Gitiles
Code Review Sign In
gerrit-public.fairphone.software / fp2-dev / platform / external / llvm / e275324efabae805f10167188bb962b49bf19e7e / . / docs / LangRef.html
blob: 84a1012ca755ffe69f71c150e39d1d176b024c4f [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">
8 <meta name="description"
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>
34 <li><a href="#linkage_linkonce">'<tt>linkonce_odr</tt>' Linkage</a></li>
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>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]46 <li><a href="#paramattrs">Parameter Attributes</a></li>
Devang Patel008cd3e2008-09-26 23:51:19 +0000[diff] [blame]47 <li><a href="#fnattrs">Function Attributes</a></li>
Gordon Henriksen13fe5e32007-12-10 03:18:06 +0000[diff] [blame]48 <li><a href="#gc">Garbage Collector Names</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]49 <li><a href="#moduleasm">Module-Level Inline Assembly</a></li>
50 <li><a href="#datalayout">Data Layout</a></li>
51 </ol>
52 </li>
53 <li><a href="#typesystem">Type System</a>
54 <ol>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]55 <li><a href="#t_classifications">Type Classifications</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]56 <li><a href="#t_primitive">Primitive Types</a>
57 <ol>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]58 <li><a href="#t_floating">Floating Point Types</a></li>
59 <li><a href="#t_void">Void Type</a></li>
60 <li><a href="#t_label">Label Type</a></li>
Nick Lewycky29aaef82009-05-30 05:06:04 +0000[diff] [blame]61 <li><a href="#t_metadata">Metadata Type</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]62 </ol>
63 </li>
64 <li><a href="#t_derived">Derived Types</a>
65 <ol>
Chris Lattner251ab812007-12-18 06:18:21 +0000[diff] [blame]66 <li><a href="#t_integer">Integer Type</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]67 <li><a href="#t_array">Array Type</a></li>
68 <li><a href="#t_function">Function Type</a></li>
69 <li><a href="#t_pointer">Pointer Type</a></li>
70 <li><a href="#t_struct">Structure Type</a></li>
71 <li><a href="#t_pstruct">Packed Structure Type</a></li>
72 <li><a href="#t_vector">Vector Type</a></li>
73 <li><a href="#t_opaque">Opaque Type</a></li>
74 </ol>
75 </li>
Chris Lattner515195a2009-02-02 07:32:36 +0000[diff] [blame]76 <li><a href="#t_uprefs">Type Up-references</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]77 </ol>
78 </li>
79 <li><a href="#constants">Constants</a>
80 <ol>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]81 <li><a href="#simpleconstants">Simple Constants</a></li>
Chris Lattner97063852009-02-28 18:32:25 +0000[diff] [blame]82 <li><a href="#complexconstants">Complex Constants</a></li>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]83 <li><a href="#globalconstants">Global Variable and Function Addresses</a></li>
84 <li><a href="#undefvalues">Undefined Values</a></li>
85 <li><a href="#constantexprs">Constant Expressions</a></li>
Nick Lewycky4dcf8102009-04-04 07:22:01 +0000[diff] [blame]86 <li><a href="#metadata">Embedded Metadata</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]87 </ol>
88 </li>
89 <li><a href="#othervalues">Other Values</a>
90 <ol>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]91 <li><a href="#inlineasm">Inline Assembler Expressions</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]92 </ol>
93 </li>
94 <li><a href="#instref">Instruction Reference</a>
95 <ol>
96 <li><a href="#terminators">Terminator Instructions</a>
97 <ol>
98 <li><a href="#i_ret">'<tt>ret</tt>' Instruction</a></li>
99 <li><a href="#i_br">'<tt>br</tt>' Instruction</a></li>
100 <li><a href="#i_switch">'<tt>switch</tt>' Instruction</a></li>
101 <li><a href="#i_invoke">'<tt>invoke</tt>' Instruction</a></li>
102 <li><a href="#i_unwind">'<tt>unwind</tt>' Instruction</a></li>
103 <li><a href="#i_unreachable">'<tt>unreachable</tt>' Instruction</a></li>
104 </ol>
105 </li>
106 <li><a href="#binaryops">Binary Operations</a>
107 <ol>
108 <li><a href="#i_add">'<tt>add</tt>' Instruction</a></li>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]109 <li><a href="#i_fadd">'<tt>fadd</tt>' Instruction</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]110 <li><a href="#i_sub">'<tt>sub</tt>' Instruction</a></li>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]111 <li><a href="#i_fsub">'<tt>fsub</tt>' Instruction</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]112 <li><a href="#i_mul">'<tt>mul</tt>' Instruction</a></li>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]113 <li><a href="#i_fmul">'<tt>fmul</tt>' Instruction</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]114 <li><a href="#i_udiv">'<tt>udiv</tt>' Instruction</a></li>
115 <li><a href="#i_sdiv">'<tt>sdiv</tt>' Instruction</a></li>
116 <li><a href="#i_fdiv">'<tt>fdiv</tt>' Instruction</a></li>
117 <li><a href="#i_urem">'<tt>urem</tt>' Instruction</a></li>
118 <li><a href="#i_srem">'<tt>srem</tt>' Instruction</a></li>
119 <li><a href="#i_frem">'<tt>frem</tt>' Instruction</a></li>
120 </ol>
121 </li>
122 <li><a href="#bitwiseops">Bitwise Binary Operations</a>
123 <ol>
124 <li><a href="#i_shl">'<tt>shl</tt>' Instruction</a></li>
125 <li><a href="#i_lshr">'<tt>lshr</tt>' Instruction</a></li>
126 <li><a href="#i_ashr">'<tt>ashr</tt>' Instruction</a></li>
127 <li><a href="#i_and">'<tt>and</tt>' Instruction</a></li>
128 <li><a href="#i_or">'<tt>or</tt>' Instruction</a></li>
129 <li><a href="#i_xor">'<tt>xor</tt>' Instruction</a></li>
130 </ol>
131 </li>
132 <li><a href="#vectorops">Vector Operations</a>
133 <ol>
134 <li><a href="#i_extractelement">'<tt>extractelement</tt>' Instruction</a></li>
135 <li><a href="#i_insertelement">'<tt>insertelement</tt>' Instruction</a></li>
136 <li><a href="#i_shufflevector">'<tt>shufflevector</tt>' Instruction</a></li>
137 </ol>
138 </li>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]139 <li><a href="#aggregateops">Aggregate Operations</a>
140 <ol>
141 <li><a href="#i_extractvalue">'<tt>extractvalue</tt>' Instruction</a></li>
142 <li><a href="#i_insertvalue">'<tt>insertvalue</tt>' Instruction</a></li>
143 </ol>
144 </li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]145 <li><a href="#memoryops">Memory Access and Addressing Operations</a>
146 <ol>
147 <li><a href="#i_malloc">'<tt>malloc</tt>' Instruction</a></li>
148 <li><a href="#i_free">'<tt>free</tt>' Instruction</a></li>
149 <li><a href="#i_alloca">'<tt>alloca</tt>' Instruction</a></li>
150 <li><a href="#i_load">'<tt>load</tt>' Instruction</a></li>
151 <li><a href="#i_store">'<tt>store</tt>' Instruction</a></li>
152 <li><a href="#i_getelementptr">'<tt>getelementptr</tt>' Instruction</a></li>
153 </ol>
154 </li>
155 <li><a href="#convertops">Conversion Operations</a>
156 <ol>
157 <li><a href="#i_trunc">'<tt>trunc .. to</tt>' Instruction</a></li>
158 <li><a href="#i_zext">'<tt>zext .. to</tt>' Instruction</a></li>
159 <li><a href="#i_sext">'<tt>sext .. to</tt>' Instruction</a></li>
160 <li><a href="#i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a></li>
161 <li><a href="#i_fpext">'<tt>fpext .. to</tt>' Instruction</a></li>
162 <li><a href="#i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a></li>
163 <li><a href="#i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a></li>
164 <li><a href="#i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a></li>
165 <li><a href="#i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a></li>
166 <li><a href="#i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a></li>
167 <li><a href="#i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a></li>
168 <li><a href="#i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a></li>
169 </ol>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]170 </li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]171 <li><a href="#otherops">Other Operations</a>
172 <ol>
173 <li><a href="#i_icmp">'<tt>icmp</tt>' Instruction</a></li>
174 <li><a href="#i_fcmp">'<tt>fcmp</tt>' Instruction</a></li>
175 <li><a href="#i_phi">'<tt>phi</tt>' Instruction</a></li>
176 <li><a href="#i_select">'<tt>select</tt>' Instruction</a></li>
177 <li><a href="#i_call">'<tt>call</tt>' Instruction</a></li>
178 <li><a href="#i_va_arg">'<tt>va_arg</tt>' Instruction</a></li>
179 </ol>
180 </li>
181 </ol>
182 </li>
183 <li><a href="#intrinsics">Intrinsic Functions</a>
184 <ol>
185 <li><a href="#int_varargs">Variable Argument Handling Intrinsics</a>
186 <ol>
187 <li><a href="#int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a></li>
188 <li><a href="#int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a></li>
189 <li><a href="#int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a></li>
190 </ol>
191 </li>
192 <li><a href="#int_gc">Accurate Garbage Collection Intrinsics</a>
193 <ol>
194 <li><a href="#int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a></li>
195 <li><a href="#int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a></li>
196 <li><a href="#int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a></li>
197 </ol>
198 </li>
199 <li><a href="#int_codegen">Code Generator Intrinsics</a>
200 <ol>
201 <li><a href="#int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a></li>
202 <li><a href="#int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a></li>
203 <li><a href="#int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a></li>
204 <li><a href="#int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a></li>
205 <li><a href="#int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a></li>
206 <li><a href="#int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a></li>
207 <li><a href="#int_readcyclecounter"><tt>llvm.readcyclecounter</tt>' Intrinsic</a></li>
208 </ol>
209 </li>
210 <li><a href="#int_libc">Standard C Library Intrinsics</a>
211 <ol>
212 <li><a href="#int_memcpy">'<tt>llvm.memcpy.*</tt>' Intrinsic</a></li>
213 <li><a href="#int_memmove">'<tt>llvm.memmove.*</tt>' Intrinsic</a></li>
214 <li><a href="#int_memset">'<tt>llvm.memset.*</tt>' Intrinsic</a></li>
215 <li><a href="#int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a></li>
216 <li><a href="#int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a></li>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]217 <li><a href="#int_sin">'<tt>llvm.sin.*</tt>' Intrinsic</a></li>
218 <li><a href="#int_cos">'<tt>llvm.cos.*</tt>' Intrinsic</a></li>
219 <li><a href="#int_pow">'<tt>llvm.pow.*</tt>' Intrinsic</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]220 </ol>
221 </li>
222 <li><a href="#int_manip">Bit Manipulation Intrinsics</a>
223 <ol>
224 <li><a href="#int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a></li>
225 <li><a href="#int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic </a></li>
226 <li><a href="#int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic </a></li>
227 <li><a href="#int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic </a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]228 </ol>
229 </li>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]230 <li><a href="#int_overflow">Arithmetic with Overflow Intrinsics</a>
231 <ol>
Bill Wendling3e1258b2009-02-08 04:04:40 +0000[diff] [blame]232 <li><a href="#int_sadd_overflow">'<tt>llvm.sadd.with.overflow.*</tt> Intrinsics</a></li>
233 <li><a href="#int_uadd_overflow">'<tt>llvm.uadd.with.overflow.*</tt> Intrinsics</a></li>
234 <li><a href="#int_ssub_overflow">'<tt>llvm.ssub.with.overflow.*</tt> Intrinsics</a></li>
235 <li><a href="#int_usub_overflow">'<tt>llvm.usub.with.overflow.*</tt> Intrinsics</a></li>
236 <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]237 <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]238 </ol>
239 </li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]240 <li><a href="#int_debugger">Debugger intrinsics</a></li>
241 <li><a href="#int_eh">Exception Handling intrinsics</a></li>
Duncan Sands7407a9f2007-09-11 14:10:23 +0000[diff] [blame]242 <li><a href="#int_trampoline">Trampoline Intrinsic</a>
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]243 <ol>
244 <li><a href="#int_it">'<tt>llvm.init.trampoline</tt>' Intrinsic</a></li>
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]245 </ol>
246 </li>
Bill Wendling9127adb2008-11-18 22:10:53 +0000[diff] [blame]247 <li><a href="#int_atomics">Atomic intrinsics</a>
248 <ol>
249 <li><a href="#int_memory_barrier"><tt>llvm.memory_barrier</tt></a></li>
250 <li><a href="#int_atomic_cmp_swap"><tt>llvm.atomic.cmp.swap</tt></a></li>
251 <li><a href="#int_atomic_swap"><tt>llvm.atomic.swap</tt></a></li>
252 <li><a href="#int_atomic_load_add"><tt>llvm.atomic.load.add</tt></a></li>
253 <li><a href="#int_atomic_load_sub"><tt>llvm.atomic.load.sub</tt></a></li>
254 <li><a href="#int_atomic_load_and"><tt>llvm.atomic.load.and</tt></a></li>
255 <li><a href="#int_atomic_load_nand"><tt>llvm.atomic.load.nand</tt></a></li>
256 <li><a href="#int_atomic_load_or"><tt>llvm.atomic.load.or</tt></a></li>
