Gitiles
Code Review Sign In
gerrit-public.fairphone.software / fp2-dev / platform / external / llvm / 24bfa21232573864633b9ac159822a2583730c80 / . / docs / LangRef.html
blob: fb3420785e248ae3c5fb76b7515a7ea130b55f24 [file] [log] [blame]
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
2 "http://www.w3.org/TR/html4/strict.dtd">
3<html>
4<head>
5 <title>LLVM Assembly Language Reference Manual</title>
6 <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
7 <meta name="author" content="Chris Lattner">
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]8 <meta name="description"
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]9 content="LLVM Assembly Language Reference Manual.">
10 <link rel="stylesheet" href="llvm.css" type="text/css">
11</head>
12
13<body>
14
15<div class="doc_title"> LLVM Language Reference Manual </div>
16<ol>
17 <li><a href="#abstract">Abstract</a></li>
18 <li><a href="#introduction">Introduction</a></li>
19 <li><a href="#identifiers">Identifiers</a></li>
20 <li><a href="#highlevel">High Level Structure</a>
21 <ol>
22 <li><a href="#modulestructure">Module Structure</a></li>
Bill Wendling41a07852009-07-20 01:03:30 +0000[diff] [blame]23 <li><a href="#linkage">Linkage Types</a>
24 <ol>
Bill Wendlinge2753242009-07-20 02:41:50 +0000[diff] [blame]25 <li><a href="#linkage_private">'<tt>private</tt>' Linkage</a></li>
26 <li><a href="#linkage_linker_private">'<tt>linker_private</tt>' Linkage</a></li>
27 <li><a href="#linkage_internal">'<tt>internal</tt>' Linkage</a></li>
28 <li><a href="#linkage_available_externally">'<tt>available_externally</tt>' Linkage</a></li>
29 <li><a href="#linkage_linkonce">'<tt>linkonce</tt>' Linkage</a></li>
30 <li><a href="#linkage_common">'<tt>common</tt>' Linkage</a></li>
31 <li><a href="#linkage_weak">'<tt>weak</tt>' Linkage</a></li>
32 <li><a href="#linkage_appending">'<tt>appending</tt>' Linkage</a></li>
33 <li><a href="#linkage_externweak">'<tt>extern_weak</tt>' Linkage</a></li>
Chris Lattner0fee5c22009-10-10 18:26:06 +0000[diff] [blame]34 <li><a href="#linkage_linkonce_odr">'<tt>linkonce_odr</tt>' Linkage</a></li>
Bill Wendlinge2753242009-07-20 02:41:50 +0000[diff] [blame]35 <li><a href="#linkage_weak">'<tt>weak_odr</tt>' Linkage</a></li>
36 <li><a href="#linkage_external">'<tt>externally visible</tt>' Linkage</a></li>
37 <li><a href="#linkage_dllimport">'<tt>dllimport</tt>' Linkage</a></li>
38 <li><a href="#linkage_dllexport">'<tt>dllexport</tt>' Linkage</a></li>
Bill Wendling41a07852009-07-20 01:03:30 +0000[diff] [blame]39 </ol>
40 </li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]41 <li><a href="#callingconv">Calling Conventions</a></li>
Chris Lattner5b6dc6e2009-01-11 20:53:49 +0000[diff] [blame]42 <li><a href="#namedtypes">Named Types</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]43 <li><a href="#globalvars">Global Variables</a></li>
44 <li><a href="#functionstructure">Functions</a></li>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]45 <li><a href="#aliasstructure">Aliases</a></li>
Devang Patela4bb6792010-01-11 19:35:55 +0000[diff] [blame]46 <li><a href="#namedmetadatastructure">Named Metadata</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]47 <li><a href="#paramattrs">Parameter Attributes</a></li>
Devang Patel008cd3e2008-09-26 23:51:19 +0000[diff] [blame]48 <li><a href="#fnattrs">Function Attributes</a></li>
Gordon Henriksen13fe5e32007-12-10 03:18:06 +0000[diff] [blame]49 <li><a href="#gc">Garbage Collector Names</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]50 <li><a href="#moduleasm">Module-Level Inline Assembly</a></li>
51 <li><a href="#datalayout">Data Layout</a></li>
Dan Gohman27b47012009-07-27 18:07:55 +0000[diff] [blame]52 <li><a href="#pointeraliasing">Pointer Aliasing Rules</a></li>
Jeffrey Yasskin6dc467f2010-04-26 21:21:24 +0000[diff] [blame]53 <li><a href="#volatile">Volatile Memory Accesses</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]54 </ol>
55 </li>
56 <li><a href="#typesystem">Type System</a>
57 <ol>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]58 <li><a href="#t_classifications">Type Classifications</a></li>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]59 <li><a href="#t_primitive">Primitive Types</a>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]60 <ol>
Nick Lewycky244cf482009-09-27 00:45:11 +0000[diff] [blame]61 <li><a href="#t_integer">Integer Type</a></li>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]62 <li><a href="#t_floating">Floating Point Types</a></li>
63 <li><a href="#t_void">Void Type</a></li>
64 <li><a href="#t_label">Label Type</a></li>
Nick Lewycky29aaef82009-05-30 05:06:04 +0000[diff] [blame]65 <li><a href="#t_metadata">Metadata Type</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]66 </ol>
67 </li>
68 <li><a href="#t_derived">Derived Types</a>
69 <ol>
Chris Lattnerd5d51722010-02-12 20:49:41 +0000[diff] [blame]70 <li><a href="#t_aggregate">Aggregate Types</a>
71 <ol>
72 <li><a href="#t_array">Array Type</a></li>
73 <li><a href="#t_struct">Structure Type</a></li>
74 <li><a href="#t_pstruct">Packed Structure Type</a></li>
75 <li><a href="#t_union">Union Type</a></li>
76 <li><a href="#t_vector">Vector Type</a></li>
77 </ol>
78 </li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]79 <li><a href="#t_function">Function Type</a></li>
80 <li><a href="#t_pointer">Pointer Type</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]81 <li><a href="#t_opaque">Opaque Type</a></li>
82 </ol>
83 </li>
Chris Lattner515195a2009-02-02 07:32:36 +0000[diff] [blame]84 <li><a href="#t_uprefs">Type Up-references</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]85 </ol>
86 </li>
87 <li><a href="#constants">Constants</a>
88 <ol>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]89 <li><a href="#simpleconstants">Simple Constants</a></li>
Chris Lattner97063852009-02-28 18:32:25 +0000[diff] [blame]90 <li><a href="#complexconstants">Complex Constants</a></li>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]91 <li><a href="#globalconstants">Global Variable and Function Addresses</a></li>
92 <li><a href="#undefvalues">Undefined Values</a></li>
Dan Gohmanb8ddf022010-04-22 23:14:21 +0000[diff] [blame]93 <li><a href="#trapvalues">Trap Values</a></li>
Chris Lattner29246b52009-10-27 21:19:13 +0000[diff] [blame]94 <li><a href="#blockaddress">Addresses of Basic Blocks</a></li>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]95 <li><a href="#constantexprs">Constant Expressions</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]96 </ol>
97 </li>
98 <li><a href="#othervalues">Other Values</a>
99 <ol>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]100 <li><a href="#inlineasm">Inline Assembler Expressions</a></li>
Devang Patela4bb6792010-01-11 19:35:55 +0000[diff] [blame]101 <li><a href="#metadata">Metadata Nodes and Metadata Strings</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]102 </ol>
103 </li>
Chris Lattner75c24e02009-07-20 05:55:19 +0000[diff] [blame]104 <li><a href="#intrinsic_globals">Intrinsic Global Variables</a>
105 <ol>
106 <li><a href="#intg_used">The '<tt>llvm.used</tt>' Global Variable</a></li>
Chris Lattner1e0e0d12009-07-20 06:14:25 +0000[diff] [blame]107 <li><a href="#intg_compiler_used">The '<tt>llvm.compiler.used</tt>'
108 Global Variable</a></li>
Chris Lattner75c24e02009-07-20 05:55:19 +0000[diff] [blame]109 <li><a href="#intg_global_ctors">The '<tt>llvm.global_ctors</tt>'
110 Global Variable</a></li>
111 <li><a href="#intg_global_dtors">The '<tt>llvm.global_dtors</tt>'
112 Global Variable</a></li>
113 </ol>
114 </li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]115 <li><a href="#instref">Instruction Reference</a>
116 <ol>
117 <li><a href="#terminators">Terminator Instructions</a>
118 <ol>
119 <li><a href="#i_ret">'<tt>ret</tt>' Instruction</a></li>
120 <li><a href="#i_br">'<tt>br</tt>' Instruction</a></li>
121 <li><a href="#i_switch">'<tt>switch</tt>' Instruction</a></li>
Chris Lattner4c3800f2009-10-28 00:19:10 +0000[diff] [blame]122 <li><a href="#i_indirectbr">'<tt>indirectbr</tt>' Instruction</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]123 <li><a href="#i_invoke">'<tt>invoke</tt>' Instruction</a></li>
124 <li><a href="#i_unwind">'<tt>unwind</tt>' Instruction</a></li>
125 <li><a href="#i_unreachable">'<tt>unreachable</tt>' Instruction</a></li>
126 </ol>
127 </li>
128 <li><a href="#binaryops">Binary Operations</a>
129 <ol>
130 <li><a href="#i_add">'<tt>add</tt>' Instruction</a></li>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]131 <li><a href="#i_fadd">'<tt>fadd</tt>' Instruction</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]132 <li><a href="#i_sub">'<tt>sub</tt>' Instruction</a></li>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]133 <li><a href="#i_fsub">'<tt>fsub</tt>' Instruction</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]134 <li><a href="#i_mul">'<tt>mul</tt>' Instruction</a></li>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]135 <li><a href="#i_fmul">'<tt>fmul</tt>' Instruction</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]136 <li><a href="#i_udiv">'<tt>udiv</tt>' Instruction</a></li>
137 <li><a href="#i_sdiv">'<tt>sdiv</tt>' Instruction</a></li>
138 <li><a href="#i_fdiv">'<tt>fdiv</tt>' Instruction</a></li>
139 <li><a href="#i_urem">'<tt>urem</tt>' Instruction</a></li>
140 <li><a href="#i_srem">'<tt>srem</tt>' Instruction</a></li>
141 <li><a href="#i_frem">'<tt>frem</tt>' Instruction</a></li>
142 </ol>
143 </li>
144 <li><a href="#bitwiseops">Bitwise Binary Operations</a>
145 <ol>
146 <li><a href="#i_shl">'<tt>shl</tt>' Instruction</a></li>
147 <li><a href="#i_lshr">'<tt>lshr</tt>' Instruction</a></li>
148 <li><a href="#i_ashr">'<tt>ashr</tt>' Instruction</a></li>
149 <li><a href="#i_and">'<tt>and</tt>' Instruction</a></li>
150 <li><a href="#i_or">'<tt>or</tt>' Instruction</a></li>
151 <li><a href="#i_xor">'<tt>xor</tt>' Instruction</a></li>
152 </ol>
153 </li>
154 <li><a href="#vectorops">Vector Operations</a>
155 <ol>
156 <li><a href="#i_extractelement">'<tt>extractelement</tt>' Instruction</a></li>
157 <li><a href="#i_insertelement">'<tt>insertelement</tt>' Instruction</a></li>
158 <li><a href="#i_shufflevector">'<tt>shufflevector</tt>' Instruction</a></li>
159 </ol>
160 </li>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]161 <li><a href="#aggregateops">Aggregate Operations</a>
162 <ol>
163 <li><a href="#i_extractvalue">'<tt>extractvalue</tt>' Instruction</a></li>
164 <li><a href="#i_insertvalue">'<tt>insertvalue</tt>' Instruction</a></li>
165 </ol>
166 </li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]167 <li><a href="#memoryops">Memory Access and Addressing Operations</a>
168 <ol>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]169 <li><a href="#i_alloca">'<tt>alloca</tt>' Instruction</a></li>
170 <li><a href="#i_load">'<tt>load</tt>' Instruction</a></li>
171 <li><a href="#i_store">'<tt>store</tt>' Instruction</a></li>
172 <li><a href="#i_getelementptr">'<tt>getelementptr</tt>' Instruction</a></li>
173 </ol>
174 </li>
175 <li><a href="#convertops">Conversion Operations</a>
176 <ol>
177 <li><a href="#i_trunc">'<tt>trunc .. to</tt>' Instruction</a></li>
178 <li><a href="#i_zext">'<tt>zext .. to</tt>' Instruction</a></li>
179 <li><a href="#i_sext">'<tt>sext .. to</tt>' Instruction</a></li>
180 <li><a href="#i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a></li>
181 <li><a href="#i_fpext">'<tt>fpext .. to</tt>' Instruction</a></li>
182 <li><a href="#i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a></li>
183 <li><a href="#i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a></li>
184 <li><a href="#i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a></li>
185 <li><a href="#i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a></li>
186 <li><a href="#i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a></li>
187 <li><a href="#i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a></li>
188 <li><a href="#i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a></li>
189 </ol>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]190 </li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]191 <li><a href="#otherops">Other Operations</a>
192 <ol>
193 <li><a href="#i_icmp">'<tt>icmp</tt>' Instruction</a></li>
194 <li><a href="#i_fcmp">'<tt>fcmp</tt>' Instruction</a></li>
195 <li><a href="#i_phi">'<tt>phi</tt>' Instruction</a></li>
196 <li><a href="#i_select">'<tt>select</tt>' Instruction</a></li>
197 <li><a href="#i_call">'<tt>call</tt>' Instruction</a></li>
198 <li><a href="#i_va_arg">'<tt>va_arg</tt>' Instruction</a></li>
199 </ol>
200 </li>
201 </ol>
202 </li>
203 <li><a href="#intrinsics">Intrinsic Functions</a>
204 <ol>
205 <li><a href="#int_varargs">Variable Argument Handling Intrinsics</a>
206 <ol>
207 <li><a href="#int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a></li>
208 <li><a href="#int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a></li>
209 <li><a href="#int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a></li>
210 </ol>
211 </li>
212 <li><a href="#int_gc">Accurate Garbage Collection Intrinsics</a>
213 <ol>
214 <li><a href="#int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a></li>
215 <li><a href="#int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a></li>
216 <li><a href="#int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a></li>
217 </ol>
218 </li>
219 <li><a href="#int_codegen">Code Generator Intrinsics</a>
220 <ol>
221 <li><a href="#int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a></li>
222 <li><a href="#int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a></li>
223 <li><a href="#int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a></li>
224 <li><a href="#int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a></li>
225 <li><a href="#int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a></li>
226 <li><a href="#int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a></li>
227 <li><a href="#int_readcyclecounter"><tt>llvm.readcyclecounter</tt>' Intrinsic</a></li>
228 </ol>
229 </li>
230 <li><a href="#int_libc">Standard C Library Intrinsics</a>
231 <ol>
232 <li><a href="#int_memcpy">'<tt>llvm.memcpy.*</tt>' Intrinsic</a></li>
233 <li><a href="#int_memmove">'<tt>llvm.memmove.*</tt>' Intrinsic</a></li>
234 <li><a href="#int_memset">'<tt>llvm.memset.*</tt>' Intrinsic</a></li>
235 <li><a href="#int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a></li>
236 <li><a href="#int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a></li>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]237 <li><a href="#int_sin">'<tt>llvm.sin.*</tt>' Intrinsic</a></li>
238 <li><a href="#int_cos">'<tt>llvm.cos.*</tt>' Intrinsic</a></li>
239 <li><a href="#int_pow">'<tt>llvm.pow.*</tt>' Intrinsic</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]240 </ol>
241 </li>
242 <li><a href="#int_manip">Bit Manipulation Intrinsics</a>
243 <ol>
244 <li><a href="#int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a></li>
245 <li><a href="#int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic </a></li>
246 <li><a href="#int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic </a></li>
247 <li><a href="#int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic </a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]248 </ol>
249 </li>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]250 <li><a href="#int_overflow">Arithmetic with Overflow Intrinsics</a>
251 <ol>
Bill Wendling3e1258b2009-02-08 04:04:40 +0000[diff] [blame]252 <li><a href="#int_sadd_overflow">'<tt>llvm.sadd.with.overflow.*</tt> Intrinsics</a></li>
253 <li><a href="#int_uadd_overflow">'<tt>llvm.uadd.with.overflow.*</tt> Intrinsics</a></li>
254 <li><a href="#int_ssub_overflow">'<tt>llvm.ssub.with.overflow.*</tt> Intrinsics</a></li>
255 <li><a href="#int_usub_overflow">'<tt>llvm.usub.with.overflow.*</tt> Intrinsics</a></li>
256 <li><a href="#int_smul_overflow">'<tt>llvm.smul.with.overflow.*</tt> Intrinsics</a></li>
Bill Wendlingbda98b62009-02-08 23:00:09 +0000[diff] [blame]257 <li><a href="#int_umul_overflow">'<tt>llvm.umul.with.overflow.*</tt> Intrinsics</a></li>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]258 </ol>
259 </li>
Anton Korobeynikov8c98a892010-03-14 18:42:47 +0000[diff] [blame]260 <li><a href="#int_fp16">Half Precision Floating Point Intrinsics</a>
261 <ol>
Chris Lattnerebc48e52010-03-14 23:03:31 +0000[diff] [blame]262 <li><a href="#int_convert_to_fp16">'<tt>llvm.convert.to.fp16</tt>' Intrinsic</a></li>
263 <li><a href="#int_convert_from_fp16">'<tt>llvm.convert.from.fp16</tt>' Intrinsic</a></li>
Anton Korobeynikov8c98a892010-03-14 18:42:47 +0000[diff] [blame]264 </ol>
265 </li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]266 <li><a href="#int_debugger">Debugger intrinsics</a></li>
267 <li><a href="#int_eh">Exception Handling intrinsics</a></li>
Duncan Sands7407a9f2007-09-11 14:10:23 +0000[diff] [blame]268 <li><a href="#int_trampoline">Trampoline Intrinsic</a>
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]269 <ol>
270 <li><a href="#int_it">'<tt>llvm.init.trampoline</tt>' Intrinsic</a></li>
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]271 </ol>
272 </li>
Bill Wendling9127adb2008-11-18 22:10:53 +0000[diff] [blame]273 <li><a href="#int_atomics">Atomic intrinsics</a>
274 <ol>
275 <li><a href="#int_memory_barrier"><tt>llvm.memory_barrier</tt></a></li>
276 <li><a href="#int_atomic_cmp_swap"><tt>llvm.atomic.cmp.swap</tt></a></li>
277 <li><a href="#int_atomic_swap"><tt>llvm.atomic.swap</tt></a></li>
278 <li><a href="#int_atomic_load_add"><tt>llvm.atomic.load.add</tt></a></li>
279 <li><a href="#int_atomic_load_sub"><tt>llvm.atomic.load.sub</tt></a></li>
280 <li><a href="#int_atomic_load_and"><tt>llvm.atomic.load.and</tt></a></li>
281 <li><a href="#int_atomic_load_nand"><tt>llvm.atomic.load.nand</tt></a></li>
282 <li><a href="#int_atomic_load_or"><tt>llvm.atomic.load.or</tt></a></li>
283 <li><a href="#int_atomic_load_xor"><tt>llvm.atomic.load.xor</tt></a></li>
284 <li><a href="#int_atomic_load_max"><tt>llvm.atomic.load.max</tt></a></li>
285 <li><a href="#int_atomic_load_min"><tt>llvm.atomic.load.min</tt></a></li>
286 <li><a href="#int_atomic_load_umax"><tt>llvm.atomic.load.umax</tt></a></li>
287 <li><a href="#int_atomic_load_umin"><tt>llvm.atomic.load.umin</tt></a></li>
288 </ol>
289 </li>
Nick Lewyckyc888d352009-10-13 07:03:23 +0000[diff] [blame]290 <li><a href="#int_memorymarkers">Memory Use Markers</a>
291 <ol>
292 <li><a href="#int_lifetime_start"><tt>llvm.lifetime.start</tt></a></li>
293 <li><a href="#int_lifetime_end"><tt>llvm.lifetime.end</tt></a></li>
294 <li><a href="#int_invariant_start"><tt>llvm.invariant.start</tt></a></li>
295 <li><a href="#int_invariant_end"><tt>llvm.invariant.end</tt></a></li>
296 </ol>
297 </li>
Reid Spencerb043f672007-07-20 19:59:11 +0000[diff] [blame]298 <li><a href="#int_general">General intrinsics</a>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]299 <ol>
Reid Spencerb043f672007-07-20 19:59:11 +0000[diff] [blame]300 <li><a href="#int_var_annotation">
Bill Wendlinge4164592008-11-19 05:56:17 +0000[diff] [blame]301 '<tt>llvm.var.annotation</tt>' Intrinsic</a></li>
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]302 <li><a href="#int_annotation">
Bill Wendlinge4164592008-11-19 05:56:17 +0000[diff] [blame]303 '<tt>llvm.annotation.*</tt>' Intrinsic</a></li>
Anton Korobeynikove6e764f2008-01-15 22:31:34 +0000[diff] [blame]304 <li><a href="#int_trap">
Bill Wendlinge4164592008-11-19 05:56:17 +0000[diff] [blame]305 '<tt>llvm.trap</tt>' Intrinsic</a></li>
306 <li><a href="#int_stackprotector">
307 '<tt>llvm.stackprotector</tt>' Intrinsic</a></li>
Eric Christopher767a3722009-11-30 08:03:53 +0000[diff] [blame]308 <li><a href="#int_objectsize">
309 '<tt>llvm.objectsize</tt>' Intrinsic</a></li>
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]310 </ol>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]311 </li>
312 </ol>
313 </li>
314</ol>
315
316<div class="doc_author">
317 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>
318 and <a href="mailto:vadve@cs.uiuc.edu">Vikram Adve</a></p>
319</div>
320
321<!-- *********************************************************************** -->
322<div class="doc_section"> <a name="abstract">Abstract </a></div>
323<!-- *********************************************************************** -->
324
325<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]326
327<p>This document is a reference manual for the LLVM assembly language. LLVM is
328 a Static Single Assignment (SSA) based representation that provides type
329 safety, low-level operations, flexibility, and the capability of representing
330 'all' high-level languages cleanly. It is the common code representation
331 used throughout all phases of the LLVM compilation strategy.</p>
332
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]333</div>
334
335<!-- *********************************************************************** -->
336<div class="doc_section"> <a name="introduction">Introduction</a> </div>
337<!-- *********************************************************************** -->
338
339<div class="doc_text">
340
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]341<p>The LLVM code representation is designed to be used in three different forms:
342 as an in-memory compiler IR, as an on-disk bitcode representation (suitable
343 for fast loading by a Just-In-Time compiler), and as a human readable
344 assembly language representation. This allows LLVM to provide a powerful
345 intermediate representation for efficient compiler transformations and
346 analysis, while providing a natural means to debug and visualize the
347 transformations. The three different forms of LLVM are all equivalent. This
348 document describes the human readable representation and notation.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]349
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]350<p>The LLVM representation aims to be light-weight and low-level while being
351 expressive, typed, and extensible at the same time. It aims to be a
352 "universal IR" of sorts, by being at a low enough level that high-level ideas
353 may be cleanly mapped to it (similar to how microprocessors are "universal
354 IR's", allowing many source languages to be mapped to them). By providing
355 type information, LLVM can be used as the target of optimizations: for
356 example, through pointer analysis, it can be proven that a C automatic
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]357 variable is never accessed outside of the current function, allowing it to
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]358 be promoted to a simple SSA value instead of a memory location.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]359
360</div>
361
362<!-- _______________________________________________________________________ -->
363<div class="doc_subsubsection"> <a name="wellformed">Well-Formedness</a> </div>
364
365<div class="doc_text">
366
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]367<p>It is important to note that this document describes 'well formed' LLVM
368 assembly language. There is a difference between what the parser accepts and
369 what is considered 'well formed'. For example, the following instruction is
370 syntactically okay, but not well formed:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]371
372<div class="doc_code">
373<pre>
374%x = <a href="#i_add">add</a> i32 1, %x
375</pre>
376</div>
377
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]378<p>because the definition of <tt>%x</tt> does not dominate all of its uses. The
379 LLVM infrastructure provides a verification pass that may be used to verify
380 that an LLVM module is well formed. This pass is automatically run by the
381 parser after parsing input assembly and by the optimizer before it outputs
382 bitcode. The violations pointed out by the verifier pass indicate bugs in
383 transformation passes or input to the parser.</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]384
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]385</div>
386
Chris Lattnera83fdc02007-10-03 17:34:29 +0000[diff] [blame]387<!-- Describe the typesetting conventions here. -->
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]388
389<!-- *********************************************************************** -->
390<div class="doc_section"> <a name="identifiers">Identifiers</a> </div>
391<!-- *********************************************************************** -->
392
393<div class="doc_text">
394
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]395<p>LLVM identifiers come in two basic types: global and local. Global
396 identifiers (functions, global variables) begin with the <tt>'@'</tt>
397 character. Local identifiers (register names, types) begin with
398 the <tt>'%'</tt> character. Additionally, there are three different formats
399 for identifiers, for different purposes:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]400
401<ol>
Reid Spencerc8245b02007-08-07 14:34:28 +0000[diff] [blame]402 <li>Named values are represented as a string of characters with their prefix.
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]403 For example, <tt>%foo</tt>, <tt>@DivisionByZero</tt>,
404 <tt>%a.really.long.identifier</tt>. The actual regular expression used is
405 '<tt>[%@][a-zA-Z$._][a-zA-Z$._0-9]*</tt>'. Identifiers which require
406 other characters in their names can be surrounded with quotes. Special
407 characters may be escaped using <tt>"\xx"</tt> where <tt>xx</tt> is the
408 ASCII code for the character in hexadecimal. In this way, any character
409 can be used in a name value, even quotes themselves.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]410
Reid Spencerc8245b02007-08-07 14:34:28 +0000[diff] [blame]411 <li>Unnamed values are represented as an unsigned numeric value with their
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]412 prefix. For example, <tt>%12</tt>, <tt>@2</tt>, <tt>%44</tt>.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]413
414 <li>Constants, which are described in a <a href="#constants">section about
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]415 constants</a>, below.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]416</ol>
417
Reid Spencerc8245b02007-08-07 14:34:28 +0000[diff] [blame]418<p>LLVM requires that values start with a prefix for two reasons: Compilers
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]419 don't need to worry about name clashes with reserved words, and the set of
420 reserved words may be expanded in the future without penalty. Additionally,
421 unnamed identifiers allow a compiler to quickly come up with a temporary
422 variable without having to avoid symbol table conflicts.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]423
424<p>Reserved words in LLVM are very similar to reserved words in other
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]425 languages. There are keywords for different opcodes
426 ('<tt><a href="#i_add">add</a></tt>',
427 '<tt><a href="#i_bitcast">bitcast</a></tt>',
428 '<tt><a href="#i_ret">ret</a></tt>', etc...), for primitive type names
429 ('<tt><a href="#t_void">void</a></tt>',
430 '<tt><a href="#t_primitive">i32</a></tt>', etc...), and others. These
431 reserved words cannot conflict with variable names, because none of them
432 start with a prefix character (<tt>'%'</tt> or <tt>'@'</tt>).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]433
434<p>Here is an example of LLVM code to multiply the integer variable
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]435 '<tt>%X</tt>' by 8:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]436
437<p>The easy way:</p>
438
439<div class="doc_code">
440<pre>
441%result = <a href="#i_mul">mul</a> i32 %X, 8
442</pre>
443</div>
444
445<p>After strength reduction:</p>
446
447<div class="doc_code">
448<pre>
449%result = <a href="#i_shl">shl</a> i32 %X, i8 3
450</pre>
451</div>
452
453<p>And the hard way:</p>
454
455<div class="doc_code">
456<pre>
Gabor Greifc0ea7672009-10-28 13:05:07 +0000[diff] [blame]457%0 = <a href="#i_add">add</a> i32 %X, %X <i>; yields {i32}:%0</i>
458%1 = <a href="#i_add">add</a> i32 %0, %0 <i>; yields {i32}:%1</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]459%result = <a href="#i_add">add</a> i32 %1, %1
460</pre>
461</div>
462
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]463<p>This last way of multiplying <tt>%X</tt> by 8 illustrates several important
464 lexical features of LLVM:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]465
466<ol>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]467 <li>Comments are delimited with a '<tt>;</tt>' and go until the end of
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]468 line.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]469
470 <li>Unnamed temporaries are created when the result of a computation is not
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]471 assigned to a named value.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]472
473 <li>Unnamed temporaries are numbered sequentially</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]474</ol>
475
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]476<p>It also shows a convention that we follow in this document. When
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]477 demonstrating instructions, we will follow an instruction with a comment that
478 defines the type and name of value produced. Comments are shown in italic
479 text.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]480
481</div>
482
483<!-- *********************************************************************** -->
484<div class="doc_section"> <a name="highlevel">High Level Structure</a> </div>
485<!-- *********************************************************************** -->
486
487<!-- ======================================================================= -->
488<div class="doc_subsection"> <a name="modulestructure">Module Structure</a>
489</div>
490
491<div class="doc_text">
492
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]493<p>LLVM programs are composed of "Module"s, each of which is a translation unit
494 of the input programs. Each module consists of functions, global variables,
495 and symbol table entries. Modules may be combined together with the LLVM
496 linker, which merges function (and global variable) definitions, resolves
497 forward declarations, and merges symbol table entries. Here is an example of
498 the "hello world" module:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]499
500<div class="doc_code">
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]501<pre>
502<i>; Declare the string constant as a global constant.</i>
503<a href="#identifiers">@.LC0</a> = <a href="#linkage_internal">internal</a> <a href="#globalvars">constant</a> <a href="#t_array">[13 x i8]</a> c"hello world\0A\00" <i>; [13 x i8]*</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]504
505<i>; External declaration of the puts function</i>
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]506<a href="#functionstructure">declare</a> i32 @puts(i8 *) <i>; i32(i8 *)* </i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]507
508<i>; Definition of main function</i>
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]509define i32 @main() { <i>; i32()* </i>
510 <i>; Convert [13 x i8]* to i8 *...</i>
511 %cast210 = <a href="#i_getelementptr">getelementptr</a> [13 x i8]* @.LC0, i64 0, i64 0 <i>; i8 *</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]512
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]513 <i>; Call puts function to write out the string to stdout.</i>
514 <a href="#i_call">call</a> i32 @puts(i8 * %cast210) <i>; i32</i>
Devang Patela4bb6792010-01-11 19:35:55 +0000[diff] [blame]515 <a href="#i_ret">ret</a> i32 0<br>}
516
517<i>; Named metadata</i>
518!1 = metadata !{i32 41}
519!foo = !{!1, null}
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]520</pre>
521</div>
522
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]523<p>This example is made up of a <a href="#globalvars">global variable</a> named
Devang Patela4bb6792010-01-11 19:35:55 +0000[diff] [blame]524 "<tt>.LC0</tt>", an external declaration of the "<tt>puts</tt>" function,
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]525 a <a href="#functionstructure">function definition</a> for
Devang Patela4bb6792010-01-11 19:35:55 +0000[diff] [blame]526 "<tt>main</tt>" and <a href="#namedmetadatastructure">named metadata</a>
527 "<tt>foo"</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]528
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]529<p>In general, a module is made up of a list of global values, where both
530 functions and global variables are global values. Global values are
531 represented by a pointer to a memory location (in this case, a pointer to an
532 array of char, and a pointer to a function), and have one of the
533 following <a href="#linkage">linkage types</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]534
535</div>
536
537<!-- ======================================================================= -->
538<div class="doc_subsection">
539 <a name="linkage">Linkage Types</a>
540</div>
541
542<div class="doc_text">
543
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]544<p>All Global Variables and Functions have one of the following types of
545 linkage:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]546
547<dl>
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]548 <dt><tt><b><a name="linkage_private">private</a></b></tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]549 <dd>Global values with private linkage are only directly accessible by objects
550 in the current module. In particular, linking code into a module with an
551 private global value may cause the private to be renamed as necessary to
552 avoid collisions. Because the symbol is private to the module, all
553 references can be updated. This doesn't show up in any symbol table in the
554 object file.</dd>
Rafael Espindolaa168fc92009-01-15 20:18:42 +0000[diff] [blame]555
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]556 <dt><tt><b><a name="linkage_linker_private">linker_private</a></b></tt></dt>
Bill Wendling41a07852009-07-20 01:03:30 +0000[diff] [blame]557 <dd>Similar to private, but the symbol is passed through the assembler and
Chris Lattner06c1ecc2009-08-24 04:32:16 +0000[diff] [blame]558 removed by the linker after evaluation. Note that (unlike private
559 symbols) linker_private symbols are subject to coalescing by the linker:
560 weak symbols get merged and redefinitions are rejected. However, unlike
561 normal strong symbols, they are removed by the linker from the final
562 linked image (executable or dynamic library).</dd>
Bill Wendling41a07852009-07-20 01:03:30 +0000[diff] [blame]563
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]564 <dt><tt><b><a name="linkage_internal">internal</a></b></tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]565 <dd>Similar to private, but the value shows as a local symbol
566 (<tt>STB_LOCAL</tt> in the case of ELF) in the object file. This
567 corresponds to the notion of the '<tt>static</tt>' keyword in C.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]568
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]569 <dt><tt><b><a name="linkage_available_externally">available_externally</a></b></tt></dt>
Chris Lattner68433442009-04-13 05:44:34 +0000[diff] [blame]570 <dd>Globals with "<tt>available_externally</tt>" linkage are never emitted
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]571 into the object file corresponding to the LLVM module. They exist to
572 allow inlining and other optimizations to take place given knowledge of
573 the definition of the global, which is known to be somewhere outside the
574 module. Globals with <tt>available_externally</tt> linkage are allowed to
575 be discarded at will, and are otherwise the same as <tt>linkonce_odr</tt>.
576 This linkage type is only allowed on definitions, not declarations.</dd>
Chris Lattner68433442009-04-13 05:44:34 +0000[diff] [blame]577
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]578 <dt><tt><b><a name="linkage_linkonce">linkonce</a></b></tt></dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]579 <dd>Globals with "<tt>linkonce</tt>" linkage are merged with other globals of
Chris Lattnerd16c5512010-01-09 19:15:14 +0000[diff] [blame]580 the same name when linkage occurs. This can be used to implement
581 some forms of inline functions, templates, or other code which must be
582 generated in each translation unit that uses it, but where the body may
583 be overridden with a more definitive definition later. Unreferenced
584 <tt>linkonce</tt> globals are allowed to be discarded. Note that
585 <tt>linkonce</tt> linkage does not actually allow the optimizer to
586 inline the body of this function into callers because it doesn't know if
587 this definition of the function is the definitive definition within the
588 program or whether it will be overridden by a stronger definition.
589 To enable inlining and other optimizations, use "<tt>linkonce_odr</tt>"
590 linkage.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]591
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]592 <dt><tt><b><a name="linkage_weak">weak</a></b></tt></dt>
Chris Lattner7215c7f2009-08-05 05:21:07 +0000[diff] [blame]593 <dd>"<tt>weak</tt>" linkage has the same merging semantics as
594 <tt>linkonce</tt> linkage, except that unreferenced globals with
595 <tt>weak</tt> linkage may not be discarded. This is used for globals that
596 are declared "weak" in C source code.</dd>
597
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]598 <dt><tt><b><a name="linkage_common">common</a></b></tt></dt>
Chris Lattner7215c7f2009-08-05 05:21:07 +0000[diff] [blame]599 <dd>"<tt>common</tt>" linkage is most similar to "<tt>weak</tt>" linkage, but
600 they are used for tentative definitions in C, such as "<tt>int X;</tt>" at
601 global scope.
602 Symbols with "<tt>common</tt>" linkage are merged in the same way as
603 <tt>weak symbols</tt>, and they may not be deleted if unreferenced.
Chris Lattner83c4a712009-08-05 05:41:44 +0000[diff] [blame]604 <tt>common</tt> symbols may not have an explicit section,
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]605 must have a zero initializer, and may not be marked '<a
Chris Lattner83c4a712009-08-05 05:41:44 +0000[diff] [blame]606 href="#globalvars"><tt>constant</tt></a>'. Functions and aliases may not
607 have common linkage.</dd>
Chris Lattner7215c7f2009-08-05 05:21:07 +0000[diff] [blame]608
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]609
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]610 <dt><tt><b><a name="linkage_appending">appending</a></b></tt></dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]611 <dd>"<tt>appending</tt>" linkage may only be applied to global variables of
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]612 pointer to array type. When two global variables with appending linkage
613 are linked together, the two global arrays are appended together. This is
614 the LLVM, typesafe, equivalent of having the system linker append together
615 "sections" with identical names when .o files are linked.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]616
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]617 <dt><tt><b><a name="linkage_externweak">extern_weak</a></b></tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]618 <dd>The semantics of this linkage follow the ELF object file model: the symbol
619 is weak until linked, if not linked, the symbol becomes null instead of
620 being an undefined reference.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]621
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]622 <dt><tt><b><a name="linkage_linkonce_odr">linkonce_odr</a></b></tt></dt>
623 <dt><tt><b><a name="linkage_weak_odr">weak_odr</a></b></tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]624 <dd>Some languages allow differing globals to be merged, such as two functions
625 with different semantics. Other languages, such as <tt>C++</tt>, ensure
626 that only equivalent globals are ever merged (the "one definition rule" -
627 "ODR"). Such languages can use the <tt>linkonce_odr</tt>
628 and <tt>weak_odr</tt> linkage types to indicate that the global will only
629 be merged with equivalent globals. These linkage types are otherwise the
630 same as their non-<tt>odr</tt> versions.</dd>
Duncan Sands19d161f2009-03-07 15:45:40 +0000[diff] [blame]631
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]632 <dt><tt><b><a name="linkage_external">externally visible</a></b></tt>:</dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]633 <dd>If none of the above identifiers are used, the global is externally
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]634 visible, meaning that it participates in linkage and can be used to
635 resolve external symbol references.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]636</dl>
637
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]638<p>The next two types of linkage are targeted for Microsoft Windows platform
639 only. They are designed to support importing (exporting) symbols from (to)
640 DLLs (Dynamic Link Libraries).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]641
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]642<dl>
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]643 <dt><tt><b><a name="linkage_dllimport">dllimport</a></b></tt></dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]644 <dd>"<tt>dllimport</tt>" linkage causes the compiler to reference a function
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]645 or variable via a global pointer to a pointer that is set up by the DLL
646 exporting the symbol. On Microsoft Windows targets, the pointer name is
647 formed by combining <code>__imp_</code> and the function or variable
648 name.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]649
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]650 <dt><tt><b><a name="linkage_dllexport">dllexport</a></b></tt></dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]651 <dd>"<tt>dllexport</tt>" linkage causes the compiler to provide a global
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]652 pointer to a pointer in a DLL, so that it can be referenced with the
653 <tt>dllimport</tt> attribute. On Microsoft Windows targets, the pointer
654 name is formed by combining <code>__imp_</code> and the function or
655 variable name.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]656</dl>
657
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]658<p>For example, since the "<tt>.LC0</tt>" variable is defined to be internal, if
659 another module defined a "<tt>.LC0</tt>" variable and was linked with this
660 one, one of the two would be renamed, preventing a collision. Since
661 "<tt>main</tt>" and "<tt>puts</tt>" are external (i.e., lacking any linkage
662 declarations), they are accessible outside of the current module.</p>
663
664<p>It is illegal for a function <i>declaration</i> to have any linkage type
665 other than "externally visible", <tt>dllimport</tt>
666 or <tt>extern_weak</tt>.</p>
667
Duncan Sands19d161f2009-03-07 15:45:40 +0000[diff] [blame]668<p>Aliases can have only <tt>external</tt>, <tt>internal</tt>, <tt>weak</tt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]669 or <tt>weak_odr</tt> linkages.</p>
670
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]671</div>
672
673<!-- ======================================================================= -->
674<div class="doc_subsection">
675 <a name="callingconv">Calling Conventions</a>
676</div>
677
678<div class="doc_text">
679
680<p>LLVM <a href="#functionstructure">functions</a>, <a href="#i_call">calls</a>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]681 and <a href="#i_invoke">invokes</a> can all have an optional calling
682 convention specified for the call. The calling convention of any pair of
683 dynamic caller/callee must match, or the behavior of the program is
684 undefined. The following calling conventions are supported by LLVM, and more
685 may be added in the future:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]686
687<dl>
688 <dt><b>"<tt>ccc</tt>" - The C calling convention</b>:</dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]689 <dd>This calling convention (the default if no other calling convention is
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]690 specified) matches the target C calling conventions. This calling
691 convention supports varargs function calls and tolerates some mismatch in
692 the declared prototype and implemented declaration of the function (as
693 does normal C).</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]694
695 <dt><b>"<tt>fastcc</tt>" - The fast calling convention</b>:</dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]696 <dd>This calling convention attempts to make calls as fast as possible
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]697 (e.g. by passing things in registers). This calling convention allows the
698 target to use whatever tricks it wants to produce fast code for the
699 target, without having to conform to an externally specified ABI
Jeffrey Yasskin3e1a3a02010-01-09 19:44:16 +0000[diff] [blame]700 (Application Binary Interface).
701 <a href="CodeGenerator.html#tailcallopt">Tail calls can only be optimized
Chris Lattnerac9a9392010-03-11 00:22:57 +0000[diff] [blame]702 when this or the GHC convention is used.</a> This calling convention
703 does not support varargs and requires the prototype of all callees to
704 exactly match the prototype of the function definition.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]705
706 <dt><b>"<tt>coldcc</tt>" - The cold calling convention</b>:</dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]707 <dd>This calling convention attempts to make code in the caller as efficient
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]708 as possible under the assumption that the call is not commonly executed.
709 As such, these calls often preserve all registers so that the call does
710 not break any live ranges in the caller side. This calling convention
711 does not support varargs and requires the prototype of all callees to
712 exactly match the prototype of the function definition.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]713
Chris Lattnerac9a9392010-03-11 00:22:57 +0000[diff] [blame]714 <dt><b>"<tt>cc <em>10</em></tt>" - GHC convention</b>:</dt>
715 <dd>This calling convention has been implemented specifically for use by the
716 <a href="http://www.haskell.org/ghc">Glasgow Haskell Compiler (GHC)</a>.
717 It passes everything in registers, going to extremes to achieve this by
718 disabling callee save registers. This calling convention should not be
719 used lightly but only for specific situations such as an alternative to
720 the <em>register pinning</em> performance technique often used when
721 implementing functional programming languages.At the moment only X86
722 supports this convention and it has the following limitations:
723 <ul>
724 <li>On <em>X86-32</em> only supports up to 4 bit type parameters. No
725 floating point types are supported.</li>
726 <li>On <em>X86-64</em> only supports up to 10 bit type parameters and
727 6 floating point parameters.</li>
728 </ul>
729 This calling convention supports
730 <a href="CodeGenerator.html#tailcallopt">tail call optimization</a> but
731 requires both the caller and callee are using it.
732 </dd>
733
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]734 <dt><b>"<tt>cc &lt;<em>n</em>&gt;</tt>" - Numbered convention</b>:</dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]735 <dd>Any calling convention may be specified by number, allowing
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]736 target-specific calling conventions to be used. Target specific calling
737 conventions start at 64.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]738</dl>
739
740<p>More calling conventions can be added/defined on an as-needed basis, to
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]741 support Pascal conventions or any other well-known target-independent
742 convention.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]743
744</div>
745
746<!-- ======================================================================= -->
747<div class="doc_subsection">
748 <a name="visibility">Visibility Styles</a>
749</div>
750
751<div class="doc_text">
752
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]753<p>All Global Variables and Functions have one of the following visibility
754 styles:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]755
756<dl>
757 <dt><b>"<tt>default</tt>" - Default style</b>:</dt>
Chris Lattner96451482008-08-05 18:29:16 +0000[diff] [blame]758 <dd>On targets that use the ELF object file format, default visibility means
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]759 that the declaration is visible to other modules and, in shared libraries,
760 means that the declared entity may be overridden. On Darwin, default
761 visibility means that the declaration is visible to other modules. Default
762 visibility corresponds to "external linkage" in the language.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]763
764 <dt><b>"<tt>hidden</tt>" - Hidden style</b>:</dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]765 <dd>Two declarations of an object with hidden visibility refer to the same
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]766 object if they are in the same shared object. Usually, hidden visibility
767 indicates that the symbol will not be placed into the dynamic symbol
768 table, so no other module (executable or shared library) can reference it
769 directly.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]770
771 <dt><b>"<tt>protected</tt>" - Protected style</b>:</dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]772 <dd>On ELF, protected visibility indicates that the symbol will be placed in
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]773 the dynamic symbol table, but that references within the defining module
774 will bind to the local symbol. That is, the symbol cannot be overridden by
775 another module.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]776</dl>
777
778</div>
779
780<!-- ======================================================================= -->
781<div class="doc_subsection">
Chris Lattner5b6dc6e2009-01-11 20:53:49 +0000[diff] [blame]782 <a name="namedtypes">Named Types</a>
783</div>
784
785<div class="doc_text">
786
787<p>LLVM IR allows you to specify name aliases for certain types. This can make
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]788 it easier to read the IR and make the IR more condensed (particularly when
789 recursive types are involved). An example of a name specification is:</p>
Chris Lattner5b6dc6e2009-01-11 20:53:49 +0000[diff] [blame]790
791<div class="doc_code">
792<pre>
793%mytype = type { %mytype*, i32 }
794</pre>
795</div>
796
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]797<p>You may give a name to any <a href="#typesystem">type</a> except
798 "<a href="t_void">void</a>". Type name aliases may be used anywhere a type
799 is expected with the syntax "%mytype".</p>
Chris Lattner5b6dc6e2009-01-11 20:53:49 +0000[diff] [blame]800
801<p>Note that type names are aliases for the structural type that they indicate,
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]802 and that you can therefore specify multiple names for the same type. This
803 often leads to confusing behavior when dumping out a .ll file. Since LLVM IR
804 uses structural typing, the name is not part of the type. When printing out
805 LLVM IR, the printer will pick <em>one name</em> to render all types of a
806 particular shape. This means that if you have code where two different
807 source types end up having the same LLVM type, that the dumper will sometimes
808 print the "wrong" or unexpected type. This is an important design point and
809 isn't going to change.</p>
Chris Lattner5b6dc6e2009-01-11 20:53:49 +0000[diff] [blame]810
811</div>
812
Chris Lattner5b6dc6e2009-01-11 20:53:49 +0000[diff] [blame]813<!-- ======================================================================= -->
814<div class="doc_subsection">
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]815 <a name="globalvars">Global Variables</a>
816</div>
817
818<div class="doc_text">
819
820<p>Global variables define regions of memory allocated at compilation time
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]821 instead of run-time. Global variables may optionally be initialized, may
822 have an explicit section to be placed in, and may have an optional explicit
823 alignment specified. A variable may be defined as "thread_local", which
824 means that it will not be shared by threads (each thread will have a
825 separated copy of the variable). A variable may be defined as a global
826 "constant," which indicates that the contents of the variable
827 will <b>never</b> be modified (enabling better optimization, allowing the
828 global data to be placed in the read-only section of an executable, etc).
