., lacking any linkage declarations), they are accessible
Reid Spencerac8d2762007-01-05 00:59:10 +0000[diff] [blame]592outside of the current module.</p>
593<p>It is illegal for a function <i>declaration</i>
594to have any linkage type other than "externally visible", <tt>dllimport</tt>,
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame]595or <tt>extern_weak</tt>.</p>
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]596<p>Aliases can have only <tt>external</tt>, <tt>internal</tt> and <tt>weak</tt>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]597linkages.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]598</div>
599
600<!-- ======================================================================= -->
601<div class="doc_subsection">
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]602 <a name="callingconv">Calling Conventions</a>
603</div>
604
605<div class="doc_text">
606
607<p>LLVM <a href="#functionstructure">functions</a>, <a href="#i_call">calls</a>
608and <a href="#i_invoke">invokes</a> can all have an optional calling convention
609specified for the call. The calling convention of any pair of dynamic
610caller/callee must match, or the behavior of the program is undefined. The
611following calling conventions are supported by LLVM, and more may be added in
612the future:</p>
613
614<dl>
615 <dt><b>"<tt>ccc</tt>" - The C calling convention</b>:</dt>
616
617 <dd>This calling convention (the default if no other calling convention is
618 specified) matches the target C calling conventions. This calling convention
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]619 supports varargs function calls and tolerates some mismatch in the declared
Reid Spencerc28d2bc2006-12-31 21:30:18 +0000[diff] [blame]620 prototype and implemented declaration of the function (as does normal C).
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]621 </dd>
622
623 <dt><b>"<tt>fastcc</tt>" - The fast calling convention</b>:</dt>
624
625 <dd>This calling convention attempts to make calls as fast as possible
626 (e.g. by passing things in registers). This calling convention allows the
627 target to use whatever tricks it wants to produce fast code for the target,
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]628 without having to conform to an externally specified ABI (Application Binary
629 Interface). Implementations of this convention should allow arbitrary
Arnold Schwaighofer9097d142008-05-14 09:17:12 +0000[diff] [blame]630 <a href="CodeGenerator.html#tailcallopt">tail call optimization</a> to be
631 supported. This calling convention does not support varargs and requires the
632 prototype of all callees to exactly match the prototype of the function
633 definition.
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]634 </dd>
635
636 <dt><b>"<tt>coldcc</tt>" - The cold calling convention</b>:</dt>
637
638 <dd>This calling convention attempts to make code in the caller as efficient
639 as possible under the assumption that the call is not commonly executed. As
640 such, these calls often preserve all registers so that the call does not break
641 any live ranges in the caller side. This calling convention does not support
642 varargs and requires the prototype of all callees to exactly match the
643 prototype of the function definition.
644 </dd>
645
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]646 <dt><b>"<tt>cc &lt;<em>n</em>&gt;</tt>" - Numbered convention</b>:</dt>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]647
648 <dd>Any calling convention may be specified by number, allowing
649 target-specific calling conventions to be used. Target specific calling
650 conventions start at 64.
651 </dd>
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]652</dl>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]653
654<p>More calling conventions can be added/defined on an as-needed basis, to
655support pascal conventions or any other well-known target-independent
656convention.</p>
657
658</div>
659
660<!-- ======================================================================= -->
661<div class="doc_subsection">
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]662 <a name="visibility">Visibility Styles</a>
663</div>
664
665<div class="doc_text">
666
667<p>
668All Global Variables and Functions have one of the following visibility styles:
669</p>
670
671<dl>
672 <dt><b>"<tt>default</tt>" - Default style</b>:</dt>
673
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]674 <dd>On targets that use the ELF object file format, default visibility means
675 that the declaration is visible to other
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]676 modules and, in shared libraries, means that the declared entity may be
677 overridden. On Darwin, default visibility means that the declaration is
678 visible to other modules. Default visibility corresponds to "external
679 linkage" in the language.
680 </dd>
681
682 <dt><b>"<tt>hidden</tt>" - Hidden style</b>:</dt>
683
684 <dd>Two declarations of an object with hidden visibility refer to the same
685 object if they are in the same shared object. Usually, hidden visibility
686 indicates that the symbol will not be placed into the dynamic symbol table,
687 so no other module (executable or shared library) can reference it
688 directly.
689 </dd>
690
Anton Korobeynikov6f9896f2007-04-29 18:35:00 +0000[diff] [blame]691 <dt><b>"<tt>protected</tt>" - Protected style</b>:</dt>
692
693 <dd>On ELF, protected visibility indicates that the symbol will be placed in
694 the dynamic symbol table, but that references within the defining module will
695 bind to the local symbol. That is, the symbol cannot be overridden by another
696 module.
697 </dd>
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]698</dl>
699
700</div>
701
702<!-- ======================================================================= -->
703<div class="doc_subsection">
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]704 <a name="namedtypes">Named Types</a>
705</div>
706
707<div class="doc_text">
708
709<p>LLVM IR allows you to specify name aliases for certain types. This can make
710it easier to read the IR and make the IR more condensed (particularly when
711recursive types are involved). An example of a name specification is:
712</p>
713
714<div class="doc_code">
715<pre>
716%mytype = type { %mytype*, i32 }
717</pre>
718</div>
719
720<p>You may give a name to any <a href="#typesystem">type</a> except "<a
721href="t_void">void</a>". Type name aliases may be used anywhere a type is
722expected with the syntax "%mytype".</p>
723
724<p>Note that type names are aliases for the structural type that they indicate,
725and that you can therefore specify multiple names for the same type. This often
726leads to confusing behavior when dumping out a .ll file. Since LLVM IR uses
727structural typing, the name is not part of the type. When printing out LLVM IR,
728the printer will pick <em>one name</em> to render all types of a particular
729shape. This means that if you have code where two different source types end up
730having the same LLVM type, that the dumper will sometimes print the "wrong" or
731unexpected type. This is an important design point and isn't going to
732change.</p>
733
734</div>
735
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]736<!-- ======================================================================= -->
737<div class="doc_subsection">
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]738 <a name="globalvars">Global Variables</a>
739</div>
740
741<div class="doc_text">
742
Chris Lattner3689a342005-02-12 19:30:21 +0000[diff] [blame]743<p>Global variables define regions of memory allocated at compilation time
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]744instead of run-time. Global variables may optionally be initialized, may have
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]745an explicit section to be placed in, and may have an optional explicit alignment
746specified. A variable may be defined as "thread_local", which means that it
747will not be shared by threads (each thread will have a separated copy of the
748variable). A variable may be defined as a global "constant," which indicates
749that the contents of the variable will <b>never</b> be modified (enabling better
Chris Lattner3689a342005-02-12 19:30:21 +0000[diff] [blame]750optimization, allowing the global data to be placed in the read-only section of
751an executable, etc). Note that variables that need runtime initialization
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]752cannot be marked "constant" as there is a store to the variable.</p>
Chris Lattner3689a342005-02-12 19:30:21 +0000[diff] [blame]753
754<p>
755LLVM explicitly allows <em>declarations</em> of global variables to be marked
756constant, even if the final definition of the global is not. This capability
757can be used to enable slightly better optimization of the program, but requires
758the language definition to guarantee that optimizations based on the
759'constantness' are valid for the translation units that do not include the
760definition.
761</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]762
763<p>As SSA values, global variables define pointer values that are in
764scope (i.e. they dominate) all basic blocks in the program. Global
765variables always define a pointer to their "content" type because they
766describe a region of memory, and all memory objects in LLVM are
767accessed through pointers.</p>
768
Christopher Lamb284d9922007-12-11 09:31:00 +0000[diff] [blame]769<p>A global variable may be declared to reside in a target-specifc numbered
770address space. For targets that support them, address spaces may affect how
771optimizations are performed and/or what target instructions are used to access
Christopher Lambd49e18d2007-12-12 08:44:39 +0000[diff] [blame]772the variable. The default address space is zero. The address space qualifier
773must precede any other attributes.</p>
Christopher Lamb284d9922007-12-11 09:31:00 +0000[diff] [blame]774
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]775<p>LLVM allows an explicit section to be specified for globals. If the target
776supports it, it will emit globals to the section specified.</p>
777
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]778<p>An explicit alignment may be specified for a global. If not present, or if
779the alignment is set to zero, the alignment of the global is set by the target
780to whatever it feels convenient. If an explicit alignment is specified, the
781global is forced to have at least that much alignment. All alignments must be
782a power of 2.</p>
783
Christopher Lamb284d9922007-12-11 09:31:00 +0000[diff] [blame]784<p>For example, the following defines a global in a numbered address space with
785an initializer, section, and alignment:</p>
Chris Lattner68027ea2007-01-14 00:27:09 +0000[diff] [blame]786
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]787<div class="doc_code">
Chris Lattner68027ea2007-01-14 00:27:09 +0000[diff] [blame]788<pre>
Dan Gohman398873c2009-01-11 00:40:00 +0000[diff] [blame]789@G = addrspace(5) constant float 1.0, section "foo", align 4
Chris Lattner68027ea2007-01-14 00:27:09 +0000[diff] [blame]790</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]791</div>
Chris Lattner68027ea2007-01-14 00:27:09 +0000[diff] [blame]792
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]793</div>
794
795
796<!-- ======================================================================= -->
797<div class="doc_subsection">
798 <a name="functionstructure">Functions</a>
799</div>
800
801<div class="doc_text">
802
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]803<p>LLVM function definitions consist of the "<tt>define</tt>" keyord,
804an optional <a href="#linkage">linkage type</a>, an optional
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]805<a href="#visibility">visibility style</a>, an optional
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]806<a href="#callingconv">calling convention</a>, a return type, an optional
807<a href="#paramattrs">parameter attribute</a> for the return type, a function
808name, a (possibly empty) argument list (each with optional
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]809<a href="#paramattrs">parameter attributes</a>), optional
810<a href="#fnattrs">function attributes</a>, an optional section,
811an optional alignment, an optional <a href="#gc">garbage collector name</a>,
Chris Lattner0c46a7d2008-10-04 18:10:21 +0000[diff] [blame]812an opening curly brace, a list of basic blocks, and a closing curly brace.
Anton Korobeynikov8cea37b2007-01-23 12:35:46 +0000[diff] [blame]813
814LLVM function declarations consist of the "<tt>declare</tt>" keyword, an
815optional <a href="#linkage">linkage type</a>, an optional
816<a href="#visibility">visibility style</a>, an optional
817<a href="#callingconv">calling convention</a>, a return type, an optional
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]818<a href="#paramattrs">parameter attribute</a> for the return type, a function
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame]819name, a possibly empty list of arguments, an optional alignment, and an optional
Gordon Henriksene754abe2007-12-10 03:30:21 +0000[diff] [blame]820<a href="#gc">garbage collector name</a>.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]821
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]822<p>A function definition contains a list of basic blocks, forming the CFG
823(Control Flow Graph) for
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]824the function. Each basic block may optionally start with a label (giving the
825basic block a symbol table entry), contains a list of instructions, and ends
826with a <a href="#terminators">terminator</a> instruction (such as a branch or
827function return).</p>
828
Chris Lattner4a3c9012007-06-08 16:52:14 +0000[diff] [blame]829<p>The first basic block in a function is special in two ways: it is immediately
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]830executed on entrance to the function, and it is not allowed to have predecessor
831basic blocks (i.e. there can not be any branches to the entry block of a
832function). Because the block can have no predecessors, it also cannot have any
833<a href="#i_phi">PHI nodes</a>.</p>
834
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]835<p>LLVM allows an explicit section to be specified for functions. If the target
836supports it, it will emit functions to the section specified.</p>
837
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]838<p>An explicit alignment may be specified for a function. If not present, or if
839the alignment is set to zero, the alignment of the function is set by the target
840to whatever it feels convenient. If an explicit alignment is specified, the
841function is forced to have at least that much alignment. All alignments must be
842a power of 2.</p>
843
Devang Patel307e8ab2008-10-07 17:48:33 +0000[diff] [blame]844 <h5>Syntax:</h5>
845
846<div class="doc_code">
Chris Lattner50ad45c2008-10-13 16:55:18 +0000[diff] [blame]847<tt>
848define [<a href="#linkage">linkage</a>] [<a href="#visibility">visibility</a>]
849 [<a href="#callingconv">cconv</a>] [<a href="#paramattrs">ret attrs</a>]
850 &lt;ResultType&gt; @&lt;FunctionName&gt; ([argument list])
851 [<a href="#fnattrs">fn Attrs</a>] [section "name"] [align N]
852 [<a href="#gc">gc</a>] { ... }
853</tt>
Devang Patel307e8ab2008-10-07 17:48:33 +0000[diff] [blame]854</div>
855
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]856</div>
857
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]858
859<!-- ======================================================================= -->
860<div class="doc_subsection">
861 <a name="aliasstructure">Aliases</a>
862</div>
863<div class="doc_text">
864 <p>Aliases act as "second name" for the aliasee value (which can be either
Anton Korobeynikov726d45c2008-03-22 08:36:14 +0000[diff] [blame]865 function, global variable, another alias or bitcast of global value). Aliases
866 may have an optional <a href="#linkage">linkage type</a>, and an
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]867 optional <a href="#visibility">visibility style</a>.</p>
868
869 <h5>Syntax:</h5>
870
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]871<div class="doc_code">
Bill Wendlingaac388b2007-05-29 09:42:13 +0000[diff] [blame]872<pre>
Duncan Sands0b23ac12008-09-12 20:48:21 +0000[diff] [blame]873@&lt;Name&gt; = alias [Linkage] [Visibility] &lt;AliaseeTy&gt; @&lt;Aliasee&gt;
Bill Wendlingaac388b2007-05-29 09:42:13 +0000[diff] [blame]874</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]875</div>
Anton Korobeynikov8b0a8c82007-04-25 14:27:10 +0000[diff] [blame]876
877</div>
878
879
880
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]881<!-- ======================================================================= -->
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]882<div class="doc_subsection"><a name="paramattrs">Parameter Attributes</a></div>
883<div class="doc_text">
884 <p>The return type and each parameter of a function type may have a set of
885 <i>parameter attributes</i> associated with them. Parameter attributes are
886 used to communicate additional information about the result or parameters of
Duncan Sandsdc024672007-11-27 13:23:08 +0000[diff] [blame]887 a function. Parameter attributes are considered to be part of the function,
888 not of the function type, so functions with different parameter attributes
889 can have the same function type.</p>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]890
Reid Spencer950e9f82007-01-15 18:27:39 +0000[diff] [blame]891 <p>Parameter attributes are simple keywords that follow the type specified. If
892 multiple parameter attributes are needed, they are space separated. For
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]893 example:</p>
894
895<div class="doc_code">
896<pre>
Nick Lewyckyb6a7d252009-02-15 23:06:14 +0000[diff] [blame]897declare i32 @printf(i8* noalias nocapture, ...)
Chris Lattner66d922c2008-10-04 18:33:34 +0000[diff] [blame]898declare i32 @atoi(i8 zeroext)
899declare signext i8 @returns_signed_char()
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]900</pre>
901</div>
902
Duncan Sandsdc024672007-11-27 13:23:08 +0000[diff] [blame]903 <p>Note that any attributes for the function result (<tt>nounwind</tt>,
904 <tt>readonly</tt>) come immediately after the argument list.</p>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]905
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame]906 <p>Currently, only the following parameter attributes are defined:</p>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]907 <dl>
Reid Spencer9445e9a2007-07-19 23:13:04 +0000[diff] [blame]908 <dt><tt>zeroext</tt></dt>
Chris Lattner66d922c2008-10-04 18:33:34 +0000[diff] [blame]909 <dd>This indicates to the code generator that the parameter or return value
910 should be zero-extended to a 32-bit value by the caller (for a parameter)
911 or the callee (for a return value).</dd>
Chris Lattner47507de2008-01-11 06:20:47 +0000[diff] [blame]912
Reid Spencer9445e9a2007-07-19 23:13:04 +0000[diff] [blame]913 <dt><tt>signext</tt></dt>
Chris Lattner66d922c2008-10-04 18:33:34 +0000[diff] [blame]914 <dd>This indicates to the code generator that the parameter or return value
915 should be sign-extended to a 32-bit value by the caller (for a parameter)
916 or the callee (for a return value).</dd>
Chris Lattner47507de2008-01-11 06:20:47 +0000[diff] [blame]917
Anton Korobeynikov34d6dea2007-01-28 14:30:45 +0000[diff] [blame]918 <dt><tt>inreg</tt></dt>
Dale Johannesenc9c6da62008-09-25 20:47:45 +0000[diff] [blame]919 <dd>This indicates that this parameter or return value should be treated
920 in a special target-dependent fashion during while emitting code for a
921 function call or return (usually, by putting it in a register as opposed
Chris Lattner66d922c2008-10-04 18:33:34 +0000[diff] [blame]922 to memory, though some targets use it to distinguish between two different
923 kinds of registers). Use of this attribute is target-specific.</dd>
Chris Lattner47507de2008-01-11 06:20:47 +0000[diff] [blame]924
Duncan Sandsedb05df2008-10-06 08:14:18 +0000[diff] [blame]925 <dt><tt><a name="byval">byval</a></tt></dt>
Chris Lattner0747baa2008-01-15 04:34:22 +0000[diff] [blame]926 <dd>This indicates that the pointer parameter should really be passed by
927 value to the function. The attribute implies that a hidden copy of the
928 pointee is made between the caller and the callee, so the callee is unable
Chris Lattnerebec6782008-08-05 18:21:08 +0000[diff] [blame]929 to modify the value in the callee. This attribute is only valid on LLVM
Chris Lattner0747baa2008-01-15 04:34:22 +0000[diff] [blame]930 pointer arguments. It is generally used to pass structs and arrays by
Duncan Sandsedb05df2008-10-06 08:14:18 +0000[diff] [blame]931 value, but is also valid on pointers to scalars. The copy is considered to
932 belong to the caller not the callee (for example,
933 <tt><a href="#readonly">readonly</a></tt> functions should not write to
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]934 <tt>byval</tt> parameters). This is not a valid attribute for return
Chris Lattnerce459b12009-02-05 05:42:28 +0000[diff] [blame]935 values. The byval attribute also supports specifying an alignment with the
936 align attribute. This has a target-specific effect on the code generator
937 that usually indicates a desired alignment for the synthesized stack
938 slot.</dd>
Chris Lattner47507de2008-01-11 06:20:47 +0000[diff] [blame]939
Anton Korobeynikov34d6dea2007-01-28 14:30:45 +0000[diff] [blame]940 <dt><tt>sret</tt></dt>
Duncan Sandse26dec62008-02-18 04:19:38 +0000[diff] [blame]941 <dd>This indicates that the pointer parameter specifies the address of a
942 structure that is the return value of the function in the source program.
Chris Lattner66d922c2008-10-04 18:33:34 +0000[diff] [blame]943 This pointer must be guaranteed by the caller to be valid: loads and stores
944 to the structure may be assumed by the callee to not to trap. This may only
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]945 be applied to the first parameter. This is not a valid attribute for
946 return values. </dd>
Chris Lattner47507de2008-01-11 06:20:47 +0000[diff] [blame]947
Zhou Shengfebca342007-06-05 05:28:26 +0000[diff] [blame]948 <dt><tt>noalias</tt></dt>
Nick Lewycky02ff3082008-11-24 03:41:24 +0000[diff] [blame]949 <dd>This indicates that the pointer does not alias any global or any other
950 parameter. The caller is responsible for ensuring that this is the
Nick Lewyckyb2b32fd2008-11-24 05:00:44 +0000[diff] [blame]951 case. On a function return value, <tt>noalias</tt> additionally indicates
952 that the pointer does not alias any other pointers visible to the
Nick Lewyckyf23d0d32008-12-19 06:39:12 +0000[diff] [blame]953 caller. For further details, please see the discussion of the NoAlias
954 response in
955 <a href="http://llvm.org/docs/AliasAnalysis.html#MustMayNo">alias
956 analysis</a>.</dd>
957
958 <dt><tt>nocapture</tt></dt>
959 <dd>This indicates that the callee does not make any copies of the pointer
960 that outlive the callee itself. This is not a valid attribute for return
961 values.</dd>
Chris Lattner47507de2008-01-11 06:20:47 +0000[diff] [blame]962
Duncan Sands50f19f52007-07-27 19:57:41 +0000[diff] [blame]963 <dt><tt>nest</tt></dt>
Duncan Sands0789b8b2008-07-08 09:27:25 +0000[diff] [blame]964 <dd>This indicates that the pointer parameter can be excised using the
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]965 <a href="#int_trampoline">trampoline intrinsics</a>. This is not a valid
966 attribute for return values.</dd>
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame]967 </dl>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]968
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]969</div>
970
971<!-- ======================================================================= -->
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]972<div class="doc_subsection">
Gordon Henriksen80a75bf2007-12-10 03:18:06 +0000[diff] [blame]973 <a name="gc">Garbage Collector Names</a>
974</div>
975
976<div class="doc_text">
977<p>Each function may specify a garbage collector name, which is simply a
978string.</p>
979
980<div class="doc_code"><pre
981>define void @f() gc "name" { ...</pre></div>
982
983<p>The compiler declares the supported values of <i>name</i>. Specifying a
984collector which will cause the compiler to alter its output in order to support
985the named garbage collection algorithm.</p>
986</div>
987
988<!-- ======================================================================= -->
989<div class="doc_subsection">
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]990 <a name="fnattrs">Function Attributes</a>
Devang Patelf8b94812008-09-04 23:05:13 +0000[diff] [blame]991</div>
992
993<div class="doc_text">
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]994
995<p>Function attributes are set to communicate additional information about
996 a function. Function attributes are considered to be part of the function,
997 not of the function type, so functions with different parameter attributes
998 can have the same function type.</p>
999
1000 <p>Function attributes are simple keywords that follow the type specified. If
1001 multiple attributes are needed, they are space separated. For
1002 example:</p>
Devang Patelf8b94812008-09-04 23:05:13 +0000[diff] [blame]1003
1004<div class="doc_code">
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1005<pre>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1006define void @f() noinline { ... }
1007define void @f() alwaysinline { ... }
1008define void @f() alwaysinline optsize { ... }
1009define void @f() optsize
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1010</pre>
Devang Patelf8b94812008-09-04 23:05:13 +0000[diff] [blame]1011</div>
1012
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1013<dl>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1014<dt><tt>alwaysinline</tt></dt>
Chris Lattner88d4b592008-10-04 18:23:17 +0000[diff] [blame]1015<dd>This attribute indicates that the inliner should attempt to inline this
1016function into callers whenever possible, ignoring any active inlining size
1017threshold for this caller.</dd>
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1018
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1019<dt><tt>noinline</tt></dt>
Chris Lattner88d4b592008-10-04 18:23:17 +0000[diff] [blame]1020<dd>This attribute indicates that the inliner should never inline this function
Chris Lattner94b5f7d2008-10-05 17:14:59 +0000[diff] [blame]1021in any situation. This attribute may not be used together with the
Chris Lattner88d4b592008-10-04 18:23:17 +0000[diff] [blame]1022<tt>alwaysinline</tt> attribute.</dd>
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1023
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1024<dt><tt>optsize</tt></dt>
Devang Patel66c6c652008-09-29 18:34:44 +0000[diff] [blame]1025<dd>This attribute suggests that optimization passes and code generator passes
Chris Lattner88d4b592008-10-04 18:23:17 +0000[diff] [blame]1026make choices that keep the code size of this function low, and otherwise do
1027optimizations specifically to reduce code size.</dd>
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1028
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1029<dt><tt>noreturn</tt></dt>
Chris Lattner88d4b592008-10-04 18:23:17 +0000[diff] [blame]1030<dd>This function attribute indicates that the function never returns normally.
1031This produces undefined behavior at runtime if the function ever does
1032dynamically return.</dd>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1033
1034<dt><tt>nounwind</tt></dt>
Chris Lattner88d4b592008-10-04 18:23:17 +0000[diff] [blame]1035<dd>This function attribute indicates that the function never returns with an
1036unwind or exceptional control flow. If the function does unwind, its runtime
1037behavior is undefined.</dd>
1038
1039<dt><tt>readnone</tt></dt>
Duncan Sandsedb05df2008-10-06 08:14:18 +0000[diff] [blame]1040<dd>This attribute indicates that the function computes its result (or the
1041exception it throws) based strictly on its arguments, without dereferencing any
1042pointer arguments or otherwise accessing any mutable state (e.g. memory, control
1043registers, etc) visible to caller functions. It does not write through any
1044pointer arguments (including <tt><a href="#byval">byval</a></tt> arguments) and
1045never changes any state visible to callers.</dd>
Devang Patel2c9c3e72008-09-26 23:51:19 +0000[diff] [blame]1046
Duncan Sandsedb05df2008-10-06 08:14:18 +0000[diff] [blame]1047<dt><tt><a name="readonly">readonly</a></tt></dt>
1048<dd>This attribute indicates that the function does not write through any
1049pointer arguments (including <tt><a href="#byval">byval</a></tt> arguments)
1050or otherwise modify any state (e.g. memory, control registers, etc) visible to
1051caller functions. It may dereference pointer arguments and read state that may
1052be set in the caller. A readonly function always returns the same value (or
1053throws the same exception) when called with the same set of arguments and global
1054state.</dd>
Bill Wendling31359ba2008-11-13 01:02:51 +0000[diff] [blame]1055
1056<dt><tt><a name="ssp">ssp</a></tt></dt>
Bill Wendlingbaa39d82008-11-26 19:19:05 +0000[diff] [blame]1057<dd>This attribute indicates that the function should emit a stack smashing
Bill Wendling31359ba2008-11-13 01:02:51 +0000[diff] [blame]1058protector. It is in the form of a "canary"&mdash;a random value placed on the
1059stack before the local variables that's checked upon return from the function to
1060see if it has been overwritten. A heuristic is used to determine if a function
Bill Wendlingbaa39d82008-11-26 19:19:05 +0000[diff] [blame]1061needs stack protectors or not.
Bill Wendling31359ba2008-11-13 01:02:51 +0000[diff] [blame]1062
Bill Wendlingfbaa7ed2008-11-26 19:07:40 +0000[diff] [blame]1063<p>If a function that has an <tt>ssp</tt> attribute is inlined into a function
1064that doesn't have an <tt>ssp</tt> attribute, then the resulting function will
1065have an <tt>ssp</tt> attribute.</p></dd>
1066
1067<dt><tt>sspreq</tt></dt>
Bill Wendlingbaa39d82008-11-26 19:19:05 +0000[diff] [blame]1068<dd>This attribute indicates that the function should <em>always</em> emit a
Bill Wendling31359ba2008-11-13 01:02:51 +0000[diff] [blame]1069stack smashing protector. This overrides the <tt><a href="#ssp">ssp</a></tt>
Bill Wendlingbaa39d82008-11-26 19:19:05 +0000[diff] [blame]1070function attribute.
Bill Wendlingfbaa7ed2008-11-26 19:07:40 +0000[diff] [blame]1071
1072<p>If a function that has an <tt>sspreq</tt> attribute is inlined into a
1073function that doesn't have an <tt>sspreq</tt> attribute or which has
1074an <tt>ssp</tt> attribute, then the resulting function will have
1075an <tt>sspreq</tt> attribute.</p></dd>
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1076</dl>
1077
Devang Patelf8b94812008-09-04 23:05:13 +0000[diff] [blame]1078</div>
1079
1080<!-- ======================================================================= -->
1081<div class="doc_subsection">
Chris Lattner1eeeb0c2006-04-08 04:40:53 +0000[diff] [blame]1082 <a name="moduleasm">Module-Level Inline Assembly</a>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]1083</div>
1084
1085<div class="doc_text">
1086<p>
1087Modules may contain "module-level inline asm" blocks, which corresponds to the
1088GCC "file scope inline asm" blocks. These blocks are internally concatenated by
1089LLVM and treated as a single unit, but may be separated in the .ll file if
1090desired. The syntax is very simple:
1091</p>
1092
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1093<div class="doc_code">
1094<pre>
1095module asm "inline asm code goes here"
1096module asm "more can go here"
1097</pre>
1098</div>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]1099
1100<p>The strings can contain any character by escaping non-printable characters.
1101 The escape sequence used is simply "\xx" where "xx" is the two digit hex code
1102 for the number.
1103</p>
1104
1105<p>
1106 The inline asm code is simply printed to the machine code .s file when
1107 assembly code is generated.
1108</p>
1109</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]1110
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1111<!-- ======================================================================= -->
1112<div class="doc_subsection">
1113 <a name="datalayout">Data Layout</a>
1114</div>
1115
1116<div class="doc_text">
1117<p>A module may specify a target specific data layout string that specifies how
Reid Spencerc8910842007-04-11 23:49:50 +0000[diff] [blame]1118data is to be laid out in memory. The syntax for the data layout is simply:</p>
1119<pre> target datalayout = "<i>layout specification</i>"</pre>
1120<p>The <i>layout specification</i> consists of a list of specifications
1121separated by the minus sign character ('-'). Each specification starts with a
1122letter and may include other information after the letter to define some
1123aspect of the data layout. The specifications accepted are as follows: </p>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1124<dl>
1125 <dt><tt>E</tt></dt>
1126 <dd>Specifies that the target lays out data in big-endian form. That is, the
1127 bits with the most significance have the lowest address location.</dd>
1128 <dt><tt>e</tt></dt>
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]1129 <dd>Specifies that the target lays out data in little-endian form. That is,
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1130 the bits with the least significance have the lowest address location.</dd>
1131 <dt><tt>p:<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1132 <dd>This specifies the <i>size</i> of a pointer and its <i>abi</i> and
1133 <i>preferred</i> alignments. All sizes are in bits. Specifying the <i>pref</i>
1134 alignment is optional. If omitted, the preceding <tt>:</tt> should be omitted
1135 too.</dd>
1136 <dt><tt>i<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1137 <dd>This specifies the alignment for an integer type of a given bit
1138 <i>size</i>. The value of <i>size</i> must be in the range [1,2^23).</dd>
1139 <dt><tt>v<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1140 <dd>This specifies the alignment for a vector type of a given bit
1141 <i>size</i>.</dd>
1142 <dt><tt>f<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1143 <dd>This specifies the alignment for a floating point type of a given bit
1144 <i>size</i>. The value of <i>size</i> must be either 32 (float) or 64
1145 (double).</dd>
1146 <dt><tt>a<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1147 <dd>This specifies the alignment for an aggregate type of a given bit
1148 <i>size</i>.</dd>
1149</dl>
1150<p>When constructing the data layout for a given target, LLVM starts with a
1151default set of specifications which are then (possibly) overriden by the
1152specifications in the <tt>datalayout</tt> keyword. The default specifications
1153are given in this list:</p>
1154<ul>
1155 <li><tt>E</tt> - big endian</li>
1156 <li><tt>p:32:64:64</tt> - 32-bit pointers with 64-bit alignment</li>
1157 <li><tt>i1:8:8</tt> - i1 is 8-bit (byte) aligned</li>
1158 <li><tt>i8:8:8</tt> - i8 is 8-bit (byte) aligned</li>
1159 <li><tt>i16:16:16</tt> - i16 is 16-bit aligned</li>
1160 <li><tt>i32:32:32</tt> - i32 is 32-bit aligned</li>
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]1161 <li><tt>i64:32:64</tt> - i64 has ABI alignment of 32-bits but preferred
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1162 alignment of 64-bits</li>
1163 <li><tt>f32:32:32</tt> - float is 32-bit aligned</li>
1164 <li><tt>f64:64:64</tt> - double is 64-bit aligned</li>
1165 <li><tt>v64:64:64</tt> - 64-bit vector is 64-bit aligned</li>
1166 <li><tt>v128:128:128</tt> - 128-bit vector is 128-bit aligned</li>
1167 <li><tt>a0:0:1</tt> - aggregates are 8-bit aligned</li>
1168</ul>
Chris Lattnerebec6782008-08-05 18:21:08 +0000[diff] [blame]1169<p>When LLVM is determining the alignment for a given type, it uses the
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]1170following rules:</p>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1171<ol>
1172 <li>If the type sought is an exact match for one of the specifications, that
1173 specification is used.</li>
1174 <li>If no match is found, and the type sought is an integer type, then the
1175 smallest integer type that is larger than the bitwidth of the sought type is
1176 used. If none of the specifications are larger than the bitwidth then the the
1177 largest integer type is used. For example, given the default specifications
1178 above, the i7 type will use the alignment of i8 (next largest) while both
1179 i65 and i256 will use the alignment of i64 (largest specified).</li>
1180 <li>If no match is found, and the type sought is a vector type, then the
1181 largest vector type that is smaller than the sought vector type will be used
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]1182 as a fall back. This happens because &lt;128 x double&gt; can be implemented
1183 in terms of 64 &lt;2 x double&gt;, for example.</li>
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]1184</ol>
1185</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]1186
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1187<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1188<div class="doc_section"> <a name="typesystem">Type System</a> </div>
1189<!-- *********************************************************************** -->
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]1190
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1191<div class="doc_text">
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]1192
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1193<p>The LLVM type system is one of the most important features of the
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1194intermediate representation. Being typed enables a number of
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]1195optimizations to be performed on the intermediate representation directly,
1196without having to do
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1197extra analyses on the side before the transformation. A strong type
1198system makes it easier to read the generated code and enables novel
1199analyses and transformations that are not feasible to perform on normal
1200three address code representations.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]1201
1202</div>
1203
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1204<!-- ======================================================================= -->
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1205<div class="doc_subsection"> <a name="t_classifications">Type
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1206Classifications</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1207<div class="doc_text">
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1208<p>The types fall into a few useful
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1209classifications:</p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1210
1211<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1212 <tbody>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1213 <tr><th>Classification</th><th>Types</th></tr>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1214 <tr>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1215 <td><a href="#t_integer">integer</a></td>
Reid Spencer2b916312007-05-16 18:44:01 +0000[diff] [blame]1216 <td><tt>i1, i2, i3, ... i8, ... i16, ... i32, ... i64, ... </tt></td>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1217 </tr>
1218 <tr>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1219 <td><a href="#t_floating">floating point</a></td>
1220 <td><tt>float, double, x86_fp80, fp128, ppc_fp128</tt></td>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1221 </tr>
1222 <tr>
1223 <td><a name="t_firstclass">first class</a></td>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1224 <td><a href="#t_integer">integer</a>,
1225 <a href="#t_floating">floating point</a>,
1226 <a href="#t_pointer">pointer</a>,
Dan Gohman0066db62008-06-18 18:42:13 +0000[diff] [blame]1227 <a href="#t_vector">vector</a>,
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]1228 <a href="#t_struct">structure</a>,
1229 <a href="#t_array">array</a>,
Dan Gohmanade5faa2008-05-23 22:50:26 +0000[diff] [blame]1230 <a href="#t_label">label</a>.
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1231 </td>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1232 </tr>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1233 <tr>
1234 <td><a href="#t_primitive">primitive</a></td>
1235 <td><a href="#t_label">label</a>,
1236 <a href="#t_void">void</a>,
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1237 <a href="#t_floating">floating point</a>.</td>
1238 </tr>
1239 <tr>
1240 <td><a href="#t_derived">derived</a></td>
1241 <td><a href="#t_integer">integer</a>,
1242 <a href="#t_array">array</a>,
1243 <a href="#t_function">function</a>,
1244 <a href="#t_pointer">pointer</a>,
1245 <a href="#t_struct">structure</a>,
1246 <a href="#t_pstruct">packed structure</a>,
1247 <a href="#t_vector">vector</a>,
1248 <a href="#t_opaque">opaque</a>.
Dan Gohman01ac1012008-10-14 16:32:04 +0000[diff] [blame]1249 </td>
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1250 </tr>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1251 </tbody>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1252</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1253
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1254<p>The <a href="#t_firstclass">first class</a> types are perhaps the
1255most important. Values of these types are the only ones which can be
1256produced by instructions, passed as arguments, or used as operands to
Dan Gohmanc4b49eb2008-05-23 21:53:15 +0000[diff] [blame]1257instructions.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1258</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1259
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1260<!-- ======================================================================= -->
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1261<div class="doc_subsection"> <a name="t_primitive">Primitive Types</a> </div>
Chris Lattner8f8c7b72008-01-04 04:34:14 +0000[diff] [blame]1262
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1263<div class="doc_text">
1264<p>The primitive types are the fundamental building blocks of the LLVM
1265system.</p>
1266
Chris Lattner8f8c7b72008-01-04 04:34:14 +0000[diff] [blame]1267</div>
1268
Chris Lattner4f69f462008-01-04 04:32:38 +0000[diff] [blame]1269<!-- _______________________________________________________________________ -->
1270<div class="doc_subsubsection"> <a name="t_floating">Floating Point Types</a> </div>
1271
1272<div class="doc_text">
1273 <table>
1274 <tbody>
1275 <tr><th>Type</th><th>Description</th></tr>
1276 <tr><td><tt>float</tt></td><td>32-bit floating point value</td></tr>
1277 <tr><td><tt>double</tt></td><td>64-bit floating point value</td></tr>
1278 <tr><td><tt>fp128</tt></td><td>128-bit floating point value (112-bit mantissa)</td></tr>
1279 <tr><td><tt>x86_fp80</tt></td><td>80-bit floating point value (X87)</td></tr>
1280 <tr><td><tt>ppc_fp128</tt></td><td>128-bit floating point value (two 64-bits)</td></tr>
1281 </tbody>
1282 </table>
1283</div>
1284
1285<!-- _______________________________________________________________________ -->
1286<div class="doc_subsubsection"> <a name="t_void">Void Type</a> </div>
1287
1288<div class="doc_text">
1289<h5>Overview:</h5>
1290<p>The void type does not represent any value and has no size.</p>
1291
1292<h5>Syntax:</h5>
1293
1294<pre>
1295 void
1296</pre>
1297</div>
1298
1299<!-- _______________________________________________________________________ -->
1300<div class="doc_subsubsection"> <a name="t_label">Label Type</a> </div>
1301
1302<div class="doc_text">
1303<h5>Overview:</h5>
1304<p>The label type represents code labels.</p>
1305
1306<h5>Syntax:</h5>
1307
1308<pre>
1309 label
1310</pre>
1311</div>
1312
1313
1314<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1315<div class="doc_subsection"> <a name="t_derived">Derived Types</a> </div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1316
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1317<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1318
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1319<p>The real power in LLVM comes from the derived types in the system.