257 <li><a href="#int_atomic_load_xor"><tt>llvm.atomic.load.xor</tt></a></li>
258 <li><a href="#int_atomic_load_max"><tt>llvm.atomic.load.max</tt></a></li>
259 <li><a href="#int_atomic_load_min"><tt>llvm.atomic.load.min</tt></a></li>
260 <li><a href="#int_atomic_load_umax"><tt>llvm.atomic.load.umax</tt></a></li>
261 <li><a href="#int_atomic_load_umin"><tt>llvm.atomic.load.umin</tt></a></li>
262 </ol>
263 </li>
Reid Spencerb043f672007-07-20 19:59:11 +0000[diff] [blame]264 <li><a href="#int_general">General intrinsics</a>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]265 <ol>
Reid Spencerb043f672007-07-20 19:59:11 +0000[diff] [blame]266 <li><a href="#int_var_annotation">
Bill Wendlinge4164592008-11-19 05:56:17 +0000[diff] [blame]267 '<tt>llvm.var.annotation</tt>' Intrinsic</a></li>
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]268 <li><a href="#int_annotation">
Bill Wendlinge4164592008-11-19 05:56:17 +0000[diff] [blame]269 '<tt>llvm.annotation.*</tt>' Intrinsic</a></li>
Anton Korobeynikove6e764f2008-01-15 22:31:34 +0000[diff] [blame]270 <li><a href="#int_trap">
Bill Wendlinge4164592008-11-19 05:56:17 +0000[diff] [blame]271 '<tt>llvm.trap</tt>' Intrinsic</a></li>
272 <li><a href="#int_stackprotector">
273 '<tt>llvm.stackprotector</tt>' Intrinsic</a></li>
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]274 </ol>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]275 </li>
276 </ol>
277 </li>
278</ol>
279
280<div class="doc_author">
281 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>
282 and <a href="mailto:vadve@cs.uiuc.edu">Vikram Adve</a></p>
283</div>
284
285<!-- *********************************************************************** -->
286<div class="doc_section"> <a name="abstract">Abstract </a></div>
287<!-- *********************************************************************** -->
288
289<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]290
291<p>This document is a reference manual for the LLVM assembly language. LLVM is
292 a Static Single Assignment (SSA) based representation that provides type
293 safety, low-level operations, flexibility, and the capability of representing
294 'all' high-level languages cleanly. It is the common code representation
295 used throughout all phases of the LLVM compilation strategy.</p>
296
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]297</div>
298
299<!-- *********************************************************************** -->
300<div class="doc_section"> <a name="introduction">Introduction</a> </div>
301<!-- *********************************************************************** -->
302
303<div class="doc_text">
304
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]305<p>The LLVM code representation is designed to be used in three different forms:
306 as an in-memory compiler IR, as an on-disk bitcode representation (suitable
307 for fast loading by a Just-In-Time compiler), and as a human readable
308 assembly language representation. This allows LLVM to provide a powerful
309 intermediate representation for efficient compiler transformations and
310 analysis, while providing a natural means to debug and visualize the
311 transformations. The three different forms of LLVM are all equivalent. This
312 document describes the human readable representation and notation.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]313
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]314<p>The LLVM representation aims to be light-weight and low-level while being
315 expressive, typed, and extensible at the same time. It aims to be a
316 "universal IR" of sorts, by being at a low enough level that high-level ideas
317 may be cleanly mapped to it (similar to how microprocessors are "universal
318 IR's", allowing many source languages to be mapped to them). By providing
319 type information, LLVM can be used as the target of optimizations: for
320 example, through pointer analysis, it can be proven that a C automatic
321 variable is never accessed outside of the current function... allowing it to
322 be promoted to a simple SSA value instead of a memory location.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]323
324</div>
325
326<!-- _______________________________________________________________________ -->
327<div class="doc_subsubsection"> <a name="wellformed">Well-Formedness</a> </div>
328
329<div class="doc_text">
330
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]331<p>It is important to note that this document describes 'well formed' LLVM
332 assembly language. There is a difference between what the parser accepts and
333 what is considered 'well formed'. For example, the following instruction is
334 syntactically okay, but not well formed:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]335
336<div class="doc_code">
337<pre>
338%x = <a href="#i_add">add</a> i32 1, %x
339</pre>
340</div>
341
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]342<p>...because the definition of <tt>%x</tt> does not dominate all of its
343 uses. The LLVM infrastructure provides a verification pass that may be used
344 to verify that an LLVM module is well formed. This pass is automatically run
345 by the parser after parsing input assembly and by the optimizer before it
346 outputs bitcode. The violations pointed out by the verifier pass indicate
347 bugs in transformation passes or input to the parser.</p>
348
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]349</div>
350
Chris Lattnera83fdc02007-10-03 17:34:29 +0000[diff] [blame]351<!-- Describe the typesetting conventions here. -->
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]352
353<!-- *********************************************************************** -->
354<div class="doc_section"> <a name="identifiers">Identifiers</a> </div>
355<!-- *********************************************************************** -->
356
357<div class="doc_text">
358
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]359<p>LLVM identifiers come in two basic types: global and local. Global
360 identifiers (functions, global variables) begin with the <tt>'@'</tt>
361 character. Local identifiers (register names, types) begin with
362 the <tt>'%'</tt> character. Additionally, there are three different formats
363 for identifiers, for different purposes:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]364
365<ol>
Reid Spencerc8245b02007-08-07 14:34:28 +0000[diff] [blame]366 <li>Named values are represented as a string of characters with their prefix.
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]367 For example, <tt>%foo</tt>, <tt>@DivisionByZero</tt>,
368 <tt>%a.really.long.identifier</tt>. The actual regular expression used is
369 '<tt>[%@][a-zA-Z$._][a-zA-Z$._0-9]*</tt>'. Identifiers which require
370 other characters in their names can be surrounded with quotes. Special
371 characters may be escaped using <tt>"\xx"</tt> where <tt>xx</tt> is the
372 ASCII code for the character in hexadecimal. In this way, any character
373 can be used in a name value, even quotes themselves.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]374
Reid Spencerc8245b02007-08-07 14:34:28 +0000[diff] [blame]375 <li>Unnamed values are represented as an unsigned numeric value with their
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]376 prefix. For example, <tt>%12</tt>, <tt>@2</tt>, <tt>%44</tt>.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]377
378 <li>Constants, which are described in a <a href="#constants">section about
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]379 constants</a>, below.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]380</ol>
381
Reid Spencerc8245b02007-08-07 14:34:28 +0000[diff] [blame]382<p>LLVM requires that values start with a prefix for two reasons: Compilers
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]383 don't need to worry about name clashes with reserved words, and the set of
384 reserved words may be expanded in the future without penalty. Additionally,
385 unnamed identifiers allow a compiler to quickly come up with a temporary
386 variable without having to avoid symbol table conflicts.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]387
388<p>Reserved words in LLVM are very similar to reserved words in other
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]389 languages. There are keywords for different opcodes
390 ('<tt><a href="#i_add">add</a></tt>',
391 '<tt><a href="#i_bitcast">bitcast</a></tt>',
392 '<tt><a href="#i_ret">ret</a></tt>', etc...), for primitive type names
393 ('<tt><a href="#t_void">void</a></tt>',
394 '<tt><a href="#t_primitive">i32</a></tt>', etc...), and others. These
395 reserved words cannot conflict with variable names, because none of them
396 start with a prefix character (<tt>'%'</tt> or <tt>'@'</tt>).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]397
398<p>Here is an example of LLVM code to multiply the integer variable
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]399 '<tt>%X</tt>' by 8:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]400
401<p>The easy way:</p>
402
403<div class="doc_code">
404<pre>
405%result = <a href="#i_mul">mul</a> i32 %X, 8
406</pre>
407</div>
408
409<p>After strength reduction:</p>
410
411<div class="doc_code">
412<pre>
413%result = <a href="#i_shl">shl</a> i32 %X, i8 3
414</pre>
415</div>
416
417<p>And the hard way:</p>
418
419<div class="doc_code">
420<pre>
421<a href="#i_add">add</a> i32 %X, %X <i>; yields {i32}:%0</i>
422<a href="#i_add">add</a> i32 %0, %0 <i>; yields {i32}:%1</i>
423%result = <a href="#i_add">add</a> i32 %1, %1
424</pre>
425</div>
426
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]427<p>This last way of multiplying <tt>%X</tt> by 8 illustrates several important
428 lexical features of LLVM:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]429
430<ol>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]431 <li>Comments are delimited with a '<tt>;</tt>' and go until the end of
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]432 line.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]433
434 <li>Unnamed temporaries are created when the result of a computation is not
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]435 assigned to a named value.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]436
437 <li>Unnamed temporaries are numbered sequentially</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]438</ol>
439
440<p>...and it also shows a convention that we follow in this document. When
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]441 demonstrating instructions, we will follow an instruction with a comment that
442 defines the type and name of value produced. Comments are shown in italic
443 text.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]444
445</div>
446
447<!-- *********************************************************************** -->
448<div class="doc_section"> <a name="highlevel">High Level Structure</a> </div>
449<!-- *********************************************************************** -->
450
451<!-- ======================================================================= -->
452<div class="doc_subsection"> <a name="modulestructure">Module Structure</a>
453</div>
454
455<div class="doc_text">
456
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]457<p>LLVM programs are composed of "Module"s, each of which is a translation unit
458 of the input programs. Each module consists of functions, global variables,
459 and symbol table entries. Modules may be combined together with the LLVM
460 linker, which merges function (and global variable) definitions, resolves
461 forward declarations, and merges symbol table entries. Here is an example of
462 the "hello world" module:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]463
464<div class="doc_code">
465<pre><i>; Declare the string constant as a global constant...</i>
466<a href="#identifiers">@.LC0</a> = <a href="#linkage_internal">internal</a> <a
467 href="#globalvars">constant</a> <a href="#t_array">[13 x i8]</a> c"hello world\0A\00" <i>; [13 x i8]*</i>
468
469<i>; External declaration of the puts function</i>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]470<a href="#functionstructure">declare</a> i32 @puts(i8 *) <i>; i32(i8 *)* </i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]471
472<i>; Definition of main function</i>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]473define i32 @main() { <i>; i32()* </i>
Dan Gohman01852382009-01-04 23:44:43 +0000[diff] [blame]474 <i>; Convert [13 x i8]* to i8 *...</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]475 %cast210 = <a
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]476 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]477
478 <i>; Call puts function to write out the string to stdout...</i>
479 <a