829 Note that variables that need runtime initialization cannot be marked
830 "constant" as there is a store to the variable.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]831
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]832<p>LLVM explicitly allows <em>declarations</em> of global variables to be marked
833 constant, even if the final definition of the global is not. This capability
834 can be used to enable slightly better optimization of the program, but
835 requires the language definition to guarantee that optimizations based on the
836 'constantness' are valid for the translation units that do not include the
837 definition.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]838
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]839<p>As SSA values, global variables define pointer values that are in scope
840 (i.e. they dominate) all basic blocks in the program. Global variables
841 always define a pointer to their "content" type because they describe a
842 region of memory, and all memory objects in LLVM are accessed through
843 pointers.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]844
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]845<p>A global variable may be declared to reside in a target-specific numbered
846 address space. For targets that support them, address spaces may affect how
847 optimizations are performed and/or what target instructions are used to
848 access the variable. The default address space is zero. The address space
849 qualifier must precede any other attributes.</p>
Christopher Lambdd0049d2007-12-11 09:31:00 +0000[diff] [blame]850
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]851<p>LLVM allows an explicit section to be specified for globals. If the target
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]852 supports it, it will emit globals to the section specified.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]853
854<p>An explicit alignment may be specified for a global. If not present, or if
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]855 the alignment is set to zero, the alignment of the global is set by the
856 target to whatever it feels convenient. If an explicit alignment is
857 specified, the global is forced to have at least that much alignment. All
858 alignments must be a power of 2.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]859
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]860<p>For example, the following defines a global in a numbered address space with
861 an initializer, section, and alignment:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]862
863<div class="doc_code">
864<pre>
Dan Gohman21ef02c2009-01-11 00:40:00 +0000[diff] [blame]865@G = addrspace(5) constant float 1.0, section "foo", align 4
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]866</pre>
867</div>
868
869</div>
870
871
872<!-- ======================================================================= -->
873<div class="doc_subsection">
874 <a name="functionstructure">Functions</a>
875</div>
876
877<div class="doc_text">
878
Dan Gohman22dc6682010-03-01 17:41:39 +0000[diff] [blame]879<p>LLVM function definitions consist of the "<tt>define</tt>" keyword, an
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]880 optional <a href="#linkage">linkage type</a>, an optional
881 <a href="#visibility">visibility style</a>, an optional
882 <a href="#callingconv">calling convention</a>, a return type, an optional
883 <a href="#paramattrs">parameter attribute</a> for the return type, a function
884 name, a (possibly empty) argument list (each with optional
885 <a href="#paramattrs">parameter attributes</a>), optional
886 <a href="#fnattrs">function attributes</a>, an optional section, an optional
887 alignment, an optional <a href="#gc">garbage collector name</a>, an opening
888 curly brace, a list of basic blocks, and a closing curly brace.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]889
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]890<p>LLVM function declarations consist of the "<tt>declare</tt>" keyword, an
891 optional <a href="#linkage">linkage type</a>, an optional
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]892 <a href="#visibility">visibility style</a>, an optional
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]893 <a href="#callingconv">calling convention</a>, a return type, an optional
894 <a href="#paramattrs">parameter attribute</a> for the return type, a function
895 name, a possibly empty list of arguments, an optional alignment, and an
896 optional <a href="#gc">garbage collector name</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]897
Chris Lattner96451482008-08-05 18:29:16 +0000[diff] [blame]898<p>A function definition contains a list of basic blocks, forming the CFG
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]899 (Control Flow Graph) for the function. Each basic block may optionally start
900 with a label (giving the basic block a symbol table entry), contains a list
901 of instructions, and ends with a <a href="#terminators">terminator</a>
902 instruction (such as a branch or function return).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]903
904<p>The first basic block in a function is special in two ways: it is immediately
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]905 executed on entrance to the function, and it is not allowed to have
906 predecessor basic blocks (i.e. there can not be any branches to the entry
907 block of a function). Because the block can have no predecessors, it also
908 cannot have any <a href="#i_phi">PHI nodes</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]909
910<p>LLVM allows an explicit section to be specified for functions. If the target
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]911 supports it, it will emit functions to the section specified.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]912
913<p>An explicit alignment may be specified for a function. If not present, or if
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]914 the alignment is set to zero, the alignment of the function is set by the
915 target to whatever it feels convenient. If an explicit alignment is
916 specified, the function is forced to have at least that much alignment. All
917 alignments must be a power of 2.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]918
Bill Wendling6ec40612009-07-20 02:39:26 +0000[diff] [blame]919<h5>Syntax:</h5>
Devang Pateld0bfcc72008-10-07 17:48:33 +0000[diff] [blame]920<div class="doc_code">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]921<pre>
Chris Lattner1e5c5cd02008-10-13 16:55:18 +0000[diff] [blame]922define [<a href="#linkage">linkage</a>] [<a href="#visibility">visibility</a>]
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]923 [<a href="#callingconv">cconv</a>] [<a href="#paramattrs">ret attrs</a>]
924 &lt;ResultType&gt; @&lt;FunctionName&gt; ([argument list])
925 [<a href="#fnattrs">fn Attrs</a>] [section "name"] [align N]
926 [<a href="#gc">gc</a>] { ... }
927</pre>
Devang Pateld0bfcc72008-10-07 17:48:33 +0000[diff] [blame]928</div>
929
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]930</div>
931
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]932<!-- ======================================================================= -->
933<div class="doc_subsection">
934 <a name="aliasstructure">Aliases</a>
935</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]936
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]937<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]938
939<p>Aliases act as "second name" for the aliasee value (which can be either
940 function, global variable, another alias or bitcast of global value). Aliases
941 may have an optional <a href="#linkage">linkage type</a>, and an
942 optional <a href="#visibility">visibility style</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]943
Bill Wendling6ec40612009-07-20 02:39:26 +0000[diff] [blame]944<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]945<div class="doc_code">
946<pre>
Duncan Sandsd7bfabf2008-09-12 20:48:21 +0000[diff] [blame]947@&lt;Name&gt; = alias [Linkage] [Visibility] &lt;AliaseeTy&gt; @&lt;Aliasee&gt;
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]948</pre>
949</div>
950
951</div>
952
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]953<!-- ======================================================================= -->
Devang Patela4bb6792010-01-11 19:35:55 +0000[diff] [blame]954<div class="doc_subsection">
955 <a name="namedmetadatastructure">Named Metadata</a>
956</div>
957
958<div class="doc_text">
959
Chris Lattnerd0d96292010-01-15 21:50:19 +0000[diff] [blame]960<p>Named metadata is a collection of metadata. <a href="#metadata">Metadata
961 nodes</a> (but not metadata strings) and null are the only valid operands for
962 a named metadata.</p>
Devang Patela4bb6792010-01-11 19:35:55 +0000[diff] [blame]963
964<h5>Syntax:</h5>
965<div class="doc_code">
966<pre>
967!1 = metadata !{metadata !"one"}
968!name = !{null, !1}
969</pre>
970</div>
971
972</div>
973
974<!-- ======================================================================= -->
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]975<div class="doc_subsection"><a name="paramattrs">Parameter Attributes</a></div>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]976
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]977<div class="doc_text">
978
979<p>The return type and each parameter of a function type may have a set of
980 <i>parameter attributes</i> associated with them. Parameter attributes are
981 used to communicate additional information about the result or parameters of
982 a function. Parameter attributes are considered to be part of the function,
983 not of the function type, so functions with different parameter attributes
984 can have the same function type.</p>
985
986<p>Parameter attributes are simple keywords that follow the type specified. If
987 multiple parameter attributes are needed, they are space separated. For
988 example:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]989
990<div class="doc_code">
991<pre>
Nick Lewycky3022a742009-02-15 23:06:14 +0000[diff] [blame]992declare i32 @printf(i8* noalias nocapture, ...)
Chris Lattnerf33b8452008-10-04 18:33:34 +0000[diff] [blame]993declare i32 @atoi(i8 zeroext)
994declare signext i8 @returns_signed_char()
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]995</pre>
996</div>
997
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]998<p>Note that any attributes for the function result (<tt>nounwind</tt>,
999 <tt>readonly</tt>) come immediately after the argument list.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1000
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1001<p>Currently, only the following parameter attributes are defined:</p>
Chris Lattner275e6be2008-01-11 06:20:47 +0000[diff] [blame]1002
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1003<dl>
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]1004 <dt><tt><b>zeroext</b></tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1005 <dd>This indicates to the code generator that the parameter or return value
1006 should be zero-extended to a 32-bit value by the caller (for a parameter)
1007 or the callee (for a return value).</dd>
Chris Lattner275e6be2008-01-11 06:20:47 +0000[diff] [blame]1008
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]1009 <dt><tt><b>signext</b></tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1010 <dd>This indicates to the code generator that the parameter or return value
1011 should be sign-extended to a 32-bit value by the caller (for a parameter)
1012 or the callee (for a return value).</dd>
Chris Lattner275e6be2008-01-11 06:20:47 +0000[diff] [blame]1013
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]1014 <dt><tt><b>inreg</b></tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1015 <dd>This indicates that this parameter or return value should be treated in a
1016 special target-dependent fashion during while emitting code for a function
1017 call or return (usually, by putting it in a register as opposed to memory,
1018 though some targets use it to distinguish between two different kinds of
1019 registers). Use of this attribute is target-specific.</dd>
Chris Lattner275e6be2008-01-11 06:20:47 +0000[diff] [blame]1020
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]1021 <dt><tt><b><a name="byval">byval</a></b></tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1022 <dd>This indicates that the pointer parameter should really be passed by value
1023 to the function. The attribute implies that a hidden copy of the pointee
1024 is made between the caller and the callee, so the callee is unable to
1025 modify the value in the callee. This attribute is only valid on LLVM
1026 pointer arguments. It is generally used to pass structs and arrays by
1027 value, but is also valid on pointers to scalars. The copy is considered
1028 to belong to the caller not the callee (for example,
1029 <tt><a href="#readonly">readonly</a></tt> functions should not write to
1030 <tt>byval</tt> parameters). This is not a valid attribute for return
1031 values. The byval attribute also supports specifying an alignment with
1032 the align attribute. This has a target-specific effect on the code
1033 generator that usually indicates a desired alignment for the synthesized
1034 stack slot.</dd>
1035
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]1036 <dt><tt><b>sret</b></tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1037 <dd>This indicates that the pointer parameter specifies the address of a
1038 structure that is the return value of the function in the source program.
1039 This pointer must be guaranteed by the caller to be valid: loads and
1040 stores to the structure may be assumed by the callee to not to trap. This
1041 may only be applied to the first parameter. This is not a valid attribute
1042 for return values. </dd>
1043
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]1044 <dt><tt><b>noalias</b></tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1045 <dd>This indicates that the pointer does not alias any global or any other
1046 parameter. The caller is responsible for ensuring that this is the
1047 case. On a function return value, <tt>noalias</tt> additionally indicates
1048 that the pointer does not alias any other pointers visible to the
1049 caller. For further details, please see the discussion of the NoAlias
1050 response in
1051 <a href="http://llvm.org/docs/AliasAnalysis.html#MustMayNo">alias
1052 analysis</a>.</dd>
1053
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]1054 <dt><tt><b>nocapture</b></tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1055 <dd>This indicates that the callee does not make any copies of the pointer
1056 that outlive the callee itself. This is not a valid attribute for return
1057 values.</dd>
1058
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]1059 <dt><tt><b>nest</b></tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1060 <dd>This indicates that the pointer parameter can be excised using the
1061 <a href="#int_trampoline">trampoline intrinsics</a>. This is not a valid
1062 attribute for return values.</dd>
1063</dl>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1064
1065</div>
1066
1067<!-- ======================================================================= -->
1068<div class="doc_subsection">
Gordon Henriksen13fe5e32007-12-10 03:18:06 +0000[diff] [blame]1069 <a name="gc">Garbage Collector Names</a>
1070</div>
1071
1072<div class="doc_text">
Gordon Henriksen13fe5e32007-12-10 03:18:06 +0000[diff] [blame]1073
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1074<p>Each function may specify a garbage collector name, which is simply a
1075 string:</p>
1076
1077<div class="doc_code">
1078<pre>
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]1079define void @f() gc "name" { ... }
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1080</pre>
1081</div>
Gordon Henriksen13fe5e32007-12-10 03:18:06 +0000[diff] [blame]1082
1083<p>The compiler declares the supported values of <i>name</i>. Specifying a
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1084 collector which will cause the compiler to alter its output in order to
1085 support the named garbage collection algorithm.</p>
1086
Gordon Henriksen13fe5e32007-12-10 03:18:06 +0000[diff] [blame]1087</div>
1088
1089<!-- ======================================================================= -->
1090<div class="doc_subsection">
Devang Patel008cd3e2008-09-26 23:51:19 +0000[diff] [blame]1091 <a name="fnattrs">Function Attributes</a>
Devang Pateld468f1c2008-09-04 23:05:13 +0000[diff] [blame]1092</div>
1093
1094<div class="doc_text">
Devang Patel008cd3e2008-09-26 23:51:19 +0000[diff] [blame]1095
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1096<p>Function attributes are set to communicate additional information about a
1097 function. Function attributes are considered to be part of the function, not
1098 of the function type, so functions with different parameter attributes can
1099 have the same function type.</p>
Devang Patel008cd3e2008-09-26 23:51:19 +0000[diff] [blame]1100
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1101<p>Function attributes are simple keywords that follow the type specified. If
1102 multiple attributes are needed, they are space separated. For example:</p>
Devang Pateld468f1c2008-09-04 23:05:13 +0000[diff] [blame]1103
1104<div class="doc_code">
Bill Wendling74d3eac2008-09-07 10:26:33 +0000[diff] [blame]1105<pre>
Devang Patel008cd3e2008-09-26 23:51:19 +0000[diff] [blame]1106define void @f() noinline { ... }
1107define void @f() alwaysinline { ... }
1108define void @f() alwaysinline optsize { ... }
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]1109define void @f() optsize { ... }
Bill Wendling74d3eac2008-09-07 10:26:33 +0000[diff] [blame]1110</pre>
Devang Pateld468f1c2008-09-04 23:05:13 +0000[diff] [blame]1111</div>
1112
Bill Wendling74d3eac2008-09-07 10:26:33 +0000[diff] [blame]1113<dl>
Charles Davisfaa8f752010-02-12 00:31:15 +0000[diff] [blame]1114 <dt><tt><b>alignstack(&lt;<em>n</em>&gt;)</b></tt></dt>
1115 <dd>This attribute indicates that, when emitting the prologue and epilogue,
1116 the backend should forcibly align the stack pointer. Specify the
1117 desired alignment, which must be a power of two, in parentheses.
1118
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]1119 <dt><tt><b>alwaysinline</b></tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1120 <dd>This attribute indicates that the inliner should attempt to inline this
1121 function into callers whenever possible, ignoring any active inlining size
1122 threshold for this caller.</dd>
Bill Wendling74d3eac2008-09-07 10:26:33 +0000[diff] [blame]1123
Jakob Stoklund Olesen77180732010-02-06 01:16:28 +0000[diff] [blame]1124 <dt><tt><b>inlinehint</b></tt></dt>
1125 <dd>This attribute indicates that the source code contained a hint that inlining
1126 this function is desirable (such as the "inline" keyword in C/C++). It
1127 is just a hint; it imposes no requirements on the inliner.</dd>
1128
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]1129 <dt><tt><b>noinline</b></tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1130 <dd>This attribute indicates that the inliner should never inline this
1131 function in any situation. This attribute may not be used together with
1132 the <tt>alwaysinline</tt> attribute.</dd>
Devang Patel008cd3e2008-09-26 23:51:19 +0000[diff] [blame]1133
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]1134 <dt><tt><b>optsize</b></tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1135 <dd>This attribute suggests that optimization passes and code generator passes
1136 make choices that keep the code size of this function low, and otherwise
1137 do optimizations specifically to reduce code size.</dd>
Devang Patel008cd3e2008-09-26 23:51:19 +0000[diff] [blame]1138
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]1139 <dt><tt><b>noreturn</b></tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1140 <dd>This function attribute indicates that the function never returns
1141 normally. This produces undefined behavior at runtime if the function
1142 ever does dynamically return.</dd>
Bill Wendlingdfaabba2008-11-13 01:02:51 +0000[diff] [blame]1143
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]1144 <dt><tt><b>nounwind</b></tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1145 <dd>This function attribute indicates that the function never returns with an
1146 unwind or exceptional control flow. If the function does unwind, its
1147 runtime behavior is undefined.</dd>
Bill Wendlingbe9ec3f2008-11-26 19:07:40 +0000[diff] [blame]1148
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]1149 <dt><tt><b>readnone</b></tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1150 <dd>This attribute indicates that the function computes its result (or decides
1151 to unwind an exception) based strictly on its arguments, without
1152 dereferencing any pointer arguments or otherwise accessing any mutable
1153 state (e.g. memory, control registers, etc) visible to caller functions.
1154 It does not write through any pointer arguments
1155 (including <tt><a href="#byval">byval</a></tt> arguments) and never
1156 changes any state visible to callers. This means that it cannot unwind
1157 exceptions by calling the <tt>C++</tt> exception throwing methods, but
1158 could use the <tt>unwind</tt> instruction.</dd>
Devang Patela2f9f412009-06-12 19:45:19 +0000[diff] [blame]1159
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]1160 <dt><tt><b><a name="readonly">readonly</a></b></tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1161 <dd>This attribute indicates that the function does not write through any
1162 pointer arguments (including <tt><a href="#byval">byval</a></tt>
1163 arguments) or otherwise modify any state (e.g. memory, control registers,
1164 etc) visible to caller functions. It may dereference pointer arguments
1165 and read state that may be set in the caller. A readonly function always
1166 returns the same value (or unwinds an exception identically) when called
1167 with the same set of arguments and global state. It cannot unwind an
1168 exception by calling the <tt>C++</tt> exception throwing methods, but may
1169 use the <tt>unwind</tt> instruction.</dd>
Anton Korobeynikovedd7d112009-07-17 18:07:26 +0000[diff] [blame]1170
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]1171 <dt><tt><b><a name="ssp">ssp</a></b></tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1172 <dd>This attribute indicates that the function should emit a stack smashing
1173 protector. It is in the form of a "canary"&mdash;a random value placed on
1174 the stack before the local variables that's checked upon return from the
1175 function to see if it has been overwritten. A heuristic is used to
1176 determine if a function needs stack protectors or not.<br>
1177<br>
1178 If a function that has an <tt>ssp</tt> attribute is inlined into a
1179 function that doesn't have an <tt>ssp</tt> attribute, then the resulting
1180 function will have an <tt>ssp</tt> attribute.</dd>
1181
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]1182 <dt><tt><b>sspreq</b></tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1183 <dd>This attribute indicates that the function should <em>always</em> emit a
1184 stack smashing protector. This overrides
Bill Wendling6ec40612009-07-20 02:39:26 +0000[diff] [blame]1185 the <tt><a href="#ssp">ssp</a></tt> function attribute.<br>
1186<br>
1187 If a function that has an <tt>sspreq</tt> attribute is inlined into a
1188 function that doesn't have an <tt>sspreq</tt> attribute or which has
1189 an <tt>ssp</tt> attribute, then the resulting function will have
1190 an <tt>sspreq</tt> attribute.</dd>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1191
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]1192 <dt><tt><b>noredzone</b></tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1193 <dd>This attribute indicates that the code generator should not use a red
1194 zone, even if the target-specific ABI normally permits it.</dd>
1195
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]1196 <dt><tt><b>noimplicitfloat</b></tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1197 <dd>This attributes disables implicit floating point instructions.</dd>
1198
Bill Wendling614b32b2009-11-02 00:24:16 +0000[diff] [blame]1199 <dt><tt><b>naked</b></tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1200 <dd>This attribute disables prologue / epilogue emission for the function.
1201 This can have very system-specific consequences.</dd>
Bill Wendling74d3eac2008-09-07 10:26:33 +0000[diff] [blame]1202</dl>
1203
Devang Pateld468f1c2008-09-04 23:05:13 +0000[diff] [blame]1204</div>
1205
1206<!-- ======================================================================= -->
1207<div class="doc_subsection">
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1208 <a name="moduleasm">Module-Level Inline Assembly</a>
1209</div>
1210
1211<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1212
1213<p>Modules may contain "module-level inline asm" blocks, which corresponds to
1214 the GCC "file scope inline asm" blocks. These blocks are internally
1215 concatenated by LLVM and treated as a single unit, but may be separated in
1216 the <tt>.ll</tt> file if desired. The syntax is very simple:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1217
1218<div class="doc_code">
1219<pre>
1220module asm "inline asm code goes here"
1221module asm "more can go here"
1222</pre>
1223</div>
1224
1225<p>The strings can contain any character by escaping non-printable characters.
1226 The escape sequence used is simply "\xx" where "xx" is the two digit hex code
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1227 for the number.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1228
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1229<p>The inline asm code is simply printed to the machine code .s file when
1230 assembly code is generated.</p>
1231
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1232</div>
1233
1234<!-- ======================================================================= -->
1235<div class="doc_subsection">
1236 <a name="datalayout">Data Layout</a>
1237</div>
1238
1239<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1240
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1241<p>A module may specify a target specific data layout string that specifies how
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1242 data is to be laid out in memory. The syntax for the data layout is
1243 simply:</p>
1244
1245<div class="doc_code">
1246<pre>
1247target datalayout = "<i>layout specification</i>"
1248</pre>
1249</div>
1250
1251<p>The <i>layout specification</i> consists of a list of specifications
1252 separated by the minus sign character ('-'). Each specification starts with
1253 a letter and may include other information after the letter to define some
1254 aspect of the data layout. The specifications accepted are as follows:</p>
1255
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1256<dl>
1257 <dt><tt>E</tt></dt>
1258 <dd>Specifies that the target lays out data in big-endian form. That is, the
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1259 bits with the most significance have the lowest address location.</dd>
1260
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1261 <dt><tt>e</tt></dt>
Chris Lattner96451482008-08-05 18:29:16 +0000[diff] [blame]1262 <dd>Specifies that the target lays out data in little-endian form. That is,
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1263 the bits with the least significance have the lowest address
1264 location.</dd>
1265
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1266 <dt><tt>p:<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]1267 <dd>This specifies the <i>size</i> of a pointer and its <i>abi</i> and
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1268 <i>preferred</i> alignments. All sizes are in bits. Specifying
1269 the <i>pref</i> alignment is optional. If omitted, the
1270 preceding <tt>:</tt> should be omitted too.</dd>
1271
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1272 <dt><tt>i<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1273 <dd>This specifies the alignment for an integer type of a given bit
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1274 <i>size</i>. The value of <i>size</i> must be in the range [1,2^23).</dd>
1275
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1276 <dt><tt>v<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]1277 <dd>This specifies the alignment for a vector type of a given bit
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1278 <i>size</i>.</dd>
1279
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1280 <dt><tt>f<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]1281 <dd>This specifies the alignment for a floating point type of a given bit
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1282 <i>size</i>. The value of <i>size</i> must be either 32 (float) or 64
1283 (double).</dd>
1284
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1285 <dt><tt>a<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1286 <dd>This specifies the alignment for an aggregate type of a given bit
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1287 <i>size</i>.</dd>
1288
Daniel Dunbard88a97b2009-06-08 22:17:53 +0000[diff] [blame]1289 <dt><tt>s<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1290 <dd>This specifies the alignment for a stack object of a given bit
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1291 <i>size</i>.</dd>
Chris Lattnerece41802009-11-07 09:35:34 +0000[diff] [blame]1292
1293 <dt><tt>n<i>size1</i>:<i>size2</i>:<i>size3</i>...</tt></dt>
1294 <dd>This specifies a set of native integer widths for the target CPU
1295 in bits. For example, it might contain "n32" for 32-bit PowerPC,
1296 "n32:64" for PowerPC 64, or "n8:16:32:64" for X86-64. Elements of
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]1297 this set are considered to support most general arithmetic
Chris Lattnerece41802009-11-07 09:35:34 +0000[diff] [blame]1298 operations efficiently.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1299</dl>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1300
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1301<p>When constructing the data layout for a given target, LLVM starts with a
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1302 default set of specifications which are then (possibly) overriden by the
1303 specifications in the <tt>datalayout</tt> keyword. The default specifications
1304 are given in this list:</p>
1305
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1306<ul>
1307 <li><tt>E</tt> - big endian</li>
Dan Gohmane78194f2010-02-23 02:44:03 +0000[diff] [blame]1308 <li><tt>p:64:64:64</tt> - 64-bit pointers with 64-bit alignment</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1309 <li><tt>i1:8:8</tt> - i1 is 8-bit (byte) aligned</li>
1310 <li><tt>i8:8:8</tt> - i8 is 8-bit (byte) aligned</li>
1311 <li><tt>i16:16:16</tt> - i16 is 16-bit aligned</li>
1312 <li><tt>i32:32:32</tt> - i32 is 32-bit aligned</li>
Chris Lattner96451482008-08-05 18:29:16 +0000[diff] [blame]1313 <li><tt>i64:32:64</tt> - i64 has ABI alignment of 32-bits but preferred
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1314 alignment of 64-bits</li>
1315 <li><tt>f32:32:32</tt> - float is 32-bit aligned</li>
1316 <li><tt>f64:64:64</tt> - double is 64-bit aligned</li>
1317 <li><tt>v64:64:64</tt> - 64-bit vector is 64-bit aligned</li>
1318 <li><tt>v128:128:128</tt> - 128-bit vector is 128-bit aligned</li>
1319 <li><tt>a0:0:1</tt> - aggregates are 8-bit aligned</li>
Daniel Dunbard88a97b2009-06-08 22:17:53 +0000[diff] [blame]1320 <li><tt>s0:64:64</tt> - stack objects are 64-bit aligned</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1321</ul>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1322
1323<p>When LLVM is determining the alignment for a given type, it uses the
1324 following rules:</p>
1325
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1326<ol>
1327 <li>If the type sought is an exact match for one of the specifications, that
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1328 specification is used.</li>
1329
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1330 <li>If no match is found, and the type sought is an integer type, then the
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1331 smallest integer type that is larger than the bitwidth of the sought type
1332 is used. If none of the specifications are larger than the bitwidth then
1333 the the largest integer type is used. For example, given the default
1334 specifications above, the i7 type will use the alignment of i8 (next
1335 largest) while both i65 and i256 will use the alignment of i64 (largest
1336 specified).</li>
1337
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1338 <li>If no match is found, and the type sought is a vector type, then the
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1339 largest vector type that is smaller than the sought vector type will be
1340 used as a fall back. This happens because &lt;128 x double&gt; can be
1341 implemented in terms of 64 &lt;2 x double&gt;, for example.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1342</ol>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1343
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1344</div>
1345
Dan Gohman27b47012009-07-27 18:07:55 +0000[diff] [blame]1346<!-- ======================================================================= -->
1347<div class="doc_subsection">
1348 <a name="pointeraliasing">Pointer Aliasing Rules</a>
1349</div>
1350
1351<div class="doc_text">
1352
Andreas Bolka11fbf432009-07-29 00:02:05 +0000[diff] [blame]1353<p>Any memory access must be done through a pointer value associated
Andreas Bolka23bece42009-07-27 20:37:10 +0000[diff] [blame]1354with an address range of the memory access, otherwise the behavior
Dan Gohman27b47012009-07-27 18:07:55 +0000[diff] [blame]1355is undefined. Pointer values are associated with address ranges
1356according to the following rules:</p>
1357
1358<ul>
Andreas Bolka11fbf432009-07-29 00:02:05 +0000[diff] [blame]1359 <li>A pointer value formed from a
1360 <tt><a href="#i_getelementptr">getelementptr</a></tt> instruction
1361 is associated with the addresses associated with the first operand
1362 of the <tt>getelementptr</tt>.</li>
1363 <li>An address of a global variable is associated with the address
Dan Gohman27b47012009-07-27 18:07:55 +0000[diff] [blame]1364 range of the variable's storage.</li>
1365 <li>The result value of an allocation instruction is associated with
1366 the address range of the allocated storage.</li>
1367 <li>A null pointer in the default address-space is associated with
Andreas Bolka11fbf432009-07-29 00:02:05 +0000[diff] [blame]1368 no address.</li>
1369 <li>A pointer value formed by an
1370 <tt><a href="#i_inttoptr">inttoptr</a></tt> is associated with all
1371 address ranges of all pointer values that contribute (directly or
1372 indirectly) to the computation of the pointer's value.</li>
1373 <li>The result value of a
1374 <tt><a href="#i_bitcast">bitcast</a></tt> is associated with all
Dan Gohman27b47012009-07-27 18:07:55 +0000[diff] [blame]1375 addresses associated with the operand of the <tt>bitcast</tt>.</li>
1376 <li>An integer constant other than zero or a pointer value returned
1377 from a function not defined within LLVM may be associated with address
1378 ranges allocated through mechanisms other than those provided by
Andreas Bolka11fbf432009-07-29 00:02:05 +0000[diff] [blame]1379 LLVM. Such ranges shall not overlap with any ranges of addresses
Dan Gohman27b47012009-07-27 18:07:55 +0000[diff] [blame]1380 allocated by mechanisms provided by LLVM.</li>
1381 </ul>
1382
1383<p>LLVM IR does not associate types with memory. The result type of a
Andreas Bolka11fbf432009-07-29 00:02:05 +0000[diff] [blame]1384<tt><a href="#i_load">load</a></tt> merely indicates the size and
1385alignment of the memory from which to load, as well as the
1386interpretation of the value. The first operand of a
1387<tt><a href="#i_store">store</a></tt> similarly only indicates the size
1388and alignment of the store.</p>
Dan Gohman27b47012009-07-27 18:07:55 +0000[diff] [blame]1389
1390<p>Consequently, type-based alias analysis, aka TBAA, aka
1391<tt>-fstrict-aliasing</tt>, is not applicable to general unadorned
1392LLVM IR. <a href="#metadata">Metadata</a> may be used to encode
1393additional information which specialized optimization passes may use
1394to implement type-based alias analysis.</p>
1395
1396</div>
1397
Jeffrey Yasskin6dc467f2010-04-26 21:21:24 +0000[diff] [blame]1398<!-- ======================================================================= -->
1399<div class="doc_subsection">
1400 <a name="volatile">Volatile Memory Accesses</a>
1401</div>
1402
1403<div class="doc_text">
1404
1405<p>Certain memory accesses, such as <a href="#i_load"><tt>load</tt></a>s, <a
1406href="#i_store"><tt>store</tt></a>s, and <a
1407href="#int_memcpy"><tt>llvm.memcpy</tt></a>s may be marked <tt>volatile</tt>.
1408The optimizers must not change the number of volatile operations or change their
1409order of execution relative to other volatile operations. The optimizers
1410<i>may</i> change the order of volatile operations relative to non-volatile
1411operations. This is not Java's "volatile" and has no cross-thread
1412synchronization behavior.</p>
1413
1414</div>
1415
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1416<!-- *********************************************************************** -->
1417<div class="doc_section"> <a name="typesystem">Type System</a> </div>
1418<!-- *********************************************************************** -->
1419
1420<div class="doc_text">
1421
1422<p>The LLVM type system is one of the most important features of the
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1423 intermediate representation. Being typed enables a number of optimizations
1424 to be performed on the intermediate representation directly, without having
1425 to do extra analyses on the side before the transformation. A strong type
1426 system makes it easier to read the generated code and enables novel analyses
1427 and transformations that are not feasible to perform on normal three address
1428 code representations.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1429
1430</div>
1431
1432<!-- ======================================================================= -->
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1433<div class="doc_subsection"> <a name="t_classifications">Type
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1434Classifications</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1435
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1436<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1437
1438<p>The types fall into a few useful classifications:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1439
1440<table border="1" cellspacing="0" cellpadding="4">
1441 <tbody>
1442 <tr><th>Classification</th><th>Types</th></tr>
1443 <tr>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1444 <td><a href="#t_integer">integer</a></td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1445 <td><tt>i1, i2, i3, ... i8, ... i16, ... i32, ... i64, ... </tt></td>
1446 </tr>
1447 <tr>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1448 <td><a href="#t_floating">floating point</a></td>
1449 <td><tt>float, double, x86_fp80, fp128, ppc_fp128</tt></td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1450 </tr>
1451 <tr>
1452 <td><a name="t_firstclass">first class</a></td>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1453 <td><a href="#t_integer">integer</a>,
1454 <a href="#t_floating">floating point</a>,
1455 <a href="#t_pointer">pointer</a>,
Dan Gohmanf6237db2008-06-18 18:42:13 +0000[diff] [blame]1456 <a href="#t_vector">vector</a>,
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]1457 <a href="#t_struct">structure</a>,
Chris Lattnerd5d51722010-02-12 20:49:41 +0000[diff] [blame]1458 <a href="#t_union">union</a>,
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]1459 <a href="#t_array">array</a>,
Nick Lewycky29aaef82009-05-30 05:06:04 +0000[diff] [blame]1460 <a href="#t_label">label</a>,
1461 <a href="#t_metadata">metadata</a>.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1462 </td>
1463 </tr>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1464 <tr>
1465 <td><a href="#t_primitive">primitive</a></td>
1466 <td><a href="#t_label">label</a>,
1467 <a href="#t_void">void</a>,
Nick Lewycky29aaef82009-05-30 05:06:04 +0000[diff] [blame]1468 <a href="#t_floating">floating point</a>,
1469 <a href="#t_metadata">metadata</a>.</td>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1470 </tr>
1471 <tr>
1472 <td><a href="#t_derived">derived</a></td>
Chris Lattnerd5d51722010-02-12 20:49:41 +0000[diff] [blame]1473 <td><a href="#t_array">array</a>,
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1474 <a href="#t_function">function</a>,
1475 <a href="#t_pointer">pointer</a>,
1476 <a href="#t_struct">structure</a>,
1477 <a href="#t_pstruct">packed structure</a>,
Chris Lattnerd5d51722010-02-12 20:49:41 +0000[diff] [blame]1478 <a href="#t_union">union</a>,
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1479 <a href="#t_vector">vector</a>,
1480 <a href="#t_opaque">opaque</a>.
Dan Gohman032ba852008-10-14 16:32:04 +0000[diff] [blame]1481 </td>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1482 </tr>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1483 </tbody>
1484</table>
1485
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1486<p>The <a href="#t_firstclass">first class</a> types are perhaps the most
1487 important. Values of these types are the only ones which can be produced by
Nick Lewycky244cf482009-09-27 00:45:11 +0000[diff] [blame]1488 instructions.</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1489
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1490</div>
1491
1492<!-- ======================================================================= -->
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1493<div class="doc_subsection"> <a name="t_primitive">Primitive Types</a> </div>
Chris Lattner86437612008-01-04 04:34:14 +0000[diff] [blame]1494
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1495<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1496
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1497<p>The primitive types are the fundamental building blocks of the LLVM
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1498 system.</p>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1499
Chris Lattner86437612008-01-04 04:34:14 +0000[diff] [blame]1500</div>
1501
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1502<!-- _______________________________________________________________________ -->
Nick Lewycky244cf482009-09-27 00:45:11 +0000[diff] [blame]1503<div class="doc_subsubsection"> <a name="t_integer">Integer Type</a> </div>
1504
1505<div class="doc_text">
1506
1507<h5>Overview:</h5>
1508<p>The integer type is a very simple type that simply specifies an arbitrary
1509 bit width for the integer type desired. Any bit width from 1 bit to
1510 2<sup>23</sup>-1 (about 8 million) can be specified.</p>
1511
1512<h5>Syntax:</h5>
1513<pre>
1514 iN
1515</pre>
1516
1517<p>The number of bits the integer will occupy is specified by the <tt>N</tt>
1518 value.</p>
1519
1520<h5>Examples:</h5>
1521<table class="layout">
1522 <tr class="layout">
1523 <td class="left"><tt>i1</tt></td>
1524 <td class="left">a single-bit integer.</td>
1525 </tr>
1526 <tr class="layout">
1527 <td class="left"><tt>i32</tt></td>
1528 <td class="left">a 32-bit integer.</td>
1529 </tr>
1530 <tr class="layout">
1531 <td class="left"><tt>i1942652</tt></td>
1532 <td class="left">a really big integer of over 1 million bits.</td>
1533 </tr>
1534</table>
1535
Nick Lewycky244cf482009-09-27 00:45:11 +0000[diff] [blame]1536</div>
1537
1538<!-- _______________________________________________________________________ -->
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1539<div class="doc_subsubsection"> <a name="t_floating">Floating Point Types</a> </div>
1540
1541<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1542
1543<table>
1544 <tbody>
1545 <tr><th>Type</th><th>Description</th></tr>
1546 <tr><td><tt>float</tt></td><td>32-bit floating point value</td></tr>
1547 <tr><td><tt>double</tt></td><td>64-bit floating point value</td></tr>
1548 <tr><td><tt>fp128</tt></td><td>128-bit floating point value (112-bit mantissa)</td></tr>
1549 <tr><td><tt>x86_fp80</tt></td><td>80-bit floating point value (X87)</td></tr>
1550 <tr><td><tt>ppc_fp128</tt></td><td>128-bit floating point value (two 64-bits)</td></tr>
1551 </tbody>
1552</table>
1553
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1554</div>
1555
1556<!-- _______________________________________________________________________ -->
1557<div class="doc_subsubsection"> <a name="t_void">Void Type</a> </div>
1558
1559<div class="doc_text">
Bill Wendling6ec40612009-07-20 02:39:26 +0000[diff] [blame]1560
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1561<h5>Overview:</h5>
1562<p>The void type does not represent any value and has no size.</p>
1563
1564<h5>Syntax:</h5>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1565<pre>
1566 void
1567</pre>
Bill Wendling6ec40612009-07-20 02:39:26 +0000[diff] [blame]1568
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1569</div>
1570
1571<!-- _______________________________________________________________________ -->
1572<div class="doc_subsubsection"> <a name="t_label">Label Type</a> </div>
1573
1574<div class="doc_text">
Bill Wendling6ec40612009-07-20 02:39:26 +0000[diff] [blame]1575
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1576<h5>Overview:</h5>
1577<p>The label type represents code labels.</p>
1578
1579<h5>Syntax:</h5>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1580<pre>
1581 label
1582</pre>
Bill Wendling6ec40612009-07-20 02:39:26 +0000[diff] [blame]1583
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1584</div>
1585
Nick Lewycky29aaef82009-05-30 05:06:04 +0000[diff] [blame]1586<!-- _______________________________________________________________________ -->
1587<div class="doc_subsubsection"> <a name="t_metadata">Metadata Type</a> </div>
1588
1589<div class="doc_text">
Bill Wendling6ec40612009-07-20 02:39:26 +0000[diff] [blame]1590
Nick Lewycky29aaef82009-05-30 05:06:04 +0000[diff] [blame]1591<h5>Overview:</h5>
Nick Lewyckyeb021882009-09-27 23:27:42 +0000[diff] [blame]1592<p>The metadata type represents embedded metadata. No derived types may be
1593 created from metadata except for <a href="#t_function">function</a>
1594 arguments.
Nick Lewycky29aaef82009-05-30 05:06:04 +0000[diff] [blame]1595
1596<h5>Syntax:</h5>
Nick Lewycky29aaef82009-05-30 05:06:04 +0000[diff] [blame]1597<pre>
1598 metadata
1599</pre>
Bill Wendling6ec40612009-07-20 02:39:26 +0000[diff] [blame]1600
Nick Lewycky29aaef82009-05-30 05:06:04 +0000[diff] [blame]1601</div>
1602
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1603
1604<!-- ======================================================================= -->
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1605<div class="doc_subsection"> <a name="t_derived">Derived Types</a> </div>
1606
1607<div class="doc_text">
1608
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1609<p>The real power in LLVM comes from the derived types in the system. This is
1610 what allows a programmer to represent arrays, functions, pointers, and other
Nick Lewycky244cf482009-09-27 00:45:11 +0000[diff] [blame]1611 useful types. Each of these types contain one or more element types which
1612 may be a primitive type, or another derived type. For example, it is
1613 possible to have a two dimensional array, using an array as the element type
1614 of another array.</p>
djge93155c2009-01-24 15:58:40 +0000[diff] [blame]1615
Chris Lattnerd5d51722010-02-12 20:49:41 +0000[diff] [blame]1616
1617</div>
1618
1619<!-- _______________________________________________________________________ -->
1620<div class="doc_subsubsection"> <a name="t_aggregate">Aggregate Types</a> </div>
1621
1622<div class="doc_text">
1623
1624<p>Aggregate Types are a subset of derived types that can contain multiple
1625 member types. <a href="#t_array">Arrays</a>,
1626 <a href="#t_struct">structs</a>, <a href="#t_vector">vectors</a> and
1627 <a href="#t_union">unions</a> are aggregate types.</p>
1628
1629</div>
1630
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1631</div>
1632
1633<!-- _______________________________________________________________________ -->
1634<div class="doc_subsubsection"> <a name="t_array">Array Type</a> </div>
1635
1636<div class="doc_text">
1637
1638<h5>Overview:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1639<p>The array type is a very simple derived type that arranges elements
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1640 sequentially in memory. The array type requires a size (number of elements)
1641 and an underlying data type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1642
1643<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1644<pre>
1645 [&lt;# elements&gt; x &lt;elementtype&gt;]
1646</pre>
1647
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1648<p>The number of elements is a constant integer value; <tt>elementtype</tt> may
1649 be any type with a size.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1650
1651<h5>Examples:</h5>
1652<table class="layout">
1653 <tr class="layout">
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1654 <td class="left"><tt>[40 x i32]</tt></td>
1655 <td class="left">Array of 40 32-bit integer values.</td>
1656 </tr>
1657 <tr class="layout">
1658 <td class="left"><tt>[41 x i32]</tt></td>
1659 <td class="left">Array of 41 32-bit integer values.</td>
1660 </tr>
1661 <tr class="layout">
1662 <td class="left"><tt>[4 x i8]</tt></td>
1663 <td class="left">Array of 4 8-bit integer values.</td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1664 </tr>
1665</table>
1666<p>Here are some examples of multidimensional arrays:</p>
1667<table class="layout">
1668 <tr class="layout">
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1669 <td class="left"><tt>[3 x [4 x i32]]</tt></td>
1670 <td class="left">3x4 array of 32-bit integer values.</td>
1671 </tr>
1672 <tr class="layout">
1673 <td class="left"><tt>[12 x [10 x float]]</tt></td>
1674 <td class="left">12x10 array of single precision floating point values.</td>
1675 </tr>
1676 <tr class="layout">
1677 <td class="left"><tt>[2 x [3 x [4 x i16]]]</tt></td>
1678 <td class="left">2x3x4 array of 16-bit integer values.</td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1679 </tr>
1680</table>
1681
Dan Gohman87ec5ad2009-11-09 19:01:53 +0000[diff] [blame]1682<p>There is no restriction on indexing beyond the end of the array implied by
1683 a static type (though there are restrictions on indexing beyond the bounds
1684 of an allocated object in some cases). This means that single-dimension
1685 'variable sized array' addressing can be implemented in LLVM with a zero
1686 length array type. An implementation of 'pascal style arrays' in LLVM could
1687 use the type "<tt>{ i32, [0 x float]}</tt>", for example.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1688
1689</div>
1690
1691<!-- _______________________________________________________________________ -->
1692<div class="doc_subsubsection"> <a name="t_function">Function Type</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1693
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1694<div class="doc_text">
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]1695
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1696<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1697<p>The function type can be thought of as a function signature. It consists of
1698 a return type and a list of formal parameter types. The return type of a
Chris Lattnerd5d51722010-02-12 20:49:41 +0000[diff] [blame]1699 function type is a scalar type, a void type, a struct type, or a union
1700 type. If the return type is a struct type then all struct elements must be
1701 of first class types, and the struct must have at least one element.</p>
Devang Patela3cc5372008-03-10 20:49:15 +0000[diff] [blame]1702
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1703<h5>Syntax:</h5>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]1704<pre>
Nick Lewycky3b3b54d2009-09-27 07:55:32 +0000[diff] [blame]1705 &lt;returntype&gt; (&lt;parameter list&gt;)
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]1706</pre>
1707
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1708<p>...where '<tt>&lt;parameter list&gt;</tt>' is a comma-separated list of type
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1709 specifiers. Optionally, the parameter list may include a type <tt>...</tt>,
1710 which indicates that the function takes a variable number of arguments.
1711 Variable argument functions can access their arguments with
1712 the <a href="#int_varargs">variable argument handling intrinsic</a>
Chris Lattner553fb1e2010-03-02 06:36:51 +0000[diff] [blame]1713 functions. '<tt>&lt;returntype&gt;</tt>' is any type except
Nick Lewyckyeb021882009-09-27 23:27:42 +0000[diff] [blame]1714 <a href="#t_label">label</a>.</p>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]1715
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1716<h5>Examples:</h5>
1717<table class="layout">
1718 <tr class="layout">
1719 <td class="left"><tt>i32 (i32)</tt></td>
1720 <td class="left">function taking an <tt>i32</tt>, returning an <tt>i32</tt>
1721 </td>
1722 </tr><tr class="layout">
Chris Lattner553fb1e2010-03-02 06:36:51 +0000[diff] [blame]1723 <td class="left"><tt>float&nbsp;(i16,&nbsp;i32&nbsp;*)&nbsp;*
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1724 </tt></td>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]1725 <td class="left"><a href="#t_pointer">Pointer</a> to a function that takes
Chris Lattner553fb1e2010-03-02 06:36:51 +0000[diff] [blame]1726 an <tt>i16</tt> and a <a href="#t_pointer">pointer</a> to <tt>i32</tt>,
1727 returning <tt>float</tt>.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1728 </td>
1729 </tr><tr class="layout">
1730 <td class="left"><tt>i32 (i8*, ...)</tt></td>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]1731 <td class="left">A vararg function that takes at least one
1732 <a href="#t_pointer">pointer</a> to <tt>i8 </tt> (char in C),
1733 which returns an integer. This is the signature for <tt>printf</tt> in
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1734 LLVM.