1320This is what allows a programmer to represent arrays, functions,
1321pointers, and other useful types. Note that these derived types may be
1322recursive: For example, it is possible to have a two dimensional array.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1323
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1324</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1325
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1326<!-- _______________________________________________________________________ -->
Reid Spencer2b916312007-05-16 18:44:01 +0000[diff] [blame]1327<div class="doc_subsubsection"> <a name="t_integer">Integer Type</a> </div>
1328
1329<div class="doc_text">
1330
1331<h5>Overview:</h5>
1332<p>The integer type is a very simple derived type that simply specifies an
1333arbitrary bit width for the integer type desired. Any bit width from 1 bit to
13342^23-1 (about 8 million) can be specified.</p>
1335
1336<h5>Syntax:</h5>
1337
1338<pre>
1339 iN
1340</pre>
1341
1342<p>The number of bits the integer will occupy is specified by the <tt>N</tt>
1343value.</p>
1344
1345<h5>Examples:</h5>
1346<table class="layout">
Chris Lattnerb9488a62007-12-18 06:18:21 +0000[diff] [blame]1347 <tbody>
1348 <tr>
1349 <td><tt>i1</tt></td>
1350 <td>a single-bit integer.</td>
1351 </tr><tr>
1352 <td><tt>i32</tt></td>
1353 <td>a 32-bit integer.</td>
1354 </tr><tr>
1355 <td><tt>i1942652</tt></td>
1356 <td>a really big integer of over 1 million bits.</td>
Reid Spencer2b916312007-05-16 18:44:01 +0000[diff] [blame]1357 </tr>
Chris Lattnerb9488a62007-12-18 06:18:21 +0000[diff] [blame]1358 </tbody>
Reid Spencer2b916312007-05-16 18:44:01 +0000[diff] [blame]1359</table>
Dan Gohmand8791e52009-01-24 15:58:40 +0000[diff] [blame]1360
1361<p>Note that the code generator does not yet support large integer types
1362to be used as function return types. The specific limit on how large a
1363return type the code generator can currently handle is target-dependent;
1364currently it's often 64 bits for 32-bit targets and 128 bits for 64-bit
1365targets.</p>
1366
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1367</div>
Reid Spencer2b916312007-05-16 18:44:01 +0000[diff] [blame]1368
1369<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1370<div class="doc_subsubsection"> <a name="t_array">Array Type</a> </div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1371
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1372<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1373
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1374<h5>Overview:</h5>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1375
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1376<p>The array type is a very simple derived type that arranges elements
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1377sequentially in memory. The array type requires a size (number of
1378elements) and an underlying data type.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1379
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1380<h5>Syntax:</h5>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1381
1382<pre>
1383 [&lt;# elements&gt; x &lt;elementtype&gt;]
1384</pre>
1385
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]1386<p>The number of elements is a constant integer value; elementtype may
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1387be any type with a size.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1388
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1389<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1390<table class="layout">
1391 <tr class="layout">
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]1392 <td class="left"><tt>[40 x i32]</tt></td>
1393 <td class="left">Array of 40 32-bit integer values.</td>
1394 </tr>
1395 <tr class="layout">
1396 <td class="left"><tt>[41 x i32]</tt></td>
1397 <td class="left">Array of 41 32-bit integer values.</td>
1398 </tr>
1399 <tr class="layout">
1400 <td class="left"><tt>[4 x i8]</tt></td>
1401 <td class="left">Array of 4 8-bit integer values.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1402 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1403</table>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1404<p>Here are some examples of multidimensional arrays:</p>
1405<table class="layout">
1406 <tr class="layout">
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]1407 <td class="left"><tt>[3 x [4 x i32]]</tt></td>
1408 <td class="left">3x4 array of 32-bit integer values.</td>
1409 </tr>
1410 <tr class="layout">
1411 <td class="left"><tt>[12 x [10 x float]]</tt></td>
1412 <td class="left">12x10 array of single precision floating point values.</td>
1413 </tr>
1414 <tr class="layout">
1415 <td class="left"><tt>[2 x [3 x [4 x i16]]]</tt></td>
1416 <td class="left">2x3x4 array of 16-bit integer values.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1417 </tr>
1418</table>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]1419
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]1420<p>Note that 'variable sized arrays' can be implemented in LLVM with a zero
1421length array. Normally, accesses past the end of an array are undefined in
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]1422LLVM (e.g. it is illegal to access the 5th element of a 3 element array).
1423As a special case, however, zero length arrays are recognized to be variable
1424length. This allows implementation of 'pascal style arrays' with the LLVM
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1425type "{ i32, [0 x float]}", for example.</p>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]1426
Dan Gohmand8791e52009-01-24 15:58:40 +0000[diff] [blame]1427<p>Note that the code generator does not yet support large aggregate types
1428to be used as function return types. The specific limit on how large an
1429aggregate return type the code generator can currently handle is
1430target-dependent, and also dependent on the aggregate element types.</p>
1431
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1432</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1433
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1434<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1435<div class="doc_subsubsection"> <a name="t_function">Function Type</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1436<div class="doc_text">
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]1437
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1438<h5>Overview:</h5>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]1439
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1440<p>The function type can be thought of as a function signature. It
Devang Patela582f402008-03-24 05:35:41 +0000[diff] [blame]1441consists of a return type and a list of formal parameter types. The
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]1442return type of a function type is a scalar type, a void type, or a struct type.
Devang Patel7a3ad1a2008-03-24 20:52:42 +0000[diff] [blame]1443If the return type is a struct type then all struct elements must be of first
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]1444class types, and the struct must have at least one element.</p>
Devang Patelc3fc6df2008-03-10 20:49:15 +0000[diff] [blame]1445
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1446<h5>Syntax:</h5>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]1447
1448<pre>
1449 &lt;returntype list&gt; (&lt;parameter list&gt;)
1450</pre>
1451
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]1452<p>...where '<tt>&lt;parameter list&gt;</tt>' is a comma-separated list of type
Misha Brukmanc24b7582004-08-12 20:16:08 +0000[diff] [blame]1453specifiers. Optionally, the parameter list may include a type <tt>...</tt>,
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]1454which indicates that the function takes a variable number of arguments.
1455Variable argument functions can access their arguments with the <a
Devang Patelc3fc6df2008-03-10 20:49:15 +0000[diff] [blame]1456 href="#int_varargs">variable argument handling intrinsic</a> functions.
1457'<tt>&lt;returntype list&gt;</tt>' is a comma-separated list of
1458<a href="#t_firstclass">first class</a> type specifiers.</p>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]1459
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1460<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1461<table class="layout">
1462 <tr class="layout">
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]1463 <td class="left"><tt>i32 (i32)</tt></td>
1464 <td class="left">function taking an <tt>i32</tt>, returning an <tt>i32</tt>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1465 </td>
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]1466 </tr><tr class="layout">
Reid Spencer9445e9a2007-07-19 23:13:04 +0000[diff] [blame]1467 <td class="left"><tt>float&nbsp;(i16&nbsp;signext,&nbsp;i32&nbsp;*)&nbsp;*
Reid Spencerf17a0b72006-12-31 07:20:23 +0000[diff] [blame]1468 </tt></td>
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]1469 <td class="left"><a href="#t_pointer">Pointer</a> to a function that takes
1470 an <tt>i16</tt> that should be sign extended and a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1471 <a href="#t_pointer">pointer</a> to <tt>i32</tt>, returning
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]1472 <tt>float</tt>.
1473 </td>
1474 </tr><tr class="layout">
1475 <td class="left"><tt>i32 (i8*, ...)</tt></td>
1476 <td class="left">A vararg function that takes at least one
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]1477 <a href="#t_pointer">pointer</a> to <tt>i8 </tt> (char in C),
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]1478 which returns an integer. This is the signature for <tt>printf</tt> in
1479 LLVM.
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1480 </td>
Devang Patela582f402008-03-24 05:35:41 +0000[diff] [blame]1481 </tr><tr class="layout">
1482 <td class="left"><tt>{i32, i32} (i32)</tt></td>
Misha Brukmanb0a57aa2008-11-27 06:41:20 +0000[diff] [blame]1483 <td class="left">A function taking an <tt>i32</tt>, returning two
1484 <tt>i32</tt> values as an aggregate of type <tt>{ i32, i32 }</tt>
Devang Patela582f402008-03-24 05:35:41 +0000[diff] [blame]1485 </td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1486 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1487</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1488
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1489</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1490<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1491<div class="doc_subsubsection"> <a name="t_struct">Structure Type</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1492<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1493<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1494<p>The structure type is used to represent a collection of data members
1495together in memory. The packing of the field types is defined to match
1496the ABI of the underlying processor. The elements of a structure may
1497be any type that has a size.</p>
1498<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
1499and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
1500field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
1501instruction.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1502<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1503<pre> { &lt;type list&gt; }<br></pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1504<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1505<table class="layout">
1506 <tr class="layout">
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]1507 <td class="left"><tt>{ i32, i32, i32 }</tt></td>
1508 <td class="left">A triple of three <tt>i32</tt> values</td>
1509 </tr><tr class="layout">
1510 <td class="left"><tt>{&nbsp;float,&nbsp;i32&nbsp;(i32)&nbsp;*&nbsp;}</tt></td>
1511 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1512 second element is a <a href="#t_pointer">pointer</a> to a
1513 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1514 an <tt>i32</tt>.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1515 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1516</table>
Dan Gohmand8791e52009-01-24 15:58:40 +0000[diff] [blame]1517
1518<p>Note that the code generator does not yet support large aggregate types
1519to be used as function return types. The specific limit on how large an
1520aggregate return type the code generator can currently handle is
1521target-dependent, and also dependent on the aggregate element types.</p>
1522
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1523</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1524
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1525<!-- _______________________________________________________________________ -->
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]1526<div class="doc_subsubsection"> <a name="t_pstruct">Packed Structure Type</a>
1527</div>
1528<div class="doc_text">
1529<h5>Overview:</h5>
1530<p>The packed structure type is used to represent a collection of data members
1531together in memory. There is no padding between fields. Further, the alignment
1532of a packed structure is 1 byte. The elements of a packed structure may
1533be any type that has a size.</p>
1534<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
1535and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
1536field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
1537instruction.</p>
1538<h5>Syntax:</h5>
1539<pre> &lt; { &lt;type list&gt; } &gt; <br></pre>
1540<h5>Examples:</h5>
1541<table class="layout">
1542 <tr class="layout">
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]1543 <td class="left"><tt>&lt; { i32, i32, i32 } &gt;</tt></td>
1544 <td class="left">A triple of three <tt>i32</tt> values</td>
1545 </tr><tr class="layout">
Bill Wendlinge36dccc2008-09-07 10:26:33 +0000[diff] [blame]1546 <td class="left">
1547<tt>&lt;&nbsp;{&nbsp;float,&nbsp;i32&nbsp;(i32)*&nbsp;}&nbsp;&gt;</tt></td>
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]1548 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1549 second element is a <a href="#t_pointer">pointer</a> to a
1550 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1551 an <tt>i32</tt>.</td>
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]1552 </tr>
1553</table>
1554</div>
1555
1556<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1557<div class="doc_subsubsection"> <a name="t_pointer">Pointer Type</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1558<div class="doc_text">
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1559<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1560<p>As in many languages, the pointer type represents a pointer or
Christopher Lamb284d9922007-12-11 09:31:00 +0000[diff] [blame]1561reference to another object, which must live in memory. Pointer types may have
1562an optional address space attribute defining the target-specific numbered
1563address space where the pointed-to object resides. The default address space is
1564zero.</p>
Chris Lattner0fd4a272009-02-08 19:53:29 +0000[diff] [blame]1565
1566<p>Note that LLVM does not permit pointers to void (<tt>void*</tt>) nor does
Chris Lattnere220e8c2009-02-08 22:21:28 +0000[diff] [blame]1567it permit pointers to labels (<tt>label*</tt>). Use <tt>i8*</tt> instead.</p>
Chris Lattner0fd4a272009-02-08 19:53:29 +0000[diff] [blame]1568
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1569<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1570<pre> &lt;type&gt; *<br></pre>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1571<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1572<table class="layout">
1573 <tr class="layout">
Dan Gohman2a08c532009-01-04 23:44:43 +0000[diff] [blame]1574 <td class="left"><tt>[4 x i32]*</tt></td>
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]1575 <td class="left">A <a href="#t_pointer">pointer</a> to <a
1576 href="#t_array">array</a> of four <tt>i32</tt> values.</td>
1577 </tr>
1578 <tr class="layout">
1579 <td class="left"><tt>i32 (i32 *) *</tt></td>
1580 <td class="left"> A <a href="#t_pointer">pointer</a> to a <a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1581 href="#t_function">function</a> that takes an <tt>i32*</tt>, returning an
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]1582 <tt>i32</tt>.</td>
1583 </tr>
1584 <tr class="layout">
1585 <td class="left"><tt>i32 addrspace(5)*</tt></td>
1586 <td class="left">A <a href="#t_pointer">pointer</a> to an <tt>i32</tt> value
1587 that resides in address space #5.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1588 </tr>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1589</table>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1590</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1591
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1592<!-- _______________________________________________________________________ -->
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1593<div class="doc_subsubsection"> <a name="t_vector">Vector Type</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1594<div class="doc_text">
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1595
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1596<h5>Overview:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1597
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1598<p>A vector type is a simple derived type that represents a vector
1599of elements. Vector types are used when multiple primitive data
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1600are operated in parallel using a single instruction (SIMD).
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1601A vector type requires a size (number of
Chris Lattnerb8d172f2005-11-10 01:44:22 +0000[diff] [blame]1602elements) and an underlying primitive data type. Vectors must have a power
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1603of two length (1, 2, 4, 8, 16 ...). Vector types are
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1604considered <a href="#t_firstclass">first class</a>.</p>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1605
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1606<h5>Syntax:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1607
1608<pre>
1609 &lt; &lt;# elements&gt; x &lt;elementtype&gt; &gt;
1610</pre>
1611
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]1612<p>The number of elements is a constant integer value; elementtype may
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]1613be any integer or floating point type.</p>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1614
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1615<h5>Examples:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1616
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1617<table class="layout">
1618 <tr class="layout">
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]1619 <td class="left"><tt>&lt;4 x i32&gt;</tt></td>
1620 <td class="left">Vector of 4 32-bit integer values.</td>
1621 </tr>
1622 <tr class="layout">
1623 <td class="left"><tt>&lt;8 x float&gt;</tt></td>
1624 <td class="left">Vector of 8 32-bit floating-point values.</td>
1625 </tr>
1626 <tr class="layout">
1627 <td class="left"><tt>&lt;2 x i64&gt;</tt></td>
1628 <td class="left">Vector of 2 64-bit integer values.</td>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1629 </tr>
1630</table>
Dan Gohmand8791e52009-01-24 15:58:40 +0000[diff] [blame]1631
1632<p>Note that the code generator does not yet support large vector types
1633to be used as function return types. The specific limit on how large a
1634vector return type codegen can currently handle is target-dependent;
1635currently it's often a few times longer than a hardware vector register.</p>
1636
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1637</div>
1638
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1639<!-- _______________________________________________________________________ -->
1640<div class="doc_subsubsection"> <a name="t_opaque">Opaque Type</a> </div>
1641<div class="doc_text">
1642
1643<h5>Overview:</h5>
1644
1645<p>Opaque types are used to represent unknown types in the system. This
Gordon Henriksen8ac04ff2007-10-14 00:34:53 +0000[diff] [blame]1646corresponds (for example) to the C notion of a forward declared structure type.
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1647In LLVM, opaque types can eventually be resolved to any type (not just a
1648structure type).</p>
1649
1650<h5>Syntax:</h5>
1651
1652<pre>
1653 opaque
1654</pre>
1655
1656<h5>Examples:</h5>
1657
1658<table class="layout">
1659 <tr class="layout">
Chris Lattner23ff1f92007-12-19 05:04:11 +0000[diff] [blame]1660 <td class="left"><tt>opaque</tt></td>
1661 <td class="left">An opaque type.</td>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1662 </tr>
1663</table>
1664</div>
1665
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]1666<!-- ======================================================================= -->
1667<div class="doc_subsection">
1668 <a name="t_uprefs">Type Up-references</a>
1669</div>
1670
1671<div class="doc_text">
1672<h5>Overview:</h5>
1673<p>
1674An "up reference" allows you to refer to a lexically enclosing type without
1675requiring it to have a name. For instance, a structure declaration may contain a
1676pointer to any of the types it is lexically a member of. Example of up
1677references (with their equivalent as named type declarations) include:</p>
1678
1679<pre>
Chris Lattner3060f5b2009-02-09 10:00:56 +0000[diff] [blame]1680 { \2 * } %x = type { %x* }
Chris Lattner242d61d2009-02-02 07:32:36 +0000[diff] [blame]1681 { \2 }* %y = type { %y }*
1682 \1* %z = type %z*
1683</pre>
1684
1685<p>
1686An up reference is needed by the asmprinter for printing out cyclic types when
1687there is no declared name for a type in the cycle. Because the asmprinter does
1688not want to print out an infinite type string, it needs a syntax to handle
1689recursive types that have no names (all names are optional in llvm IR).
1690</p>
1691
1692<h5>Syntax:</h5>
1693<pre>
1694 \&lt;level&gt;
1695</pre>
1696
1697<p>
1698The level is the count of the lexical type that is being referred to.
1699</p>
1700
1701<h5>Examples:</h5>
1702
1703<table class="layout">
1704 <tr class="layout">
1705 <td class="left"><tt>\1*</tt></td>
1706 <td class="left">Self-referential pointer.</td>
1707 </tr>
1708 <tr class="layout">
1709 <td class="left"><tt>{ { \3*, i8 }, i32 }</tt></td>
1710 <td class="left">Recursive structure where the upref refers to the out-most
1711 structure.</td>
1712 </tr>
1713</table>
1714</div>
1715
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1716
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1717<!-- *********************************************************************** -->
1718<div class="doc_section"> <a name="constants">Constants</a> </div>
1719<!-- *********************************************************************** -->
1720
1721<div class="doc_text">
1722
1723<p>LLVM has several different basic types of constants. This section describes
1724them all and their syntax.</p>
1725
1726</div>
1727
1728<!-- ======================================================================= -->
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]1729<div class="doc_subsection"><a name="simpleconstants">Simple Constants</a></div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1730
1731<div class="doc_text">
1732
1733<dl>
1734 <dt><b>Boolean constants</b></dt>
1735
1736 <dd>The two strings '<tt>true</tt>' and '<tt>false</tt>' are both valid
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]1737 constants of the <tt><a href="#t_primitive">i1</a></tt> type.
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1738 </dd>
1739
1740 <dt><b>Integer constants</b></dt>
1741
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]1742 <dd>Standard integers (such as '4') are constants of the <a
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]1743 href="#t_integer">integer</a> type. Negative numbers may be used with
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1744 integer types.
1745 </dd>
1746
1747 <dt><b>Floating point constants</b></dt>
1748
1749 <dd>Floating point constants use standard decimal notation (e.g. 123.421),
1750 exponential notation (e.g. 1.23421e+2), or a more precise hexadecimal
Chris Lattnera73afe02008-04-01 18:45:27 +0000[diff] [blame]1751 notation (see below). The assembler requires the exact decimal value of
1752 a floating-point constant. For example, the assembler accepts 1.25 but
1753 rejects 1.3 because 1.3 is a repeating decimal in binary. Floating point
1754 constants must have a <a href="#t_floating">floating point</a> type. </dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1755
1756 <dt><b>Null pointer constants</b></dt>
1757
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1758 <dd>The identifier '<tt>null</tt>' is recognized as a null pointer constant
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1759 and must be of <a href="#t_pointer">pointer type</a>.</dd>
1760
1761</dl>
1762
Dale Johannesenbd5e5a82009-02-11 22:14:51 +0000[diff] [blame]1763<p>The one non-intuitive notation for constants is the hexadecimal form
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1764of floating point constants. For example, the form '<tt>double
17650x432ff973cafa8000</tt>' is equivalent to (but harder to read than) '<tt>double
17664.5e+15</tt>'. The only time hexadecimal floating point constants are required
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]1767(and the only time that they are generated by the disassembler) is when a
1768floating point constant must be emitted but it cannot be represented as a
Dale Johannesenbd5e5a82009-02-11 22:14:51 +0000[diff] [blame]1769decimal floating point number in a reasonable number of digits. For example,
1770NaN's, infinities, and other
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]1771special values are represented in their IEEE hexadecimal format so that
1772assembly and disassembly do not cause any bits to change in the constants.</p>
Dale Johannesenbd5e5a82009-02-11 22:14:51 +0000[diff] [blame]1773<p>When using the hexadecimal form, constants of types float and double are
1774represented using the 16-digit form shown above (which matches the IEEE754
1775representation for double); float values must, however, be exactly representable
1776as IEE754 single precision.
1777Hexadecimal format is always used for long
1778double, and there are three forms of long double. The 80-bit
1779format used by x86 is represented as <tt>0xK</tt>
1780followed by 20 hexadecimal digits.
1781The 128-bit format used by PowerPC (two adjacent doubles) is represented
1782by <tt>0xM</tt> followed by 32 hexadecimal digits. The IEEE 128-bit
1783format is represented
1784by <tt>0xL</tt> followed by 32 hexadecimal digits; no currently supported
1785target uses this format. Long doubles will only work if they match
1786the long double format on your target. All hexadecimal formats are big-endian
1787(sign bit at the left).</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1788</div>
1789
1790<!-- ======================================================================= -->
1791<div class="doc_subsection"><a name="aggregateconstants">Aggregate Constants</a>
1792</div>
1793
1794<div class="doc_text">
Chris Lattnerd4f6b172005-03-07 22:13:59 +0000[diff] [blame]1795<p>Aggregate constants arise from aggregation of simple constants
1796and smaller aggregate constants.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1797
1798<dl>
1799 <dt><b>Structure constants</b></dt>
1800
1801 <dd>Structure constants are represented with notation similar to structure
1802 type definitions (a comma separated list of elements, surrounded by braces
Chris Lattner64910ee2007-12-25 20:34:52 +0000[diff] [blame]1803 (<tt>{}</tt>)). For example: "<tt>{ i32 4, float 17.0, i32* @G }</tt>",
1804 where "<tt>@G</tt>" is declared as "<tt>@G = external global i32</tt>". Structure constants
Chris Lattnerd4f6b172005-03-07 22:13:59 +0000[diff] [blame]1805 must have <a href="#t_struct">structure type</a>, and the number and
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1806 types of elements must match those specified by the type.
1807 </dd>
1808
1809 <dt><b>Array constants</b></dt>
1810
1811 <dd>Array constants are represented with notation similar to array type
1812 definitions (a comma separated list of elements, surrounded by square brackets
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1813 (<tt>[]</tt>)). For example: "<tt>[ i32 42, i32 11, i32 74 ]</tt>". Array
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1814 constants must have <a href="#t_array">array type</a>, and the number and
1815 types of elements must match those specified by the type.
1816 </dd>
1817
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1818 <dt><b>Vector constants</b></dt>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1819
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1820 <dd>Vector constants are represented with notation similar to vector type
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1821 definitions (a comma separated list of elements, surrounded by
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1822 less-than/greater-than's (<tt>&lt;&gt;</tt>)). For example: "<tt>&lt; i32 42,
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]1823 i32 11, i32 74, i32 100 &gt;</tt>". Vector constants must have <a
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]1824 href="#t_vector">vector type</a>, and the number and types of elements must
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1825 match those specified by the type.
1826 </dd>
1827
1828 <dt><b>Zero initialization</b></dt>
1829
1830 <dd>The string '<tt>zeroinitializer</tt>' can be used to zero initialize a
1831 value to zero of <em>any</em> type, including scalar and aggregate types.
1832 This is often used to avoid having to print large zero initializers (e.g. for
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]1833 large arrays) and is always exactly equivalent to using explicit zero
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1834 initializers.
1835 </dd>
1836</dl>
1837
1838</div>
1839
1840<!-- ======================================================================= -->
1841<div class="doc_subsection">
1842 <a name="globalconstants">Global Variable and Function Addresses</a>
1843</div>
1844
1845<div class="doc_text">
1846
1847<p>The addresses of <a href="#globalvars">global variables</a> and <a
1848href="#functionstructure">functions</a> are always implicitly valid (link-time)
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1849constants. These constants are explicitly referenced when the <a
1850href="#identifiers">identifier for the global</a> is used and always have <a
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1851href="#t_pointer">pointer</a> type. For example, the following is a legal LLVM
1852file:</p>
1853
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1854<div class="doc_code">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1855<pre>
Chris Lattnera18a4242007-06-06 18:28:13 +0000[diff] [blame]1856@X = global i32 17
1857@Y = global i32 42
1858@Z = global [2 x i32*] [ i32* @X, i32* @Y ]
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1859</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]1860</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1861
1862</div>
1863
1864<!-- ======================================================================= -->
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]1865<div class="doc_subsection"><a name="undefvalues">Undefined Values</a></div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1866<div class="doc_text">
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]1867 <p>The string '<tt>undef</tt>' is recognized as a type-less constant that has
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]1868 no specific value. Undefined values may be of any type and be used anywhere
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]1869 a constant is permitted.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1870
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]1871 <p>Undefined values indicate to the compiler that the program is well defined
1872 no matter what value is used, giving the compiler more freedom to optimize.
1873 </p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1874</div>
1875
1876<!-- ======================================================================= -->
1877<div class="doc_subsection"><a name="constantexprs">Constant Expressions</a>
1878</div>
1879
1880<div class="doc_text">
1881
1882<p>Constant expressions are used to allow expressions involving other constants
1883to be used as constants. Constant expressions may be of any <a
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]1884href="#t_firstclass">first class</a> type and may involve any LLVM operation
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1885that does not have side effects (e.g. load and call are not supported). The
1886following is the syntax for constant expressions:</p>
1887
1888<dl>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1889 <dt><b><tt>trunc ( CST to TYPE )</tt></b></dt>
1890 <dd>Truncate a constant to another type. The bit size of CST must be larger
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]1891 than the bit size of TYPE. Both types must be integers.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1892
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1893 <dt><b><tt>zext ( CST to TYPE )</tt></b></dt>
1894 <dd>Zero extend a constant to another type. The bit size of CST must be
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]1895 smaller or equal to the bit size of TYPE. Both types must be integers.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1896
1897 <dt><b><tt>sext ( CST to TYPE )</tt></b></dt>
1898 <dd>Sign extend a constant to another type. The bit size of CST must be
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]1899 smaller or equal to the bit size of TYPE. Both types must be integers.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1900
1901 <dt><b><tt>fptrunc ( CST to TYPE )</tt></b></dt>
1902 <dd>Truncate a floating point constant to another floating point type. The
1903 size of CST must be larger than the size of TYPE. Both types must be
1904 floating point.</dd>
1905
1906 <dt><b><tt>fpext ( CST to TYPE )</tt></b></dt>
1907 <dd>Floating point extend a constant to another type. The size of CST must be
1908 smaller or equal to the size of TYPE. Both types must be floating point.</dd>
1909
Reid Spencer1539a1c2007-07-31 14:40:14 +0000[diff] [blame]1910 <dt><b><tt>fptoui ( CST to TYPE )</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1911 <dd>Convert a floating point constant to the corresponding unsigned integer
Nate Begemanb348d182007-11-17 03:58:34 +0000[diff] [blame]1912 constant. TYPE must be a scalar or vector integer type. CST must be of scalar
1913 or vector floating point type. Both CST and TYPE must be scalars, or vectors
1914 of the same number of elements. If the value won't fit in the integer type,
1915 the results are undefined.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1916
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]1917 <dt><b><tt>fptosi ( CST to TYPE )</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1918 <dd>Convert a floating point constant to the corresponding signed integer
Nate Begemanb348d182007-11-17 03:58:34 +0000[diff] [blame]1919 constant. TYPE must be a scalar or vector integer type. CST must be of scalar
1920 or vector floating point type. Both CST and TYPE must be scalars, or vectors
1921 of the same number of elements. If the value won't fit in the integer type,
1922 the results are undefined.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1923
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]1924 <dt><b><tt>uitofp ( CST to TYPE )</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1925 <dd>Convert an unsigned integer constant to the corresponding floating point
Nate Begemanb348d182007-11-17 03:58:34 +0000[diff] [blame]1926 constant. TYPE must be a scalar or vector floating point type. CST must be of
1927 scalar or vector integer type. Both CST and TYPE must be scalars, or vectors
1928 of the same number of elements. If the value won't fit in the floating point
1929 type, the results are undefined.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1930
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]1931 <dt><b><tt>sitofp ( CST to TYPE )</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1932 <dd>Convert a signed integer constant to the corresponding floating point
Nate Begemanb348d182007-11-17 03:58:34 +0000[diff] [blame]1933 constant. TYPE must be a scalar or vector floating point type. CST must be of
1934 scalar or vector integer type. Both CST and TYPE must be scalars, or vectors
1935 of the same number of elements. If the value won't fit in the floating point
1936 type, the results are undefined.</dd>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1937
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]1938 <dt><b><tt>ptrtoint ( CST to TYPE )</tt></b></dt>
1939 <dd>Convert a pointer typed constant to the corresponding integer constant
1940 TYPE must be an integer type. CST must be of pointer type. The CST value is
1941 zero extended, truncated, or unchanged to make it fit in TYPE.</dd>
1942
1943 <dt><b><tt>inttoptr ( CST to TYPE )</tt></b></dt>
1944 <dd>Convert a integer constant to a pointer constant. TYPE must be a
1945 pointer type. CST must be of integer type. The CST value is zero extended,
1946 truncated, or unchanged to make it fit in a pointer size. This one is
1947 <i>really</i> dangerous!</dd>
1948
1949 <dt><b><tt>bitcast ( CST to TYPE )</tt></b></dt>
Chris Lattner03bbad62009-02-28 18:27:03 +0000[diff] [blame^]1950 <dd>Convert a constant, CST, to another TYPE. The constraints of the operands
1951 are the same as those for the <a href="#i_bitcast">bitcast
1952 instruction</a>.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1953
1954 <dt><b><tt>getelementptr ( CSTPTR, IDX0, IDX1, ... )</tt></b></dt>
1955
1956 <dd>Perform the <a href="#i_getelementptr">getelementptr operation</a> on
1957 constants. As with the <a href="#i_getelementptr">getelementptr</a>
1958 instruction, the index list may have zero or more indexes, which are required
1959 to make sense for the type of "CSTPTR".</dd>
1960
Robert Bocchino9fbe1452006-01-10 19:31:34 +0000[diff] [blame]1961 <dt><b><tt>select ( COND, VAL1, VAL2 )</tt></b></dt>
1962
1963 <dd>Perform the <a href="#i_select">select operation</a> on
Reid Spencer01c42592006-12-04 19:23:19 +0000[diff] [blame]1964 constants.</dd>
1965
1966 <dt><b><tt>icmp COND ( VAL1, VAL2 )</tt></b></dt>
1967 <dd>Performs the <a href="#i_icmp">icmp operation</a> on constants.</dd>
1968
1969 <dt><b><tt>fcmp COND ( VAL1, VAL2 )</tt></b></dt>
1970 <dd>Performs the <a href="#i_fcmp">fcmp operation</a> on constants.</dd>
Robert Bocchino9fbe1452006-01-10 19:31:34 +0000[diff] [blame]1971
Nate Begemanac80ade2008-05-12 19:01:56 +0000[diff] [blame]1972 <dt><b><tt>vicmp COND ( VAL1, VAL2 )</tt></b></dt>
1973 <dd>Performs the <a href="#i_vicmp">vicmp operation</a> on constants.</dd>
1974
1975 <dt><b><tt>vfcmp COND ( VAL1, VAL2 )</tt></b></dt>
1976 <dd>Performs the <a href="#i_vfcmp">vfcmp operation</a> on constants.</dd>
1977
Robert Bocchino9fbe1452006-01-10 19:31:34 +0000[diff] [blame]1978 <dt><b><tt>extractelement ( VAL, IDX )</tt></b></dt>
1979
1980 <dd>Perform the <a href="#i_extractelement">extractelement
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]1981 operation</a> on constants.</dd>
Robert Bocchino9fbe1452006-01-10 19:31:34 +0000[diff] [blame]1982
Robert Bocchino05ccd702006-01-15 20:48:27 +0000[diff] [blame]1983 <dt><b><tt>insertelement ( VAL, ELT, IDX )</tt></b></dt>
1984
1985 <dd>Perform the <a href="#i_insertelement">insertelement
Reid Spencer01c42592006-12-04 19:23:19 +0000[diff] [blame]1986 operation</a> on constants.</dd>
Robert Bocchino05ccd702006-01-15 20:48:27 +0000[diff] [blame]1987
Chris Lattnerc1989542006-04-08 00:13:41 +0000[diff] [blame]1988
1989 <dt><b><tt>shufflevector ( VEC1, VEC2, IDXMASK )</tt></b></dt>
1990
1991 <dd>Perform the <a href="#i_shufflevector">shufflevector
Reid Spencer01c42592006-12-04 19:23:19 +0000[diff] [blame]1992 operation</a> on constants.</dd>
Chris Lattnerc1989542006-04-08 00:13:41 +0000[diff] [blame]1993
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1994 <dt><b><tt>OPCODE ( LHS, RHS )</tt></b></dt>
1995
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]1996 <dd>Perform the specified operation of the LHS and RHS constants. OPCODE may
1997 be any of the <a href="#binaryops">binary</a> or <a href="#bitwiseops">bitwise
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1998 binary</a> operations. The constraints on operands are the same as those for
1999 the corresponding instruction (e.g. no bitwise operations on floating point
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]2000 values are allowed).</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2001</dl>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2002</div>
Chris Lattner9ee5d222004-03-08 16:49:10 +0000[diff] [blame]2003
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2004<!-- *********************************************************************** -->
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2005<div class="doc_section"> <a name="othervalues">Other Values</a> </div>
2006<!-- *********************************************************************** -->
2007
2008<!-- ======================================================================= -->
2009<div class="doc_subsection">
2010<a name="inlineasm">Inline Assembler Expressions</a>
2011</div>
2012
2013<div class="doc_text">
2014
2015<p>
2016LLVM supports inline assembler expressions (as opposed to <a href="#moduleasm">
2017Module-Level Inline Assembly</a>) through the use of a special value. This
2018value represents the inline assembler as a string (containing the instructions
2019to emit), a list of operand constraints (stored as a string), and a flag that
2020indicates whether or not the inline asm expression has side effects. An example
2021inline assembler expression is:
2022</p>
2023
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2024<div class="doc_code">
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2025<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2026i32 (i32) asm "bswap $0", "=r,r"
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2027</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2028</div>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2029
2030<p>
2031Inline assembler expressions may <b>only</b> be used as the callee operand of
2032a <a href="#i_call"><tt>call</tt> instruction</a>. Thus, typically we have:
2033</p>
2034
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2035<div class="doc_code">
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2036<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2037%X = call i32 asm "<a href="#int_bswap">bswap</a> $0", "=r,r"(i32 %Y)
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2038</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2039</div>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2040
2041<p>
2042Inline asms with side effects not visible in the constraint list must be marked
2043as having side effects. This is done through the use of the
2044'<tt>sideeffect</tt>' keyword, like so:
2045</p>
2046
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2047<div class="doc_code">
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2048<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2049call void asm sideeffect "eieio", ""()
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2050</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]2051</div>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2052
2053<p>TODO: The format of the asm and constraints string still need to be
2054documented here. Constraints on what can be done (e.g. duplication, moving, etc
Chris Lattner4f993352008-10-04 18:36:02 +0000[diff] [blame]2055need to be documented). This is probably best done by reference to another
2056document that covers inline asm from a holistic perspective.
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]2057</p>
2058
2059</div>
2060
2061<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2062<div class="doc_section"> <a name="instref">Instruction Reference</a> </div>
2063<!-- *********************************************************************** -->
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2064
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2065<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2066
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2067<p>The LLVM instruction set consists of several different
2068classifications of instructions: <a href="#terminators">terminator
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]2069instructions</a>, <a href="#binaryops">binary instructions</a>,
2070<a href="#bitwiseops">bitwise binary instructions</a>, <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2071 href="#memoryops">memory instructions</a>, and <a href="#otherops">other
2072instructions</a>.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2073
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2074</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2075
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2076<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2077<div class="doc_subsection"> <a name="terminators">Terminator
2078Instructions</a> </div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2079
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2080<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2081
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2082<p>As mentioned <a href="#functionstructure">previously</a>, every
2083basic block in a program ends with a "Terminator" instruction, which
2084indicates which block should be executed after the current block is
2085finished. These terminator instructions typically yield a '<tt>void</tt>'
2086value: they produce control flow, not values (the one exception being
2087the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction).</p>
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]2088<p>There are six different terminator instructions: the '<a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2089 href="#i_ret"><tt>ret</tt></a>' instruction, the '<a href="#i_br"><tt>br</tt></a>'
2090instruction, the '<a href="#i_switch"><tt>switch</tt></a>' instruction,
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]2091the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction, the '<a
2092 href="#i_unwind"><tt>unwind</tt></a>' instruction, and the '<a
2093 href="#i_unreachable"><tt>unreachable</tt></a>' instruction.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2094
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2095</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]2096
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2097<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2098<div class="doc_subsubsection"> <a name="i_ret">'<tt>ret</tt>'
2099Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2100<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2101<h5>Syntax:</h5>
Dan Gohmanb1e6b962008-10-04 19:00:07 +0000[diff] [blame]2102<pre>
2103 ret &lt;type&gt; &lt;value&gt; <i>; Return a value from a non-void function</i>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2104 ret void <i>; Return from void function</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2105</pre>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]2106
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2107<h5>Overview:</h5>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]2108
Dan Gohmanb1e6b962008-10-04 19:00:07 +0000[diff] [blame]2109<p>The '<tt>ret</tt>' instruction is used to return control flow (and
2110optionally a value) from a function back to the caller.</p>
John Criswell4457dc92004-04-09 16:48:45 +0000[diff] [blame]2111<p>There are two forms of the '<tt>ret</tt>' instruction: one that
Dan Gohmanb1e6b962008-10-04 19:00:07 +0000[diff] [blame]2112returns a value and then causes control flow, and one that just causes
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2113control flow to occur.</p>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]2114
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2115<h5>Arguments:</h5>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]2116
Dan Gohmanb1e6b962008-10-04 19:00:07 +0000[diff] [blame]2117<p>The '<tt>ret</tt>' instruction optionally accepts a single argument,
2118the return value. The type of the return value must be a
2119'<a href="#t_firstclass">first class</a>' type.</p>
2120
2121<p>A function is not <a href="#wellformed">well formed</a> if
2122it it has a non-void return type and contains a '<tt>ret</tt>'
2123instruction with no return value or a return value with a type that
2124does not match its type, or if it has a void return type and contains
2125a '<tt>ret</tt>' instruction with a return value.</p>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]2126
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2127<h5>Semantics:</h5>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]2128
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2129<p>When the '<tt>ret</tt>' instruction is executed, control flow
2130returns back to the calling function's context. If the caller is a "<a
John Criswellfa081872004-06-25 15:16:57 +0000[diff] [blame]2131 href="#i_call"><tt>call</tt></a>" instruction, execution continues at
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2132the instruction after the call. If the caller was an "<a
2133 href="#i_invoke"><tt>invoke</tt></a>" instruction, execution continues
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]2134at the beginning of the "normal" destination block. If the instruction
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2135returns a value, that value shall set the call or invoke instruction's
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]2136return value.</p>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]2137
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2138<h5>Example:</h5>
Chris Lattnerf4cde4e2008-04-23 04:59:35 +0000[diff] [blame]2139
2140<pre>
2141 ret i32 5 <i>; Return an integer value of 5</i>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2142 ret void <i>; Return from a void function</i>
Dan Gohmanb1e6b962008-10-04 19:00:07 +0000[diff] [blame]2143 ret { i32, i8 } { i32 4, i8 2 } <i>; Return an aggregate of values 4 and 2</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2144</pre>
Dan Gohmanf3e60bd2009-01-12 23:12:39 +0000[diff] [blame]2145
Dan Gohmand8791e52009-01-24 15:58:40 +0000[diff] [blame]2146<p>Note that the code generator does not yet fully support large
2147 return values. The specific sizes that are currently supported are
2148 dependent on the target. For integers, on 32-bit targets the limit
2149 is often 64 bits, and on 64-bit targets the limit is often 128 bits.