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]480 href="#i_call">call</a> i32 @puts(i8 * %cast210) <i>; i32</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]481 <a
482 href="#i_ret">ret</a> i32 0<br>}<br>
483</pre>
484</div>
485
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]486<p>This example is made up of a <a href="#globalvars">global variable</a> named
487 "<tt>.LC0</tt>", an external declaration of the "<tt>puts</tt>" function, and
488 a <a href="#functionstructure">function definition</a> for
489 "<tt>main</tt>".</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]490
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]491<p>In general, a module is made up of a list of global values, where both
492 functions and global variables are global values. Global values are
493 represented by a pointer to a memory location (in this case, a pointer to an
494 array of char, and a pointer to a function), and have one of the
495 following <a href="#linkage">linkage types</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]496
497</div>
498
499<!-- ======================================================================= -->
500<div class="doc_subsection">
501 <a name="linkage">Linkage Types</a>
502</div>
503
504<div class="doc_text">
505
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]506<p>All Global Variables and Functions have one of the following types of
507 linkage:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]508
509<dl>
Rafael Espindolaa168fc92009-01-15 20:18:42 +0000[diff] [blame]510 <dt><tt><b><a name="linkage_private">private</a></b></tt>: </dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]511 <dd>Global values with private linkage are only directly accessible by objects
512 in the current module. In particular, linking code into a module with an
513 private global value may cause the private to be renamed as necessary to
514 avoid collisions. Because the symbol is private to the module, all
515 references can be updated. This doesn't show up in any symbol table in the
516 object file.</dd>
Rafael Espindolaa168fc92009-01-15 20:18:42 +0000[diff] [blame]517
Bill Wendling41a07852009-07-20 01:03:30 +0000[diff] [blame]518 <dt><tt><b><a name="linkage_linker_private">linker_private</a></b></tt>: </dt>
Bill Wendling41a07852009-07-20 01:03:30 +0000[diff] [blame]519 <dd>Similar to private, but the symbol is passed through the assembler and
520 removed by the linker after evaluation.</dd>
521
Dale Johannesen96e7e092008-05-23 23:13:41 +0000[diff] [blame]522 <dt><tt><b><a name="linkage_internal">internal</a></b></tt>: </dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]523 <dd>Similar to private, but the value shows as a local symbol
524 (<tt>STB_LOCAL</tt> in the case of ELF) in the object file. This
525 corresponds to the notion of the '<tt>static</tt>' keyword in C.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]526
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]527 <dt><tt><b><a name="linkage_available_externally">available_externally</a></b></tt>: </dt>
Chris Lattner68433442009-04-13 05:44:34 +0000[diff] [blame]528 <dd>Globals with "<tt>available_externally</tt>" linkage are never emitted
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]529 into the object file corresponding to the LLVM module. They exist to
530 allow inlining and other optimizations to take place given knowledge of
531 the definition of the global, which is known to be somewhere outside the
532 module. Globals with <tt>available_externally</tt> linkage are allowed to
533 be discarded at will, and are otherwise the same as <tt>linkonce_odr</tt>.
534 This linkage type is only allowed on definitions, not declarations.</dd>
Chris Lattner68433442009-04-13 05:44:34 +0000[diff] [blame]535
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]536 <dt><tt><b><a name="linkage_linkonce">linkonce</a></b></tt>: </dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]537 <dd>Globals with "<tt>linkonce</tt>" linkage are merged with other globals of
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]538 the same name when linkage occurs. This is typically used to implement
539 inline functions, templates, or other code which must be generated in each
540 translation unit that uses it. Unreferenced <tt>linkonce</tt> globals are
541 allowed to be discarded.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]542
Dale Johannesen96e7e092008-05-23 23:13:41 +0000[diff] [blame]543 <dt><tt><b><a name="linkage_common">common</a></b></tt>: </dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]544 <dd>"<tt>common</tt>" linkage is exactly the same as <tt>linkonce</tt>
545 linkage, except that unreferenced <tt>common</tt> globals may not be
546 discarded. This is used for globals that may be emitted in multiple
547 translation units, but that are not guaranteed to be emitted into every
548 translation unit that uses them. One example of this is tentative
549 definitions in C, such as "<tt>int X;</tt>" at global scope.</dd>
Dale Johannesen96e7e092008-05-23 23:13:41 +0000[diff] [blame]550
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]551 <dt><tt><b><a name="linkage_weak">weak</a></b></tt>: </dt>
Dale Johannesen96e7e092008-05-23 23:13:41 +0000[diff] [blame]552 <dd>"<tt>weak</tt>" linkage is the same as <tt>common</tt> linkage, except
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]553 that some targets may choose to emit different assembly sequences for them
554 for target-dependent reasons. This is used for globals that are declared
555 "weak" in C source code.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]556
557 <dt><tt><b><a name="linkage_appending">appending</a></b></tt>: </dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]558 <dd>"<tt>appending</tt>" linkage may only be applied to global variables of
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]559 pointer to array type. When two global variables with appending linkage
560 are linked together, the two global arrays are appended together. This is
561 the LLVM, typesafe, equivalent of having the system linker append together
562 "sections" with identical names when .o files are linked.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]563
564 <dt><tt><b><a name="linkage_externweak">extern_weak</a></b></tt>: </dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]565 <dd>The semantics of this linkage follow the ELF object file model: the symbol
566 is weak until linked, if not linked, the symbol becomes null instead of
567 being an undefined reference.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]568
Duncan Sands19d161f2009-03-07 15:45:40 +0000[diff] [blame]569 <dt><tt><b><a name="linkage_linkonce">linkonce_odr</a></b></tt>: </dt>
Duncan Sands19d161f2009-03-07 15:45:40 +0000[diff] [blame]570 <dt><tt><b><a name="linkage_weak">weak_odr</a></b></tt>: </dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]571 <dd>Some languages allow differing globals to be merged, such as two functions
572 with different semantics. Other languages, such as <tt>C++</tt>, ensure
573 that only equivalent globals are ever merged (the "one definition rule" -
574 "ODR"). Such languages can use the <tt>linkonce_odr</tt>
575 and <tt>weak_odr</tt> linkage types to indicate that the global will only
576 be merged with equivalent globals. These linkage types are otherwise the
577 same as their non-<tt>odr</tt> versions.</dd>
Duncan Sands19d161f2009-03-07 15:45:40 +0000[diff] [blame]578
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]579 <dt><tt><b><a name="linkage_external">externally visible</a></b></tt>:</dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]580 <dd>If none of the above identifiers are used, the global is externally
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]581 visible, meaning that it participates in linkage and can be used to
582 resolve external symbol references.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]583</dl>
584
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]585<p>The next two types of linkage are targeted for Microsoft Windows platform
586 only. They are designed to support importing (exporting) symbols from (to)
587 DLLs (Dynamic Link Libraries).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]588
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]589<dl>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]590 <dt><tt><b><a name="linkage_dllimport">dllimport</a></b></tt>: </dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]591 <dd>"<tt>dllimport</tt>" linkage causes the compiler to reference a function
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]592 or variable via a global pointer to a pointer that is set up by the DLL
593 exporting the symbol. On Microsoft Windows targets, the pointer name is
594 formed by combining <code>__imp_</code> and the function or variable
595 name.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]596
597 <dt><tt><b><a name="linkage_dllexport">dllexport</a></b></tt>: </dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]598 <dd>"<tt>dllexport</tt>" linkage causes the compiler to provide a global
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]599 pointer to a pointer in a DLL, so that it can be referenced with the
600 <tt>dllimport</tt> attribute. On Microsoft Windows targets, the pointer
601 name is formed by combining <code>__imp_</code> and the function or
602 variable name.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]603</dl>
604
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]605<p>For example, since the "<tt>.LC0</tt>" variable is defined to be internal, if
606 another module defined a "<tt>.LC0</tt>" variable and was linked with this
607 one, one of the two would be renamed, preventing a collision. Since
608 "<tt>main</tt>" and "<tt>puts</tt>" are external (i.e., lacking any linkage
609 declarations), they are accessible outside of the current module.</p>
610
611<p>It is illegal for a function <i>declaration</i> to have any linkage type
612 other than "externally visible", <tt>dllimport</tt>
613 or <tt>extern_weak</tt>.</p>
614
Duncan Sands19d161f2009-03-07 15:45:40 +0000[diff] [blame]615<p>Aliases can have only <tt>external</tt>, <tt>internal</tt>, <tt>weak</tt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]616 or <tt>weak_odr</tt> linkages.</p>
617
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]618</div>
619
620<!-- ======================================================================= -->
621<div class="doc_subsection">
622 <a name="callingconv">Calling Conventions</a>
623</div>
624
625<div class="doc_text">
626
627<p>LLVM <a href="#functionstructure">functions</a>, <a href="#i_call">calls</a>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]628 and <a href="#i_invoke">invokes</a> can all have an optional calling
629 convention specified for the call. The calling convention of any pair of
630 dynamic caller/callee must match, or the behavior of the program is
631 undefined. The following calling conventions are supported by LLVM, and more
632 may be added in the future:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]633
634<dl>
635 <dt><b>"<tt>ccc</tt>" - The C calling convention</b>:</dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]636 <dd>This calling convention (the default if no other calling convention is
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]637 specified) matches the target C calling conventions. This calling
638 convention supports varargs function calls and tolerates some mismatch in
639 the declared prototype and implemented declaration of the function (as
640 does normal C).</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]641
642 <dt><b>"<tt>fastcc</tt>" - The fast calling convention</b>:</dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]643 <dd>This calling convention attempts to make calls as fast as possible
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]644 (e.g. by passing things in registers). This calling convention allows the
645 target to use whatever tricks it wants to produce fast code for the
646 target, without having to conform to an externally specified ABI
647 (Application Binary Interface). Implementations of this convention should
648 allow arbitrary <a href="CodeGenerator.html#tailcallopt">tail call
649 optimization</a> to be supported. This calling convention does not
650 support varargs and requires the prototype of all callees to exactly match
651 the prototype of the function definition.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]652
653 <dt><b>"<tt>coldcc</tt>" - The cold calling convention</b>:</dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]654 <dd>This calling convention attempts to make code in the caller as efficient
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]655 as possible under the assumption that the call is not commonly executed.