1735 </td>
Devang Pateld4ba41d2008-03-24 05:35:41 +0000[diff] [blame]1736 </tr><tr class="layout">
1737 <td class="left"><tt>{i32, i32} (i32)</tt></td>
Nick Lewycky3b3b54d2009-09-27 07:55:32 +0000[diff] [blame]1738 <td class="left">A function taking an <tt>i32</tt>, returning a
1739 <a href="#t_struct">structure</a> containing two <tt>i32</tt> values
Devang Pateld4ba41d2008-03-24 05:35:41 +0000[diff] [blame]1740 </td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1741 </tr>
1742</table>
1743
1744</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1745
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1746<!-- _______________________________________________________________________ -->
1747<div class="doc_subsubsection"> <a name="t_struct">Structure Type</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1748
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1749<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1750
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1751<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1752<p>The structure type is used to represent a collection of data members together
1753 in memory. The packing of the field types is defined to match the ABI of the
1754 underlying processor. The elements of a structure may be any type that has a
1755 size.</p>
1756
Jeffrey Yasskine05d7732010-01-11 19:19:26 +0000[diff] [blame]1757<p>Structures in memory are accessed using '<tt><a href="#i_load">load</a></tt>'
1758 and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a field
1759 with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.
1760 Structures in registers are accessed using the
1761 '<tt><a href="#i_extractvalue">extractvalue</a></tt>' and
1762 '<tt><a href="#i_insertvalue">insertvalue</a></tt>' instructions.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1763<h5>Syntax:</h5>
Bill Wendling6ec40612009-07-20 02:39:26 +0000[diff] [blame]1764<pre>
1765 { &lt;type list&gt; }
1766</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1767
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1768<h5>Examples:</h5>
1769<table class="layout">
1770 <tr class="layout">
1771 <td class="left"><tt>{ i32, i32, i32 }</tt></td>
1772 <td class="left">A triple of three <tt>i32</tt> values</td>
1773 </tr><tr class="layout">
1774 <td class="left"><tt>{&nbsp;float,&nbsp;i32&nbsp;(i32)&nbsp;*&nbsp;}</tt></td>
1775 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1776 second element is a <a href="#t_pointer">pointer</a> to a
1777 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1778 an <tt>i32</tt>.</td>
1779 </tr>
1780</table>
djge93155c2009-01-24 15:58:40 +0000[diff] [blame]1781
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1782</div>
1783
1784<!-- _______________________________________________________________________ -->
1785<div class="doc_subsubsection"> <a name="t_pstruct">Packed Structure Type</a>
1786</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1787
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1788<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1789
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1790<h5>Overview:</h5>
1791<p>The packed structure type is used to represent a collection of data members
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1792 together in memory. There is no padding between fields. Further, the
1793 alignment of a packed structure is 1 byte. The elements of a packed
1794 structure may be any type that has a size.</p>
1795
1796<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt> and
1797 '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a field with
1798 the '<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.</p>
1799
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1800<h5>Syntax:</h5>
Bill Wendling6ec40612009-07-20 02:39:26 +0000[diff] [blame]1801<pre>
1802 &lt; { &lt;type list&gt; } &gt;
1803</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1804
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1805<h5>Examples:</h5>
1806<table class="layout">
1807 <tr class="layout">
1808 <td class="left"><tt>&lt; { i32, i32, i32 } &gt;</tt></td>
1809 <td class="left">A triple of three <tt>i32</tt> values</td>
1810 </tr><tr class="layout">
Bill Wendling74d3eac2008-09-07 10:26:33 +0000[diff] [blame]1811 <td class="left">
1812<tt>&lt;&nbsp;{&nbsp;float,&nbsp;i32&nbsp;(i32)*&nbsp;}&nbsp;&gt;</tt></td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1813 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1814 second element is a <a href="#t_pointer">pointer</a> to a
1815 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1816 an <tt>i32</tt>.</td>
1817 </tr>
1818</table>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1819
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1820</div>
1821
1822<!-- _______________________________________________________________________ -->
Chris Lattnerd5d51722010-02-12 20:49:41 +0000[diff] [blame]1823<div class="doc_subsubsection"> <a name="t_union">Union Type</a> </div>
1824
1825<div class="doc_text">
1826
1827<h5>Overview:</h5>
1828<p>A union type describes an object with size and alignment suitable for
1829 an object of any one of a given set of types (also known as an "untagged"
1830 union). It is similar in concept and usage to a
1831 <a href="#t_struct">struct</a>, except that all members of the union
1832 have an offset of zero. The elements of a union may be any type that has a
1833 size. Unions must have at least one member - empty unions are not allowed.
1834 </p>
1835
1836<p>The size of the union as a whole will be the size of its largest member,
1837 and the alignment requirements of the union as a whole will be the largest
1838 alignment requirement of any member.</p>
1839
Dan Gohmanef8400c2010-02-25 16:51:31 +0000[diff] [blame]1840<p>Union members are accessed using '<tt><a href="#i_load">load</a></tt> and
Chris Lattnerd5d51722010-02-12 20:49:41 +0000[diff] [blame]1841 '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a field with
1842 the '<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.
1843 Since all members are at offset zero, the getelementptr instruction does
1844 not affect the address, only the type of the resulting pointer.</p>
1845
1846<h5>Syntax:</h5>
1847<pre>
1848 union { &lt;type list&gt; }
1849</pre>
1850
1851<h5>Examples:</h5>
1852<table class="layout">
1853 <tr class="layout">
1854 <td class="left"><tt>union { i32, i32*, float }</tt></td>
1855 <td class="left">A union of three types: an <tt>i32</tt>, a pointer to
1856 an <tt>i32</tt>, and a <tt>float</tt>.</td>
1857 </tr><tr class="layout">
1858 <td class="left">
1859 <tt>union {&nbsp;float,&nbsp;i32&nbsp;(i32)&nbsp;*&nbsp;}</tt></td>
1860 <td class="left">A union, where the first element is a <tt>float</tt> and the
1861 second element is a <a href="#t_pointer">pointer</a> to a
1862 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1863 an <tt>i32</tt>.</td>
1864 </tr>
1865</table>
1866
1867</div>
1868
1869<!-- _______________________________________________________________________ -->
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1870<div class="doc_subsubsection"> <a name="t_pointer">Pointer Type</a> </div>
Chris Lattner96edbd32009-02-08 19:53:29 +0000[diff] [blame]1871
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1872<div class="doc_text">
1873
1874<h5>Overview:</h5>
Dan Gohmanb2f72c82010-02-25 16:50:07 +0000[diff] [blame]1875<p>The pointer type is used to specify memory locations.
1876 Pointers are commonly used to reference objects in memory.</p>
1877
1878<p>Pointer types may have an optional address space attribute defining the
1879 numbered address space where the pointed-to object resides. The default
1880 address space is number zero. The semantics of non-zero address
1881 spaces are target-specific.</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1882
1883<p>Note that LLVM does not permit pointers to void (<tt>void*</tt>) nor does it
1884 permit pointers to labels (<tt>label*</tt>). Use <tt>i8*</tt> instead.</p>
Chris Lattner96edbd32009-02-08 19:53:29 +0000[diff] [blame]1885
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1886<h5>Syntax:</h5>
Bill Wendling6ec40612009-07-20 02:39:26 +0000[diff] [blame]1887<pre>
1888 &lt;type&gt; *
1889</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1890
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1891<h5>Examples:</h5>
1892<table class="layout">
1893 <tr class="layout">
Dan Gohman01852382009-01-04 23:44:43 +0000[diff] [blame]1894 <td class="left"><tt>[4 x i32]*</tt></td>
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1895 <td class="left">A <a href="#t_pointer">pointer</a> to <a
1896 href="#t_array">array</a> of four <tt>i32</tt> values.</td>
1897 </tr>
1898 <tr class="layout">
1899 <td class="left"><tt>i32 (i32 *) *</tt></td>
1900 <td class="left"> A <a href="#t_pointer">pointer</a> to a <a
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1901 href="#t_function">function</a> that takes an <tt>i32*</tt>, returning an
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1902 <tt>i32</tt>.</td>
1903 </tr>
1904 <tr class="layout">
1905 <td class="left"><tt>i32 addrspace(5)*</tt></td>
1906 <td class="left">A <a href="#t_pointer">pointer</a> to an <tt>i32</tt> value
1907 that resides in address space #5.</td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1908 </tr>
1909</table>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1910
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1911</div>
1912
1913<!-- _______________________________________________________________________ -->
1914<div class="doc_subsubsection"> <a name="t_vector">Vector Type</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1915
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1916<div class="doc_text">
1917
1918<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1919<p>A vector type is a simple derived type that represents a vector of elements.
1920 Vector types are used when multiple primitive data are operated in parallel
1921 using a single instruction (SIMD). A vector type requires a size (number of
Duncan Sands87598b02009-11-27 13:38:03 +0000[diff] [blame]1922 elements) and an underlying primitive data type. Vector types are considered
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1923 <a href="#t_firstclass">first class</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1924
1925<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1926<pre>
1927 &lt; &lt;# elements&gt; x &lt;elementtype&gt; &gt;
1928</pre>
1929
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1930<p>The number of elements is a constant integer value; elementtype may be any
1931 integer or floating point type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1932
1933<h5>Examples:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1934<table class="layout">
1935 <tr class="layout">
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1936 <td class="left"><tt>&lt;4 x i32&gt;</tt></td>
1937 <td class="left">Vector of 4 32-bit integer values.</td>
1938 </tr>
1939 <tr class="layout">
1940 <td class="left"><tt>&lt;8 x float&gt;</tt></td>
1941 <td class="left">Vector of 8 32-bit floating-point values.</td>
1942 </tr>
1943 <tr class="layout">
1944 <td class="left"><tt>&lt;2 x i64&gt;</tt></td>
1945 <td class="left">Vector of 2 64-bit integer values.</td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1946 </tr>
1947</table>
djge93155c2009-01-24 15:58:40 +0000[diff] [blame]1948
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1949</div>
1950
1951<!-- _______________________________________________________________________ -->
1952<div class="doc_subsubsection"> <a name="t_opaque">Opaque Type</a> </div>
1953<div class="doc_text">
1954
1955<h5>Overview:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1956<p>Opaque types are used to represent unknown types in the system. This
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1957 corresponds (for example) to the C notion of a forward declared structure
1958 type. In LLVM, opaque types can eventually be resolved to any type (not just
1959 a structure type).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1960
1961<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1962<pre>
1963 opaque
1964</pre>
1965
1966<h5>Examples:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1967<table class="layout">
1968 <tr class="layout">
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1969 <td class="left"><tt>opaque</tt></td>
1970 <td class="left">An opaque type.</td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1971 </tr>
1972</table>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1973
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1974</div>
1975
Chris Lattner515195a2009-02-02 07:32:36 +0000[diff] [blame]1976<!-- ======================================================================= -->
1977<div class="doc_subsection">
1978 <a name="t_uprefs">Type Up-references</a>
1979</div>
1980
1981<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1982
Chris Lattner515195a2009-02-02 07:32:36 +0000[diff] [blame]1983<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1984<p>An "up reference" allows you to refer to a lexically enclosing type without
1985 requiring it to have a name. For instance, a structure declaration may
1986 contain a pointer to any of the types it is lexically a member of. Example
1987 of up references (with their equivalent as named type declarations)
1988 include:</p>
Chris Lattner515195a2009-02-02 07:32:36 +0000[diff] [blame]1989
1990<pre>
Chris Lattner5ad632d2009-02-09 10:00:56 +0000[diff] [blame]1991 { \2 * } %x = type { %x* }
Chris Lattner515195a2009-02-02 07:32:36 +0000[diff] [blame]1992 { \2 }* %y = type { %y }*
1993 \1* %z = type %z*
1994</pre>
1995
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1996<p>An up reference is needed by the asmprinter for printing out cyclic types
1997 when there is no declared name for a type in the cycle. Because the
1998 asmprinter does not want to print out an infinite type string, it needs a
1999 syntax to handle recursive types that have no names (all names are optional
2000 in llvm IR).</p>
Chris Lattner515195a2009-02-02 07:32:36 +0000[diff] [blame]2001
2002<h5>Syntax:</h5>
2003<pre>
2004 \&lt;level&gt;
2005</pre>
2006
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2007<p>The level is the count of the lexical type that is being referred to.</p>
Chris Lattner515195a2009-02-02 07:32:36 +0000[diff] [blame]2008
2009<h5>Examples:</h5>
Chris Lattner515195a2009-02-02 07:32:36 +0000[diff] [blame]2010<table class="layout">
2011 <tr class="layout">
2012 <td class="left"><tt>\1*</tt></td>
2013 <td class="left">Self-referential pointer.</td>
2014 </tr>
2015 <tr class="layout">
2016 <td class="left"><tt>{ { \3*, i8 }, i32 }</tt></td>
2017 <td class="left">Recursive structure where the upref refers to the out-most
2018 structure.</td>
2019 </tr>
2020</table>
Chris Lattner515195a2009-02-02 07:32:36 +0000[diff] [blame]2021
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2022</div>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2023
2024<!-- *********************************************************************** -->
2025<div class="doc_section"> <a name="constants">Constants</a> </div>
2026<!-- *********************************************************************** -->
2027
2028<div class="doc_text">
2029
2030<p>LLVM has several different basic types of constants. This section describes
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2031 them all and their syntax.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2032
2033</div>
2034
2035<!-- ======================================================================= -->
2036<div class="doc_subsection"><a name="simpleconstants">Simple Constants</a></div>
2037
2038<div class="doc_text">
2039
2040<dl>
2041 <dt><b>Boolean constants</b></dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2042 <dd>The two strings '<tt>true</tt>' and '<tt>false</tt>' are both valid
Nick Lewycky244cf482009-09-27 00:45:11 +0000[diff] [blame]2043 constants of the <tt><a href="#t_integer">i1</a></tt> type.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2044
2045 <dt><b>Integer constants</b></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2046 <dd>Standard integers (such as '4') are constants of
2047 the <a href="#t_integer">integer</a> type. Negative numbers may be used
2048 with integer types.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2049
2050 <dt><b>Floating point constants</b></dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2051 <dd>Floating point constants use standard decimal notation (e.g. 123.421),
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2052 exponential notation (e.g. 1.23421e+2), or a more precise hexadecimal
2053 notation (see below). The assembler requires the exact decimal value of a
2054 floating-point constant. For example, the assembler accepts 1.25 but
2055 rejects 1.3 because 1.3 is a repeating decimal in binary. Floating point
2056 constants must have a <a href="#t_floating">floating point</a> type. </dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2057
2058 <dt><b>Null pointer constants</b></dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2059 <dd>The identifier '<tt>null</tt>' is recognized as a null pointer constant
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2060 and must be of <a href="#t_pointer">pointer type</a>.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2061</dl>
2062
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2063<p>The one non-intuitive notation for constants is the hexadecimal form of
2064 floating point constants. For example, the form '<tt>double
2065 0x432ff973cafa8000</tt>' is equivalent to (but harder to read than)
2066 '<tt>double 4.5e+15</tt>'. The only time hexadecimal floating point
2067 constants are required (and the only time that they are generated by the
2068 disassembler) is when a floating point constant must be emitted but it cannot
2069 be represented as a decimal floating point number in a reasonable number of
2070 digits. For example, NaN's, infinities, and other special values are
2071 represented in their IEEE hexadecimal format so that assembly and disassembly
2072 do not cause any bits to change in the constants.</p>
2073
Dale Johannesenf82a52f2009-02-11 22:14:51 +0000[diff] [blame]2074<p>When using the hexadecimal form, constants of types float and double are
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2075 represented using the 16-digit form shown above (which matches the IEEE754
2076 representation for double); float values must, however, be exactly
2077 representable as IEE754 single precision. Hexadecimal format is always used
2078 for long double, and there are three forms of long double. The 80-bit format
2079 used by x86 is represented as <tt>0xK</tt> followed by 20 hexadecimal digits.
2080 The 128-bit format used by PowerPC (two adjacent doubles) is represented
2081 by <tt>0xM</tt> followed by 32 hexadecimal digits. The IEEE 128-bit format
2082 is represented by <tt>0xL</tt> followed by 32 hexadecimal digits; no
2083 currently supported target uses this format. Long doubles will only work if
2084 they match the long double format on your target. All hexadecimal formats
2085 are big-endian (sign bit at the left).</p>
2086
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2087</div>
2088
2089<!-- ======================================================================= -->
Chris Lattner97063852009-02-28 18:32:25 +0000[diff] [blame]2090<div class="doc_subsection">
Bill Wendling1a2630a2009-07-20 02:32:41 +0000[diff] [blame]2091<a name="aggregateconstants"></a> <!-- old anchor -->
2092<a name="complexconstants">Complex Constants</a>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2093</div>
2094
2095<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2096
Chris Lattner97063852009-02-28 18:32:25 +0000[diff] [blame]2097<p>Complex constants are a (potentially recursive) combination of simple
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2098 constants and smaller complex constants.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2099
2100<dl>
2101 <dt><b>Structure constants</b></dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2102 <dd>Structure constants are represented with notation similar to structure
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2103 type definitions (a comma separated list of elements, surrounded by braces
2104 (<tt>{}</tt>)). For example: "<tt>{ i32 4, float 17.0, i32* @G }</tt>",
2105 where "<tt>@G</tt>" is declared as "<tt>@G = external global i32</tt>".
2106 Structure constants must have <a href="#t_struct">structure type</a>, and
2107 the number and types of elements must match those specified by the
2108 type.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2109
Chris Lattnerd5d51722010-02-12 20:49:41 +0000[diff] [blame]2110 <dt><b>Union constants</b></dt>
2111 <dd>Union constants are represented with notation similar to a structure with
2112 a single element - that is, a single typed element surrounded
2113 by braces (<tt>{}</tt>)). For example: "<tt>{ i32 4 }</tt>". The
2114 <a href="#t_union">union type</a> can be initialized with a single-element
2115 struct as long as the type of the struct element matches the type of
2116 one of the union members.</dd>
2117
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2118 <dt><b>Array constants</b></dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2119 <dd>Array constants are represented with notation similar to array type
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2120 definitions (a comma separated list of elements, surrounded by square
2121 brackets (<tt>[]</tt>)). For example: "<tt>[ i32 42, i32 11, i32 74
2122 ]</tt>". Array constants must have <a href="#t_array">array type</a>, and
2123 the number and types of elements must match those specified by the
2124 type.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2125
2126 <dt><b>Vector constants</b></dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2127 <dd>Vector constants are represented with notation similar to vector type
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2128 definitions (a comma separated list of elements, surrounded by
2129 less-than/greater-than's (<tt>&lt;&gt;</tt>)). For example: "<tt>&lt; i32
2130 42, i32 11, i32 74, i32 100 &gt;</tt>". Vector constants must
2131 have <a href="#t_vector">vector type</a>, and the number and types of
2132 elements must match those specified by the type.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2133
2134 <dt><b>Zero initialization</b></dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2135 <dd>The string '<tt>zeroinitializer</tt>' can be used to zero initialize a
Chris Lattnerd5d51722010-02-12 20:49:41 +0000[diff] [blame]2136 value to zero of <em>any</em> type, including scalar and
2137 <a href="#t_aggregate">aggregate</a> types.
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2138 This is often used to avoid having to print large zero initializers
2139 (e.g. for large arrays) and is always exactly equivalent to using explicit
2140 zero initializers.</dd>
Nick Lewycky4dcf8102009-04-04 07:22:01 +0000[diff] [blame]2141
2142 <dt><b>Metadata node</b></dt>
Nick Lewyckyf122c7e2009-05-30 16:08:30 +0000[diff] [blame]2143 <dd>A metadata node is a structure-like constant with
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2144 <a href="#t_metadata">metadata type</a>. For example: "<tt>metadata !{
2145 i32 0, metadata !"test" }</tt>". Unlike other constants that are meant to
2146 be interpreted as part of the instruction stream, metadata is a place to
2147 attach additional information such as debug info.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2148</dl>
2149
2150</div>
2151
2152<!-- ======================================================================= -->
2153<div class="doc_subsection">
2154 <a name="globalconstants">Global Variable and Function Addresses</a>
2155</div>
2156
2157<div class="doc_text">
2158
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2159<p>The addresses of <a href="#globalvars">global variables</a>
2160 and <a href="#functionstructure">functions</a> are always implicitly valid
2161 (link-time) constants. These constants are explicitly referenced when
2162 the <a href="#identifiers">identifier for the global</a> is used and always
2163 have <a href="#t_pointer">pointer</a> type. For example, the following is a
2164 legal LLVM file:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2165
2166<div class="doc_code">
2167<pre>
2168@X = global i32 17
2169@Y = global i32 42
2170@Z = global [2 x i32*] [ i32* @X, i32* @Y ]
2171</pre>
2172</div>
2173
2174</div>
2175
2176<!-- ======================================================================= -->
2177<div class="doc_subsection"><a name="undefvalues">Undefined Values</a></div>
2178<div class="doc_text">
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2179
Chris Lattner3d72cd82009-09-07 22:52:39 +0000[diff] [blame]2180<p>The string '<tt>undef</tt>' can be used anywhere a constant is expected, and
Benjamin Kramer5fb9d7e2009-10-12 14:46:08 +0000[diff] [blame]2181 indicates that the user of the value may receive an unspecified bit-pattern.
Chris Lattner3d72cd82009-09-07 22:52:39 +0000[diff] [blame]2182 Undefined values may be of any type (other than label or void) and be used
2183 anywhere a constant is permitted.</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2184
Chris Lattner9d1e4fa2009-09-11 01:49:31 +0000[diff] [blame]2185<p>Undefined values are useful because they indicate to the compiler that the
Chris Lattner3d72cd82009-09-07 22:52:39 +0000[diff] [blame]2186 program is well defined no matter what value is used. This gives the
2187 compiler more freedom to optimize. Here are some examples of (potentially
2188 surprising) transformations that are valid (in pseudo IR):</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2189
Chris Lattner3d72cd82009-09-07 22:52:39 +0000[diff] [blame]2190
2191<div class="doc_code">
2192<pre>
2193 %A = add %X, undef
2194 %B = sub %X, undef
2195 %C = xor %X, undef
2196Safe:
2197 %A = undef
2198 %B = undef
2199 %C = undef
2200</pre>
2201</div>
2202
2203<p>This is safe because all of the output bits are affected by the undef bits.
2204Any output bit can have a zero or one depending on the input bits.</p>
2205
2206<div class="doc_code">
2207<pre>
2208 %A = or %X, undef
2209 %B = and %X, undef
2210Safe:
2211 %A = -1
2212 %B = 0
2213Unsafe:
2214 %A = undef
2215 %B = undef
2216</pre>
2217</div>
2218
2219<p>These logical operations have bits that are not always affected by the input.
2220For example, if "%X" has a zero bit, then the output of the 'and' operation will
2221always be a zero, no matter what the corresponding bit from the undef is. As
Chris Lattner9d1e4fa2009-09-11 01:49:31 +0000[diff] [blame]2222such, it is unsafe to optimize or assume that the result of the and is undef.
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]2223However, it is safe to assume that all bits of the undef could be 0, and
2224optimize the and to 0. Likewise, it is safe to assume that all the bits of
2225the undef operand to the or could be set, allowing the or to be folded to
Chris Lattner9d1e4fa2009-09-11 01:49:31 +0000[diff] [blame]2226-1.</p>
Chris Lattner3d72cd82009-09-07 22:52:39 +0000[diff] [blame]2227
2228<div class="doc_code">
2229<pre>
2230 %A = select undef, %X, %Y
2231 %B = select undef, 42, %Y
2232 %C = select %X, %Y, undef
2233Safe:
2234 %A = %X (or %Y)
2235 %B = 42 (or %Y)
2236 %C = %Y
2237Unsafe:
2238 %A = undef
2239 %B = undef
2240 %C = undef
2241</pre>
2242</div>
2243
2244<p>This set of examples show that undefined select (and conditional branch)
2245conditions can go "either way" but they have to come from one of the two
2246operands. In the %A example, if %X and %Y were both known to have a clear low
2247bit, then %A would have to have a cleared low bit. However, in the %C example,
2248the optimizer is allowed to assume that the undef operand could be the same as
2249%Y, allowing the whole select to be eliminated.</p>
2250
2251
2252<div class="doc_code">
2253<pre>
2254 %A = xor undef, undef
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]2255
Chris Lattner3d72cd82009-09-07 22:52:39 +0000[diff] [blame]2256 %B = undef
2257 %C = xor %B, %B
2258
2259 %D = undef
2260 %E = icmp lt %D, 4
2261 %F = icmp gte %D, 4
2262
2263Safe:
2264 %A = undef
2265 %B = undef
2266 %C = undef
2267 %D = undef
2268 %E = undef
2269 %F = undef
2270</pre>
2271</div>
2272
2273<p>This example points out that two undef operands are not necessarily the same.
2274This can be surprising to people (and also matches C semantics) where they
2275assume that "X^X" is always zero, even if X is undef. This isn't true for a
2276number of reasons, but the short answer is that an undef "variable" can
2277arbitrarily change its value over its "live range". This is true because the
2278"variable" doesn't actually <em>have a live range</em>. Instead, the value is
2279logically read from arbitrary registers that happen to be around when needed,
Benjamin Kramer5fb9d7e2009-10-12 14:46:08 +0000[diff] [blame]2280so the value is not necessarily consistent over time. In fact, %A and %C need
Chris Lattnerc802e482009-09-08 15:13:16 +0000[diff] [blame]2281to have the same semantics or the core LLVM "replace all uses with" concept
Chris Lattner3d72cd82009-09-07 22:52:39 +0000[diff] [blame]2282would not hold.</p>
Chris Lattner466291f2009-09-07 23:33:52 +0000[diff] [blame]2283
2284<div class="doc_code">
2285<pre>
2286 %A = fdiv undef, %X
2287 %B = fdiv %X, undef
2288Safe:
2289 %A = undef
2290b: unreachable
2291</pre>
2292</div>
2293
2294<p>These examples show the crucial difference between an <em>undefined
2295value</em> and <em>undefined behavior</em>. An undefined value (like undef) is
2296allowed to have an arbitrary bit-pattern. This means that the %A operation
2297can be constant folded to undef because the undef could be an SNaN, and fdiv is
2298not (currently) defined on SNaN's. However, in the second example, we can make
2299a more aggressive assumption: because the undef is allowed to be an arbitrary
2300value, we are allowed to assume that it could be zero. Since a divide by zero
Chris Lattner83d45d72009-09-08 19:45:34 +0000[diff] [blame]2301has <em>undefined behavior</em>, we are allowed to assume that the operation
Chris Lattner466291f2009-09-07 23:33:52 +0000[diff] [blame]2302does not execute at all. This allows us to delete the divide and all code after
2303it: since the undefined operation "can't happen", the optimizer can assume that
2304it occurs in dead code.
2305</p>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]2306
Chris Lattner466291f2009-09-07 23:33:52 +0000[diff] [blame]2307<div class="doc_code">
2308<pre>
2309a: store undef -> %X
2310b: store %X -> undef
2311Safe:
2312a: &lt;deleted&gt;
2313b: unreachable
2314</pre>
2315</div>
2316
2317<p>These examples reiterate the fdiv example: a store "of" an undefined value
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]2318can be assumed to not have any effect: we can assume that the value is
Chris Lattner466291f2009-09-07 23:33:52 +0000[diff] [blame]2319overwritten with bits that happen to match what was already there. However, a
2320store "to" an undefined location could clobber arbitrary memory, therefore, it
2321has undefined behavior.</p>
2322
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2323</div>
2324
2325<!-- ======================================================================= -->
Dan Gohmanb8ddf022010-04-22 23:14:21 +0000[diff] [blame]2326<div class="doc_subsection"><a name="trapvalues">Trap Values</a></div>
2327<div class="doc_text">
2328
Dan Gohman67bf37f2010-04-26 20:21:21 +0000[diff] [blame]2329<p>Trap values are similar to <a href="#undefvalues">undef values</a>, however
Dan Gohmanb8ddf022010-04-22 23:14:21 +0000[diff] [blame]2330 instead of representing an unspecified bit pattern, they represent the
2331 fact that an instruction or constant expression which cannot evoke side
2332 effects has nevertheless detected a condition which results in undefined
Dan Gohman67bf37f2010-04-26 20:21:21 +0000[diff] [blame]2333 behavior.</p>
Dan Gohmanb8ddf022010-04-22 23:14:21 +0000[diff] [blame]2334
Dan Gohman49ddd982010-04-24 22:15:58 +0000[diff] [blame]2335<p>Any non-void instruction or constant expression other than a non-intrinsic
2336 call, invoke, or phi with a trap operand has trap as its result value.
Dan Gohmanb8ddf022010-04-22 23:14:21 +0000[diff] [blame]2337 Any instruction with a trap operand which may have side effects emits
2338 those side effects as if it had an undef operand instead.</p>
2339
Dan Gohman54884272010-04-26 20:54:53 +0000[diff] [blame]2340<p>If a <a href="#i_br"><tt>br</tt></a> or
2341 <a href="#i_switch"><tt>switch</tt></a> instruction has a trap value
2342 operand, all non-phi non-void instructions which control-depend on it
2343 have trap as their result value. If any instruction which
2344 control-depends on the <tt>br</tt> or <tt>switch</tt> invokes externally
2345 visible side effects, the behavior of the program is undefined.</p>
2346
2347<!-- FIXME: What about exceptions thrown from control-dependent instrs? -->
2348
Dan Gohmanb8ddf022010-04-22 23:14:21 +0000[diff] [blame]2349<p>For example, an <a href="#i_and"><tt>and</tt></a> of a trap value with
2350 zero still has a trap value result. Using that value as an index in a
2351 <a href="#i_getelementptr"><tt>getelementptr</tt></a> yields a trap
2352 result. Using that result as the address of a
2353 <a href="#i_store"><tt>store</tt></a> produces undefined behavior.</p>
2354
2355<p>There is currently no way of representing a trap constant in the IR; they
2356 only exist when produced by certain instructions, such as an
2357 <a href="#i_add"><tt>add</tt></a> with the <tt>nsw</tt> flag
2358 set, when overflow occurs.</p>
2359
2360</div>
2361
2362<!-- ======================================================================= -->
Chris Lattner29246b52009-10-27 21:19:13 +0000[diff] [blame]2363<div class="doc_subsection"><a name="blockaddress">Addresses of Basic
2364 Blocks</a></div>
Chris Lattnerd07c8372009-10-27 21:01:34 +0000[diff] [blame]2365<div class="doc_text">
2366
Chris Lattner620cead2009-11-01 01:27:45 +0000[diff] [blame]2367<p><b><tt>blockaddress(@function, %block)</tt></b></p>
Chris Lattnerd07c8372009-10-27 21:01:34 +0000[diff] [blame]2368
2369<p>The '<tt>blockaddress</tt>' constant computes the address of the specified
Chris Lattner24f2e852009-10-27 21:49:40 +0000[diff] [blame]2370 basic block in the specified function, and always has an i8* type. Taking
Chris Lattner620cead2009-11-01 01:27:45 +0000[diff] [blame]2371 the address of the entry block is illegal.</p>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]2372
Chris Lattnerd07c8372009-10-27 21:01:34 +0000[diff] [blame]2373<p>This value only has defined behavior when used as an operand to the
Chris Lattner4c3800f2009-10-28 00:19:10 +0000[diff] [blame]2374 '<a href="#i_indirectbr"><tt>indirectbr</tt></a>' instruction or for comparisons
Chris Lattnerd07c8372009-10-27 21:01:34 +0000[diff] [blame]2375 against null. Pointer equality tests between labels addresses is undefined
2376 behavior - though, again, comparison against null is ok, and no label is
Chris Lattner29246b52009-10-27 21:19:13 +0000[diff] [blame]2377 equal to the null pointer. This may also be passed around as an opaque
2378 pointer sized value as long as the bits are not inspected. This allows
Chris Lattner0bae7b32009-10-27 21:44:20 +0000[diff] [blame]2379 <tt>ptrtoint</tt> and arithmetic to be performed on these values so long as
Chris Lattner4c3800f2009-10-28 00:19:10 +0000[diff] [blame]2380 the original value is reconstituted before the <tt>indirectbr</tt>.</p>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]2381
Chris Lattner29246b52009-10-27 21:19:13 +0000[diff] [blame]2382<p>Finally, some targets may provide defined semantics when
Chris Lattnerd07c8372009-10-27 21:01:34 +0000[diff] [blame]2383 using the value as the operand to an inline assembly, but that is target
2384 specific.
2385 </p>
2386
2387</div>
2388
2389
2390<!-- ======================================================================= -->
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2391<div class="doc_subsection"><a name="constantexprs">Constant Expressions</a>
2392</div>
2393
2394<div class="doc_text">
2395
2396<p>Constant expressions are used to allow expressions involving other constants
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2397 to be used as constants. Constant expressions may be of
2398 any <a href="#t_firstclass">first class</a> type and may involve any LLVM
2399 operation that does not have side effects (e.g. load and call are not
2400 supported). The following is the syntax for constant expressions:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2401
2402<dl>
2403 <dt><b><tt>trunc ( CST to TYPE )</tt></b></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2404 <dd>Truncate a constant to another type. The bit size of CST must be larger
2405 than the bit size of TYPE. Both types must be integers.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2406
2407 <dt><b><tt>zext ( CST to TYPE )</tt></b></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2408 <dd>Zero extend a constant to another type. The bit size of CST must be
2409 smaller or equal to the bit size of TYPE. Both types must be
2410 integers.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2411
2412 <dt><b><tt>sext ( CST to TYPE )</tt></b></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2413 <dd>Sign extend a constant to another type. The bit size of CST must be
2414 smaller or equal to the bit size of TYPE. Both types must be
2415 integers.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2416
2417 <dt><b><tt>fptrunc ( CST to TYPE )</tt></b></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2418 <dd>Truncate a floating point constant to another floating point type. The
2419 size of CST must be larger than the size of TYPE. Both types must be
2420 floating point.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2421
2422 <dt><b><tt>fpext ( CST to TYPE )</tt></b></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2423 <dd>Floating point extend a constant to another type. The size of CST must be
2424 smaller or equal to the size of TYPE. Both types must be floating
2425 point.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2426
Reid Spencere6adee82007-07-31 14:40:14 +0000[diff] [blame]2427 <dt><b><tt>fptoui ( CST to TYPE )</tt></b></dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2428 <dd>Convert a floating point constant to the corresponding unsigned integer
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2429 constant. TYPE must be a scalar or vector integer type. CST must be of
2430 scalar or vector floating point type. Both CST and TYPE must be scalars,
2431 or vectors of the same number of elements. If the value won't fit in the
2432 integer type, the results are undefined.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2433
2434 <dt><b><tt>fptosi ( CST to TYPE )</tt></b></dt>
2435 <dd>Convert a floating point constant to the corresponding signed integer
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2436 constant. TYPE must be a scalar or vector integer type. CST must be of
2437 scalar or vector floating point type. Both CST and TYPE must be scalars,
2438 or vectors of the same number of elements. If the value won't fit in the
2439 integer type, the results are undefined.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2440
2441 <dt><b><tt>uitofp ( CST to TYPE )</tt></b></dt>
2442 <dd>Convert an unsigned integer constant to the corresponding floating point
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2443 constant. TYPE must be a scalar or vector floating point type. CST must be
2444 of scalar or vector integer type. Both CST and TYPE must be scalars, or
2445 vectors of the same number of elements. If the value won't fit in the
2446 floating point type, the results are undefined.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2447
2448 <dt><b><tt>sitofp ( CST to TYPE )</tt></b></dt>
2449 <dd>Convert a signed integer constant to the corresponding floating point
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2450 constant. TYPE must be a scalar or vector floating point type. CST must be
2451 of scalar or vector integer type. Both CST and TYPE must be scalars, or
2452 vectors of the same number of elements. If the value won't fit in the
2453 floating point type, the results are undefined.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2454
2455 <dt><b><tt>ptrtoint ( CST to TYPE )</tt></b></dt>
2456 <dd>Convert a pointer typed constant to the corresponding integer constant
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2457 <tt>TYPE</tt> must be an integer type. <tt>CST</tt> must be of pointer
2458 type. The <tt>CST</tt> value is zero extended, truncated, or unchanged to
2459 make it fit in <tt>TYPE</tt>.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2460
2461 <dt><b><tt>inttoptr ( CST to TYPE )</tt></b></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2462 <dd>Convert a integer constant to a pointer constant. TYPE must be a pointer
2463 type. CST must be of integer type. The CST value is zero extended,
2464 truncated, or unchanged to make it fit in a pointer size. This one is
2465 <i>really</i> dangerous!</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2466
2467 <dt><b><tt>bitcast ( CST to TYPE )</tt></b></dt>
Chris Lattner557bc5d2009-02-28 18:27:03 +0000[diff] [blame]2468 <dd>Convert a constant, CST, to another TYPE. The constraints of the operands
2469 are the same as those for the <a href="#i_bitcast">bitcast
2470 instruction</a>.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2471
2472 <dt><b><tt>getelementptr ( CSTPTR, IDX0, IDX1, ... )</tt></b></dt>
Dan Gohman106b2ae2009-07-27 21:53:46 +0000[diff] [blame]2473 <dt><b><tt>getelementptr inbounds ( CSTPTR, IDX0, IDX1, ... )</tt></b></dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2474 <dd>Perform the <a href="#i_getelementptr">getelementptr operation</a> on
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2475 constants. As with the <a href="#i_getelementptr">getelementptr</a>
2476 instruction, the index list may have zero or more indexes, which are
2477 required to make sense for the type of "CSTPTR".</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2478
2479 <dt><b><tt>select ( COND, VAL1, VAL2 )</tt></b></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2480 <dd>Perform the <a href="#i_select">select operation</a> on constants.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2481
2482 <dt><b><tt>icmp COND ( VAL1, VAL2 )</tt></b></dt>
2483 <dd>Performs the <a href="#i_icmp">icmp operation</a> on constants.</dd>
2484
2485 <dt><b><tt>fcmp COND ( VAL1, VAL2 )</tt></b></dt>
2486 <dd>Performs the <a href="#i_fcmp">fcmp operation</a> on constants.</dd>
2487
2488 <dt><b><tt>extractelement ( VAL, IDX )</tt></b></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2489 <dd>Perform the <a href="#i_extractelement">extractelement operation</a> on
2490 constants.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2491
2492 <dt><b><tt>insertelement ( VAL, ELT, IDX )</tt></b></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2493 <dd>Perform the <a href="#i_insertelement">insertelement operation</a> on
2494 constants.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2495
2496 <dt><b><tt>shufflevector ( VEC1, VEC2, IDXMASK )</tt></b></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2497 <dd>Perform the <a href="#i_shufflevector">shufflevector operation</a> on
2498 constants.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2499
2500 <dt><b><tt>OPCODE ( LHS, RHS )</tt></b></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2501 <dd>Perform the specified operation of the LHS and RHS constants. OPCODE may
2502 be any of the <a href="#binaryops">binary</a>
2503 or <a href="#bitwiseops">bitwise binary</a> operations. The constraints
2504 on operands are the same as those for the corresponding instruction
2505 (e.g. no bitwise operations on floating point values are allowed).</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2506</dl>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2507
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2508</div>
2509
2510<!-- *********************************************************************** -->
2511<div class="doc_section"> <a name="othervalues">Other Values</a> </div>
2512<!-- *********************************************************************** -->
2513
2514<!-- ======================================================================= -->
2515<div class="doc_subsection">
2516<a name="inlineasm">Inline Assembler Expressions</a>
2517</div>
2518
2519<div class="doc_text">
2520
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2521<p>LLVM supports inline assembler expressions (as opposed
2522 to <a href="#moduleasm"> Module-Level Inline Assembly</a>) through the use of
2523 a special value. This value represents the inline assembler as a string
2524 (containing the instructions to emit), a list of operand constraints (stored
Dale Johannesen648950f2009-10-13 21:56:55 +0000[diff] [blame]2525 as a string), a flag that indicates whether or not the inline asm
Dale Johannesen5ee3e4b2009-10-21 23:28:00 +0000[diff] [blame]2526 expression has side effects, and a flag indicating whether the function
2527 containing the asm needs to align its stack conservatively. An example
2528 inline assembler expression is:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2529
2530<div class="doc_code">
2531<pre>
2532i32 (i32) asm "bswap $0", "=r,r"
2533</pre>
2534</div>
2535
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2536<p>Inline assembler expressions may <b>only</b> be used as the callee operand of
2537 a <a href="#i_call"><tt>call</tt> instruction</a>. Thus, typically we
2538 have:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2539
2540<div class="doc_code">
2541<pre>
2542%X = call i32 asm "<a href="#int_bswap">bswap</a> $0", "=r,r"(i32 %Y)
2543</pre>
2544</div>
2545
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2546<p>Inline asms with side effects not visible in the constraint list must be
2547 marked as having side effects. This is done through the use of the
2548 '<tt>sideeffect</tt>' keyword, like so:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2549
2550<div class="doc_code">
2551<pre>
2552call void asm sideeffect "eieio", ""()
2553</pre>
2554</div>
2555
Dale Johannesen5ee3e4b2009-10-21 23:28:00 +0000[diff] [blame]2556<p>In some cases inline asms will contain code that will not work unless the
2557 stack is aligned in some way, such as calls or SSE instructions on x86,
2558 yet will not contain code that does that alignment within the asm.
2559 The compiler should make conservative assumptions about what the asm might
2560 contain and should generate its usual stack alignment code in the prologue
2561 if the '<tt>alignstack</tt>' keyword is present:</p>
Dale Johannesen648950f2009-10-13 21:56:55 +0000[diff] [blame]2562
2563<div class="doc_code">
2564<pre>
Dale Johannesen5ee3e4b2009-10-21 23:28:00 +0000[diff] [blame]2565call void asm alignstack "eieio", ""()
Dale Johannesen648950f2009-10-13 21:56:55 +0000[diff] [blame]2566</pre>
2567</div>
2568
2569<p>If both keywords appear the '<tt>sideeffect</tt>' keyword must come
2570 first.</p>
2571
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2572<p>TODO: The format of the asm and constraints string still need to be
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2573 documented here. Constraints on what can be done (e.g. duplication, moving,
2574 etc need to be documented). This is probably best done by reference to
2575 another document that covers inline asm from a holistic perspective.</p>
Chris Lattnerbafc8372010-04-07 05:38:05 +0000[diff] [blame]2576</div>
2577
2578<div class="doc_subsubsection">
2579<a name="inlineasm_md">Inline Asm Metadata</a>
2580</div>
2581
2582<div class="doc_text">
2583
2584<p>The call instructions that wrap inline asm nodes may have a "!srcloc" MDNode
2585 attached to it that contains a constant integer. If present, the code
2586 generator will use the integer as the location cookie value when report
2587 errors through the LLVMContext error reporting mechanisms. This allows a
2588 front-end to corrolate backend errors that occur with inline asm back to the
2589 source code that produced it. For example:</p>
2590
2591<div class="doc_code">
2592<pre>
2593call void asm sideeffect "something bad", ""()<b>, !srcloc !42</b>
2594...
2595!42 = !{ i32 1234567 }
2596</pre>
2597</div>
2598
2599<p>It is up to the front-end to make sense of the magic numbers it places in the
2600 IR.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2601
2602</div>
2603
Chris Lattnerd0d96292010-01-15 21:50:19 +0000[diff] [blame]2604<!-- ======================================================================= -->
2605<div class="doc_subsection"><a name="metadata">Metadata Nodes and Metadata
2606 Strings</a>
2607</div>
2608
2609<div class="doc_text">
2610
2611<p>LLVM IR allows metadata to be attached to instructions in the program that
2612 can convey extra information about the code to the optimizers and code
2613 generator. One example application of metadata is source-level debug
2614 information. There are two metadata primitives: strings and nodes. All
2615 metadata has the <tt>metadata</tt> type and is identified in syntax by a
2616 preceding exclamation point ('<tt>!</tt>').</p>
2617
2618<p>A metadata string is a string surrounded by double quotes. It can contain
2619 any character by escaping non-printable characters with "\xx" where "xx" is
2620 the two digit hex code. For example: "<tt>!"test\00"</tt>".</p>
2621
2622<p>Metadata nodes are represented with notation similar to structure constants
2623 (a comma separated list of elements, surrounded by braces and preceded by an
2624 exclamation point). For example: "<tt>!{ metadata !"test\00", i32
2625 10}</tt>". Metadata nodes can have any values as their operand.</p>
2626
2627<p>A <a href="#namedmetadatastructure">named metadata</a> is a collection of
2628 metadata nodes, which can be looked up in the module symbol table. For
2629 example: "<tt>!foo = metadata !{!4, !3}</tt>".
2630
Devang Patelb1586922010-03-04 23:44:48 +0000[diff] [blame]2631<p>Metadata can be used as function arguments. Here <tt>llvm.dbg.value</tt>
2632 function is using two metadata arguments.
2633
2634 <div class="doc_code">
2635 <pre>
2636 call void @llvm.dbg.value(metadata !24, i64 0, metadata !25)
2637 </pre>
2638 </div></p>
2639
2640<p>Metadata can be attached with an instruction. Here metadata <tt>!21</tt> is
2641 attached with <tt>add</tt> instruction using <tt>!dbg</tt> identifier.