2150 For aggregate types, the current limits are dependent on the element
2151 types; for example targets are often limited to 2 total integer
2152 elements and 2 total floating-point elements.</p>
Dan Gohmanf3e60bd2009-01-12 23:12:39 +0000[diff] [blame]2153
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2154</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2155<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2156<div class="doc_subsubsection"> <a name="i_br">'<tt>br</tt>' Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2157<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2158<h5>Syntax:</h5>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]2159<pre> br i1 &lt;cond&gt;, label &lt;iftrue&gt;, label &lt;iffalse&gt;<br> br label &lt;dest&gt; <i>; Unconditional branch</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2160</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2161<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2162<p>The '<tt>br</tt>' instruction is used to cause control flow to
2163transfer to a different basic block in the current function. There are
2164two forms of this instruction, corresponding to a conditional branch
2165and an unconditional branch.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2166<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2167<p>The conditional branch form of the '<tt>br</tt>' instruction takes a
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]2168single '<tt>i1</tt>' value and two '<tt>label</tt>' values. The
Reid Spencerde151942007-02-19 23:54:10 +0000[diff] [blame]2169unconditional form of the '<tt>br</tt>' instruction takes a single
2170'<tt>label</tt>' value as a target.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2171<h5>Semantics:</h5>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]2172<p>Upon execution of a conditional '<tt>br</tt>' instruction, the '<tt>i1</tt>'
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2173argument is evaluated. If the value is <tt>true</tt>, control flows
2174to the '<tt>iftrue</tt>' <tt>label</tt> argument. If "cond" is <tt>false</tt>,
2175control flows to the '<tt>iffalse</tt>' <tt>label</tt> argument.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2176<h5>Example:</h5>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]2177<pre>Test:<br> %cond = <a href="#i_icmp">icmp</a> eq, i32 %a, %b<br> br i1 %cond, label %IfEqual, label %IfUnequal<br>IfEqual:<br> <a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2178 href="#i_ret">ret</a> i32 1<br>IfUnequal:<br> <a href="#i_ret">ret</a> i32 0<br></pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2179</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2180<!-- _______________________________________________________________________ -->
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2181<div class="doc_subsubsection">
2182 <a name="i_switch">'<tt>switch</tt>' Instruction</a>
2183</div>
2184
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2185<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2186<h5>Syntax:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2187
2188<pre>
2189 switch &lt;intty&gt; &lt;value&gt;, label &lt;defaultdest&gt; [ &lt;intty&gt; &lt;val&gt;, label &lt;dest&gt; ... ]
2190</pre>
2191
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2192<h5>Overview:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2193
2194<p>The '<tt>switch</tt>' instruction is used to transfer control flow to one of
2195several different places. It is a generalization of the '<tt>br</tt>'
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2196instruction, allowing a branch to occur to one of many possible
2197destinations.</p>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2198
2199
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2200<h5>Arguments:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2201
2202<p>The '<tt>switch</tt>' instruction uses three parameters: an integer
2203comparison value '<tt>value</tt>', a default '<tt>label</tt>' destination, and
2204an array of pairs of comparison value constants and '<tt>label</tt>'s. The
2205table is not allowed to contain duplicate constant entries.</p>
2206
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2207<h5>Semantics:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2208
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2209<p>The <tt>switch</tt> instruction specifies a table of values and
2210destinations. When the '<tt>switch</tt>' instruction is executed, this
John Criswell84114752004-06-25 16:05:06 +0000[diff] [blame]2211table is searched for the given value. If the value is found, control flow is
2212transfered to the corresponding destination; otherwise, control flow is
2213transfered to the default destination.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2214
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2215<h5>Implementation:</h5>
2216
2217<p>Depending on properties of the target machine and the particular
2218<tt>switch</tt> instruction, this instruction may be code generated in different
John Criswell84114752004-06-25 16:05:06 +0000[diff] [blame]2219ways. For example, it could be generated as a series of chained conditional
2220branches or with a lookup table.</p>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2221
2222<h5>Example:</h5>
2223
2224<pre>
2225 <i>; Emulate a conditional br instruction</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]2226 %Val = <a href="#i_zext">zext</a> i1 %value to i32
Dan Gohman2a08c532009-01-04 23:44:43 +0000[diff] [blame]2227 switch i32 %Val, label %truedest [ i32 0, label %falsedest ]
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2228
2229 <i>; Emulate an unconditional br instruction</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2230 switch i32 0, label %dest [ ]
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]2231
2232 <i>; Implement a jump table:</i>
Dan Gohman2a08c532009-01-04 23:44:43 +0000[diff] [blame]2233 switch i32 %val, label %otherwise [ i32 0, label %onzero
2234 i32 1, label %onone
2235 i32 2, label %ontwo ]
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2236</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2237</div>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2238
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2239<!-- _______________________________________________________________________ -->
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2240<div class="doc_subsubsection">
2241 <a name="i_invoke">'<tt>invoke</tt>' Instruction</a>
2242</div>
2243
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2244<div class="doc_text">
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2245
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2246<h5>Syntax:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2247
2248<pre>
Devang Patel307e8ab2008-10-07 17:48:33 +0000[diff] [blame]2249 &lt;result&gt; = invoke [<a href="#callingconv">cconv</a>] [<a href="#paramattrs">ret attrs</a>] &lt;ptr to function ty&gt; &lt;function ptr val&gt;(&lt;function args&gt;) [<a href="#fnattrs">fn attrs</a>]
Chris Lattner76b8a332006-05-14 18:23:06 +0000[diff] [blame]2250 to label &lt;normal label&gt; unwind label &lt;exception label&gt;
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2251</pre>
2252
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]2253<h5>Overview:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2254
2255<p>The '<tt>invoke</tt>' instruction causes control to transfer to a specified
2256function, with the possibility of control flow transfer to either the
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]2257'<tt>normal</tt>' label or the
2258'<tt>exception</tt>' label. If the callee function returns with the
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2259"<tt><a href="#i_ret">ret</a></tt>" instruction, control flow will return to the
2260"normal" label. If the callee (or any indirect callees) returns with the "<a
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]2261href="#i_unwind"><tt>unwind</tt></a>" instruction, control is interrupted and
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]2262continued at the dynamically nearest "exception" label.</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2263
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2264<h5>Arguments:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2265
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2266<p>This instruction requires several arguments:</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2267
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2268<ol>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2269 <li>
Duncan Sands8036ca42007-03-30 12:22:09 +0000[diff] [blame]2270 The optional "cconv" marker indicates which <a href="#callingconv">calling
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2271 convention</a> the call should use. If none is specified, the call defaults
2272 to using C calling conventions.
2273 </li>
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]2274
2275 <li>The optional <a href="#paramattrs">Parameter Attributes</a> list for
2276 return values. Only '<tt>zeroext</tt>', '<tt>signext</tt>',
2277 and '<tt>inreg</tt>' attributes are valid here.</li>
2278
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2279 <li>'<tt>ptr to function ty</tt>': shall be the signature of the pointer to
2280 function value being invoked. In most cases, this is a direct function
2281 invocation, but indirect <tt>invoke</tt>s are just as possible, branching off
2282 an arbitrary pointer to function value.
2283 </li>
2284
2285 <li>'<tt>function ptr val</tt>': An LLVM value containing a pointer to a
2286 function to be invoked. </li>
2287
2288 <li>'<tt>function args</tt>': argument list whose types match the function
2289 signature argument types. If the function signature indicates the function
2290 accepts a variable number of arguments, the extra arguments can be
2291 specified. </li>
2292
2293 <li>'<tt>normal label</tt>': the label reached when the called function
2294 executes a '<tt><a href="#i_ret">ret</a></tt>' instruction. </li>
2295
2296 <li>'<tt>exception label</tt>': the label reached when a callee returns with
2297 the <a href="#i_unwind"><tt>unwind</tt></a> instruction. </li>
2298
Devang Patel307e8ab2008-10-07 17:48:33 +0000[diff] [blame]2299 <li>The optional <a href="#fnattrs">function attributes</a> list. Only
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]2300 '<tt>noreturn</tt>', '<tt>nounwind</tt>', '<tt>readonly</tt>' and
2301 '<tt>readnone</tt>' attributes are valid here.</li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2302</ol>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2303
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2304<h5>Semantics:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2305
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2306<p>This instruction is designed to operate as a standard '<tt><a
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2307href="#i_call">call</a></tt>' instruction in most regards. The primary
2308difference is that it establishes an association with a label, which is used by
2309the runtime library to unwind the stack.</p>
2310
2311<p>This instruction is used in languages with destructors to ensure that proper
2312cleanup is performed in the case of either a <tt>longjmp</tt> or a thrown
2313exception. Additionally, this is important for implementation of
2314'<tt>catch</tt>' clauses in high-level languages that support them.</p>
2315
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2316<h5>Example:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]2317<pre>
Nick Lewyckyd703f652008-03-16 07:18:12 +0000[diff] [blame]2318 %retval = invoke i32 @Test(i32 15) to label %Continue
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]2319 unwind label %TestCleanup <i>; {i32}:retval set</i>
Nick Lewyckyd703f652008-03-16 07:18:12 +0000[diff] [blame]2320 %retval = invoke <a href="#callingconv">coldcc</a> i32 %Testfnptr(i32 15) to label %Continue
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]2321 unwind label %TestCleanup <i>; {i32}:retval set</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2322</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2323</div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]2324
2325
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]2326<!-- _______________________________________________________________________ -->
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]2327
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2328<div class="doc_subsubsection"> <a name="i_unwind">'<tt>unwind</tt>'
2329Instruction</a> </div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]2330
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2331<div class="doc_text">
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]2332
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]2333<h5>Syntax:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]2334<pre>
2335 unwind
2336</pre>
2337
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]2338<h5>Overview:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]2339
2340<p>The '<tt>unwind</tt>' instruction unwinds the stack, continuing control flow
2341at the first callee in the dynamic call stack which used an <a
2342href="#i_invoke"><tt>invoke</tt></a> instruction to perform the call. This is
2343primarily used to implement exception handling.</p>
2344
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]2345<h5>Semantics:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]2346
Chris Lattner72ed2002008-04-19 21:01:16 +0000[diff] [blame]2347<p>The '<tt>unwind</tt>' instruction causes execution of the current function to
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]2348immediately halt. The dynamic call stack is then searched for the first <a
2349href="#i_invoke"><tt>invoke</tt></a> instruction on the call stack. Once found,
2350execution continues at the "exceptional" destination block specified by the
2351<tt>invoke</tt> instruction. If there is no <tt>invoke</tt> instruction in the
2352dynamic call chain, undefined behavior results.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2353</div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]2354
2355<!-- _______________________________________________________________________ -->
2356
2357<div class="doc_subsubsection"> <a name="i_unreachable">'<tt>unreachable</tt>'
2358Instruction</a> </div>
2359
2360<div class="doc_text">
2361
2362<h5>Syntax:</h5>
2363<pre>
2364 unreachable
2365</pre>
2366
2367<h5>Overview:</h5>
2368
2369<p>The '<tt>unreachable</tt>' instruction has no defined semantics. This
2370instruction is used to inform the optimizer that a particular portion of the
2371code is not reachable. This can be used to indicate that the code after a
2372no-return function cannot be reached, and other facts.</p>
2373
2374<h5>Semantics:</h5>
2375
2376<p>The '<tt>unreachable</tt>' instruction has no defined semantics.</p>
2377</div>
2378
2379
2380
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2381<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2382<div class="doc_subsection"> <a name="binaryops">Binary Operations</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2383<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2384<p>Binary operators are used to do most of the computation in a
Chris Lattner5a158142008-04-01 18:47:32 +0000[diff] [blame]2385program. They require two operands of the same type, execute an operation on them, and
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]2386produce a single value. The operands might represent
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]2387multiple data, as is the case with the <a href="#t_vector">vector</a> data type.
Chris Lattner5a158142008-04-01 18:47:32 +0000[diff] [blame]2388The result value has the same type as its operands.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2389<p>There are several different binary operators:</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2390</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2391<!-- _______________________________________________________________________ -->
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2392<div class="doc_subsubsection">
2393 <a name="i_add">'<tt>add</tt>' Instruction</a>
2394</div>
2395
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2396<div class="doc_text">
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2397
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2398<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2399
2400<pre>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]2401 &lt;result&gt; = add &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2402</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2403
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2404<h5>Overview:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2405
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2406<p>The '<tt>add</tt>' instruction returns the sum of its two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2407
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2408<h5>Arguments:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2409
2410<p>The two arguments to the '<tt>add</tt>' instruction must be <a
2411 href="#t_integer">integer</a>, <a href="#t_floating">floating point</a>, or
2412 <a href="#t_vector">vector</a> values. Both arguments must have identical
2413 types.</p>
2414
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2415<h5>Semantics:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2416
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2417<p>The value produced is the integer or floating point sum of the two
2418operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2419
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]2420<p>If an integer sum has unsigned overflow, the result returned is the
2421mathematical result modulo 2<sup>n</sup>, where n is the bit width of
2422the result.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2423
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]2424<p>Because LLVM integers use a two's complement representation, this
2425instruction is appropriate for both signed and unsigned integers.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2426
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2427<h5>Example:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2428
2429<pre>
2430 &lt;result&gt; = add i32 4, %var <i>; yields {i32}:result = 4 + %var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2431</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2432</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2433<!-- _______________________________________________________________________ -->
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2434<div class="doc_subsubsection">
2435 <a name="i_sub">'<tt>sub</tt>' Instruction</a>
2436</div>
2437
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2438<div class="doc_text">
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2439
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2440<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2441
2442<pre>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]2443 &lt;result&gt; = sub &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2444</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2445
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2446<h5>Overview:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2447
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2448<p>The '<tt>sub</tt>' instruction returns the difference of its two
2449operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2450
2451<p>Note that the '<tt>sub</tt>' instruction is used to represent the
2452'<tt>neg</tt>' instruction present in most other intermediate
2453representations.</p>
2454
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2455<h5>Arguments:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2456
2457<p>The two arguments to the '<tt>sub</tt>' instruction must be <a
2458 href="#t_integer">integer</a>, <a href="#t_floating">floating point</a>,
2459 or <a href="#t_vector">vector</a> values. Both arguments must have identical
2460 types.</p>
2461
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2462<h5>Semantics:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2463
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2464<p>The value produced is the integer or floating point difference of
2465the two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2466
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]2467<p>If an integer difference has unsigned overflow, the result returned is the
2468mathematical result modulo 2<sup>n</sup>, where n is the bit width of
2469the result.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2470
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]2471<p>Because LLVM integers use a two's complement representation, this
2472instruction is appropriate for both signed and unsigned integers.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2473
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2474<h5>Example:</h5>
Bill Wendlingaac388b2007-05-29 09:42:13 +0000[diff] [blame]2475<pre>
2476 &lt;result&gt; = sub i32 4, %var <i>; yields {i32}:result = 4 - %var</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2477 &lt;result&gt; = sub i32 0, %val <i>; yields {i32}:result = -%var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2478</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2479</div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2480
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2481<!-- _______________________________________________________________________ -->
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2482<div class="doc_subsubsection">
2483 <a name="i_mul">'<tt>mul</tt>' Instruction</a>
2484</div>
2485
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2486<div class="doc_text">
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2487
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2488<h5>Syntax:</h5>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]2489<pre> &lt;result&gt; = mul &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2490</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2491<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2492<p>The '<tt>mul</tt>' instruction returns the product of its two
2493operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2494
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2495<h5>Arguments:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2496
2497<p>The two arguments to the '<tt>mul</tt>' instruction must be <a
2498href="#t_integer">integer</a>, <a href="#t_floating">floating point</a>,
2499or <a href="#t_vector">vector</a> values. Both arguments must have identical
2500types.</p>
2501
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2502<h5>Semantics:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2503
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2504<p>The value produced is the integer or floating point product of the
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2505two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2506
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]2507<p>If the result of an integer multiplication has unsigned overflow,
2508the result returned is the mathematical result modulo
25092<sup>n</sup>, where n is the bit width of the result.</p>
2510<p>Because LLVM integers use a two's complement representation, and the
2511result is the same width as the operands, this instruction returns the
2512correct result for both signed and unsigned integers. If a full product
2513(e.g. <tt>i32</tt>x<tt>i32</tt>-><tt>i64</tt>) is needed, the operands
2514should be sign-extended or zero-extended as appropriate to the
2515width of the full product.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2516<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2517<pre> &lt;result&gt; = mul i32 4, %var <i>; yields {i32}:result = 4 * %var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2518</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2519</div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2520
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2521<!-- _______________________________________________________________________ -->
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2522<div class="doc_subsubsection"> <a name="i_udiv">'<tt>udiv</tt>' Instruction
2523</a></div>
2524<div class="doc_text">
2525<h5>Syntax:</h5>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]2526<pre> &lt;result&gt; = udiv &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2527</pre>
2528<h5>Overview:</h5>
2529<p>The '<tt>udiv</tt>' instruction returns the quotient of its two
2530operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2531
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2532<h5>Arguments:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2533
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2534<p>The two arguments to the '<tt>udiv</tt>' instruction must be
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2535<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2536values. Both arguments must have identical types.</p>
2537
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2538<h5>Semantics:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2539
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]2540<p>The value produced is the unsigned integer quotient of the two operands.</p>
2541<p>Note that unsigned integer division and signed integer division are distinct
2542operations; for signed integer division, use '<tt>sdiv</tt>'.</p>
2543<p>Division by zero leads to undefined behavior.</p>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2544<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2545<pre> &lt;result&gt; = udiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2546</pre>
2547</div>
2548<!-- _______________________________________________________________________ -->
2549<div class="doc_subsubsection"> <a name="i_sdiv">'<tt>sdiv</tt>' Instruction
2550</a> </div>
2551<div class="doc_text">
2552<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2553<pre>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]2554 &lt;result&gt; = sdiv &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2555</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2556
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2557<h5>Overview:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2558
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2559<p>The '<tt>sdiv</tt>' instruction returns the quotient of its two
2560operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2561
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2562<h5>Arguments:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2563
2564<p>The two arguments to the '<tt>sdiv</tt>' instruction must be
2565<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2566values. Both arguments must have identical types.</p>
2567
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2568<h5>Semantics:</h5>
Chris Lattnera73afe02008-04-01 18:45:27 +0000[diff] [blame]2569<p>The value produced is the signed integer quotient of the two operands rounded towards zero.</p>
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]2570<p>Note that signed integer division and unsigned integer division are distinct
2571operations; for unsigned integer division, use '<tt>udiv</tt>'.</p>
2572<p>Division by zero leads to undefined behavior. Overflow also leads to
2573undefined behavior; this is a rare case, but can occur, for example,
2574by doing a 32-bit division of -2147483648 by -1.</p>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2575<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2576<pre> &lt;result&gt; = sdiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2577</pre>
2578</div>
2579<!-- _______________________________________________________________________ -->
2580<div class="doc_subsubsection"> <a name="i_fdiv">'<tt>fdiv</tt>'
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2581Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2582<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2583<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2584<pre>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]2585 &lt;result&gt; = fdiv &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2586</pre>
2587<h5>Overview:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2588
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2589<p>The '<tt>fdiv</tt>' instruction returns the quotient of its two
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2590operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2591
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2592<h5>Arguments:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2593
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]2594<p>The two arguments to the '<tt>fdiv</tt>' instruction must be
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2595<a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a>
2596of floating point values. Both arguments must have identical types.</p>
2597
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2598<h5>Semantics:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2599
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]2600<p>The value produced is the floating point quotient of the two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2601
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2602<h5>Example:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2603
2604<pre>
2605 &lt;result&gt; = fdiv float 4.0, %var <i>; yields {float}:result = 4.0 / %var</i>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2606</pre>
2607</div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2608
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2609<!-- _______________________________________________________________________ -->
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]2610<div class="doc_subsubsection"> <a name="i_urem">'<tt>urem</tt>' Instruction</a>
2611</div>
2612<div class="doc_text">
2613<h5>Syntax:</h5>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]2614<pre> &lt;result&gt; = urem &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]2615</pre>
2616<h5>Overview:</h5>
2617<p>The '<tt>urem</tt>' instruction returns the remainder from the
2618unsigned division of its two arguments.</p>
2619<h5>Arguments:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2620<p>The two arguments to the '<tt>urem</tt>' instruction must be
2621<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2622values. Both arguments must have identical types.</p>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]2623<h5>Semantics:</h5>
2624<p>This instruction returns the unsigned integer <i>remainder</i> of a division.
Chris Lattnera73afe02008-04-01 18:45:27 +0000[diff] [blame]2625This instruction always performs an unsigned division to get the remainder.</p>
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]2626<p>Note that unsigned integer remainder and signed integer remainder are
2627distinct operations; for signed integer remainder, use '<tt>srem</tt>'.</p>
2628<p>Taking the remainder of a division by zero leads to undefined behavior.</p>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]2629<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2630<pre> &lt;result&gt; = urem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]2631</pre>
2632
2633</div>
2634<!-- _______________________________________________________________________ -->
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2635<div class="doc_subsubsection">
2636 <a name="i_srem">'<tt>srem</tt>' Instruction</a>
2637</div>
2638
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2639<div class="doc_text">
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2640
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2641<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2642
2643<pre>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]2644 &lt;result&gt; = srem &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2645</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2646
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2647<h5>Overview:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2648
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]2649<p>The '<tt>srem</tt>' instruction returns the remainder from the
Dan Gohman80176312007-11-05 23:35:22 +0000[diff] [blame]2650signed division of its two operands. This instruction can also take
2651<a href="#t_vector">vector</a> versions of the values in which case
2652the elements must be integers.</p>
Chris Lattnerc7d3ab32008-01-04 04:33:49 +0000[diff] [blame]2653
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2654<h5>Arguments:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2655
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]2656<p>The two arguments to the '<tt>srem</tt>' instruction must be
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2657<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2658values. Both arguments must have identical types.</p>
2659
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2660<h5>Semantics:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2661
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]2662<p>This instruction returns the <i>remainder</i> of a division (where the result
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]2663has the same sign as the dividend, <tt>op1</tt>), not the <i>modulo</i>
2664operator (where the result has the same sign as the divisor, <tt>op2</tt>) of
Reid Spencerc9fdfc82007-03-24 22:23:39 +0000[diff] [blame]2665a value. For more information about the difference, see <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2666 href="http://mathforum.org/dr.math/problems/anne.4.28.99.html">The
Reid Spencerc9fdfc82007-03-24 22:23:39 +0000[diff] [blame]2667Math Forum</a>. For a table of how this is implemented in various languages,
Reid Spencer64f5c6c2007-03-24 22:40:44 +0000[diff] [blame]2668please see <a href="http://en.wikipedia.org/wiki/Modulo_operation">
Reid Spencerc9fdfc82007-03-24 22:23:39 +0000[diff] [blame]2669Wikipedia: modulo operation</a>.</p>
Chris Lattner5ec89832008-01-28 00:36:27 +0000[diff] [blame]2670<p>Note that signed integer remainder and unsigned integer remainder are
2671distinct operations; for unsigned integer remainder, use '<tt>urem</tt>'.</p>
2672<p>Taking the remainder of a division by zero leads to undefined behavior.
2673Overflow also leads to undefined behavior; this is a rare case, but can occur,
2674for example, by taking the remainder of a 32-bit division of -2147483648 by -1.
2675(The remainder doesn't actually overflow, but this rule lets srem be
2676implemented using instructions that return both the result of the division
2677and the remainder.)</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2678<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2679<pre> &lt;result&gt; = srem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]2680</pre>
2681
2682</div>
2683<!-- _______________________________________________________________________ -->
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2684<div class="doc_subsubsection">
2685 <a name="i_frem">'<tt>frem</tt>' Instruction</a> </div>
2686
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]2687<div class="doc_text">
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2688
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]2689<h5>Syntax:</h5>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]2690<pre> &lt;result&gt; = frem &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]2691</pre>
2692<h5>Overview:</h5>
2693<p>The '<tt>frem</tt>' instruction returns the remainder from the
2694division of its two operands.</p>
2695<h5>Arguments:</h5>
2696<p>The two arguments to the '<tt>frem</tt>' instruction must be
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2697<a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a>
2698of floating point values. Both arguments must have identical types.</p>
2699
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]2700<h5>Semantics:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2701
Chris Lattnera73afe02008-04-01 18:45:27 +0000[diff] [blame]2702<p>This instruction returns the <i>remainder</i> of a division.
2703The remainder has the same sign as the dividend.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2704
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]2705<h5>Example:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2706
2707<pre>
2708 &lt;result&gt; = frem float 4.0, %var <i>; yields {float}:result = 4.0 % %var</i>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2709</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2710</div>
Robert Bocchino7b81c752006-02-17 21:18:08 +0000[diff] [blame]2711
Reid Spencer8e11bf82007-02-02 13:57:07 +0000[diff] [blame]2712<!-- ======================================================================= -->
2713<div class="doc_subsection"> <a name="bitwiseops">Bitwise Binary
2714Operations</a> </div>
2715<div class="doc_text">
2716<p>Bitwise binary operators are used to do various forms of
2717bit-twiddling in a program. They are generally very efficient
2718instructions and can commonly be strength reduced from other
Chris Lattnera73afe02008-04-01 18:45:27 +0000[diff] [blame]2719instructions. They require two operands of the same type, execute an operation on them,
2720and produce a single value. The resulting value is the same type as its operands.</p>
Reid Spencer8e11bf82007-02-02 13:57:07 +0000[diff] [blame]2721</div>
2722
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2723<!-- _______________________________________________________________________ -->
2724<div class="doc_subsubsection"> <a name="i_shl">'<tt>shl</tt>'
2725Instruction</a> </div>
2726<div class="doc_text">
2727<h5>Syntax:</h5>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]2728<pre> &lt;result&gt; = shl &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2729</pre>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]2730
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2731<h5>Overview:</h5>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]2732
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2733<p>The '<tt>shl</tt>' instruction returns the first operand shifted to
2734the left a specified number of bits.</p>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]2735
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2736<h5>Arguments:</h5>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]2737
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2738<p>Both arguments to the '<tt>shl</tt>' instruction must be the same <a
Nate Begeman5bc1ea02008-07-29 15:49:41 +0000[diff] [blame]2739 href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]2740type. '<tt>op2</tt>' is treated as an unsigned value.</p>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]2741
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2742<h5>Semantics:</h5>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]2743
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]2744<p>The value produced is <tt>op1</tt> * 2<sup><tt>op2</tt></sup> mod 2<sup>n</sup>,
2745where n is the width of the result. If <tt>op2</tt> is (statically or dynamically) negative or
Mon P Wang01f8d092008-12-10 08:55:09 +0000[diff] [blame]2746equal to or larger than the number of bits in <tt>op1</tt>, the result is undefined.
2747If the arguments are vectors, each vector element of <tt>op1</tt> is shifted by the
2748corresponding shift amount in <tt>op2</tt>.</p>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]2749
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2750<h5>Example:</h5><pre>
2751 &lt;result&gt; = shl i32 4, %var <i>; yields {i32}: 4 &lt;&lt; %var</i>
2752 &lt;result&gt; = shl i32 4, 2 <i>; yields {i32}: 16</i>
2753 &lt;result&gt; = shl i32 1, 10 <i>; yields {i32}: 1024</i>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]2754 &lt;result&gt; = shl i32 1, 32 <i>; undefined</i>
Mon P Wange9f10152008-12-09 05:46:39 +0000[diff] [blame]2755 &lt;result&gt; = shl &lt;2 x i32&gt; &lt; i32 1, i32 1&gt;, &lt; i32 1, i32 2&gt; <i>; yields: result=&lt;2 x i32&gt; &lt; i32 2, i32 4&gt;</i>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2756</pre>
2757</div>
2758<!-- _______________________________________________________________________ -->
2759<div class="doc_subsubsection"> <a name="i_lshr">'<tt>lshr</tt>'
2760Instruction</a> </div>
2761<div class="doc_text">
2762<h5>Syntax:</h5>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]2763<pre> &lt;result&gt; = lshr &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2764</pre>
2765
2766<h5>Overview:</h5>
2767<p>The '<tt>lshr</tt>' instruction (logical shift right) returns the first
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]2768operand shifted to the right a specified number of bits with zero fill.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2769
2770<h5>Arguments:</h5>
2771<p>Both arguments to the '<tt>lshr</tt>' instruction must be the same
Nate Begeman5bc1ea02008-07-29 15:49:41 +0000[diff] [blame]2772<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]2773type. '<tt>op2</tt>' is treated as an unsigned value.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2774
2775<h5>Semantics:</h5>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]2776
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2777<p>This instruction always performs a logical shift right operation. The most
2778significant bits of the result will be filled with zero bits after the
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]2779shift. If <tt>op2</tt> is (statically or dynamically) equal to or larger than
Mon P Wang01f8d092008-12-10 08:55:09 +0000[diff] [blame]2780the number of bits in <tt>op1</tt>, the result is undefined. If the arguments are
2781vectors, each vector element of <tt>op1</tt> is shifted by the corresponding shift
2782amount in <tt>op2</tt>.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2783
2784<h5>Example:</h5>
2785<pre>
2786 &lt;result&gt; = lshr i32 4, 1 <i>; yields {i32}:result = 2</i>
2787 &lt;result&gt; = lshr i32 4, 2 <i>; yields {i32}:result = 1</i>
2788 &lt;result&gt; = lshr i8 4, 3 <i>; yields {i8}:result = 0</i>
2789 &lt;result&gt; = lshr i8 -2, 1 <i>; yields {i8}:result = 0x7FFFFFFF </i>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]2790 &lt;result&gt; = lshr i32 1, 32 <i>; undefined</i>
Mon P Wange9f10152008-12-09 05:46:39 +0000[diff] [blame]2791 &lt;result&gt; = lshr &lt;2 x i32&gt; &lt; i32 -2, i32 4&gt;, &lt; i32 1, i32 2&gt; <i>; yields: result=&lt;2 x i32&gt; &lt; i32 0x7FFFFFFF, i32 1&gt;</i>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2792</pre>
2793</div>
2794
Reid Spencer8e11bf82007-02-02 13:57:07 +0000[diff] [blame]2795<!-- _______________________________________________________________________ -->
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2796<div class="doc_subsubsection"> <a name="i_ashr">'<tt>ashr</tt>'
2797Instruction</a> </div>
2798<div class="doc_text">
2799
2800<h5>Syntax:</h5>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]2801<pre> &lt;result&gt; = ashr &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2802</pre>
2803
2804<h5>Overview:</h5>
2805<p>The '<tt>ashr</tt>' instruction (arithmetic shift right) returns the first
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]2806operand shifted to the right a specified number of bits with sign extension.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2807
2808<h5>Arguments:</h5>
2809<p>Both arguments to the '<tt>ashr</tt>' instruction must be the same
Nate Begeman5bc1ea02008-07-29 15:49:41 +0000[diff] [blame]2810<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]2811type. '<tt>op2</tt>' is treated as an unsigned value.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2812
2813<h5>Semantics:</h5>
2814<p>This instruction always performs an arithmetic shift right operation,
2815The most significant bits of the result will be filled with the sign bit
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]2816of <tt>op1</tt>. If <tt>op2</tt> is (statically or dynamically) equal to or
Mon P Wang01f8d092008-12-10 08:55:09 +0000[diff] [blame]2817larger than the number of bits in <tt>op1</tt>, the result is undefined. If the
2818arguments are vectors, each vector element of <tt>op1</tt> is shifted by the
2819corresponding shift amount in <tt>op2</tt>.</p>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2820
2821<h5>Example:</h5>
2822<pre>
2823 &lt;result&gt; = ashr i32 4, 1 <i>; yields {i32}:result = 2</i>
2824 &lt;result&gt; = ashr i32 4, 2 <i>; yields {i32}:result = 1</i>
2825 &lt;result&gt; = ashr i8 4, 3 <i>; yields {i8}:result = 0</i>
2826 &lt;result&gt; = ashr i8 -2, 1 <i>; yields {i8}:result = -1</i>
Chris Lattner6ccc2d52007-10-03 21:01:14 +0000[diff] [blame]2827 &lt;result&gt; = ashr i32 1, 32 <i>; undefined</i>
Mon P Wange9f10152008-12-09 05:46:39 +0000[diff] [blame]2828 &lt;result&gt; = ashr &lt;2 x i32&gt; &lt; i32 -2, i32 4&gt;, &lt; i32 1, i32 3&gt; <i>; yields: result=&lt;2 x i32&gt; &lt; i32 -1, i32 0&gt;</i>
Reid Spencer569f2fa2007-01-31 21:39:12 +0000[diff] [blame]2829</pre>
2830</div>
2831
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2832<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2833<div class="doc_subsubsection"> <a name="i_and">'<tt>and</tt>'
2834Instruction</a> </div>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2835
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2836<div class="doc_text">
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2837
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2838<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2839
2840<pre>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]2841 &lt;result&gt; = and &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2842</pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2843
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2844<h5>Overview:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2845
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2846<p>The '<tt>and</tt>' instruction returns the bitwise logical and of
2847its two operands.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2848
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2849<h5>Arguments:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2850
2851<p>The two arguments to the '<tt>and</tt>' instruction must be
2852<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2853values. Both arguments must have identical types.</p>
2854
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2855<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2856<p>The truth table used for the '<tt>and</tt>' instruction is:</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2857<p> </p>
Bill Wendlingc7e4c4d2008-09-07 10:29:20 +0000[diff] [blame]2858<div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2859<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2860 <tbody>
2861 <tr>
2862 <td>In0</td>
2863 <td>In1</td>
2864 <td>Out</td>
2865 </tr>
2866 <tr>
2867 <td>0</td>
2868 <td>0</td>
2869 <td>0</td>
2870 </tr>
2871 <tr>
2872 <td>0</td>
2873 <td>1</td>
2874 <td>0</td>
2875 </tr>
2876 <tr>
2877 <td>1</td>
2878 <td>0</td>
2879 <td>0</td>
2880 </tr>
2881 <tr>
2882 <td>1</td>
2883 <td>1</td>
2884 <td>1</td>
2885 </tr>
2886 </tbody>
2887</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]2888</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2889<h5>Example:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2890<pre>
2891 &lt;result&gt; = and i32 4, %var <i>; yields {i32}:result = 4 &amp; %var</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2892 &lt;result&gt; = and i32 15, 40 <i>; yields {i32}:result = 8</i>
2893 &lt;result&gt; = and i32 4, 8 <i>; yields {i32}:result = 0</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2894</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2895</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2896<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2897<div class="doc_subsubsection"> <a name="i_or">'<tt>or</tt>' Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2898<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2899<h5>Syntax:</h5>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]2900<pre> &lt;result&gt; = or &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2901</pre>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2902<h5>Overview:</h5>
2903<p>The '<tt>or</tt>' instruction returns the bitwise logical inclusive
2904or of its two operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2905<h5>Arguments:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2906
2907<p>The two arguments to the '<tt>or</tt>' instruction must be
2908<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2909values. Both arguments must have identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2910<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2911<p>The truth table used for the '<tt>or</tt>' instruction is:</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2912<p> </p>
Bill Wendlingc7e4c4d2008-09-07 10:29:20 +0000[diff] [blame]2913<div>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2914<table border="1" cellspacing="0" cellpadding="4">
2915 <tbody>
2916 <tr>
2917 <td>In0</td>
2918 <td>In1</td>
2919 <td>Out</td>
2920 </tr>
2921 <tr>
2922 <td>0</td>
2923 <td>0</td>
2924 <td>0</td>
2925 </tr>
2926 <tr>
2927 <td>0</td>
2928 <td>1</td>
2929 <td>1</td>
2930 </tr>
2931 <tr>
2932 <td>1</td>
2933 <td>0</td>
2934 <td>1</td>
2935 </tr>
2936 <tr>
2937 <td>1</td>
2938 <td>1</td>
2939 <td>1</td>
2940 </tr>
2941 </tbody>
2942</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]2943</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2944<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2945<pre> &lt;result&gt; = or i32 4, %var <i>; yields {i32}:result = 4 | %var</i>
2946 &lt;result&gt; = or i32 15, 40 <i>; yields {i32}:result = 47</i>
2947 &lt;result&gt; = or i32 4, 8 <i>; yields {i32}:result = 12</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2948</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2949</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2950<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2951<div class="doc_subsubsection"> <a name="i_xor">'<tt>xor</tt>'
2952Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2953<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2954<h5>Syntax:</h5>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]2955<pre> &lt;result&gt; = xor &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2956</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2957<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2958<p>The '<tt>xor</tt>' instruction returns the bitwise logical exclusive
2959or of its two operands. The <tt>xor</tt> is used to implement the
2960"one's complement" operation, which is the "~" operator in C.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2961<h5>Arguments:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2962<p>The two arguments to the '<tt>xor</tt>' instruction must be
2963<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2964values. Both arguments must have identical types.</p>
2965
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2966<h5>Semantics:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]2967
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2968<p>The truth table used for the '<tt>xor</tt>' instruction is:</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2969<p> </p>
Bill Wendlingc7e4c4d2008-09-07 10:29:20 +0000[diff] [blame]2970<div>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2971<table border="1" cellspacing="0" cellpadding="4">
2972 <tbody>
2973 <tr>
2974 <td>In0</td>
2975 <td>In1</td>
2976 <td>Out</td>
2977 </tr>
2978 <tr>
2979 <td>0</td>
2980 <td>0</td>
2981 <td>0</td>
2982 </tr>
2983 <tr>
2984 <td>0</td>
2985 <td>1</td>
2986 <td>1</td>
2987 </tr>
2988 <tr>
2989 <td>1</td>
2990 <td>0</td>
2991 <td>1</td>
2992 </tr>
2993 <tr>
2994 <td>1</td>
2995 <td>1</td>
2996 <td>0</td>
2997 </tr>
2998 </tbody>
2999</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]3000</div>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3001<p> </p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3002<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3003<pre> &lt;result&gt; = xor i32 4, %var <i>; yields {i32}:result = 4 ^ %var</i>
3004 &lt;result&gt; = xor i32 15, 40 <i>; yields {i32}:result = 39</i>
3005 &lt;result&gt; = xor i32 4, 8 <i>; yields {i32}:result = 12</i>
3006 &lt;result&gt; = xor i32 %V, -1 <i>; yields {i32}:result = ~%V</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3007</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3008</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3009
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3010<!-- ======================================================================= -->
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3011<div class="doc_subsection">
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3012 <a name="vectorops">Vector Operations</a>
3013</div>
3014
3015<div class="doc_text">
3016
3017<p>LLVM supports several instructions to represent vector operations in a
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]3018target-independent manner. These instructions cover the element-access and
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3019vector-specific operations needed to process vectors effectively. While LLVM
3020does directly support these vector operations, many sophisticated algorithms
3021will want to use target-specific intrinsics to take full advantage of a specific
3022target.</p>
3023
3024</div>
3025
3026<!-- _______________________________________________________________________ -->
3027<div class="doc_subsubsection">
3028 <a name="i_extractelement">'<tt>extractelement</tt>' Instruction</a>
3029</div>
3030
3031<div class="doc_text">
3032
3033<h5>Syntax:</h5>
3034
3035<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3036 &lt;result&gt; = extractelement &lt;n x &lt;ty&gt;&gt; &lt;val&gt;, i32 &lt;idx&gt; <i>; yields &lt;ty&gt;</i>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3037</pre>
3038
3039<h5>Overview:</h5>
3040
3041<p>
3042The '<tt>extractelement</tt>' instruction extracts a single scalar
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]3043element from a vector at a specified index.