656 As such, these calls often preserve all registers so that the call does
657 not break any live ranges in the caller side. This calling convention
658 does not support varargs and requires the prototype of all callees to
659 exactly match the prototype of the function definition.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]660
661 <dt><b>"<tt>cc &lt;<em>n</em>&gt;</tt>" - Numbered convention</b>:</dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]662 <dd>Any calling convention may be specified by number, allowing
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]663 target-specific calling conventions to be used. Target specific calling
664 conventions start at 64.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]665</dl>
666
667<p>More calling conventions can be added/defined on an as-needed basis, to
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]668 support Pascal conventions or any other well-known target-independent
669 convention.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]670
671</div>
672
673<!-- ======================================================================= -->
674<div class="doc_subsection">
675 <a name="visibility">Visibility Styles</a>
676</div>
677
678<div class="doc_text">
679
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]680<p>All Global Variables and Functions have one of the following visibility
681 styles:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]682
683<dl>
684 <dt><b>"<tt>default</tt>" - Default style</b>:</dt>
Chris Lattner96451482008-08-05 18:29:16 +0000[diff] [blame]685 <dd>On targets that use the ELF object file format, default visibility means
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]686 that the declaration is visible to other modules and, in shared libraries,
687 means that the declared entity may be overridden. On Darwin, default
688 visibility means that the declaration is visible to other modules. Default
689 visibility corresponds to "external linkage" in the language.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]690
691 <dt><b>"<tt>hidden</tt>" - Hidden style</b>:</dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]692 <dd>Two declarations of an object with hidden visibility refer to the same
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]693 object if they are in the same shared object. Usually, hidden visibility
694 indicates that the symbol will not be placed into the dynamic symbol
695 table, so no other module (executable or shared library) can reference it
696 directly.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]697
698 <dt><b>"<tt>protected</tt>" - Protected style</b>:</dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]699 <dd>On ELF, protected visibility indicates that the symbol will be placed in
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]700 the dynamic symbol table, but that references within the defining module
701 will bind to the local symbol. That is, the symbol cannot be overridden by
702 another module.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]703</dl>
704
705</div>
706
707<!-- ======================================================================= -->
708<div class="doc_subsection">
Chris Lattner5b6dc6e2009-01-11 20:53:49 +0000[diff] [blame]709 <a name="namedtypes">Named Types</a>
710</div>
711
712<div class="doc_text">
713
714<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]715 it easier to read the IR and make the IR more condensed (particularly when
716 recursive types are involved). An example of a name specification is:</p>
Chris Lattner5b6dc6e2009-01-11 20:53:49 +0000[diff] [blame]717
718<div class="doc_code">
719<pre>
720%mytype = type { %mytype*, i32 }
721</pre>
722</div>
723
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]724<p>You may give a name to any <a href="#typesystem">type</a> except
725 "<a href="t_void">void</a>". Type name aliases may be used anywhere a type
726 is expected with the syntax "%mytype".</p>
Chris Lattner5b6dc6e2009-01-11 20:53:49 +0000[diff] [blame]727
728<p>Note that type names are aliases for the structural type that they indicate,
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]729 and that you can therefore specify multiple names for the same type. This
730 often leads to confusing behavior when dumping out a .ll file. Since LLVM IR
731 uses structural typing, the name is not part of the type. When printing out
732 LLVM IR, the printer will pick <em>one name</em> to render all types of a
733 particular shape. This means that if you have code where two different
734 source types end up having the same LLVM type, that the dumper will sometimes
735 print the "wrong" or unexpected type. This is an important design point and
736 isn't going to change.</p>
Chris Lattner5b6dc6e2009-01-11 20:53:49 +0000[diff] [blame]737
738</div>
739
Chris Lattner5b6dc6e2009-01-11 20:53:49 +0000[diff] [blame]740<!-- ======================================================================= -->
741<div class="doc_subsection">
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]742 <a name="globalvars">Global Variables</a>
743</div>
744
745<div class="doc_text">
746
747<p>Global variables define regions of memory allocated at compilation time
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]748 instead of run-time. Global variables may optionally be initialized, may
749 have an explicit section to be placed in, and may have an optional explicit
750 alignment specified. A variable may be defined as "thread_local", which
751 means that it will not be shared by threads (each thread will have a
752 separated copy of the variable). A variable may be defined as a global
753 "constant," which indicates that the contents of the variable
754 will <b>never</b> be modified (enabling better optimization, allowing the
755 global data to be placed in the read-only section of an executable, etc).
756 Note that variables that need runtime initialization cannot be marked
757 "constant" as there is a store to the variable.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]758
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]759<p>LLVM explicitly allows <em>declarations</em> of global variables to be marked
760 constant, even if the final definition of the global is not. This capability
761 can be used to enable slightly better optimization of the program, but
762 requires the language definition to guarantee that optimizations based on the
763 'constantness' are valid for the translation units that do not include the
764 definition.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]765
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]766<p>As SSA values, global variables define pointer values that are in scope
767 (i.e. they dominate) all basic blocks in the program. Global variables
768 always define a pointer to their "content" type because they describe a
769 region of memory, and all memory objects in LLVM are accessed through
770 pointers.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]771
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]772<p>A global variable may be declared to reside in a target-specific numbered
773 address space. For targets that support them, address spaces may affect how
774 optimizations are performed and/or what target instructions are used to
775 access the variable. The default address space is zero. The address space
776 qualifier must precede any other attributes.</p>
Christopher Lambdd0049d2007-12-11 09:31:00 +0000[diff] [blame]777
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]778<p>LLVM allows an explicit section to be specified for globals. If the target
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]779 supports it, it will emit globals to the section specified.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]780
781<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]782 the alignment is set to zero, the alignment of the global is set by the
783 target to whatever it feels convenient. If an explicit alignment is
784 specified, the global is forced to have at least that much alignment. All
785 alignments must be a power of 2.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]786
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]787<p>For example, the following defines a global in a numbered address space with
788 an initializer, section, and alignment:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]789
790<div class="doc_code">
791<pre>
Dan Gohman21ef02c2009-01-11 00:40:00 +0000[diff] [blame]792@G = addrspace(5) constant float 1.0, section "foo", align 4
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]793</pre>
794</div>
795
796</div>
797
798
799<!-- ======================================================================= -->
800<div class="doc_subsection">
801 <a name="functionstructure">Functions</a>
802</div>
803
804<div class="doc_text">
805
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]806<p>LLVM function definitions consist of the "<tt>define</tt>" keyord, an
807 optional <a href="#linkage">linkage type</a>, an optional
808 <a href="#visibility">visibility style</a>, an optional
809 <a href="#callingconv">calling convention</a>, a return type, an optional
810 <a href="#paramattrs">parameter attribute</a> for the return type, a function
811 name, a (possibly empty) argument list (each with optional
812 <a href="#paramattrs">parameter attributes</a>), optional
813 <a href="#fnattrs">function attributes</a>, an optional section, an optional
814 alignment, an optional <a href="#gc">garbage collector name</a>, an opening
815 curly brace, a list of basic blocks, and a closing curly brace.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]816
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]817<p>LLVM function declarations consist of the "<tt>declare</tt>" keyword, an
818 optional <a href="#linkage">linkage type</a>, an optional
819 <a href="#visibility">visibility style</a>, an optional
820 <a href="#callingconv">calling convention</a>, a return type, an optional
821 <a href="#paramattrs">parameter attribute</a> for the return type, a function
822 name, a possibly empty list of arguments, an optional alignment, and an
823 optional <a href="#gc">garbage collector name</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]824
Chris Lattner96451482008-08-05 18:29:16 +0000[diff] [blame]825<p>A function definition contains a list of basic blocks, forming the CFG
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]826 (Control Flow Graph) for the function. Each basic block may optionally start
827 with a label (giving the basic block a symbol table entry), contains a list
828 of instructions, and ends with a <a href="#terminators">terminator</a>
829 instruction (such as a branch or function return).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]830
831<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]832 executed on entrance to the function, and it is not allowed to have
833 predecessor basic blocks (i.e. there can not be any branches to the entry
834 block of a function). Because the block can have no predecessors, it also
835 cannot have any <a href="#i_phi">PHI nodes</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]836
837<p>LLVM allows an explicit section to be specified for functions. If the target
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]838 supports it, it will emit functions to the section specified.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]839
840<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]841 the alignment is set to zero, the alignment of the function is set by the
842 target to whatever it feels convenient. If an explicit alignment is
843 specified, the function is forced to have at least that much alignment. All
844 alignments must be a power of 2.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]845
Bill Wendling6ec40612009-07-20 02:39:26 +0000[diff] [blame]846<h5>Syntax:</h5>
Devang Pateld0bfcc72008-10-07 17:48:33 +0000[diff] [blame]847<div class="doc_code">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]848<pre>
Chris Lattner1e5c5cd02008-10-13 16:55:18 +0000[diff] [blame]849define [<a href="#linkage">linkage</a>] [<a href="#visibility">visibility</a>]
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]850 [<a href="#callingconv">cconv</a>] [<a href="#paramattrs">ret attrs</a>]
851 &lt;ResultType&gt; @&lt;FunctionName&gt; ([argument list])
852 [<a href="#fnattrs">fn Attrs</a>] [section "name"] [align N]
853 [<a href="#gc">gc</a>] { ... }
854</pre>
Devang Pateld0bfcc72008-10-07 17:48:33 +0000[diff] [blame]855</div>
856
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]857</div>
858
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]859<!-- ======================================================================= -->
860<div class="doc_subsection">
861 <a name="aliasstructure">Aliases</a>
862</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]863
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]864<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]865
866<p>Aliases act as "second name" for the aliasee value (which can be either
867 function, global variable, another alias or bitcast of global value). Aliases
868 may have an optional <a href="#linkage">linkage type</a>, and an
869 optional <a href="#visibility">visibility style</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]870
Bill Wendling6ec40612009-07-20 02:39:26 +0000[diff] [blame]871<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]872<div class="doc_code">
873<pre>
Duncan Sandsd7bfabf2008-09-12 20:48:21 +0000[diff] [blame]874@&lt;Name&gt; = alias [Linkage] [Visibility] &lt;AliaseeTy&gt; @&lt;Aliasee&gt;
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]875</pre>
876</div>
877
878</div>
879
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]880<!-- ======================================================================= -->
881<div class="doc_subsection"><a name="paramattrs">Parameter Attributes</a></div>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]882
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]883<div class="doc_text">
884
885<p>The return type and each parameter of a function type may have a set of
886 <i>parameter attributes</i> associated with them. Parameter attributes are
887 used to communicate additional information about the result or parameters of
888 a function. Parameter attributes are considered to be part of the function,
889 not of the function type, so functions with different parameter attributes
890 can have the same function type.</p>
891
892<p>Parameter attributes are simple keywords that follow the type specified. If
893 multiple parameter attributes are needed, they are space separated. For
894 example:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]895
896<div class="doc_code">
897<pre>
Nick Lewycky3022a742009-02-15 23:06:14 +0000[diff] [blame]898declare i32 @printf(i8* noalias nocapture, ...)