2642
2643 <div class="doc_code">
2644 <pre>
2645 %indvar.next = add i64 %indvar, 1, !dbg !21
2646 </pre>
2647 </div></p>
Chris Lattnerd0d96292010-01-15 21:50:19 +0000[diff] [blame]2648</div>
2649
Chris Lattner75c24e02009-07-20 05:55:19 +0000[diff] [blame]2650
2651<!-- *********************************************************************** -->
2652<div class="doc_section">
2653 <a name="intrinsic_globals">Intrinsic Global Variables</a>
2654</div>
2655<!-- *********************************************************************** -->
2656
2657<p>LLVM has a number of "magic" global variables that contain data that affect
2658code generation or other IR semantics. These are documented here. All globals
Chris Lattner1e0e0d12009-07-20 06:14:25 +0000[diff] [blame]2659of this sort should have a section specified as "<tt>llvm.metadata</tt>". This
2660section and all globals that start with "<tt>llvm.</tt>" are reserved for use
2661by LLVM.</p>
Chris Lattner75c24e02009-07-20 05:55:19 +0000[diff] [blame]2662
2663<!-- ======================================================================= -->
2664<div class="doc_subsection">
2665<a name="intg_used">The '<tt>llvm.used</tt>' Global Variable</a>
2666</div>
2667
2668<div class="doc_text">
2669
2670<p>The <tt>@llvm.used</tt> global is an array with i8* element type which has <a
2671href="#linkage_appending">appending linkage</a>. This array contains a list of
2672pointers to global variables and functions which may optionally have a pointer
2673cast formed of bitcast or getelementptr. For example, a legal use of it is:</p>
2674
2675<pre>
2676 @X = global i8 4
2677 @Y = global i32 123
2678
2679 @llvm.used = appending global [2 x i8*] [
2680 i8* @X,
2681 i8* bitcast (i32* @Y to i8*)
2682 ], section "llvm.metadata"
2683</pre>
2684
2685<p>If a global variable appears in the <tt>@llvm.used</tt> list, then the
2686compiler, assembler, and linker are required to treat the symbol as if there is
2687a reference to the global that it cannot see. For example, if a variable has
2688internal linkage and no references other than that from the <tt>@llvm.used</tt>
2689list, it cannot be deleted. This is commonly used to represent references from
2690inline asms and other things the compiler cannot "see", and corresponds to
2691"attribute((used))" in GNU C.</p>
2692
2693<p>On some targets, the code generator must emit a directive to the assembler or
2694object file to prevent the assembler and linker from molesting the symbol.</p>
2695
2696</div>
2697
2698<!-- ======================================================================= -->
2699<div class="doc_subsection">
Chris Lattner1e0e0d12009-07-20 06:14:25 +0000[diff] [blame]2700<a name="intg_compiler_used">The '<tt>llvm.compiler.used</tt>' Global Variable</a>
2701</div>
2702
2703<div class="doc_text">
2704
2705<p>The <tt>@llvm.compiler.used</tt> directive is the same as the
2706<tt>@llvm.used</tt> directive, except that it only prevents the compiler from
2707touching the symbol. On targets that support it, this allows an intelligent
2708linker to optimize references to the symbol without being impeded as it would be
2709by <tt>@llvm.used</tt>.</p>
2710
2711<p>This is a rare construct that should only be used in rare circumstances, and
2712should not be exposed to source languages.</p>
2713
2714</div>
2715
2716<!-- ======================================================================= -->
2717<div class="doc_subsection">
Chris Lattner75c24e02009-07-20 05:55:19 +0000[diff] [blame]2718<a name="intg_global_ctors">The '<tt>llvm.global_ctors</tt>' Global Variable</a>
2719</div>
2720
2721<div class="doc_text">
2722
2723<p>TODO: Describe this.</p>
2724
2725</div>
2726
2727<!-- ======================================================================= -->
2728<div class="doc_subsection">
2729<a name="intg_global_dtors">The '<tt>llvm.global_dtors</tt>' Global Variable</a>
2730</div>
2731
2732<div class="doc_text">
2733
2734<p>TODO: Describe this.</p>
2735
2736</div>
2737
2738
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2739<!-- *********************************************************************** -->
2740<div class="doc_section"> <a name="instref">Instruction Reference</a> </div>
2741<!-- *********************************************************************** -->
2742
2743<div class="doc_text">
2744
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2745<p>The LLVM instruction set consists of several different classifications of
2746 instructions: <a href="#terminators">terminator
2747 instructions</a>, <a href="#binaryops">binary instructions</a>,
2748 <a href="#bitwiseops">bitwise binary instructions</a>,
2749 <a href="#memoryops">memory instructions</a>, and
2750 <a href="#otherops">other instructions</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2751
2752</div>
2753
2754<!-- ======================================================================= -->
2755<div class="doc_subsection"> <a name="terminators">Terminator
2756Instructions</a> </div>
2757
2758<div class="doc_text">
2759
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2760<p>As mentioned <a href="#functionstructure">previously</a>, every basic block
2761 in a program ends with a "Terminator" instruction, which indicates which
2762 block should be executed after the current block is finished. These
2763 terminator instructions typically yield a '<tt>void</tt>' value: they produce
2764 control flow, not values (the one exception being the
2765 '<a href="#i_invoke"><tt>invoke</tt></a>' instruction).</p>
2766
Duncan Sands048d8062010-04-15 20:35:54 +0000[diff] [blame]2767<p>There are seven different terminator instructions: the
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2768 '<a href="#i_ret"><tt>ret</tt></a>' instruction, the
2769 '<a href="#i_br"><tt>br</tt></a>' instruction, the
2770 '<a href="#i_switch"><tt>switch</tt></a>' instruction, the
Bill Wendlingb4ae2292009-11-02 00:25:26 +0000[diff] [blame]2771 '<a href="#i_indirectbr">'<tt>indirectbr</tt></a>' Instruction, the
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2772 '<a href="#i_invoke"><tt>invoke</tt></a>' instruction, the
2773 '<a href="#i_unwind"><tt>unwind</tt></a>' instruction, and the
2774 '<a href="#i_unreachable"><tt>unreachable</tt></a>' instruction.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2775
2776</div>
2777
2778<!-- _______________________________________________________________________ -->
2779<div class="doc_subsubsection"> <a name="i_ret">'<tt>ret</tt>'
2780Instruction</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2781
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2782<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2783
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2784<h5>Syntax:</h5>
Dan Gohman3e700032008-10-04 19:00:07 +0000[diff] [blame]2785<pre>
2786 ret &lt;type&gt; &lt;value&gt; <i>; Return a value from a non-void function</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2787 ret void <i>; Return from void function</i>
2788</pre>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]2789
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2790<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2791<p>The '<tt>ret</tt>' instruction is used to return control flow (and optionally
2792 a value) from a function back to the caller.</p>
2793
2794<p>There are two forms of the '<tt>ret</tt>' instruction: one that returns a
2795 value and then causes control flow, and one that just causes control flow to
2796 occur.</p>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]2797
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2798<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2799<p>The '<tt>ret</tt>' instruction optionally accepts a single argument, the
2800 return value. The type of the return value must be a
2801 '<a href="#t_firstclass">first class</a>' type.</p>
Dan Gohman3e700032008-10-04 19:00:07 +0000[diff] [blame]2802
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2803<p>A function is not <a href="#wellformed">well formed</a> if it it has a
2804 non-void return type and contains a '<tt>ret</tt>' instruction with no return
2805 value or a return value with a type that does not match its type, or if it
2806 has a void return type and contains a '<tt>ret</tt>' instruction with a
2807 return value.</p>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]2808
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2809<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2810<p>When the '<tt>ret</tt>' instruction is executed, control flow returns back to
2811 the calling function's context. If the caller is a
2812 "<a href="#i_call"><tt>call</tt></a>" instruction, execution continues at the
2813 instruction after the call. If the caller was an
2814 "<a href="#i_invoke"><tt>invoke</tt></a>" instruction, execution continues at
2815 the beginning of the "normal" destination block. If the instruction returns
2816 a value, that value shall set the call or invoke instruction's return
2817 value.</p>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]2818
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2819<h5>Example:</h5>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]2820<pre>
2821 ret i32 5 <i>; Return an integer value of 5</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2822 ret void <i>; Return from a void function</i>
Bill Wendlingd163e2d2009-02-28 22:12:54 +0000[diff] [blame]2823 ret { i32, i8 } { i32 4, i8 2 } <i>; Return a struct of values 4 and 2</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2824</pre>
Dan Gohman60967192009-01-12 23:12:39 +0000[diff] [blame]2825
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2826</div>
2827<!-- _______________________________________________________________________ -->
2828<div class="doc_subsubsection"> <a name="i_br">'<tt>br</tt>' Instruction</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2829
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2830<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2831
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2832<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2833<pre>
2834 br i1 &lt;cond&gt;, label &lt;iftrue&gt;, label &lt;iffalse&gt;<br> br label &lt;dest&gt; <i>; Unconditional branch</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2835</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2836
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2837<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2838<p>The '<tt>br</tt>' instruction is used to cause control flow to transfer to a
2839 different basic block in the current function. There are two forms of this
2840 instruction, corresponding to a conditional branch and an unconditional
2841 branch.</p>
2842
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2843<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2844<p>The conditional branch form of the '<tt>br</tt>' instruction takes a single
2845 '<tt>i1</tt>' value and two '<tt>label</tt>' values. The unconditional form
2846 of the '<tt>br</tt>' instruction takes a single '<tt>label</tt>' value as a
2847 target.</p>
2848
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2849<h5>Semantics:</h5>
2850<p>Upon execution of a conditional '<tt>br</tt>' instruction, the '<tt>i1</tt>'
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2851 argument is evaluated. If the value is <tt>true</tt>, control flows to the
2852 '<tt>iftrue</tt>' <tt>label</tt> argument. If "cond" is <tt>false</tt>,
2853 control flows to the '<tt>iffalse</tt>' <tt>label</tt> argument.</p>
2854
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2855<h5>Example:</h5>
Bill Wendling6ec40612009-07-20 02:39:26 +0000[diff] [blame]2856<pre>
2857Test:
2858 %cond = <a href="#i_icmp">icmp</a> eq i32 %a, %b
2859 br i1 %cond, label %IfEqual, label %IfUnequal
2860IfEqual:
2861 <a href="#i_ret">ret</a> i32 1
2862IfUnequal:
2863 <a href="#i_ret">ret</a> i32 0
2864</pre>
2865
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2866</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2867
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2868<!-- _______________________________________________________________________ -->
2869<div class="doc_subsubsection">
2870 <a name="i_switch">'<tt>switch</tt>' Instruction</a>
2871</div>
2872
2873<div class="doc_text">
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2874
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2875<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2876<pre>
2877 switch &lt;intty&gt; &lt;value&gt;, label &lt;defaultdest&gt; [ &lt;intty&gt; &lt;val&gt;, label &lt;dest&gt; ... ]
2878</pre>
2879
2880<h5>Overview:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2881<p>The '<tt>switch</tt>' instruction is used to transfer control flow to one of
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2882 several different places. It is a generalization of the '<tt>br</tt>'
2883 instruction, allowing a branch to occur to one of many possible
2884 destinations.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2885
2886<h5>Arguments:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2887<p>The '<tt>switch</tt>' instruction uses three parameters: an integer
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2888 comparison value '<tt>value</tt>', a default '<tt>label</tt>' destination,
2889 and an array of pairs of comparison value constants and '<tt>label</tt>'s.
2890 The table is not allowed to contain duplicate constant entries.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2891
2892<h5>Semantics:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2893<p>The <tt>switch</tt> instruction specifies a table of values and
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2894 destinations. When the '<tt>switch</tt>' instruction is executed, this table
2895 is searched for the given value. If the value is found, control flow is
Benjamin Kramer5fb9d7e2009-10-12 14:46:08 +0000[diff] [blame]2896 transferred to the corresponding destination; otherwise, control flow is
2897 transferred to the default destination.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2898
2899<h5>Implementation:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2900<p>Depending on properties of the target machine and the particular
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2901 <tt>switch</tt> instruction, this instruction may be code generated in
2902 different ways. For example, it could be generated as a series of chained
2903 conditional branches or with a lookup table.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2904
2905<h5>Example:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2906<pre>
2907 <i>; Emulate a conditional br instruction</i>
2908 %Val = <a href="#i_zext">zext</a> i1 %value to i32
Dan Gohman01852382009-01-04 23:44:43 +0000[diff] [blame]2909 switch i32 %Val, label %truedest [ i32 0, label %falsedest ]
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2910
2911 <i>; Emulate an unconditional br instruction</i>
2912 switch i32 0, label %dest [ ]
2913
2914 <i>; Implement a jump table:</i>
Dan Gohman01852382009-01-04 23:44:43 +0000[diff] [blame]2915 switch i32 %val, label %otherwise [ i32 0, label %onzero
2916 i32 1, label %onone
2917 i32 2, label %ontwo ]
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2918</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2919
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2920</div>
2921
Chris Lattnere0787282009-10-27 19:13:16 +0000[diff] [blame]2922
2923<!-- _______________________________________________________________________ -->
2924<div class="doc_subsubsection">
Chris Lattner4c3800f2009-10-28 00:19:10 +0000[diff] [blame]2925 <a name="i_indirectbr">'<tt>indirectbr</tt>' Instruction</a>
Chris Lattnere0787282009-10-27 19:13:16 +0000[diff] [blame]2926</div>
2927
2928<div class="doc_text">
2929
2930<h5>Syntax:</h5>
2931<pre>
Chris Lattner4c3800f2009-10-28 00:19:10 +0000[diff] [blame]2932 indirectbr &lt;somety&gt;* &lt;address&gt;, [ label &lt;dest1&gt;, label &lt;dest2&gt;, ... ]
Chris Lattnere0787282009-10-27 19:13:16 +0000[diff] [blame]2933</pre>
2934
2935<h5>Overview:</h5>
2936
Chris Lattner4c3800f2009-10-28 00:19:10 +0000[diff] [blame]2937<p>The '<tt>indirectbr</tt>' instruction implements an indirect branch to a label
Chris Lattnere0787282009-10-27 19:13:16 +0000[diff] [blame]2938 within the current function, whose address is specified by
Chris Lattnerd07c8372009-10-27 21:01:34 +0000[diff] [blame]2939 "<tt>address</tt>". Address must be derived from a <a
2940 href="#blockaddress">blockaddress</a> constant.</p>
Chris Lattnere0787282009-10-27 19:13:16 +0000[diff] [blame]2941
2942<h5>Arguments:</h5>
2943
2944<p>The '<tt>address</tt>' argument is the address of the label to jump to. The
2945 rest of the arguments indicate the full set of possible destinations that the
2946 address may point to. Blocks are allowed to occur multiple times in the
2947 destination list, though this isn't particularly useful.</p>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]2948
Chris Lattnere0787282009-10-27 19:13:16 +0000[diff] [blame]2949<p>This destination list is required so that dataflow analysis has an accurate
2950 understanding of the CFG.</p>
2951
2952<h5>Semantics:</h5>
2953
2954<p>Control transfers to the block specified in the address argument. All
2955 possible destination blocks must be listed in the label list, otherwise this
2956 instruction has undefined behavior. This implies that jumps to labels
2957 defined in other functions have undefined behavior as well.</p>
2958
2959<h5>Implementation:</h5>
2960
2961<p>This is typically implemented with a jump through a register.</p>
2962
2963<h5>Example:</h5>
2964<pre>
Chris Lattner4c3800f2009-10-28 00:19:10 +0000[diff] [blame]2965 indirectbr i8* %Addr, [ label %bb1, label %bb2, label %bb3 ]
Chris Lattnere0787282009-10-27 19:13:16 +0000[diff] [blame]2966</pre>
2967
2968</div>
2969
2970
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2971<!-- _______________________________________________________________________ -->
2972<div class="doc_subsubsection">
2973 <a name="i_invoke">'<tt>invoke</tt>' Instruction</a>
2974</div>
2975
2976<div class="doc_text">
2977
2978<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2979<pre>
Devang Pateld0bfcc72008-10-07 17:48:33 +0000[diff] [blame]2980 &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>]
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2981 to label &lt;normal label&gt; unwind label &lt;exception label&gt;
2982</pre>
2983
2984<h5>Overview:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2985<p>The '<tt>invoke</tt>' instruction causes control to transfer to a specified
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2986 function, with the possibility of control flow transfer to either the
2987 '<tt>normal</tt>' label or the '<tt>exception</tt>' label. If the callee
2988 function returns with the "<tt><a href="#i_ret">ret</a></tt>" instruction,
2989 control flow will return to the "normal" label. If the callee (or any
2990 indirect callees) returns with the "<a href="#i_unwind"><tt>unwind</tt></a>"
2991 instruction, control is interrupted and continued at the dynamically nearest
2992 "exception" label.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2993
2994<h5>Arguments:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2995<p>This instruction requires several arguments:</p>
2996
2997<ol>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2998 <li>The optional "cconv" marker indicates which <a href="#callingconv">calling
2999 convention</a> the call should use. If none is specified, the call
3000 defaults to using C calling conventions.</li>
Devang Patelac2fc272008-10-06 18:50:38 +0000[diff] [blame]3001
3002 <li>The optional <a href="#paramattrs">Parameter Attributes</a> list for
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3003 return values. Only '<tt>zeroext</tt>', '<tt>signext</tt>', and
3004 '<tt>inreg</tt>' attributes are valid here.</li>
Devang Patelac2fc272008-10-06 18:50:38 +0000[diff] [blame]3005
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3006 <li>'<tt>ptr to function ty</tt>': shall be the signature of the pointer to
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3007 function value being invoked. In most cases, this is a direct function
3008 invocation, but indirect <tt>invoke</tt>s are just as possible, branching
3009 off an arbitrary pointer to function value.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3010
3011 <li>'<tt>function ptr val</tt>': An LLVM value containing a pointer to a
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3012 function to be invoked. </li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3013
3014 <li>'<tt>function args</tt>': argument list whose types match the function
Chris Lattner553fb1e2010-03-02 06:36:51 +0000[diff] [blame]3015 signature argument types and parameter attributes. All arguments must be
3016 of <a href="#t_firstclass">first class</a> type. If the function
3017 signature indicates the function accepts a variable number of arguments,
3018 the extra arguments can be specified.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3019
3020 <li>'<tt>normal label</tt>': the label reached when the called function
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3021 executes a '<tt><a href="#i_ret">ret</a></tt>' instruction. </li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3022
3023 <li>'<tt>exception label</tt>': the label reached when a callee returns with
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3024 the <a href="#i_unwind"><tt>unwind</tt></a> instruction. </li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3025
Devang Pateld0bfcc72008-10-07 17:48:33 +0000[diff] [blame]3026 <li>The optional <a href="#fnattrs">function attributes</a> list. Only
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3027 '<tt>noreturn</tt>', '<tt>nounwind</tt>', '<tt>readonly</tt>' and
3028 '<tt>readnone</tt>' attributes are valid here.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3029</ol>
3030
3031<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3032<p>This instruction is designed to operate as a standard
3033 '<tt><a href="#i_call">call</a></tt>' instruction in most regards. The
3034 primary difference is that it establishes an association with a label, which
3035 is used by the runtime library to unwind the stack.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3036
3037<p>This instruction is used in languages with destructors to ensure that proper
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3038 cleanup is performed in the case of either a <tt>longjmp</tt> or a thrown
3039 exception. Additionally, this is important for implementation of
3040 '<tt>catch</tt>' clauses in high-level languages that support them.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3041
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3042<p>For the purposes of the SSA form, the definition of the value returned by the
3043 '<tt>invoke</tt>' instruction is deemed to occur on the edge from the current
3044 block to the "normal" label. If the callee unwinds then no return value is
3045 available.</p>
Dan Gohman140ba5d2009-05-22 21:47:08 +0000[diff] [blame]3046
Chris Lattner4a91ef42010-01-15 18:08:37 +0000[diff] [blame]3047<p>Note that the code generator does not yet completely support unwind, and
3048that the invoke/unwind semantics are likely to change in future versions.</p>
3049
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3050<h5>Example:</h5>
3051<pre>
Nick Lewyckya1c11a12008-03-16 07:18:12 +0000[diff] [blame]3052 %retval = invoke i32 @Test(i32 15) to label %Continue
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3053 unwind label %TestCleanup <i>; {i32}:retval set</i>
Nick Lewyckya1c11a12008-03-16 07:18:12 +0000[diff] [blame]3054 %retval = invoke <a href="#callingconv">coldcc</a> i32 %Testfnptr(i32 15) to label %Continue
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3055 unwind label %TestCleanup <i>; {i32}:retval set</i>
3056</pre>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3057
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3058</div>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3059
3060<!-- _______________________________________________________________________ -->
3061
3062<div class="doc_subsubsection"> <a name="i_unwind">'<tt>unwind</tt>'
3063Instruction</a> </div>
3064
3065<div class="doc_text">
3066
3067<h5>Syntax:</h5>
3068<pre>
3069 unwind
3070</pre>
3071
3072<h5>Overview:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3073<p>The '<tt>unwind</tt>' instruction unwinds the stack, continuing control flow
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3074 at the first callee in the dynamic call stack which used
3075 an <a href="#i_invoke"><tt>invoke</tt></a> instruction to perform the call.
3076 This is primarily used to implement exception handling.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3077
3078<h5>Semantics:</h5>
Chris Lattner8b094fc2008-04-19 21:01:16 +0000[diff] [blame]3079<p>The '<tt>unwind</tt>' instruction causes execution of the current function to
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3080 immediately halt. The dynamic call stack is then searched for the
3081 first <a href="#i_invoke"><tt>invoke</tt></a> instruction on the call stack.
3082 Once found, execution continues at the "exceptional" destination block
3083 specified by the <tt>invoke</tt> instruction. If there is no <tt>invoke</tt>
3084 instruction in the dynamic call chain, undefined behavior results.</p>
3085
Chris Lattner4a91ef42010-01-15 18:08:37 +0000[diff] [blame]3086<p>Note that the code generator does not yet completely support unwind, and
3087that the invoke/unwind semantics are likely to change in future versions.</p>
3088
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3089</div>
3090
3091<!-- _______________________________________________________________________ -->
3092
3093<div class="doc_subsubsection"> <a name="i_unreachable">'<tt>unreachable</tt>'
3094Instruction</a> </div>
3095
3096<div class="doc_text">
3097
3098<h5>Syntax:</h5>
3099<pre>
3100 unreachable
3101</pre>
3102
3103<h5>Overview:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3104<p>The '<tt>unreachable</tt>' instruction has no defined semantics. This
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3105 instruction is used to inform the optimizer that a particular portion of the
3106 code is not reachable. This can be used to indicate that the code after a
3107 no-return function cannot be reached, and other facts.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3108
3109<h5>Semantics:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3110<p>The '<tt>unreachable</tt>' instruction has no defined semantics.</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3111
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3112</div>
3113
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3114<!-- ======================================================================= -->
3115<div class="doc_subsection"> <a name="binaryops">Binary Operations</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3116
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3117<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3118
3119<p>Binary operators are used to do most of the computation in a program. They
3120 require two operands of the same type, execute an operation on them, and
3121 produce a single value. The operands might represent multiple data, as is
3122 the case with the <a href="#t_vector">vector</a> data type. The result value
3123 has the same type as its operands.</p>
3124
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3125<p>There are several different binary operators:</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3126
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3127</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3128
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3129<!-- _______________________________________________________________________ -->
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3130<div class="doc_subsubsection">
3131 <a name="i_add">'<tt>add</tt>' Instruction</a>
3132</div>
3133
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3134<div class="doc_text">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3135
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3136<h5>Syntax:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3137<pre>
Dan Gohmaned5fcb22009-07-20 22:41:19 +0000[diff] [blame]3138 &lt;result&gt; = add &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanb3b03482009-09-02 17:31:42 +0000[diff] [blame]3139 &lt;result&gt; = add nuw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3140 &lt;result&gt; = add nsw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3141 &lt;result&gt; = add nuw nsw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3142</pre>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3143
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3144<h5>Overview:</h5>
3145<p>The '<tt>add</tt>' instruction returns the sum of its two operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3146
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3147<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3148<p>The two arguments to the '<tt>add</tt>' instruction must
3149 be <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of
3150 integer values. Both arguments must have identical types.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3151
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3152<h5>Semantics:</h5>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]3153<p>The value produced is the integer sum of the two operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3154
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3155<p>If the sum has unsigned overflow, the result returned is the mathematical
3156 result modulo 2<sup>n</sup>, where n is the bit width of the result.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3157
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3158<p>Because LLVM integers use a two's complement representation, this instruction
3159 is appropriate for both signed and unsigned integers.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3160
Dan Gohman46e96012009-07-22 22:44:56 +0000[diff] [blame]3161<p><tt>nuw</tt> and <tt>nsw</tt> stand for &quot;No Unsigned Wrap&quot;
3162 and &quot;No Signed Wrap&quot;, respectively. If the <tt>nuw</tt> and/or
3163 <tt>nsw</tt> keywords are present, the result value of the <tt>add</tt>
Dan Gohmanb8ddf022010-04-22 23:14:21 +0000[diff] [blame]3164 is a <a href="#trapvalues">trap value</a> if unsigned and/or signed overflow,
3165 respectively, occurs.</p>
Dan Gohmaned5fcb22009-07-20 22:41:19 +0000[diff] [blame]3166
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3167<h5>Example:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3168<pre>
3169 &lt;result&gt; = add i32 4, %var <i>; yields {i32}:result = 4 + %var</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3170</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3171
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3172</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3173
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3174<!-- _______________________________________________________________________ -->
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3175<div class="doc_subsubsection">
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]3176 <a name="i_fadd">'<tt>fadd</tt>' Instruction</a>
3177</div>
3178
3179<div class="doc_text">
3180
3181<h5>Syntax:</h5>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]3182<pre>
3183 &lt;result&gt; = fadd &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3184</pre>
3185
3186<h5>Overview:</h5>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]3187<p>The '<tt>fadd</tt>' instruction returns the sum of its two operands.</p>
3188
3189<h5>Arguments:</h5>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]3190<p>The two arguments to the '<tt>fadd</tt>' instruction must be
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3191 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
3192 floating point values. Both arguments must have identical types.</p>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]3193
3194<h5>Semantics:</h5>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]3195<p>The value produced is the floating point sum of the two operands.</p>
3196
3197<h5>Example:</h5>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]3198<pre>
3199 &lt;result&gt; = fadd float 4.0, %var <i>; yields {float}:result = 4.0 + %var</i>
3200</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3201
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]3202</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3203
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]3204<!-- _______________________________________________________________________ -->
3205<div class="doc_subsubsection">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3206 <a name="i_sub">'<tt>sub</tt>' Instruction</a>
3207</div>
3208
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3209<div class="doc_text">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3210
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3211<h5>Syntax:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3212<pre>
Dan Gohman46e96012009-07-22 22:44:56 +0000[diff] [blame]3213 &lt;result&gt; = sub &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanb3b03482009-09-02 17:31:42 +0000[diff] [blame]3214 &lt;result&gt; = sub nuw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3215 &lt;result&gt; = sub nsw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3216 &lt;result&gt; = sub nuw nsw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3217</pre>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3218
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3219<h5>Overview:</h5>
3220<p>The '<tt>sub</tt>' instruction returns the difference of its two
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3221 operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3222
3223<p>Note that the '<tt>sub</tt>' instruction is used to represent the
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3224 '<tt>neg</tt>' instruction present in most other intermediate
3225 representations.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3226
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3227<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3228<p>The two arguments to the '<tt>sub</tt>' instruction must
3229 be <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of
3230 integer values. Both arguments must have identical types.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3231
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3232<h5>Semantics:</h5>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]3233<p>The value produced is the integer difference of the two operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3234
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]3235<p>If the difference has unsigned overflow, the result returned is the
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3236 mathematical result modulo 2<sup>n</sup>, where n is the bit width of the
3237 result.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3238
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3239<p>Because LLVM integers use a two's complement representation, this instruction
3240 is appropriate for both signed and unsigned integers.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3241
Dan Gohman46e96012009-07-22 22:44:56 +0000[diff] [blame]3242<p><tt>nuw</tt> and <tt>nsw</tt> stand for &quot;No Unsigned Wrap&quot;
3243 and &quot;No Signed Wrap&quot;, respectively. If the <tt>nuw</tt> and/or
3244 <tt>nsw</tt> keywords are present, the result value of the <tt>sub</tt>
Dan Gohmanb8ddf022010-04-22 23:14:21 +0000[diff] [blame]3245 is a <a href="#trapvalues">trap value</a> if unsigned and/or signed overflow,
3246 respectively, occurs.</p>
Dan Gohmaned5fcb22009-07-20 22:41:19 +0000[diff] [blame]3247
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3248<h5>Example:</h5>
3249<pre>
3250 &lt;result&gt; = sub i32 4, %var <i>; yields {i32}:result = 4 - %var</i>
3251 &lt;result&gt; = sub i32 0, %val <i>; yields {i32}:result = -%var</i>
3252</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3253
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3254</div>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3255
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3256<!-- _______________________________________________________________________ -->
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3257<div class="doc_subsubsection">
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]3258 <a name="i_fsub">'<tt>fsub</tt>' Instruction</a>
3259</div>
3260
3261<div class="doc_text">
3262
3263<h5>Syntax:</h5>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]3264<pre>
3265 &lt;result&gt; = fsub &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3266</pre>
3267
3268<h5>Overview:</h5>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]3269<p>The '<tt>fsub</tt>' instruction returns the difference of its two
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3270 operands.</p>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]3271
3272<p>Note that the '<tt>fsub</tt>' instruction is used to represent the
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3273 '<tt>fneg</tt>' instruction present in most other intermediate
3274 representations.</p>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]3275
3276<h5>Arguments:</h5>
Bill Wendling1a2630a2009-07-20 02:32:41 +0000[diff] [blame]3277<p>The two arguments to the '<tt>fsub</tt>' instruction must be
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3278 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
3279 floating point values. Both arguments must have identical types.</p>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]3280
3281<h5>Semantics:</h5>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]3282<p>The value produced is the floating point difference of the two operands.</p>
3283
3284<h5>Example:</h5>
3285<pre>
3286 &lt;result&gt; = fsub float 4.0, %var <i>; yields {float}:result = 4.0 - %var</i>
3287 &lt;result&gt; = fsub float -0.0, %val <i>; yields {float}:result = -%var</i>
3288</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3289
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]3290</div>
3291
3292<!-- _______________________________________________________________________ -->
3293<div class="doc_subsubsection">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3294 <a name="i_mul">'<tt>mul</tt>' Instruction</a>
3295</div>
3296
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3297<div class="doc_text">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3298
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3299<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3300<pre>
Dan Gohman46e96012009-07-22 22:44:56 +0000[diff] [blame]3301 &lt;result&gt; = mul &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanb3b03482009-09-02 17:31:42 +0000[diff] [blame]3302 &lt;result&gt; = mul nuw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3303 &lt;result&gt; = mul nsw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3304 &lt;result&gt; = mul nuw nsw &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3305</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3306
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3307<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3308<p>The '<tt>mul</tt>' instruction returns the product of its two operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3309
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3310<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3311<p>The two arguments to the '<tt>mul</tt>' instruction must
3312 be <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of
3313 integer values. Both arguments must have identical types.</p>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]3314
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3315<h5>Semantics:</h5>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]3316<p>The value produced is the integer product of the two operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3317
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3318<p>If the result of the multiplication has unsigned overflow, the result
3319 returned is the mathematical result modulo 2<sup>n</sup>, where n is the bit
3320 width of the result.</p>
3321
3322<p>Because LLVM integers use a two's complement representation, and the result
3323 is the same width as the operands, this instruction returns the correct
3324 result for both signed and unsigned integers. If a full product
3325 (e.g. <tt>i32</tt>x<tt>i32</tt>-><tt>i64</tt>) is needed, the operands should
3326 be sign-extended or zero-extended as appropriate to the width of the full
3327 product.</p>
3328
Dan Gohman46e96012009-07-22 22:44:56 +0000[diff] [blame]3329<p><tt>nuw</tt> and <tt>nsw</tt> stand for &quot;No Unsigned Wrap&quot;
3330 and &quot;No Signed Wrap&quot;, respectively. If the <tt>nuw</tt> and/or
3331 <tt>nsw</tt> keywords are present, the result value of the <tt>mul</tt>
Dan Gohmanb8ddf022010-04-22 23:14:21 +0000[diff] [blame]3332 is a <a href="#trapvalues">trap value</a> if unsigned and/or signed overflow,
3333 respectively, occurs.</p>
Dan Gohmaned5fcb22009-07-20 22:41:19 +0000[diff] [blame]3334
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3335<h5>Example:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3336<pre>
3337 &lt;result&gt; = mul i32 4, %var <i>; yields {i32}:result = 4 * %var</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3338</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3339
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3340</div>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3341
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3342<!-- _______________________________________________________________________ -->
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]3343<div class="doc_subsubsection">
3344 <a name="i_fmul">'<tt>fmul</tt>' Instruction</a>
3345</div>
3346
3347<div class="doc_text">
3348
3349<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3350<pre>
3351 &lt;result&gt; = fmul &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]3352</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3353
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]3354<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3355<p>The '<tt>fmul</tt>' instruction returns the product of its two operands.</p>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]3356
3357<h5>Arguments:</h5>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]3358<p>The two arguments to the '<tt>fmul</tt>' instruction must be
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3359 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
3360 floating point values. Both arguments must have identical types.</p>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]3361
3362<h5>Semantics:</h5>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]3363<p>The value produced is the floating point product of the two operands.</p>
3364
3365<h5>Example:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3366<pre>
3367 &lt;result&gt; = fmul float 4.0, %var <i>; yields {float}:result = 4.0 * %var</i>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]3368</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3369
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]3370</div>
3371
3372<!-- _______________________________________________________________________ -->
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3373<div class="doc_subsubsection"> <a name="i_udiv">'<tt>udiv</tt>' Instruction
3374</a></div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3375
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3376<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3377
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3378<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3379<pre>
3380 &lt;result&gt; = udiv &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3381</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3382
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3383<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3384<p>The '<tt>udiv</tt>' instruction returns the quotient of its two operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3385
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3386<h5>Arguments:</h5>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]3387<p>The two arguments to the '<tt>udiv</tt>' instruction must be
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3388 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3389 values. Both arguments must have identical types.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3390
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3391<h5>Semantics:</h5>
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]3392<p>The value produced is the unsigned integer quotient of the two operands.</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3393
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]3394<p>Note that unsigned integer division and signed integer division are distinct
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3395 operations; for signed integer division, use '<tt>sdiv</tt>'.</p>
3396
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]3397<p>Division by zero leads to undefined behavior.</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3398
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3399<h5>Example:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3400<pre>
3401 &lt;result&gt; = udiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3402</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3403
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3404</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3405
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3406<!-- _______________________________________________________________________ -->
3407<div class="doc_subsubsection"> <a name="i_sdiv">'<tt>sdiv</tt>' Instruction
3408</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3409
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3410<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3411
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3412<h5>Syntax:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3413<pre>
Dan Gohmaned5fcb22009-07-20 22:41:19 +0000[diff] [blame]3414 &lt;result&gt; = sdiv &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanb3b03482009-09-02 17:31:42 +0000[diff] [blame]3415 &lt;result&gt; = sdiv exact &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3416</pre>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3417
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3418<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3419<p>The '<tt>sdiv</tt>' instruction returns the quotient of its two operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3420
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3421<h5>Arguments:</h5>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]3422<p>The two arguments to the '<tt>sdiv</tt>' instruction must be
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3423 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3424 values. Both arguments must have identical types.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3425
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3426<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3427<p>The value produced is the signed integer quotient of the two operands rounded
3428 towards zero.</p>
3429
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]3430<p>Note that signed integer division and unsigned integer division are distinct
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3431 operations; for unsigned integer division, use '<tt>udiv</tt>'.</p>
3432
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]3433<p>Division by zero leads to undefined behavior. Overflow also leads to
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3434 undefined behavior; this is a rare case, but can occur, for example, by doing
3435 a 32-bit division of -2147483648 by -1.</p>
3436
Dan Gohman67fa48e2009-07-22 00:04:19 +0000[diff] [blame]3437<p>If the <tt>exact</tt> keyword is present, the result value of the
Dan Gohmanf7263862010-04-23 15:23:32 +0000[diff] [blame]3438 <tt>sdiv</tt> is a <a href="#trapvalues">trap value</a> if the result would
3439 be rounded or if overflow would occur.</p>
Dan Gohmaned5fcb22009-07-20 22:41:19 +0000[diff] [blame]3440
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3441<h5>Example:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3442<pre>
3443 &lt;result&gt; = sdiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3444</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3445
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3446</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3447
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3448<!-- _______________________________________________________________________ -->
3449<div class="doc_subsubsection"> <a name="i_fdiv">'<tt>fdiv</tt>'
3450Instruction</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3451
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3452<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3453
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3454<h5>Syntax:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3455<pre>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]3456 &lt;result&gt; = fdiv &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3457</pre>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3458
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3459<h5>Overview:</h5>
3460<p>The '<tt>fdiv</tt>' instruction returns the quotient of its two operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3461
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3462<h5>Arguments:</h5>
3463<p>The two arguments to the '<tt>fdiv</tt>' instruction must be
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3464 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
3465 floating point values. Both arguments must have identical types.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3466
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3467<h5>Semantics:</h5>
3468<p>The value produced is the floating point quotient of the two operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3469
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3470<h5>Example:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3471<pre>
3472 &lt;result&gt; = fdiv float 4.0, %var <i>; yields {float}:result = 4.0 / %var</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3473</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3474
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3475</div>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3476
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3477<!-- _______________________________________________________________________ -->
3478<div class="doc_subsubsection"> <a name="i_urem">'<tt>urem</tt>' Instruction</a>
3479</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3480
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3481<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3482
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3483<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3484<pre>
3485 &lt;result&gt; = urem &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3486</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3487
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3488<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3489<p>The '<tt>urem</tt>' instruction returns the remainder from the unsigned
3490 division of its two arguments.</p>
3491
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3492<h5>Arguments:</h5>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]3493<p>The two arguments to the '<tt>urem</tt>' instruction must be
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3494 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3495 values. Both arguments must have identical types.</p>
3496
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3497<h5>Semantics:</h5>
3498<p>This instruction returns the unsigned integer <i>remainder</i> of a division.
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3499 This instruction always performs an unsigned division to get the
3500 remainder.</p>
3501
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]3502<p>Note that unsigned integer remainder and signed integer remainder are
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3503 distinct operations; for signed integer remainder, use '<tt>srem</tt>'.</p>
3504
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]3505<p>Taking the remainder of a division by zero leads to undefined behavior.</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3506
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3507<h5>Example:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3508<pre>
3509 &lt;result&gt; = urem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3510</pre>
3511
3512</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3513
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3514<!-- _______________________________________________________________________ -->
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3515<div class="doc_subsubsection">
3516 <a name="i_srem">'<tt>srem</tt>' Instruction</a>
3517</div>
3518
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3519<div class="doc_text">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3520
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3521<h5>Syntax:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3522<pre>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]3523 &lt;result&gt; = srem &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3524</pre>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3525
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3526<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3527<p>The '<tt>srem</tt>' instruction returns the remainder from the signed
3528 division of its two operands. This instruction can also take
3529 <a href="#t_vector">vector</a> versions of the values in which case the
3530 elements must be integers.</p>
Chris Lattner08497ce2008-01-04 04:33:49 +0000[diff] [blame]3531
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3532<h5>Arguments:</h5>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]3533<p>The two arguments to the '<tt>srem</tt>' instruction must be
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3534 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3535 values. Both arguments must have identical types.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3536
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3537<h5>Semantics:</h5>
3538<p>This instruction returns the <i>remainder</i> of a division (where the result
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3539 has the same sign as the dividend, <tt>op1</tt>), not the <i>modulo</i>
3540 operator (where the result has the same sign as the divisor, <tt>op2</tt>) of
3541 a value. For more information about the difference,
3542 see <a href="http://mathforum.org/dr.math/problems/anne.4.28.99.html">The
3543 Math Forum</a>. For a table of how this is implemented in various languages,
3544 please see <a href="http://en.wikipedia.org/wiki/Modulo_operation">
3545 Wikipedia: modulo operation</a>.</p>
3546
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]3547<p>Note that signed integer remainder and unsigned integer remainder are
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3548 distinct operations; for unsigned integer remainder, use '<tt>urem</tt>'.</p>
3549
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]3550<p>Taking the remainder of a division by zero leads to undefined behavior.
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3551 Overflow also leads to undefined behavior; this is a rare case, but can
3552 occur, for example, by taking the remainder of a 32-bit division of
3553 -2147483648 by -1. (The remainder doesn't actually overflow, but this rule
3554 lets srem be implemented using instructions that return both the result of
3555 the division and the remainder.)</p>
3556
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3557<h5>Example:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3558<pre>
3559 &lt;result&gt; = srem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3560</pre>
3561
3562</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3563
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3564<!-- _______________________________________________________________________ -->
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3565<div class="doc_subsubsection">
3566 <a name="i_frem">'<tt>frem</tt>' Instruction</a> </div>
3567
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3568<div class="doc_text">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3569
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3570<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3571<pre>
3572 &lt;result&gt; = frem &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3573</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3574
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3575<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3576<p>The '<tt>frem</tt>' instruction returns the remainder from the division of
3577 its two operands.</p>
3578
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3579<h5>Arguments:</h5>
3580<p>The two arguments to the '<tt>frem</tt>' instruction must be
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3581 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
3582 floating point values. Both arguments must have identical types.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3583
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3584<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3585<p>This instruction returns the <i>remainder</i> of a division. The remainder
3586 has the same sign as the dividend.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3587
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3588<h5>Example:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3589<pre>
3590 &lt;result&gt; = frem float 4.0, %var <i>; yields {float}:result = 4.0 % %var</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3591</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3592
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3593</div>
3594
3595<!-- ======================================================================= -->
3596<div class="doc_subsection"> <a name="bitwiseops">Bitwise Binary
3597Operations</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3598
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3599<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3600
3601<p>Bitwise binary operators are used to do various forms of bit-twiddling in a
3602 program. They are generally very efficient instructions and can commonly be
3603 strength reduced from other instructions. They require two operands of the
3604 same type, execute an operation on them, and produce a single value. The
3605 resulting value is the same type as its operands.</p>
3606
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3607</div>
3608
3609<!-- _______________________________________________________________________ -->
3610<div class="doc_subsubsection"> <a name="i_shl">'<tt>shl</tt>'
3611Instruction</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3612
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3613<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3614
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3615<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3616<pre>
3617 &lt;result&gt; = shl &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3618</pre>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]3619
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3620<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3621<p>The '<tt>shl</tt>' instruction returns the first operand shifted to the left
3622 a specified number of bits.</p>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]3623
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3624<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3625<p>Both arguments to the '<tt>shl</tt>' instruction must be the
3626 same <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of
3627 integer type. '<tt>op2</tt>' is treated as an unsigned value.</p>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]3628
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3629<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3630<p>The value produced is <tt>op1</tt> * 2<sup><tt>op2</tt></sup> mod
3631 2<sup>n</sup>, where <tt>n</tt> is the width of the result. If <tt>op2</tt>
3632 is (statically or dynamically) negative or equal to or larger than the number
3633 of bits in <tt>op1</tt>, the result is undefined. If the arguments are
3634 vectors, each vector element of <tt>op1</tt> is shifted by the corresponding
3635 shift amount in <tt>op2</tt>.</p>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]3636
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3637<h5>Example:</h5>
3638<pre>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3639 &lt;result&gt; = shl i32 4, %var <i>; yields {i32}: 4 &lt;&lt; %var</i>
3640 &lt;result&gt; = shl i32 4, 2 <i>; yields {i32}: 16</i>
3641 &lt;result&gt; = shl i32 1, 10 <i>; yields {i32}: 1024</i>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]3642 &lt;result&gt; = shl i32 1, 32 <i>; undefined</i>
Mon P Wang9901e732008-12-09 05:46:39 +0000[diff] [blame]3643 &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>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3644</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3645
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3646</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3647
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3648<!-- _______________________________________________________________________ -->
3649<div class="doc_subsubsection"> <a name="i_lshr">'<tt>lshr</tt>'
3650Instruction</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3651
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3652<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3653
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3654<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3655<pre>
3656 &lt;result&gt; = lshr &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3657</pre>
3658
3659<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3660<p>The '<tt>lshr</tt>' instruction (logical shift right) returns the first
3661 operand shifted to the right a specified number of bits with zero fill.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3662
3663<h5>Arguments:</h5>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]3664<p>Both arguments to the '<tt>lshr</tt>' instruction must be the same
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3665 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3666 type. '<tt>op2</tt>' is treated as an unsigned value.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3667
3668<h5>Semantics:</h5>
3669<p>This instruction always performs a logical shift right operation. The most
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3670 significant bits of the result will be filled with zero bits after the shift.