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3044</p>
3045
3046
3047<h5>Arguments:</h5>
3048
3049<p>
3050The first operand of an '<tt>extractelement</tt>' instruction is a
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]3051value of <a href="#t_vector">vector</a> type. The second operand is
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3052an index indicating the position from which to extract the element.
3053The index may be a variable.</p>
3054
3055<h5>Semantics:</h5>
3056
3057<p>
3058The result is a scalar of the same type as the element type of
3059<tt>val</tt>. Its value is the value at position <tt>idx</tt> of
3060<tt>val</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
3061results are undefined.
3062</p>
3063
3064<h5>Example:</h5>
3065
3066<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3067 %result = extractelement &lt;4 x i32&gt; %vec, i32 0 <i>; yields i32</i>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3068</pre>
3069</div>
3070
3071
3072<!-- _______________________________________________________________________ -->
3073<div class="doc_subsubsection">
3074 <a name="i_insertelement">'<tt>insertelement</tt>' Instruction</a>
3075</div>
3076
3077<div class="doc_text">
3078
3079<h5>Syntax:</h5>
3080
3081<pre>
Dan Gohmanf3480b92008-05-12 23:38:42 +0000[diff] [blame]3082 &lt;result&gt; = insertelement &lt;n x &lt;ty&gt;&gt; &lt;val&gt;, &lt;ty&gt; &lt;elt&gt;, i32 &lt;idx&gt; <i>; yields &lt;n x &lt;ty&gt;&gt;</i>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3083</pre>
3084
3085<h5>Overview:</h5>
3086
3087<p>
3088The '<tt>insertelement</tt>' instruction inserts a scalar
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]3089element into a vector at a specified index.
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3090</p>
3091
3092
3093<h5>Arguments:</h5>
3094
3095<p>
3096The first operand of an '<tt>insertelement</tt>' instruction is a
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]3097value of <a href="#t_vector">vector</a> type. The second operand is a
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3098scalar value whose type must equal the element type of the first
3099operand. The third operand is an index indicating the position at
3100which to insert the value. The index may be a variable.</p>
3101
3102<h5>Semantics:</h5>
3103
3104<p>
Reid Spencer485bad12007-02-15 03:07:05 +0000[diff] [blame]3105The result is a vector of the same type as <tt>val</tt>. Its
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3106element values are those of <tt>val</tt> except at position
3107<tt>idx</tt>, where it gets the value <tt>elt</tt>. If <tt>idx</tt>
3108exceeds the length of <tt>val</tt>, the results are undefined.
3109</p>
3110
3111<h5>Example:</h5>
3112
3113<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3114 %result = insertelement &lt;4 x i32&gt; %vec, i32 1, i32 0 <i>; yields &lt;4 x i32&gt;</i>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3115</pre>
3116</div>
3117
3118<!-- _______________________________________________________________________ -->
3119<div class="doc_subsubsection">
3120 <a name="i_shufflevector">'<tt>shufflevector</tt>' Instruction</a>
3121</div>
3122
3123<div class="doc_text">
3124
3125<h5>Syntax:</h5>
3126
3127<pre>
Mon P Wangaeb06d22008-11-10 04:46:22 +0000[diff] [blame]3128 &lt;result&gt; = shufflevector &lt;n x &lt;ty&gt;&gt; &lt;v1&gt;, &lt;n x &lt;ty&gt;&gt; &lt;v2&gt;, &lt;m x i32&gt; &lt;mask&gt; <i>; yields &lt;m x &lt;ty&gt;&gt;</i>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3129</pre>
3130
3131<h5>Overview:</h5>
3132
3133<p>
3134The '<tt>shufflevector</tt>' instruction constructs a permutation of elements
Mon P Wangaeb06d22008-11-10 04:46:22 +0000[diff] [blame]3135from two input vectors, returning a vector with the same element type as
3136the input and length that is the same as the shuffle mask.
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3137</p>
3138
3139<h5>Arguments:</h5>
3140
3141<p>
Mon P Wangaeb06d22008-11-10 04:46:22 +0000[diff] [blame]3142The first two operands of a '<tt>shufflevector</tt>' instruction are vectors
3143with types that match each other. The third argument is a shuffle mask whose
3144element type is always 'i32'. The result of the instruction is a vector whose
3145length is the same as the shuffle mask and whose element type is the same as
3146the element type of the first two operands.
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3147</p>
3148
3149<p>
3150The shuffle mask operand is required to be a constant vector with either
3151constant integer or undef values.
3152</p>
3153
3154<h5>Semantics:</h5>
3155
3156<p>
3157The elements of the two input vectors are numbered from left to right across
3158both of the vectors. The shuffle mask operand specifies, for each element of
Mon P Wangaeb06d22008-11-10 04:46:22 +0000[diff] [blame]3159the result vector, which element of the two input vectors the result element
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3160gets. The element selector may be undef (meaning "don't care") and the second
3161operand may be undef if performing a shuffle from only one vector.
3162</p>
3163
3164<h5>Example:</h5>
3165
3166<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3167 %result = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; %v2,
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]3168 &lt;4 x i32&gt; &lt;i32 0, i32 4, i32 1, i32 5&gt; <i>; yields &lt;4 x i32&gt;</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3169 %result = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; undef,
3170 &lt;4 x i32&gt; &lt;i32 0, i32 1, i32 2, i32 3&gt; <i>; yields &lt;4 x i32&gt;</i> - Identity shuffle.
Mon P Wangaeb06d22008-11-10 04:46:22 +0000[diff] [blame]3171 %result = shufflevector &lt;8 x i32&gt; %v1, &lt;8 x i32&gt; undef,
3172 &lt;4 x i32&gt; &lt;i32 0, i32 1, i32 2, i32 3&gt; <i>; yields &lt;4 x i32&gt;</i>
3173 %result = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; %v2,
3174 &lt;8 x i32&gt; &lt;i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7 &gt; <i>; yields &lt;8 x i32&gt;</i>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3175</pre>
3176</div>
3177
Tanya Lattner09474292006-04-14 19:24:33 +0000[diff] [blame]3178
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]3179<!-- ======================================================================= -->
3180<div class="doc_subsection">
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]3181 <a name="aggregateops">Aggregate Operations</a>
3182</div>
3183
3184<div class="doc_text">
3185
3186<p>LLVM supports several instructions for working with aggregate values.
3187</p>
3188
3189</div>
3190
3191<!-- _______________________________________________________________________ -->
3192<div class="doc_subsubsection">
3193 <a name="i_extractvalue">'<tt>extractvalue</tt>' Instruction</a>
3194</div>
3195
3196<div class="doc_text">
3197
3198<h5>Syntax:</h5>
3199
3200<pre>
3201 &lt;result&gt; = extractvalue &lt;aggregate type&gt; &lt;val&gt;, &lt;idx&gt;{, &lt;idx&gt;}*
3202</pre>
3203
3204<h5>Overview:</h5>
3205
3206<p>
Dan Gohmanc3dac5c2008-05-13 18:16:06 +0000[diff] [blame]3207The '<tt>extractvalue</tt>' instruction extracts the value of a struct field
3208or array element from an aggregate value.
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]3209</p>
3210
3211
3212<h5>Arguments:</h5>
3213
3214<p>
3215The first operand of an '<tt>extractvalue</tt>' instruction is a
3216value of <a href="#t_struct">struct</a> or <a href="#t_array">array</a>
Dan Gohmanc3dac5c2008-05-13 18:16:06 +0000[diff] [blame]3217type. The operands are constant indices to specify which value to extract
Dan Gohman81a0c0b2008-05-31 00:58:22 +0000[diff] [blame]3218in a similar manner as indices in a
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]3219'<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.
3220</p>
3221
3222<h5>Semantics:</h5>
3223
3224<p>
3225The result is the value at the position in the aggregate specified by
3226the index operands.
3227</p>
3228
3229<h5>Example:</h5>
3230
3231<pre>
Dan Gohman81a0c0b2008-05-31 00:58:22 +0000[diff] [blame]3232 %result = extractvalue {i32, float} %agg, 0 <i>; yields i32</i>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]3233</pre>
3234</div>
3235
3236
3237<!-- _______________________________________________________________________ -->
3238<div class="doc_subsubsection">
3239 <a name="i_insertvalue">'<tt>insertvalue</tt>' Instruction</a>
3240</div>
3241
3242<div class="doc_text">
3243
3244<h5>Syntax:</h5>
3245
3246<pre>
Dan Gohman81a0c0b2008-05-31 00:58:22 +0000[diff] [blame]3247 &lt;result&gt; = insertvalue &lt;aggregate type&gt; &lt;val&gt;, &lt;ty&gt; &lt;val&gt;, &lt;idx&gt; <i>; yields &lt;n x &lt;ty&gt;&gt;</i>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]3248</pre>
3249
3250<h5>Overview:</h5>
3251
3252<p>
3253The '<tt>insertvalue</tt>' instruction inserts a value
Dan Gohmanc3dac5c2008-05-13 18:16:06 +0000[diff] [blame]3254into a struct field or array element in an aggregate.
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]3255</p>
3256
3257
3258<h5>Arguments:</h5>
3259
3260<p>
3261The first operand of an '<tt>insertvalue</tt>' instruction is a
3262value of <a href="#t_struct">struct</a> or <a href="#t_array">array</a> type.
3263The second operand is a first-class value to insert.
Dan Gohmanc4b49eb2008-05-23 21:53:15 +0000[diff] [blame]3264The following operands are constant indices
Dan Gohman81a0c0b2008-05-31 00:58:22 +0000[diff] [blame]3265indicating the position at which to insert the value in a similar manner as
Dan Gohmanc3dac5c2008-05-13 18:16:06 +0000[diff] [blame]3266indices in a
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]3267'<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.
3268The value to insert must have the same type as the value identified
Dan Gohmanc3dac5c2008-05-13 18:16:06 +0000[diff] [blame]3269by the indices.
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]3270</p>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]3271
3272<h5>Semantics:</h5>
3273
3274<p>
3275The result is an aggregate of the same type as <tt>val</tt>. Its
3276value is that of <tt>val</tt> except that the value at the position
Dan Gohmanc3dac5c2008-05-13 18:16:06 +0000[diff] [blame]3277specified by the indices is that of <tt>elt</tt>.
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]3278</p>
3279
3280<h5>Example:</h5>
3281
3282<pre>
Dan Gohman52bb2db2008-06-23 15:26:37 +0000[diff] [blame]3283 %result = insertvalue {i32, float} %agg, i32 1, 0 <i>; yields {i32, float}</i>
Dan Gohmana334d5f2008-05-12 23:51:09 +0000[diff] [blame]3284</pre>
3285</div>
3286
3287
3288<!-- ======================================================================= -->
3289<div class="doc_subsection">
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]3290 <a name="memoryops">Memory Access and Addressing Operations</a>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3291</div>
3292
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3293<div class="doc_text">
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3294
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3295<p>A key design point of an SSA-based representation is how it
3296represents memory. In LLVM, no memory locations are in SSA form, which
3297makes things very simple. This section describes how to read, write,
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]3298allocate, and free memory in LLVM.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3299
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3300</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3301
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3302<!-- _______________________________________________________________________ -->
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3303<div class="doc_subsubsection">
3304 <a name="i_malloc">'<tt>malloc</tt>' Instruction</a>
3305</div>
3306
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3307<div class="doc_text">
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3308
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3309<h5>Syntax:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3310
3311<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3312 &lt;result&gt; = malloc &lt;type&gt;[, i32 &lt;NumElements&gt;][, align &lt;alignment&gt;] <i>; yields {type*}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3313</pre>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3314
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3315<h5>Overview:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3316
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3317<p>The '<tt>malloc</tt>' instruction allocates memory from the system
Christopher Lamb303dae92007-12-17 01:00:21 +0000[diff] [blame]3318heap and returns a pointer to it. The object is always allocated in the generic
3319address space (address space zero).</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3320
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3321<h5>Arguments:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3322
3323<p>The '<tt>malloc</tt>' instruction allocates
3324<tt>sizeof(&lt;type&gt;)*NumElements</tt>
John Criswell6e4ca612004-02-24 16:13:56 +0000[diff] [blame]3325bytes of memory from the operating system and returns a pointer of the
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3326appropriate type to the program. If "NumElements" is specified, it is the
Gabor Greif1acd2ee2008-02-09 22:24:34 +0000[diff] [blame]3327number of elements allocated, otherwise "NumElements" is defaulted to be one.
Chris Lattner4316dec2008-04-02 00:38:26 +0000[diff] [blame]3328If a constant alignment is specified, the value result of the allocation is guaranteed to
Gabor Greif1acd2ee2008-02-09 22:24:34 +0000[diff] [blame]3329be aligned to at least that boundary. If not specified, or if zero, the target can
3330choose to align the allocation on any convenient boundary.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3331
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3332<p>'<tt>type</tt>' must be a sized type.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3333
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3334<h5>Semantics:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3335
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3336<p>Memory is allocated using the system "<tt>malloc</tt>" function, and
Nick Lewycky02ff3082008-11-24 03:41:24 +0000[diff] [blame]3337a pointer is returned. The result of a zero byte allocation is undefined. The
Chris Lattner72ed2002008-04-19 21:01:16 +0000[diff] [blame]3338result is null if there is insufficient memory available.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3339
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3340<h5>Example:</h5>
3341
3342<pre>
Dan Gohman81e21672009-01-04 23:49:44 +0000[diff] [blame]3343 %array = malloc [4 x i8] <i>; yields {[%4 x i8]*}:array</i>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3344
Bill Wendlingaac388b2007-05-29 09:42:13 +0000[diff] [blame]3345 %size = <a href="#i_add">add</a> i32 2, 2 <i>; yields {i32}:size = i32 4</i>
3346 %array1 = malloc i8, i32 4 <i>; yields {i8*}:array1</i>
3347 %array2 = malloc [12 x i8], i32 %size <i>; yields {[12 x i8]*}:array2</i>
3348 %array3 = malloc i32, i32 4, align 1024 <i>; yields {i32*}:array3</i>
3349 %array4 = malloc i32, align 1024 <i>; yields {i32*}:array4</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3350</pre>
Dan Gohmanf3e60bd2009-01-12 23:12:39 +0000[diff] [blame]3351
3352<p>Note that the code generator does not yet respect the
3353 alignment value.</p>
3354
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3355</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3356
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3357<!-- _______________________________________________________________________ -->
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3358<div class="doc_subsubsection">
3359 <a name="i_free">'<tt>free</tt>' Instruction</a>
3360</div>
3361
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3362<div class="doc_text">
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3363
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3364<h5>Syntax:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3365
3366<pre>
Dan Gohman81e21672009-01-04 23:49:44 +0000[diff] [blame]3367 free &lt;type&gt; &lt;value&gt; <i>; yields {void}</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3368</pre>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3369
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3370<h5>Overview:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3371
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3372<p>The '<tt>free</tt>' instruction returns memory back to the unused
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]3373memory heap to be reallocated in the future.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3374
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3375<h5>Arguments:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3376
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3377<p>'<tt>value</tt>' shall be a pointer value that points to a value
3378that was allocated with the '<tt><a href="#i_malloc">malloc</a></tt>'
3379instruction.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3380
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3381<h5>Semantics:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3382
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]3383<p>Access to the memory pointed to by the pointer is no longer defined
Chris Lattnere0db56d2008-04-19 22:41:32 +0000[diff] [blame]3384after this instruction executes. If the pointer is null, the operation
3385is a noop.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3386
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3387<h5>Example:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3388
3389<pre>
Dan Gohman81e21672009-01-04 23:49:44 +0000[diff] [blame]3390 %array = <a href="#i_malloc">malloc</a> [4 x i8] <i>; yields {[4 x i8]*}:array</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3391 free [4 x i8]* %array
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3392</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3393</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3394
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3395<!-- _______________________________________________________________________ -->
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3396<div class="doc_subsubsection">
3397 <a name="i_alloca">'<tt>alloca</tt>' Instruction</a>
3398</div>
3399
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3400<div class="doc_text">
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3401
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3402<h5>Syntax:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3403
3404<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3405 &lt;result&gt; = alloca &lt;type&gt;[, i32 &lt;NumElements&gt;][, align &lt;alignment&gt;] <i>; yields {type*}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3406</pre>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3407
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3408<h5>Overview:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3409
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]3410<p>The '<tt>alloca</tt>' instruction allocates memory on the stack frame of the
3411currently executing function, to be automatically released when this function
Christopher Lamb303dae92007-12-17 01:00:21 +0000[diff] [blame]3412returns to its caller. The object is always allocated in the generic address
3413space (address space zero).</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3414
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3415<h5>Arguments:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3416
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]3417<p>The '<tt>alloca</tt>' instruction allocates <tt>sizeof(&lt;type&gt;)*NumElements</tt>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3418bytes of memory on the runtime stack, returning a pointer of the
Gabor Greif1acd2ee2008-02-09 22:24:34 +0000[diff] [blame]3419appropriate type to the program. If "NumElements" is specified, it is the
3420number of elements allocated, otherwise "NumElements" is defaulted to be one.
Chris Lattner4316dec2008-04-02 00:38:26 +0000[diff] [blame]3421If a constant alignment is specified, the value result of the allocation is guaranteed
Gabor Greif1acd2ee2008-02-09 22:24:34 +0000[diff] [blame]3422to be aligned to at least that boundary. If not specified, or if zero, the target
3423can choose to align the allocation on any convenient boundary.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3424
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3425<p>'<tt>type</tt>' may be any sized type.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3426
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3427<h5>Semantics:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3428
Chris Lattner72ed2002008-04-19 21:01:16 +0000[diff] [blame]3429<p>Memory is allocated; a pointer is returned. The operation is undefiend if
3430there is insufficient stack space for the allocation. '<tt>alloca</tt>'d
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3431memory is automatically released when the function returns. The '<tt>alloca</tt>'
3432instruction is commonly used to represent automatic variables that must
3433have an address available. When the function returns (either with the <tt><a
John Criswelldae2e932005-05-12 16:55:34 +0000[diff] [blame]3434 href="#i_ret">ret</a></tt> or <tt><a href="#i_unwind">unwind</a></tt>
Chris Lattner4316dec2008-04-02 00:38:26 +0000[diff] [blame]3435instructions), the memory is reclaimed. Allocating zero bytes
3436is legal, but the result is undefined.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3437
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3438<h5>Example:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3439
3440<pre>
Dan Gohman81e21672009-01-04 23:49:44 +0000[diff] [blame]3441 %ptr = alloca i32 <i>; yields {i32*}:ptr</i>
3442 %ptr = alloca i32, i32 4 <i>; yields {i32*}:ptr</i>
3443 %ptr = alloca i32, i32 4, align 1024 <i>; yields {i32*}:ptr</i>
3444 %ptr = alloca i32, align 1024 <i>; yields {i32*}:ptr</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3445</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3446</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]3447
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3448<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3449<div class="doc_subsubsection"> <a name="i_load">'<tt>load</tt>'
3450Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3451<div class="doc_text">
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]3452<h5>Syntax:</h5>
Christopher Lamb2330e4d2007-04-21 08:16:25 +0000[diff] [blame]3453<pre> &lt;result&gt; = load &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;]<br> &lt;result&gt; = volatile load &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;]<br></pre>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]3454<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3455<p>The '<tt>load</tt>' instruction is used to read from memory.</p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]3456<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3457<p>The argument to the '<tt>load</tt>' instruction specifies the memory
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]3458address from which to load. The pointer must point to a <a
Chris Lattnere53e5082004-06-03 22:57:15 +0000[diff] [blame]3459 href="#t_firstclass">first class</a> type. If the <tt>load</tt> is
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]3460marked as <tt>volatile</tt>, then the optimizer is not allowed to modify
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3461the number or order of execution of this <tt>load</tt> with other
3462volatile <tt>load</tt> and <tt><a href="#i_store">store</a></tt>
3463instructions. </p>
Chris Lattnera31d1d72008-01-06 21:04:43 +0000[diff] [blame]3464<p>
Chris Lattner4316dec2008-04-02 00:38:26 +0000[diff] [blame]3465The optional constant "align" argument specifies the alignment of the operation
Chris Lattnera31d1d72008-01-06 21:04:43 +0000[diff] [blame]3466(that is, the alignment of the memory address). A value of 0 or an
3467omitted "align" argument means that the operation has the preferential
3468alignment for the target. It is the responsibility of the code emitter
3469to ensure that the alignment information is correct. Overestimating
3470the alignment results in an undefined behavior. Underestimating the
3471alignment may produce less efficient code. An alignment of 1 is always
3472safe.
3473</p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]3474<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3475<p>The location of memory pointed to is loaded.</p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]3476<h5>Examples:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3477<pre> %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3478 <a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3479 href="#i_store">store</a> i32 3, i32* %ptr <i>; yields {void}</i>
3480 %val = load i32* %ptr <i>; yields {i32}:val = i32 3</i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]3481</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3482</div>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]3483<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3484<div class="doc_subsubsection"> <a name="i_store">'<tt>store</tt>'
3485Instruction</a> </div>
Reid Spencer035ab572006-11-09 21:18:01 +0000[diff] [blame]3486<div class="doc_text">
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]3487<h5>Syntax:</h5>
Christopher Lamb2330e4d2007-04-21 08:16:25 +0000[diff] [blame]3488<pre> store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;] <i>; yields {void}</i>
3489 volatile store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;] <i>; yields {void}</i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]3490</pre>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]3491<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3492<p>The '<tt>store</tt>' instruction is used to write to memory.</p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]3493<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3494<p>There are two arguments to the '<tt>store</tt>' instruction: a value
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]3495to store and an address at which to store it. The type of the '<tt>&lt;pointer&gt;</tt>'
Chris Lattner4316dec2008-04-02 00:38:26 +0000[diff] [blame]3496operand must be a pointer to the <a href="#t_firstclass">first class</a> type
3497of the '<tt>&lt;value&gt;</tt>'
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]3498operand. If the <tt>store</tt> is marked as <tt>volatile</tt>, then the
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3499optimizer is not allowed to modify the number or order of execution of
3500this <tt>store</tt> with other volatile <tt>load</tt> and <tt><a
3501 href="#i_store">store</a></tt> instructions.</p>
Chris Lattnera31d1d72008-01-06 21:04:43 +0000[diff] [blame]3502<p>
Chris Lattner4316dec2008-04-02 00:38:26 +0000[diff] [blame]3503The optional constant "align" argument specifies the alignment of the operation
Chris Lattnera31d1d72008-01-06 21:04:43 +0000[diff] [blame]3504(that is, the alignment of the memory address). A value of 0 or an
3505omitted "align" argument means that the operation has the preferential
3506alignment for the target. It is the responsibility of the code emitter
3507to ensure that the alignment information is correct. Overestimating
3508the alignment results in an undefined behavior. Underestimating the
3509alignment may produce less efficient code. An alignment of 1 is always
3510safe.
3511</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3512<h5>Semantics:</h5>
3513<p>The contents of memory are updated to contain '<tt>&lt;value&gt;</tt>'
3514at the location specified by the '<tt>&lt;pointer&gt;</tt>' operand.</p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]3515<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3516<pre> %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
Bill Wendling8c6c72d2007-10-22 05:10:05 +0000[diff] [blame]3517 store i32 3, i32* %ptr <i>; yields {void}</i>
3518 %val = <a href="#i_load">load</a> i32* %ptr <i>; yields {i32}:val = i32 3</i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]3519</pre>
Reid Spencer47ce1792006-11-09 21:15:49 +0000[diff] [blame]3520</div>
3521
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]3522<!-- _______________________________________________________________________ -->
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]3523<div class="doc_subsubsection">
3524 <a name="i_getelementptr">'<tt>getelementptr</tt>' Instruction</a>
3525</div>
3526
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3527<div class="doc_text">
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]3528<h5>Syntax:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]3529<pre>
Matthijs Kooijmane49d0bc2008-10-13 13:44:15 +0000[diff] [blame]3530 &lt;result&gt; = getelementptr &lt;pty&gt;* &lt;ptrval&gt;{, &lt;ty&gt; &lt;idx&gt;}*
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]3531</pre>
3532
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]3533<h5>Overview:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]3534
3535<p>
3536The '<tt>getelementptr</tt>' instruction is used to get the address of a
Matthijs Kooijmane49d0bc2008-10-13 13:44:15 +0000[diff] [blame]3537subelement of an aggregate data structure. It performs address calculation only
3538and does not access memory.</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]3539
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]3540<h5>Arguments:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]3541
Matthijs Kooijmane49d0bc2008-10-13 13:44:15 +0000[diff] [blame]3542<p>The first argument is always a pointer, and forms the basis of the
3543calculation. The remaining arguments are indices, that indicate which of the
3544elements of the aggregate object are indexed. The interpretation of each index
3545is dependent on the type being indexed into. The first index always indexes the
3546pointer value given as the first argument, the second index indexes a value of
3547the type pointed to (not necessarily the value directly pointed to, since the
3548first index can be non-zero), etc. The first type indexed into must be a pointer
3549value, subsequent types can be arrays, vectors and structs. Note that subsequent
3550types being indexed into can never be pointers, since that would require loading
3551the pointer before continuing calculation.</p>
3552
3553<p>The type of each index argument depends on the type it is indexing into.
3554When indexing into a (packed) structure, only <tt>i32</tt> integer
3555<b>constants</b> are allowed. When indexing into an array, pointer or vector,
3556only integers of 32 or 64 bits are allowed (also non-constants). 32-bit values
3557will be sign extended to 64-bits if required.</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]3558
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3559<p>For example, let's consider a C code fragment and how it gets
3560compiled to LLVM:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]3561
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]3562<div class="doc_code">
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]3563<pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]3564struct RT {
3565 char A;
Chris Lattnercabc8462007-05-29 15:43:56 +0000[diff] [blame]3566 int B[10][20];
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]3567 char C;
3568};
3569struct ST {
Chris Lattnercabc8462007-05-29 15:43:56 +0000[diff] [blame]3570 int X;
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]3571 double Y;
3572 struct RT Z;
3573};
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]3574
Chris Lattnercabc8462007-05-29 15:43:56 +0000[diff] [blame]3575int *foo(struct ST *s) {
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]3576 return &amp;s[1].Z.B[5][13];
3577}
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]3578</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]3579</div>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]3580
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3581<p>The LLVM code generated by the GCC frontend is:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]3582
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]3583<div class="doc_code">
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]3584<pre>
Chris Lattnere7886e42009-01-11 20:53:49 +0000[diff] [blame]3585%RT = <a href="#namedtypes">type</a> { i8 , [10 x [20 x i32]], i8 }
3586%ST = <a href="#namedtypes">type</a> { i32, double, %RT }
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]3587
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]3588define i32* %foo(%ST* %s) {
3589entry:
3590 %reg = getelementptr %ST* %s, i32 1, i32 2, i32 1, i32 5, i32 13
3591 ret i32* %reg
3592}
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]3593</pre>
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]3594</div>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]3595
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]3596<h5>Semantics:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]3597
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3598<p>In the example above, the first index is indexing into the '<tt>%ST*</tt>'
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3599type, which is a pointer, yielding a '<tt>%ST</tt>' = '<tt>{ i32, double, %RT
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]3600}</tt>' type, a structure. The second index indexes into the third element of
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3601the structure, yielding a '<tt>%RT</tt>' = '<tt>{ i8 , [10 x [20 x i32]],
3602i8 }</tt>' type, another structure. The third index indexes into the second
3603element of the structure, yielding a '<tt>[10 x [20 x i32]]</tt>' type, an
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]3604array. The two dimensions of the array are subscripted into, yielding an
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3605'<tt>i32</tt>' type. The '<tt>getelementptr</tt>' instruction returns a pointer
3606to this element, thus computing a value of '<tt>i32*</tt>' type.</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]3607
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3608<p>Note that it is perfectly legal to index partially through a
3609structure, returning a pointer to an inner element. Because of this,
3610the LLVM code for the given testcase is equivalent to:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]3611
3612<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3613 define i32* %foo(%ST* %s) {
3614 %t1 = getelementptr %ST* %s, i32 1 <i>; yields %ST*:%t1</i>
Jeff Cohen6f1cc772007-04-22 01:17:39 +0000[diff] [blame]3615 %t2 = getelementptr %ST* %t1, i32 0, i32 2 <i>; yields %RT*:%t2</i>
3616 %t3 = getelementptr %RT* %t2, i32 0, i32 1 <i>; yields [10 x [20 x i32]]*:%t3</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3617 %t4 = getelementptr [10 x [20 x i32]]* %t3, i32 0, i32 5 <i>; yields [20 x i32]*:%t4</i>
3618 %t5 = getelementptr [20 x i32]* %t4, i32 0, i32 13 <i>; yields i32*:%t5</i>
3619 ret i32* %t5
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]3620 }
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]3621</pre>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]3622
3623<p>Note that it is undefined to access an array out of bounds: array and
3624pointer indexes must always be within the defined bounds of the array type.
Chris Lattner05d67092008-04-24 05:59:56 +0000[diff] [blame]3625The one exception for this rule is zero length arrays. These arrays are
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]3626defined to be accessible as variable length arrays, which requires access
3627beyond the zero'th element.</p>
3628
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]3629<p>The getelementptr instruction is often confusing. For some more insight
3630into how it works, see <a href="GetElementPtr.html">the getelementptr
3631FAQ</a>.</p>
3632
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]3633<h5>Example:</h5>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]3634
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]3635<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3636 <i>; yields [12 x i8]*:aptr</i>
Matthijs Kooijmane49d0bc2008-10-13 13:44:15 +0000[diff] [blame]3637 %aptr = getelementptr {i32, [12 x i8]}* %saptr, i64 0, i32 1
3638 <i>; yields i8*:vptr</i>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]3639 %vptr = getelementptr {i32, &lt;2 x i8&gt;}* %svptr, i64 0, i32 1, i32 1
Matthijs Kooijmane49d0bc2008-10-13 13:44:15 +0000[diff] [blame]3640 <i>; yields i8*:eptr</i>
3641 %eptr = getelementptr [12 x i8]* %aptr, i64 0, i32 1
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]3642</pre>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]3643</div>
Reid Spencer47ce1792006-11-09 21:15:49 +0000[diff] [blame]3644
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3645<!-- ======================================================================= -->
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]3646<div class="doc_subsection"> <a name="convertops">Conversion Operations</a>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3647</div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3648<div class="doc_text">
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]3649<p>The instructions in this category are the conversion instructions (casting)
3650which all take a single operand and a type. They perform various bit conversions
3651on the operand.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3652</div>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3653
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]3654<!-- _______________________________________________________________________ -->
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3655<div class="doc_subsubsection">
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3656 <a name="i_trunc">'<tt>trunc .. to</tt>' Instruction</a>
3657</div>
3658<div class="doc_text">
3659
3660<h5>Syntax:</h5>
3661<pre>
3662 &lt;result&gt; = trunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3663</pre>
3664
3665<h5>Overview:</h5>
3666<p>
3667The '<tt>trunc</tt>' instruction truncates its operand to the type <tt>ty2</tt>.
3668</p>
3669
3670<h5>Arguments:</h5>
3671<p>
3672The '<tt>trunc</tt>' instruction takes a <tt>value</tt> to trunc, which must
3673be an <a href="#t_integer">integer</a> type, and a type that specifies the size
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]3674and type of the result, which must be an <a href="#t_integer">integer</a>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3675type. The bit size of <tt>value</tt> must be larger than the bit size of
3676<tt>ty2</tt>. Equal sized types are not allowed.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3677
3678<h5>Semantics:</h5>
3679<p>
3680The '<tt>trunc</tt>' instruction truncates the high order bits in <tt>value</tt>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3681and converts the remaining bits to <tt>ty2</tt>. Since the source size must be
3682larger than the destination size, <tt>trunc</tt> cannot be a <i>no-op cast</i>.