Chris Lattnerf33b8452008-10-04 18:33:34 +0000[diff] [blame]899declare i32 @atoi(i8 zeroext)
900declare signext i8 @returns_signed_char()
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]901</pre>
902</div>
903
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]904<p>Note that any attributes for the function result (<tt>nounwind</tt>,
905 <tt>readonly</tt>) come immediately after the argument list.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]906
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]907<p>Currently, only the following parameter attributes are defined:</p>
Chris Lattner275e6be2008-01-11 06:20:47 +0000[diff] [blame]908
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]909<dl>
910 <dt><tt>zeroext</tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]911 <dd>This indicates to the code generator that the parameter or return value
912 should be zero-extended to a 32-bit value by the caller (for a parameter)
913 or the callee (for a return value).</dd>
Chris Lattner275e6be2008-01-11 06:20:47 +0000[diff] [blame]914
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]915 <dt><tt>signext</tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]916 <dd>This indicates to the code generator that the parameter or return value
917 should be sign-extended to a 32-bit value by the caller (for a parameter)
918 or the callee (for a return value).</dd>
Chris Lattner275e6be2008-01-11 06:20:47 +0000[diff] [blame]919
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]920 <dt><tt>inreg</tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]921 <dd>This indicates that this parameter or return value should be treated in a
922 special target-dependent fashion during while emitting code for a function
923 call or return (usually, by putting it in a register as opposed to memory,
924 though some targets use it to distinguish between two different kinds of
925 registers). Use of this attribute is target-specific.</dd>
Chris Lattner275e6be2008-01-11 06:20:47 +0000[diff] [blame]926
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]927 <dt><tt><a name="byval">byval</a></tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]928 <dd>This indicates that the pointer parameter should really be passed by value
929 to the function. The attribute implies that a hidden copy of the pointee
930 is made between the caller and the callee, so the callee is unable to
931 modify the value in the callee. This attribute is only valid on LLVM
932 pointer arguments. It is generally used to pass structs and arrays by
933 value, but is also valid on pointers to scalars. The copy is considered
934 to belong to the caller not the callee (for example,
935 <tt><a href="#readonly">readonly</a></tt> functions should not write to
936 <tt>byval</tt> parameters). This is not a valid attribute for return
937 values. The byval attribute also supports specifying an alignment with
938 the align attribute. This has a target-specific effect on the code
939 generator that usually indicates a desired alignment for the synthesized
940 stack slot.</dd>
941
942 <dt><tt>sret</tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]943 <dd>This indicates that the pointer parameter specifies the address of a
944 structure that is the return value of the function in the source program.
945 This pointer must be guaranteed by the caller to be valid: loads and
946 stores to the structure may be assumed by the callee to not to trap. This
947 may only be applied to the first parameter. This is not a valid attribute
948 for return values. </dd>
949
950 <dt><tt>noalias</tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]951 <dd>This indicates that the pointer does not alias any global or any other
952 parameter. The caller is responsible for ensuring that this is the
953 case. On a function return value, <tt>noalias</tt> additionally indicates
954 that the pointer does not alias any other pointers visible to the
955 caller. For further details, please see the discussion of the NoAlias
956 response in
957 <a href="http://llvm.org/docs/AliasAnalysis.html#MustMayNo">alias
958 analysis</a>.</dd>
959
960 <dt><tt>nocapture</tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]961 <dd>This indicates that the callee does not make any copies of the pointer
962 that outlive the callee itself. This is not a valid attribute for return
963 values.</dd>
964
965 <dt><tt>nest</tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]966 <dd>This indicates that the pointer parameter can be excised using the
967 <a href="#int_trampoline">trampoline intrinsics</a>. This is not a valid
968 attribute for return values.</dd>
969</dl>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]970
971</div>
972
973<!-- ======================================================================= -->
974<div class="doc_subsection">
Gordon Henriksen13fe5e32007-12-10 03:18:06 +0000[diff] [blame]975 <a name="gc">Garbage Collector Names</a>
976</div>
977
978<div class="doc_text">
Gordon Henriksen13fe5e32007-12-10 03:18:06 +0000[diff] [blame]979
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]980<p>Each function may specify a garbage collector name, which is simply a
981 string:</p>
982
983<div class="doc_code">
984<pre>
985define void @f() gc "name" { ...
986</pre>
987</div>
Gordon Henriksen13fe5e32007-12-10 03:18:06 +0000[diff] [blame]988
989<p>The compiler declares the supported values of <i>name</i>. Specifying a
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]990 collector which will cause the compiler to alter its output in order to
991 support the named garbage collection algorithm.</p>
992
Gordon Henriksen13fe5e32007-12-10 03:18:06 +0000[diff] [blame]993</div>
994
995<!-- ======================================================================= -->
996<div class="doc_subsection">
Devang Patel008cd3e2008-09-26 23:51:19 +0000[diff] [blame]997 <a name="fnattrs">Function Attributes</a>
Devang Pateld468f1c2008-09-04 23:05:13 +0000[diff] [blame]998</div>
999
1000<div class="doc_text">
Devang Patel008cd3e2008-09-26 23:51:19 +0000[diff] [blame]1001
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1002<p>Function attributes are set to communicate additional information about a
1003 function. Function attributes are considered to be part of the function, not
1004 of the function type, so functions with different parameter attributes can
1005 have the same function type.</p>
Devang Patel008cd3e2008-09-26 23:51:19 +0000[diff] [blame]1006
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1007<p>Function attributes are simple keywords that follow the type specified. If
1008 multiple attributes are needed, they are space separated. For example:</p>
Devang Pateld468f1c2008-09-04 23:05:13 +0000[diff] [blame]1009
1010<div class="doc_code">
Bill Wendling74d3eac2008-09-07 10:26:33 +0000[diff] [blame]1011<pre>
Devang Patel008cd3e2008-09-26 23:51:19 +0000[diff] [blame]1012define void @f() noinline { ... }
1013define void @f() alwaysinline { ... }
1014define void @f() alwaysinline optsize { ... }
1015define void @f() optsize
Bill Wendling74d3eac2008-09-07 10:26:33 +0000[diff] [blame]1016</pre>
Devang Pateld468f1c2008-09-04 23:05:13 +0000[diff] [blame]1017</div>
1018
Bill Wendling74d3eac2008-09-07 10:26:33 +0000[diff] [blame]1019<dl>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1020 <dt><tt>alwaysinline</tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1021 <dd>This attribute indicates that the inliner should attempt to inline this
1022 function into callers whenever possible, ignoring any active inlining size
1023 threshold for this caller.</dd>
Bill Wendling74d3eac2008-09-07 10:26:33 +0000[diff] [blame]1024
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1025 <dt><tt>noinline</tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1026 <dd>This attribute indicates that the inliner should never inline this
1027 function in any situation. This attribute may not be used together with
1028 the <tt>alwaysinline</tt> attribute.</dd>
Devang Patel008cd3e2008-09-26 23:51:19 +0000[diff] [blame]1029
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1030 <dt><tt>optsize</tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1031 <dd>This attribute suggests that optimization passes and code generator passes
1032 make choices that keep the code size of this function low, and otherwise
1033 do optimizations specifically to reduce code size.</dd>
Devang Patel008cd3e2008-09-26 23:51:19 +0000[diff] [blame]1034
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1035 <dt><tt>noreturn</tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1036 <dd>This function attribute indicates that the function never returns
1037 normally. This produces undefined behavior at runtime if the function
1038 ever does dynamically return.</dd>
Bill Wendlingdfaabba2008-11-13 01:02:51 +0000[diff] [blame]1039
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1040 <dt><tt>nounwind</tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1041 <dd>This function attribute indicates that the function never returns with an
1042 unwind or exceptional control flow. If the function does unwind, its
1043 runtime behavior is undefined.</dd>
Bill Wendlingbe9ec3f2008-11-26 19:07:40 +0000[diff] [blame]1044
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1045 <dt><tt>readnone</tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1046 <dd>This attribute indicates that the function computes its result (or decides
1047 to unwind an exception) based strictly on its arguments, without
1048 dereferencing any pointer arguments or otherwise accessing any mutable
1049 state (e.g. memory, control registers, etc) visible to caller functions.
1050 It does not write through any pointer arguments
1051 (including <tt><a href="#byval">byval</a></tt> arguments) and never
1052 changes any state visible to callers. This means that it cannot unwind
1053 exceptions by calling the <tt>C++</tt> exception throwing methods, but
1054 could use the <tt>unwind</tt> instruction.</dd>
Devang Patela2f9f412009-06-12 19:45:19 +0000[diff] [blame]1055
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1056 <dt><tt><a name="readonly">readonly</a></tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1057 <dd>This attribute indicates that the function does not write through any
1058 pointer arguments (including <tt><a href="#byval">byval</a></tt>
1059 arguments) or otherwise modify any state (e.g. memory, control registers,
1060 etc) visible to caller functions. It may dereference pointer arguments
1061 and read state that may be set in the caller. A readonly function always
1062 returns the same value (or unwinds an exception identically) when called
1063 with the same set of arguments and global state. It cannot unwind an
1064 exception by calling the <tt>C++</tt> exception throwing methods, but may
1065 use the <tt>unwind</tt> instruction.</dd>
Anton Korobeynikovedd7d112009-07-17 18:07:26 +0000[diff] [blame]1066
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1067 <dt><tt><a name="ssp">ssp</a></tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1068 <dd>This attribute indicates that the function should emit a stack smashing
1069 protector. It is in the form of a "canary"&mdash;a random value placed on
1070 the stack before the local variables that's checked upon return from the
1071 function to see if it has been overwritten. A heuristic is used to
1072 determine if a function needs stack protectors or not.<br>
1073<br>
1074 If a function that has an <tt>ssp</tt> attribute is inlined into a
1075 function that doesn't have an <tt>ssp</tt> attribute, then the resulting
1076 function will have an <tt>ssp</tt> attribute.</dd>
1077
1078 <dt><tt>sspreq</tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1079 <dd>This attribute indicates that the function should <em>always</em> emit a
1080 stack smashing protector. This overrides
Bill Wendling6ec40612009-07-20 02:39:26 +0000[diff] [blame]1081 the <tt><a href="#ssp">ssp</a></tt> function attribute.<br>
1082<br>
1083 If a function that has an <tt>sspreq</tt> attribute is inlined into a
1084 function that doesn't have an <tt>sspreq</tt> attribute or which has
1085 an <tt>ssp</tt> attribute, then the resulting function will have
1086 an <tt>sspreq</tt> attribute.</dd>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1087
1088 <dt><tt>noredzone</tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1089 <dd>This attribute indicates that the code generator should not use a red
1090 zone, even if the target-specific ABI normally permits it.</dd>
1091
1092 <dt><tt>noimplicitfloat</tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1093 <dd>This attributes disables implicit floating point instructions.</dd>
1094
1095 <dt><tt>naked</tt></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1096 <dd>This attribute disables prologue / epilogue emission for the function.
1097 This can have very system-specific consequences.</dd>
Bill Wendling74d3eac2008-09-07 10:26:33 +0000[diff] [blame]1098</dl>
1099
Devang Pateld468f1c2008-09-04 23:05:13 +0000[diff] [blame]1100</div>
1101
1102<!-- ======================================================================= -->
1103<div class="doc_subsection">
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1104 <a name="moduleasm">Module-Level Inline Assembly</a>
1105</div>
1106
1107<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1108
1109<p>Modules may contain "module-level inline asm" blocks, which corresponds to
1110 the GCC "file scope inline asm" blocks. These blocks are internally
1111 concatenated by LLVM and treated as a single unit, but may be separated in
1112 the <tt>.ll</tt> file if desired. The syntax is very simple:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1113
1114<div class="doc_code">
1115<pre>
1116module asm "inline asm code goes here"
1117module asm "more can go here"
1118</pre>
1119</div>
1120
1121<p>The strings can contain any character by escaping non-printable characters.