3671 If <tt>op2</tt> is (statically or dynamically) equal to or larger than the
3672 number of bits in <tt>op1</tt>, the result is undefined. If the arguments are
3673 vectors, each vector element of <tt>op1</tt> is shifted by the corresponding
3674 shift amount in <tt>op2</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3675
3676<h5>Example:</h5>
3677<pre>
3678 &lt;result&gt; = lshr i32 4, 1 <i>; yields {i32}:result = 2</i>
3679 &lt;result&gt; = lshr i32 4, 2 <i>; yields {i32}:result = 1</i>
3680 &lt;result&gt; = lshr i8 4, 3 <i>; yields {i8}:result = 0</i>
3681 &lt;result&gt; = lshr i8 -2, 1 <i>; yields {i8}:result = 0x7FFFFFFF </i>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]3682 &lt;result&gt; = lshr i32 1, 32 <i>; undefined</i>
Mon P Wang9901e732008-12-09 05:46:39 +0000[diff] [blame]3683 &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>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3684</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3685
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3686</div>
3687
3688<!-- _______________________________________________________________________ -->
3689<div class="doc_subsubsection"> <a name="i_ashr">'<tt>ashr</tt>'
3690Instruction</a> </div>
3691<div class="doc_text">
3692
3693<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3694<pre>
3695 &lt;result&gt; = ashr &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3696</pre>
3697
3698<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3699<p>The '<tt>ashr</tt>' instruction (arithmetic shift right) returns the first
3700 operand shifted to the right a specified number of bits with sign
3701 extension.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3702
3703<h5>Arguments:</h5>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]3704<p>Both arguments to the '<tt>ashr</tt>' instruction must be the same
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3705 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3706 type. '<tt>op2</tt>' is treated as an unsigned value.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3707
3708<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3709<p>This instruction always performs an arithmetic shift right operation, The
3710 most significant bits of the result will be filled with the sign bit
3711 of <tt>op1</tt>. If <tt>op2</tt> is (statically or dynamically) equal to or
3712 larger than the number of bits in <tt>op1</tt>, the result is undefined. If
3713 the arguments are vectors, each vector element of <tt>op1</tt> is shifted by
3714 the corresponding shift amount in <tt>op2</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3715
3716<h5>Example:</h5>
3717<pre>
3718 &lt;result&gt; = ashr i32 4, 1 <i>; yields {i32}:result = 2</i>
3719 &lt;result&gt; = ashr i32 4, 2 <i>; yields {i32}:result = 1</i>
3720 &lt;result&gt; = ashr i8 4, 3 <i>; yields {i8}:result = 0</i>
3721 &lt;result&gt; = ashr i8 -2, 1 <i>; yields {i8}:result = -1</i>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]3722 &lt;result&gt; = ashr i32 1, 32 <i>; undefined</i>
Mon P Wang9901e732008-12-09 05:46:39 +0000[diff] [blame]3723 &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>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3724</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3725
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3726</div>
3727
3728<!-- _______________________________________________________________________ -->
3729<div class="doc_subsubsection"> <a name="i_and">'<tt>and</tt>'
3730Instruction</a> </div>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3731
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3732<div class="doc_text">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3733
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3734<h5>Syntax:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3735<pre>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]3736 &lt;result&gt; = and &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3737</pre>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3738
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3739<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3740<p>The '<tt>and</tt>' instruction returns the bitwise logical and of its two
3741 operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3742
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3743<h5>Arguments:</h5>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]3744<p>The two arguments to the '<tt>and</tt>' instruction must be
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3745 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3746 values. Both arguments must have identical types.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3747
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3748<h5>Semantics:</h5>
3749<p>The truth table used for the '<tt>and</tt>' instruction is:</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3750
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3751<table border="1" cellspacing="0" cellpadding="4">
3752 <tbody>
3753 <tr>
3754 <td>In0</td>
3755 <td>In1</td>
3756 <td>Out</td>
3757 </tr>
3758 <tr>
3759 <td>0</td>
3760 <td>0</td>
3761 <td>0</td>
3762 </tr>
3763 <tr>
3764 <td>0</td>
3765 <td>1</td>
3766 <td>0</td>
3767 </tr>
3768 <tr>
3769 <td>1</td>
3770 <td>0</td>
3771 <td>0</td>
3772 </tr>
3773 <tr>
3774 <td>1</td>
3775 <td>1</td>
3776 <td>1</td>
3777 </tr>
3778 </tbody>
3779</table>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3780
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3781<h5>Example:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3782<pre>
3783 &lt;result&gt; = and i32 4, %var <i>; yields {i32}:result = 4 &amp; %var</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3784 &lt;result&gt; = and i32 15, 40 <i>; yields {i32}:result = 8</i>
3785 &lt;result&gt; = and i32 4, 8 <i>; yields {i32}:result = 0</i>
3786</pre>
3787</div>
3788<!-- _______________________________________________________________________ -->
3789<div class="doc_subsubsection"> <a name="i_or">'<tt>or</tt>' Instruction</a> </div>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3790
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3791<div class="doc_text">
3792
3793<h5>Syntax:</h5>
3794<pre>
3795 &lt;result&gt; = or &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3796</pre>
3797
3798<h5>Overview:</h5>
3799<p>The '<tt>or</tt>' instruction returns the bitwise logical inclusive or of its
3800 two operands.</p>
3801
3802<h5>Arguments:</h5>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]3803<p>The two arguments to the '<tt>or</tt>' instruction must be
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3804 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3805 values. Both arguments must have identical types.</p>
3806
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3807<h5>Semantics:</h5>
3808<p>The truth table used for the '<tt>or</tt>' instruction is:</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3809
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3810<table border="1" cellspacing="0" cellpadding="4">
3811 <tbody>
3812 <tr>
3813 <td>In0</td>
3814 <td>In1</td>
3815 <td>Out</td>
3816 </tr>
3817 <tr>
3818 <td>0</td>
3819 <td>0</td>
3820 <td>0</td>
3821 </tr>
3822 <tr>
3823 <td>0</td>
3824 <td>1</td>
3825 <td>1</td>
3826 </tr>
3827 <tr>
3828 <td>1</td>
3829 <td>0</td>
3830 <td>1</td>
3831 </tr>
3832 <tr>
3833 <td>1</td>
3834 <td>1</td>
3835 <td>1</td>
3836 </tr>
3837 </tbody>
3838</table>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3839
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3840<h5>Example:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3841<pre>
3842 &lt;result&gt; = or i32 4, %var <i>; yields {i32}:result = 4 | %var</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3843 &lt;result&gt; = or i32 15, 40 <i>; yields {i32}:result = 47</i>
3844 &lt;result&gt; = or i32 4, 8 <i>; yields {i32}:result = 12</i>
3845</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3846
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3847</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3848
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3849<!-- _______________________________________________________________________ -->
3850<div class="doc_subsubsection"> <a name="i_xor">'<tt>xor</tt>'
3851Instruction</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3852
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3853<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3854
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3855<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3856<pre>
3857 &lt;result&gt; = xor &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3858</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3859
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3860<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3861<p>The '<tt>xor</tt>' instruction returns the bitwise logical exclusive or of
3862 its two operands. The <tt>xor</tt> is used to implement the "one's
3863 complement" operation, which is the "~" operator in C.</p>
3864
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3865<h5>Arguments:</h5>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]3866<p>The two arguments to the '<tt>xor</tt>' instruction must be
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3867 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3868 values. Both arguments must have identical types.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3869
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3870<h5>Semantics:</h5>
3871<p>The truth table used for the '<tt>xor</tt>' instruction is:</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3872
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3873<table border="1" cellspacing="0" cellpadding="4">
3874 <tbody>
3875 <tr>
3876 <td>In0</td>
3877 <td>In1</td>
3878 <td>Out</td>
3879 </tr>
3880 <tr>
3881 <td>0</td>
3882 <td>0</td>
3883 <td>0</td>
3884 </tr>
3885 <tr>
3886 <td>0</td>
3887 <td>1</td>
3888 <td>1</td>
3889 </tr>
3890 <tr>
3891 <td>1</td>
3892 <td>0</td>
3893 <td>1</td>
3894 </tr>
3895 <tr>
3896 <td>1</td>
3897 <td>1</td>
3898 <td>0</td>
3899 </tr>
3900 </tbody>
3901</table>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3902
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3903<h5>Example:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3904<pre>
3905 &lt;result&gt; = xor i32 4, %var <i>; yields {i32}:result = 4 ^ %var</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3906 &lt;result&gt; = xor i32 15, 40 <i>; yields {i32}:result = 39</i>
3907 &lt;result&gt; = xor i32 4, 8 <i>; yields {i32}:result = 12</i>
3908 &lt;result&gt; = xor i32 %V, -1 <i>; yields {i32}:result = ~%V</i>
3909</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3910
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3911</div>
3912
3913<!-- ======================================================================= -->
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]3914<div class="doc_subsection">
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3915 <a name="vectorops">Vector Operations</a>
3916</div>
3917
3918<div class="doc_text">
3919
3920<p>LLVM supports several instructions to represent vector operations in a
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3921 target-independent manner. These instructions cover the element-access and
3922 vector-specific operations needed to process vectors effectively. While LLVM
3923 does directly support these vector operations, many sophisticated algorithms
3924 will want to use target-specific intrinsics to take full advantage of a
3925 specific target.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3926
3927</div>
3928
3929<!-- _______________________________________________________________________ -->
3930<div class="doc_subsubsection">
3931 <a name="i_extractelement">'<tt>extractelement</tt>' Instruction</a>
3932</div>
3933
3934<div class="doc_text">
3935
3936<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3937<pre>
3938 &lt;result&gt; = extractelement &lt;n x &lt;ty&gt;&gt; &lt;val&gt;, i32 &lt;idx&gt; <i>; yields &lt;ty&gt;</i>
3939</pre>
3940
3941<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3942<p>The '<tt>extractelement</tt>' instruction extracts a single scalar element
3943 from a vector at a specified index.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3944
3945
3946<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3947<p>The first operand of an '<tt>extractelement</tt>' instruction is a value
3948 of <a href="#t_vector">vector</a> type. The second operand is an index
3949 indicating the position from which to extract the element. The index may be
3950 a variable.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3951
3952<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3953<p>The result is a scalar of the same type as the element type of
3954 <tt>val</tt>. Its value is the value at position <tt>idx</tt> of
3955 <tt>val</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
3956 results are undefined.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3957
3958<h5>Example:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3959<pre>
Gabor Greifb2c73142009-10-28 13:14:50 +0000[diff] [blame]3960 &lt;result&gt; = extractelement &lt;4 x i32&gt; %vec, i32 0 <i>; yields i32</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3961</pre>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3962
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3963</div>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3964
3965<!-- _______________________________________________________________________ -->
3966<div class="doc_subsubsection">
3967 <a name="i_insertelement">'<tt>insertelement</tt>' Instruction</a>
3968</div>
3969
3970<div class="doc_text">
3971
3972<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3973<pre>
Dan Gohmanbcc3c502008-05-12 23:38:42 +0000[diff] [blame]3974 &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>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3975</pre>
3976
3977<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3978<p>The '<tt>insertelement</tt>' instruction inserts a scalar element into a
3979 vector at a specified index.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3980
3981<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3982<p>The first operand of an '<tt>insertelement</tt>' instruction is a value
3983 of <a href="#t_vector">vector</a> type. The second operand is a scalar value
3984 whose type must equal the element type of the first operand. The third
3985 operand is an index indicating the position at which to insert the value.
3986 The index may be a variable.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3987
3988<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3989<p>The result is a vector of the same type as <tt>val</tt>. Its element values
3990 are those of <tt>val</tt> except at position <tt>idx</tt>, where it gets the
3991 value <tt>elt</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
3992 results are undefined.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3993
3994<h5>Example:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3995<pre>
Gabor Greifb2c73142009-10-28 13:14:50 +0000[diff] [blame]3996 &lt;result&gt; = insertelement &lt;4 x i32&gt; %vec, i32 1, i32 0 <i>; yields &lt;4 x i32&gt;</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3997</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3998
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3999</div>
4000
4001<!-- _______________________________________________________________________ -->
4002<div class="doc_subsubsection">
4003 <a name="i_shufflevector">'<tt>shufflevector</tt>' Instruction</a>
4004</div>
4005
4006<div class="doc_text">
4007
4008<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4009<pre>
Mon P Wangbff5d9c2008-11-10 04:46:22 +0000[diff] [blame]4010 &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>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4011</pre>
4012
4013<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4014<p>The '<tt>shufflevector</tt>' instruction constructs a permutation of elements
4015 from two input vectors, returning a vector with the same element type as the
4016 input and length that is the same as the shuffle mask.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4017
4018<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4019<p>The first two operands of a '<tt>shufflevector</tt>' instruction are vectors
4020 with types that match each other. The third argument is a shuffle mask whose
4021 element type is always 'i32'. The result of the instruction is a vector
4022 whose length is the same as the shuffle mask and whose element type is the
4023 same as the element type of the first two operands.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4024
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4025<p>The shuffle mask operand is required to be a constant vector with either
4026 constant integer or undef values.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4027
4028<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4029<p>The elements of the two input vectors are numbered from left to right across
4030 both of the vectors. The shuffle mask operand specifies, for each element of
4031 the result vector, which element of the two input vectors the result element
4032 gets. The element selector may be undef (meaning "don't care") and the
4033 second operand may be undef if performing a shuffle from only one vector.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4034
4035<h5>Example:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4036<pre>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]4037 &lt;result&gt; = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; %v2,
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4038 &lt;4 x i32&gt; &lt;i32 0, i32 4, i32 1, i32 5&gt; <i>; yields &lt;4 x i32&gt;</i>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]4039 &lt;result&gt; = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; undef,
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4040 &lt;4 x i32&gt; &lt;i32 0, i32 1, i32 2, i32 3&gt; <i>; yields &lt;4 x i32&gt;</i> - Identity shuffle.
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]4041 &lt;result&gt; = shufflevector &lt;8 x i32&gt; %v1, &lt;8 x i32&gt; undef,
Mon P Wangbff5d9c2008-11-10 04:46:22 +0000[diff] [blame]4042 &lt;4 x i32&gt; &lt;i32 0, i32 1, i32 2, i32 3&gt; <i>; yields &lt;4 x i32&gt;</i>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]4043 &lt;result&gt; = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; %v2,
Mon P Wangbff5d9c2008-11-10 04:46:22 +0000[diff] [blame]4044 &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>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4045</pre>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4046
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4047</div>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4048
4049<!-- ======================================================================= -->
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]4050<div class="doc_subsection">
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]4051 <a name="aggregateops">Aggregate Operations</a>
4052</div>
4053
4054<div class="doc_text">
4055
Chris Lattnerd5d51722010-02-12 20:49:41 +0000[diff] [blame]4056<p>LLVM supports several instructions for working with
4057 <a href="#t_aggregate">aggregate</a> values.</p>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]4058
4059</div>
4060
4061<!-- _______________________________________________________________________ -->
4062<div class="doc_subsubsection">
4063 <a name="i_extractvalue">'<tt>extractvalue</tt>' Instruction</a>
4064</div>
4065
4066<div class="doc_text">
4067
4068<h5>Syntax:</h5>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]4069<pre>
4070 &lt;result&gt; = extractvalue &lt;aggregate type&gt; &lt;val&gt;, &lt;idx&gt;{, &lt;idx&gt;}*
4071</pre>
4072
4073<h5>Overview:</h5>
Chris Lattnerd5d51722010-02-12 20:49:41 +0000[diff] [blame]4074<p>The '<tt>extractvalue</tt>' instruction extracts the value of a member field
4075 from an <a href="#t_aggregate">aggregate</a> value.</p>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]4076
4077<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4078<p>The first operand of an '<tt>extractvalue</tt>' instruction is a value
Chris Lattnerd5d51722010-02-12 20:49:41 +0000[diff] [blame]4079 of <a href="#t_struct">struct</a>, <a href="#t_union">union</a> or
4080 <a href="#t_array">array</a> type. The operands are constant indices to
4081 specify which value to extract in a similar manner as indices in a
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4082 '<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.</p>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]4083
4084<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4085<p>The result is the value at the position in the aggregate specified by the
4086 index operands.</p>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]4087
4088<h5>Example:</h5>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]4089<pre>
Gabor Greifb2c73142009-10-28 13:14:50 +0000[diff] [blame]4090 &lt;result&gt; = extractvalue {i32, float} %agg, 0 <i>; yields i32</i>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]4091</pre>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]4092
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4093</div>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]4094
4095<!-- _______________________________________________________________________ -->
4096<div class="doc_subsubsection">
4097 <a name="i_insertvalue">'<tt>insertvalue</tt>' Instruction</a>
4098</div>
4099
4100<div class="doc_text">
4101
4102<h5>Syntax:</h5>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]4103<pre>
Jeffrey Yasskine05d7732010-01-11 19:19:26 +0000[diff] [blame]4104 &lt;result&gt; = insertvalue &lt;aggregate type&gt; &lt;val&gt;, &lt;ty&gt; &lt;elt&gt;, &lt;idx&gt; <i>; yields &lt;aggregate type&gt;</i>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]4105</pre>
4106
4107<h5>Overview:</h5>
Chris Lattnerd5d51722010-02-12 20:49:41 +0000[diff] [blame]4108<p>The '<tt>insertvalue</tt>' instruction inserts a value into a member field
4109 in an <a href="#t_aggregate">aggregate</a> value.</p>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]4110
4111<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4112<p>The first operand of an '<tt>insertvalue</tt>' instruction is a value
Chris Lattnerd5d51722010-02-12 20:49:41 +0000[diff] [blame]4113 of <a href="#t_struct">struct</a>, <a href="#t_union">union</a> or
4114 <a href="#t_array">array</a> type. The second operand is a first-class
4115 value to insert. The following operands are constant indices indicating
4116 the position at which to insert the value in a similar manner as indices in a
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4117 '<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction. The
4118 value to insert must have the same type as the value identified by the
4119 indices.</p>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]4120
4121<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4122<p>The result is an aggregate of the same type as <tt>val</tt>. Its value is
4123 that of <tt>val</tt> except that the value at the position specified by the
4124 indices is that of <tt>elt</tt>.</p>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]4125
4126<h5>Example:</h5>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]4127<pre>
Jeffrey Yasskine05d7732010-01-11 19:19:26 +0000[diff] [blame]4128 %agg1 = insertvalue {i32, float} undef, i32 1, 0 <i>; yields {i32 1, float undef}</i>
4129 %agg2 = insertvalue {i32, float} %agg1, float %val, 1 <i>; yields {i32 1, float %val}</i>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]4130</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4131
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]4132</div>
4133
4134
4135<!-- ======================================================================= -->
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]4136<div class="doc_subsection">
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4137 <a name="memoryops">Memory Access and Addressing Operations</a>
4138</div>
4139
4140<div class="doc_text">
4141
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4142<p>A key design point of an SSA-based representation is how it represents
4143 memory. In LLVM, no memory locations are in SSA form, which makes things
Victor Hernandez6b054092009-10-26 23:44:29 +0000[diff] [blame]4144 very simple. This section describes how to read, write, and allocate
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4145 memory in LLVM.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4146
4147</div>
4148
4149<!-- _______________________________________________________________________ -->
4150<div class="doc_subsubsection">
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4151 <a name="i_alloca">'<tt>alloca</tt>' Instruction</a>
4152</div>
4153
4154<div class="doc_text">
4155
4156<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4157<pre>
4158 &lt;result&gt; = alloca &lt;type&gt;[, i32 &lt;NumElements&gt;][, align &lt;alignment&gt;] <i>; yields {type*}:result</i>
4159</pre>
4160
4161<h5>Overview:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4162<p>The '<tt>alloca</tt>' instruction allocates memory on the stack frame of the
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4163 currently executing function, to be automatically released when this function
4164 returns to its caller. The object is always allocated in the generic address
4165 space (address space zero).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4166
4167<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4168<p>The '<tt>alloca</tt>' instruction
4169 allocates <tt>sizeof(&lt;type&gt;)*NumElements</tt> bytes of memory on the
4170 runtime stack, returning a pointer of the appropriate type to the program.
4171 If "NumElements" is specified, it is the number of elements allocated,
4172 otherwise "NumElements" is defaulted to be one. If a constant alignment is
4173 specified, the value result of the allocation is guaranteed to be aligned to
4174 at least that boundary. If not specified, or if zero, the target can choose
4175 to align the allocation on any convenient boundary compatible with the
4176 type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4177
4178<p>'<tt>type</tt>' may be any sized type.</p>
4179
4180<h5>Semantics:</h5>
Bill Wendling2a454572009-05-08 20:49:29 +0000[diff] [blame]4181<p>Memory is allocated; a pointer is returned. The operation is undefined if
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4182 there is insufficient stack space for the allocation. '<tt>alloca</tt>'d
4183 memory is automatically released when the function returns. The
4184 '<tt>alloca</tt>' instruction is commonly used to represent automatic
4185 variables that must have an address available. When the function returns
4186 (either with the <tt><a href="#i_ret">ret</a></tt>
4187 or <tt><a href="#i_unwind">unwind</a></tt> instructions), the memory is
4188 reclaimed. Allocating zero bytes is legal, but the result is undefined.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4189
4190<h5>Example:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4191<pre>
Dan Gohmanf54f50a2009-01-04 23:49:44 +0000[diff] [blame]4192 %ptr = alloca i32 <i>; yields {i32*}:ptr</i>
4193 %ptr = alloca i32, i32 4 <i>; yields {i32*}:ptr</i>
4194 %ptr = alloca i32, i32 4, align 1024 <i>; yields {i32*}:ptr</i>
4195 %ptr = alloca i32, align 1024 <i>; yields {i32*}:ptr</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4196</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4197
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4198</div>
4199
4200<!-- _______________________________________________________________________ -->
4201<div class="doc_subsubsection"> <a name="i_load">'<tt>load</tt>'
4202Instruction</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4203
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4204<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4205
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4206<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4207<pre>
Bill Wendling4197e452010-02-25 21:23:24 +0000[diff] [blame]4208 &lt;result&gt; = load &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;][, !nontemporal !&lt;index&gt;]
4209 &lt;result&gt; = volatile load &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;][, !nontemporal !&lt;index&gt;]
4210 !&lt;index&gt; = !{ i32 1 }
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4211</pre>
4212
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4213<h5>Overview:</h5>
4214<p>The '<tt>load</tt>' instruction is used to read from memory.</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4215
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4216<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4217<p>The argument to the '<tt>load</tt>' instruction specifies the memory address
4218 from which to load. The pointer must point to
4219 a <a href="#t_firstclass">first class</a> type. If the <tt>load</tt> is
4220 marked as <tt>volatile</tt>, then the optimizer is not allowed to modify the
Jeffrey Yasskin6dc467f2010-04-26 21:21:24 +0000[diff] [blame]4221 number or order of execution of this <tt>load</tt> with other <a
4222 href="#volatile">volatile operations</a>.</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4223
Bill Wendling4197e452010-02-25 21:23:24 +0000[diff] [blame]4224<p>The optional constant <tt>align</tt> argument specifies the alignment of the
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4225 operation (that is, the alignment of the memory address). A value of 0 or an
Bill Wendling4197e452010-02-25 21:23:24 +0000[diff] [blame]4226 omitted <tt>align</tt> argument means that the operation has the preferential
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4227 alignment for the target. It is the responsibility of the code emitter to
4228 ensure that the alignment information is correct. Overestimating the
Bill Wendling4197e452010-02-25 21:23:24 +0000[diff] [blame]4229 alignment results in undefined behavior. Underestimating the alignment may
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4230 produce less efficient code. An alignment of 1 is always safe.</p>
4231
Bill Wendling4197e452010-02-25 21:23:24 +0000[diff] [blame]4232<p>The optional <tt>!nontemporal</tt> metadata must reference a single
4233 metatadata name &lt;index&gt; corresponding to a metadata node with
Dan Gohman22dc6682010-03-01 17:41:39 +0000[diff] [blame]4234 one <tt>i32</tt> entry of value 1. The existence of
Bill Wendling4197e452010-02-25 21:23:24 +0000[diff] [blame]4235 the <tt>!nontemporal</tt> metatadata on the instruction tells the optimizer
4236 and code generator that this load is not expected to be reused in the cache.
4237 The code generator may select special instructions to save cache bandwidth,
Dan Gohman22dc6682010-03-01 17:41:39 +0000[diff] [blame]4238 such as the <tt>MOVNT</tt> instruction on x86.</p>
David Greene02dfe202010-02-16 20:50:18 +0000[diff] [blame]4239
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4240<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4241<p>The location of memory pointed to is loaded. If the value being loaded is of
4242 scalar type then the number of bytes read does not exceed the minimum number
4243 of bytes needed to hold all bits of the type. For example, loading an
4244 <tt>i24</tt> reads at most three bytes. When loading a value of a type like
4245 <tt>i20</tt> with a size that is not an integral number of bytes, the result
4246 is undefined if the value was not originally written using a store of the
4247 same type.</p>
4248
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4249<h5>Examples:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4250<pre>
4251 %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
4252 <a href="#i_store">store</a> i32 3, i32* %ptr <i>; yields {void}</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4253 %val = load i32* %ptr <i>; yields {i32}:val = i32 3</i>
4254</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4255
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4256</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4257
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4258<!-- _______________________________________________________________________ -->
4259<div class="doc_subsubsection"> <a name="i_store">'<tt>store</tt>'
4260Instruction</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4261
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4262<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4263
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4264<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4265<pre>
David Greene02dfe202010-02-16 20:50:18 +0000[diff] [blame]4266 store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;][, !nontemporal !<index>] <i>; yields {void}</i>
4267 volatile store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;][, !nontemporal !<index>] <i>; yields {void}</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4268</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4269
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4270<h5>Overview:</h5>
4271<p>The '<tt>store</tt>' instruction is used to write to memory.</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4272
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4273<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4274<p>There are two arguments to the '<tt>store</tt>' instruction: a value to store
4275 and an address at which to store it. The type of the
4276 '<tt>&lt;pointer&gt;</tt>' operand must be a pointer to
4277 the <a href="#t_firstclass">first class</a> type of the
Jeffrey Yasskin6dc467f2010-04-26 21:21:24 +0000[diff] [blame]4278 '<tt>&lt;value&gt;</tt>' operand. If the <tt>store</tt> is marked as
4279 <tt>volatile</tt>, then the optimizer is not allowed to modify the number or
4280 order of execution of this <tt>store</tt> with other <a
4281 href="#volatile">volatile operations</a>.</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4282
4283<p>The optional constant "align" argument specifies the alignment of the
4284 operation (that is, the alignment of the memory address). A value of 0 or an
4285 omitted "align" argument means that the operation has the preferential
4286 alignment for the target. It is the responsibility of the code emitter to
4287 ensure that the alignment information is correct. Overestimating the
4288 alignment results in an undefined behavior. Underestimating the alignment may
4289 produce less efficient code. An alignment of 1 is always safe.</p>
4290
David Greene02dfe202010-02-16 20:50:18 +0000[diff] [blame]4291<p>The optional !nontemporal metadata must reference a single metatadata
4292 name <index> corresponding to a metadata node with one i32 entry of
Dan Gohman22dc6682010-03-01 17:41:39 +0000[diff] [blame]4293 value 1. The existence of the !nontemporal metatadata on the
David Greene02dfe202010-02-16 20:50:18 +0000[diff] [blame]4294 instruction tells the optimizer and code generator that this load is
4295 not expected to be reused in the cache. The code generator may
4296 select special instructions to save cache bandwidth, such as the
Dan Gohman22dc6682010-03-01 17:41:39 +0000[diff] [blame]4297 MOVNT instruction on x86.</p>
David Greene02dfe202010-02-16 20:50:18 +0000[diff] [blame]4298
4299
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4300<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4301<p>The contents of memory are updated to contain '<tt>&lt;value&gt;</tt>' at the
4302 location specified by the '<tt>&lt;pointer&gt;</tt>' operand. If
4303 '<tt>&lt;value&gt;</tt>' is of scalar type then the number of bytes written
4304 does not exceed the minimum number of bytes needed to hold all bits of the
4305 type. For example, storing an <tt>i24</tt> writes at most three bytes. When
4306 writing a value of a type like <tt>i20</tt> with a size that is not an
4307 integral number of bytes, it is unspecified what happens to the extra bits
4308 that do not belong to the type, but they will typically be overwritten.</p>
4309
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4310<h5>Example:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4311<pre>
4312 %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
Bill Wendling63ffa142007-10-22 05:10:05 +0000[diff] [blame]4313 store i32 3, i32* %ptr <i>; yields {void}</i>
4314 %val = <a href="#i_load">load</a> i32* %ptr <i>; yields {i32}:val = i32 3</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4315</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4316
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4317</div>
4318
4319<!-- _______________________________________________________________________ -->
4320<div class="doc_subsubsection">
4321 <a name="i_getelementptr">'<tt>getelementptr</tt>' Instruction</a>
4322</div>
4323
4324<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4325
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4326<h5>Syntax:</h5>
4327<pre>
Matthijs Kooijman32a080e2008-10-13 13:44:15 +0000[diff] [blame]4328 &lt;result&gt; = getelementptr &lt;pty&gt;* &lt;ptrval&gt;{, &lt;ty&gt; &lt;idx&gt;}*
Dan Gohman106b2ae2009-07-27 21:53:46 +0000[diff] [blame]4329 &lt;result&gt; = getelementptr inbounds &lt;pty&gt;* &lt;ptrval&gt;{, &lt;ty&gt; &lt;idx&gt;}*
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4330</pre>
4331
4332<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4333<p>The '<tt>getelementptr</tt>' instruction is used to get the address of a
Chris Lattnerd5d51722010-02-12 20:49:41 +0000[diff] [blame]4334 subelement of an <a href="#t_aggregate">aggregate</a> data structure.
4335 It performs address calculation only and does not access memory.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4336
4337<h5>Arguments:</h5>
Matthijs Kooijman32a080e2008-10-13 13:44:15 +0000[diff] [blame]4338<p>The first argument is always a pointer, and forms the basis of the
Chris Lattnere92fc832009-07-29 06:44:13 +0000[diff] [blame]4339 calculation. The remaining arguments are indices that indicate which of the
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4340 elements of the aggregate object are indexed. The interpretation of each
4341 index is dependent on the type being indexed into. The first index always
4342 indexes the pointer value given as the first argument, the second index
4343 indexes a value of the type pointed to (not necessarily the value directly
4344 pointed to, since the first index can be non-zero), etc. The first type
Chris Lattnerd5d51722010-02-12 20:49:41 +0000[diff] [blame]4345 indexed into must be a pointer value, subsequent types can be arrays,
4346 vectors, structs and unions. Note that subsequent types being indexed into
4347 can never be pointers, since that would require loading the pointer before
4348 continuing calculation.</p>
Matthijs Kooijman32a080e2008-10-13 13:44:15 +0000[diff] [blame]4349
4350<p>The type of each index argument depends on the type it is indexing into.
Chris Lattnerd5d51722010-02-12 20:49:41 +0000[diff] [blame]4351 When indexing into a (optionally packed) structure or union, only <tt>i32</tt>
4352 integer <b>constants</b> are allowed. When indexing into an array, pointer
4353 or vector, integers of any width are allowed, and they are not required to be
Chris Lattnere92fc832009-07-29 06:44:13 +0000[diff] [blame]4354 constant.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4355
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4356<p>For example, let's consider a C code fragment and how it gets compiled to
4357 LLVM:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4358
4359<div class="doc_code">
4360<pre>
4361struct RT {
4362 char A;
4363 int B[10][20];
4364 char C;
4365};
4366struct ST {
4367 int X;
4368 double Y;
4369 struct RT Z;
4370};
4371
4372int *foo(struct ST *s) {
4373 return &amp;s[1].Z.B[5][13];
4374}
4375</pre>
4376</div>
4377
4378<p>The LLVM code generated by the GCC frontend is:</p>
4379
4380<div class="doc_code">
4381<pre>
Chris Lattner5b6dc6e2009-01-11 20:53:49 +0000[diff] [blame]4382%RT = <a href="#namedtypes">type</a> { i8 , [10 x [20 x i32]], i8 }
4383%ST = <a href="#namedtypes">type</a> { i32, double, %RT }
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4384
Dan Gohman47360842009-07-25 02:23:48 +0000[diff] [blame]4385define i32* @foo(%ST* %s) {
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4386entry:
4387 %reg = getelementptr %ST* %s, i32 1, i32 2, i32 1, i32 5, i32 13
4388 ret i32* %reg
4389}
4390</pre>
4391</div>
4392
4393<h5>Semantics:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4394<p>In the example above, the first index is indexing into the '<tt>%ST*</tt>'
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4395 type, which is a pointer, yielding a '<tt>%ST</tt>' = '<tt>{ i32, double, %RT
4396 }</tt>' type, a structure. The second index indexes into the third element
4397 of the structure, yielding a '<tt>%RT</tt>' = '<tt>{ i8 , [10 x [20 x i32]],
4398 i8 }</tt>' type, another structure. The third index indexes into the second
4399 element of the structure, yielding a '<tt>[10 x [20 x i32]]</tt>' type, an
4400 array. The two dimensions of the array are subscripted into, yielding an
4401 '<tt>i32</tt>' type. The '<tt>getelementptr</tt>' instruction returns a
4402 pointer to this element, thus computing a value of '<tt>i32*</tt>' type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4403
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4404<p>Note that it is perfectly legal to index partially through a structure,
4405 returning a pointer to an inner element. Because of this, the LLVM code for
4406 the given testcase is equivalent to:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4407
4408<pre>
Dan Gohman47360842009-07-25 02:23:48 +0000[diff] [blame]4409 define i32* @foo(%ST* %s) {
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4410 %t1 = getelementptr %ST* %s, i32 1 <i>; yields %ST*:%t1</i>
4411 %t2 = getelementptr %ST* %t1, i32 0, i32 2 <i>; yields %RT*:%t2</i>
4412 %t3 = getelementptr %RT* %t2, i32 0, i32 1 <i>; yields [10 x [20 x i32]]*:%t3</i>
4413 %t4 = getelementptr [10 x [20 x i32]]* %t3, i32 0, i32 5 <i>; yields [20 x i32]*:%t4</i>
4414 %t5 = getelementptr [20 x i32]* %t4, i32 0, i32 13 <i>; yields i32*:%t5</i>
4415 ret i32* %t5
4416 }
4417</pre>
4418
Dan Gohman106b2ae2009-07-27 21:53:46 +0000[diff] [blame]4419<p>If the <tt>inbounds</tt> keyword is present, the result value of the
Dan Gohmanf7263862010-04-23 15:23:32 +0000[diff] [blame]4420 <tt>getelementptr</tt> is a <a href="#trapvalues">trap value</a> if the
4421 base pointer is not an <i>in bounds</i> address of an allocated object,
4422 or if any of the addresses that would be formed by successive addition of
4423 the offsets implied by the indices to the base address with infinitely
4424 precise arithmetic are not an <i>in bounds</i> address of that allocated
4425 object. The <i>in bounds</i> addresses for an allocated object are all
4426 the addresses that point into the object, plus the address one byte past
4427 the end.</p>
Dan Gohman106b2ae2009-07-27 21:53:46 +0000[diff] [blame]4428
4429<p>If the <tt>inbounds</tt> keyword is not present, the offsets are added to
4430 the base address with silently-wrapping two's complement arithmetic, and
4431 the result value of the <tt>getelementptr</tt> may be outside the object
4432 pointed to by the base pointer. The result value may not necessarily be
4433 used to access memory though, even if it happens to point into allocated
4434 storage. See the <a href="#pointeraliasing">Pointer Aliasing Rules</a>
4435 section for more information.</p>
4436
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4437<p>The getelementptr instruction is often confusing. For some more insight into
4438 how it works, see <a href="GetElementPtr.html">the getelementptr FAQ</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4439
4440<h5>Example:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4441<pre>
4442 <i>; yields [12 x i8]*:aptr</i>
Matthijs Kooijman32a080e2008-10-13 13:44:15 +0000[diff] [blame]4443 %aptr = getelementptr {i32, [12 x i8]}* %saptr, i64 0, i32 1
4444 <i>; yields i8*:vptr</i>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]4445 %vptr = getelementptr {i32, &lt;2 x i8&gt;}* %svptr, i64 0, i32 1, i32 1
Matthijs Kooijman32a080e2008-10-13 13:44:15 +0000[diff] [blame]4446 <i>; yields i8*:eptr</i>
4447 %eptr = getelementptr [12 x i8]* %aptr, i64 0, i32 1
Sanjiv Gupta4f9a0dc2009-04-25 07:27:44 +0000[diff] [blame]4448 <i>; yields i32*:iptr</i>
Sanjiv Gupta1e46c582009-04-24 16:38:13 +0000[diff] [blame]4449 %iptr = getelementptr [10 x i32]* @arr, i16 0, i16 0
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4450</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4451
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4452</div>
4453
4454<!-- ======================================================================= -->
4455<div class="doc_subsection"> <a name="convertops">Conversion Operations</a>
4456</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4457
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4458<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4459
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4460<p>The instructions in this category are the conversion instructions (casting)
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4461 which all take a single operand and a type. They perform various bit
4462 conversions on the operand.</p>
4463
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4464</div>
4465
4466<!-- _______________________________________________________________________ -->
4467<div class="doc_subsubsection">
4468 <a name="i_trunc">'<tt>trunc .. to</tt>' Instruction</a>
4469</div>
4470<div class="doc_text">
4471
4472<h5>Syntax:</h5>
4473<pre>
4474 &lt;result&gt; = trunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4475</pre>
4476
4477<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4478<p>The '<tt>trunc</tt>' instruction truncates its operand to the
4479 type <tt>ty2</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4480
4481<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4482<p>The '<tt>trunc</tt>' instruction takes a <tt>value</tt> to trunc, which must
4483 be an <a href="#t_integer">integer</a> type, and a type that specifies the
4484 size and type of the result, which must be
4485 an <a href="#t_integer">integer</a> type. The bit size of <tt>value</tt> must
4486 be larger than the bit size of <tt>ty2</tt>. Equal sized types are not
4487 allowed.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4488
4489<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4490<p>The '<tt>trunc</tt>' instruction truncates the high order bits
4491 in <tt>value</tt> and converts the remaining bits to <tt>ty2</tt>. Since the
4492 source size must be larger than the destination size, <tt>trunc</tt> cannot
4493 be a <i>no-op cast</i>. It will always truncate bits.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4494
4495<h5>Example:</h5>
4496<pre>
4497 %X = trunc i32 257 to i8 <i>; yields i8:1</i>
4498 %Y = trunc i32 123 to i1 <i>; yields i1:true</i>
Gabor Greif463c9342009-10-28 09:21:30 +0000[diff] [blame]4499 %Z = trunc i32 122 to i1 <i>; yields i1:false</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4500</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4501
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4502</div>
4503
4504<!-- _______________________________________________________________________ -->
4505<div class="doc_subsubsection">
4506 <a name="i_zext">'<tt>zext .. to</tt>' Instruction</a>
4507</div>
4508<div class="doc_text">
4509
4510<h5>Syntax:</h5>
4511<pre>
4512 &lt;result&gt; = zext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4513</pre>
4514
4515<h5>Overview:</h5>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]4516<p>The '<tt>zext</tt>' instruction zero extends its operand to type
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4517 <tt>ty2</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4518
4519
4520<h5>Arguments:</h5>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]4521<p>The '<tt>zext</tt>' instruction takes a value to cast, which must be of
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4522 <a href="#t_integer">integer</a> type, and a type to cast it to, which must
4523 also be of <a href="#t_integer">integer</a> type. The bit size of the
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]4524 <tt>value</tt> must be smaller than the bit size of the destination type,
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4525 <tt>ty2</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4526
4527<h5>Semantics:</h5>
4528<p>The <tt>zext</tt> fills the high order bits of the <tt>value</tt> with zero
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4529 bits until it reaches the size of the destination type, <tt>ty2</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4530
4531<p>When zero extending from i1, the result will always be either 0 or 1.</p>
4532
4533<h5>Example:</h5>
4534<pre>
4535 %X = zext i32 257 to i64 <i>; yields i64:257</i>
4536 %Y = zext i1 true to i32 <i>; yields i32:1</i>
4537</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4538
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4539</div>
4540
4541<!-- _______________________________________________________________________ -->
4542<div class="doc_subsubsection">
4543 <a name="i_sext">'<tt>sext .. to</tt>' Instruction</a>
4544</div>
4545<div class="doc_text">
4546
4547<h5>Syntax:</h5>
4548<pre>
4549 &lt;result&gt; = sext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4550</pre>
4551
4552<h5>Overview:</h5>
4553<p>The '<tt>sext</tt>' sign extends <tt>value</tt> to the type <tt>ty2</tt>.</p>
4554
4555<h5>Arguments:</h5>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]4556<p>The '<tt>sext</tt>' instruction takes a value to cast, which must be of
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4557 <a href="#t_integer">integer</a> type, and a type to cast it to, which must
4558 also be of <a href="#t_integer">integer</a> type. The bit size of the
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]4559 <tt>value</tt> must be smaller than the bit size of the destination type,
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4560 <tt>ty2</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4561
4562<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4563<p>The '<tt>sext</tt>' instruction performs a sign extension by copying the sign
4564 bit (highest order bit) of the <tt>value</tt> until it reaches the bit size
4565 of the type <tt>ty2</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4566
4567<p>When sign extending from i1, the extension always results in -1 or 0.</p>
4568
4569<h5>Example:</h5>
4570<pre>
4571 %X = sext i8 -1 to i16 <i>; yields i16 :65535</i>
4572 %Y = sext i1 true to i32 <i>; yields i32:-1</i>
4573</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4574
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4575</div>
4576
4577<!-- _______________________________________________________________________ -->
4578<div class="doc_subsubsection">
4579 <a name="i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a>
4580</div>
4581
4582<div class="doc_text">
4583
4584<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4585<pre>
4586 &lt;result&gt; = fptrunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4587</pre>
4588
4589<h5>Overview:</h5>
4590<p>The '<tt>fptrunc</tt>' instruction truncates <tt>value</tt> to type
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4591 <tt>ty2</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4592
4593<h5>Arguments:</h5>
4594<p>The '<tt>fptrunc</tt>' instruction takes a <a href="#t_floating">floating
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4595 point</a> value to cast and a <a href="#t_floating">floating point</a> type
4596 to cast it to. The size of <tt>value</tt> must be larger than the size of
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]4597 <tt>ty2</tt>. This implies that <tt>fptrunc</tt> cannot be used to make a
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4598 <i>no-op cast</i>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4599
4600<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4601<p>The '<tt>fptrunc</tt>' instruction truncates a <tt>value</tt> from a larger
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]4602 <a href="#t_floating">floating point</a> type to a smaller
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4603 <a href="#t_floating">floating point</a> type. If the value cannot fit
4604 within the destination type, <tt>ty2</tt>, then the results are
4605 undefined.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4606
4607<h5>Example:</h5>
4608<pre>
4609 %X = fptrunc double 123.0 to float <i>; yields float:123.0</i>
4610 %Y = fptrunc double 1.0E+300 to float <i>; yields undefined</i>
4611</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4612
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4613</div>
4614
4615<!-- _______________________________________________________________________ -->
4616<div class="doc_subsubsection">
4617 <a name="i_fpext">'<tt>fpext .. to</tt>' Instruction</a>
4618</div>
4619<div class="doc_text">
4620
4621<h5>Syntax:</h5>
4622<pre>
4623 &lt;result&gt; = fpext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4624</pre>
4625
4626<h5>Overview:</h5>
4627<p>The '<tt>fpext</tt>' extends a floating point <tt>value</tt> to a larger
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4628 floating point value.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4629
4630<h5>Arguments:</h5>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]4631<p>The '<tt>fpext</tt>' instruction takes a
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4632 <a href="#t_floating">floating point</a> <tt>value</tt> to cast, and
4633 a <a href="#t_floating">floating point</a> type to cast it to. The source
4634 type must be smaller than the destination type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4635
4636<h5>Semantics:</h5>
4637<p>The '<tt>fpext</tt>' instruction extends the <tt>value</tt> from a smaller
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4638 <a href="#t_floating">floating point</a> type to a larger
4639 <a href="#t_floating">floating point</a> type. The <tt>fpext</tt> cannot be
4640 used to make a <i>no-op cast</i> because it always changes bits. Use
4641 <tt>bitcast</tt> to make a <i>no-op cast</i> for a floating point cast.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4642
4643<h5>Example:</h5>
4644<pre>
4645 %X = fpext float 3.1415 to double <i>; yields double:3.1415</i>
4646 %Y = fpext float 1.0 to float <i>; yields float:1.0 (no-op)</i>
4647</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4648
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4649</div>
4650
4651<!-- _______________________________________________________________________ -->
4652<div class="doc_subsubsection">
4653 <a name="i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a>
4654</div>
4655<div class="doc_text">
4656
4657<h5>Syntax:</h5>
4658<pre>
Reid Spencere6adee82007-07-31 14:40:14 +0000[diff] [blame]4659 &lt;result&gt; = fptoui &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4660</pre>
4661
4662<h5>Overview:</h5>
Reid Spencere6adee82007-07-31 14:40:14 +0000[diff] [blame]4663<p>The '<tt>fptoui</tt>' converts a floating point <tt>value</tt> to its
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4664 unsigned integer equivalent of type <tt>ty2</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4665
4666<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4667<p>The '<tt>fptoui</tt>' instruction takes a value to cast, which must be a
4668 scalar or vector <a href="#t_floating">floating point</a> value, and a type
4669 to cast it to <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a>
4670 type. If <tt>ty</tt> is a vector floating point type, <tt>ty2</tt> must be a
4671 vector integer type with the same number of elements as <tt>ty</tt></p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4672
4673<h5>Semantics:</h5>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]4674<p>The '<tt>fptoui</tt>' instruction converts its
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4675 <a href="#t_floating">floating point</a> operand into the nearest (rounding
4676 towards zero) unsigned integer value. If the value cannot fit
4677 in <tt>ty2</tt>, the results are undefined.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4678
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4679<h5>Example:</h5>
4680<pre>
Reid Spencere6adee82007-07-31 14:40:14 +0000[diff] [blame]4681 %X = fptoui double 123.0 to i32 <i>; yields i32:123</i>
Chris Lattner681f1e82007-09-22 03:17:52 +0000[diff] [blame]4682 %Y = fptoui float 1.0E+300 to i1 <i>; yields undefined:1</i>
Gabor Greif463c9342009-10-28 09:21:30 +0000[diff] [blame]4683 %Z = fptoui float 1.04E+17 to i8 <i>; yields undefined:1</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4684</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4685
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4686</div>
4687
4688<!-- _______________________________________________________________________ -->
4689<div class="doc_subsubsection">
4690 <a name="i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a>
4691</div>
4692<div class="doc_text">
4693
4694<h5>Syntax:</h5>
4695<pre>
4696 &lt;result&gt; = fptosi &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4697</pre>
4698
4699<h5>Overview:</h5>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]4700<p>The '<tt>fptosi</tt>' instruction converts
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4701 <a href="#t_floating">floating point</a> <tt>value</tt> to
4702 type <tt>ty2</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4703
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4704<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4705<p>The '<tt>fptosi</tt>' instruction takes a value to cast, which must be a
4706 scalar or vector <a href="#t_floating">floating point</a> value, and a type
4707 to cast it to <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a>
4708 type. If <tt>ty</tt> is a vector floating point type, <tt>ty2</tt> must be a
4709 vector integer type with the same number of elements as <tt>ty</tt></p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4710
4711<h5>Semantics:</h5>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]4712<p>The '<tt>fptosi</tt>' instruction converts its
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4713 <a href="#t_floating">floating point</a> operand into the nearest (rounding
4714 towards zero) signed integer value. If the value cannot fit in <tt>ty2</tt>,
4715 the results are undefined.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4716
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4717<h5>Example:</h5>
4718<pre>
4719 %X = fptosi double -123.0 to i32 <i>; yields i32:-123</i>
Chris Lattner681f1e82007-09-22 03:17:52 +0000[diff] [blame]4720 %Y = fptosi float 1.0E-247 to i1 <i>; yields undefined:1</i>