3683It will always truncate bits.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3684
3685<h5>Example:</h5>
3686<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3687 %X = trunc i32 257 to i8 <i>; yields i8:1</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3688 %Y = trunc i32 123 to i1 <i>; yields i1:true</i>
3689 %Y = trunc i32 122 to i1 <i>; yields i1:false</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3690</pre>
3691</div>
3692
3693<!-- _______________________________________________________________________ -->
3694<div class="doc_subsubsection">
3695 <a name="i_zext">'<tt>zext .. to</tt>' Instruction</a>
3696</div>
3697<div class="doc_text">
3698
3699<h5>Syntax:</h5>
3700<pre>
3701 &lt;result&gt; = zext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3702</pre>
3703
3704<h5>Overview:</h5>
3705<p>The '<tt>zext</tt>' instruction zero extends its operand to type
3706<tt>ty2</tt>.</p>
3707
3708
3709<h5>Arguments:</h5>
3710<p>The '<tt>zext</tt>' instruction takes a value to cast, which must be of
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]3711<a href="#t_integer">integer</a> type, and a type to cast it to, which must
3712also be of <a href="#t_integer">integer</a> type. The bit size of the
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3713<tt>value</tt> must be smaller than the bit size of the destination type,
3714<tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3715
3716<h5>Semantics:</h5>
3717<p>The <tt>zext</tt> fills the high order bits of the <tt>value</tt> with zero
Chris Lattnerd1d25172007-05-24 19:13:27 +0000[diff] [blame]3718bits until it reaches the size of the destination type, <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3719
Reid Spencerb5929522007-01-12 15:46:11 +0000[diff] [blame]3720<p>When zero extending from i1, the result will always be either 0 or 1.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3721
3722<h5>Example:</h5>
3723<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3724 %X = zext i32 257 to i64 <i>; yields i64:257</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3725 %Y = zext i1 true to i32 <i>; yields i32:1</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3726</pre>
3727</div>
3728
3729<!-- _______________________________________________________________________ -->
3730<div class="doc_subsubsection">
3731 <a name="i_sext">'<tt>sext .. to</tt>' Instruction</a>
3732</div>
3733<div class="doc_text">
3734
3735<h5>Syntax:</h5>
3736<pre>
3737 &lt;result&gt; = sext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3738</pre>
3739
3740<h5>Overview:</h5>
3741<p>The '<tt>sext</tt>' sign extends <tt>value</tt> to the type <tt>ty2</tt>.</p>
3742
3743<h5>Arguments:</h5>
3744<p>
3745The '<tt>sext</tt>' instruction takes a value to cast, which must be of
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]3746<a href="#t_integer">integer</a> type, and a type to cast it to, which must
3747also be of <a href="#t_integer">integer</a> type. The bit size of the
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3748<tt>value</tt> must be smaller than the bit size of the destination type,
3749<tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3750
3751<h5>Semantics:</h5>
3752<p>
3753The '<tt>sext</tt>' instruction performs a sign extension by copying the sign
3754bit (highest order bit) of the <tt>value</tt> until it reaches the bit size of
Chris Lattnerd1d25172007-05-24 19:13:27 +0000[diff] [blame]3755the type <tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3756
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3757<p>When sign extending from i1, the extension always results in -1 or 0.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3758
3759<h5>Example:</h5>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3760<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3761 %X = sext i8 -1 to i16 <i>; yields i16 :65535</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3762 %Y = sext i1 true to i32 <i>; yields i32:-1</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3763</pre>
3764</div>
3765
3766<!-- _______________________________________________________________________ -->
3767<div class="doc_subsubsection">
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]3768 <a name="i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a>
3769</div>
3770
3771<div class="doc_text">
3772
3773<h5>Syntax:</h5>
3774
3775<pre>
3776 &lt;result&gt; = fptrunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3777</pre>
3778
3779<h5>Overview:</h5>
3780<p>The '<tt>fptrunc</tt>' instruction truncates <tt>value</tt> to type
3781<tt>ty2</tt>.</p>
3782
3783
3784<h5>Arguments:</h5>
3785<p>The '<tt>fptrunc</tt>' instruction takes a <a href="#t_floating">floating
3786 point</a> value to cast and a <a href="#t_floating">floating point</a> type to
3787cast it to. The size of <tt>value</tt> must be larger than the size of
3788<tt>ty2</tt>. This implies that <tt>fptrunc</tt> cannot be used to make a
3789<i>no-op cast</i>.</p>
3790
3791<h5>Semantics:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3792<p> The '<tt>fptrunc</tt>' instruction truncates a <tt>value</tt> from a larger
3793<a href="#t_floating">floating point</a> type to a smaller
3794<a href="#t_floating">floating point</a> type. If the value cannot fit within
3795the destination type, <tt>ty2</tt>, then the results are undefined.</p>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]3796
3797<h5>Example:</h5>
3798<pre>
3799 %X = fptrunc double 123.0 to float <i>; yields float:123.0</i>
3800 %Y = fptrunc double 1.0E+300 to float <i>; yields undefined</i>
3801</pre>
3802</div>
3803
3804<!-- _______________________________________________________________________ -->
3805<div class="doc_subsubsection">
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3806 <a name="i_fpext">'<tt>fpext .. to</tt>' Instruction</a>
3807</div>
3808<div class="doc_text">
3809
3810<h5>Syntax:</h5>
3811<pre>
3812 &lt;result&gt; = fpext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3813</pre>
3814
3815<h5>Overview:</h5>
3816<p>The '<tt>fpext</tt>' extends a floating point <tt>value</tt> to a larger
3817floating point value.</p>
3818
3819<h5>Arguments:</h5>
3820<p>The '<tt>fpext</tt>' instruction takes a
3821<a href="#t_floating">floating point</a> <tt>value</tt> to cast,
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3822and a <a href="#t_floating">floating point</a> type to cast it to. The source
3823type must be smaller than the destination type.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3824
3825<h5>Semantics:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3826<p>The '<tt>fpext</tt>' instruction extends the <tt>value</tt> from a smaller
Duncan Sands8036ca42007-03-30 12:22:09 +0000[diff] [blame]3827<a href="#t_floating">floating point</a> type to a larger
3828<a href="#t_floating">floating point</a> type. The <tt>fpext</tt> cannot be
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3829used to make a <i>no-op cast</i> because it always changes bits. Use
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]3830<tt>bitcast</tt> to make a <i>no-op cast</i> for a floating point cast.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3831
3832<h5>Example:</h5>
3833<pre>
3834 %X = fpext float 3.1415 to double <i>; yields double:3.1415</i>
3835 %Y = fpext float 1.0 to float <i>; yields float:1.0 (no-op)</i>
3836</pre>
3837</div>
3838
3839<!-- _______________________________________________________________________ -->
3840<div class="doc_subsubsection">
Reid Spencer24d6da52007-01-21 00:29:26 +0000[diff] [blame]3841 <a name="i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3842</div>
3843<div class="doc_text">
3844
3845<h5>Syntax:</h5>
3846<pre>
Reid Spencer1539a1c2007-07-31 14:40:14 +0000[diff] [blame]3847 &lt;result&gt; = fptoui &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3848</pre>
3849
3850<h5>Overview:</h5>
Reid Spencer1539a1c2007-07-31 14:40:14 +0000[diff] [blame]3851<p>The '<tt>fptoui</tt>' converts a floating point <tt>value</tt> to its
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3852unsigned integer equivalent of type <tt>ty2</tt>.
3853</p>
3854
3855<h5>Arguments:</h5>
Reid Spencer1539a1c2007-07-31 14:40:14 +0000[diff] [blame]3856<p>The '<tt>fptoui</tt>' instruction takes a value to cast, which must be a
Nate Begemanb348d182007-11-17 03:58:34 +0000[diff] [blame]3857scalar or vector <a href="#t_floating">floating point</a> value, and a type
3858to cast it to <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a>
3859type. If <tt>ty</tt> is a vector floating point type, <tt>ty2</tt> must be a
3860vector integer type with the same number of elements as <tt>ty</tt></p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3861
3862<h5>Semantics:</h5>
Reid Spencer1539a1c2007-07-31 14:40:14 +0000[diff] [blame]3863<p> The '<tt>fptoui</tt>' instruction converts its
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3864<a href="#t_floating">floating point</a> operand into the nearest (rounding
3865towards zero) unsigned integer value. If the value cannot fit in <tt>ty2</tt>,
3866the results are undefined.</p>
3867
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3868<h5>Example:</h5>
3869<pre>
Reid Spencer1539a1c2007-07-31 14:40:14 +0000[diff] [blame]3870 %X = fptoui double 123.0 to i32 <i>; yields i32:123</i>
Chris Lattner88519042007-09-22 03:17:52 +0000[diff] [blame]3871 %Y = fptoui float 1.0E+300 to i1 <i>; yields undefined:1</i>
Reid Spencer1539a1c2007-07-31 14:40:14 +0000[diff] [blame]3872 %X = fptoui float 1.04E+17 to i8 <i>; yields undefined:1</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3873</pre>
3874</div>
3875
3876<!-- _______________________________________________________________________ -->
3877<div class="doc_subsubsection">
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3878 <a name="i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3879</div>
3880<div class="doc_text">
3881
3882<h5>Syntax:</h5>
3883<pre>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3884 &lt;result&gt; = fptosi &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3885</pre>
3886
3887<h5>Overview:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3888<p>The '<tt>fptosi</tt>' instruction converts
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3889<a href="#t_floating">floating point</a> <tt>value</tt> to type <tt>ty2</tt>.
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3890</p>
3891
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]3892<h5>Arguments:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3893<p> The '<tt>fptosi</tt>' instruction takes a value to cast, which must be a
Nate Begemanb348d182007-11-17 03:58:34 +0000[diff] [blame]3894scalar or vector <a href="#t_floating">floating point</a> value, and a type
3895to cast it to <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a>
3896type. If <tt>ty</tt> is a vector floating point type, <tt>ty2</tt> must be a
3897vector integer type with the same number of elements as <tt>ty</tt></p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3898
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]3899<h5>Semantics:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3900<p>The '<tt>fptosi</tt>' instruction converts its
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3901<a href="#t_floating">floating point</a> operand into the nearest (rounding
3902towards zero) signed integer value. If the value cannot fit in <tt>ty2</tt>,
3903the results are undefined.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3904
Chris Lattner33ba0d92001-07-09 00:26:23 +0000[diff] [blame]3905<h5>Example:</h5>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3906<pre>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3907 %X = fptosi double -123.0 to i32 <i>; yields i32:-123</i>
Chris Lattner88519042007-09-22 03:17:52 +0000[diff] [blame]3908 %Y = fptosi float 1.0E-247 to i1 <i>; yields undefined:1</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3909 %X = fptosi float 1.04E+17 to i8 <i>; yields undefined:1</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3910</pre>
3911</div>
3912
3913<!-- _______________________________________________________________________ -->
3914<div class="doc_subsubsection">
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3915 <a name="i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3916</div>
3917<div class="doc_text">
3918
3919<h5>Syntax:</h5>
3920<pre>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3921 &lt;result&gt; = uitofp &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3922</pre>
3923
3924<h5>Overview:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3925<p>The '<tt>uitofp</tt>' instruction regards <tt>value</tt> as an unsigned
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3926integer and converts that value to the <tt>ty2</tt> type.</p>
3927
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3928<h5>Arguments:</h5>
Nate Begemanb348d182007-11-17 03:58:34 +0000[diff] [blame]3929<p>The '<tt>uitofp</tt>' instruction takes a value to cast, which must be a
3930scalar or vector <a href="#t_integer">integer</a> value, and a type to cast it
3931to <tt>ty2</tt>, which must be an <a href="#t_floating">floating point</a>
3932type. If <tt>ty</tt> is a vector integer type, <tt>ty2</tt> must be a vector
3933floating point type with the same number of elements as <tt>ty</tt></p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3934
3935<h5>Semantics:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3936<p>The '<tt>uitofp</tt>' instruction interprets its operand as an unsigned
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3937integer quantity and converts it to the corresponding floating point value. If
Jeff Cohencb757312007-04-22 14:56:37 +0000[diff] [blame]3938the value cannot fit in the floating point value, the results are undefined.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3939
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3940<h5>Example:</h5>
3941<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3942 %X = uitofp i32 257 to float <i>; yields float:257.0</i>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]3943 %Y = uitofp i8 -1 to double <i>; yields double:255.0</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3944</pre>
3945</div>
3946
3947<!-- _______________________________________________________________________ -->
3948<div class="doc_subsubsection">
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3949 <a name="i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3950</div>
3951<div class="doc_text">
3952
3953<h5>Syntax:</h5>
3954<pre>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3955 &lt;result&gt; = sitofp &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3956</pre>
3957
3958<h5>Overview:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3959<p>The '<tt>sitofp</tt>' instruction regards <tt>value</tt> as a signed
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3960integer and converts that value to the <tt>ty2</tt> type.</p>
3961
3962<h5>Arguments:</h5>
Nate Begemanb348d182007-11-17 03:58:34 +0000[diff] [blame]3963<p>The '<tt>sitofp</tt>' instruction takes a value to cast, which must be a
3964scalar or vector <a href="#t_integer">integer</a> value, and a type to cast it
3965to <tt>ty2</tt>, which must be an <a href="#t_floating">floating point</a>
3966type. If <tt>ty</tt> is a vector integer type, <tt>ty2</tt> must be a vector
3967floating point type with the same number of elements as <tt>ty</tt></p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3968
3969<h5>Semantics:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]3970<p>The '<tt>sitofp</tt>' instruction interprets its operand as a signed
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3971integer quantity and converts it to the corresponding floating point value. If
Jeff Cohencb757312007-04-22 14:56:37 +0000[diff] [blame]3972the value cannot fit in the floating point value, the results are undefined.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3973
3974<h5>Example:</h5>
3975<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3976 %X = sitofp i32 257 to float <i>; yields float:257.0</i>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]3977 %Y = sitofp i8 -1 to double <i>; yields double:-1.0</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3978</pre>
3979</div>
3980
3981<!-- _______________________________________________________________________ -->
3982<div class="doc_subsubsection">
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]3983 <a name="i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a>
3984</div>
3985<div class="doc_text">
3986
3987<h5>Syntax:</h5>
3988<pre>
3989 &lt;result&gt; = ptrtoint &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3990</pre>
3991
3992<h5>Overview:</h5>
3993<p>The '<tt>ptrtoint</tt>' instruction converts the pointer <tt>value</tt> to
3994the integer type <tt>ty2</tt>.</p>
3995
3996<h5>Arguments:</h5>
3997<p>The '<tt>ptrtoint</tt>' instruction takes a <tt>value</tt> to cast, which
Duncan Sands8036ca42007-03-30 12:22:09 +0000[diff] [blame]3998must be a <a href="#t_pointer">pointer</a> value, and a type to cast it to
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]3999<tt>ty2</tt>, which must be an <a href="#t_integer">integer</a> type.</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]4000
4001<h5>Semantics:</h5>
4002<p>The '<tt>ptrtoint</tt>' instruction converts <tt>value</tt> to integer type
4003<tt>ty2</tt> by interpreting the pointer value as an integer and either
4004truncating or zero extending that value to the size of the integer type. If
4005<tt>value</tt> is smaller than <tt>ty2</tt> then a zero extension is done. If
4006<tt>value</tt> is larger than <tt>ty2</tt> then a truncation is done. If they
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4007are the same size, then nothing is done (<i>no-op cast</i>) other than a type
4008change.</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]4009
4010<h5>Example:</h5>
4011<pre>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4012 %X = ptrtoint i32* %X to i8 <i>; yields truncation on 32-bit architecture</i>
4013 %Y = ptrtoint i32* %x to i64 <i>; yields zero extension on 32-bit architecture</i>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]4014</pre>
4015</div>
4016
4017<!-- _______________________________________________________________________ -->
4018<div class="doc_subsubsection">
4019 <a name="i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a>
4020</div>
4021<div class="doc_text">
4022
4023<h5>Syntax:</h5>
4024<pre>
4025 &lt;result&gt; = inttoptr &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4026</pre>
4027
4028<h5>Overview:</h5>
4029<p>The '<tt>inttoptr</tt>' instruction converts an integer <tt>value</tt> to
4030a pointer type, <tt>ty2</tt>.</p>
4031
4032<h5>Arguments:</h5>
Duncan Sands8036ca42007-03-30 12:22:09 +0000[diff] [blame]4033<p>The '<tt>inttoptr</tt>' instruction takes an <a href="#t_integer">integer</a>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]4034value to cast, and a type to cast it to, which must be a
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]4035<a href="#t_pointer">pointer</a> type.</p>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]4036
4037<h5>Semantics:</h5>
4038<p>The '<tt>inttoptr</tt>' instruction converts <tt>value</tt> to type
4039<tt>ty2</tt> by applying either a zero extension or a truncation depending on
4040the size of the integer <tt>value</tt>. If <tt>value</tt> is larger than the
4041size of a pointer then a truncation is done. If <tt>value</tt> is smaller than
4042the size of a pointer then a zero extension is done. If they are the same size,
4043nothing is done (<i>no-op cast</i>).</p>
4044
4045<h5>Example:</h5>
4046<pre>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4047 %X = inttoptr i32 255 to i32* <i>; yields zero extension on 64-bit architecture</i>
4048 %X = inttoptr i32 255 to i32* <i>; yields no-op on 32-bit architecture</i>
4049 %Y = inttoptr i64 0 to i32* <i>; yields truncation on 32-bit architecture</i>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]4050</pre>
4051</div>
4052
4053<!-- _______________________________________________________________________ -->
4054<div class="doc_subsubsection">
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]4055 <a name="i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4056</div>
4057<div class="doc_text">
4058
4059<h5>Syntax:</h5>
4060<pre>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]4061 &lt;result&gt; = bitcast &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4062</pre>
4063
4064<h5>Overview:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4065
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]4066<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4067<tt>ty2</tt> without changing any bits.</p>
4068
4069<h5>Arguments:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4070
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]4071<p>The '<tt>bitcast</tt>' instruction takes a value to cast, which must be
Dan Gohman500233a2008-09-08 16:45:59 +0000[diff] [blame]4072a non-aggregate first class value, and a type to cast it to, which must also be
4073a non-aggregate <a href="#t_firstclass">first class</a> type. The bit sizes of
4074<tt>value</tt>
Reid Spencer19b569f2007-01-09 20:08:58 +0000[diff] [blame]4075and the destination type, <tt>ty2</tt>, must be identical. If the source
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4076type is a pointer, the destination type must also be a pointer. This
4077instruction supports bitwise conversion of vectors to integers and to vectors
4078of other types (as long as they have the same size).</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4079
4080<h5>Semantics:</h5>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]4081<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]4082<tt>ty2</tt>. It is always a <i>no-op cast</i> because no bits change with
4083this conversion. The conversion is done as if the <tt>value</tt> had been
4084stored to memory and read back as type <tt>ty2</tt>. Pointer types may only be
4085converted to other pointer types with this instruction. To convert pointers to
4086other types, use the <a href="#i_inttoptr">inttoptr</a> or
4087<a href="#i_ptrtoint">ptrtoint</a> instructions first.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]4088
4089<h5>Example:</h5>
4090<pre>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4091 %X = bitcast i8 255 to i8 <i>; yields i8 :-1</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4092 %Y = bitcast i32* %x to sint* <i>; yields sint*:%x</i>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]4093 %Z = bitcast &lt;2 x int&gt; %V to i64; <i>; yields i64: %V</i>
Chris Lattner33ba0d92001-07-09 00:26:23 +0000[diff] [blame]4094</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4095</div>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]4096
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]4097<!-- ======================================================================= -->
4098<div class="doc_subsection"> <a name="otherops">Other Operations</a> </div>
4099<div class="doc_text">
4100<p>The instructions in this category are the "miscellaneous"
4101instructions, which defy better classification.</p>
4102</div>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4103
4104<!-- _______________________________________________________________________ -->
4105<div class="doc_subsubsection"><a name="i_icmp">'<tt>icmp</tt>' Instruction</a>
4106</div>
4107<div class="doc_text">
4108<h5>Syntax:</h5>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]4109<pre> &lt;result&gt; = icmp &lt;cond&gt; &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {i1} or {&lt;N x i1&gt;}:result</i>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4110</pre>
4111<h5>Overview:</h5>
Dan Gohmanf72fb672008-09-09 01:02:47 +0000[diff] [blame]4112<p>The '<tt>icmp</tt>' instruction returns a boolean value or
4113a vector of boolean values based on comparison
4114of its two integer, integer vector, or pointer operands.</p>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4115<h5>Arguments:</h5>
4116<p>The '<tt>icmp</tt>' instruction takes three operands. The first operand is
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4117the condition code indicating the kind of comparison to perform. It is not
4118a value, just a keyword. The possible condition code are:
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]4119</p>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4120<ol>
4121 <li><tt>eq</tt>: equal</li>
4122 <li><tt>ne</tt>: not equal </li>
4123 <li><tt>ugt</tt>: unsigned greater than</li>
4124 <li><tt>uge</tt>: unsigned greater or equal</li>
4125 <li><tt>ult</tt>: unsigned less than</li>
4126 <li><tt>ule</tt>: unsigned less or equal</li>
4127 <li><tt>sgt</tt>: signed greater than</li>
4128 <li><tt>sge</tt>: signed greater or equal</li>
4129 <li><tt>slt</tt>: signed less than</li>
4130 <li><tt>sle</tt>: signed less or equal</li>
4131</ol>
Chris Lattner3b19d652007-01-15 01:54:13 +0000[diff] [blame]4132<p>The remaining two arguments must be <a href="#t_integer">integer</a> or
Dan Gohmanf72fb672008-09-09 01:02:47 +0000[diff] [blame]4133<a href="#t_pointer">pointer</a>
4134or integer <a href="#t_vector">vector</a> typed.
4135They must also be identical types.</p>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4136<h5>Semantics:</h5>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4137<p>The '<tt>icmp</tt>' compares <tt>op1</tt> and <tt>op2</tt> according to
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4138the condition code given as <tt>cond</tt>. The comparison performed always
Dan Gohmanf72fb672008-09-09 01:02:47 +0000[diff] [blame]4139yields either an <a href="#t_primitive"><tt>i1</tt></a> or vector of <tt>i1</tt> result, as follows:
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]4140</p>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4141<ol>
4142 <li><tt>eq</tt>: yields <tt>true</tt> if the operands are equal,
4143 <tt>false</tt> otherwise. No sign interpretation is necessary or performed.
4144 </li>
4145 <li><tt>ne</tt>: yields <tt>true</tt> if the operands are unequal,
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]4146 <tt>false</tt> otherwise. No sign interpretation is necessary or performed.</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4147 <li><tt>ugt</tt>: interprets the operands as unsigned values and yields
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4148 <tt>true</tt> if <tt>op1</tt> is greater than <tt>op2</tt>.</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4149 <li><tt>uge</tt>: interprets the operands as unsigned values and yields
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4150 <tt>true</tt> if <tt>op1</tt> is greater than or equal to <tt>op2</tt>.</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4151 <li><tt>ult</tt>: interprets the operands as unsigned values and yields
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4152 <tt>true</tt> if <tt>op1</tt> is less than <tt>op2</tt>.</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4153 <li><tt>ule</tt>: interprets the operands as unsigned values and yields
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4154 <tt>true</tt> if <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4155 <li><tt>sgt</tt>: interprets the operands as signed values and yields
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4156 <tt>true</tt> if <tt>op1</tt> is greater than <tt>op2</tt>.</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4157 <li><tt>sge</tt>: interprets the operands as signed values and yields
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4158 <tt>true</tt> if <tt>op1</tt> is greater than or equal to <tt>op2</tt>.</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4159 <li><tt>slt</tt>: interprets the operands as signed values and yields
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4160 <tt>true</tt> if <tt>op1</tt> is less than <tt>op2</tt>.</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4161 <li><tt>sle</tt>: interprets the operands as signed values and yields
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4162 <tt>true</tt> if <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4163</ol>
4164<p>If the operands are <a href="#t_pointer">pointer</a> typed, the pointer
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4165values are compared as if they were integers.</p>
Dan Gohmanf72fb672008-09-09 01:02:47 +0000[diff] [blame]4166<p>If the operands are integer vectors, then they are compared
4167element by element. The result is an <tt>i1</tt> vector with
4168the same number of elements as the values being compared.
4169Otherwise, the result is an <tt>i1</tt>.
4170</p>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4171
4172<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4173<pre> &lt;result&gt; = icmp eq i32 4, 5 <i>; yields: result=false</i>
4174 &lt;result&gt; = icmp ne float* %X, %X <i>; yields: result=false</i>
4175 &lt;result&gt; = icmp ult i16 4, 5 <i>; yields: result=true</i>
4176 &lt;result&gt; = icmp sgt i16 4, 5 <i>; yields: result=false</i>
4177 &lt;result&gt; = icmp ule i16 -4, 5 <i>; yields: result=false</i>
4178 &lt;result&gt; = icmp sge i16 4, 5 <i>; yields: result=false</i>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4179</pre>
Dan Gohmanc766f722009-01-22 01:39:38 +0000[diff] [blame]4180
4181<p>Note that the code generator does not yet support vector types with
4182 the <tt>icmp</tt> instruction.</p>
4183
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4184</div>
4185
4186<!-- _______________________________________________________________________ -->
4187<div class="doc_subsubsection"><a name="i_fcmp">'<tt>fcmp</tt>' Instruction</a>
4188</div>
4189<div class="doc_text">
4190<h5>Syntax:</h5>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]4191<pre> &lt;result&gt; = fcmp &lt;cond&gt; &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {i1} or {&lt;N x i1&gt;}:result</i>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4192</pre>
4193<h5>Overview:</h5>
Dan Gohmanf72fb672008-09-09 01:02:47 +0000[diff] [blame]4194<p>The '<tt>fcmp</tt>' instruction returns a boolean value
4195or vector of boolean values based on comparison
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]4196of its operands.</p>
Dan Gohmanf72fb672008-09-09 01:02:47 +0000[diff] [blame]4197<p>
4198If the operands are floating point scalars, then the result
4199type is a boolean (<a href="#t_primitive"><tt>i1</tt></a>).
4200</p>
4201<p>If the operands are floating point vectors, then the result type
4202is a vector of boolean with the same number of elements as the
4203operands being compared.</p>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4204<h5>Arguments:</h5>
4205<p>The '<tt>fcmp</tt>' instruction takes three operands. The first operand is
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4206the condition code indicating the kind of comparison to perform. It is not
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]4207a value, just a keyword. The possible condition code are:</p>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4208<ol>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]4209 <li><tt>false</tt>: no comparison, always returns false</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4210 <li><tt>oeq</tt>: ordered and equal</li>
4211 <li><tt>ogt</tt>: ordered and greater than </li>
4212 <li><tt>oge</tt>: ordered and greater than or equal</li>
4213 <li><tt>olt</tt>: ordered and less than </li>
4214 <li><tt>ole</tt>: ordered and less than or equal</li>
4215 <li><tt>one</tt>: ordered and not equal</li>
4216 <li><tt>ord</tt>: ordered (no nans)</li>
4217 <li><tt>ueq</tt>: unordered or equal</li>
4218 <li><tt>ugt</tt>: unordered or greater than </li>
4219 <li><tt>uge</tt>: unordered or greater than or equal</li>
4220 <li><tt>ult</tt>: unordered or less than </li>
4221 <li><tt>ule</tt>: unordered or less than or equal</li>
4222 <li><tt>une</tt>: unordered or not equal</li>
4223 <li><tt>uno</tt>: unordered (either nans)</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]4224 <li><tt>true</tt>: no comparison, always returns true</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4225</ol>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4226<p><i>Ordered</i> means that neither operand is a QNAN while
Reid Spencer93a49852006-12-06 07:08:07 +0000[diff] [blame]4227<i>unordered</i> means that either operand may be a QNAN.</p>
Dan Gohmanf72fb672008-09-09 01:02:47 +0000[diff] [blame]4228<p>Each of <tt>val1</tt> and <tt>val2</tt> arguments must be
4229either a <a href="#t_floating">floating point</a> type
4230or a <a href="#t_vector">vector</a> of floating point type.
4231They must have identical types.</p>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4232<h5>Semantics:</h5>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4233<p>The '<tt>fcmp</tt>' instruction compares <tt>op1</tt> and <tt>op2</tt>
Dan Gohmanf72fb672008-09-09 01:02:47 +0000[diff] [blame]4234according to the condition code given as <tt>cond</tt>.
4235If the operands are vectors, then the vectors are compared
4236element by element.
4237Each comparison performed
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]4238always yields an <a href="#t_primitive">i1</a> result, as follows:</p>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4239<ol>
4240 <li><tt>false</tt>: always yields <tt>false</tt>, regardless of operands.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]4241 <li><tt>oeq</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4242 <tt>op1</tt> is equal to <tt>op2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]4243 <li><tt>ogt</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4244 <tt>op1</tt> is greather than <tt>op2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]4245 <li><tt>oge</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4246 <tt>op1</tt> is greater than or equal to <tt>op2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]4247 <li><tt>olt</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4248 <tt>op1</tt> is less than <tt>op2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]4249 <li><tt>ole</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4250 <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]4251 <li><tt>one</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4252 <tt>op1</tt> is not equal to <tt>op2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]4253 <li><tt>ord</tt>: yields <tt>true</tt> if both operands are not a QNAN.</li>
4254 <li><tt>ueq</tt>: yields <tt>true</tt> if either operand is a QNAN or
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4255 <tt>op1</tt> is equal to <tt>op2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]4256 <li><tt>ugt</tt>: yields <tt>true</tt> if either operand is a QNAN or
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4257 <tt>op1</tt> is greater than <tt>op2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]4258 <li><tt>uge</tt>: yields <tt>true</tt> if either operand is a QNAN or
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4259 <tt>op1</tt> is greater than or equal to <tt>op2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]4260 <li><tt>ult</tt>: yields <tt>true</tt> if either operand is a QNAN or
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4261 <tt>op1</tt> is less than <tt>op2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]4262 <li><tt>ule</tt>: yields <tt>true</tt> if either operand is a QNAN or
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4263 <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]4264 <li><tt>une</tt>: yields <tt>true</tt> if either operand is a QNAN or
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4265 <tt>op1</tt> is not equal to <tt>op2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]4266 <li><tt>uno</tt>: yields <tt>true</tt> if either operand is a QNAN.</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4267 <li><tt>true</tt>: always yields <tt>true</tt>, regardless of operands.</li>
4268</ol>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4269
4270<h5>Example:</h5>
4271<pre> &lt;result&gt; = fcmp oeq float 4.0, 5.0 <i>; yields: result=false</i>
Dan Gohmanf72fb672008-09-09 01:02:47 +0000[diff] [blame]4272 &lt;result&gt; = fcmp one float 4.0, 5.0 <i>; yields: result=true</i>
4273 &lt;result&gt; = fcmp olt float 4.0, 5.0 <i>; yields: result=true</i>
4274 &lt;result&gt; = fcmp ueq double 1.0, 2.0 <i>; yields: result=false</i>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4275</pre>
Dan Gohmanc766f722009-01-22 01:39:38 +0000[diff] [blame]4276
4277<p>Note that the code generator does not yet support vector types with
4278 the <tt>fcmp</tt> instruction.</p>
4279
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]4280</div>
4281
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]4282<!-- _______________________________________________________________________ -->
Nate Begemanac80ade2008-05-12 19:01:56 +0000[diff] [blame]4283<div class="doc_subsubsection">
4284 <a name="i_vicmp">'<tt>vicmp</tt>' Instruction</a>
4285</div>
4286<div class="doc_text">
4287<h5>Syntax:</h5>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4288<pre> &lt;result&gt; = vicmp &lt;cond&gt; &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Nate Begemanac80ade2008-05-12 19:01:56 +0000[diff] [blame]4289</pre>
4290<h5>Overview:</h5>
4291<p>The '<tt>vicmp</tt>' instruction returns an integer vector value based on
4292element-wise comparison of its two integer vector operands.</p>
4293<h5>Arguments:</h5>
4294<p>The '<tt>vicmp</tt>' instruction takes three operands. The first operand is
4295the condition code indicating the kind of comparison to perform. It is not
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]4296a value, just a keyword. The possible condition code are:</p>
Nate Begemanac80ade2008-05-12 19:01:56 +0000[diff] [blame]4297<ol>
4298 <li><tt>eq</tt>: equal</li>
4299 <li><tt>ne</tt>: not equal </li>
4300 <li><tt>ugt</tt>: unsigned greater than</li>
4301 <li><tt>uge</tt>: unsigned greater or equal</li>
4302 <li><tt>ult</tt>: unsigned less than</li>
4303 <li><tt>ule</tt>: unsigned less or equal</li>
4304 <li><tt>sgt</tt>: signed greater than</li>
4305 <li><tt>sge</tt>: signed greater or equal</li>
4306 <li><tt>slt</tt>: signed less than</li>
4307 <li><tt>sle</tt>: signed less or equal</li>
4308</ol>
Dan Gohmanf72fb672008-09-09 01:02:47 +0000[diff] [blame]4309<p>The remaining two arguments must be <a href="#t_vector">vector</a> or
Nate Begemanac80ade2008-05-12 19:01:56 +0000[diff] [blame]4310<a href="#t_integer">integer</a> typed. They must also be identical types.</p>
4311<h5>Semantics:</h5>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4312<p>The '<tt>vicmp</tt>' instruction compares <tt>op1</tt> and <tt>op2</tt>
Nate Begemanac80ade2008-05-12 19:01:56 +0000[diff] [blame]4313according to the condition code given as <tt>cond</tt>. The comparison yields a
4314<a href="#t_vector">vector</a> of <a href="#t_integer">integer</a> result, of
4315identical type as the values being compared. The most significant bit in each
4316element is 1 if the element-wise comparison evaluates to true, and is 0
4317otherwise. All other bits of the result are undefined. The condition codes
4318are evaluated identically to the <a href="#i_icmp">'<tt>icmp</tt>'
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]4319instruction</a>.</p>
Nate Begemanac80ade2008-05-12 19:01:56 +0000[diff] [blame]4320
4321<h5>Example:</h5>
4322<pre>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4323 &lt;result&gt; = vicmp eq &lt;2 x i32&gt; &lt; i32 4, i32 0&gt;, &lt; i32 5, i32 0&gt; <i>; yields: result=&lt;2 x i32&gt; &lt; i32 0, i32 -1 &gt;</i>
4324 &lt;result&gt; = vicmp ult &lt;2 x i8 &gt; &lt; i8 1, i8 2&gt;, &lt; i8 2, i8 2 &gt; <i>; yields: result=&lt;2 x i8&gt; &lt; i8 -1, i8 0 &gt;</i>
Nate Begemanac80ade2008-05-12 19:01:56 +0000[diff] [blame]4325</pre>
4326</div>
4327
4328<!-- _______________________________________________________________________ -->
4329<div class="doc_subsubsection">
4330 <a name="i_vfcmp">'<tt>vfcmp</tt>' Instruction</a>
4331</div>
4332<div class="doc_text">
4333<h5>Syntax:</h5>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4334<pre> &lt;result&gt; = vfcmp &lt;cond&gt; &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt;</pre>
Nate Begemanac80ade2008-05-12 19:01:56 +0000[diff] [blame]4335<h5>Overview:</h5>
4336<p>The '<tt>vfcmp</tt>' instruction returns an integer vector value based on
4337element-wise comparison of its two floating point vector operands. The output
4338elements have the same width as the input elements.</p>
4339<h5>Arguments:</h5>
4340<p>The '<tt>vfcmp</tt>' instruction takes three operands. The first operand is
4341the condition code indicating the kind of comparison to perform. It is not
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]4342a value, just a keyword. The possible condition code are:</p>
Nate Begemanac80ade2008-05-12 19:01:56 +0000[diff] [blame]4343<ol>
4344 <li><tt>false</tt>: no comparison, always returns false</li>
4345 <li><tt>oeq</tt>: ordered and equal</li>
4346 <li><tt>ogt</tt>: ordered and greater than </li>
4347 <li><tt>oge</tt>: ordered and greater than or equal</li>
4348 <li><tt>olt</tt>: ordered and less than </li>
4349 <li><tt>ole</tt>: ordered and less than or equal</li>
4350 <li><tt>one</tt>: ordered and not equal</li>
4351 <li><tt>ord</tt>: ordered (no nans)</li>
4352 <li><tt>ueq</tt>: unordered or equal</li>
4353 <li><tt>ugt</tt>: unordered or greater than </li>
4354 <li><tt>uge</tt>: unordered or greater than or equal</li>
4355 <li><tt>ult</tt>: unordered or less than </li>
4356 <li><tt>ule</tt>: unordered or less than or equal</li>
4357 <li><tt>une</tt>: unordered or not equal</li>
4358 <li><tt>uno</tt>: unordered (either nans)</li>
4359 <li><tt>true</tt>: no comparison, always returns true</li>
4360</ol>
4361<p>The remaining two arguments must be <a href="#t_vector">vector</a> of
4362<a href="#t_floating">floating point</a> typed. They must also be identical
4363types.</p>
4364<h5>Semantics:</h5>
Gabor Greiffb224a22008-08-07 21:46:00 +0000[diff] [blame]4365<p>The '<tt>vfcmp</tt>' instruction compares <tt>op1</tt> and <tt>op2</tt>
Nate Begemanac80ade2008-05-12 19:01:56 +0000[diff] [blame]4366according to the condition code given as <tt>cond</tt>. The comparison yields a
4367<a href="#t_vector">vector</a> of <a href="#t_integer">integer</a> result, with
4368an identical number of elements as the values being compared, and each element
4369having identical with to the width of the floating point elements. The most
4370significant bit in each element is 1 if the element-wise comparison evaluates to
4371true, and is 0 otherwise. All other bits of the result are undefined. The
4372condition codes are evaluated identically to the
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]4373<a href="#i_fcmp">'<tt>fcmp</tt>' instruction</a>.</p>
Nate Begemanac80ade2008-05-12 19:01:56 +0000[diff] [blame]4374
4375<h5>Example:</h5>
4376<pre>
Chris Lattner50ad45c2008-10-13 16:55:18 +0000[diff] [blame]4377 <i>; yields: result=&lt;2 x i32&gt; &lt; i32 0, i32 -1 &gt;</i>
4378 &lt;result&gt; = vfcmp oeq &lt;2 x float&gt; &lt; float 4, float 0 &gt;, &lt; float 5, float 0 &gt;
4379
4380 <i>; yields: result=&lt;2 x i64&gt; &lt; i64 -1, i64 0 &gt;</i>
4381 &lt;result&gt; = vfcmp ult &lt;2 x double&gt; &lt; double 1, double 2 &gt;, &lt; double 2, double 2&gt;
Nate Begemanac80ade2008-05-12 19:01:56 +0000[diff] [blame]4382</pre>
4383</div>
4384
4385<!-- _______________________________________________________________________ -->
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4386<div class="doc_subsubsection">
4387 <a name="i_phi">'<tt>phi</tt>' Instruction</a>
4388</div>
4389
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]4390<div class="doc_text">
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4391
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]4392<h5>Syntax:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4393
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]4394<pre> &lt;result&gt; = phi &lt;ty&gt; [ &lt;val0&gt;, &lt;label0&gt;], ...<br></pre>
4395<h5>Overview:</h5>
4396<p>The '<tt>phi</tt>' instruction is used to implement the &#966; node in
4397the SSA graph representing the function.</p>
4398<h5>Arguments:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4399
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4400<p>The type of the incoming values is specified with the first type
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]4401field. After this, the '<tt>phi</tt>' instruction takes a list of pairs
4402as arguments, with one pair for each predecessor basic block of the
4403current block. Only values of <a href="#t_firstclass">first class</a>
4404type may be used as the value arguments to the PHI node. Only labels
4405may be used as the label arguments.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4406
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]4407<p>There must be no non-phi instructions between the start of a basic
4408block and the PHI instructions: i.e. PHI instructions must be first in
4409a basic block.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4410
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]4411<h5>Semantics:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4412
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4413<p>At runtime, the '<tt>phi</tt>' instruction logically takes on the value
4414specified by the pair corresponding to the predecessor basic block that executed
4415just prior to the current block.</p>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4416
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]4417<h5>Example:</h5>
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4418<pre>
4419Loop: ; Infinite loop that counts from 0 on up...
4420 %indvar = phi i32 [ 0, %LoopHeader ], [ %nextindvar, %Loop ]
4421 %nextindvar = add i32 %indvar, 1
4422 br label %Loop
4423</pre>
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]4424</div>
4425
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]4426<!-- _______________________________________________________________________ -->
4427<div class="doc_subsubsection">
4428 <a name="i_select">'<tt>select</tt>' Instruction</a>
4429</div>
4430
4431<div class="doc_text">
4432
4433<h5>Syntax:</h5>
4434
4435<pre>
Dan Gohmanf72fb672008-09-09 01:02:47 +0000[diff] [blame]4436 &lt;result&gt; = select <i>selty</i> &lt;cond&gt;, &lt;ty&gt; &lt;val1&gt;, &lt;ty&gt; &lt;val2&gt; <i>; yields ty</i>
4437
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]4438 <i>selty</i> is either i1 or {&lt;N x i1&gt;}
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]4439</pre>
4440
4441<h5>Overview:</h5>
4442
4443<p>
4444The '<tt>select</tt>' instruction is used to choose one value based on a
4445condition, without branching.