1122 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]1123 for the number.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1124
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1125<p>The inline asm code is simply printed to the machine code .s file when
1126 assembly code is generated.</p>
1127
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1128</div>
1129
1130<!-- ======================================================================= -->
1131<div class="doc_subsection">
1132 <a name="datalayout">Data Layout</a>
1133</div>
1134
1135<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1136
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1137<p>A module may specify a target specific data layout string that specifies how
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1138 data is to be laid out in memory. The syntax for the data layout is
1139 simply:</p>
1140
1141<div class="doc_code">
1142<pre>
1143target datalayout = "<i>layout specification</i>"
1144</pre>
1145</div>
1146
1147<p>The <i>layout specification</i> consists of a list of specifications
1148 separated by the minus sign character ('-'). Each specification starts with
1149 a letter and may include other information after the letter to define some
1150 aspect of the data layout. The specifications accepted are as follows:</p>
1151
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1152<dl>
1153 <dt><tt>E</tt></dt>
1154 <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]1155 bits with the most significance have the lowest address location.</dd>
1156
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1157 <dt><tt>e</tt></dt>
Chris Lattner96451482008-08-05 18:29:16 +0000[diff] [blame]1158 <dd>Specifies that the target lays out data in little-endian form. That is,
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1159 the bits with the least significance have the lowest address
1160 location.</dd>
1161
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1162 <dt><tt>p:<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1163 <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]1164 <i>preferred</i> alignments. All sizes are in bits. Specifying
1165 the <i>pref</i> alignment is optional. If omitted, the
1166 preceding <tt>:</tt> should be omitted too.</dd>
1167
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1168 <dt><tt>i<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1169 <dd>This specifies the alignment for an integer type of a given bit
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1170 <i>size</i>. The value of <i>size</i> must be in the range [1,2^23).</dd>
1171
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1172 <dt><tt>v<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1173 <dd>This specifies the alignment for a vector type of a given bit
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1174 <i>size</i>.</dd>
1175
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1176 <dt><tt>f<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1177 <dd>This specifies the alignment for a floating point type of a given bit
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1178 <i>size</i>. The value of <i>size</i> must be either 32 (float) or 64
1179 (double).</dd>
1180
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1181 <dt><tt>a<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1182 <dd>This specifies the alignment for an aggregate type of a given bit
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1183 <i>size</i>.</dd>
1184
Daniel Dunbard88a97b2009-06-08 22:17:53 +0000[diff] [blame]1185 <dt><tt>s<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1186 <dd>This specifies the alignment for a stack object of a given bit
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1187 <i>size</i>.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1188</dl>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1189
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1190<p>When constructing the data layout for a given target, LLVM starts with a
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1191 default set of specifications which are then (possibly) overriden by the
1192 specifications in the <tt>datalayout</tt> keyword. The default specifications
1193 are given in this list:</p>
1194
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1195<ul>
1196 <li><tt>E</tt> - big endian</li>
1197 <li><tt>p:32:64:64</tt> - 32-bit pointers with 64-bit alignment</li>
1198 <li><tt>i1:8:8</tt> - i1 is 8-bit (byte) aligned</li>
1199 <li><tt>i8:8:8</tt> - i8 is 8-bit (byte) aligned</li>
1200 <li><tt>i16:16:16</tt> - i16 is 16-bit aligned</li>
1201 <li><tt>i32:32:32</tt> - i32 is 32-bit aligned</li>
Chris Lattner96451482008-08-05 18:29:16 +0000[diff] [blame]1202 <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]1203 alignment of 64-bits</li>
1204 <li><tt>f32:32:32</tt> - float is 32-bit aligned</li>
1205 <li><tt>f64:64:64</tt> - double is 64-bit aligned</li>
1206 <li><tt>v64:64:64</tt> - 64-bit vector is 64-bit aligned</li>
1207 <li><tt>v128:128:128</tt> - 128-bit vector is 128-bit aligned</li>
1208 <li><tt>a0:0:1</tt> - aggregates are 8-bit aligned</li>
Daniel Dunbard88a97b2009-06-08 22:17:53 +0000[diff] [blame]1209 <li><tt>s0:64:64</tt> - stack objects are 64-bit aligned</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1210</ul>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1211
1212<p>When LLVM is determining the alignment for a given type, it uses the
1213 following rules:</p>
1214
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1215<ol>
1216 <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]1217 specification is used.</li>
1218
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1219 <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]1220 smallest integer type that is larger than the bitwidth of the sought type
1221 is used. If none of the specifications are larger than the bitwidth then
1222 the the largest integer type is used. For example, given the default
1223 specifications above, the i7 type will use the alignment of i8 (next
1224 largest) while both i65 and i256 will use the alignment of i64 (largest
1225 specified).</li>
1226
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1227 <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]1228 largest vector type that is smaller than the sought vector type will be
1229 used as a fall back. This happens because &lt;128 x double&gt; can be
1230 implemented in terms of 64 &lt;2 x double&gt;, for example.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1231</ol>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1232
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1233</div>
1234
1235<!-- *********************************************************************** -->
1236<div class="doc_section"> <a name="typesystem">Type System</a> </div>
1237<!-- *********************************************************************** -->
1238
1239<div class="doc_text">
1240
1241<p>The LLVM type system is one of the most important features of the
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1242 intermediate representation. Being typed enables a number of optimizations
1243 to be performed on the intermediate representation directly, without having
1244 to do extra analyses on the side before the transformation. A strong type
1245 system makes it easier to read the generated code and enables novel analyses
1246 and transformations that are not feasible to perform on normal three address
1247 code representations.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1248
1249</div>
1250
1251<!-- ======================================================================= -->
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1252<div class="doc_subsection"> <a name="t_classifications">Type
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1253Classifications</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1254
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1255<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1256
1257<p>The types fall into a few useful classifications:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1258
1259<table border="1" cellspacing="0" cellpadding="4">
1260 <tbody>
1261 <tr><th>Classification</th><th>Types</th></tr>
1262 <tr>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1263 <td><a href="#t_integer">integer</a></td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1264 <td><tt>i1, i2, i3, ... i8, ... i16, ... i32, ... i64, ... </tt></td>
1265 </tr>
1266 <tr>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1267 <td><a href="#t_floating">floating point</a></td>
1268 <td><tt>float, double, x86_fp80, fp128, ppc_fp128</tt></td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1269 </tr>
1270 <tr>
1271 <td><a name="t_firstclass">first class</a></td>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1272 <td><a href="#t_integer">integer</a>,
1273 <a href="#t_floating">floating point</a>,
1274 <a href="#t_pointer">pointer</a>,
Dan Gohmanf6237db2008-06-18 18:42:13 +0000[diff] [blame]1275 <a href="#t_vector">vector</a>,
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]1276 <a href="#t_struct">structure</a>,
1277 <a href="#t_array">array</a>,
Nick Lewycky29aaef82009-05-30 05:06:04 +0000[diff] [blame]1278 <a href="#t_label">label</a>,
1279 <a href="#t_metadata">metadata</a>.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1280 </td>
1281 </tr>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1282 <tr>
1283 <td><a href="#t_primitive">primitive</a></td>
1284 <td><a href="#t_label">label</a>,
1285 <a href="#t_void">void</a>,
Nick Lewycky29aaef82009-05-30 05:06:04 +0000[diff] [blame]1286 <a href="#t_floating">floating point</a>,
1287 <a href="#t_metadata">metadata</a>.</td>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1288 </tr>
1289 <tr>
1290 <td><a href="#t_derived">derived</a></td>
1291 <td><a href="#t_integer">integer</a>,
1292 <a href="#t_array">array</a>,
1293 <a href="#t_function">function</a>,
1294 <a href="#t_pointer">pointer</a>,
1295 <a href="#t_struct">structure</a>,
1296 <a href="#t_pstruct">packed structure</a>,
1297 <a href="#t_vector">vector</a>,
1298 <a href="#t_opaque">opaque</a>.