Gabor Greif463c9342009-10-28 09:21:30 +0000[diff] [blame]4721 %Z = fptosi float 1.04E+17 to i8 <i>; yields undefined:1</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4722</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4723
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4724</div>
4725
4726<!-- _______________________________________________________________________ -->
4727<div class="doc_subsubsection">
4728 <a name="i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a>
4729</div>
4730<div class="doc_text">
4731
4732<h5>Syntax:</h5>
4733<pre>
4734 &lt;result&gt; = uitofp &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4735</pre>
4736
4737<h5>Overview:</h5>
4738<p>The '<tt>uitofp</tt>' instruction regards <tt>value</tt> as an unsigned
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4739 integer and converts that value to the <tt>ty2</tt> type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4740
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4741<h5>Arguments:</h5>
Nate Begeman78246ca2007-11-17 03:58:34 +0000[diff] [blame]4742<p>The '<tt>uitofp</tt>' instruction takes a value to cast, which must be a
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4743 scalar or vector <a href="#t_integer">integer</a> value, and a type to cast
4744 it to <tt>ty2</tt>, which must be an <a href="#t_floating">floating point</a>
4745 type. If <tt>ty</tt> is a vector integer type, <tt>ty2</tt> must be a vector
4746 floating point type with the same number of elements as <tt>ty</tt></p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4747
4748<h5>Semantics:</h5>
4749<p>The '<tt>uitofp</tt>' instruction interprets its operand as an unsigned
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4750 integer quantity and converts it to the corresponding floating point
4751 value. If the value cannot fit in the floating point value, the results are
4752 undefined.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4753
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4754<h5>Example:</h5>
4755<pre>
4756 %X = uitofp i32 257 to float <i>; yields float:257.0</i>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]4757 %Y = uitofp i8 -1 to double <i>; yields double:255.0</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4758</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4759
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4760</div>
4761
4762<!-- _______________________________________________________________________ -->
4763<div class="doc_subsubsection">
4764 <a name="i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a>
4765</div>
4766<div class="doc_text">
4767
4768<h5>Syntax:</h5>
4769<pre>
4770 &lt;result&gt; = sitofp &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4771</pre>
4772
4773<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4774<p>The '<tt>sitofp</tt>' instruction regards <tt>value</tt> as a signed integer
4775 and converts that value to the <tt>ty2</tt> type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4776
4777<h5>Arguments:</h5>
Nate Begeman78246ca2007-11-17 03:58:34 +0000[diff] [blame]4778<p>The '<tt>sitofp</tt>' instruction takes a value to cast, which must be a
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4779 scalar or vector <a href="#t_integer">integer</a> value, and a type to cast
4780 it to <tt>ty2</tt>, which must be an <a href="#t_floating">floating point</a>
4781 type. If <tt>ty</tt> is a vector integer type, <tt>ty2</tt> must be a vector
4782 floating point type with the same number of elements as <tt>ty</tt></p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4783
4784<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4785<p>The '<tt>sitofp</tt>' instruction interprets its operand as a signed integer
4786 quantity and converts it to the corresponding floating point value. If the
4787 value cannot fit in the floating point value, the results are undefined.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4788
4789<h5>Example:</h5>
4790<pre>
4791 %X = sitofp i32 257 to float <i>; yields float:257.0</i>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]4792 %Y = sitofp i8 -1 to double <i>; yields double:-1.0</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4793</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4794
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4795</div>
4796
4797<!-- _______________________________________________________________________ -->
4798<div class="doc_subsubsection">
4799 <a name="i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a>
4800</div>
4801<div class="doc_text">
4802
4803<h5>Syntax:</h5>
4804<pre>
4805 &lt;result&gt; = ptrtoint &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4806</pre>
4807
4808<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4809<p>The '<tt>ptrtoint</tt>' instruction converts the pointer <tt>value</tt> to
4810 the integer type <tt>ty2</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4811
4812<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4813<p>The '<tt>ptrtoint</tt>' instruction takes a <tt>value</tt> to cast, which
4814 must be a <a href="#t_pointer">pointer</a> value, and a type to cast it to
4815 <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a> type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4816
4817<h5>Semantics:</h5>
4818<p>The '<tt>ptrtoint</tt>' instruction converts <tt>value</tt> to integer type
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4819 <tt>ty2</tt> by interpreting the pointer value as an integer and either
4820 truncating or zero extending that value to the size of the integer type. If
4821 <tt>value</tt> is smaller than <tt>ty2</tt> then a zero extension is done. If
4822 <tt>value</tt> is larger than <tt>ty2</tt> then a truncation is done. If they
4823 are the same size, then nothing is done (<i>no-op cast</i>) other than a type
4824 change.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4825
4826<h5>Example:</h5>
4827<pre>
4828 %X = ptrtoint i32* %X to i8 <i>; yields truncation on 32-bit architecture</i>
4829 %Y = ptrtoint i32* %x to i64 <i>; yields zero extension on 32-bit architecture</i>
4830</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4831
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4832</div>
4833
4834<!-- _______________________________________________________________________ -->
4835<div class="doc_subsubsection">
4836 <a name="i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a>
4837</div>
4838<div class="doc_text">
4839
4840<h5>Syntax:</h5>
4841<pre>
4842 &lt;result&gt; = inttoptr &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4843</pre>
4844
4845<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4846<p>The '<tt>inttoptr</tt>' instruction converts an integer <tt>value</tt> to a
4847 pointer type, <tt>ty2</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4848
4849<h5>Arguments:</h5>
4850<p>The '<tt>inttoptr</tt>' instruction takes an <a href="#t_integer">integer</a>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4851 value to cast, and a type to cast it to, which must be a
4852 <a href="#t_pointer">pointer</a> type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4853
4854<h5>Semantics:</h5>
4855<p>The '<tt>inttoptr</tt>' instruction converts <tt>value</tt> to type
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4856 <tt>ty2</tt> by applying either a zero extension or a truncation depending on
4857 the size of the integer <tt>value</tt>. If <tt>value</tt> is larger than the
4858 size of a pointer then a truncation is done. If <tt>value</tt> is smaller
4859 than the size of a pointer then a zero extension is done. If they are the
4860 same size, nothing is done (<i>no-op cast</i>).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4861
4862<h5>Example:</h5>
4863<pre>
4864 %X = inttoptr i32 255 to i32* <i>; yields zero extension on 64-bit architecture</i>
Gabor Greif463c9342009-10-28 09:21:30 +0000[diff] [blame]4865 %Y = inttoptr i32 255 to i32* <i>; yields no-op on 32-bit architecture</i>
4866 %Z = inttoptr i64 0 to i32* <i>; yields truncation on 32-bit architecture</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4867</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4868
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4869</div>
4870
4871<!-- _______________________________________________________________________ -->
4872<div class="doc_subsubsection">
4873 <a name="i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a>
4874</div>
4875<div class="doc_text">
4876
4877<h5>Syntax:</h5>
4878<pre>
4879 &lt;result&gt; = bitcast &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4880</pre>
4881
4882<h5>Overview:</h5>
4883<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4884 <tt>ty2</tt> without changing any bits.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4885
4886<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4887<p>The '<tt>bitcast</tt>' instruction takes a value to cast, which must be a
4888 non-aggregate first class value, and a type to cast it to, which must also be
4889 a non-aggregate <a href="#t_firstclass">first class</a> type. The bit sizes
4890 of <tt>value</tt> and the destination type, <tt>ty2</tt>, must be
4891 identical. If the source type is a pointer, the destination type must also be
4892 a pointer. This instruction supports bitwise conversion of vectors to
4893 integers and to vectors of other types (as long as they have the same
4894 size).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4895
4896<h5>Semantics:</h5>
4897<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4898 <tt>ty2</tt>. It is always a <i>no-op cast</i> because no bits change with
4899 this conversion. The conversion is done as if the <tt>value</tt> had been
4900 stored to memory and read back as type <tt>ty2</tt>. Pointer types may only
4901 be converted to other pointer types with this instruction. To convert
4902 pointers to other types, use the <a href="#i_inttoptr">inttoptr</a> or
4903 <a href="#i_ptrtoint">ptrtoint</a> instructions first.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4904
4905<h5>Example:</h5>
4906<pre>
4907 %X = bitcast i8 255 to i8 <i>; yields i8 :-1</i>
4908 %Y = bitcast i32* %x to sint* <i>; yields sint*:%x</i>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]4909 %Z = bitcast &lt;2 x int&gt; %V to i64; <i>; yields i64: %V</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4910</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4911
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4912</div>
4913
4914<!-- ======================================================================= -->
4915<div class="doc_subsection"> <a name="otherops">Other Operations</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4916
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4917<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4918
4919<p>The instructions in this category are the "miscellaneous" instructions, which
4920 defy better classification.</p>
4921
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4922</div>
4923
4924<!-- _______________________________________________________________________ -->
4925<div class="doc_subsubsection"><a name="i_icmp">'<tt>icmp</tt>' Instruction</a>
4926</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4927
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4928<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4929
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4930<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4931<pre>
4932 &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>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4933</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4934
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4935<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4936<p>The '<tt>icmp</tt>' instruction returns a boolean value or a vector of
4937 boolean values based on comparison of its two integer, integer vector, or
4938 pointer operands.</p>
4939
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4940<h5>Arguments:</h5>
4941<p>The '<tt>icmp</tt>' instruction takes three operands. The first operand is
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4942 the condition code indicating the kind of comparison to perform. It is not a
4943 value, just a keyword. The possible condition code are:</p>
4944
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4945<ol>
4946 <li><tt>eq</tt>: equal</li>
4947 <li><tt>ne</tt>: not equal </li>
4948 <li><tt>ugt</tt>: unsigned greater than</li>
4949 <li><tt>uge</tt>: unsigned greater or equal</li>
4950 <li><tt>ult</tt>: unsigned less than</li>
4951 <li><tt>ule</tt>: unsigned less or equal</li>
4952 <li><tt>sgt</tt>: signed greater than</li>
4953 <li><tt>sge</tt>: signed greater or equal</li>
4954 <li><tt>slt</tt>: signed less than</li>
4955 <li><tt>sle</tt>: signed less or equal</li>
4956</ol>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4957
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4958<p>The remaining two arguments must be <a href="#t_integer">integer</a> or
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4959 <a href="#t_pointer">pointer</a> or integer <a href="#t_vector">vector</a>
4960 typed. They must also be identical types.</p>
4961
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4962<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4963<p>The '<tt>icmp</tt>' compares <tt>op1</tt> and <tt>op2</tt> according to the
4964 condition code given as <tt>cond</tt>. The comparison performed always yields
Nick Lewycky244cf482009-09-27 00:45:11 +0000[diff] [blame]4965 either an <a href="#t_integer"><tt>i1</tt></a> or vector of <tt>i1</tt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4966 result, as follows:</p>
4967
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4968<ol>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]4969 <li><tt>eq</tt>: yields <tt>true</tt> if the operands are equal,
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4970 <tt>false</tt> otherwise. No sign interpretation is necessary or
4971 performed.</li>
4972
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]4973 <li><tt>ne</tt>: yields <tt>true</tt> if the operands are unequal,
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4974 <tt>false</tt> otherwise. No sign interpretation is necessary or
4975 performed.</li>
4976
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4977 <li><tt>ugt</tt>: interprets the operands as unsigned values and yields
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4978 <tt>true</tt> if <tt>op1</tt> is greater than <tt>op2</tt>.</li>
4979
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4980 <li><tt>uge</tt>: interprets the operands as unsigned values and yields
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4981 <tt>true</tt> if <tt>op1</tt> is greater than or equal
4982 to <tt>op2</tt>.</li>
4983
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4984 <li><tt>ult</tt>: interprets the operands as unsigned values and yields
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4985 <tt>true</tt> if <tt>op1</tt> is less than <tt>op2</tt>.</li>
4986
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4987 <li><tt>ule</tt>: interprets the operands as unsigned values and yields
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4988 <tt>true</tt> if <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
4989
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4990 <li><tt>sgt</tt>: interprets the operands as signed values and yields
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4991 <tt>true</tt> if <tt>op1</tt> is greater than <tt>op2</tt>.</li>
4992
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4993 <li><tt>sge</tt>: interprets the operands as signed values and yields
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4994 <tt>true</tt> if <tt>op1</tt> is greater than or equal
4995 to <tt>op2</tt>.</li>
4996
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4997 <li><tt>slt</tt>: interprets the operands as signed values and yields
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4998 <tt>true</tt> if <tt>op1</tt> is less than <tt>op2</tt>.</li>
4999
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5000 <li><tt>sle</tt>: interprets the operands as signed values and yields
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5001 <tt>true</tt> if <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5002</ol>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5003
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5004<p>If the operands are <a href="#t_pointer">pointer</a> typed, the pointer
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5005 values are compared as if they were integers.</p>
5006
5007<p>If the operands are integer vectors, then they are compared element by
5008 element. The result is an <tt>i1</tt> vector with the same number of elements
5009 as the values being compared. Otherwise, the result is an <tt>i1</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5010
5011<h5>Example:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5012<pre>
5013 &lt;result&gt; = icmp eq i32 4, 5 <i>; yields: result=false</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5014 &lt;result&gt; = icmp ne float* %X, %X <i>; yields: result=false</i>
5015 &lt;result&gt; = icmp ult i16 4, 5 <i>; yields: result=true</i>
5016 &lt;result&gt; = icmp sgt i16 4, 5 <i>; yields: result=false</i>
5017 &lt;result&gt; = icmp ule i16 -4, 5 <i>; yields: result=false</i>
5018 &lt;result&gt; = icmp sge i16 4, 5 <i>; yields: result=false</i>
5019</pre>
Dan Gohmana53eb382009-01-22 01:39:38 +0000[diff] [blame]5020
5021<p>Note that the code generator does not yet support vector types with
5022 the <tt>icmp</tt> instruction.</p>
5023
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5024</div>
5025
5026<!-- _______________________________________________________________________ -->
5027<div class="doc_subsubsection"><a name="i_fcmp">'<tt>fcmp</tt>' Instruction</a>
5028</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5029
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5030<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5031
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5032<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5033<pre>
5034 &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>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5035</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5036
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5037<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5038<p>The '<tt>fcmp</tt>' instruction returns a boolean value or vector of boolean
5039 values based on comparison of its operands.</p>
5040
5041<p>If the operands are floating point scalars, then the result type is a boolean
Nick Lewycky244cf482009-09-27 00:45:11 +0000[diff] [blame]5042(<a href="#t_integer"><tt>i1</tt></a>).</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5043
5044<p>If the operands are floating point vectors, then the result type is a vector
5045 of boolean with the same number of elements as the operands being
5046 compared.</p>
5047
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5048<h5>Arguments:</h5>
5049<p>The '<tt>fcmp</tt>' instruction takes three operands. The first operand is
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5050 the condition code indicating the kind of comparison to perform. It is not a
5051 value, just a keyword. The possible condition code are:</p>
5052
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5053<ol>
5054 <li><tt>false</tt>: no comparison, always returns false</li>
5055 <li><tt>oeq</tt>: ordered and equal</li>
5056 <li><tt>ogt</tt>: ordered and greater than </li>
5057 <li><tt>oge</tt>: ordered and greater than or equal</li>
5058 <li><tt>olt</tt>: ordered and less than </li>
5059 <li><tt>ole</tt>: ordered and less than or equal</li>
5060 <li><tt>one</tt>: ordered and not equal</li>
5061 <li><tt>ord</tt>: ordered (no nans)</li>
5062 <li><tt>ueq</tt>: unordered or equal</li>
5063 <li><tt>ugt</tt>: unordered or greater than </li>
5064 <li><tt>uge</tt>: unordered or greater than or equal</li>
5065 <li><tt>ult</tt>: unordered or less than </li>
5066 <li><tt>ule</tt>: unordered or less than or equal</li>
5067 <li><tt>une</tt>: unordered or not equal</li>
5068 <li><tt>uno</tt>: unordered (either nans)</li>
5069 <li><tt>true</tt>: no comparison, always returns true</li>
5070</ol>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5071
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5072<p><i>Ordered</i> means that neither operand is a QNAN while
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5073 <i>unordered</i> means that either operand may be a QNAN.</p>
5074
5075<p>Each of <tt>val1</tt> and <tt>val2</tt> arguments must be either
5076 a <a href="#t_floating">floating point</a> type or
5077 a <a href="#t_vector">vector</a> of floating point type. They must have
5078 identical types.</p>
5079
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5080<h5>Semantics:</h5>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]5081<p>The '<tt>fcmp</tt>' instruction compares <tt>op1</tt> and <tt>op2</tt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5082 according to the condition code given as <tt>cond</tt>. If the operands are
5083 vectors, then the vectors are compared element by element. Each comparison
Nick Lewycky244cf482009-09-27 00:45:11 +0000[diff] [blame]5084 performed always yields an <a href="#t_integer">i1</a> result, as
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5085 follows:</p>
5086
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5087<ol>
5088 <li><tt>false</tt>: always yields <tt>false</tt>, regardless of operands.</li>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5089
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]5090 <li><tt>oeq</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5091 <tt>op1</tt> is equal to <tt>op2</tt>.</li>
5092
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5093 <li><tt>ogt</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Dan Gohman22dc6682010-03-01 17:41:39 +0000[diff] [blame]5094 <tt>op1</tt> is greater than <tt>op2</tt>.</li>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5095
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]5096 <li><tt>oge</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5097 <tt>op1</tt> is greater than or equal to <tt>op2</tt>.</li>
5098
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]5099 <li><tt>olt</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5100 <tt>op1</tt> is less than <tt>op2</tt>.</li>
5101
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]5102 <li><tt>ole</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5103 <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
5104
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]5105 <li><tt>one</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5106 <tt>op1</tt> is not equal to <tt>op2</tt>.</li>
5107
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5108 <li><tt>ord</tt>: yields <tt>true</tt> if both operands are not a QNAN.</li>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5109
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]5110 <li><tt>ueq</tt>: yields <tt>true</tt> if either operand is a QNAN or
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5111 <tt>op1</tt> is equal to <tt>op2</tt>.</li>
5112
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]5113 <li><tt>ugt</tt>: yields <tt>true</tt> if either operand is a QNAN or
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5114 <tt>op1</tt> is greater than <tt>op2</tt>.</li>
5115
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]5116 <li><tt>uge</tt>: yields <tt>true</tt> if either operand is a QNAN or
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5117 <tt>op1</tt> is greater than or equal to <tt>op2</tt>.</li>
5118
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]5119 <li><tt>ult</tt>: yields <tt>true</tt> if either operand is a QNAN or
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5120 <tt>op1</tt> is less than <tt>op2</tt>.</li>
5121
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]5122 <li><tt>ule</tt>: yields <tt>true</tt> if either operand is a QNAN or
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5123 <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
5124
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]5125 <li><tt>une</tt>: yields <tt>true</tt> if either operand is a QNAN or
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5126 <tt>op1</tt> is not equal to <tt>op2</tt>.</li>
5127
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5128 <li><tt>uno</tt>: yields <tt>true</tt> if either operand is a QNAN.</li>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5129
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5130 <li><tt>true</tt>: always yields <tt>true</tt>, regardless of operands.</li>
5131</ol>
5132
5133<h5>Example:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5134<pre>
5135 &lt;result&gt; = fcmp oeq float 4.0, 5.0 <i>; yields: result=false</i>
Dan Gohmanb60ca3c2008-09-09 01:02:47 +0000[diff] [blame]5136 &lt;result&gt; = fcmp one float 4.0, 5.0 <i>; yields: result=true</i>
5137 &lt;result&gt; = fcmp olt float 4.0, 5.0 <i>; yields: result=true</i>
5138 &lt;result&gt; = fcmp ueq double 1.0, 2.0 <i>; yields: result=false</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5139</pre>
Dan Gohmana53eb382009-01-22 01:39:38 +0000[diff] [blame]5140
5141<p>Note that the code generator does not yet support vector types with
5142 the <tt>fcmp</tt> instruction.</p>
5143
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5144</div>
5145
5146<!-- _______________________________________________________________________ -->
Nate Begeman646fa482008-05-12 19:01:56 +0000[diff] [blame]5147<div class="doc_subsubsection">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]5148 <a name="i_phi">'<tt>phi</tt>' Instruction</a>
5149</div>
5150
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5151<div class="doc_text">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]5152
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5153<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5154<pre>
5155 &lt;result&gt; = phi &lt;ty&gt; [ &lt;val0&gt;, &lt;label0&gt;], ...
5156</pre>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]5157
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5158<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5159<p>The '<tt>phi</tt>' instruction is used to implement the &#966; node in the
5160 SSA graph representing the function.</p>
5161
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5162<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5163<p>The type of the incoming values is specified with the first type field. After
5164 this, the '<tt>phi</tt>' instruction takes a list of pairs as arguments, with
5165 one pair for each predecessor basic block of the current block. Only values
5166 of <a href="#t_firstclass">first class</a> type may be used as the value
5167 arguments to the PHI node. Only labels may be used as the label
5168 arguments.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]5169
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5170<p>There must be no non-phi instructions between the start of a basic block and
5171 the PHI instructions: i.e. PHI instructions must be first in a basic
5172 block.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]5173
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5174<p>For the purposes of the SSA form, the use of each incoming value is deemed to
5175 occur on the edge from the corresponding predecessor block to the current
5176 block (but after any definition of an '<tt>invoke</tt>' instruction's return
5177 value on the same edge).</p>
Jay Foad8e2fd2c2009-06-03 10:20:10 +0000[diff] [blame]5178
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5179<h5>Semantics:</h5>
5180<p>At runtime, the '<tt>phi</tt>' instruction logically takes on the value
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5181 specified by the pair corresponding to the predecessor basic block that
5182 executed just prior to the current block.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]5183
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5184<h5>Example:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]5185<pre>
5186Loop: ; Infinite loop that counts from 0 on up...
5187 %indvar = phi i32 [ 0, %LoopHeader ], [ %nextindvar, %Loop ]
5188 %nextindvar = add i32 %indvar, 1
5189 br label %Loop
5190</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5191
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5192</div>
5193
5194<!-- _______________________________________________________________________ -->
5195<div class="doc_subsubsection">
5196 <a name="i_select">'<tt>select</tt>' Instruction</a>
5197</div>
5198
5199<div class="doc_text">
5200
5201<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5202<pre>
Dan Gohmanb60ca3c2008-09-09 01:02:47 +0000[diff] [blame]5203 &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>
5204
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]5205 <i>selty</i> is either i1 or {&lt;N x i1&gt;}
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5206</pre>
5207
5208<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5209<p>The '<tt>select</tt>' instruction is used to choose one value based on a
5210 condition, without branching.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5211
5212
5213<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5214<p>The '<tt>select</tt>' instruction requires an 'i1' value or a vector of 'i1'
5215 values indicating the condition, and two values of the
5216 same <a href="#t_firstclass">first class</a> type. If the val1/val2 are
5217 vectors and the condition is a scalar, then entire vectors are selected, not
5218 individual elements.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5219
5220<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5221<p>If the condition is an i1 and it evaluates to 1, the instruction returns the
5222 first value argument; otherwise, it returns the second value argument.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5223
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5224<p>If the condition is a vector of i1, then the value arguments must be vectors
5225 of the same size, and the selection is done element by element.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5226
5227<h5>Example:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5228<pre>
5229 %X = select i1 true, i8 17, i8 42 <i>; yields i8:17</i>
5230</pre>
Dan Gohmana53eb382009-01-22 01:39:38 +0000[diff] [blame]5231
5232<p>Note that the code generator does not yet support conditions
5233 with vector type.</p>
5234
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5235</div>
5236
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5237<!-- _______________________________________________________________________ -->
5238<div class="doc_subsubsection">
5239 <a name="i_call">'<tt>call</tt>' Instruction</a>
5240</div>
5241
5242<div class="doc_text">
5243
5244<h5>Syntax:</h5>
5245<pre>
Devang Pateld0bfcc72008-10-07 17:48:33 +0000[diff] [blame]5246 &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>]
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5247</pre>
5248
5249<h5>Overview:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5250<p>The '<tt>call</tt>' instruction represents a simple function call.</p>
5251
5252<h5>Arguments:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5253<p>This instruction requires several arguments:</p>
5254
5255<ol>
Jeffrey Yasskin3e1a3a02010-01-09 19:44:16 +0000[diff] [blame]5256 <li>The optional "tail" marker indicates that the callee function does not
5257 access any allocas or varargs in the caller. Note that calls may be
5258 marked "tail" even if they do not occur before
5259 a <a href="#i_ret"><tt>ret</tt></a> instruction. If the "tail" marker is
5260 present, the function call is eligible for tail call optimization,
5261 but <a href="CodeGenerator.html#tailcallopt">might not in fact be
Evan Chengcc7495c2010-03-08 21:05:02 +0000[diff] [blame]5262 optimized into a jump</a>. The code generator may optimize calls marked
5263 "tail" with either 1) automatic <a href="CodeGenerator.html#sibcallopt">
5264 sibling call optimization</a> when the caller and callee have
5265 matching signatures, or 2) forced tail call optimization when the
5266 following extra requirements are met:
Jeffrey Yasskin3e1a3a02010-01-09 19:44:16 +0000[diff] [blame]5267 <ul>
5268 <li>Caller and callee both have the calling
5269 convention <tt>fastcc</tt>.</li>
5270 <li>The call is in tail position (ret immediately follows call and ret
5271 uses value of call or is void).</li>
5272 <li>Option <tt>-tailcallopt</tt> is enabled,
Dan Gohman1be84f02010-03-02 01:08:11 +0000[diff] [blame]5273 or <code>llvm::GuaranteedTailCallOpt</code> is <code>true</code>.</li>
Jeffrey Yasskin3e1a3a02010-01-09 19:44:16 +0000[diff] [blame]5274 <li><a href="CodeGenerator.html#tailcallopt">Platform specific
5275 constraints are met.</a></li>
5276 </ul>
5277 </li>
Devang Patelac2fc272008-10-06 18:50:38 +0000[diff] [blame]5278
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5279 <li>The optional "cconv" marker indicates which <a href="#callingconv">calling
5280 convention</a> the call should use. If none is specified, the call
Jeffrey Yasskin3e1a3a02010-01-09 19:44:16 +0000[diff] [blame]5281 defaults to using C calling conventions. The calling convention of the
5282 call must match the calling convention of the target function, or else the
5283 behavior is undefined.</li>
Devang Patelac2fc272008-10-06 18:50:38 +0000[diff] [blame]5284
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5285 <li>The optional <a href="#paramattrs">Parameter Attributes</a> list for
5286 return values. Only '<tt>zeroext</tt>', '<tt>signext</tt>', and
5287 '<tt>inreg</tt>' attributes are valid here.</li>
5288
5289 <li>'<tt>ty</tt>': the type of the call instruction itself which is also the
5290 type of the return value. Functions that return no value are marked
5291 <tt><a href="#t_void">void</a></tt>.</li>
5292
5293 <li>'<tt>fnty</tt>': shall be the signature of the pointer to function value
5294 being invoked. The argument types must match the types implied by this
5295 signature. This type can be omitted if the function is not varargs and if
5296 the function type does not return a pointer to a function.</li>
5297
5298 <li>'<tt>fnptrval</tt>': An LLVM value containing a pointer to a function to
5299 be invoked. In most cases, this is a direct function invocation, but
5300 indirect <tt>call</tt>s are just as possible, calling an arbitrary pointer
5301 to function value.</li>
5302
5303 <li>'<tt>function args</tt>': argument list whose types match the function
Chris Lattner553fb1e2010-03-02 06:36:51 +0000[diff] [blame]5304 signature argument types and parameter attributes. All arguments must be
5305 of <a href="#t_firstclass">first class</a> type. If the function
5306 signature indicates the function accepts a variable number of arguments,
5307 the extra arguments can be specified.</li>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5308
5309 <li>The optional <a href="#fnattrs">function attributes</a> list. Only
5310 '<tt>noreturn</tt>', '<tt>nounwind</tt>', '<tt>readonly</tt>' and
5311 '<tt>readnone</tt>' attributes are valid here.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5312</ol>
5313
5314<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5315<p>The '<tt>call</tt>' instruction is used to cause control flow to transfer to
5316 a specified function, with its incoming arguments bound to the specified
5317 values. Upon a '<tt><a href="#i_ret">ret</a></tt>' instruction in the called
5318 function, control flow continues with the instruction after the function
5319 call, and the return value of the function is bound to the result
5320 argument.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5321
5322<h5>Example:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5323<pre>
Nick Lewycky93082fc2007-09-08 13:57:50 +0000[diff] [blame]5324 %retval = call i32 @test(i32 %argc)
Chris Lattner5e893ef2008-03-21 17:24:17 +0000[diff] [blame]5325 call i32 (i8 *, ...)* @printf(i8 * %msg, i32 12, i8 42) <i>; yields i32</i>
5326 %X = tail call i32 @foo() <i>; yields i32</i>
5327 %Y = tail call <a href="#callingconv">fastcc</a> i32 @foo() <i>; yields i32</i>
5328 call void %foo(i8 97 signext)
Devang Patela3cc5372008-03-10 20:49:15 +0000[diff] [blame]5329
5330 %struct.A = type { i32, i8 }
Devang Patelac2fc272008-10-06 18:50:38 +0000[diff] [blame]5331 %r = call %struct.A @foo() <i>; yields { 32, i8 }</i>
Dan Gohman3e700032008-10-04 19:00:07 +0000[diff] [blame]5332 %gr = extractvalue %struct.A %r, 0 <i>; yields i32</i>
5333 %gr1 = extractvalue %struct.A %r, 1 <i>; yields i8</i>
Chris Lattnerac454b32008-10-08 06:26:11 +0000[diff] [blame]5334 %Z = call void @foo() noreturn <i>; indicates that %foo never returns normally</i>
Matthijs Kooijman2c4e05a2008-10-07 10:03:45 +0000[diff] [blame]5335 %ZZ = call zeroext i32 @bar() <i>; Return value is %zero extended</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5336</pre>
5337
Dale Johannesen9fcf4a92009-09-24 18:38:21 +0000[diff] [blame]5338<p>llvm treats calls to some functions with names and arguments that match the
Dale Johannesen4730cd32009-09-25 17:04:42 +0000[diff] [blame]5339standard C99 library as being the C99 library functions, and may perform
5340optimizations or generate code for them under that assumption. This is
5341something we'd like to change in the future to provide better support for
Dan Gohman22dc6682010-03-01 17:41:39 +0000[diff] [blame]5342freestanding environments and non-C-based languages.</p>
Dale Johannesen9fcf4a92009-09-24 18:38:21 +0000[diff] [blame]5343
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5344</div>
5345
5346<!-- _______________________________________________________________________ -->
5347<div class="doc_subsubsection">
5348 <a name="i_va_arg">'<tt>va_arg</tt>' Instruction</a>
5349</div>
5350
5351<div class="doc_text">
5352
5353<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5354<pre>
5355 &lt;resultval&gt; = va_arg &lt;va_list*&gt; &lt;arglist&gt;, &lt;argty&gt;
5356</pre>
5357
5358<h5>Overview:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5359<p>The '<tt>va_arg</tt>' instruction is used to access arguments passed through
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5360 the "variable argument" area of a function call. It is used to implement the
5361 <tt>va_arg</tt> macro in C.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5362
5363<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5364<p>This instruction takes a <tt>va_list*</tt> value and the type of the
5365 argument. It returns a value of the specified argument type and increments
5366 the <tt>va_list</tt> to point to the next argument. The actual type
5367 of <tt>va_list</tt> is target specific.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5368
5369<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5370<p>The '<tt>va_arg</tt>' instruction loads an argument of the specified type
5371 from the specified <tt>va_list</tt> and causes the <tt>va_list</tt> to point
5372 to the next argument. For more information, see the variable argument
5373 handling <a href="#int_varargs">Intrinsic Functions</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5374
5375<p>It is legal for this instruction to be called in a function which does not
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5376 take a variable number of arguments, for example, the <tt>vfprintf</tt>
5377 function.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5378
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5379<p><tt>va_arg</tt> is an LLVM instruction instead of
5380 an <a href="#intrinsics">intrinsic function</a> because it takes a type as an
5381 argument.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5382
5383<h5>Example:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5384<p>See the <a href="#int_varargs">variable argument processing</a> section.</p>
5385
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5386<p>Note that the code generator does not yet fully support va_arg on many
5387 targets. Also, it does not currently support va_arg with aggregate types on
5388 any target.</p>
Dan Gohman60967192009-01-12 23:12:39 +0000[diff] [blame]5389
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5390</div>
5391
5392<!-- *********************************************************************** -->
5393<div class="doc_section"> <a name="intrinsics">Intrinsic Functions</a> </div>
5394<!-- *********************************************************************** -->
5395
5396<div class="doc_text">
5397
5398<p>LLVM supports the notion of an "intrinsic function". These functions have
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5399 well known names and semantics and are required to follow certain
5400 restrictions. Overall, these intrinsics represent an extension mechanism for
5401 the LLVM language that does not require changing all of the transformations
5402 in LLVM when adding to the language (or the bitcode reader/writer, the
5403 parser, etc...).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5404
5405<p>Intrinsic function names must all start with an "<tt>llvm.</tt>" prefix. This
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5406 prefix is reserved in LLVM for intrinsic names; thus, function names may not
5407 begin with this prefix. Intrinsic functions must always be external
5408 functions: you cannot define the body of intrinsic functions. Intrinsic
5409 functions may only be used in call or invoke instructions: it is illegal to
5410 take the address of an intrinsic function. Additionally, because intrinsic
5411 functions are part of the LLVM language, it is required if any are added that
5412 they be documented here.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5413
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5414<p>Some intrinsic functions can be overloaded, i.e., the intrinsic represents a
5415 family of functions that perform the same operation but on different data
5416 types. Because LLVM can represent over 8 million different integer types,
5417 overloading is used commonly to allow an intrinsic function to operate on any
5418 integer type. One or more of the argument types or the result type can be
5419 overloaded to accept any integer type. Argument types may also be defined as
5420 exactly matching a previous argument's type or the result type. This allows
5421 an intrinsic function which accepts multiple arguments, but needs all of them
5422 to be of the same type, to only be overloaded with respect to a single
5423 argument or the result.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5424
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5425<p>Overloaded intrinsics will have the names of its overloaded argument types
5426 encoded into its function name, each preceded by a period. Only those types
5427 which are overloaded result in a name suffix. Arguments whose type is matched
5428 against another type do not. For example, the <tt>llvm.ctpop</tt> function
5429 can take an integer of any width and returns an integer of exactly the same
5430 integer width. This leads to a family of functions such as
5431 <tt>i8 @llvm.ctpop.i8(i8 %val)</tt> and <tt>i29 @llvm.ctpop.i29(i29
5432 %val)</tt>. Only one type, the return type, is overloaded, and only one type
5433 suffix is required. Because the argument's type is matched against the return
5434 type, it does not require its own name suffix.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5435
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]5436<p>To learn how to add an intrinsic function, please see the
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5437 <a href="ExtendingLLVM.html">Extending LLVM Guide</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5438
5439</div>
5440
5441<!-- ======================================================================= -->
5442<div class="doc_subsection">
5443 <a name="int_varargs">Variable Argument Handling Intrinsics</a>
5444</div>
5445
5446<div class="doc_text">
5447
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5448<p>Variable argument support is defined in LLVM with
5449 the <a href="#i_va_arg"><tt>va_arg</tt></a> instruction and these three
5450 intrinsic functions. These functions are related to the similarly named
5451 macros defined in the <tt>&lt;stdarg.h&gt;</tt> header file.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5452
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5453<p>All of these functions operate on arguments that use a target-specific value
5454 type "<tt>va_list</tt>". The LLVM assembly language reference manual does
5455 not define what this type is, so all transformations should be prepared to
5456 handle these functions regardless of the type used.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5457
5458<p>This example shows how the <a href="#i_va_arg"><tt>va_arg</tt></a>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5459 instruction and the variable argument handling intrinsic functions are
5460 used.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5461
5462<div class="doc_code">
5463<pre>
5464define i32 @test(i32 %X, ...) {
5465 ; Initialize variable argument processing
5466 %ap = alloca i8*
5467 %ap2 = bitcast i8** %ap to i8*
5468 call void @llvm.va_start(i8* %ap2)
5469
5470 ; Read a single integer argument
5471 %tmp = va_arg i8** %ap, i32
5472
5473 ; Demonstrate usage of llvm.va_copy and llvm.va_end
5474 %aq = alloca i8*
5475 %aq2 = bitcast i8** %aq to i8*
5476 call void @llvm.va_copy(i8* %aq2, i8* %ap2)
5477 call void @llvm.va_end(i8* %aq2)
5478
5479 ; Stop processing of arguments.
5480 call void @llvm.va_end(i8* %ap2)
5481 ret i32 %tmp
5482}
5483
5484declare void @llvm.va_start(i8*)
5485declare void @llvm.va_copy(i8*, i8*)
5486declare void @llvm.va_end(i8*)
5487</pre>
5488</div>
5489
5490</div>
5491
5492<!-- _______________________________________________________________________ -->
5493<div class="doc_subsubsection">
5494 <a name="int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a>
5495</div>
5496
5497
5498<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5499
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5500<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5501<pre>
5502 declare void %llvm.va_start(i8* &lt;arglist&gt;)
5503</pre>
5504
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5505<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5506<p>The '<tt>llvm.va_start</tt>' intrinsic initializes <tt>*&lt;arglist&gt;</tt>
5507 for subsequent use by <tt><a href="#i_va_arg">va_arg</a></tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5508
5509<h5>Arguments:</h5>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]5510<p>The argument is a pointer to a <tt>va_list</tt> element to initialize.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5511
5512<h5>Semantics:</h5>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]5513<p>The '<tt>llvm.va_start</tt>' intrinsic works just like the <tt>va_start</tt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5514 macro available in C. In a target-dependent way, it initializes
5515 the <tt>va_list</tt> element to which the argument points, so that the next
5516 call to <tt>va_arg</tt> will produce the first variable argument passed to
5517 the function. Unlike the C <tt>va_start</tt> macro, this intrinsic does not
5518 need to know the last argument of the function as the compiler can figure
5519 that out.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5520
5521</div>
5522
5523<!-- _______________________________________________________________________ -->
5524<div class="doc_subsubsection">
5525 <a name="int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a>
5526</div>
5527
5528<div class="doc_text">
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5529
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5530<h5>Syntax:</h5>
5531<pre>
5532 declare void @llvm.va_end(i8* &lt;arglist&gt;)
5533</pre>
5534
5535<h5>Overview:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5536<p>The '<tt>llvm.va_end</tt>' intrinsic destroys <tt>*&lt;arglist&gt;</tt>,
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5537 which has been initialized previously
5538 with <tt><a href="#int_va_start">llvm.va_start</a></tt>
5539 or <tt><a href="#i_va_copy">llvm.va_copy</a></tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5540
5541<h5>Arguments:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5542<p>The argument is a pointer to a <tt>va_list</tt> to destroy.</p>
5543
5544<h5>Semantics:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5545<p>The '<tt>llvm.va_end</tt>' intrinsic works just like the <tt>va_end</tt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5546 macro available in C. In a target-dependent way, it destroys
5547 the <tt>va_list</tt> element to which the argument points. Calls
5548 to <a href="#int_va_start"><tt>llvm.va_start</tt></a>
5549 and <a href="#int_va_copy"> <tt>llvm.va_copy</tt></a> must be matched exactly
5550 with calls to <tt>llvm.va_end</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5551
5552</div>
5553
5554<!-- _______________________________________________________________________ -->
5555<div class="doc_subsubsection">
5556 <a name="int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a>
5557</div>
5558
5559<div class="doc_text">
5560
5561<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5562<pre>
5563 declare void @llvm.va_copy(i8* &lt;destarglist&gt;, i8* &lt;srcarglist&gt;)
5564</pre>
5565
5566<h5>Overview:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5567<p>The '<tt>llvm.va_copy</tt>' intrinsic copies the current argument position
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5568 from the source argument list to the destination argument list.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5569
5570<h5>Arguments:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5571<p>The first argument is a pointer to a <tt>va_list</tt> element to initialize.
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5572 The second argument is a pointer to a <tt>va_list</tt> element to copy
5573 from.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5574
5575<h5>Semantics:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5576<p>The '<tt>llvm.va_copy</tt>' intrinsic works just like the <tt>va_copy</tt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5577 macro available in C. In a target-dependent way, it copies the
5578 source <tt>va_list</tt> element into the destination <tt>va_list</tt>
5579 element. This intrinsic is necessary because
5580 the <tt><a href="#int_va_start"> llvm.va_start</a></tt> intrinsic may be
5581 arbitrarily complex and require, for example, memory allocation.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5582
5583</div>
5584
5585<!-- ======================================================================= -->
5586<div class="doc_subsection">
5587 <a name="int_gc">Accurate Garbage Collection Intrinsics</a>
5588</div>
5589
5590<div class="doc_text">
5591
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5592<p>LLVM support for <a href="GarbageCollection.html">Accurate Garbage
Chris Lattner96451482008-08-05 18:29:16 +0000[diff] [blame]5593Collection</a> (GC) requires the implementation and generation of these
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5594intrinsics. These intrinsics allow identification of <a href="#int_gcroot">GC
5595roots on the stack</a>, as well as garbage collector implementations that
5596require <a href="#int_gcread">read</a> and <a href="#int_gcwrite">write</a>
5597barriers. Front-ends for type-safe garbage collected languages should generate
5598these intrinsics to make use of the LLVM garbage collectors. For more details,
5599see <a href="GarbageCollection.html">Accurate Garbage Collection with
5600LLVM</a>.</p>
Christopher Lambcfe00962007-12-17 01:00:21 +0000[diff] [blame]5601
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5602<p>The garbage collection intrinsics only operate on objects in the generic
5603 address space (address space zero).</p>
Christopher Lambcfe00962007-12-17 01:00:21 +0000[diff] [blame]5604
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5605</div>
5606
5607<!-- _______________________________________________________________________ -->
5608<div class="doc_subsubsection">
5609 <a name="int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a>
5610</div>
5611
5612<div class="doc_text">
5613
5614<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5615<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]5616 declare void @llvm.gcroot(i8** %ptrloc, i8* %metadata)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5617</pre>
5618
5619<h5>Overview:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5620<p>The '<tt>llvm.gcroot</tt>' intrinsic declares the existence of a GC root to
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5621 the code generator, and allows some metadata to be associated with it.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5622
5623<h5>Arguments:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5624<p>The first argument specifies the address of a stack object that contains the
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5625 root pointer. The second pointer (which must be either a constant or a
5626 global value address) contains the meta-data to be associated with the
5627 root.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5628
5629<h5>Semantics:</h5>
Chris Lattnera7d94ba2008-04-24 05:59:56 +0000[diff] [blame]5630<p>At runtime, a call to this intrinsic stores a null pointer into the "ptrloc"
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5631 location. At compile-time, the code generator generates information to allow
5632 the runtime to find the pointer at GC safe points. The '<tt>llvm.gcroot</tt>'
5633 intrinsic may only be used in a function which <a href="#gc">specifies a GC
5634 algorithm</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5635
5636</div>
5637
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5638<!-- _______________________________________________________________________ -->
5639<div class="doc_subsubsection">
5640 <a name="int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a>
5641</div>
5642
5643<div class="doc_text">
5644
5645<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5646<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]5647 declare i8* @llvm.gcread(i8* %ObjPtr, i8** %Ptr)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5648</pre>
5649
5650<h5>Overview:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5651<p>The '<tt>llvm.gcread</tt>' intrinsic identifies reads of references from heap
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5652 locations, allowing garbage collector implementations that require read
5653 barriers.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5654
5655<h5>Arguments:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5656<p>The second argument is the address to read from, which should be an address
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5657 allocated from the garbage collector. The first object is a pointer to the
5658 start of the referenced object, if needed by the language runtime (otherwise
5659 null).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5660
5661<h5>Semantics:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5662<p>The '<tt>llvm.gcread</tt>' intrinsic has the same semantics as a load
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5663 instruction, but may be replaced with substantially more complex code by the
5664 garbage collector runtime, as needed. The '<tt>llvm.gcread</tt>' intrinsic
5665 may only be used in a function which <a href="#gc">specifies a GC
5666 algorithm</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5667
5668</div>
5669
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5670<!-- _______________________________________________________________________ -->
5671<div class="doc_subsubsection">
5672 <a name="int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a>
5673</div>
5674
5675<div class="doc_text">
5676
5677<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5678<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]5679 declare void @llvm.gcwrite(i8* %P1, i8* %Obj, i8** %P2)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5680</pre>
5681
5682<h5>Overview:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5683<p>The '<tt>llvm.gcwrite</tt>' intrinsic identifies writes of references to heap
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5684 locations, allowing garbage collector implementations that require write
5685 barriers (such as generational or reference counting collectors).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5686
5687<h5>Arguments:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5688<p>The first argument is the reference to store, the second is the start of the
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5689 object to store it to, and the third is the address of the field of Obj to
5690 store to. If the runtime does not require a pointer to the object, Obj may
5691 be null.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5692
5693<h5>Semantics:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5694<p>The '<tt>llvm.gcwrite</tt>' intrinsic has the same semantics as a store
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5695 instruction, but may be replaced with substantially more complex code by the
5696 garbage collector runtime, as needed. The '<tt>llvm.gcwrite</tt>' intrinsic
5697 may only be used in a function which <a href="#gc">specifies a GC
5698 algorithm</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5699
5700</div>
5701
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5702<!-- ======================================================================= -->
5703<div class="doc_subsection">
5704 <a name="int_codegen">Code Generator Intrinsics</a>
5705</div>
5706
5707<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5708
5709<p>These intrinsics are provided by LLVM to expose special features that may
5710 only be implemented with code generator support.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5711
5712</div>
5713
5714<!-- _______________________________________________________________________ -->
5715<div class="doc_subsubsection">
5716 <a name="int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a>
5717</div>
5718
5719<div class="doc_text">
5720
5721<h5>Syntax:</h5>
5722<pre>
5723 declare i8 *@llvm.returnaddress(i32 &lt;level&gt;)
5724</pre>
5725
5726<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5727<p>The '<tt>llvm.returnaddress</tt>' intrinsic attempts to compute a
5728 target-specific value indicating the return address of the current function
5729 or one of its callers.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5730
5731<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5732<p>The argument to this intrinsic indicates which function to return the address
5733 for. Zero indicates the calling function, one indicates its caller, etc.