4446</p>
4447
4448
4449<h5>Arguments:</h5>
4450
4451<p>
Dan Gohmanf72fb672008-09-09 01:02:47 +0000[diff] [blame]4452The '<tt>select</tt>' instruction requires an 'i1' value or
4453a vector of 'i1' values indicating the
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4454condition, and two values of the same <a href="#t_firstclass">first class</a>
Dan Gohmanf72fb672008-09-09 01:02:47 +0000[diff] [blame]4455type. If the val1/val2 are vectors and
4456the condition is a scalar, then entire vectors are selected, not
Chris Lattner5568e942008-05-20 20:48:21 +0000[diff] [blame]4457individual elements.
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]4458</p>
4459
4460<h5>Semantics:</h5>
4461
4462<p>
Dan Gohmanf72fb672008-09-09 01:02:47 +0000[diff] [blame]4463If the condition is an i1 and it evaluates to 1, the instruction returns the first
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]4464value argument; otherwise, it returns the second value argument.
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]4465</p>
Dan Gohmanf72fb672008-09-09 01:02:47 +0000[diff] [blame]4466<p>
4467If the condition is a vector of i1, then the value arguments must
4468be vectors of the same size, and the selection is done element
4469by element.
4470</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]4471
4472<h5>Example:</h5>
4473
4474<pre>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]4475 %X = select i1 true, i8 17, i8 42 <i>; yields i8:17</i>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]4476</pre>
Dan Gohmanc766f722009-01-22 01:39:38 +0000[diff] [blame]4477
4478<p>Note that the code generator does not yet support conditions
4479 with vector type.</p>
4480
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]4481</div>
4482
Robert Bocchino05ccd702006-01-15 20:48:27 +0000[diff] [blame]4483
4484<!-- _______________________________________________________________________ -->
4485<div class="doc_subsubsection">
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]4486 <a name="i_call">'<tt>call</tt>' Instruction</a>
4487</div>
4488
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4489<div class="doc_text">
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]4490
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4491<h5>Syntax:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]4492<pre>
Devang Patel307e8ab2008-10-07 17:48:33 +0000[diff] [blame]4493 &lt;result&gt; = [tail] call [<a href="#callingconv">cconv</a>] [<a href="#paramattrs">ret attrs</a>] &lt;ty&gt; [&lt;fnty&gt;*] &lt;fnptrval&gt;(&lt;function args&gt;) [<a href="#fnattrs">fn attrs</a>]
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]4494</pre>
4495
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4496<h5>Overview:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]4497
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4498<p>The '<tt>call</tt>' instruction represents a simple function call.</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]4499
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4500<h5>Arguments:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]4501
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4502<p>This instruction requires several arguments:</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]4503
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]4504<ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4505 <li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]4506 <p>The optional "tail" marker indicates whether the callee function accesses
4507 any allocas or varargs in the caller. If the "tail" marker is present, the
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]4508 function call is eligible for tail call optimization. Note that calls may
4509 be marked "tail" even if they do not occur before a <a
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]4510 href="#i_ret"><tt>ret</tt></a> instruction.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4511 </li>
4512 <li>
Duncan Sands8036ca42007-03-30 12:22:09 +0000[diff] [blame]4513 <p>The optional "cconv" marker indicates which <a href="#callingconv">calling
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]4514 convention</a> the call should use. If none is specified, the call defaults
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]4515 to using C calling conventions.</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]4516 </li>
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]4517
4518 <li>
4519 <p>The optional <a href="#paramattrs">Parameter Attributes</a> list for
4520 return values. Only '<tt>zeroext</tt>', '<tt>signext</tt>',
4521 and '<tt>inreg</tt>' attributes are valid here.</p>
4522 </li>
4523
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]4524 <li>
Nick Lewyckydb7e3c92007-09-08 13:57:50 +0000[diff] [blame]4525 <p>'<tt>ty</tt>': the type of the call instruction itself which is also
4526 the type of the return value. Functions that return no value are marked
4527 <tt><a href="#t_void">void</a></tt>.</p>
4528 </li>
4529 <li>
4530 <p>'<tt>fnty</tt>': shall be the signature of the pointer to function
4531 value being invoked. The argument types must match the types implied by
4532 this signature. This type can be omitted if the function is not varargs
4533 and if the function type does not return a pointer to a function.</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]4534 </li>
4535 <li>
4536 <p>'<tt>fnptrval</tt>': An LLVM value containing a pointer to a function to
4537 be invoked. In most cases, this is a direct function invocation, but
4538 indirect <tt>call</tt>s are just as possible, calling an arbitrary pointer
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]4539 to function value.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4540 </li>
4541 <li>
4542 <p>'<tt>function args</tt>': argument list whose types match the
Reid Spencera7e302a2005-05-01 22:22:57 +0000[diff] [blame]4543 function signature argument types. All arguments must be of
4544 <a href="#t_firstclass">first class</a> type. If the function signature
4545 indicates the function accepts a variable number of arguments, the extra
4546 arguments can be specified.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4547 </li>
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]4548 <li>
Devang Patel307e8ab2008-10-07 17:48:33 +0000[diff] [blame]4549 <p>The optional <a href="#fnattrs">function attributes</a> list. Only
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]4550 '<tt>noreturn</tt>', '<tt>nounwind</tt>', '<tt>readonly</tt>' and
4551 '<tt>readnone</tt>' attributes are valid here.</p>
4552 </li>
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]4553</ol>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]4554
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4555<h5>Semantics:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]4556
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4557<p>The '<tt>call</tt>' instruction is used to cause control flow to
4558transfer to a specified function, with its incoming arguments bound to
4559the specified values. Upon a '<tt><a href="#i_ret">ret</a></tt>'
4560instruction in the called function, control flow continues with the
4561instruction after the function call, and the return value of the
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]4562function is bound to the result argument.</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]4563
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4564<h5>Example:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]4565
4566<pre>
Nick Lewyckydb7e3c92007-09-08 13:57:50 +0000[diff] [blame]4567 %retval = call i32 @test(i32 %argc)
Chris Lattner772fccf2008-03-21 17:24:17 +0000[diff] [blame]4568 call i32 (i8 *, ...)* @printf(i8 * %msg, i32 12, i8 42) <i>; yields i32</i>
4569 %X = tail call i32 @foo() <i>; yields i32</i>
4570 %Y = tail call <a href="#callingconv">fastcc</a> i32 @foo() <i>; yields i32</i>
4571 call void %foo(i8 97 signext)
Devang Patelc3fc6df2008-03-10 20:49:15 +0000[diff] [blame]4572
4573 %struct.A = type { i32, i8 }
Devang Patelf642f472008-10-06 18:50:38 +0000[diff] [blame]4574 %r = call %struct.A @foo() <i>; yields { 32, i8 }</i>
Dan Gohmanb1e6b962008-10-04 19:00:07 +0000[diff] [blame]4575 %gr = extractvalue %struct.A %r, 0 <i>; yields i32</i>
4576 %gr1 = extractvalue %struct.A %r, 1 <i>; yields i8</i>
Chris Lattner85a350f2008-10-08 06:26:11 +0000[diff] [blame]4577 %Z = call void @foo() noreturn <i>; indicates that %foo never returns normally</i>
Matthijs Kooijmancb73d192008-10-07 10:03:45 +0000[diff] [blame]4578 %ZZ = call zeroext i32 @bar() <i>; Return value is %zero extended</i>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]4579</pre>
4580
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4581</div>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]4582
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]4583<!-- _______________________________________________________________________ -->
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]4584<div class="doc_subsubsection">
Chris Lattnerfb6977d2006-01-13 23:26:01 +0000[diff] [blame]4585 <a name="i_va_arg">'<tt>va_arg</tt>' Instruction</a>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]4586</div>
4587
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4588<div class="doc_text">
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]4589
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]4590<h5>Syntax:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]4591
4592<pre>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]4593 &lt;resultval&gt; = va_arg &lt;va_list*&gt; &lt;arglist&gt;, &lt;argty&gt;
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]4594</pre>
4595
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]4596<h5>Overview:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]4597
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]4598<p>The '<tt>va_arg</tt>' instruction is used to access arguments passed through
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]4599the "variable argument" area of a function call. It is used to implement the
4600<tt>va_arg</tt> macro in C.</p>
4601
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]4602<h5>Arguments:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]4603
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]4604<p>This instruction takes a <tt>va_list*</tt> value and the type of
4605the argument. It returns a value of the specified argument type and
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4606increments the <tt>va_list</tt> to point to the next argument. The
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]4607actual type of <tt>va_list</tt> is target specific.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]4608
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]4609<h5>Semantics:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]4610
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]4611<p>The '<tt>va_arg</tt>' instruction loads an argument of the specified
4612type from the specified <tt>va_list</tt> and causes the
4613<tt>va_list</tt> to point to the next argument. For more information,
4614see the variable argument handling <a href="#int_varargs">Intrinsic
4615Functions</a>.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]4616
4617<p>It is legal for this instruction to be called in a function which does not
4618take a variable number of arguments, for example, the <tt>vfprintf</tt>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4619function.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]4620
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]4621<p><tt>va_arg</tt> is an LLVM instruction instead of an <a
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]4622href="#intrinsics">intrinsic function</a> because it takes a type as an
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]4623argument.</p>
4624
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]4625<h5>Example:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]4626
4627<p>See the <a href="#int_varargs">variable argument processing</a> section.</p>
4628
Dan Gohmanf3e60bd2009-01-12 23:12:39 +0000[diff] [blame]4629<p>Note that the code generator does not yet fully support va_arg
4630 on many targets. Also, it does not currently support va_arg with
4631 aggregate types on any target.</p>
4632
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4633</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]4634
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]4635<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4636<div class="doc_section"> <a name="intrinsics">Intrinsic Functions</a> </div>
4637<!-- *********************************************************************** -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]4638
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4639<div class="doc_text">
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4640
4641<p>LLVM supports the notion of an "intrinsic function". These functions have
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]4642well known names and semantics and are required to follow certain restrictions.
4643Overall, these intrinsics represent an extension mechanism for the LLVM
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4644language that does not require changing all of the transformations in LLVM when
Gabor Greif04367bf2007-07-06 22:07:22 +0000[diff] [blame]4645adding to the language (or the bitcode reader/writer, the parser, etc...).</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4646
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]4647<p>Intrinsic function names must all start with an "<tt>llvm.</tt>" prefix. This
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4648prefix is reserved in LLVM for intrinsic names; thus, function names may not
4649begin with this prefix. Intrinsic functions must always be external functions:
4650you cannot define the body of intrinsic functions. Intrinsic functions may
4651only be used in call or invoke instructions: it is illegal to take the address
4652of an intrinsic function. Additionally, because intrinsic functions are part
4653of the LLVM language, it is required if any are added that they be documented
4654here.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4655
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]4656<p>Some intrinsic functions can be overloaded, i.e., the intrinsic represents
4657a family of functions that perform the same operation but on different data
4658types. Because LLVM can represent over 8 million different integer types,
4659overloading is used commonly to allow an intrinsic function to operate on any
4660integer type. One or more of the argument types or the result type can be
4661overloaded to accept any integer type. Argument types may also be defined as
4662exactly matching a previous argument's type or the result type. This allows an
4663intrinsic function which accepts multiple arguments, but needs all of them to
4664be of the same type, to only be overloaded with respect to a single argument or
4665the result.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4666
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]4667<p>Overloaded intrinsics will have the names of its overloaded argument types
4668encoded into its function name, each preceded by a period. Only those types
4669which are overloaded result in a name suffix. Arguments whose type is matched
4670against another type do not. For example, the <tt>llvm.ctpop</tt> function can
4671take an integer of any width and returns an integer of exactly the same integer
4672width. This leads to a family of functions such as
4673<tt>i8 @llvm.ctpop.i8(i8 %val)</tt> and <tt>i29 @llvm.ctpop.i29(i29 %val)</tt>.
4674Only one type, the return type, is overloaded, and only one type suffix is
4675required. Because the argument's type is matched against the return type, it
4676does not require its own name suffix.</p>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]4677
4678<p>To learn how to add an intrinsic function, please see the
4679<a href="ExtendingLLVM.html">Extending LLVM Guide</a>.
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4680</p>
4681
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4682</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]4683
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]4684<!-- ======================================================================= -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]4685<div class="doc_subsection">
4686 <a name="int_varargs">Variable Argument Handling Intrinsics</a>
4687</div>
4688
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4689<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4690
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4691<p>Variable argument support is defined in LLVM with the <a
Chris Lattnerfb6977d2006-01-13 23:26:01 +0000[diff] [blame]4692 href="#i_va_arg"><tt>va_arg</tt></a> instruction and these three
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4693intrinsic functions. These functions are related to the similarly
4694named macros defined in the <tt>&lt;stdarg.h&gt;</tt> header file.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4695
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4696<p>All of these functions operate on arguments that use a
4697target-specific value type "<tt>va_list</tt>". The LLVM assembly
4698language reference manual does not define what this type is, so all
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4699transformations should be prepared to handle these functions regardless of
4700the type used.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4701
Chris Lattner374ab302006-05-15 17:26:46 +0000[diff] [blame]4702<p>This example shows how the <a href="#i_va_arg"><tt>va_arg</tt></a>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4703instruction and the variable argument handling intrinsic functions are
4704used.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4705
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]4706<div class="doc_code">
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4707<pre>
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]4708define i32 @test(i32 %X, ...) {
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4709 ; Initialize variable argument processing
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4710 %ap = alloca i8*
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]4711 %ap2 = bitcast i8** %ap to i8*
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]4712 call void @llvm.va_start(i8* %ap2)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4713
4714 ; Read a single integer argument
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4715 %tmp = va_arg i8** %ap, i32
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4716
4717 ; Demonstrate usage of llvm.va_copy and llvm.va_end
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4718 %aq = alloca i8*
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]4719 %aq2 = bitcast i8** %aq to i8*
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4720 call void @llvm.va_copy(i8* %aq2, i8* %ap2)
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]4721 call void @llvm.va_end(i8* %aq2)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4722
4723 ; Stop processing of arguments.
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]4724 call void @llvm.va_end(i8* %ap2)
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4725 ret i32 %tmp
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4726}
Anton Korobeynikov5d522f32007-03-21 23:58:04 +0000[diff] [blame]4727
4728declare void @llvm.va_start(i8*)
4729declare void @llvm.va_copy(i8*, i8*)
4730declare void @llvm.va_end(i8*)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4731</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4732</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]4733
Bill Wendling2f7a8b02007-05-29 09:04:49 +0000[diff] [blame]4734</div>
4735
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]4736<!-- _______________________________________________________________________ -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]4737<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4738 <a name="int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]4739</div>
4740
4741
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4742<div class="doc_text">
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]4743<h5>Syntax:</h5>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]4744<pre> declare void %llvm.va_start(i8* &lt;arglist&gt;)<br></pre>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]4745<h5>Overview:</h5>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]4746<p>The '<tt>llvm.va_start</tt>' intrinsic initializes
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]4747<tt>*&lt;arglist&gt;</tt> for subsequent use by <tt><a
4748href="#i_va_arg">va_arg</a></tt>.</p>
4749
4750<h5>Arguments:</h5>
4751
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]4752<p>The argument is a pointer to a <tt>va_list</tt> element to initialize.</p>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]4753
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]4754<h5>Semantics:</h5>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]4755
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]4756<p>The '<tt>llvm.va_start</tt>' intrinsic works just like the <tt>va_start</tt>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]4757macro available in C. In a target-dependent way, it initializes the
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4758<tt>va_list</tt> element to which the argument points, so that the next call to
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]4759<tt>va_arg</tt> will produce the first variable argument passed to the function.
4760Unlike the C <tt>va_start</tt> macro, this intrinsic does not need to know the
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4761last argument of the function as the compiler can figure that out.</p>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]4762
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4763</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]4764
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]4765<!-- _______________________________________________________________________ -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]4766<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4767 <a name="int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]4768</div>
4769
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4770<div class="doc_text">
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]4771<h5>Syntax:</h5>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4772<pre> declare void @llvm.va_end(i8* &lt;arglist&gt;)<br></pre>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]4773<h5>Overview:</h5>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]4774
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4775<p>The '<tt>llvm.va_end</tt>' intrinsic destroys <tt>*&lt;arglist&gt;</tt>,
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4776which has been initialized previously with <tt><a href="#int_va_start">llvm.va_start</a></tt>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]4777or <tt><a href="#i_va_copy">llvm.va_copy</a></tt>.</p>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]4778
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]4779<h5>Arguments:</h5>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]4780
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4781<p>The argument is a pointer to a <tt>va_list</tt> to destroy.</p>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]4782
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]4783<h5>Semantics:</h5>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]4784
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4785<p>The '<tt>llvm.va_end</tt>' intrinsic works just like the <tt>va_end</tt>
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4786macro available in C. In a target-dependent way, it destroys the
4787<tt>va_list</tt> element to which the argument points. Calls to <a
4788href="#int_va_start"><tt>llvm.va_start</tt></a> and <a href="#int_va_copy">
4789<tt>llvm.va_copy</tt></a> must be matched exactly with calls to
4790<tt>llvm.va_end</tt>.</p>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]4791
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4792</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]4793
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]4794<!-- _______________________________________________________________________ -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]4795<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4796 <a name="int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]4797</div>
4798
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4799<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4800
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]4801<h5>Syntax:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4802
4803<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4804 declare void @llvm.va_copy(i8* &lt;destarglist&gt;, i8* &lt;srcarglist&gt;)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4805</pre>
4806
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]4807<h5>Overview:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4808
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4809<p>The '<tt>llvm.va_copy</tt>' intrinsic copies the current argument position
4810from the source argument list to the destination argument list.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4811
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]4812<h5>Arguments:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4813
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]4814<p>The first argument is a pointer to a <tt>va_list</tt> element to initialize.
Andrew Lenharthd0a4c622005-06-22 20:38:11 +0000[diff] [blame]4815The second argument is a pointer to a <tt>va_list</tt> element to copy from.</p>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]4816
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4817
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]4818<h5>Semantics:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4819
Jeff Cohenb627eab2007-04-29 01:07:00 +0000[diff] [blame]4820<p>The '<tt>llvm.va_copy</tt>' intrinsic works just like the <tt>va_copy</tt>
4821macro available in C. In a target-dependent way, it copies the source
4822<tt>va_list</tt> element into the destination <tt>va_list</tt> element. This
4823intrinsic is necessary because the <tt><a href="#int_va_start">
4824llvm.va_start</a></tt> intrinsic may be arbitrarily complex and require, for
4825example, memory allocation.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4826
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4827</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]4828
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4829<!-- ======================================================================= -->
4830<div class="doc_subsection">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4831 <a name="int_gc">Accurate Garbage Collection Intrinsics</a>
4832</div>
4833
4834<div class="doc_text">
4835
4836<p>
4837LLVM support for <a href="GarbageCollection.html">Accurate Garbage
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]4838Collection</a> (GC) requires the implementation and generation of these
4839intrinsics.
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4840These intrinsics allow identification of <a href="#int_gcroot">GC roots on the
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4841stack</a>, as well as garbage collector implementations that require <a
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4842href="#int_gcread">read</a> and <a href="#int_gcwrite">write</a> barriers.
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4843Front-ends for type-safe garbage collected languages should generate these
4844intrinsics to make use of the LLVM garbage collectors. For more details, see <a
4845href="GarbageCollection.html">Accurate Garbage Collection with LLVM</a>.
4846</p>
Christopher Lamb303dae92007-12-17 01:00:21 +0000[diff] [blame]4847
4848<p>The garbage collection intrinsics only operate on objects in the generic
4849 address space (address space zero).</p>
4850
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4851</div>
4852
4853<!-- _______________________________________________________________________ -->
4854<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4855 <a name="int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4856</div>
4857
4858<div class="doc_text">
4859
4860<h5>Syntax:</h5>
4861
4862<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]4863 declare void @llvm.gcroot(i8** %ptrloc, i8* %metadata)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4864</pre>
4865
4866<h5>Overview:</h5>
4867
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]4868<p>The '<tt>llvm.gcroot</tt>' intrinsic declares the existence of a GC root to
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4869the code generator, and allows some metadata to be associated with it.</p>
4870
4871<h5>Arguments:</h5>
4872
4873<p>The first argument specifies the address of a stack object that contains the
4874root pointer. The second pointer (which must be either a constant or a global
4875value address) contains the meta-data to be associated with the root.</p>
4876
4877<h5>Semantics:</h5>
4878
Chris Lattner05d67092008-04-24 05:59:56 +0000[diff] [blame]4879<p>At runtime, a call to this intrinsic stores a null pointer into the "ptrloc"
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4880location. At compile-time, the code generator generates information to allow
Gordon Henriksene1433f22007-12-25 02:31:26 +0000[diff] [blame]4881the runtime to find the pointer at GC safe points. The '<tt>llvm.gcroot</tt>'
4882intrinsic may only be used in a function which <a href="#gc">specifies a GC
4883algorithm</a>.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4884
4885</div>
4886
4887
4888<!-- _______________________________________________________________________ -->
4889<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4890 <a name="int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4891</div>
4892
4893<div class="doc_text">
4894
4895<h5>Syntax:</h5>
4896
4897<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]4898 declare i8* @llvm.gcread(i8* %ObjPtr, i8** %Ptr)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4899</pre>
4900
4901<h5>Overview:</h5>
4902
4903<p>The '<tt>llvm.gcread</tt>' intrinsic identifies reads of references from heap
4904locations, allowing garbage collector implementations that require read
4905barriers.</p>
4906
4907<h5>Arguments:</h5>
4908
Chris Lattner80626e92006-03-14 20:02:51 +0000[diff] [blame]4909<p>The second argument is the address to read from, which should be an address
4910allocated from the garbage collector. The first object is a pointer to the
4911start of the referenced object, if needed by the language runtime (otherwise
4912null).</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4913
4914<h5>Semantics:</h5>
4915
4916<p>The '<tt>llvm.gcread</tt>' intrinsic has the same semantics as a load
4917instruction, but may be replaced with substantially more complex code by the
Gordon Henriksene1433f22007-12-25 02:31:26 +0000[diff] [blame]4918garbage collector runtime, as needed. The '<tt>llvm.gcread</tt>' intrinsic
4919may only be used in a function which <a href="#gc">specifies a GC
4920algorithm</a>.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4921
4922</div>
4923
4924
4925<!-- _______________________________________________________________________ -->
4926<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4927 <a name="int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4928</div>
4929
4930<div class="doc_text">
4931
4932<h5>Syntax:</h5>
4933
4934<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]4935 declare void @llvm.gcwrite(i8* %P1, i8* %Obj, i8** %P2)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4936</pre>
4937
4938<h5>Overview:</h5>
4939
4940<p>The '<tt>llvm.gcwrite</tt>' intrinsic identifies writes of references to heap
4941locations, allowing garbage collector implementations that require write
4942barriers (such as generational or reference counting collectors).</p>
4943
4944<h5>Arguments:</h5>
4945
Chris Lattner80626e92006-03-14 20:02:51 +0000[diff] [blame]4946<p>The first argument is the reference to store, the second is the start of the
4947object to store it to, and the third is the address of the field of Obj to
4948store to. If the runtime does not require a pointer to the object, Obj may be
4949null.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4950
4951<h5>Semantics:</h5>
4952
4953<p>The '<tt>llvm.gcwrite</tt>' intrinsic has the same semantics as a store
4954instruction, but may be replaced with substantially more complex code by the
Gordon Henriksene1433f22007-12-25 02:31:26 +0000[diff] [blame]4955garbage collector runtime, as needed. The '<tt>llvm.gcwrite</tt>' intrinsic
4956may only be used in a function which <a href="#gc">specifies a GC
4957algorithm</a>.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]4958
4959</div>
4960
4961
4962
4963<!-- ======================================================================= -->
4964<div class="doc_subsection">
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4965 <a name="int_codegen">Code Generator Intrinsics</a>
4966</div>
4967
4968<div class="doc_text">
4969<p>
4970These intrinsics are provided by LLVM to expose special features that may only
4971be implemented with code generator support.
4972</p>
4973
4974</div>
4975
4976<!-- _______________________________________________________________________ -->
4977<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]4978 <a name="int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4979</div>
4980
4981<div class="doc_text">
4982
4983<h5>Syntax:</h5>
4984<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]4985 declare i8 *@llvm.returnaddress(i32 &lt;level&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4986</pre>
4987
4988<h5>Overview:</h5>
4989
4990<p>
Chris Lattner32b5d712006-10-15 20:05:59 +0000[diff] [blame]4991The '<tt>llvm.returnaddress</tt>' intrinsic attempts to compute a
4992target-specific value indicating the return address of the current function
4993or one of its callers.
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4994</p>
4995
4996<h5>Arguments:</h5>
4997
4998<p>
4999The argument to this intrinsic indicates which function to return the address
5000for. Zero indicates the calling function, one indicates its caller, etc. The
5001argument is <b>required</b> to be a constant integer value.
5002</p>
5003
5004<h5>Semantics:</h5>
5005
5006<p>
5007The '<tt>llvm.returnaddress</tt>' intrinsic either returns a pointer indicating
5008the return address of the specified call frame, or zero if it cannot be
5009identified. The value returned by this intrinsic is likely to be incorrect or 0
5010for arguments other than zero, so it should only be used for debugging purposes.
5011</p>
5012
5013<p>
5014Note that calling this intrinsic does not prevent function inlining or other
Chris Lattnerb40bb382005-03-07 20:30:51 +0000[diff] [blame]5015aggressive transformations, so the value returned may not be that of the obvious
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5016source-language caller.
5017</p>
5018</div>
5019
5020
5021<!-- _______________________________________________________________________ -->
5022<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5023 <a name="int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5024</div>
5025
5026<div class="doc_text">
5027
5028<h5>Syntax:</h5>
5029<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]5030 declare i8 *@llvm.frameaddress(i32 &lt;level&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5031</pre>
5032
5033<h5>Overview:</h5>
5034
5035<p>
Chris Lattner32b5d712006-10-15 20:05:59 +0000[diff] [blame]5036The '<tt>llvm.frameaddress</tt>' intrinsic attempts to return the
5037target-specific frame pointer value for the specified stack frame.
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5038</p>
5039
5040<h5>Arguments:</h5>
5041
5042<p>
5043The argument to this intrinsic indicates which function to return the frame
5044pointer for. Zero indicates the calling function, one indicates its caller,
5045etc. The argument is <b>required</b> to be a constant integer value.
5046</p>
5047
5048<h5>Semantics:</h5>
5049
5050<p>
5051The '<tt>llvm.frameaddress</tt>' intrinsic either returns a pointer indicating
5052the frame address of the specified call frame, or zero if it cannot be
5053identified. The value returned by this intrinsic is likely to be incorrect or 0
5054for arguments other than zero, so it should only be used for debugging purposes.
5055</p>
5056
5057<p>
5058Note that calling this intrinsic does not prevent function inlining or other
Chris Lattnerb40bb382005-03-07 20:30:51 +0000[diff] [blame]5059aggressive transformations, so the value returned may not be that of the obvious
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5060source-language caller.
5061</p>
5062</div>
5063
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]5064<!-- _______________________________________________________________________ -->
5065<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5066 <a name="int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]5067</div>
5068
5069<div class="doc_text">
5070
5071<h5>Syntax:</h5>
5072<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]5073 declare i8 *@llvm.stacksave()
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]5074</pre>
5075
5076<h5>Overview:</h5>
5077
5078<p>
5079The '<tt>llvm.stacksave</tt>' intrinsic is used to remember the current state of
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5080the function stack, for use with <a href="#int_stackrestore">
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]5081<tt>llvm.stackrestore</tt></a>. This is useful for implementing language
5082features like scoped automatic variable sized arrays in C99.
5083</p>
5084
5085<h5>Semantics:</h5>
5086
5087<p>
5088This intrinsic returns a opaque pointer value that can be passed to <a
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5089href="#int_stackrestore"><tt>llvm.stackrestore</tt></a>. When an
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]5090<tt>llvm.stackrestore</tt> intrinsic is executed with a value saved from
5091<tt>llvm.stacksave</tt>, it effectively restores the state of the stack to the
5092state it was in when the <tt>llvm.stacksave</tt> intrinsic executed. In
5093practice, this pops any <a href="#i_alloca">alloca</a> blocks from the stack
5094that were allocated after the <tt>llvm.stacksave</tt> was executed.
5095</p>
5096
5097</div>
5098
5099<!-- _______________________________________________________________________ -->
5100<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5101 <a name="int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]5102</div>
5103
5104<div class="doc_text">
5105
5106<h5>Syntax:</h5>
5107<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]5108 declare void @llvm.stackrestore(i8 * %ptr)
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]5109</pre>
5110
5111<h5>Overview:</h5>
5112
5113<p>
5114The '<tt>llvm.stackrestore</tt>' intrinsic is used to restore the state of
5115the function stack to the state it was in when the corresponding <a
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5116href="#int_stacksave"><tt>llvm.stacksave</tt></a> intrinsic executed. This is
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]5117useful for implementing language features like scoped automatic variable sized
5118arrays in C99.
5119</p>
5120
5121<h5>Semantics:</h5>
5122
5123<p>
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5124See the description for <a href="#int_stacksave"><tt>llvm.stacksave</tt></a>.
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]5125</p>
5126
5127</div>
5128
5129
5130<!-- _______________________________________________________________________ -->
5131<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5132 <a name="int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a>
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]5133</div>
5134
5135<div class="doc_text">
5136
5137<h5>Syntax:</h5>
5138<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]5139 declare void @llvm.prefetch(i8* &lt;address&gt;, i32 &lt;rw&gt;, i32 &lt;locality&gt;)
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]5140</pre>
5141
5142<h5>Overview:</h5>
5143
5144
5145<p>
5146The '<tt>llvm.prefetch</tt>' intrinsic is a hint to the code generator to insert
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]5147a prefetch instruction if supported; otherwise, it is a noop. Prefetches have
5148no
5149effect on the behavior of the program but can change its performance
Chris Lattner2a615362005-02-28 19:47:14 +0000[diff] [blame]5150characteristics.
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]5151</p>
5152
5153<h5>Arguments:</h5>
5154
5155<p>
5156<tt>address</tt> is the address to be prefetched, <tt>rw</tt> is the specifier
5157determining if the fetch should be for a read (0) or write (1), and
5158<tt>locality</tt> is a temporal locality specifier ranging from (0) - no
Chris Lattneraeffb4a2005-03-07 20:31:38 +0000[diff] [blame]5159locality, to (3) - extremely local keep in cache. The <tt>rw</tt> and
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]5160<tt>locality</tt> arguments must be constant integers.
5161</p>
5162
5163<h5>Semantics:</h5>
5164
5165<p>
5166This intrinsic does not modify the behavior of the program. In particular,
5167prefetches cannot trap and do not produce a value. On targets that support this
5168intrinsic, the prefetch can provide hints to the processor cache for better
5169performance.
5170</p>
5171
5172</div>
5173
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]5174<!-- _______________________________________________________________________ -->
5175<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5176 <a name="int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a>
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]5177</div>
5178
5179<div class="doc_text">
5180
5181<h5>Syntax:</h5>
5182<pre>
Chris Lattner1df4f752007-09-21 17:30:40 +0000[diff] [blame]5183 declare void @llvm.pcmarker(i32 &lt;id&gt;)
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]5184</pre>
5185
5186<h5>Overview:</h5>
5187
5188
5189<p>
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]5190The '<tt>llvm.pcmarker</tt>' intrinsic is a method to export a Program Counter
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]5191(PC) in a region of
5192code to simulators and other tools. The method is target specific, but it is
5193expected that the marker will use exported symbols to transmit the PC of the
5194marker.
5195The marker makes no guarantees that it will remain with any specific instruction
5196after optimizations. It is possible that the presence of a marker will inhibit
Chris Lattnerb3e7afd2006-03-24 07:16:10 +0000[diff] [blame]5197optimizations. The intended use is to be inserted after optimizations to allow
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]5198correlations of simulation runs.
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]5199</p>
5200
5201<h5>Arguments:</h5>
5202
5203<p>
5204<tt>id</tt> is a numerical id identifying the marker.
5205</p>
5206
5207<h5>Semantics:</h5>
5208
5209<p>
5210This intrinsic does not modify the behavior of the program. Backends that do not
5211support this intrinisic may ignore it.
5212</p>
5213
5214</div>
5215
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]5216<!-- _______________________________________________________________________ -->
5217<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5218 <a name="int_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a>
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]5219</div>
5220
5221<div class="doc_text">
5222
5223<h5>Syntax:</h5>
5224<pre>
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]5225 declare i64 @llvm.readcyclecounter( )
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]5226</pre>
5227
5228<h5>Overview:</h5>
5229
5230
5231<p>
5232The '<tt>llvm.readcyclecounter</tt>' intrinsic provides access to the cycle
5233counter register (or similar low latency, high accuracy clocks) on those targets
5234that support it. On X86, it should map to RDTSC. On Alpha, it should map to RPCC.
5235As the backing counters overflow quickly (on the order of 9 seconds on alpha), this
5236should only be used for small timings.
5237</p>
5238
5239<h5>Semantics:</h5>
5240
5241<p>
5242When directly supported, reading the cycle counter should not modify any memory.
5243Implementations are allowed to either return a application specific value or a
5244system wide value. On backends without support, this is lowered to a constant 0.
5245</p>
5246
5247</div>
5248
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5249<!-- ======================================================================= -->
5250<div class="doc_subsection">
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5251 <a name="int_libc">Standard C Library Intrinsics</a>
5252</div>
5253
5254<div class="doc_text">
5255<p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5256LLVM provides intrinsics for a few important standard C library functions.
5257These intrinsics allow source-language front-ends to pass information about the
5258alignment of the pointer arguments to the code generator, providing opportunity
5259for more efficient code generation.
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5260</p>
5261
5262</div>
5263
5264<!-- _______________________________________________________________________ -->
5265<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5266 <a name="int_memcpy">'<tt>llvm.memcpy</tt>' Intrinsic</a>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5267</div>
5268
5269<div class="doc_text">
5270
5271<h5>Syntax:</h5>
Chris Lattner824b9582008-11-21 16:42:48 +0000[diff] [blame]5272<p>This is an overloaded intrinsic. You can use llvm.memcpy on any integer bit
5273width. Not all targets support all bit widths however.</p>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5274<pre>
Chris Lattner824b9582008-11-21 16:42:48 +0000[diff] [blame]5275 declare void @llvm.memcpy.i8(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
5276 i8 &lt;len&gt;, i32 &lt;align&gt;)
5277 declare void @llvm.memcpy.i16(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
5278 i16 &lt;len&gt;, i32 &lt;align&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]5279 declare void @llvm.memcpy.i32(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5280 i32 &lt;len&gt;, i32 &lt;align&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]5281 declare void @llvm.memcpy.i64(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5282 i64 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5283</pre>
5284
5285<h5>Overview:</h5>
5286
5287<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]5288The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5289location to the destination location.
5290</p>
5291
5292<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]5293Note that, unlike the standard libc function, the <tt>llvm.memcpy.*</tt>
5294intrinsics do not return a value, and takes an extra alignment argument.
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5295</p>
5296
5297<h5>Arguments:</h5>
5298
5299<p>
5300The first argument is a pointer to the destination, the second is a pointer to
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]5301the source. The third argument is an integer argument
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5302specifying the number of bytes to copy, and the fourth argument is the alignment
5303of the source and destination locations.
5304</p>
5305
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]5306<p>
5307If the call to this intrinisic has an alignment value that is not 0 or 1, then
Chris Lattnerf0afc2c2006-03-04 00:02:10 +0000[diff] [blame]5308the caller guarantees that both the source and destination pointers are aligned
5309to that boundary.
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]5310</p>
5311
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5312<h5>Semantics:</h5>
5313
5314<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]5315The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]5316location to the destination location, which are not allowed to overlap. It
5317copies "len" bytes of memory over. If the argument is known to be aligned to
5318some boundary, this can be specified as the fourth argument, otherwise it should
5319be set to 0 or 1.
5320</p>
5321</div>
5322
5323
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]5324<!-- _______________________________________________________________________ -->
5325<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5326 <a name="int_memmove">'<tt>llvm.memmove</tt>' Intrinsic</a>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]5327</div>
5328
5329<div class="doc_text">
5330
5331<h5>Syntax:</h5>
Chris Lattner824b9582008-11-21 16:42:48 +0000[diff] [blame]5332<p>This is an overloaded intrinsic. You can use llvm.memmove on any integer bit
5333width. Not all targets support all bit widths however.</p>
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]5334<pre>
Chris Lattner824b9582008-11-21 16:42:48 +0000[diff] [blame]5335 declare void @llvm.memmove.i8(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
5336 i8 &lt;len&gt;, i32 &lt;align&gt;)
5337 declare void @llvm.memmove.i16(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
5338 i16 &lt;len&gt;, i32 &lt;align&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]5339 declare void @llvm.memmove.i32(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5340 i32 &lt;len&gt;, i32 &lt;align&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]5341 declare void @llvm.memmove.i64(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5342 i64 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]5343</pre>
5344
5345<h5>Overview:</h5>
5346
5347<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]5348The '<tt>llvm.memmove.*</tt>' intrinsics move a block of memory from the source
5349location to the destination location. It is similar to the
Chris Lattner4b2cbcf2008-01-06 19:51:52 +0000[diff] [blame]5350'<tt>llvm.memcpy</tt>' intrinsic but allows the two memory locations to overlap.
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]5351</p>
5352
5353<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]5354Note that, unlike the standard libc function, the <tt>llvm.memmove.*</tt>
5355intrinsics do not return a value, and takes an extra alignment argument.
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]5356</p>
5357
5358<h5>Arguments:</h5>
5359
5360<p>
5361The first argument is a pointer to the destination, the second is a pointer to
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]5362the source. The third argument is an integer argument
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]5363specifying the number of bytes to copy, and the fourth argument is the alignment
5364of the source and destination locations.
5365</p>
5366
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]5367<p>
5368If the call to this intrinisic has an alignment value that is not 0 or 1, then
Chris Lattnerf0afc2c2006-03-04 00:02:10 +0000[diff] [blame]5369the caller guarantees that the source and destination pointers are aligned to
5370that boundary.