Dan Gohman032ba852008-10-14 16:32:04 +0000[diff] [blame]1299 </td>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1300 </tr>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1301 </tbody>
1302</table>
1303
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1304<p>The <a href="#t_firstclass">first class</a> types are perhaps the most
1305 important. Values of these types are the only ones which can be produced by
1306 instructions, passed as arguments, or used as operands to instructions.</p>
1307
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1308</div>
1309
1310<!-- ======================================================================= -->
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1311<div class="doc_subsection"> <a name="t_primitive">Primitive Types</a> </div>
Chris Lattner86437612008-01-04 04:34:14 +0000[diff] [blame]1312
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1313<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1314
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1315<p>The primitive types are the fundamental building blocks of the LLVM
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1316 system.</p>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1317
Chris Lattner86437612008-01-04 04:34:14 +0000[diff] [blame]1318</div>
1319
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1320<!-- _______________________________________________________________________ -->
1321<div class="doc_subsubsection"> <a name="t_floating">Floating Point Types</a> </div>
1322
1323<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1324
1325<table>
1326 <tbody>
1327 <tr><th>Type</th><th>Description</th></tr>
1328 <tr><td><tt>float</tt></td><td>32-bit floating point value</td></tr>
1329 <tr><td><tt>double</tt></td><td>64-bit floating point value</td></tr>
1330 <tr><td><tt>fp128</tt></td><td>128-bit floating point value (112-bit mantissa)</td></tr>
1331 <tr><td><tt>x86_fp80</tt></td><td>80-bit floating point value (X87)</td></tr>
1332 <tr><td><tt>ppc_fp128</tt></td><td>128-bit floating point value (two 64-bits)</td></tr>
1333 </tbody>
1334</table>
1335
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1336</div>
1337
1338<!-- _______________________________________________________________________ -->
1339<div class="doc_subsubsection"> <a name="t_void">Void Type</a> </div>
1340
1341<div class="doc_text">
Bill Wendling6ec40612009-07-20 02:39:26 +0000[diff] [blame]1342
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1343<h5>Overview:</h5>
1344<p>The void type does not represent any value and has no size.</p>
1345
1346<h5>Syntax:</h5>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1347<pre>
1348 void
1349</pre>
Bill Wendling6ec40612009-07-20 02:39:26 +0000[diff] [blame]1350
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1351</div>
1352
1353<!-- _______________________________________________________________________ -->
1354<div class="doc_subsubsection"> <a name="t_label">Label Type</a> </div>
1355
1356<div class="doc_text">
Bill Wendling6ec40612009-07-20 02:39:26 +0000[diff] [blame]1357
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1358<h5>Overview:</h5>
1359<p>The label type represents code labels.</p>
1360
1361<h5>Syntax:</h5>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1362<pre>
1363 label
1364</pre>
Bill Wendling6ec40612009-07-20 02:39:26 +0000[diff] [blame]1365
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1366</div>
1367
Nick Lewycky29aaef82009-05-30 05:06:04 +0000[diff] [blame]1368<!-- _______________________________________________________________________ -->
1369<div class="doc_subsubsection"> <a name="t_metadata">Metadata Type</a> </div>
1370
1371<div class="doc_text">
Bill Wendling6ec40612009-07-20 02:39:26 +0000[diff] [blame]1372
Nick Lewycky29aaef82009-05-30 05:06:04 +0000[diff] [blame]1373<h5>Overview:</h5>
1374<p>The metadata type represents embedded metadata. The only derived type that
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1375 may contain metadata is <tt>metadata*</tt> or a function type that returns or
1376 takes metadata typed parameters, but not pointer to metadata types.</p>
Nick Lewycky29aaef82009-05-30 05:06:04 +0000[diff] [blame]1377
1378<h5>Syntax:</h5>
Nick Lewycky29aaef82009-05-30 05:06:04 +0000[diff] [blame]1379<pre>
1380 metadata
1381</pre>
Bill Wendling6ec40612009-07-20 02:39:26 +0000[diff] [blame]1382
Nick Lewycky29aaef82009-05-30 05:06:04 +0000[diff] [blame]1383</div>
1384
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1385
1386<!-- ======================================================================= -->
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1387<div class="doc_subsection"> <a name="t_derived">Derived Types</a> </div>
1388
1389<div class="doc_text">
1390
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1391<p>The real power in LLVM comes from the derived types in the system. This is
1392 what allows a programmer to represent arrays, functions, pointers, and other
1393 useful types. Note that these derived types may be recursive: For example,
1394 it is possible to have a two dimensional array.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1395
1396</div>
1397
1398<!-- _______________________________________________________________________ -->
1399<div class="doc_subsubsection"> <a name="t_integer">Integer Type</a> </div>
1400
1401<div class="doc_text">
1402
1403<h5>Overview:</h5>
1404<p>The integer type is a very simple derived type that simply specifies an
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1405 arbitrary bit width for the integer type desired. Any bit width from 1 bit to
1406 2^23-1 (about 8 million) can be specified.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1407
1408<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1409<pre>
1410 iN
1411</pre>
1412
1413<p>The number of bits the integer will occupy is specified by the <tt>N</tt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1414 value.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1415
1416<h5>Examples:</h5>
1417<table class="layout">
Nick Lewycky39382d62009-05-24 02:46:06 +0000[diff] [blame]1418 <tr class="layout">
1419 <td class="left"><tt>i1</tt></td>
1420 <td class="left">a single-bit integer.</td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1421 </tr>
Nick Lewycky39382d62009-05-24 02:46:06 +0000[diff] [blame]1422 <tr class="layout">
1423 <td class="left"><tt>i32</tt></td>
1424 <td class="left">a 32-bit integer.</td>
1425 </tr>
1426 <tr class="layout">
1427 <td class="left"><tt>i1942652</tt></td>
1428 <td class="left">a really big integer of over 1 million bits.</td>
1429 </tr>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1430</table>
djge93155c2009-01-24 15:58:40 +0000[diff] [blame]1431
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1432<p>Note that the code generator does not yet support large integer types to be
1433 used as function return types. The specific limit on how large a return type
1434 the code generator can currently handle is target-dependent; currently it's
1435 often 64 bits for 32-bit targets and 128 bits for 64-bit targets.</p>
djge93155c2009-01-24 15:58:40 +0000[diff] [blame]1436
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1437</div>
1438
1439<!-- _______________________________________________________________________ -->
1440<div class="doc_subsubsection"> <a name="t_array">Array Type</a> </div>
1441
1442<div class="doc_text">
1443
1444<h5>Overview:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1445<p>The array type is a very simple derived type that arranges elements
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1446 sequentially in memory. The array type requires a size (number of elements)
1447 and an underlying data type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1448
1449<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1450<pre>
1451 [&lt;# elements&gt; x &lt;elementtype&gt;]
1452</pre>
1453
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1454<p>The number of elements is a constant integer value; <tt>elementtype</tt> may
1455 be any type with a size.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1456
1457<h5>Examples:</h5>
1458<table class="layout">
1459 <tr class="layout">
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1460 <td class="left"><tt>[40 x i32]</tt></td>
1461 <td class="left">Array of 40 32-bit integer values.</td>
1462 </tr>
1463 <tr class="layout">
1464 <td class="left"><tt>[41 x i32]</tt></td>
1465 <td class="left">Array of 41 32-bit integer values.</td>
1466 </tr>
1467 <tr class="layout">
1468 <td class="left"><tt>[4 x i8]</tt></td>
1469 <td class="left">Array of 4 8-bit integer values.</td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1470 </tr>
1471</table>
1472<p>Here are some examples of multidimensional arrays:</p>
1473<table class="layout">
1474 <tr class="layout">
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1475 <td class="left"><tt>[3 x [4 x i32]]</tt></td>
1476 <td class="left">3x4 array of 32-bit integer values.</td>
1477 </tr>
1478 <tr class="layout">
1479 <td class="left"><tt>[12 x [10 x float]]</tt></td>
1480 <td class="left">12x10 array of single precision floating point values.</td>
1481 </tr>
1482 <tr class="layout">
1483 <td class="left"><tt>[2 x [3 x [4 x i16]]]</tt></td>
1484 <td class="left">2x3x4 array of 16-bit integer values.</td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1485 </tr>
1486</table>
1487
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1488<p>Note that 'variable sized arrays' can be implemented in LLVM with a zero
1489 length array. Normally, accesses past the end of an array are undefined in
1490 LLVM (e.g. it is illegal to access the 5th element of a 3 element array). As
1491 a special case, however, zero length arrays are recognized to be variable
1492 length. This allows implementation of 'pascal style arrays' with the LLVM
1493 type "<tt>{ i32, [0 x float]}</tt>", for example.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1494
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1495<p>Note that the code generator does not yet support large aggregate types to be
1496 used as function return types. The specific limit on how large an aggregate
1497 return type the code generator can currently handle is target-dependent, and
1498 also dependent on the aggregate element types.</p>
djge93155c2009-01-24 15:58:40 +0000[diff] [blame]1499
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1500</div>
1501
1502<!-- _______________________________________________________________________ -->
1503<div class="doc_subsubsection"> <a name="t_function">Function Type</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1504
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1505<div class="doc_text">
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]1506
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1507<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1508<p>The function type can be thought of as a function signature. It consists of
1509 a return type and a list of formal parameter types. The return type of a
1510 function type is a scalar type, a void type, or a struct type. If the return
1511 type is a struct type then all struct elements must be of first class types,
1512 and the struct must have at least one element.</p>
Devang Patela3cc5372008-03-10 20:49:15 +0000[diff] [blame]1513
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1514<h5>Syntax:</h5>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]1515<pre>
1516 &lt;returntype list&gt; (&lt;parameter list&gt;)
1517</pre>
1518
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1519<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]1520 specifiers. Optionally, the parameter list may include a type <tt>...</tt>,
1521 which indicates that the function takes a variable number of arguments.
1522 Variable argument functions can access their arguments with
1523 the <a href="#int_varargs">variable argument handling intrinsic</a>
1524 functions. '<tt>&lt;returntype list&gt;</tt>' is a comma-separated list of
1525 <a href="#t_firstclass">first class</a> type specifiers.</p>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]1526
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1527<h5>Examples:</h5>
1528<table class="layout">
1529 <tr class="layout">
1530 <td class="left"><tt>i32 (i32)</tt></td>
1531 <td class="left">function taking an <tt>i32</tt>, returning an <tt>i32</tt>
1532 </td>
1533 </tr><tr class="layout">
Reid Spencerf234bed2007-07-19 23:13:04 +0000[diff] [blame]1534 <td class="left"><tt>float&nbsp;(i16&nbsp;signext,&nbsp;i32&nbsp;*)&nbsp;*
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1535 </tt></td>
1536 <td class="left"><a href="#t_pointer">Pointer</a> to a function that takes
1537 an <tt>i16</tt> that should be sign extended and a
1538 <a href="#t_pointer">pointer</a> to <tt>i32</tt>, returning
1539 <tt>float</tt>.