5734 The argument is <b>required</b> to be a constant integer value.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5735
5736<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5737<p>The '<tt>llvm.returnaddress</tt>' intrinsic either returns a pointer
5738 indicating the return address of the specified call frame, or zero if it
5739 cannot be identified. The value returned by this intrinsic is likely to be
5740 incorrect or 0 for arguments other than zero, so it should only be used for
5741 debugging purposes.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5742
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5743<p>Note that calling this intrinsic does not prevent function inlining or other
5744 aggressive transformations, so the value returned may not be that of the
5745 obvious source-language caller.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5746
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5747</div>
5748
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5749<!-- _______________________________________________________________________ -->
5750<div class="doc_subsubsection">
5751 <a name="int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a>
5752</div>
5753
5754<div class="doc_text">
5755
5756<h5>Syntax:</h5>
5757<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]5758 declare i8 *@llvm.frameaddress(i32 &lt;level&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5759</pre>
5760
5761<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5762<p>The '<tt>llvm.frameaddress</tt>' intrinsic attempts to return the
5763 target-specific frame pointer value for the specified stack frame.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5764
5765<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5766<p>The argument to this intrinsic indicates which function to return the frame
5767 pointer for. Zero indicates the calling function, one indicates its caller,
5768 etc. The argument is <b>required</b> to be a constant integer value.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5769
5770<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5771<p>The '<tt>llvm.frameaddress</tt>' intrinsic either returns a pointer
5772 indicating the frame address of the specified call frame, or zero if it
5773 cannot be identified. The value returned by this intrinsic is likely to be
5774 incorrect or 0 for arguments other than zero, so it should only be used for
5775 debugging purposes.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5776
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5777<p>Note that calling this intrinsic does not prevent function inlining or other
5778 aggressive transformations, so the value returned may not be that of the
5779 obvious source-language caller.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5780
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5781</div>
5782
5783<!-- _______________________________________________________________________ -->
5784<div class="doc_subsubsection">
5785 <a name="int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a>
5786</div>
5787
5788<div class="doc_text">
5789
5790<h5>Syntax:</h5>
5791<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]5792 declare i8 *@llvm.stacksave()
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5793</pre>
5794
5795<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5796<p>The '<tt>llvm.stacksave</tt>' intrinsic is used to remember the current state
5797 of the function stack, for use
5798 with <a href="#int_stackrestore"> <tt>llvm.stackrestore</tt></a>. This is
5799 useful for implementing language features like scoped automatic variable
5800 sized arrays in C99.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5801
5802<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5803<p>This intrinsic returns a opaque pointer value that can be passed
5804 to <a href="#int_stackrestore"><tt>llvm.stackrestore</tt></a>. When
5805 an <tt>llvm.stackrestore</tt> intrinsic is executed with a value saved
5806 from <tt>llvm.stacksave</tt>, it effectively restores the state of the stack
5807 to the state it was in when the <tt>llvm.stacksave</tt> intrinsic executed.
5808 In practice, this pops any <a href="#i_alloca">alloca</a> blocks from the
5809 stack that were allocated after the <tt>llvm.stacksave</tt> was executed.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5810
5811</div>
5812
5813<!-- _______________________________________________________________________ -->
5814<div class="doc_subsubsection">
5815 <a name="int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a>
5816</div>
5817
5818<div class="doc_text">
5819
5820<h5>Syntax:</h5>
5821<pre>
5822 declare void @llvm.stackrestore(i8 * %ptr)
5823</pre>
5824
5825<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5826<p>The '<tt>llvm.stackrestore</tt>' intrinsic is used to restore the state of
5827 the function stack to the state it was in when the
5828 corresponding <a href="#int_stacksave"><tt>llvm.stacksave</tt></a> intrinsic
5829 executed. This is useful for implementing language features like scoped
5830 automatic variable sized arrays in C99.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5831
5832<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5833<p>See the description
5834 for <a href="#int_stacksave"><tt>llvm.stacksave</tt></a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5835
5836</div>
5837
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5838<!-- _______________________________________________________________________ -->
5839<div class="doc_subsubsection">
5840 <a name="int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a>
5841</div>
5842
5843<div class="doc_text">
5844
5845<h5>Syntax:</h5>
5846<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]5847 declare void @llvm.prefetch(i8* &lt;address&gt;, i32 &lt;rw&gt;, i32 &lt;locality&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5848</pre>
5849
5850<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5851<p>The '<tt>llvm.prefetch</tt>' intrinsic is a hint to the code generator to
5852 insert a prefetch instruction if supported; otherwise, it is a noop.
5853 Prefetches have no effect on the behavior of the program but can change its
5854 performance characteristics.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5855
5856<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5857<p><tt>address</tt> is the address to be prefetched, <tt>rw</tt> is the
5858 specifier determining if the fetch should be for a read (0) or write (1),
5859 and <tt>locality</tt> is a temporal locality specifier ranging from (0) - no
5860 locality, to (3) - extremely local keep in cache. The <tt>rw</tt>
5861 and <tt>locality</tt> arguments must be constant integers.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5862
5863<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5864<p>This intrinsic does not modify the behavior of the program. In particular,
5865 prefetches cannot trap and do not produce a value. On targets that support
5866 this intrinsic, the prefetch can provide hints to the processor cache for
5867 better performance.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5868
5869</div>
5870
5871<!-- _______________________________________________________________________ -->
5872<div class="doc_subsubsection">
5873 <a name="int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a>
5874</div>
5875
5876<div class="doc_text">
5877
5878<h5>Syntax:</h5>
5879<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]5880 declare void @llvm.pcmarker(i32 &lt;id&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5881</pre>
5882
5883<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5884<p>The '<tt>llvm.pcmarker</tt>' intrinsic is a method to export a Program
5885 Counter (PC) in a region of code to simulators and other tools. The method
5886 is target specific, but it is expected that the marker will use exported
5887 symbols to transmit the PC of the marker. The marker makes no guarantees
5888 that it will remain with any specific instruction after optimizations. It is
5889 possible that the presence of a marker will inhibit optimizations. The
5890 intended use is to be inserted after optimizations to allow correlations of
5891 simulation runs.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5892
5893<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5894<p><tt>id</tt> is a numerical id identifying the marker.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5895
5896<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5897<p>This intrinsic does not modify the behavior of the program. Backends that do
Dan Gohman22dc6682010-03-01 17:41:39 +0000[diff] [blame]5898 not support this intrinsic may ignore it.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5899
5900</div>
5901
5902<!-- _______________________________________________________________________ -->
5903<div class="doc_subsubsection">
5904 <a name="int_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a>
5905</div>
5906
5907<div class="doc_text">
5908
5909<h5>Syntax:</h5>
5910<pre>
5911 declare i64 @llvm.readcyclecounter( )
5912</pre>
5913
5914<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5915<p>The '<tt>llvm.readcyclecounter</tt>' intrinsic provides access to the cycle
5916 counter register (or similar low latency, high accuracy clocks) on those
5917 targets that support it. On X86, it should map to RDTSC. On Alpha, it
5918 should map to RPCC. As the backing counters overflow quickly (on the order
5919 of 9 seconds on alpha), this should only be used for small timings.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5920
5921<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5922<p>When directly supported, reading the cycle counter should not modify any
5923 memory. Implementations are allowed to either return a application specific
5924 value or a system wide value. On backends without support, this is lowered
5925 to a constant 0.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5926
5927</div>
5928
5929<!-- ======================================================================= -->
5930<div class="doc_subsection">
5931 <a name="int_libc">Standard C Library Intrinsics</a>
5932</div>
5933
5934<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5935
5936<p>LLVM provides intrinsics for a few important standard C library functions.
5937 These intrinsics allow source-language front-ends to pass information about
5938 the alignment of the pointer arguments to the code generator, providing
5939 opportunity for more efficient code generation.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5940
5941</div>
5942
5943<!-- _______________________________________________________________________ -->
5944<div class="doc_subsubsection">
5945 <a name="int_memcpy">'<tt>llvm.memcpy</tt>' Intrinsic</a>
5946</div>
5947
5948<div class="doc_text">
5949
5950<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5951<p>This is an overloaded intrinsic. You can use <tt>llvm.memcpy</tt> on any
Mon P Wang238462c2010-04-07 06:35:53 +0000[diff] [blame]5952 integer bit width and for different address spaces. Not all targets support
5953 all bit widths however.</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5954
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5955<pre>
Chris Lattner5aa5aab2010-04-08 00:53:57 +0000[diff] [blame]5956 declare void @llvm.memcpy.p0i8.p0i8.i32(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
5957 i32 &lt;len&gt;, i32 &lt;align&gt;, i1 &lt;isvolatile&gt;)
5958 declare void @llvm.memcpy.p0i8.p0i8.i64(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
5959 i64 &lt;len&gt;, i32 &lt;align&gt;, i1 &lt;isvolatile&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5960</pre>
5961
5962<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5963<p>The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the
5964 source location to the destination location.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5965
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5966<p>Note that, unlike the standard libc function, the <tt>llvm.memcpy.*</tt>
Chris Lattner5aa5aab2010-04-08 00:53:57 +0000[diff] [blame]5967 intrinsics do not return a value, takes extra alignment/isvolatile arguments
5968 and the pointers can be in specified address spaces.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5969
5970<h5>Arguments:</h5>
Chris Lattner5aa5aab2010-04-08 00:53:57 +0000[diff] [blame]5971
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5972<p>The first argument is a pointer to the destination, the second is a pointer
5973 to the source. The third argument is an integer argument specifying the
Chris Lattner5aa5aab2010-04-08 00:53:57 +0000[diff] [blame]5974 number of bytes to copy, the fourth argument is the alignment of the
5975 source and destination locations, and the fifth is a boolean indicating a
5976 volatile access.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5977
Dan Gohman22dc6682010-03-01 17:41:39 +0000[diff] [blame]5978<p>If the call to this intrinsic has an alignment value that is not 0 or 1,
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5979 then the caller guarantees that both the source and destination pointers are
5980 aligned to that boundary.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5981
Jeffrey Yasskin6dc467f2010-04-26 21:21:24 +0000[diff] [blame]5982<p>If the <tt>isvolatile</tt> parameter is <tt>true</tt>, the
5983 <tt>llvm.memcpy</tt> call is a <a href="#volatile">volatile operation</a>.
5984 The detailed access behavior is not very cleanly specified and it is unwise
5985 to depend on it.</p>
Chris Lattner5aa5aab2010-04-08 00:53:57 +0000[diff] [blame]5986
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5987<h5>Semantics:</h5>
Chris Lattner5aa5aab2010-04-08 00:53:57 +0000[diff] [blame]5988
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5989<p>The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the
5990 source location to the destination location, which are not allowed to
5991 overlap. It copies "len" bytes of memory over. If the argument is known to
5992 be aligned to some boundary, this can be specified as the fourth argument,
5993 otherwise it should be set to 0 or 1.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5994
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5995</div>
5996
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5997<!-- _______________________________________________________________________ -->
5998<div class="doc_subsubsection">
5999 <a name="int_memmove">'<tt>llvm.memmove</tt>' Intrinsic</a>
6000</div>
6001
6002<div class="doc_text">
6003
6004<h5>Syntax:</h5>
Chris Lattner82c2e432008-11-21 16:42:48 +0000[diff] [blame]6005<p>This is an overloaded intrinsic. You can use llvm.memmove on any integer bit
Mon P Wang238462c2010-04-07 06:35:53 +0000[diff] [blame]6006 width and for different address space. Not all targets support all bit
6007 widths however.</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6008
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6009<pre>
Chris Lattner5aa5aab2010-04-08 00:53:57 +0000[diff] [blame]6010 declare void @llvm.memmove.p0i8.p0i8.i32(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
6011 i32 &lt;len&gt;, i32 &lt;align&gt;, i1 &lt;isvolatile&gt;)
6012 declare void @llvm.memmove.p0i8.p0i8.i64(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
6013 i64 &lt;len&gt;, i32 &lt;align&gt;, i1 &lt;isvolatile&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6014</pre>
6015
6016<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6017<p>The '<tt>llvm.memmove.*</tt>' intrinsics move a block of memory from the
6018 source location to the destination location. It is similar to the
6019 '<tt>llvm.memcpy</tt>' intrinsic but allows the two memory locations to
6020 overlap.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6021
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6022<p>Note that, unlike the standard libc function, the <tt>llvm.memmove.*</tt>
Chris Lattner5aa5aab2010-04-08 00:53:57 +0000[diff] [blame]6023 intrinsics do not return a value, takes extra alignment/isvolatile arguments
6024 and the pointers can be in specified address spaces.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6025
6026<h5>Arguments:</h5>
Chris Lattner5aa5aab2010-04-08 00:53:57 +0000[diff] [blame]6027
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6028<p>The first argument is a pointer to the destination, the second is a pointer
6029 to the source. The third argument is an integer argument specifying the
Chris Lattner5aa5aab2010-04-08 00:53:57 +0000[diff] [blame]6030 number of bytes to copy, the fourth argument is the alignment of the
6031 source and destination locations, and the fifth is a boolean indicating a
6032 volatile access.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6033
Dan Gohman22dc6682010-03-01 17:41:39 +0000[diff] [blame]6034<p>If the call to this intrinsic has an alignment value that is not 0 or 1,
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6035 then the caller guarantees that the source and destination pointers are
6036 aligned to that boundary.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6037
Jeffrey Yasskin6dc467f2010-04-26 21:21:24 +0000[diff] [blame]6038<p>If the <tt>isvolatile</tt> parameter is <tt>true</tt>, the
6039 <tt>llvm.memmove</tt> call is a <a href="#volatile">volatile operation</a>.
6040 The detailed access behavior is not very cleanly specified and it is unwise
6041 to depend on it.</p>
Chris Lattner5aa5aab2010-04-08 00:53:57 +0000[diff] [blame]6042
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6043<h5>Semantics:</h5>
Chris Lattner5aa5aab2010-04-08 00:53:57 +0000[diff] [blame]6044
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6045<p>The '<tt>llvm.memmove.*</tt>' intrinsics copy a block of memory from the
6046 source location to the destination location, which may overlap. It copies
6047 "len" bytes of memory over. If the argument is known to be aligned to some
6048 boundary, this can be specified as the fourth argument, otherwise it should
6049 be set to 0 or 1.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6050
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6051</div>
6052
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6053<!-- _______________________________________________________________________ -->
6054<div class="doc_subsubsection">
6055 <a name="int_memset">'<tt>llvm.memset.*</tt>' Intrinsics</a>
6056</div>
6057
6058<div class="doc_text">
6059
6060<h5>Syntax:</h5>
Chris Lattner82c2e432008-11-21 16:42:48 +0000[diff] [blame]6061<p>This is an overloaded intrinsic. You can use llvm.memset on any integer bit
Mon P Wang238462c2010-04-07 06:35:53 +0000[diff] [blame]6062 width and for different address spaces. Not all targets support all bit
6063 widths however.</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6064
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6065<pre>
Chris Lattner5aa5aab2010-04-08 00:53:57 +0000[diff] [blame]6066 declare void @llvm.memset.p0i8.i32(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
Chris Lattnerc0dfafe2010-04-08 00:54:34 +0000[diff] [blame]6067 i32 &lt;len&gt;, i32 &lt;align&gt;, i1 &lt;isvolatile&gt;)
Chris Lattner5aa5aab2010-04-08 00:53:57 +0000[diff] [blame]6068 declare void @llvm.memset.p0i8.i64(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
Chris Lattnerc0dfafe2010-04-08 00:54:34 +0000[diff] [blame]6069 i64 &lt;len&gt;, i32 &lt;align&gt;, i1 &lt;isvolatile&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6070</pre>
6071
6072<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6073<p>The '<tt>llvm.memset.*</tt>' intrinsics fill a block of memory with a
6074 particular byte value.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6075
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6076<p>Note that, unlike the standard libc function, the <tt>llvm.memset</tt>
Chris Lattner5aa5aab2010-04-08 00:53:57 +0000[diff] [blame]6077 intrinsic does not return a value, takes extra alignment/volatile arguments,
6078 and the destination can be in an arbitrary address space.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6079
6080<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6081<p>The first argument is a pointer to the destination to fill, the second is the
6082 byte value to fill it with, the third argument is an integer argument
6083 specifying the number of bytes to fill, and the fourth argument is the known
6084 alignment of destination location.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6085
Dan Gohman22dc6682010-03-01 17:41:39 +0000[diff] [blame]6086<p>If the call to this intrinsic has an alignment value that is not 0 or 1,
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6087 then the caller guarantees that the destination pointer is aligned to that
6088 boundary.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6089
Jeffrey Yasskin6dc467f2010-04-26 21:21:24 +0000[diff] [blame]6090<p>If the <tt>isvolatile</tt> parameter is <tt>true</tt>, the
6091 <tt>llvm.memset</tt> call is a <a href="#volatile">volatile operation</a>.
6092 The detailed access behavior is not very cleanly specified and it is unwise
6093 to depend on it.</p>
Chris Lattner5aa5aab2010-04-08 00:53:57 +0000[diff] [blame]6094
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6095<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6096<p>The '<tt>llvm.memset.*</tt>' intrinsics fill "len" bytes of memory starting
6097 at the destination location. If the argument is known to be aligned to some
6098 boundary, this can be specified as the fourth argument, otherwise it should
6099 be set to 0 or 1.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6100
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6101</div>
6102
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6103<!-- _______________________________________________________________________ -->
6104<div class="doc_subsubsection">
6105 <a name="int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a>
6106</div>
6107
6108<div class="doc_text">
6109
6110<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6111<p>This is an overloaded intrinsic. You can use <tt>llvm.sqrt</tt> on any
6112 floating point or vector of floating point type. Not all targets support all
6113 types however.</p>
6114
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6115<pre>
Dale Johannesenf9adbb62007-10-02 17:47:38 +0000[diff] [blame]6116 declare float @llvm.sqrt.f32(float %Val)
6117 declare double @llvm.sqrt.f64(double %Val)
6118 declare x86_fp80 @llvm.sqrt.f80(x86_fp80 %Val)
6119 declare fp128 @llvm.sqrt.f128(fp128 %Val)
6120 declare ppc_fp128 @llvm.sqrt.ppcf128(ppc_fp128 %Val)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6121</pre>
6122
6123<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6124<p>The '<tt>llvm.sqrt</tt>' intrinsics return the sqrt of the specified operand,
6125 returning the same value as the libm '<tt>sqrt</tt>' functions would.
6126 Unlike <tt>sqrt</tt> in libm, however, <tt>llvm.sqrt</tt> has undefined
6127 behavior for negative numbers other than -0.0 (which allows for better
6128 optimization, because there is no need to worry about errno being
6129 set). <tt>llvm.sqrt(-0.0)</tt> is defined to return -0.0 like IEEE sqrt.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6130
6131<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6132<p>The argument and return value are floating point numbers of the same
6133 type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6134
6135<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6136<p>This function returns the sqrt of the specified operand if it is a
6137 nonnegative floating point number.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6138
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6139</div>
6140
6141<!-- _______________________________________________________________________ -->
6142<div class="doc_subsubsection">
6143 <a name="int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a>
6144</div>
6145
6146<div class="doc_text">
6147
6148<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6149<p>This is an overloaded intrinsic. You can use <tt>llvm.powi</tt> on any
6150 floating point or vector of floating point type. Not all targets support all
6151 types however.</p>
6152
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6153<pre>
Dale Johannesenf9adbb62007-10-02 17:47:38 +0000[diff] [blame]6154 declare float @llvm.powi.f32(float %Val, i32 %power)
6155 declare double @llvm.powi.f64(double %Val, i32 %power)
6156 declare x86_fp80 @llvm.powi.f80(x86_fp80 %Val, i32 %power)
6157 declare fp128 @llvm.powi.f128(fp128 %Val, i32 %power)
6158 declare ppc_fp128 @llvm.powi.ppcf128(ppc_fp128 %Val, i32 %power)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6159</pre>
6160
6161<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6162<p>The '<tt>llvm.powi.*</tt>' intrinsics return the first operand raised to the
6163 specified (positive or negative) power. The order of evaluation of
6164 multiplications is not defined. When a vector of floating point type is
6165 used, the second argument remains a scalar integer value.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6166
6167<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6168<p>The second argument is an integer power, and the first is a value to raise to
6169 that power.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6170
6171<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6172<p>This function returns the first value raised to the second power with an
6173 unspecified sequence of rounding operations.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6174
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6175</div>
6176
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]6177<!-- _______________________________________________________________________ -->
6178<div class="doc_subsubsection">
6179 <a name="int_sin">'<tt>llvm.sin.*</tt>' Intrinsic</a>
6180</div>
6181
6182<div class="doc_text">
6183
6184<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6185<p>This is an overloaded intrinsic. You can use <tt>llvm.sin</tt> on any
6186 floating point or vector of floating point type. Not all targets support all
6187 types however.</p>
6188
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]6189<pre>
6190 declare float @llvm.sin.f32(float %Val)
6191 declare double @llvm.sin.f64(double %Val)
6192 declare x86_fp80 @llvm.sin.f80(x86_fp80 %Val)
6193 declare fp128 @llvm.sin.f128(fp128 %Val)
6194 declare ppc_fp128 @llvm.sin.ppcf128(ppc_fp128 %Val)
6195</pre>
6196
6197<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6198<p>The '<tt>llvm.sin.*</tt>' intrinsics return the sine of the operand.</p>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]6199
6200<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6201<p>The argument and return value are floating point numbers of the same
6202 type.</p>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]6203
6204<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6205<p>This function returns the sine of the specified operand, returning the same
6206 values as the libm <tt>sin</tt> functions would, and handles error conditions
6207 in the same way.</p>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]6208
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]6209</div>
6210
6211<!-- _______________________________________________________________________ -->
6212<div class="doc_subsubsection">
6213 <a name="int_cos">'<tt>llvm.cos.*</tt>' Intrinsic</a>
6214</div>
6215
6216<div class="doc_text">
6217
6218<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6219<p>This is an overloaded intrinsic. You can use <tt>llvm.cos</tt> on any
6220 floating point or vector of floating point type. Not all targets support all
6221 types however.</p>
6222
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]6223<pre>
6224 declare float @llvm.cos.f32(float %Val)
6225 declare double @llvm.cos.f64(double %Val)
6226 declare x86_fp80 @llvm.cos.f80(x86_fp80 %Val)
6227 declare fp128 @llvm.cos.f128(fp128 %Val)
6228 declare ppc_fp128 @llvm.cos.ppcf128(ppc_fp128 %Val)
6229</pre>
6230
6231<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6232<p>The '<tt>llvm.cos.*</tt>' intrinsics return the cosine of the operand.</p>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]6233
6234<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6235<p>The argument and return value are floating point numbers of the same
6236 type.</p>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]6237
6238<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6239<p>This function returns the cosine of the specified operand, returning the same
6240 values as the libm <tt>cos</tt> functions would, and handles error conditions
6241 in the same way.</p>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]6242
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]6243</div>
6244
6245<!-- _______________________________________________________________________ -->
6246<div class="doc_subsubsection">
6247 <a name="int_pow">'<tt>llvm.pow.*</tt>' Intrinsic</a>
6248</div>
6249
6250<div class="doc_text">
6251
6252<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6253<p>This is an overloaded intrinsic. You can use <tt>llvm.pow</tt> on any
6254 floating point or vector of floating point type. Not all targets support all
6255 types however.</p>
6256
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]6257<pre>
6258 declare float @llvm.pow.f32(float %Val, float %Power)
6259 declare double @llvm.pow.f64(double %Val, double %Power)
6260 declare x86_fp80 @llvm.pow.f80(x86_fp80 %Val, x86_fp80 %Power)
6261 declare fp128 @llvm.pow.f128(fp128 %Val, fp128 %Power)
6262 declare ppc_fp128 @llvm.pow.ppcf128(ppc_fp128 %Val, ppc_fp128 Power)
6263</pre>
6264
6265<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6266<p>The '<tt>llvm.pow.*</tt>' intrinsics return the first operand raised to the
6267 specified (positive or negative) power.</p>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]6268
6269<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6270<p>The second argument is a floating point power, and the first is a value to
6271 raise to that power.</p>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]6272
6273<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6274<p>This function returns the first value raised to the second power, returning
6275 the same values as the libm <tt>pow</tt> functions would, and handles error
6276 conditions in the same way.</p>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]6277
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]6278</div>
6279
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6280<!-- ======================================================================= -->
6281<div class="doc_subsection">
6282 <a name="int_manip">Bit Manipulation Intrinsics</a>
6283</div>
6284
6285<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6286
6287<p>LLVM provides intrinsics for a few important bit manipulation operations.
6288 These allow efficient code generation for some algorithms.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6289
6290</div>
6291
6292<!-- _______________________________________________________________________ -->
6293<div class="doc_subsubsection">
6294 <a name="int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a>
6295</div>
6296
6297<div class="doc_text">
6298
6299<h5>Syntax:</h5>
6300<p>This is an overloaded intrinsic function. You can use bswap on any integer
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6301 type that is an even number of bytes (i.e. BitWidth % 16 == 0).</p>
6302
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6303<pre>
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]6304 declare i16 @llvm.bswap.i16(i16 &lt;id&gt;)
6305 declare i32 @llvm.bswap.i32(i32 &lt;id&gt;)
6306 declare i64 @llvm.bswap.i64(i64 &lt;id&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6307</pre>
6308
6309<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6310<p>The '<tt>llvm.bswap</tt>' family of intrinsics is used to byte swap integer
6311 values with an even number of bytes (positive multiple of 16 bits). These
6312 are useful for performing operations on data that is not in the target's
6313 native byte order.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6314
6315<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6316<p>The <tt>llvm.bswap.i16</tt> intrinsic returns an i16 value that has the high
6317 and low byte of the input i16 swapped. Similarly,
6318 the <tt>llvm.bswap.i32</tt> intrinsic returns an i32 value that has the four
6319 bytes of the input i32 swapped, so that if the input bytes are numbered 0, 1,
6320 2, 3 then the returned i32 will have its bytes in 3, 2, 1, 0 order.
6321 The <tt>llvm.bswap.i48</tt>, <tt>llvm.bswap.i64</tt> and other intrinsics
6322 extend this concept to additional even-byte lengths (6 bytes, 8 bytes and
6323 more, respectively).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6324
6325</div>
6326
6327<!-- _______________________________________________________________________ -->
6328<div class="doc_subsubsection">
6329 <a name="int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic</a>
6330</div>
6331
6332<div class="doc_text">
6333
6334<h5>Syntax:</h5>
6335<p>This is an overloaded intrinsic. You can use llvm.ctpop on any integer bit
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6336 width. Not all targets support all bit widths however.</p>
6337
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6338<pre>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6339 declare i8 @llvm.ctpop.i8(i8 &lt;src&gt;)
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]6340 declare i16 @llvm.ctpop.i16(i16 &lt;src&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6341 declare i32 @llvm.ctpop.i32(i32 &lt;src&gt;)
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]6342 declare i64 @llvm.ctpop.i64(i64 &lt;src&gt;)
6343 declare i256 @llvm.ctpop.i256(i256 &lt;src&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6344</pre>
6345
6346<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6347<p>The '<tt>llvm.ctpop</tt>' family of intrinsics counts the number of bits set
6348 in a value.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6349
6350<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6351<p>The only argument is the value to be counted. The argument may be of any
6352 integer type. The return type must match the argument type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6353
6354<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6355<p>The '<tt>llvm.ctpop</tt>' intrinsic counts the 1's in a variable.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6356
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6357</div>
6358
6359<!-- _______________________________________________________________________ -->
6360<div class="doc_subsubsection">
6361 <a name="int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic</a>
6362</div>
6363
6364<div class="doc_text">
6365
6366<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6367<p>This is an overloaded intrinsic. You can use <tt>llvm.ctlz</tt> on any
6368 integer bit width. Not all targets support all bit widths however.</p>
6369
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6370<pre>
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]6371 declare i8 @llvm.ctlz.i8 (i8 &lt;src&gt;)
6372 declare i16 @llvm.ctlz.i16(i16 &lt;src&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6373 declare i32 @llvm.ctlz.i32(i32 &lt;src&gt;)
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]6374 declare i64 @llvm.ctlz.i64(i64 &lt;src&gt;)
6375 declare i256 @llvm.ctlz.i256(i256 &lt;src&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6376</pre>
6377
6378<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6379<p>The '<tt>llvm.ctlz</tt>' family of intrinsic functions counts the number of
6380 leading zeros in a variable.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6381
6382<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6383<p>The only argument is the value to be counted. The argument may be of any
6384 integer type. The return type must match the argument type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6385
6386<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6387<p>The '<tt>llvm.ctlz</tt>' intrinsic counts the leading (most significant)
6388 zeros in a variable. If the src == 0 then the result is the size in bits of
6389 the type of src. For example, <tt>llvm.ctlz(i32 2) = 30</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6390
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6391</div>
6392
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6393<!-- _______________________________________________________________________ -->
6394<div class="doc_subsubsection">
6395 <a name="int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic</a>
6396</div>
6397
6398<div class="doc_text">
6399
6400<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6401<p>This is an overloaded intrinsic. You can use <tt>llvm.cttz</tt> on any
6402 integer bit width. Not all targets support all bit widths however.</p>
6403
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6404<pre>
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]6405 declare i8 @llvm.cttz.i8 (i8 &lt;src&gt;)
6406 declare i16 @llvm.cttz.i16(i16 &lt;src&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6407 declare i32 @llvm.cttz.i32(i32 &lt;src&gt;)
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]6408 declare i64 @llvm.cttz.i64(i64 &lt;src&gt;)
6409 declare i256 @llvm.cttz.i256(i256 &lt;src&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6410</pre>
6411
6412<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6413<p>The '<tt>llvm.cttz</tt>' family of intrinsic functions counts the number of
6414 trailing zeros.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6415
6416<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6417<p>The only argument is the value to be counted. The argument may be of any
6418 integer type. The return type must match the argument type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6419
6420<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6421<p>The '<tt>llvm.cttz</tt>' intrinsic counts the trailing (least significant)
6422 zeros in a variable. If the src == 0 then the result is the size in bits of
6423 the type of src. For example, <tt>llvm.cttz(2) = 1</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6424
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6425</div>
6426
Bill Wendling3e1258b2009-02-08 04:04:40 +0000[diff] [blame]6427<!-- ======================================================================= -->
6428<div class="doc_subsection">
6429 <a name="int_overflow">Arithmetic with Overflow Intrinsics</a>
6430</div>
6431
6432<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6433
6434<p>LLVM provides intrinsics for some arithmetic with overflow operations.</p>
Bill Wendling3e1258b2009-02-08 04:04:40 +0000[diff] [blame]6435
6436</div>
6437
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6438<!-- _______________________________________________________________________ -->
6439<div class="doc_subsubsection">
Bill Wendling3e1258b2009-02-08 04:04:40 +0000[diff] [blame]6440 <a name="int_sadd_overflow">'<tt>llvm.sadd.with.overflow.*</tt>' Intrinsics</a>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6441</div>
6442
6443<div class="doc_text">
6444
6445<h5>Syntax:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6446<p>This is an overloaded intrinsic. You can use <tt>llvm.sadd.with.overflow</tt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6447 on any integer bit width.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6448
6449<pre>
6450 declare {i16, i1} @llvm.sadd.with.overflow.i16(i16 %a, i16 %b)
6451 declare {i32, i1} @llvm.sadd.with.overflow.i32(i32 %a, i32 %b)
6452 declare {i64, i1} @llvm.sadd.with.overflow.i64(i64 %a, i64 %b)
6453</pre>
6454
6455<h5>Overview:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6456<p>The '<tt>llvm.sadd.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6457 a signed addition of the two arguments, and indicate whether an overflow
6458 occurred during the signed summation.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6459
6460<h5>Arguments:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6461<p>The arguments (%a and %b) and the first element of the result structure may
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6462 be of integer types of any bit width, but they must have the same bit
6463 width. The second element of the result structure must be of
6464 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
6465 undergo signed addition.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6466
6467<h5>Semantics:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6468<p>The '<tt>llvm.sadd.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6469 a signed addition of the two variables. They return a structure &mdash; the
6470 first element of which is the signed summation, and the second element of
6471 which is a bit specifying if the signed summation resulted in an
6472 overflow.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6473
6474<h5>Examples:</h5>
6475<pre>
6476 %res = call {i32, i1} @llvm.sadd.with.overflow.i32(i32 %a, i32 %b)
6477 %sum = extractvalue {i32, i1} %res, 0
6478 %obit = extractvalue {i32, i1} %res, 1
6479 br i1 %obit, label %overflow, label %normal
6480</pre>
6481
6482</div>
6483
6484<!-- _______________________________________________________________________ -->
6485<div class="doc_subsubsection">
Bill Wendling3e1258b2009-02-08 04:04:40 +0000[diff] [blame]6486 <a name="int_uadd_overflow">'<tt>llvm.uadd.with.overflow.*</tt>' Intrinsics</a>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6487</div>
6488
6489<div class="doc_text">
6490
6491<h5>Syntax:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6492<p>This is an overloaded intrinsic. You can use <tt>llvm.uadd.with.overflow</tt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6493 on any integer bit width.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6494
6495<pre>
6496 declare {i16, i1} @llvm.uadd.with.overflow.i16(i16 %a, i16 %b)
6497 declare {i32, i1} @llvm.uadd.with.overflow.i32(i32 %a, i32 %b)
6498 declare {i64, i1} @llvm.uadd.with.overflow.i64(i64 %a, i64 %b)
6499</pre>
6500
6501<h5>Overview:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6502<p>The '<tt>llvm.uadd.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6503 an unsigned addition of the two arguments, and indicate whether a carry
6504 occurred during the unsigned summation.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6505
6506<h5>Arguments:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6507<p>The arguments (%a and %b) and the first element of the result structure may
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6508 be of integer types of any bit width, but they must have the same bit
6509 width. The second element of the result structure must be of
6510 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
6511 undergo unsigned addition.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6512
6513<h5>Semantics:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6514<p>The '<tt>llvm.uadd.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6515 an unsigned addition of the two arguments. They return a structure &mdash;
6516 the first element of which is the sum, and the second element of which is a
6517 bit specifying if the unsigned summation resulted in a carry.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6518
6519<h5>Examples:</h5>
6520<pre>
6521 %res = call {i32, i1} @llvm.uadd.with.overflow.i32(i32 %a, i32 %b)
6522 %sum = extractvalue {i32, i1} %res, 0
6523 %obit = extractvalue {i32, i1} %res, 1
6524 br i1 %obit, label %carry, label %normal
6525</pre>
6526
6527</div>
6528
6529<!-- _______________________________________________________________________ -->
6530<div class="doc_subsubsection">
Bill Wendling3e1258b2009-02-08 04:04:40 +0000[diff] [blame]6531 <a name="int_ssub_overflow">'<tt>llvm.ssub.with.overflow.*</tt>' Intrinsics</a>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6532</div>
6533
6534<div class="doc_text">
6535
6536<h5>Syntax:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6537<p>This is an overloaded intrinsic. You can use <tt>llvm.ssub.with.overflow</tt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6538 on any integer bit width.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6539
6540<pre>
6541 declare {i16, i1} @llvm.ssub.with.overflow.i16(i16 %a, i16 %b)
6542 declare {i32, i1} @llvm.ssub.with.overflow.i32(i32 %a, i32 %b)
6543 declare {i64, i1} @llvm.ssub.with.overflow.i64(i64 %a, i64 %b)
6544</pre>
6545
6546<h5>Overview:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6547<p>The '<tt>llvm.ssub.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6548 a signed subtraction of the two arguments, and indicate whether an overflow
6549 occurred during the signed subtraction.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6550
6551<h5>Arguments:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6552<p>The arguments (%a and %b) and the first element of the result structure may
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6553 be of integer types of any bit width, but they must have the same bit
6554 width. The second element of the result structure must be of
6555 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
6556 undergo signed subtraction.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6557
6558<h5>Semantics:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6559<p>The '<tt>llvm.ssub.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6560 a signed subtraction of the two arguments. They return a structure &mdash;
6561 the first element of which is the subtraction, and the second element of
6562 which is a bit specifying if the signed subtraction resulted in an
6563 overflow.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6564
6565<h5>Examples:</h5>
6566<pre>
6567 %res = call {i32, i1} @llvm.ssub.with.overflow.i32(i32 %a, i32 %b)
6568 %sum = extractvalue {i32, i1} %res, 0
6569 %obit = extractvalue {i32, i1} %res, 1
6570 br i1 %obit, label %overflow, label %normal
6571</pre>
6572
6573</div>
6574
6575<!-- _______________________________________________________________________ -->
6576<div class="doc_subsubsection">
Bill Wendling3e1258b2009-02-08 04:04:40 +0000[diff] [blame]6577 <a name="int_usub_overflow">'<tt>llvm.usub.with.overflow.*</tt>' Intrinsics</a>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6578</div>
6579
6580<div class="doc_text">
6581
6582<h5>Syntax:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6583<p>This is an overloaded intrinsic. You can use <tt>llvm.usub.with.overflow</tt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6584 on any integer bit width.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6585
6586<pre>
6587 declare {i16, i1} @llvm.usub.with.overflow.i16(i16 %a, i16 %b)
6588 declare {i32, i1} @llvm.usub.with.overflow.i32(i32 %a, i32 %b)
6589 declare {i64, i1} @llvm.usub.with.overflow.i64(i64 %a, i64 %b)
6590</pre>
6591
6592<h5>Overview:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6593<p>The '<tt>llvm.usub.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6594 an unsigned subtraction of the two arguments, and indicate whether an
6595 overflow occurred during the unsigned subtraction.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6596
6597<h5>Arguments:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6598<p>The arguments (%a and %b) and the first element of the result structure may
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6599 be of integer types of any bit width, but they must have the same bit
6600 width. The second element of the result structure must be of
6601 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
6602 undergo unsigned subtraction.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6603
6604<h5>Semantics:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6605<p>The '<tt>llvm.usub.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6606 an unsigned subtraction of the two arguments. They return a structure &mdash;
6607 the first element of which is the subtraction, and the second element of
6608 which is a bit specifying if the unsigned subtraction resulted in an
6609 overflow.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6610
6611<h5>Examples:</h5>
6612<pre>
6613 %res = call {i32, i1} @llvm.usub.with.overflow.i32(i32 %a, i32 %b)
6614 %sum = extractvalue {i32, i1} %res, 0
6615 %obit = extractvalue {i32, i1} %res, 1
6616 br i1 %obit, label %overflow, label %normal
6617</pre>
6618
6619</div>
6620
6621<!-- _______________________________________________________________________ -->
6622<div class="doc_subsubsection">
Bill Wendling3e1258b2009-02-08 04:04:40 +0000[diff] [blame]6623 <a name="int_smul_overflow">'<tt>llvm.smul.with.overflow.*</tt>' Intrinsics</a>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6624</div>
6625
6626<div class="doc_text">
6627
6628<h5>Syntax:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6629<p>This is an overloaded intrinsic. You can use <tt>llvm.smul.with.overflow</tt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6630 on any integer bit width.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6631
6632<pre>
6633 declare {i16, i1} @llvm.smul.with.overflow.i16(i16 %a, i16 %b)
6634 declare {i32, i1} @llvm.smul.with.overflow.i32(i32 %a, i32 %b)
6635 declare {i64, i1} @llvm.smul.with.overflow.i64(i64 %a, i64 %b)
6636</pre>
6637
6638<h5>Overview:</h5>
6639
6640<p>The '<tt>llvm.smul.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6641 a signed multiplication of the two arguments, and indicate whether an
6642 overflow occurred during the signed multiplication.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6643
6644<h5>Arguments:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6645<p>The arguments (%a and %b) and the first element of the result structure may
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6646 be of integer types of any bit width, but they must have the same bit
6647 width. The second element of the result structure must be of
6648 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
6649 undergo signed multiplication.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6650
6651<h5>Semantics:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6652<p>The '<tt>llvm.smul.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6653 a signed multiplication of the two arguments. They return a structure &mdash;
6654 the first element of which is the multiplication, and the second element of
6655 which is a bit specifying if the signed multiplication resulted in an
6656 overflow.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6657
6658<h5>Examples:</h5>
6659<pre>
6660 %res = call {i32, i1} @llvm.smul.with.overflow.i32(i32 %a, i32 %b)
6661 %sum = extractvalue {i32, i1} %res, 0
6662 %obit = extractvalue {i32, i1} %res, 1
6663 br i1 %obit, label %overflow, label %normal
6664</pre>
6665
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6666</div>
6667
Bill Wendlingbda98b62009-02-08 23:00:09 +0000[diff] [blame]6668<!-- _______________________________________________________________________ -->
6669<div class="doc_subsubsection">
6670 <a name="int_umul_overflow">'<tt>llvm.umul.with.overflow.*</tt>' Intrinsics</a>
6671</div>
6672
6673<div class="doc_text">
6674
6675<h5>Syntax:</h5>
Bill Wendlingbda98b62009-02-08 23:00:09 +0000[diff] [blame]6676<p>This is an overloaded intrinsic. You can use <tt>llvm.umul.with.overflow</tt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6677 on any integer bit width.</p>
Bill Wendlingbda98b62009-02-08 23:00:09 +0000[diff] [blame]6678
6679<pre>
6680 declare {i16, i1} @llvm.umul.with.overflow.i16(i16 %a, i16 %b)
6681 declare {i32, i1} @llvm.umul.with.overflow.i32(i32 %a, i32 %b)
6682 declare {i64, i1} @llvm.umul.with.overflow.i64(i64 %a, i64 %b)
6683</pre>
6684
6685<h5>Overview:</h5>
Bill Wendlingbda98b62009-02-08 23:00:09 +0000[diff] [blame]6686<p>The '<tt>llvm.umul.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6687 a unsigned multiplication of the two arguments, and indicate whether an
6688 overflow occurred during the unsigned multiplication.</p>
Bill Wendlingbda98b62009-02-08 23:00:09 +0000[diff] [blame]6689
6690<h5>Arguments:</h5>
Bill Wendlingbda98b62009-02-08 23:00:09 +0000[diff] [blame]6691<p>The arguments (%a and %b) and the first element of the result structure may
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6692 be of integer types of any bit width, but they must have the same bit
6693 width. The second element of the result structure must be of
6694 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
6695 undergo unsigned multiplication.</p>
Bill Wendlingbda98b62009-02-08 23:00:09 +0000[diff] [blame]6696
6697<h5>Semantics:</h5>
Bill Wendlingbda98b62009-02-08 23:00:09 +0000[diff] [blame]6698<p>The '<tt>llvm.umul.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6699 an unsigned multiplication of the two arguments. They return a structure
6700 &mdash; the first element of which is the multiplication, and the second
6701 element of which is a bit specifying if the unsigned multiplication resulted
6702 in an overflow.</p>
Bill Wendlingbda98b62009-02-08 23:00:09 +0000[diff] [blame]6703
6704<h5>Examples:</h5>
6705<pre>
6706 %res = call {i32, i1} @llvm.umul.with.overflow.i32(i32 %a, i32 %b)
6707 %sum = extractvalue {i32, i1} %res, 0
6708 %obit = extractvalue {i32, i1} %res, 1
6709 br i1 %obit, label %overflow, label %normal
6710</pre>
6711
6712</div>
6713
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6714<!-- ======================================================================= -->
6715<div class="doc_subsection">
Anton Korobeynikov8c98a892010-03-14 18:42:47 +0000[diff] [blame]6716 <a name="int_fp16">Half Precision Floating Point Intrinsics</a>
6717</div>
6718
6719<div class="doc_text">
6720
Chris Lattnere5969c62010-03-15 04:12:21 +0000[diff] [blame]6721<p>Half precision floating point is a storage-only format. This means that it is
6722 a dense encoding (in memory) but does not support computation in the
6723 format.</p>
Chris Lattnerebc48e52010-03-14 23:03:31 +0000[diff] [blame]6724
Chris Lattnere5969c62010-03-15 04:12:21 +0000[diff] [blame]6725<p>This means that code must first load the half-precision floating point
Chris Lattnerebc48e52010-03-14 23:03:31 +0000[diff] [blame]6726 value as an i16, then convert it to float with <a
6727 href="#int_convert_from_fp16"><tt>llvm.convert.from.fp16</tt></a>.