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]5371</p>
5372
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]5373<h5>Semantics:</h5>
5374
5375<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]5376The '<tt>llvm.memmove.*</tt>' intrinsics copy a block of memory from the source
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]5377location to the destination location, which may overlap. It
5378copies "len" bytes of memory over. If the argument is known to be aligned to
5379some boundary, this can be specified as the fourth argument, otherwise it should
5380be set to 0 or 1.
5381</p>
5382</div>
5383
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]5384
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5385<!-- _______________________________________________________________________ -->
5386<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5387 <a name="int_memset">'<tt>llvm.memset.*</tt>' Intrinsics</a>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5388</div>
5389
5390<div class="doc_text">
5391
5392<h5>Syntax:</h5>
Chris Lattner824b9582008-11-21 16:42:48 +0000[diff] [blame]5393<p>This is an overloaded intrinsic. You can use llvm.memset on any integer bit
5394width. Not all targets support all bit widths however.</p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5395<pre>
Chris Lattner824b9582008-11-21 16:42:48 +0000[diff] [blame]5396 declare void @llvm.memset.i8(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
5397 i8 &lt;len&gt;, i32 &lt;align&gt;)
5398 declare void @llvm.memset.i16(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
5399 i16 &lt;len&gt;, i32 &lt;align&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]5400 declare void @llvm.memset.i32(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5401 i32 &lt;len&gt;, i32 &lt;align&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]5402 declare void @llvm.memset.i64(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5403 i64 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5404</pre>
5405
5406<h5>Overview:</h5>
5407
5408<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]5409The '<tt>llvm.memset.*</tt>' intrinsics fill a block of memory with a particular
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5410byte value.
5411</p>
5412
5413<p>
5414Note that, unlike the standard libc function, the <tt>llvm.memset</tt> intrinsic
5415does not return a value, and takes an extra alignment argument.
5416</p>
5417
5418<h5>Arguments:</h5>
5419
5420<p>
5421The first argument is a pointer to the destination to fill, the second is the
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]5422byte value to fill it with, the third argument is an integer
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5423argument specifying the number of bytes to fill, and the fourth argument is the
5424known alignment of destination location.
5425</p>
5426
5427<p>
5428If the call to this intrinisic has an alignment value that is not 0 or 1, then
Chris Lattnerf0afc2c2006-03-04 00:02:10 +0000[diff] [blame]5429the caller guarantees that the destination pointer is aligned to that boundary.
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5430</p>
5431
5432<h5>Semantics:</h5>
5433
5434<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]5435The '<tt>llvm.memset.*</tt>' intrinsics fill "len" bytes of memory starting at
5436the
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]5437destination location. If the argument is known to be aligned to some boundary,
5438this can be specified as the fourth argument, otherwise it should be set to 0 or
54391.
5440</p>
5441</div>
5442
5443
Chris Lattner32006282004-06-11 02:28:03 +0000[diff] [blame]5444<!-- _______________________________________________________________________ -->
5445<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5446 <a name="int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a>
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]5447</div>
5448
5449<div class="doc_text">
5450
5451<h5>Syntax:</h5>
Dale Johannesen408f9c12007-10-02 17:47:38 +0000[diff] [blame]5452<p>This is an overloaded intrinsic. You can use <tt>llvm.sqrt</tt> on any
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]5453floating point or vector of floating point type. Not all targets support all
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5454types however.</p>
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]5455<pre>
Dale Johannesen408f9c12007-10-02 17:47:38 +0000[diff] [blame]5456 declare float @llvm.sqrt.f32(float %Val)
5457 declare double @llvm.sqrt.f64(double %Val)
5458 declare x86_fp80 @llvm.sqrt.f80(x86_fp80 %Val)
5459 declare fp128 @llvm.sqrt.f128(fp128 %Val)
5460 declare ppc_fp128 @llvm.sqrt.ppcf128(ppc_fp128 %Val)
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]5461</pre>
5462
5463<h5>Overview:</h5>
5464
5465<p>
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]5466The '<tt>llvm.sqrt</tt>' intrinsics return the sqrt of the specified operand,
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]5467returning the same value as the libm '<tt>sqrt</tt>' functions would. Unlike
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]5468<tt>sqrt</tt> in libm, however, <tt>llvm.sqrt</tt> has undefined behavior for
Chris Lattner103e2d72008-01-29 07:00:44 +0000[diff] [blame]5469negative numbers other than -0.0 (which allows for better optimization, because
5470there is no need to worry about errno being set). <tt>llvm.sqrt(-0.0)</tt> is
5471defined to return -0.0 like IEEE sqrt.
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]5472</p>
5473
5474<h5>Arguments:</h5>
5475
5476<p>
5477The argument and return value are floating point numbers of the same type.
5478</p>
5479
5480<h5>Semantics:</h5>
5481
5482<p>
Dan Gohmand6257fe2007-07-16 14:37:41 +0000[diff] [blame]5483This function returns the sqrt of the specified operand if it is a nonnegative
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]5484floating point number.
5485</p>
5486</div>
5487
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]5488<!-- _______________________________________________________________________ -->
5489<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5490 <a name="int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]5491</div>
5492
5493<div class="doc_text">
5494
5495<h5>Syntax:</h5>
Dale Johannesen408f9c12007-10-02 17:47:38 +0000[diff] [blame]5496<p>This is an overloaded intrinsic. You can use <tt>llvm.powi</tt> on any
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]5497floating point or vector of floating point type. Not all targets support all
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5498types however.</p>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]5499<pre>
Dale Johannesen408f9c12007-10-02 17:47:38 +0000[diff] [blame]5500 declare float @llvm.powi.f32(float %Val, i32 %power)
5501 declare double @llvm.powi.f64(double %Val, i32 %power)
5502 declare x86_fp80 @llvm.powi.f80(x86_fp80 %Val, i32 %power)
5503 declare fp128 @llvm.powi.f128(fp128 %Val, i32 %power)
5504 declare ppc_fp128 @llvm.powi.ppcf128(ppc_fp128 %Val, i32 %power)
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]5505</pre>
5506
5507<h5>Overview:</h5>
5508
5509<p>
5510The '<tt>llvm.powi.*</tt>' intrinsics return the first operand raised to the
5511specified (positive or negative) power. The order of evaluation of
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]5512multiplications is not defined. When a vector of floating point type is
5513used, the second argument remains a scalar integer value.
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]5514</p>
5515
5516<h5>Arguments:</h5>
5517
5518<p>
5519The second argument is an integer power, and the first is a value to raise to
5520that power.
5521</p>
5522
5523<h5>Semantics:</h5>
5524
5525<p>
5526This function returns the first value raised to the second power with an
5527unspecified sequence of rounding operations.</p>
5528</div>
5529
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]5530<!-- _______________________________________________________________________ -->
5531<div class="doc_subsubsection">
5532 <a name="int_sin">'<tt>llvm.sin.*</tt>' Intrinsic</a>
5533</div>
5534
5535<div class="doc_text">
5536
5537<h5>Syntax:</h5>
5538<p>This is an overloaded intrinsic. You can use <tt>llvm.sin</tt> on any
5539floating point or vector of floating point type. Not all targets support all
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5540types however.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]5541<pre>
5542 declare float @llvm.sin.f32(float %Val)
5543 declare double @llvm.sin.f64(double %Val)
5544 declare x86_fp80 @llvm.sin.f80(x86_fp80 %Val)
5545 declare fp128 @llvm.sin.f128(fp128 %Val)
5546 declare ppc_fp128 @llvm.sin.ppcf128(ppc_fp128 %Val)
5547</pre>
5548
5549<h5>Overview:</h5>
5550
5551<p>
5552The '<tt>llvm.sin.*</tt>' intrinsics return the sine of the operand.
5553</p>
5554
5555<h5>Arguments:</h5>
5556
5557<p>
5558The argument and return value are floating point numbers of the same type.
5559</p>
5560
5561<h5>Semantics:</h5>
5562
5563<p>
5564This function returns the sine of the specified operand, returning the
5565same values as the libm <tt>sin</tt> functions would, and handles error
Dan Gohmanba83b7e2007-10-17 18:05:13 +0000[diff] [blame]5566conditions in the same way.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]5567</div>
5568
5569<!-- _______________________________________________________________________ -->
5570<div class="doc_subsubsection">
5571 <a name="int_cos">'<tt>llvm.cos.*</tt>' Intrinsic</a>
5572</div>
5573
5574<div class="doc_text">
5575
5576<h5>Syntax:</h5>
5577<p>This is an overloaded intrinsic. You can use <tt>llvm.cos</tt> on any
5578floating point or vector of floating point type. Not all targets support all
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5579types however.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]5580<pre>
5581 declare float @llvm.cos.f32(float %Val)
5582 declare double @llvm.cos.f64(double %Val)
5583 declare x86_fp80 @llvm.cos.f80(x86_fp80 %Val)
5584 declare fp128 @llvm.cos.f128(fp128 %Val)
5585 declare ppc_fp128 @llvm.cos.ppcf128(ppc_fp128 %Val)
5586</pre>
5587
5588<h5>Overview:</h5>
5589
5590<p>
5591The '<tt>llvm.cos.*</tt>' intrinsics return the cosine of the operand.
5592</p>
5593
5594<h5>Arguments:</h5>
5595
5596<p>
5597The argument and return value are floating point numbers of the same type.
5598</p>
5599
5600<h5>Semantics:</h5>
5601
5602<p>
5603This function returns the cosine of the specified operand, returning the
5604same values as the libm <tt>cos</tt> functions would, and handles error
Dan Gohmanba83b7e2007-10-17 18:05:13 +0000[diff] [blame]5605conditions in the same way.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]5606</div>
5607
5608<!-- _______________________________________________________________________ -->
5609<div class="doc_subsubsection">
5610 <a name="int_pow">'<tt>llvm.pow.*</tt>' Intrinsic</a>
5611</div>
5612
5613<div class="doc_text">
5614
5615<h5>Syntax:</h5>
5616<p>This is an overloaded intrinsic. You can use <tt>llvm.pow</tt> on any
5617floating point or vector of floating point type. Not all targets support all
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5618types however.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]5619<pre>
5620 declare float @llvm.pow.f32(float %Val, float %Power)
5621 declare double @llvm.pow.f64(double %Val, double %Power)
5622 declare x86_fp80 @llvm.pow.f80(x86_fp80 %Val, x86_fp80 %Power)
5623 declare fp128 @llvm.pow.f128(fp128 %Val, fp128 %Power)
5624 declare ppc_fp128 @llvm.pow.ppcf128(ppc_fp128 %Val, ppc_fp128 Power)
5625</pre>
5626
5627<h5>Overview:</h5>
5628
5629<p>
5630The '<tt>llvm.pow.*</tt>' intrinsics return the first operand raised to the
5631specified (positive or negative) power.
5632</p>
5633
5634<h5>Arguments:</h5>
5635
5636<p>
5637The second argument is a floating point power, and the first is a value to
5638raise to that power.
5639</p>
5640
5641<h5>Semantics:</h5>
5642
5643<p>
5644This function returns the first value raised to the second power,
5645returning the
5646same values as the libm <tt>pow</tt> functions would, and handles error
Dan Gohmanba83b7e2007-10-17 18:05:13 +0000[diff] [blame]5647conditions in the same way.</p>
Dan Gohman91c284c2007-10-15 20:30:11 +0000[diff] [blame]5648</div>
5649
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]5650
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]5651<!-- ======================================================================= -->
5652<div class="doc_subsection">
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]5653 <a name="int_manip">Bit Manipulation Intrinsics</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]5654</div>
5655
5656<div class="doc_text">
5657<p>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]5658LLVM provides intrinsics for a few important bit manipulation operations.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]5659These allow efficient code generation for some algorithms.
5660</p>
5661
5662</div>
5663
5664<!-- _______________________________________________________________________ -->
5665<div class="doc_subsubsection">
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5666 <a name="int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]5667</div>
5668
5669<div class="doc_text">
5670
5671<h5>Syntax:</h5>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]5672<p>This is an overloaded intrinsic function. You can use bswap on any integer
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5673type that is an even number of bytes (i.e. BitWidth % 16 == 0).</p>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]5674<pre>
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]5675 declare i16 @llvm.bswap.i16(i16 &lt;id&gt;)
5676 declare i32 @llvm.bswap.i32(i32 &lt;id&gt;)
5677 declare i64 @llvm.bswap.i64(i64 &lt;id&gt;)
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]5678</pre>
5679
5680<h5>Overview:</h5>
5681
5682<p>
Reid Spencer338ea092007-04-02 02:25:19 +0000[diff] [blame]5683The '<tt>llvm.bswap</tt>' family of intrinsics is used to byte swap integer
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]5684values with an even number of bytes (positive multiple of 16 bits). These are
5685useful for performing operations on data that is not in the target's native
5686byte order.
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]5687</p>
5688
5689<h5>Semantics:</h5>
5690
5691<p>
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]5692The <tt>llvm.bswap.i16</tt> intrinsic returns an i16 value that has the high
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5693and low byte of the input i16 swapped. Similarly, the <tt>llvm.bswap.i32</tt>
5694intrinsic returns an i32 value that has the four bytes of the input i32
5695swapped, so that if the input bytes are numbered 0, 1, 2, 3 then the returned
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]5696i32 will have its bytes in 3, 2, 1, 0 order. The <tt>llvm.bswap.i48</tt>,
5697<tt>llvm.bswap.i64</tt> and other intrinsics extend this concept to
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]5698additional even-byte lengths (6 bytes, 8 bytes and more, respectively).
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]5699</p>
5700
5701</div>
5702
5703<!-- _______________________________________________________________________ -->
5704<div class="doc_subsubsection">
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]5705 <a name="int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]5706</div>
5707
5708<div class="doc_text">
5709
5710<h5>Syntax:</h5>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]5711<p>This is an overloaded intrinsic. You can use llvm.ctpop on any integer bit
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5712width. Not all targets support all bit widths however.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]5713<pre>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]5714 declare i8 @llvm.ctpop.i8(i8 &lt;src&gt;)
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]5715 declare i16 @llvm.ctpop.i16(i16 &lt;src&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]5716 declare i32 @llvm.ctpop.i32(i32 &lt;src&gt;)
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]5717 declare i64 @llvm.ctpop.i64(i64 &lt;src&gt;)
5718 declare i256 @llvm.ctpop.i256(i256 &lt;src&gt;)
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]5719</pre>
5720
5721<h5>Overview:</h5>
5722
5723<p>
Chris Lattnerec6cb612006-01-16 22:38:59 +0000[diff] [blame]5724The '<tt>llvm.ctpop</tt>' family of intrinsics counts the number of bits set in a
5725value.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]5726</p>
5727
5728<h5>Arguments:</h5>
5729
5730<p>
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]5731The only argument is the value to be counted. The argument may be of any
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]5732integer type. The return type must match the argument type.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]5733</p>
5734
5735<h5>Semantics:</h5>
5736
5737<p>
5738The '<tt>llvm.ctpop</tt>' intrinsic counts the 1's in a variable.
5739</p>
5740</div>
5741
5742<!-- _______________________________________________________________________ -->
5743<div class="doc_subsubsection">
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]5744 <a name="int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]5745</div>
5746
5747<div class="doc_text">
5748
5749<h5>Syntax:</h5>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]5750<p>This is an overloaded intrinsic. You can use <tt>llvm.ctlz</tt> on any
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5751integer bit width. Not all targets support all bit widths however.</p>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]5752<pre>
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]5753 declare i8 @llvm.ctlz.i8 (i8 &lt;src&gt;)
5754 declare i16 @llvm.ctlz.i16(i16 &lt;src&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]5755 declare i32 @llvm.ctlz.i32(i32 &lt;src&gt;)
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]5756 declare i64 @llvm.ctlz.i64(i64 &lt;src&gt;)
5757 declare i256 @llvm.ctlz.i256(i256 &lt;src&gt;)
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]5758</pre>
5759
5760<h5>Overview:</h5>
5761
5762<p>
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]5763The '<tt>llvm.ctlz</tt>' family of intrinsic functions counts the number of
5764leading zeros in a variable.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]5765</p>
5766
5767<h5>Arguments:</h5>
5768
5769<p>
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]5770The only argument is the value to be counted. The argument may be of any
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]5771integer type. The return type must match the argument type.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]5772</p>
5773
5774<h5>Semantics:</h5>
5775
5776<p>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]5777The '<tt>llvm.ctlz</tt>' intrinsic counts the leading (most significant) zeros
5778in a variable. If the src == 0 then the result is the size in bits of the type
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]5779of src. For example, <tt>llvm.ctlz(i32 2) = 30</tt>.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]5780</p>
5781</div>
Chris Lattner32006282004-06-11 02:28:03 +0000[diff] [blame]5782
5783
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]5784
5785<!-- _______________________________________________________________________ -->
5786<div class="doc_subsubsection">
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]5787 <a name="int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic</a>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]5788</div>
5789
5790<div class="doc_text">
5791
5792<h5>Syntax:</h5>
Reid Spencer409e28f2007-04-01 08:04:23 +0000[diff] [blame]5793<p>This is an overloaded intrinsic. You can use <tt>llvm.cttz</tt> on any
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5794integer bit width. Not all targets support all bit widths however.</p>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]5795<pre>
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]5796 declare i8 @llvm.cttz.i8 (i8 &lt;src&gt;)
5797 declare i16 @llvm.cttz.i16(i16 &lt;src&gt;)
Anton Korobeynikovec43a062007-03-22 00:02:17 +0000[diff] [blame]5798 declare i32 @llvm.cttz.i32(i32 &lt;src&gt;)
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]5799 declare i64 @llvm.cttz.i64(i64 &lt;src&gt;)
5800 declare i256 @llvm.cttz.i256(i256 &lt;src&gt;)
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]5801</pre>
5802
5803<h5>Overview:</h5>
5804
5805<p>
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]5806The '<tt>llvm.cttz</tt>' family of intrinsic functions counts the number of
5807trailing zeros.
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]5808</p>
5809
5810<h5>Arguments:</h5>
5811
5812<p>
5813The only argument is the value to be counted. The argument may be of any
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]5814integer type. The return type must match the argument type.
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]5815</p>
5816
5817<h5>Semantics:</h5>
5818
5819<p>
5820The '<tt>llvm.cttz</tt>' intrinsic counts the trailing (least significant) zeros
5821in a variable. If the src == 0 then the result is the size in bits of the type
5822of src. For example, <tt>llvm.cttz(2) = 1</tt>.
5823</p>
5824</div>
5825
Reid Spencer497d93e2007-04-01 08:27:01 +0000[diff] [blame]5826<!-- _______________________________________________________________________ -->
5827<div class="doc_subsubsection">
Reid Spencerbeacf662007-04-10 02:51:31 +0000[diff] [blame]5828 <a name="int_part_select">'<tt>llvm.part.select.*</tt>' Intrinsic</a>
Reid Spencera13ba7d2007-04-01 19:00:37 +0000[diff] [blame]5829</div>
5830
5831<div class="doc_text">
5832
5833<h5>Syntax:</h5>
Reid Spencerbeacf662007-04-10 02:51:31 +0000[diff] [blame]5834<p>This is an overloaded intrinsic. You can use <tt>llvm.part.select</tt>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5835on any integer bit width.</p>
Reid Spencera13ba7d2007-04-01 19:00:37 +0000[diff] [blame]5836<pre>
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]5837 declare i17 @llvm.part.select.i17 (i17 %val, i32 %loBit, i32 %hiBit)
5838 declare i29 @llvm.part.select.i29 (i29 %val, i32 %loBit, i32 %hiBit)
Reid Spencera13ba7d2007-04-01 19:00:37 +0000[diff] [blame]5839</pre>
5840
5841<h5>Overview:</h5>
Reid Spencerbeacf662007-04-10 02:51:31 +0000[diff] [blame]5842<p>The '<tt>llvm.part.select</tt>' family of intrinsic functions selects a
Reid Spencera13ba7d2007-04-01 19:00:37 +0000[diff] [blame]5843range of bits from an integer value and returns them in the same bit width as
5844the original value.</p>
5845
5846<h5>Arguments:</h5>
5847<p>The first argument, <tt>%val</tt> and the result may be integer types of
5848any bit width but they must have the same bit width. The second and third
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5849arguments must be <tt>i32</tt> type since they specify only a bit index.</p>
Reid Spencera13ba7d2007-04-01 19:00:37 +0000[diff] [blame]5850
5851<h5>Semantics:</h5>
Reid Spencerbeacf662007-04-10 02:51:31 +0000[diff] [blame]5852<p>The operation of the '<tt>llvm.part.select</tt>' intrinsic has two modes
Reid Spencera3e435f2007-04-04 02:42:35 +0000[diff] [blame]5853of operation: forwards and reverse. If <tt>%loBit</tt> is greater than
5854<tt>%hiBits</tt> then the intrinsic operates in reverse mode. Otherwise it
5855operates in forward mode.</p>
5856<p>In forward mode, this intrinsic is the equivalent of shifting <tt>%val</tt>
5857right by <tt>%loBit</tt> bits and then ANDing it with a mask with
Reid Spencera13ba7d2007-04-01 19:00:37 +0000[diff] [blame]5858only the <tt>%hiBit - %loBit</tt> bits set, as follows:</p>
5859<ol>
5860 <li>The <tt>%val</tt> is shifted right (LSHR) by the number of bits specified
5861 by <tt>%loBits</tt>. This normalizes the value to the low order bits.</li>
5862 <li>The <tt>%loBits</tt> value is subtracted from the <tt>%hiBits</tt> value
5863 to determine the number of bits to retain.</li>
5864 <li>A mask of the retained bits is created by shifting a -1 value.</li>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5865 <li>The mask is ANDed with <tt>%val</tt> to produce the result.</li>
Reid Spencera13ba7d2007-04-01 19:00:37 +0000[diff] [blame]5866</ol>
Reid Spencerd6a85b52007-05-14 16:14:57 +0000[diff] [blame]5867<p>In reverse mode, a similar computation is made except that the bits are
5868returned in the reverse order. So, for example, if <tt>X</tt> has the value
5869<tt>i16 0x0ACF (101011001111)</tt> and we apply
5870<tt>part.select(i16 X, 8, 3)</tt> to it, we get back the value
5871<tt>i16 0x0026 (000000100110)</tt>.</p>
Reid Spencera13ba7d2007-04-01 19:00:37 +0000[diff] [blame]5872</div>
5873
Reid Spencerf86037f2007-04-11 23:23:49 +0000[diff] [blame]5874<div class="doc_subsubsection">
5875 <a name="int_part_set">'<tt>llvm.part.set.*</tt>' Intrinsic</a>
5876</div>
5877
5878<div class="doc_text">
5879
5880<h5>Syntax:</h5>
5881<p>This is an overloaded intrinsic. You can use <tt>llvm.part.set</tt>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5882on any integer bit width.</p>
Reid Spencerf86037f2007-04-11 23:23:49 +0000[diff] [blame]5883<pre>
Chandler Carruth69940402007-08-04 01:51:18 +0000[diff] [blame]5884 declare i17 @llvm.part.set.i17.i9 (i17 %val, i9 %repl, i32 %lo, i32 %hi)
5885 declare i29 @llvm.part.set.i29.i9 (i29 %val, i9 %repl, i32 %lo, i32 %hi)
Reid Spencerf86037f2007-04-11 23:23:49 +0000[diff] [blame]5886</pre>
5887
5888<h5>Overview:</h5>
5889<p>The '<tt>llvm.part.set</tt>' family of intrinsic functions replaces a range
5890of bits in an integer value with another integer value. It returns the integer
5891with the replaced bits.</p>
5892
5893<h5>Arguments:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]5894<p>The first argument, <tt>%val</tt>, and the result may be integer types of
5895any bit width, but they must have the same bit width. <tt>%val</tt> is the value
Reid Spencerf86037f2007-04-11 23:23:49 +0000[diff] [blame]5896whose bits will be replaced. The second argument, <tt>%repl</tt> may be an
5897integer of any bit width. The third and fourth arguments must be <tt>i32</tt>
5898type since they specify only a bit index.</p>
5899
5900<h5>Semantics:</h5>
5901<p>The operation of the '<tt>llvm.part.set</tt>' intrinsic has two modes
5902of operation: forwards and reverse. If <tt>%lo</tt> is greater than
5903<tt>%hi</tt> then the intrinsic operates in reverse mode. Otherwise it
5904operates in forward mode.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]5905
Reid Spencerf86037f2007-04-11 23:23:49 +0000[diff] [blame]5906<p>For both modes, the <tt>%repl</tt> value is prepared for use by either
5907truncating it down to the size of the replacement area or zero extending it
5908up to that size.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]5909
Reid Spencerf86037f2007-04-11 23:23:49 +0000[diff] [blame]5910<p>In forward mode, the bits between <tt>%lo</tt> and <tt>%hi</tt> (inclusive)
5911are replaced with corresponding bits from <tt>%repl</tt>. That is the 0th bit
5912in <tt>%repl</tt> replaces the <tt>%lo</tt>th bit in <tt>%val</tt> and etc. up
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5913to the <tt>%hi</tt>th bit.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]5914
Reid Spencerc6749c42007-05-14 16:50:20 +0000[diff] [blame]5915<p>In reverse mode, a similar computation is made except that the bits are
5916reversed. That is, the <tt>0</tt>th bit in <tt>%repl</tt> replaces the
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]5917<tt>%hi</tt> bit in <tt>%val</tt> and etc. down to the <tt>%lo</tt>th bit.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]5918
Reid Spencerf86037f2007-04-11 23:23:49 +0000[diff] [blame]5919<h5>Examples:</h5>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]5920
Reid Spencerf86037f2007-04-11 23:23:49 +0000[diff] [blame]5921<pre>
Reid Spencerf0dbf642007-04-12 01:03:03 +0000[diff] [blame]5922 llvm.part.set(0xFFFF, 0, 4, 7) -&gt; 0xFF0F
Reid Spencerc6749c42007-05-14 16:50:20 +0000[diff] [blame]5923 llvm.part.set(0xFFFF, 0, 7, 4) -&gt; 0xFF0F
5924 llvm.part.set(0xFFFF, 1, 7, 4) -&gt; 0xFF8F
5925 llvm.part.set(0xFFFF, F, 8, 3) -&gt; 0xFFE7
Reid Spencerf0dbf642007-04-12 01:03:03 +0000[diff] [blame]5926 llvm.part.set(0xFFFF, 0, 3, 8) -&gt; 0xFE07
Reid Spencerc8910842007-04-11 23:49:50 +0000[diff] [blame]5927</pre>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]5928
5929</div>
5930
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]5931<!-- ======================================================================= -->
5932<div class="doc_subsection">
5933 <a name="int_overflow">Arithmetic with Overflow Intrinsics</a>
5934</div>
5935
5936<div class="doc_text">
5937<p>
5938LLVM provides intrinsics for some arithmetic with overflow operations.
5939</p>
5940
5941</div>
5942
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]5943<!-- _______________________________________________________________________ -->
5944<div class="doc_subsubsection">
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]5945 <a name="int_sadd_overflow">'<tt>llvm.sadd.with.overflow.*</tt>' Intrinsics</a>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]5946</div>
5947
5948<div class="doc_text">
5949
5950<h5>Syntax:</h5>
5951
5952<p>This is an overloaded intrinsic. You can use <tt>llvm.sadd.with.overflow</tt>
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]5953on any integer bit width.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]5954
5955<pre>
5956 declare {i16, i1} @llvm.sadd.with.overflow.i16(i16 %a, i16 %b)
5957 declare {i32, i1} @llvm.sadd.with.overflow.i32(i32 %a, i32 %b)
5958 declare {i64, i1} @llvm.sadd.with.overflow.i64(i64 %a, i64 %b)
5959</pre>
5960
5961<h5>Overview:</h5>
5962
5963<p>The '<tt>llvm.sadd.with.overflow</tt>' family of intrinsic functions perform
5964a signed addition of the two arguments, and indicate whether an overflow
5965occurred during the signed summation.</p>
5966
5967<h5>Arguments:</h5>
5968
5969<p>The arguments (%a and %b) and the first element of the result structure may
5970be of integer types of any bit width, but they must have the same bit width. The
5971second element of the result structure must be of type <tt>i1</tt>. <tt>%a</tt>
5972and <tt>%b</tt> are the two values that will undergo signed addition.</p>
5973
5974<h5>Semantics:</h5>
5975
5976<p>The '<tt>llvm.sadd.with.overflow</tt>' family of intrinsic functions perform
5977a signed addition of the two variables. They return a structure &mdash; the
5978first element of which is the signed summation, and the second element of which
5979is a bit specifying if the signed summation resulted in an overflow.</p>
5980
5981<h5>Examples:</h5>
5982<pre>
5983 %res = call {i32, i1} @llvm.sadd.with.overflow.i32(i32 %a, i32 %b)
5984 %sum = extractvalue {i32, i1} %res, 0
5985 %obit = extractvalue {i32, i1} %res, 1
5986 br i1 %obit, label %overflow, label %normal
5987</pre>
5988
5989</div>
5990
5991<!-- _______________________________________________________________________ -->
5992<div class="doc_subsubsection">
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]5993 <a name="int_uadd_overflow">'<tt>llvm.uadd.with.overflow.*</tt>' Intrinsics</a>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]5994</div>
5995
5996<div class="doc_text">
5997
5998<h5>Syntax:</h5>
5999
6000<p>This is an overloaded intrinsic. You can use <tt>llvm.uadd.with.overflow</tt>
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]6001on any integer bit width.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6002
6003<pre>
6004 declare {i16, i1} @llvm.uadd.with.overflow.i16(i16 %a, i16 %b)
6005 declare {i32, i1} @llvm.uadd.with.overflow.i32(i32 %a, i32 %b)
6006 declare {i64, i1} @llvm.uadd.with.overflow.i64(i64 %a, i64 %b)
6007</pre>
6008
6009<h5>Overview:</h5>
6010
6011<p>The '<tt>llvm.uadd.with.overflow</tt>' family of intrinsic functions perform
6012an unsigned addition of the two arguments, and indicate whether a carry occurred
6013during the unsigned summation.</p>
6014
6015<h5>Arguments:</h5>
6016
6017<p>The arguments (%a and %b) and the first element of the result structure may
6018be of integer types of any bit width, but they must have the same bit width. The
6019second element of the result structure must be of type <tt>i1</tt>. <tt>%a</tt>
6020and <tt>%b</tt> are the two values that will undergo unsigned addition.</p>
6021
6022<h5>Semantics:</h5>
6023
6024<p>The '<tt>llvm.uadd.with.overflow</tt>' family of intrinsic functions perform
6025an unsigned addition of the two arguments. They return a structure &mdash; the
6026first element of which is the sum, and the second element of which is a bit
6027specifying if the unsigned summation resulted in a carry.</p>
6028
6029<h5>Examples:</h5>
6030<pre>
6031 %res = call {i32, i1} @llvm.uadd.with.overflow.i32(i32 %a, i32 %b)
6032 %sum = extractvalue {i32, i1} %res, 0
6033 %obit = extractvalue {i32, i1} %res, 1
6034 br i1 %obit, label %carry, label %normal
6035</pre>
6036
6037</div>
6038
6039<!-- _______________________________________________________________________ -->
6040<div class="doc_subsubsection">
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]6041 <a name="int_ssub_overflow">'<tt>llvm.ssub.with.overflow.*</tt>' Intrinsics</a>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6042</div>
6043
6044<div class="doc_text">
6045
6046<h5>Syntax:</h5>
6047
6048<p>This is an overloaded intrinsic. You can use <tt>llvm.ssub.with.overflow</tt>
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]6049on any integer bit width.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6050
6051<pre>
6052 declare {i16, i1} @llvm.ssub.with.overflow.i16(i16 %a, i16 %b)
6053 declare {i32, i1} @llvm.ssub.with.overflow.i32(i32 %a, i32 %b)
6054 declare {i64, i1} @llvm.ssub.with.overflow.i64(i64 %a, i64 %b)
6055</pre>
6056
6057<h5>Overview:</h5>
6058
6059<p>The '<tt>llvm.ssub.with.overflow</tt>' family of intrinsic functions perform
6060a signed subtraction of the two arguments, and indicate whether an overflow
6061occurred during the signed subtraction.</p>
6062
6063<h5>Arguments:</h5>
6064
6065<p>The arguments (%a and %b) and the first element of the result structure may
6066be of integer types of any bit width, but they must have the same bit width. The
6067second element of the result structure must be of type <tt>i1</tt>. <tt>%a</tt>
6068and <tt>%b</tt> are the two values that will undergo signed subtraction.</p>
6069
6070<h5>Semantics:</h5>
6071
6072<p>The '<tt>llvm.ssub.with.overflow</tt>' family of intrinsic functions perform
6073a signed subtraction of the two arguments. They return a structure &mdash; the
6074first element of which is the subtraction, and the second element of which is a bit
6075specifying if the signed subtraction resulted in an overflow.</p>
6076
6077<h5>Examples:</h5>
6078<pre>
6079 %res = call {i32, i1} @llvm.ssub.with.overflow.i32(i32 %a, i32 %b)
6080 %sum = extractvalue {i32, i1} %res, 0
6081 %obit = extractvalue {i32, i1} %res, 1
6082 br i1 %obit, label %overflow, label %normal
6083</pre>
6084
6085</div>
6086
6087<!-- _______________________________________________________________________ -->
6088<div class="doc_subsubsection">
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]6089 <a name="int_usub_overflow">'<tt>llvm.usub.with.overflow.*</tt>' Intrinsics</a>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6090</div>
6091
6092<div class="doc_text">
6093
6094<h5>Syntax:</h5>
6095
6096<p>This is an overloaded intrinsic. You can use <tt>llvm.usub.with.overflow</tt>
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]6097on any integer bit width.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6098
6099<pre>
6100 declare {i16, i1} @llvm.usub.with.overflow.i16(i16 %a, i16 %b)
6101 declare {i32, i1} @llvm.usub.with.overflow.i32(i32 %a, i32 %b)
6102 declare {i64, i1} @llvm.usub.with.overflow.i64(i64 %a, i64 %b)
6103</pre>
6104
6105<h5>Overview:</h5>
6106
6107<p>The '<tt>llvm.usub.with.overflow</tt>' family of intrinsic functions perform
6108an unsigned subtraction of the two arguments, and indicate whether an overflow
6109occurred during the unsigned subtraction.</p>
6110
6111<h5>Arguments:</h5>
6112
6113<p>The arguments (%a and %b) and the first element of the result structure may
6114be of integer types of any bit width, but they must have the same bit width. The
6115second element of the result structure must be of type <tt>i1</tt>. <tt>%a</tt>
6116and <tt>%b</tt> are the two values that will undergo unsigned subtraction.</p>
6117
6118<h5>Semantics:</h5>
6119
6120<p>The '<tt>llvm.usub.with.overflow</tt>' family of intrinsic functions perform
6121an unsigned subtraction of the two arguments. They return a structure &mdash; the
6122first element of which is the subtraction, and the second element of which is a bit
6123specifying if the unsigned subtraction resulted in an overflow.</p>
6124
6125<h5>Examples:</h5>
6126<pre>
6127 %res = call {i32, i1} @llvm.usub.with.overflow.i32(i32 %a, i32 %b)
6128 %sum = extractvalue {i32, i1} %res, 0
6129 %obit = extractvalue {i32, i1} %res, 1
6130 br i1 %obit, label %overflow, label %normal
6131</pre>
6132
6133</div>
6134
6135<!-- _______________________________________________________________________ -->
6136<div class="doc_subsubsection">
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]6137 <a name="int_smul_overflow">'<tt>llvm.smul.with.overflow.*</tt>' Intrinsics</a>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6138</div>
6139
6140<div class="doc_text">
6141
6142<h5>Syntax:</h5>
6143
6144<p>This is an overloaded intrinsic. You can use <tt>llvm.smul.with.overflow</tt>
Bill Wendlingda01af72009-02-08 04:04:40 +0000[diff] [blame]6145on any integer bit width.</p>
Bill Wendlingac1df8e2009-02-08 01:40:31 +0000[diff] [blame]6146
6147<pre>
6148 declare {i16, i1} @llvm.smul.with.overflow.i16(i16 %a, i16 %b)
6149 declare {i32, i1} @llvm.smul.with.overflow.i32(i32 %a, i32 %b)
6150 declare {i64, i1} @llvm.smul.with.overflow.i64(i64 %a, i64 %b)
6151</pre>
6152
6153<h5>Overview:</h5>
6154
6155<p>The '<tt>llvm.smul.with.overflow</tt>' family of intrinsic functions perform
6156a signed multiplication of the two arguments, and indicate whether an overflow
6157occurred during the signed multiplication.</p>
6158
6159<h5>Arguments:</h5>
6160
6161<p>The arguments (%a and %b) and the first element of the result structure may
6162be of integer types of any bit width, but they must have the same bit width. The
6163second element of the result structure must be of type <tt>i1</tt>. <tt>%a</tt>
6164and <tt>%b</tt> are the two values that will undergo signed multiplication.</p>
6165
6166<h5>Semantics:</h5>
6167
6168<p>The '<tt>llvm.smul.with.overflow</tt>' family of intrinsic functions perform
6169a signed multiplication of the two arguments. They return a structure &mdash;
6170the first element of which is the multiplication, and the second element of
6171which is a bit specifying if the signed multiplication resulted in an
6172overflow.</p>
6173
6174<h5>Examples:</h5>
6175<pre>
6176 %res = call {i32, i1} @llvm.smul.with.overflow.i32(i32 %a, i32 %b)
6177 %sum = extractvalue {i32, i1} %res, 0
6178 %obit = extractvalue {i32, i1} %res, 1
6179 br i1 %obit, label %overflow, label %normal
6180</pre>
6181
Reid Spencerf86037f2007-04-11 23:23:49 +0000[diff] [blame]6182</div>
6183
Bill Wendling41b485c2009-02-08 23:00:09 +0000[diff] [blame]6184<!-- _______________________________________________________________________ -->
6185<div class="doc_subsubsection">
6186 <a name="int_umul_overflow">'<tt>llvm.umul.with.overflow.*</tt>' Intrinsics</a>
6187</div>
6188
6189<div class="doc_text">
6190
6191<h5>Syntax:</h5>
6192
6193<p>This is an overloaded intrinsic. You can use <tt>llvm.umul.with.overflow</tt>
6194on any integer bit width.</p>
6195
6196<pre>
6197 declare {i16, i1} @llvm.umul.with.overflow.i16(i16 %a, i16 %b)
6198 declare {i32, i1} @llvm.umul.with.overflow.i32(i32 %a, i32 %b)
6199 declare {i64, i1} @llvm.umul.with.overflow.i64(i64 %a, i64 %b)
6200</pre>
6201
6202<h5>Overview:</h5>
6203
6204<p><i><b>Warning:</b> '<tt>llvm.umul.with.overflow</tt>' is badly broken. It is
6205actively being fixed, but it should not currently be used!</i></p>
6206
6207<p>The '<tt>llvm.umul.with.overflow</tt>' family of intrinsic functions perform
6208a unsigned multiplication of the two arguments, and indicate whether an overflow
6209occurred during the unsigned multiplication.</p>
6210
6211<h5>Arguments:</h5>
6212
6213<p>The arguments (%a and %b) and the first element of the result structure may
6214be of integer types of any bit width, but they must have the same bit width. The
6215second element of the result structure must be of type <tt>i1</tt>. <tt>%a</tt>
6216and <tt>%b</tt> are the two values that will undergo unsigned
6217multiplication.</p>
6218
6219<h5>Semantics:</h5>
6220
6221<p>The '<tt>llvm.umul.with.overflow</tt>' family of intrinsic functions perform
6222an unsigned multiplication of the two arguments. They return a structure &mdash;
6223the first element of which is the multiplication, and the second element of
6224which is a bit specifying if the unsigned multiplication resulted in an
6225overflow.</p>
6226
6227<h5>Examples:</h5>
6228<pre>
6229 %res = call {i32, i1} @llvm.umul.with.overflow.i32(i32 %a, i32 %b)
6230 %sum = extractvalue {i32, i1} %res, 0
6231 %obit = extractvalue {i32, i1} %res, 1
6232 br i1 %obit, label %overflow, label %normal
6233</pre>
6234
6235</div>
6236
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]6237<!-- ======================================================================= -->
6238<div class="doc_subsection">
6239 <a name="int_debugger">Debugger Intrinsics</a>
6240</div>
6241
6242<div class="doc_text">
6243<p>
6244The LLVM debugger intrinsics (which all start with <tt>llvm.dbg.</tt> prefix),
6245are described in the <a
6246href="SourceLevelDebugging.html#format_common_intrinsics">LLVM Source Level
6247Debugging</a> document.
6248</p>
6249</div>
6250
6251
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]6252<!-- ======================================================================= -->
6253<div class="doc_subsection">
6254 <a name="int_eh">Exception Handling Intrinsics</a>
6255</div>
6256
6257<div class="doc_text">
6258<p> The LLVM exception handling intrinsics (which all start with
6259<tt>llvm.eh.</tt> prefix), are described in the <a
6260href="ExceptionHandling.html#format_common_intrinsics">LLVM Exception
6261Handling</a> document. </p>
6262</div>
6263
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]6264<!-- ======================================================================= -->
6265<div class="doc_subsection">
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]6266 <a name="int_trampoline">Trampoline Intrinsic</a>
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]6267</div>
6268
6269<div class="doc_text">
6270<p>
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]6271 This intrinsic makes it possible to excise one parameter, marked with
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]6272 the <tt>nest</tt> attribute, from a function. The result is a callable
6273 function pointer lacking the nest parameter - the caller does not need
6274 to provide a value for it. Instead, the value to use is stored in
6275 advance in a "trampoline", a block of memory usually allocated
6276 on the stack, which also contains code to splice the nest value into the
6277 argument list. This is used to implement the GCC nested function address
6278 extension.
6279</p>
6280<p>
6281 For example, if the function is
6282 <tt>i32 f(i8* nest %c, i32 %x, i32 %y)</tt> then the resulting function
Bill Wendling03295ca2007-09-22 09:23:55 +0000[diff] [blame]6283 pointer has signature <tt>i32 (i32, i32)*</tt>. It can be created as follows:</p>
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]6284<pre>
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]6285 %tramp = alloca [10 x i8], align 4 ; size and alignment only correct for X86
6286 %tramp1 = getelementptr [10 x i8]* %tramp, i32 0, i32 0
6287 %p = call i8* @llvm.init.trampoline( i8* %tramp1, i8* bitcast (i32 (i8* nest , i32, i32)* @f to i8*), i8* %nval )
6288 %fp = bitcast i8* %p to i32 (i32, i32)*
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]6289</pre>
Bill Wendling03295ca2007-09-22 09:23:55 +0000[diff] [blame]6290 <p>The call <tt>%val = call i32 %fp( i32 %x, i32 %y )</tt> is then equivalent
6291 to <tt>%val = call i32 %f( i8* %nval, i32 %x, i32 %y )</tt>.</p>
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]6292</div>
6293
6294<!-- _______________________________________________________________________ -->
6295<div class="doc_subsubsection">
6296 <a name="int_it">'<tt>llvm.init.trampoline</tt>' Intrinsic</a>
6297</div>
6298<div class="doc_text">
6299<h5>Syntax:</h5>
6300<pre>
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]6301declare i8* @llvm.init.trampoline(i8* &lt;tramp&gt;, i8* &lt;func&gt;, i8* &lt;nval&gt;)
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]6302</pre>
6303<h5>Overview:</h5>
6304<p>
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]6305 This fills the memory pointed to by <tt>tramp</tt> with code
6306 and returns a function pointer suitable for executing it.
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]6307</p>
6308<h5>Arguments:</h5>
6309<p>
6310 The <tt>llvm.init.trampoline</tt> intrinsic takes three arguments, all
6311 pointers. The <tt>tramp</tt> argument must point to a sufficiently large
6312 and sufficiently aligned block of memory; this memory is written to by the
Duncan Sandsc00c2ba2007-08-22 23:39:54 +0000[diff] [blame]6313 intrinsic. Note that the size and the alignment are target-specific - LLVM
6314 currently provides no portable way of determining them, so a front-end that
6315 generates this intrinsic needs to have some target-specific knowledge.
6316 The <tt>func</tt> argument must hold a function bitcast to an <tt>i8*</tt>.
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]6317</p>
6318<h5>Semantics:</h5>
6319<p>
6320 The block of memory pointed to by <tt>tramp</tt> is filled with target
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]6321 dependent code, turning it into a function. A pointer to this function is
6322 returned, but needs to be bitcast to an
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]6323 <a href="#int_trampoline">appropriate function pointer type</a>
Duncan Sandsf7331b32007-09-11 14:10:23 +0000[diff] [blame]6324 before being called. The new function's signature is the same as that of
6325 <tt>func</tt> with any arguments marked with the <tt>nest</tt> attribute
6326 removed. At most one such <tt>nest</tt> argument is allowed, and it must be
6327 of pointer type. Calling the new function is equivalent to calling
6328 <tt>func</tt> with the same argument list, but with <tt>nval</tt> used for the
6329 missing <tt>nest</tt> argument. If, after calling
6330 <tt>llvm.init.trampoline</tt>, the memory pointed to by <tt>tramp</tt> is
6331 modified, then the effect of any later call to the returned function pointer is
6332 undefined.
Duncan Sands36397f52007-07-27 12:58:54 +0000[diff] [blame]6333</p>
6334</div>
6335
6336<!-- ======================================================================= -->
6337<div class="doc_subsection">
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]6338 <a name="int_atomics">Atomic Operations and Synchronization Intrinsics</a>
6339</div>
6340
6341<div class="doc_text">
6342<p>
6343 These intrinsic functions expand the "universal IR" of LLVM to represent
6344 hardware constructs for atomic operations and memory synchronization. This
6345 provides an interface to the hardware, not an interface to the programmer. It
Chris Lattnerd3eda892008-08-05 18:29:16 +0000[diff] [blame]6346 is aimed at a low enough level to allow any programming models or APIs
6347 (Application Programming Interfaces) which
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]6348 need atomic behaviors to map cleanly onto it. It is also modeled primarily on
6349 hardware behavior. Just as hardware provides a "universal IR" for source
6350 languages, it also provides a starting point for developing a "universal"
6351 atomic operation and synchronization IR.
6352</p>
6353<p>
6354 These do <em>not</em> form an API such as high-level threading libraries,
6355 software transaction memory systems, atomic primitives, and intrinsic
6356 functions as found in BSD, GNU libc, atomic_ops, APR, and other system and
6357 application libraries. The hardware interface provided by LLVM should allow
6358 a clean implementation of all of these APIs and parallel programming models.
6359 No one model or paradigm should be selected above others unless the hardware
6360 itself ubiquitously does so.
6361
6362</p>
6363</div>
6364
6365<!-- _______________________________________________________________________ -->
6366<div class="doc_subsubsection">
6367 <a name="int_memory_barrier">'<tt>llvm.memory.barrier</tt>' Intrinsic</a>
6368</div>
6369<div class="doc_text">
6370<h5>Syntax:</h5>
6371<pre>
6372declare void @llvm.memory.barrier( i1 &lt;ll&gt;, i1 &lt;ls&gt;, i1 &lt;sl&gt;, i1 &lt;ss&gt;,
6373i1 &lt;device&gt; )
6374
6375</pre>
6376<h5>Overview:</h5>
6377<p>
6378 The <tt>llvm.memory.barrier</tt> intrinsic guarantees ordering between
6379 specific pairs of memory access types.
6380</p>
6381<h5>Arguments:</h5>
6382<p>
6383 The <tt>llvm.memory.barrier</tt> intrinsic requires five boolean arguments.
6384 The first four arguments enables a specific barrier as listed below. The fith
6385 argument specifies that the barrier applies to io or device or uncached memory.
6386
6387</p>
6388 <ul>
6389 <li><tt>ll</tt>: load-load barrier</li>
6390 <li><tt>ls</tt>: load-store barrier</li>
6391 <li><tt>sl</tt>: store-load barrier</li>
6392 <li><tt>ss</tt>: store-store barrier</li>
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]6393 <li><tt>device</tt>: barrier applies to device and uncached memory also.</li>
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]6394 </ul>
6395<h5>Semantics:</h5>
6396<p>
6397 This intrinsic causes the system to enforce some ordering constraints upon
6398 the loads and stores of the program. This barrier does not indicate
6399 <em>when</em> any events will occur, it only enforces an <em>order</em> in
6400 which they occur. For any of the specified pairs of load and store operations
6401 (f.ex. load-load, or store-load), all of the first operations preceding the
6402 barrier will complete before any of the second operations succeeding the
6403 barrier begin. Specifically the semantics for each pairing is as follows:
6404</p>
6405 <ul>
6406 <li><tt>ll</tt>: All loads before the barrier must complete before any load
6407 after the barrier begins.</li>
6408
6409 <li><tt>ls</tt>: All loads before the barrier must complete before any
6410 store after the barrier begins.</li>
6411 <li><tt>ss</tt>: All stores before the barrier must complete before any
6412 store after the barrier begins.</li>
6413 <li><tt>sl</tt>: All stores before the barrier must complete before any
6414 load after the barrier begins.</li>
6415 </ul>
6416<p>
6417 These semantics are applied with a logical "and" behavior when more than one
6418 is enabled in a single memory barrier intrinsic.
6419</p>
6420<p>
6421 Backends may implement stronger barriers than those requested when they do not
6422 support as fine grained a barrier as requested. Some architectures do not
6423 need all types of barriers and on such architectures, these become noops.
6424</p>
6425<h5>Example:</h5>
6426<pre>
6427%ptr = malloc i32
6428 store i32 4, %ptr
6429
6430%result1 = load i32* %ptr <i>; yields {i32}:result1 = 4</i>
6431 call void @llvm.memory.barrier( i1 false, i1 true, i1 false, i1 false )
6432 <i>; guarantee the above finishes</i>
6433 store i32 8, %ptr <i>; before this begins</i>
6434</pre>
6435</div>
6436
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6437<!-- _______________________________________________________________________ -->
6438<div class="doc_subsubsection">
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6439 <a name="int_atomic_cmp_swap">'<tt>llvm.atomic.cmp.swap.*</tt>' Intrinsic</a>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6440</div>
6441<div class="doc_text">
6442<h5>Syntax:</h5>
6443<p>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6444 This is an overloaded intrinsic. You can use <tt>llvm.atomic.cmp.swap</tt> on
6445 any integer bit width and for different address spaces. Not all targets
6446 support all bit widths however.</p>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6447
6448<pre>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6449declare i8 @llvm.atomic.cmp.swap.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;cmp&gt;, i8 &lt;val&gt; )
6450declare i16 @llvm.atomic.cmp.swap.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;cmp&gt;, i16 &lt;val&gt; )
6451declare i32 @llvm.atomic.cmp.swap.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;cmp&gt;, i32 &lt;val&gt; )
6452declare i64 @llvm.atomic.cmp.swap.i64.p0i64( i64* &lt;ptr&gt;, i64 &lt;cmp&gt;, i64 &lt;val&gt; )
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6453
6454</pre>
6455<h5>Overview:</h5>
6456<p>
6457 This loads a value in memory and compares it to a given value. If they are
6458 equal, it stores a new value into the memory.
6459</p>
6460<h5>Arguments:</h5>
6461<p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6462 The <tt>llvm.atomic.cmp.swap</tt> intrinsic takes three arguments. The result as
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6463 well as both <tt>cmp</tt> and <tt>val</tt> must be integer values with the
6464 same bit width. The <tt>ptr</tt> argument must be a pointer to a value of
6465 this integer type. While any bit width integer may be used, targets may only
6466 lower representations they support in hardware.
6467
6468</p>
6469<h5>Semantics:</h5>
6470<p>
6471 This entire intrinsic must be executed atomically. It first loads the value
6472 in memory pointed to by <tt>ptr</tt> and compares it with the value
6473 <tt>cmp</tt>. If they are equal, <tt>val</tt> is stored into the memory. The
6474 loaded value is yielded in all cases. This provides the equivalent of an
6475 atomic compare-and-swap operation within the SSA framework.
6476</p>
6477<h5>Examples:</h5>
6478
6479<pre>
6480%ptr = malloc i32
6481 store i32 4, %ptr
6482
6483%val1 = add i32 4, 4
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6484%result1 = call i32 @llvm.atomic.cmp.swap.i32.p0i32( i32* %ptr, i32 4, %val1 )
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6485 <i>; yields {i32}:result1 = 4</i>
6486%stored1 = icmp eq i32 %result1, 4 <i>; yields {i1}:stored1 = true</i>
6487%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 8</i>
6488
6489%val2 = add i32 1, 1
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6490%result2 = call i32 @llvm.atomic.cmp.swap.i32.p0i32( i32* %ptr, i32 5, %val2 )
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6491 <i>; yields {i32}:result2 = 8</i>
6492%stored2 = icmp eq i32 %result2, 5 <i>; yields {i1}:stored2 = false</i>
6493
6494%memval2 = load i32* %ptr <i>; yields {i32}:memval2 = 8</i>
6495</pre>
6496</div>
6497
6498<!-- _______________________________________________________________________ -->
6499<div class="doc_subsubsection">
6500 <a name="int_atomic_swap">'<tt>llvm.atomic.swap.*</tt>' Intrinsic</a>
6501</div>
6502<div class="doc_text">
6503<h5>Syntax:</h5>
6504
6505<p>
6506 This is an overloaded intrinsic. You can use <tt>llvm.atomic.swap</tt> on any
6507 integer bit width. Not all targets support all bit widths however.</p>
6508<pre>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6509declare i8 @llvm.atomic.swap.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;val&gt; )
6510declare i16 @llvm.atomic.swap.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;val&gt; )
6511declare i32 @llvm.atomic.swap.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;val&gt; )
6512declare i64 @llvm.atomic.swap.i64.p0i64( i64* &lt;ptr&gt;, i64 &lt;val&gt; )
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6513
6514</pre>
6515<h5>Overview:</h5>
6516<p>
6517 This intrinsic loads the value stored in memory at <tt>ptr</tt> and yields
6518 the value from memory. It then stores the value in <tt>val</tt> in the memory
6519 at <tt>ptr</tt>.
6520</p>
6521<h5>Arguments:</h5>
6522
6523<p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6524 The <tt>llvm.atomic.swap</tt> intrinsic takes two arguments. Both the
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6525 <tt>val</tt> argument and the result must be integers of the same bit width.
6526 The first argument, <tt>ptr</tt>, must be a pointer to a value of this
6527 integer type. The targets may only lower integer representations they
6528 support.
6529</p>
6530<h5>Semantics:</h5>
6531<p>
6532 This intrinsic loads the value pointed to by <tt>ptr</tt>, yields it, and
6533 stores <tt>val</tt> back into <tt>ptr</tt> atomically. This provides the
6534 equivalent of an atomic swap operation within the SSA framework.
6535
6536</p>
6537<h5>Examples:</h5>
6538<pre>
6539%ptr = malloc i32
6540 store i32 4, %ptr
6541
6542%val1 = add i32 4, 4
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6543%result1 = call i32 @llvm.atomic.swap.i32.p0i32( i32* %ptr, i32 %val1 )
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6544 <i>; yields {i32}:result1 = 4</i>
6545%stored1 = icmp eq i32 %result1, 4 <i>; yields {i1}:stored1 = true</i>
6546%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 8</i>
6547
6548%val2 = add i32 1, 1
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6549%result2 = call i32 @llvm.atomic.swap.i32.p0i32( i32* %ptr, i32 %val2 )
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6550 <i>; yields {i32}:result2 = 8</i>
6551
6552%stored2 = icmp eq i32 %result2, 8 <i>; yields {i1}:stored2 = true</i>
6553%memval2 = load i32* %ptr <i>; yields {i32}:memval2 = 2</i>
6554</pre>
6555</div>
6556
6557<!-- _______________________________________________________________________ -->
6558<div class="doc_subsubsection">
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6559 <a name="int_atomic_load_add">'<tt>llvm.atomic.load.add.*</tt>' Intrinsic</a>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6560
6561</div>
6562<div class="doc_text">
6563<h5>Syntax:</h5>
6564<p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6565 This is an overloaded intrinsic. You can use <tt>llvm.atomic.load.add</tt> on any
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6566 integer bit width. Not all targets support all bit widths however.</p>
6567<pre>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6568declare i8 @llvm.atomic.load.add.i8..p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6569declare i16 @llvm.atomic.load.add.i16..p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6570declare i32 @llvm.atomic.load.add.i32..p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6571declare i64 @llvm.atomic.load.add.i64..p0i64( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6572
6573</pre>
6574<h5>Overview:</h5>
6575<p>
6576 This intrinsic adds <tt>delta</tt> to the value stored in memory at
6577 <tt>ptr</tt>. It yields the original value at <tt>ptr</tt>.
6578</p>
6579<h5>Arguments:</h5>
6580<p>
6581
6582 The intrinsic takes two arguments, the first a pointer to an integer value
6583 and the second an integer value. The result is also an integer value. These
6584 integer types can have any bit width, but they must all have the same bit
6585 width. The targets may only lower integer representations they support.
6586</p>
6587<h5>Semantics:</h5>
6588<p>
6589 This intrinsic does a series of operations atomically. It first loads the
6590 value stored at <tt>ptr</tt>. It then adds <tt>delta</tt>, stores the result
6591 to <tt>ptr</tt>. It yields the original value stored at <tt>ptr</tt>.
6592</p>
6593
6594<h5>Examples:</h5>
6595<pre>
6596%ptr = malloc i32
6597 store i32 4, %ptr
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6598%result1 = call i32 @llvm.atomic.load.add.i32.p0i32( i32* %ptr, i32 4 )
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6599 <i>; yields {i32}:result1 = 4</i>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6600%result2 = call i32 @llvm.atomic.load.add.i32.p0i32( i32* %ptr, i32 2 )
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6601 <i>; yields {i32}:result2 = 8</i>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6602%result3 = call i32 @llvm.atomic.load.add.i32.p0i32( i32* %ptr, i32 5 )
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6603 <i>; yields {i32}:result3 = 10</i>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6604%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 15</i>
Andrew Lenharthab0b9492008-02-21 06:45:13 +0000[diff] [blame]6605</pre>
6606</div>
6607
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6608<!-- _______________________________________________________________________ -->
6609<div class="doc_subsubsection">
6610 <a name="int_atomic_load_sub">'<tt>llvm.atomic.load.sub.*</tt>' Intrinsic</a>
6611
6612</div>
6613<div class="doc_text">
6614<h5>Syntax:</h5>
6615<p>
6616 This is an overloaded intrinsic. You can use <tt>llvm.atomic.load.sub</tt> on
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6617 any integer bit width and for different address spaces. Not all targets
6618 support all bit widths however.</p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6619<pre>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6620declare i8 @llvm.atomic.load.sub.i8.p0i32( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6621declare i16 @llvm.atomic.load.sub.i16.p0i32( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6622declare i32 @llvm.atomic.load.sub.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6623declare i64 @llvm.atomic.load.sub.i64.p0i32( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6624
6625</pre>
6626<h5>Overview:</h5>
6627<p>
6628 This intrinsic subtracts <tt>delta</tt> to the value stored in memory at
6629 <tt>ptr</tt>. It yields the original value at <tt>ptr</tt>.
6630</p>
6631<h5>Arguments:</h5>
6632<p>
6633
6634 The intrinsic takes two arguments, the first a pointer to an integer value
6635 and the second an integer value. The result is also an integer value. These
6636 integer types can have any bit width, but they must all have the same bit
6637 width. The targets may only lower integer representations they support.
6638</p>
6639<h5>Semantics:</h5>
6640<p>
6641 This intrinsic does a series of operations atomically. It first loads the
6642 value stored at <tt>ptr</tt>. It then subtracts <tt>delta</tt>, stores the
6643 result to <tt>ptr</tt>. It yields the original value stored at <tt>ptr</tt>.
6644</p>
6645
6646<h5>Examples:</h5>
6647<pre>
6648%ptr = malloc i32
6649 store i32 8, %ptr
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6650%result1 = call i32 @llvm.atomic.load.sub.i32.p0i32( i32* %ptr, i32 4 )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6651 <i>; yields {i32}:result1 = 8</i>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6652%result2 = call i32 @llvm.atomic.load.sub.i32.p0i32( i32* %ptr, i32 2 )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6653 <i>; yields {i32}:result2 = 4</i>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6654%result3 = call i32 @llvm.atomic.load.sub.i32.p0i32( i32* %ptr, i32 5 )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6655 <i>; yields {i32}:result3 = 2</i>
6656%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = -3</i>
6657</pre>
6658</div>
6659
6660<!-- _______________________________________________________________________ -->
6661<div class="doc_subsubsection">
6662 <a name="int_atomic_load_and">'<tt>llvm.atomic.load.and.*</tt>' Intrinsic</a><br>
6663 <a name="int_atomic_load_nand">'<tt>llvm.atomic.load.nand.*</tt>' Intrinsic</a><br>
6664 <a name="int_atomic_load_or">'<tt>llvm.atomic.load.or.*</tt>' Intrinsic</a><br>
6665 <a name="int_atomic_load_xor">'<tt>llvm.atomic.load.xor.*</tt>' Intrinsic</a><br>
6666
6667</div>
6668<div class="doc_text">
6669<h5>Syntax:</h5>
6670<p>
6671 These are overloaded intrinsics. You can use <tt>llvm.atomic.load_and</tt>,
6672 <tt>llvm.atomic.load_nand</tt>, <tt>llvm.atomic.load_or</tt>, and
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6673 <tt>llvm.atomic.load_xor</tt> on any integer bit width and for different
6674 address spaces. Not all targets support all bit widths however.</p>
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6675<pre>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6676declare i8 @llvm.atomic.load.and.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6677declare i16 @llvm.atomic.load.and.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6678declare i32 @llvm.atomic.load.and.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6679declare i64 @llvm.atomic.load.and.i64.p0i64( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6680
6681</pre>
6682
6683<pre>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6684declare i8 @llvm.atomic.load.or.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6685declare i16 @llvm.atomic.load.or.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6686declare i32 @llvm.atomic.load.or.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6687declare i64 @llvm.atomic.load.or.i64.p0i64( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6688
6689</pre>
6690
6691<pre>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6692declare i8 @llvm.atomic.load.nand.i8.p0i32( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6693declare i16 @llvm.atomic.load.nand.i16.p0i32( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6694declare i32 @llvm.atomic.load.nand.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6695declare i64 @llvm.atomic.load.nand.i64.p0i32( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6696
6697</pre>
6698
6699<pre>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6700declare i8 @llvm.atomic.load.xor.i8.p0i32( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6701declare i16 @llvm.atomic.load.xor.i16.p0i32( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6702declare i32 @llvm.atomic.load.xor.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6703declare i64 @llvm.atomic.load.xor.i64.p0i32( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6704
6705</pre>
6706<h5>Overview:</h5>
6707<p>
6708 These intrinsics bitwise the operation (and, nand, or, xor) <tt>delta</tt> to
6709 the value stored in memory at <tt>ptr</tt>. It yields the original value
6710 at <tt>ptr</tt>.
6711</p>
6712<h5>Arguments:</h5>
6713<p>
6714
6715 These intrinsics take two arguments, the first a pointer to an integer value
6716 and the second an integer value. The result is also an integer value. These
6717 integer types can have any bit width, but they must all have the same bit
6718 width. The targets may only lower integer representations they support.
6719</p>
6720<h5>Semantics:</h5>
6721<p>
6722 These intrinsics does a series of operations atomically. They first load the
6723 value stored at <tt>ptr</tt>. They then do the bitwise operation
6724 <tt>delta</tt>, store the result to <tt>ptr</tt>. They yield the original
6725 value stored at <tt>ptr</tt>.
6726</p>
6727
6728<h5>Examples:</h5>
6729<pre>
6730%ptr = malloc i32
6731 store i32 0x0F0F, %ptr
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6732%result0 = call i32 @llvm.atomic.load.nand.i32.p0i32( i32* %ptr, i32 0xFF )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6733 <i>; yields {i32}:result0 = 0x0F0F</i>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6734%result1 = call i32 @llvm.atomic.load.and.i32.p0i32( i32* %ptr, i32 0xFF )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6735 <i>; yields {i32}:result1 = 0xFFFFFFF0</i>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6736%result2 = call i32 @llvm.atomic.load.or.i32.p0i32( i32* %ptr, i32 0F )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6737 <i>; yields {i32}:result2 = 0xF0</i>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6738%result3 = call i32 @llvm.atomic.load.xor.i32.p0i32( i32* %ptr, i32 0F )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6739 <i>; yields {i32}:result3 = FF</i>
6740%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = F0</i>
6741</pre>
6742</div>
6743
6744
6745<!-- _______________________________________________________________________ -->
6746<div class="doc_subsubsection">
6747 <a name="int_atomic_load_max">'<tt>llvm.atomic.load.max.*</tt>' Intrinsic</a><br>
6748 <a name="int_atomic_load_min">'<tt>llvm.atomic.load.min.*</tt>' Intrinsic</a><br>
6749 <a name="int_atomic_load_umax">'<tt>llvm.atomic.load.umax.*</tt>' Intrinsic</a><br>
6750 <a name="int_atomic_load_umin">'<tt>llvm.atomic.load.umin.*</tt>' Intrinsic</a><br>
6751
6752</div>
6753<div class="doc_text">
6754<h5>Syntax:</h5>
6755<p>
6756 These are overloaded intrinsics. You can use <tt>llvm.atomic.load_max</tt>,
6757 <tt>llvm.atomic.load_min</tt>, <tt>llvm.atomic.load_umax</tt>, and
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6758 <tt>llvm.atomic.load_umin</tt> on any integer bit width and for different
6759 address spaces. Not all targets
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6760 support all bit widths however.</p>
6761<pre>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6762declare i8 @llvm.atomic.load.max.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6763declare i16 @llvm.atomic.load.max.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6764declare i32 @llvm.atomic.load.max.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6765declare i64 @llvm.atomic.load.max.i64.p0i64( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6766
6767</pre>
6768
6769<pre>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6770declare i8 @llvm.atomic.load.min.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6771declare i16 @llvm.atomic.load.min.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6772declare i32 @llvm.atomic.load.min.i32..p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6773declare i64 @llvm.atomic.load.min.i64..p0i64( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6774
6775</pre>
6776
6777<pre>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6778declare i8 @llvm.atomic.load.umax.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6779declare i16 @llvm.atomic.load.umax.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6780declare i32 @llvm.atomic.load.umax.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6781declare i64 @llvm.atomic.load.umax.i64.p0i64( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6782
6783</pre>
6784
6785<pre>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6786declare i8 @llvm.atomic.load.umin.i8..p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6787declare i16 @llvm.atomic.load.umin.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6788declare i32 @llvm.atomic.load.umin.i32..p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6789declare i64 @llvm.atomic.load.umin.i64..p0i64( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6790
6791</pre>
6792<h5>Overview:</h5>
6793<p>
6794 These intrinsics takes the signed or unsigned minimum or maximum of
6795 <tt>delta</tt> and the value stored in memory at <tt>ptr</tt>. It yields the
6796 original value at <tt>ptr</tt>.
6797</p>
6798<h5>Arguments:</h5>
6799<p>
6800
6801 These intrinsics take two arguments, the first a pointer to an integer value
6802 and the second an integer value. The result is also an integer value. These
6803 integer types can have any bit width, but they must all have the same bit
6804 width. The targets may only lower integer representations they support.
6805</p>
6806<h5>Semantics:</h5>
6807<p>
6808 These intrinsics does a series of operations atomically. They first load the
6809 value stored at <tt>ptr</tt>. They then do the signed or unsigned min or max
6810 <tt>delta</tt> and the value, store the result to <tt>ptr</tt>. They yield
6811 the original value stored at <tt>ptr</tt>.
6812</p>
6813
6814<h5>Examples:</h5>
6815<pre>
6816%ptr = malloc i32
6817 store i32 7, %ptr
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6818%result0 = call i32 @llvm.atomic.load.min.i32.p0i32( i32* %ptr, i32 -2 )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6819 <i>; yields {i32}:result0 = 7</i>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6820%result1 = call i32 @llvm.atomic.load.max.i32.p0i32( i32* %ptr, i32 8 )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6821 <i>; yields {i32}:result1 = -2</i>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6822%result2 = call i32 @llvm.atomic.load.umin.i32.p0i32( i32* %ptr, i32 10 )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6823 <i>; yields {i32}:result2 = 8</i>
Mon P Wange3b3a722008-07-30 04:36:53 +0000[diff] [blame]6824%result3 = call i32 @llvm.atomic.load.umax.i32.p0i32( i32* %ptr, i32 30 )
Mon P Wang28873102008-06-25 08:15:39 +0000[diff] [blame]6825 <i>; yields {i32}:result3 = 8</i>
6826%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 30</i>
6827</pre>
6828</div>
Andrew Lenharth22c5c1b2008-02-16 01:24:58 +0000[diff] [blame]6829
6830<!-- ======================================================================= -->
6831<div class="doc_subsection">
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]6832 <a name="int_general">General Intrinsics</a>
6833</div>
6834
6835<div class="doc_text">
6836<p> This class of intrinsics is designed to be generic and has
6837no specific purpose. </p>
6838</div>
6839
6840<!-- _______________________________________________________________________ -->
6841<div class="doc_subsubsection">
6842 <a name="int_var_annotation">'<tt>llvm.var.annotation</tt>' Intrinsic</a>
6843</div>
6844
6845<div class="doc_text">
6846
6847<h5>Syntax:</h5>
6848<pre>
Tanya Lattnerd2e84422007-06-18 23:42:37 +0000[diff] [blame]6849 declare void @llvm.var.annotation(i8* &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]6850</pre>
6851
6852<h5>Overview:</h5>
6853
6854<p>
6855The '<tt>llvm.var.annotation</tt>' intrinsic
6856</p>
6857
6858<h5>Arguments:</h5>
6859
6860<p>
Tanya Lattnerd2e84422007-06-18 23:42:37 +0000[diff] [blame]6861The first argument is a pointer to a value, the second is a pointer to a
6862global string, the third is a pointer to a global string which is the source
6863file name, and the last argument is the line number.
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]6864</p>
6865
6866<h5>Semantics:</h5>
6867
6868<p>
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]6869This intrinsic allows annotation of local variables with arbitrary strings.
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]6870This can be useful for special purpose optimizations that want to look for these
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]6871annotations. These have no other defined use, they are ignored by code
6872generation and optimization.
6873</p>
Tanya Lattner6d806e92007-06-15 20:50:54 +0000[diff] [blame]6874</div>
6875
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]6876<!-- _______________________________________________________________________ -->
6877<div class="doc_subsubsection">
Tanya Lattnere1a8da02007-09-21 23:57:59 +0000[diff] [blame]6878 <a name="int_annotation">'<tt>llvm.annotation.*</tt>' Intrinsic</a>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]6879</div>
6880
6881<div class="doc_text">
6882
6883<h5>Syntax:</h5>
Tanya Lattner39cfba62007-09-21 23:56:27 +0000[diff] [blame]6884<p>This is an overloaded intrinsic. You can use '<tt>llvm.annotation</tt>' on
6885any integer bit width.
6886</p>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]6887<pre>
Tanya Lattnerd3989a82007-09-22 00:03:01 +0000[diff] [blame]6888 declare i8 @llvm.annotation.i8(i8 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
6889 declare i16 @llvm.annotation.i16(i16 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
6890 declare i32 @llvm.annotation.i32(i32 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
6891 declare i64 @llvm.annotation.i64(i64 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
6892 declare i256 @llvm.annotation.i256(i256 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]6893</pre>
6894
6895<h5>Overview:</h5>
Tanya Lattner39cfba62007-09-21 23:56:27 +0000[diff] [blame]6896
6897<p>
6898The '<tt>llvm.annotation</tt>' intrinsic.
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]6899</p>
6900
6901<h5>Arguments:</h5>
6902
6903<p>
6904The first argument is an integer value (result of some expression),
6905the second is a pointer to a global string, the third is a pointer to a global
6906string which is the source file name, and the last argument is the line number.
Tanya Lattner39cfba62007-09-21 23:56:27 +0000[diff] [blame]6907It returns the value of the first argument.
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]6908</p>
6909
6910<h5>Semantics:</h5>
6911
6912<p>
6913This intrinsic allows annotations to be put on arbitrary expressions
6914with arbitrary strings. This can be useful for special purpose optimizations
6915that want to look for these annotations. These have no other defined use, they
6916are ignored by code generation and optimization.
Dan Gohman0e451ce2008-10-14 16:51:45 +0000[diff] [blame]6917</p>
Tanya Lattnerb6367882007-09-21 22:59:12 +0000[diff] [blame]6918</div>
Jim Laskeydd4ef1b2007-03-14 19:31:19 +0000[diff] [blame]6919
Anton Korobeynikov4cb86182008-01-15 22:31:34 +0000[diff] [blame]6920<!-- _______________________________________________________________________ -->
6921<div class="doc_subsubsection">
6922 <a name="int_trap">'<tt>llvm.trap</tt>' Intrinsic</a>
6923</div>
6924
6925<div class="doc_text">
6926
6927<h5>Syntax:</h5>
6928<pre>
6929 declare void @llvm.trap()
6930</pre>
6931
6932<h5>Overview:</h5>
6933
6934<p>
6935The '<tt>llvm.trap</tt>' intrinsic
6936</p>
6937
6938<h5>Arguments:</h5>
6939
6940<p>
6941None
6942</p>
6943
6944<h5>Semantics:</h5>
6945
6946<p>
6947This intrinsics is lowered to the target dependent trap instruction. If the
6948target does not have a trap instruction, this intrinsic will be lowered to the
6949call of the abort() function.
6950</p>
6951</div>
6952
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]6953<!-- _______________________________________________________________________ -->
6954<div class="doc_subsubsection">
Misha Brukmandccb0252008-11-22 23:55:29 +0000[diff] [blame]6955 <a name="int_stackprotector">'<tt>llvm.stackprotector</tt>' Intrinsic</a>
Bill Wendling69e4adb2008-11-19 05:56:17 +0000[diff] [blame]6956</div>
6957<div class="doc_text">
6958<h5>Syntax:</h5>
6959<pre>
6960declare void @llvm.stackprotector( i8* &lt;guard&gt;, i8** &lt;slot&gt; )
6961
6962</pre>
6963<h5>Overview:</h5>
6964<p>
6965 The <tt>llvm.stackprotector</tt> intrinsic takes the <tt>guard</tt> and stores
6966 it onto the stack at <tt>slot</tt>. The stack slot is adjusted to ensure that
6967 it is placed on the stack before local variables.
6968</p>
6969<h5>Arguments:</h5>
6970<p>
6971 The <tt>llvm.stackprotector</tt> intrinsic requires two pointer arguments. The
6972 first argument is the value loaded from the stack guard
6973 <tt>@__stack_chk_guard</tt>. The second variable is an <tt>alloca</tt> that
6974 has enough space to hold the value of the guard.
6975</p>
6976<h5>Semantics:</h5>
6977<p>
6978 This intrinsic causes the prologue/epilogue inserter to force the position of
6979 the <tt>AllocaInst</tt> stack slot to be before local variables on the
6980 stack. This is to ensure that if a local variable on the stack is overwritten,
6981 it will destroy the value of the guard. When the function exits, the guard on
6982 the stack is checked against the original guard. If they're different, then
6983 the program aborts by calling the <tt>__stack_chk_fail()</tt> function.
6984</p>
6985</div>
6986
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]6987<!-- *********************************************************************** -->
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]6988<hr>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]6989<address>
6990 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
Misha Brukman44408702008-12-11 17:34:48 +0000[diff] [blame]6991 src="http://jigsaw.w3.org/css-validator/images/vcss-blue" alt="Valid CSS"></a>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]6992 <a href="http://validator.w3.org/check/referer"><img
Misha Brukman44408702008-12-11 17:34:48 +0000[diff] [blame]6993 src="http://www.w3.org/Icons/valid-html401-blue" alt="Valid HTML 4.01"></a>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]6994
6995 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
Reid Spencer05fe4b02006-03-14 05:39:39 +0000[diff] [blame]6996 <a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]6997 Last modified: $Date$
6998</address>
Chris Lattnerc7d3ab32008-01-04 04:33:49 +0000[diff] [blame]6999
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]7000</body>
7001</html>