6728 Computation can then be performed on the float value (including extending to
Chris Lattnere5969c62010-03-15 04:12:21 +0000[diff] [blame]6729 double etc). To store the value back to memory, it is first converted to
6730 float if needed, then converted to i16 with
Chris Lattnerebc48e52010-03-14 23:03:31 +0000[diff] [blame]6731 <a href="#int_convert_to_fp16"><tt>llvm.convert.to.fp16</tt></a>, then
6732 storing as an i16 value.</p>
Anton Korobeynikov8c98a892010-03-14 18:42:47 +0000[diff] [blame]6733</div>
6734
6735<!-- _______________________________________________________________________ -->
6736<div class="doc_subsubsection">
Chris Lattnerebc48e52010-03-14 23:03:31 +0000[diff] [blame]6737 <a name="int_convert_to_fp16">'<tt>llvm.convert.to.fp16</tt>' Intrinsic</a>
Anton Korobeynikov8c98a892010-03-14 18:42:47 +0000[diff] [blame]6738</div>
6739
6740<div class="doc_text">
6741
6742<h5>Syntax:</h5>
6743<pre>
6744 declare i16 @llvm.convert.to.fp16(f32 %a)
6745</pre>
6746
6747<h5>Overview:</h5>
6748<p>The '<tt>llvm.convert.to.fp16</tt>' intrinsic function performs
6749 a conversion from single precision floating point format to half precision
6750 floating point format.</p>
6751
6752<h5>Arguments:</h5>
6753<p>The intrinsic function contains single argument - the value to be
6754 converted.</p>
6755
6756<h5>Semantics:</h5>
6757<p>The '<tt>llvm.convert.to.fp16</tt>' intrinsic function performs
6758 a conversion from single precision floating point format to half precision
Chris Lattnere5969c62010-03-15 04:12:21 +0000[diff] [blame]6759 floating point format. The return value is an <tt>i16</tt> which
Chris Lattnerebc48e52010-03-14 23:03:31 +0000[diff] [blame]6760 contains the converted number.</p>
Anton Korobeynikov8c98a892010-03-14 18:42:47 +0000[diff] [blame]6761
6762<h5>Examples:</h5>
6763<pre>
6764 %res = call i16 @llvm.convert.to.fp16(f32 %a)
6765 store i16 %res, i16* @x, align 2
6766</pre>
6767
6768</div>
6769
6770<!-- _______________________________________________________________________ -->
6771<div class="doc_subsubsection">
Chris Lattnerebc48e52010-03-14 23:03:31 +0000[diff] [blame]6772 <a name="int_convert_from_fp16">'<tt>llvm.convert.from.fp16</tt>' Intrinsic</a>
Anton Korobeynikov8c98a892010-03-14 18:42:47 +0000[diff] [blame]6773</div>
6774
6775<div class="doc_text">
6776
6777<h5>Syntax:</h5>
6778<pre>
6779 declare f32 @llvm.convert.from.fp16(i16 %a)
6780</pre>
6781
6782<h5>Overview:</h5>
6783<p>The '<tt>llvm.convert.from.fp16</tt>' intrinsic function performs
6784 a conversion from half precision floating point format to single precision
6785 floating point format.</p>
6786
6787<h5>Arguments:</h5>
6788<p>The intrinsic function contains single argument - the value to be
6789 converted.</p>
6790
6791<h5>Semantics:</h5>
6792<p>The '<tt>llvm.convert.from.fp16</tt>' intrinsic function performs a
Chris Lattnere5969c62010-03-15 04:12:21 +0000[diff] [blame]6793 conversion from half single precision floating point format to single
Chris Lattnerebc48e52010-03-14 23:03:31 +0000[diff] [blame]6794 precision floating point format. The input half-float value is represented by
6795 an <tt>i16</tt> value.</p>
Anton Korobeynikov8c98a892010-03-14 18:42:47 +0000[diff] [blame]6796
6797<h5>Examples:</h5>
6798<pre>
6799 %a = load i16* @x, align 2
6800 %res = call f32 @llvm.convert.from.fp16(i16 %a)
6801</pre>
6802
6803</div>
6804
6805<!-- ======================================================================= -->
6806<div class="doc_subsection">
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6807 <a name="int_debugger">Debugger Intrinsics</a>
6808</div>
6809
6810<div class="doc_text">
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6811
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6812<p>The LLVM debugger intrinsics (which all start with <tt>llvm.dbg.</tt>
6813 prefix), are described in
6814 the <a href="SourceLevelDebugging.html#format_common_intrinsics">LLVM Source
6815 Level Debugging</a> document.</p>
6816
6817</div>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6818
6819<!-- ======================================================================= -->
6820<div class="doc_subsection">
6821 <a name="int_eh">Exception Handling Intrinsics</a>
6822</div>
6823
6824<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6825
6826<p>The LLVM exception handling intrinsics (which all start with
6827 <tt>llvm.eh.</tt> prefix), are described in
6828 the <a href="ExceptionHandling.html#format_common_intrinsics">LLVM Exception
6829 Handling</a> document.</p>
6830
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6831</div>
6832
6833<!-- ======================================================================= -->
6834<div class="doc_subsection">
Duncan Sands7407a9f2007-09-11 14:10:23 +0000[diff] [blame]6835 <a name="int_trampoline">Trampoline Intrinsic</a>
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]6836</div>
6837
6838<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6839
6840<p>This intrinsic makes it possible to excise one parameter, marked with
6841 the <tt>nest</tt> attribute, from a function. The result is a callable
6842 function pointer lacking the nest parameter - the caller does not need to
6843 provide a value for it. Instead, the value to use is stored in advance in a
6844 "trampoline", a block of memory usually allocated on the stack, which also
6845 contains code to splice the nest value into the argument list. This is used
6846 to implement the GCC nested function address extension.</p>
6847
6848<p>For example, if the function is
6849 <tt>i32 f(i8* nest %c, i32 %x, i32 %y)</tt> then the resulting function
6850 pointer has signature <tt>i32 (i32, i32)*</tt>. It can be created as
6851 follows:</p>
6852
6853<div class="doc_code">
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]6854<pre>
Duncan Sands7407a9f2007-09-11 14:10:23 +0000[diff] [blame]6855 %tramp = alloca [10 x i8], align 4 ; size and alignment only correct for X86
6856 %tramp1 = getelementptr [10 x i8]* %tramp, i32 0, i32 0
6857 %p = call i8* @llvm.init.trampoline( i8* %tramp1, i8* bitcast (i32 (i8* nest , i32, i32)* @f to i8*), i8* %nval )
6858 %fp = bitcast i8* %p to i32 (i32, i32)*
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]6859</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6860</div>
6861
6862<p>The call <tt>%val = call i32 %fp( i32 %x, i32 %y )</tt> is then equivalent
6863 to <tt>%val = call i32 %f( i8* %nval, i32 %x, i32 %y )</tt>.</p>
6864
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]6865</div>
6866
6867<!-- _______________________________________________________________________ -->
6868<div class="doc_subsubsection">
6869 <a name="int_it">'<tt>llvm.init.trampoline</tt>' Intrinsic</a>
6870</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6871
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]6872<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6873
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]6874<h5>Syntax:</h5>
6875<pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6876 declare i8* @llvm.init.trampoline(i8* &lt;tramp&gt;, i8* &lt;func&gt;, i8* &lt;nval&gt;)
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]6877</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6878
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]6879<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6880<p>This fills the memory pointed to by <tt>tramp</tt> with code and returns a
6881 function pointer suitable for executing it.</p>
6882
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]6883<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6884<p>The <tt>llvm.init.trampoline</tt> intrinsic takes three arguments, all
6885 pointers. The <tt>tramp</tt> argument must point to a sufficiently large and
6886 sufficiently aligned block of memory; this memory is written to by the
6887 intrinsic. Note that the size and the alignment are target-specific - LLVM
6888 currently provides no portable way of determining them, so a front-end that
6889 generates this intrinsic needs to have some target-specific knowledge.
6890 The <tt>func</tt> argument must hold a function bitcast to
6891 an <tt>i8*</tt>.</p>
6892
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]6893<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6894<p>The block of memory pointed to by <tt>tramp</tt> is filled with target
6895 dependent code, turning it into a function. A pointer to this function is
6896 returned, but needs to be bitcast to an <a href="#int_trampoline">appropriate
6897 function pointer type</a> before being called. The new function's signature
6898 is the same as that of <tt>func</tt> with any arguments marked with
6899 the <tt>nest</tt> attribute removed. At most one such <tt>nest</tt> argument
6900 is allowed, and it must be of pointer type. Calling the new function is
6901 equivalent to calling <tt>func</tt> with the same argument list, but
6902 with <tt>nval</tt> used for the missing <tt>nest</tt> argument. If, after
6903 calling <tt>llvm.init.trampoline</tt>, the memory pointed to
6904 by <tt>tramp</tt> is modified, then the effect of any later call to the
6905 returned function pointer is undefined.</p>
6906
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]6907</div>
6908
6909<!-- ======================================================================= -->
6910<div class="doc_subsection">
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]6911 <a name="int_atomics">Atomic Operations and Synchronization Intrinsics</a>
6912</div>
6913
6914<div class="doc_text">
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]6915
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6916<p>These intrinsic functions expand the "universal IR" of LLVM to represent
6917 hardware constructs for atomic operations and memory synchronization. This
6918 provides an interface to the hardware, not an interface to the programmer. It
6919 is aimed at a low enough level to allow any programming models or APIs
6920 (Application Programming Interfaces) which need atomic behaviors to map
6921 cleanly onto it. It is also modeled primarily on hardware behavior. Just as
6922 hardware provides a "universal IR" for source languages, it also provides a
6923 starting point for developing a "universal" atomic operation and
6924 synchronization IR.</p>
6925
6926<p>These do <em>not</em> form an API such as high-level threading libraries,
6927 software transaction memory systems, atomic primitives, and intrinsic
6928 functions as found in BSD, GNU libc, atomic_ops, APR, and other system and
6929 application libraries. The hardware interface provided by LLVM should allow
6930 a clean implementation of all of these APIs and parallel programming models.
6931 No one model or paradigm should be selected above others unless the hardware
6932 itself ubiquitously does so.</p>
6933
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]6934</div>
6935
6936<!-- _______________________________________________________________________ -->
6937<div class="doc_subsubsection">
6938 <a name="int_memory_barrier">'<tt>llvm.memory.barrier</tt>' Intrinsic</a>
6939</div>
6940<div class="doc_text">
6941<h5>Syntax:</h5>
6942<pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6943 declare void @llvm.memory.barrier( i1 &lt;ll&gt;, i1 &lt;ls&gt;, i1 &lt;sl&gt;, i1 &lt;ss&gt;, i1 &lt;device&gt; )
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]6944</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6945
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]6946<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6947<p>The <tt>llvm.memory.barrier</tt> intrinsic guarantees ordering between
6948 specific pairs of memory access types.</p>
6949
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]6950<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6951<p>The <tt>llvm.memory.barrier</tt> intrinsic requires five boolean arguments.
6952 The first four arguments enables a specific barrier as listed below. The
Dan Gohman22dc6682010-03-01 17:41:39 +0000[diff] [blame]6953 fifth argument specifies that the barrier applies to io or device or uncached
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6954 memory.</p>
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]6955
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6956<ul>
6957 <li><tt>ll</tt>: load-load barrier</li>
6958 <li><tt>ls</tt>: load-store barrier</li>
6959 <li><tt>sl</tt>: store-load barrier</li>
6960 <li><tt>ss</tt>: store-store barrier</li>
6961 <li><tt>device</tt>: barrier applies to device and uncached memory also.</li>
6962</ul>
6963
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]6964<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6965<p>This intrinsic causes the system to enforce some ordering constraints upon
6966 the loads and stores of the program. This barrier does not
6967 indicate <em>when</em> any events will occur, it only enforces
6968 an <em>order</em> in which they occur. For any of the specified pairs of load
6969 and store operations (f.ex. load-load, or store-load), all of the first
6970 operations preceding the barrier will complete before any of the second
6971 operations succeeding the barrier begin. Specifically the semantics for each
6972 pairing is as follows:</p>
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]6973
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6974<ul>
6975 <li><tt>ll</tt>: All loads before the barrier must complete before any load
6976 after the barrier begins.</li>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]6977 <li><tt>ls</tt>: All loads before the barrier must complete before any
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6978 store after the barrier begins.</li>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]6979 <li><tt>ss</tt>: All stores before the barrier must complete before any
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6980 store after the barrier begins.</li>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]6981 <li><tt>sl</tt>: All stores before the barrier must complete before any
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6982 load after the barrier begins.</li>
6983</ul>
6984
6985<p>These semantics are applied with a logical "and" behavior when more than one
6986 is enabled in a single memory barrier intrinsic.</p>
6987
6988<p>Backends may implement stronger barriers than those requested when they do
6989 not support as fine grained a barrier as requested. Some architectures do
6990 not need all types of barriers and on such architectures, these become
6991 noops.</p>
6992
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]6993<h5>Example:</h5>
6994<pre>
Victor Hernandez6b054092009-10-26 23:44:29 +0000[diff] [blame]6995%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
6996%ptr = bitcast i8* %mallocP to i32*
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]6997 store i32 4, %ptr
6998
6999%result1 = load i32* %ptr <i>; yields {i32}:result1 = 4</i>
7000 call void @llvm.memory.barrier( i1 false, i1 true, i1 false, i1 false )
7001 <i>; guarantee the above finishes</i>
7002 store i32 8, %ptr <i>; before this begins</i>
7003</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7004
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]7005</div>
7006
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7007<!-- _______________________________________________________________________ -->
7008<div class="doc_subsubsection">
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7009 <a name="int_atomic_cmp_swap">'<tt>llvm.atomic.cmp.swap.*</tt>' Intrinsic</a>
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7010</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7011
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7012<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7013
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7014<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7015<p>This is an overloaded intrinsic. You can use <tt>llvm.atomic.cmp.swap</tt> on
7016 any integer bit width and for different address spaces. Not all targets
7017 support all bit widths however.</p>
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7018
7019<pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7020 declare i8 @llvm.atomic.cmp.swap.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;cmp&gt;, i8 &lt;val&gt; )
7021 declare i16 @llvm.atomic.cmp.swap.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;cmp&gt;, i16 &lt;val&gt; )
7022 declare i32 @llvm.atomic.cmp.swap.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;cmp&gt;, i32 &lt;val&gt; )
7023 declare i64 @llvm.atomic.cmp.swap.i64.p0i64( i64* &lt;ptr&gt;, i64 &lt;cmp&gt;, i64 &lt;val&gt; )
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7024</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7025
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7026<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7027<p>This loads a value in memory and compares it to a given value. If they are
7028 equal, it stores a new value into the memory.</p>
7029
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7030<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7031<p>The <tt>llvm.atomic.cmp.swap</tt> intrinsic takes three arguments. The result
7032 as well as both <tt>cmp</tt> and <tt>val</tt> must be integer values with the
7033 same bit width. The <tt>ptr</tt> argument must be a pointer to a value of
7034 this integer type. While any bit width integer may be used, targets may only
7035 lower representations they support in hardware.</p>
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7036
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7037<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7038<p>This entire intrinsic must be executed atomically. It first loads the value
7039 in memory pointed to by <tt>ptr</tt> and compares it with the
7040 value <tt>cmp</tt>. If they are equal, <tt>val</tt> is stored into the
7041 memory. The loaded value is yielded in all cases. This provides the
7042 equivalent of an atomic compare-and-swap operation within the SSA
7043 framework.</p>
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7044
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7045<h5>Examples:</h5>
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7046<pre>
Victor Hernandez6b054092009-10-26 23:44:29 +0000[diff] [blame]7047%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
7048%ptr = bitcast i8* %mallocP to i32*
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7049 store i32 4, %ptr
7050
7051%val1 = add i32 4, 4
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]7052%result1 = call i32 @llvm.atomic.cmp.swap.i32.p0i32( i32* %ptr, i32 4, %val1 )
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7053 <i>; yields {i32}:result1 = 4</i>
7054%stored1 = icmp eq i32 %result1, 4 <i>; yields {i1}:stored1 = true</i>
7055%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 8</i>
7056
7057%val2 = add i32 1, 1
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]7058%result2 = call i32 @llvm.atomic.cmp.swap.i32.p0i32( i32* %ptr, i32 5, %val2 )
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7059 <i>; yields {i32}:result2 = 8</i>
7060%stored2 = icmp eq i32 %result2, 5 <i>; yields {i1}:stored2 = false</i>
7061
7062%memval2 = load i32* %ptr <i>; yields {i32}:memval2 = 8</i>
7063</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7064
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7065</div>
7066
7067<!-- _______________________________________________________________________ -->
7068<div class="doc_subsubsection">
7069 <a name="int_atomic_swap">'<tt>llvm.atomic.swap.*</tt>' Intrinsic</a>
7070</div>
7071<div class="doc_text">
7072<h5>Syntax:</h5>
7073
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7074<p>This is an overloaded intrinsic. You can use <tt>llvm.atomic.swap</tt> on any
7075 integer bit width. Not all targets support all bit widths however.</p>
7076
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7077<pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7078 declare i8 @llvm.atomic.swap.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;val&gt; )
7079 declare i16 @llvm.atomic.swap.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;val&gt; )
7080 declare i32 @llvm.atomic.swap.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;val&gt; )
7081 declare i64 @llvm.atomic.swap.i64.p0i64( i64* &lt;ptr&gt;, i64 &lt;val&gt; )
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7082</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7083
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7084<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7085<p>This intrinsic loads the value stored in memory at <tt>ptr</tt> and yields
7086 the value from memory. It then stores the value in <tt>val</tt> in the memory
7087 at <tt>ptr</tt>.</p>
7088
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7089<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7090<p>The <tt>llvm.atomic.swap</tt> intrinsic takes two arguments. Both
7091 the <tt>val</tt> argument and the result must be integers of the same bit
7092 width. The first argument, <tt>ptr</tt>, must be a pointer to a value of this
7093 integer type. The targets may only lower integer representations they
7094 support.</p>
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7095
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7096<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7097<p>This intrinsic loads the value pointed to by <tt>ptr</tt>, yields it, and
7098 stores <tt>val</tt> back into <tt>ptr</tt> atomically. This provides the
7099 equivalent of an atomic swap operation within the SSA framework.</p>
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7100
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7101<h5>Examples:</h5>
7102<pre>
Victor Hernandez6b054092009-10-26 23:44:29 +0000[diff] [blame]7103%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
7104%ptr = bitcast i8* %mallocP to i32*
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7105 store i32 4, %ptr
7106
7107%val1 = add i32 4, 4
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]7108%result1 = call i32 @llvm.atomic.swap.i32.p0i32( i32* %ptr, i32 %val1 )
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7109 <i>; yields {i32}:result1 = 4</i>
7110%stored1 = icmp eq i32 %result1, 4 <i>; yields {i1}:stored1 = true</i>
7111%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 8</i>
7112
7113%val2 = add i32 1, 1
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]7114%result2 = call i32 @llvm.atomic.swap.i32.p0i32( i32* %ptr, i32 %val2 )
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7115 <i>; yields {i32}:result2 = 8</i>
7116
7117%stored2 = icmp eq i32 %result2, 8 <i>; yields {i1}:stored2 = true</i>
7118%memval2 = load i32* %ptr <i>; yields {i32}:memval2 = 2</i>
7119</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7120
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7121</div>
7122
7123<!-- _______________________________________________________________________ -->
7124<div class="doc_subsubsection">
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7125 <a name="int_atomic_load_add">'<tt>llvm.atomic.load.add.*</tt>' Intrinsic</a>
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7126
7127</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7128
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7129<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7130
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7131<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7132<p>This is an overloaded intrinsic. You can use <tt>llvm.atomic.load.add</tt> on
7133 any integer bit width. Not all targets support all bit widths however.</p>
7134
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7135<pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7136 declare i8 @llvm.atomic.load.add.i8..p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
7137 declare i16 @llvm.atomic.load.add.i16..p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
7138 declare i32 @llvm.atomic.load.add.i32..p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
7139 declare i64 @llvm.atomic.load.add.i64..p0i64( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7140</pre>
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7141
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7142<h5>Overview:</h5>
7143<p>This intrinsic adds <tt>delta</tt> to the value stored in memory
7144 at <tt>ptr</tt>. It yields the original value at <tt>ptr</tt>.</p>
7145
7146<h5>Arguments:</h5>
7147<p>The intrinsic takes two arguments, the first a pointer to an integer value
7148 and the second an integer value. The result is also an integer value. These
7149 integer types can have any bit width, but they must all have the same bit
7150 width. The targets may only lower integer representations they support.</p>
7151
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7152<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7153<p>This intrinsic does a series of operations atomically. It first loads the
7154 value stored at <tt>ptr</tt>. It then adds <tt>delta</tt>, stores the result
7155 to <tt>ptr</tt>. It yields the original value stored at <tt>ptr</tt>.</p>
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7156
7157<h5>Examples:</h5>
7158<pre>
Victor Hernandez6b054092009-10-26 23:44:29 +0000[diff] [blame]7159%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
7160%ptr = bitcast i8* %mallocP to i32*
7161 store i32 4, %ptr
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]7162%result1 = call i32 @llvm.atomic.load.add.i32.p0i32( i32* %ptr, i32 4 )
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7163 <i>; yields {i32}:result1 = 4</i>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]7164%result2 = call i32 @llvm.atomic.load.add.i32.p0i32( i32* %ptr, i32 2 )
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7165 <i>; yields {i32}:result2 = 8</i>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]7166%result3 = call i32 @llvm.atomic.load.add.i32.p0i32( i32* %ptr, i32 5 )
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7167 <i>; yields {i32}:result3 = 10</i>
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7168%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 15</i>
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7169</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7170
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]7171</div>
7172
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7173<!-- _______________________________________________________________________ -->
7174<div class="doc_subsubsection">
7175 <a name="int_atomic_load_sub">'<tt>llvm.atomic.load.sub.*</tt>' Intrinsic</a>
7176
7177</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7178
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7179<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7180
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7181<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7182<p>This is an overloaded intrinsic. You can use <tt>llvm.atomic.load.sub</tt> on
7183 any integer bit width and for different address spaces. Not all targets
7184 support all bit widths however.</p>
7185
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7186<pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7187 declare i8 @llvm.atomic.load.sub.i8.p0i32( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
7188 declare i16 @llvm.atomic.load.sub.i16.p0i32( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
7189 declare i32 @llvm.atomic.load.sub.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
7190 declare i64 @llvm.atomic.load.sub.i64.p0i32( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7191</pre>
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7192
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7193<h5>Overview:</h5>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]7194<p>This intrinsic subtracts <tt>delta</tt> to the value stored in memory at
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7195 <tt>ptr</tt>. It yields the original value at <tt>ptr</tt>.</p>
7196
7197<h5>Arguments:</h5>
7198<p>The intrinsic takes two arguments, the first a pointer to an integer value
7199 and the second an integer value. The result is also an integer value. These
7200 integer types can have any bit width, but they must all have the same bit
7201 width. The targets may only lower integer representations they support.</p>
7202
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7203<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7204<p>This intrinsic does a series of operations atomically. It first loads the
7205 value stored at <tt>ptr</tt>. It then subtracts <tt>delta</tt>, stores the
7206 result to <tt>ptr</tt>. It yields the original value stored
7207 at <tt>ptr</tt>.</p>
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7208
7209<h5>Examples:</h5>
7210<pre>
Victor Hernandez6b054092009-10-26 23:44:29 +0000[diff] [blame]7211%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
7212%ptr = bitcast i8* %mallocP to i32*
7213 store i32 8, %ptr
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]7214%result1 = call i32 @llvm.atomic.load.sub.i32.p0i32( i32* %ptr, i32 4 )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7215 <i>; yields {i32}:result1 = 8</i>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]7216%result2 = call i32 @llvm.atomic.load.sub.i32.p0i32( i32* %ptr, i32 2 )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7217 <i>; yields {i32}:result2 = 4</i>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]7218%result3 = call i32 @llvm.atomic.load.sub.i32.p0i32( i32* %ptr, i32 5 )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7219 <i>; yields {i32}:result3 = 2</i>
7220%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = -3</i>
7221</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7222
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7223</div>
7224
7225<!-- _______________________________________________________________________ -->
7226<div class="doc_subsubsection">
7227 <a name="int_atomic_load_and">'<tt>llvm.atomic.load.and.*</tt>' Intrinsic</a><br>
7228 <a name="int_atomic_load_nand">'<tt>llvm.atomic.load.nand.*</tt>' Intrinsic</a><br>
7229 <a name="int_atomic_load_or">'<tt>llvm.atomic.load.or.*</tt>' Intrinsic</a><br>
7230 <a name="int_atomic_load_xor">'<tt>llvm.atomic.load.xor.*</tt>' Intrinsic</a><br>
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7231</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7232
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7233<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7234
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7235<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7236<p>These are overloaded intrinsics. You can
7237 use <tt>llvm.atomic.load_and</tt>, <tt>llvm.atomic.load_nand</tt>,
7238 <tt>llvm.atomic.load_or</tt>, and <tt>llvm.atomic.load_xor</tt> on any integer
7239 bit width and for different address spaces. Not all targets support all bit
7240 widths however.</p>
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7241
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7242<pre>
7243 declare i8 @llvm.atomic.load.and.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
7244 declare i16 @llvm.atomic.load.and.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
7245 declare i32 @llvm.atomic.load.and.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
7246 declare i64 @llvm.atomic.load.and.i64.p0i64( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7247</pre>
7248
7249<pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7250 declare i8 @llvm.atomic.load.or.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
7251 declare i16 @llvm.atomic.load.or.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
7252 declare i32 @llvm.atomic.load.or.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
7253 declare i64 @llvm.atomic.load.or.i64.p0i64( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7254</pre>
7255
7256<pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7257 declare i8 @llvm.atomic.load.nand.i8.p0i32( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
7258 declare i16 @llvm.atomic.load.nand.i16.p0i32( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
7259 declare i32 @llvm.atomic.load.nand.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
7260 declare i64 @llvm.atomic.load.nand.i64.p0i32( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7261</pre>
7262
7263<pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7264 declare i8 @llvm.atomic.load.xor.i8.p0i32( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
7265 declare i16 @llvm.atomic.load.xor.i16.p0i32( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
7266 declare i32 @llvm.atomic.load.xor.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
7267 declare i64 @llvm.atomic.load.xor.i64.p0i32( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7268</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7269
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7270<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7271<p>These intrinsics bitwise the operation (and, nand, or, xor) <tt>delta</tt> to
7272 the value stored in memory at <tt>ptr</tt>. It yields the original value
7273 at <tt>ptr</tt>.</p>
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7274
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7275<h5>Arguments:</h5>
7276<p>These intrinsics take two arguments, the first a pointer to an integer value
7277 and the second an integer value. The result is also an integer value. These
7278 integer types can have any bit width, but they must all have the same bit
7279 width. The targets may only lower integer representations they support.</p>
7280
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7281<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7282<p>These intrinsics does a series of operations atomically. They first load the
7283 value stored at <tt>ptr</tt>. They then do the bitwise
7284 operation <tt>delta</tt>, store the result to <tt>ptr</tt>. They yield the
7285 original value stored at <tt>ptr</tt>.</p>
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7286
7287<h5>Examples:</h5>
7288<pre>
Victor Hernandez6b054092009-10-26 23:44:29 +0000[diff] [blame]7289%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
7290%ptr = bitcast i8* %mallocP to i32*
7291 store i32 0x0F0F, %ptr
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]7292%result0 = call i32 @llvm.atomic.load.nand.i32.p0i32( i32* %ptr, i32 0xFF )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7293 <i>; yields {i32}:result0 = 0x0F0F</i>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]7294%result1 = call i32 @llvm.atomic.load.and.i32.p0i32( i32* %ptr, i32 0xFF )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7295 <i>; yields {i32}:result1 = 0xFFFFFFF0</i>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]7296%result2 = call i32 @llvm.atomic.load.or.i32.p0i32( i32* %ptr, i32 0F )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7297 <i>; yields {i32}:result2 = 0xF0</i>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]7298%result3 = call i32 @llvm.atomic.load.xor.i32.p0i32( i32* %ptr, i32 0F )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7299 <i>; yields {i32}:result3 = FF</i>
7300%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = F0</i>
7301</pre>
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7302
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7303</div>
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7304
7305<!-- _______________________________________________________________________ -->
7306<div class="doc_subsubsection">
7307 <a name="int_atomic_load_max">'<tt>llvm.atomic.load.max.*</tt>' Intrinsic</a><br>
7308 <a name="int_atomic_load_min">'<tt>llvm.atomic.load.min.*</tt>' Intrinsic</a><br>
7309 <a name="int_atomic_load_umax">'<tt>llvm.atomic.load.umax.*</tt>' Intrinsic</a><br>
7310 <a name="int_atomic_load_umin">'<tt>llvm.atomic.load.umin.*</tt>' Intrinsic</a><br>
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7311</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7312
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7313<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7314
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7315<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7316<p>These are overloaded intrinsics. You can use <tt>llvm.atomic.load_max</tt>,
7317 <tt>llvm.atomic.load_min</tt>, <tt>llvm.atomic.load_umax</tt>, and
7318 <tt>llvm.atomic.load_umin</tt> on any integer bit width and for different
7319 address spaces. Not all targets support all bit widths however.</p>
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7320
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7321<pre>
7322 declare i8 @llvm.atomic.load.max.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
7323 declare i16 @llvm.atomic.load.max.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
7324 declare i32 @llvm.atomic.load.max.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
7325 declare i64 @llvm.atomic.load.max.i64.p0i64( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7326</pre>
7327
7328<pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7329 declare i8 @llvm.atomic.load.min.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
7330 declare i16 @llvm.atomic.load.min.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
7331 declare i32 @llvm.atomic.load.min.i32..p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
7332 declare i64 @llvm.atomic.load.min.i64..p0i64( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7333</pre>
7334
7335<pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7336 declare i8 @llvm.atomic.load.umax.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
7337 declare i16 @llvm.atomic.load.umax.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
7338 declare i32 @llvm.atomic.load.umax.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
7339 declare i64 @llvm.atomic.load.umax.i64.p0i64( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7340</pre>
7341
7342<pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7343 declare i8 @llvm.atomic.load.umin.i8..p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
7344 declare i16 @llvm.atomic.load.umin.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
7345 declare i32 @llvm.atomic.load.umin.i32..p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
7346 declare i64 @llvm.atomic.load.umin.i64..p0i64( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7347</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7348
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7349<h5>Overview:</h5>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]7350<p>These intrinsics takes the signed or unsigned minimum or maximum of
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7351 <tt>delta</tt> and the value stored in memory at <tt>ptr</tt>. It yields the
7352 original value at <tt>ptr</tt>.</p>
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7353
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7354<h5>Arguments:</h5>
7355<p>These intrinsics take two arguments, the first a pointer to an integer value
7356 and the second an integer value. The result is also an integer value. These
7357 integer types can have any bit width, but they must all have the same bit
7358 width. The targets may only lower integer representations they support.</p>
7359
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7360<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7361<p>These intrinsics does a series of operations atomically. They first load the
7362 value stored at <tt>ptr</tt>. They then do the signed or unsigned min or
7363 max <tt>delta</tt> and the value, store the result to <tt>ptr</tt>. They
7364 yield the original value stored at <tt>ptr</tt>.</p>
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7365
7366<h5>Examples:</h5>
7367<pre>
Victor Hernandez6b054092009-10-26 23:44:29 +0000[diff] [blame]7368%mallocP = tail call i8* @malloc(i32 ptrtoint (i32* getelementptr (i32* null, i32 1) to i32))
7369%ptr = bitcast i8* %mallocP to i32*
7370 store i32 7, %ptr
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]7371%result0 = call i32 @llvm.atomic.load.min.i32.p0i32( i32* %ptr, i32 -2 )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7372 <i>; yields {i32}:result0 = 7</i>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]7373%result1 = call i32 @llvm.atomic.load.max.i32.p0i32( i32* %ptr, i32 8 )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7374 <i>; yields {i32}:result1 = -2</i>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]7375%result2 = call i32 @llvm.atomic.load.umin.i32.p0i32( i32* %ptr, i32 10 )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7376 <i>; yields {i32}:result2 = 8</i>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]7377%result3 = call i32 @llvm.atomic.load.umax.i32.p0i32( i32* %ptr, i32 30 )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7378 <i>; yields {i32}:result3 = 8</i>
7379%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 30</i>
7380</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7381
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]7382</div>
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]7383
Nick Lewyckyc888d352009-10-13 07:03:23 +0000[diff] [blame]7384
7385<!-- ======================================================================= -->
7386<div class="doc_subsection">
7387 <a name="int_memorymarkers">Memory Use Markers</a>
7388</div>
7389
7390<div class="doc_text">
7391
7392<p>This class of intrinsics exists to information about the lifetime of memory
7393 objects and ranges where variables are immutable.</p>
7394
7395</div>
7396
7397<!-- _______________________________________________________________________ -->
7398<div class="doc_subsubsection">
7399 <a name="int_lifetime_start">'<tt>llvm.lifetime.start</tt>' Intrinsic</a>
7400</div>
7401
7402<div class="doc_text">
7403
7404<h5>Syntax:</h5>
7405<pre>
7406 declare void @llvm.lifetime.start(i64 &lt;size&gt;, i8* nocapture &lt;ptr&gt;)
7407</pre>
7408
7409<h5>Overview:</h5>
7410<p>The '<tt>llvm.lifetime.start</tt>' intrinsic specifies the start of a memory
7411 object's lifetime.</p>
7412
7413<h5>Arguments:</h5>
Nick Lewycky23f49982009-10-13 07:57:33 +0000[diff] [blame]7414<p>The first argument is a constant integer representing the size of the
7415 object, or -1 if it is variable sized. The second argument is a pointer to
7416 the object.</p>
Nick Lewyckyc888d352009-10-13 07:03:23 +0000[diff] [blame]7417
7418<h5>Semantics:</h5>
7419<p>This intrinsic indicates that before this point in the code, the value of the
7420 memory pointed to by <tt>ptr</tt> is dead. This means that it is known to
Nick Lewyckyf8a5f302009-10-27 16:56:58 +0000[diff] [blame]7421 never be used and has an undefined value. A load from the pointer that
7422 precedes this intrinsic can be replaced with
Nick Lewyckyc888d352009-10-13 07:03:23 +0000[diff] [blame]7423 <tt>'<a href="#undefvalues">undef</a>'</tt>.</p>
7424
7425</div>
7426
7427<!-- _______________________________________________________________________ -->
7428<div class="doc_subsubsection">
7429 <a name="int_lifetime_end">'<tt>llvm.lifetime.end</tt>' Intrinsic</a>
7430</div>
7431
7432<div class="doc_text">
7433
7434<h5>Syntax:</h5>
7435<pre>
7436 declare void @llvm.lifetime.end(i64 &lt;size&gt;, i8* nocapture &lt;ptr&gt;)
7437</pre>
7438
7439<h5>Overview:</h5>
7440<p>The '<tt>llvm.lifetime.end</tt>' intrinsic specifies the end of a memory
7441 object's lifetime.</p>
7442
7443<h5>Arguments:</h5>
Nick Lewycky23f49982009-10-13 07:57:33 +0000[diff] [blame]7444<p>The first argument is a constant integer representing the size of the
7445 object, or -1 if it is variable sized. The second argument is a pointer to
7446 the object.</p>
Nick Lewyckyc888d352009-10-13 07:03:23 +0000[diff] [blame]7447
7448<h5>Semantics:</h5>
7449<p>This intrinsic indicates that after this point in the code, the value of the
7450 memory pointed to by <tt>ptr</tt> is dead. This means that it is known to
7451 never be used and has an undefined value. Any stores into the memory object
7452 following this intrinsic may be removed as dead.
7453
7454</div>
7455
7456<!-- _______________________________________________________________________ -->
7457<div class="doc_subsubsection">
7458 <a name="int_invariant_start">'<tt>llvm.invariant.start</tt>' Intrinsic</a>
7459</div>
7460
7461<div class="doc_text">
7462
7463<h5>Syntax:</h5>
7464<pre>
7465 declare {}* @llvm.invariant.start(i64 &lt;size&gt;, i8* nocapture &lt;ptr&gt;) readonly
7466</pre>
7467
7468<h5>Overview:</h5>
7469<p>The '<tt>llvm.invariant.start</tt>' intrinsic specifies that the contents of
7470 a memory object will not change.</p>
7471
7472<h5>Arguments:</h5>
Nick Lewycky23f49982009-10-13 07:57:33 +0000[diff] [blame]7473<p>The first argument is a constant integer representing the size of the
7474 object, or -1 if it is variable sized. The second argument is a pointer to
7475 the object.</p>
Nick Lewyckyc888d352009-10-13 07:03:23 +0000[diff] [blame]7476
7477<h5>Semantics:</h5>
7478<p>This intrinsic indicates that until an <tt>llvm.invariant.end</tt> that uses
7479 the return value, the referenced memory location is constant and
7480 unchanging.</p>
7481
7482</div>
7483
7484<!-- _______________________________________________________________________ -->
7485<div class="doc_subsubsection">
7486 <a name="int_invariant_end">'<tt>llvm.invariant.end</tt>' Intrinsic</a>
7487</div>
7488
7489<div class="doc_text">
7490
7491<h5>Syntax:</h5>
7492<pre>
7493 declare void @llvm.invariant.end({}* &lt;start&gt;, i64 &lt;size&gt;, i8* nocapture &lt;ptr&gt;)
7494</pre>
7495
7496<h5>Overview:</h5>
7497<p>The '<tt>llvm.invariant.end</tt>' intrinsic specifies that the contents of
7498 a memory object are mutable.</p>
7499
7500<h5>Arguments:</h5>
7501<p>The first argument is the matching <tt>llvm.invariant.start</tt> intrinsic.
Nick Lewycky23f49982009-10-13 07:57:33 +0000[diff] [blame]7502 The second argument is a constant integer representing the size of the
7503 object, or -1 if it is variable sized and the third argument is a pointer
7504 to the object.</p>
Nick Lewyckyc888d352009-10-13 07:03:23 +0000[diff] [blame]7505
7506<h5>Semantics:</h5>
7507<p>This intrinsic indicates that the memory is mutable again.</p>
7508
7509</div>
7510
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]7511<!-- ======================================================================= -->
7512<div class="doc_subsection">
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]7513 <a name="int_general">General Intrinsics</a>
7514</div>
7515
7516<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7517
7518<p>This class of intrinsics is designed to be generic and has no specific
7519 purpose.</p>
7520
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]7521</div>
7522
7523<!-- _______________________________________________________________________ -->
7524<div class="doc_subsubsection">
7525 <a name="int_var_annotation">'<tt>llvm.var.annotation</tt>' Intrinsic</a>
7526</div>
7527
7528<div class="doc_text">
7529
7530<h5>Syntax:</h5>
7531<pre>
7532 declare void @llvm.var.annotation(i8* &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
7533</pre>
7534
7535<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7536<p>The '<tt>llvm.var.annotation</tt>' intrinsic.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]7537
7538<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7539<p>The first argument is a pointer to a value, the second is a pointer to a
7540 global string, the third is a pointer to a global string which is the source
7541 file name, and the last argument is the line number.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]7542
7543<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7544<p>This intrinsic allows annotation of local variables with arbitrary strings.
7545 This can be useful for special purpose optimizations that want to look for
7546 these annotations. These have no other defined use, they are ignored by code
7547 generation and optimization.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]7548
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]7549</div>
7550
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]7551<!-- _______________________________________________________________________ -->
7552<div class="doc_subsubsection">
Tanya Lattnerc9869b12007-09-21 23:57:59 +0000[diff] [blame]7553 <a name="int_annotation">'<tt>llvm.annotation.*</tt>' Intrinsic</a>
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]7554</div>
7555
7556<div class="doc_text">
7557
7558<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7559<p>This is an overloaded intrinsic. You can use '<tt>llvm.annotation</tt>' on
7560 any integer bit width.</p>
7561
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]7562<pre>
Tanya Lattner09161fe2007-09-22 00:03:01 +0000[diff] [blame]7563 declare i8 @llvm.annotation.i8(i8 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
7564 declare i16 @llvm.annotation.i16(i16 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
7565 declare i32 @llvm.annotation.i32(i32 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
7566 declare i64 @llvm.annotation.i64(i64 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
7567 declare i256 @llvm.annotation.i256(i256 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]7568</pre>
7569
7570<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7571<p>The '<tt>llvm.annotation</tt>' intrinsic.</p>
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]7572
7573<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7574<p>The first argument is an integer value (result of some expression), the
7575 second is a pointer to a global string, the third is a pointer to a global
7576 string which is the source file name, and the last argument is the line
7577 number. It returns the value of the first argument.</p>
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]7578
7579<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7580<p>This intrinsic allows annotations to be put on arbitrary expressions with
7581 arbitrary strings. This can be useful for special purpose optimizations that
7582 want to look for these annotations. These have no other defined use, they
7583 are ignored by code generation and optimization.</p>
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]7584
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]7585</div>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]7586
Anton Korobeynikove6e764f2008-01-15 22:31:34 +0000[diff] [blame]7587<!-- _______________________________________________________________________ -->
7588<div class="doc_subsubsection">
7589 <a name="int_trap">'<tt>llvm.trap</tt>' Intrinsic</a>
7590</div>
7591
7592<div class="doc_text">
7593
7594<h5>Syntax:</h5>
7595<pre>
7596 declare void @llvm.trap()
7597</pre>
7598
7599<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7600<p>The '<tt>llvm.trap</tt>' intrinsic.</p>
Anton Korobeynikove6e764f2008-01-15 22:31:34 +0000[diff] [blame]7601
7602<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7603<p>None.</p>
Anton Korobeynikove6e764f2008-01-15 22:31:34 +0000[diff] [blame]7604
7605<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7606<p>This intrinsics is lowered to the target dependent trap instruction. If the
7607 target does not have a trap instruction, this intrinsic will be lowered to
7608 the call of the <tt>abort()</tt> function.</p>
Anton Korobeynikove6e764f2008-01-15 22:31:34 +0000[diff] [blame]7609
Anton Korobeynikove6e764f2008-01-15 22:31:34 +0000[diff] [blame]7610</div>
7611
Bill Wendlinge4164592008-11-19 05:56:17 +0000[diff] [blame]7612<!-- _______________________________________________________________________ -->
7613<div class="doc_subsubsection">
Misha Brukman5dd7f4d2008-11-22 23:55:29 +0000[diff] [blame]7614 <a name="int_stackprotector">'<tt>llvm.stackprotector</tt>' Intrinsic</a>
Bill Wendlinge4164592008-11-19 05:56:17 +0000[diff] [blame]7615</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7616
Bill Wendlinge4164592008-11-19 05:56:17 +0000[diff] [blame]7617<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7618
Bill Wendlinge4164592008-11-19 05:56:17 +0000[diff] [blame]7619<h5>Syntax:</h5>
7620<pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7621 declare void @llvm.stackprotector( i8* &lt;guard&gt;, i8** &lt;slot&gt; )
Bill Wendlinge4164592008-11-19 05:56:17 +0000[diff] [blame]7622</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7623
Bill Wendlinge4164592008-11-19 05:56:17 +0000[diff] [blame]7624<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7625<p>The <tt>llvm.stackprotector</tt> intrinsic takes the <tt>guard</tt> and
7626 stores it onto the stack at <tt>slot</tt>. The stack slot is adjusted to
7627 ensure that it is placed on the stack before local variables.</p>
7628
Bill Wendlinge4164592008-11-19 05:56:17 +0000[diff] [blame]7629<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7630<p>The <tt>llvm.stackprotector</tt> intrinsic requires two pointer
7631 arguments. The first argument is the value loaded from the stack
7632 guard <tt>@__stack_chk_guard</tt>. The second variable is an <tt>alloca</tt>
7633 that has enough space to hold the value of the guard.</p>
7634
Bill Wendlinge4164592008-11-19 05:56:17 +0000[diff] [blame]7635<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]7636<p>This intrinsic causes the prologue/epilogue inserter to force the position of
7637 the <tt>AllocaInst</tt> stack slot to be before local variables on the
7638 stack. This is to ensure that if a local variable on the stack is
7639 overwritten, it will destroy the value of the guard. When the function exits,
7640 the guard on the stack is checked against the original guard. If they're
7641 different, then the program aborts by calling the <tt>__stack_chk_fail()</tt>
7642 function.</p>
7643
Bill Wendlinge4164592008-11-19 05:56:17 +0000[diff] [blame]7644</div>
7645
Eric Christopher767a3722009-11-30 08:03:53 +0000[diff] [blame]7646<!-- _______________________________________________________________________ -->
7647<div class="doc_subsubsection">
7648 <a name="int_objectsize">'<tt>llvm.objectsize</tt>' Intrinsic</a>
7649</div>
7650
7651<div class="doc_text">
7652
7653<h5>Syntax:</h5>
7654<pre>
Eric Christopher0101f9d2009-12-23 00:29:49 +0000[diff] [blame]7655 declare i32 @llvm.objectsize.i32( i8* &lt;object&gt;, i1 &lt;type&gt; )
7656 declare i64 @llvm.objectsize.i64( i8* &lt;object&gt;, i1 &lt;type&gt; )
Eric Christopher767a3722009-11-30 08:03:53 +0000[diff] [blame]7657</pre>
7658
7659<h5>Overview:</h5>
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]7660<p>The <tt>llvm.objectsize</tt> intrinsic is designed to provide information
Eric Christopher184f44d2010-01-08 21:42:39 +0000[diff] [blame]7661 to the optimizers to discover at compile time either a) when an
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]7662 operation like memcpy will either overflow a buffer that corresponds to
7663 an object, or b) to determine that a runtime check for overflow isn't
7664 necessary. An object in this context means an allocation of a
Eric Christopher0101f9d2009-12-23 00:29:49 +0000[diff] [blame]7665 specific class, structure, array, or other object.</p>
Eric Christopher767a3722009-11-30 08:03:53 +0000[diff] [blame]7666
7667<h5>Arguments:</h5>
7668<p>The <tt>llvm.objectsize</tt> intrinsic takes two arguments. The first
Eric Christopher0101f9d2009-12-23 00:29:49 +0000[diff] [blame]7669 argument is a pointer to or into the <tt>object</tt>. The second argument
7670 is a boolean 0 or 1. This argument determines whether you want the
7671 maximum (0) or minimum (1) bytes remaining. This needs to be a literal 0 or
7672 1, variables are not allowed.</p>
7673
Eric Christopher767a3722009-11-30 08:03:53 +0000[diff] [blame]7674<h5>Semantics:</h5>
7675<p>The <tt>llvm.objectsize</tt> intrinsic is lowered to either a constant
Eric Christophera1151bf2009-12-05 02:46:03 +0000[diff] [blame]7676 representing the size of the object concerned or <tt>i32/i64 -1 or 0</tt>
7677 (depending on the <tt>type</tt> argument if the size cannot be determined
7678 at compile time.</p>
Eric Christopher767a3722009-11-30 08:03:53 +0000[diff] [blame]7679
7680</div>
7681
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]7682<!-- *********************************************************************** -->
7683<hr>
7684<address>
7685 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
Misha Brukman947321d2008-12-11 17:34:48 +0000[diff] [blame]7686 src="http://jigsaw.w3.org/css-validator/images/vcss-blue" alt="Valid CSS"></a>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]7687 <a href="http://validator.w3.org/check/referer"><img
Misha Brukman947321d2008-12-11 17:34:48 +0000[diff] [blame]7688 src="http://www.w3.org/Icons/valid-html401-blue" alt="Valid HTML 4.01"></a>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]7689
7690 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
7691 <a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br>
7692 Last modified: $Date$
7693</address>
Chris Lattner08497ce2008-01-04 04:33:49 +0000[diff] [blame]7694
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]7695</body>
7696</html>