1540 </td>
1541 </tr><tr class="layout">
1542 <td class="left"><tt>i32 (i8*, ...)</tt></td>
1543 <td class="left">A vararg function that takes at least one
1544 <a href="#t_pointer">pointer</a> to <tt>i8 </tt> (char in C),
1545 which returns an integer. This is the signature for <tt>printf</tt> in
1546 LLVM.
1547 </td>
Devang Pateld4ba41d2008-03-24 05:35:41 +0000[diff] [blame]1548 </tr><tr class="layout">
1549 <td class="left"><tt>{i32, i32} (i32)</tt></td>
Misha Brukmanafc88b02008-11-27 06:41:20 +0000[diff] [blame]1550 <td class="left">A function taking an <tt>i32</tt>, returning two
1551 <tt>i32</tt> values as an aggregate of type <tt>{ i32, i32 }</tt>
Devang Pateld4ba41d2008-03-24 05:35:41 +0000[diff] [blame]1552 </td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1553 </tr>
1554</table>
1555
1556</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1557
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1558<!-- _______________________________________________________________________ -->
1559<div class="doc_subsubsection"> <a name="t_struct">Structure Type</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1560
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1561<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1562
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1563<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1564<p>The structure type is used to represent a collection of data members together
1565 in memory. The packing of the field types is defined to match the ABI of the
1566 underlying processor. The elements of a structure may be any type that has a
1567 size.</p>
1568
1569<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt> and
1570 '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a field with
1571 the '<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.</p>
1572
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1573<h5>Syntax:</h5>
Bill Wendling6ec40612009-07-20 02:39:26 +0000[diff] [blame]1574<pre>
1575 { &lt;type list&gt; }
1576</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1577
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1578<h5>Examples:</h5>
1579<table class="layout">
1580 <tr class="layout">
1581 <td class="left"><tt>{ i32, i32, i32 }</tt></td>
1582 <td class="left">A triple of three <tt>i32</tt> values</td>
1583 </tr><tr class="layout">
1584 <td class="left"><tt>{&nbsp;float,&nbsp;i32&nbsp;(i32)&nbsp;*&nbsp;}</tt></td>
1585 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1586 second element is a <a href="#t_pointer">pointer</a> to a
1587 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1588 an <tt>i32</tt>.</td>
1589 </tr>
1590</table>
djge93155c2009-01-24 15:58:40 +0000[diff] [blame]1591
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1592<p>Note that the code generator does not yet support large aggregate types to be
1593 used as function return types. The specific limit on how large an aggregate
1594 return type the code generator can currently handle is target-dependent, and
1595 also dependent on the aggregate element types.</p>
djge93155c2009-01-24 15:58:40 +0000[diff] [blame]1596
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1597</div>
1598
1599<!-- _______________________________________________________________________ -->
1600<div class="doc_subsubsection"> <a name="t_pstruct">Packed Structure Type</a>
1601</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1602
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1603<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1604
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1605<h5>Overview:</h5>
1606<p>The packed structure type is used to represent a collection of data members
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1607 together in memory. There is no padding between fields. Further, the
1608 alignment of a packed structure is 1 byte. The elements of a packed
1609 structure may be any type that has a size.</p>
1610
1611<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt> and
1612 '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a field with
1613 the '<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.</p>
1614
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1615<h5>Syntax:</h5>
Bill Wendling6ec40612009-07-20 02:39:26 +0000[diff] [blame]1616<pre>
1617 &lt; { &lt;type list&gt; } &gt;
1618</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1619
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1620<h5>Examples:</h5>
1621<table class="layout">
1622 <tr class="layout">
1623 <td class="left"><tt>&lt; { i32, i32, i32 } &gt;</tt></td>
1624 <td class="left">A triple of three <tt>i32</tt> values</td>
1625 </tr><tr class="layout">
Bill Wendling74d3eac2008-09-07 10:26:33 +0000[diff] [blame]1626 <td class="left">
1627<tt>&lt;&nbsp;{&nbsp;float,&nbsp;i32&nbsp;(i32)*&nbsp;}&nbsp;&gt;</tt></td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1628 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1629 second element is a <a href="#t_pointer">pointer</a> to a
1630 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1631 an <tt>i32</tt>.</td>
1632 </tr>
1633</table>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1634
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1635</div>
1636
1637<!-- _______________________________________________________________________ -->
1638<div class="doc_subsubsection"> <a name="t_pointer">Pointer Type</a> </div>
Chris Lattner96edbd32009-02-08 19:53:29 +0000[diff] [blame]1639
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1640<div class="doc_text">
1641
1642<h5>Overview:</h5>
1643<p>As in many languages, the pointer type represents a pointer or reference to
1644 another object, which must live in memory. Pointer types may have an optional
1645 address space attribute defining the target-specific numbered address space
1646 where the pointed-to object resides. The default address space is zero.</p>
1647
1648<p>Note that LLVM does not permit pointers to void (<tt>void*</tt>) nor does it
1649 permit pointers to labels (<tt>label*</tt>). Use <tt>i8*</tt> instead.</p>
Chris Lattner96edbd32009-02-08 19:53:29 +0000[diff] [blame]1650
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1651<h5>Syntax:</h5>
Bill Wendling6ec40612009-07-20 02:39:26 +0000[diff] [blame]1652<pre>
1653 &lt;type&gt; *
1654</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1655
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1656<h5>Examples:</h5>
1657<table class="layout">
1658 <tr class="layout">
Dan Gohman01852382009-01-04 23:44:43 +0000[diff] [blame]1659 <td class="left"><tt>[4 x i32]*</tt></td>
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1660 <td class="left">A <a href="#t_pointer">pointer</a> to <a
1661 href="#t_array">array</a> of four <tt>i32</tt> values.</td>
1662 </tr>
1663 <tr class="layout">
1664 <td class="left"><tt>i32 (i32 *) *</tt></td>
1665 <td class="left"> A <a href="#t_pointer">pointer</a> to a <a
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1666 href="#t_function">function</a> that takes an <tt>i32*</tt>, returning an
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1667 <tt>i32</tt>.</td>
1668 </tr>
1669 <tr class="layout">
1670 <td class="left"><tt>i32 addrspace(5)*</tt></td>
1671 <td class="left">A <a href="#t_pointer">pointer</a> to an <tt>i32</tt> value
1672 that resides in address space #5.</td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1673 </tr>
1674</table>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1675
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1676</div>
1677
1678<!-- _______________________________________________________________________ -->
1679<div class="doc_subsubsection"> <a name="t_vector">Vector Type</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1680
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1681<div class="doc_text">
1682
1683<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1684<p>A vector type is a simple derived type that represents a vector of elements.
1685 Vector types are used when multiple primitive data are operated in parallel
1686 using a single instruction (SIMD). A vector type requires a size (number of
1687 elements) and an underlying primitive data type. Vectors must have a power
1688 of two length (1, 2, 4, 8, 16 ...). Vector types are considered
1689 <a href="#t_firstclass">first class</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1690
1691<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1692<pre>
1693 &lt; &lt;# elements&gt; x &lt;elementtype&gt; &gt;
1694</pre>
1695
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1696<p>The number of elements is a constant integer value; elementtype may be any
1697 integer or floating point type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1698
1699<h5>Examples:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1700<table class="layout">
1701 <tr class="layout">
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1702 <td class="left"><tt>&lt;4 x i32&gt;</tt></td>
1703 <td class="left">Vector of 4 32-bit integer values.</td>
1704 </tr>
1705 <tr class="layout">
1706 <td class="left"><tt>&lt;8 x float&gt;</tt></td>
1707 <td class="left">Vector of 8 32-bit floating-point values.</td>
1708 </tr>
1709 <tr class="layout">
1710 <td class="left"><tt>&lt;2 x i64&gt;</tt></td>
1711 <td class="left">Vector of 2 64-bit integer values.</td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1712 </tr>
1713</table>
djge93155c2009-01-24 15:58:40 +0000[diff] [blame]1714
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1715<p>Note that the code generator does not yet support large vector types to be
1716 used as function return types. The specific limit on how large a vector
1717 return type codegen can currently handle is target-dependent; currently it's
1718 often a few times longer than a hardware vector register.</p>
djge93155c2009-01-24 15:58:40 +0000[diff] [blame]1719
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1720</div>
1721
1722<!-- _______________________________________________________________________ -->
1723<div class="doc_subsubsection"> <a name="t_opaque">Opaque Type</a> </div>
1724<div class="doc_text">
1725
1726<h5>Overview:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1727<p>Opaque types are used to represent unknown types in the system. This
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1728 corresponds (for example) to the C notion of a forward declared structure
1729 type. In LLVM, opaque types can eventually be resolved to any type (not just
1730 a structure type).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1731
1732<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1733<pre>
1734 opaque
1735</pre>
1736
1737<h5>Examples:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1738<table class="layout">
1739 <tr class="layout">
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1740 <td class="left"><tt>opaque</tt></td>
1741 <td class="left">An opaque type.</td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1742 </tr>
1743</table>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1744
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1745</div>
1746
Chris Lattner515195a2009-02-02 07:32:36 +0000[diff] [blame]1747<!-- ======================================================================= -->
1748<div class="doc_subsection">
1749 <a name="t_uprefs">Type Up-references</a>
1750</div>
1751
1752<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1753
Chris Lattner515195a2009-02-02 07:32:36 +0000[diff] [blame]1754<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1755<p>An "up reference" allows you to refer to a lexically enclosing type without
1756 requiring it to have a name. For instance, a structure declaration may
1757 contain a pointer to any of the types it is lexically a member of. Example
1758 of up references (with their equivalent as named type declarations)
1759 include:</p>
Chris Lattner515195a2009-02-02 07:32:36 +0000[diff] [blame]1760
1761<pre>
Chris Lattner5ad632d2009-02-09 10:00:56 +0000[diff] [blame]1762 { \2 * } %x = type { %x* }
Chris Lattner515195a2009-02-02 07:32:36 +0000[diff] [blame]1763 { \2 }* %y = type { %y }*
1764 \1* %z = type %z*
1765</pre>
1766
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1767<p>An up reference is needed by the asmprinter for printing out cyclic types
1768 when there is no declared name for a type in the cycle. Because the
1769 asmprinter does not want to print out an infinite type string, it needs a
1770 syntax to handle recursive types that have no names (all names are optional
1771 in llvm IR).</p>
Chris Lattner515195a2009-02-02 07:32:36 +0000[diff] [blame]1772
1773<h5>Syntax:</h5>
1774<pre>
1775 \&lt;level&gt;
1776</pre>
1777
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1778<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]1779
1780<h5>Examples:</h5>
Chris Lattner515195a2009-02-02 07:32:36 +0000[diff] [blame]1781<table class="layout">
1782 <tr class="layout">
1783 <td class="left"><tt>\1*</tt></td>
1784 <td class="left">Self-referential pointer.</td>
1785 </tr>
1786 <tr class="layout">
1787 <td class="left"><tt>{ { \3*, i8 }, i32 }</tt></td>
1788 <td class="left">Recursive structure where the upref refers to the out-most
1789 structure.</td>
1790 </tr>
1791</table>
Chris Lattner515195a2009-02-02 07:32:36 +0000[diff] [blame]1792
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1793</div>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1794
1795<!-- *********************************************************************** -->
1796<div class="doc_section"> <a name="constants">Constants</a> </div>
1797<!-- *********************************************************************** -->
1798
1799<div class="doc_text">
1800
1801<p>LLVM has several different basic types of constants. This section describes
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1802 them all and their syntax.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1803
1804</div>
1805
1806<!-- ======================================================================= -->
1807<div class="doc_subsection"><a name="simpleconstants">Simple Constants</a></div>
1808
1809<div class="doc_text">
1810
1811<dl>
1812 <dt><b>Boolean constants</b></dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1813 <dd>The two strings '<tt>true</tt>' and '<tt>false</tt>' are both valid
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1814 constants of the <tt><a href="#t_primitive">i1</a></tt> type.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1815
1816 <dt><b>Integer constants</b></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1817 <dd>Standard integers (such as '4') are constants of
1818 the <a href="#t_integer">integer</a> type. Negative numbers may be used
1819 with integer types.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1820
1821 <dt><b>Floating point constants</b></dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1822 <dd>Floating point constants use standard decimal notation (e.g. 123.421),
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1823 exponential notation (e.g. 1.23421e+2), or a more precise hexadecimal
1824 notation (see below). The assembler requires the exact decimal value of a
1825 floating-point constant. For example, the assembler accepts 1.25 but
1826 rejects 1.3 because 1.3 is a repeating decimal in binary. Floating point
1827 constants must have a <a href="#t_floating">floating point</a> type. </dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1828
1829 <dt><b>Null pointer constants</b></dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1830 <dd>The identifier '<tt>null</tt>' is recognized as a null pointer constant
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1831 and must be of <a href="#t_pointer">pointer type</a>.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1832</dl>
1833
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1834<p>The one non-intuitive notation for constants is the hexadecimal form of
1835 floating point constants. For example, the form '<tt>double
1836 0x432ff973cafa8000</tt>' is equivalent to (but harder to read than)
1837 '<tt>double 4.5e+15</tt>'. The only time hexadecimal floating point
1838 constants are required (and the only time that they are generated by the
1839 disassembler) is when a floating point constant must be emitted but it cannot
1840 be represented as a decimal floating point number in a reasonable number of
1841 digits. For example, NaN's, infinities, and other special values are
1842 represented in their IEEE hexadecimal format so that assembly and disassembly
1843 do not cause any bits to change in the constants.</p>
1844
Dale Johannesenf82a52f2009-02-11 22:14:51 +0000[diff] [blame]1845<p>When using the hexadecimal form, constants of types float and double are
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1846 represented using the 16-digit form shown above (which matches the IEEE754
1847 representation for double); float values must, however, be exactly
1848 representable as IEE754 single precision. Hexadecimal format is always used
1849 for long double, and there are three forms of long double. The 80-bit format
1850 used by x86 is represented as <tt>0xK</tt> followed by 20 hexadecimal digits.
1851 The 128-bit format used by PowerPC (two adjacent doubles) is represented
1852 by <tt>0xM</tt> followed by 32 hexadecimal digits. The IEEE 128-bit format
1853 is represented by <tt>0xL</tt> followed by 32 hexadecimal digits; no
1854 currently supported target uses this format. Long doubles will only work if
1855 they match the long double format on your target. All hexadecimal formats
1856 are big-endian (sign bit at the left).</p>
1857
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1858</div>
1859
1860<!-- ======================================================================= -->
Chris Lattner97063852009-02-28 18:32:25 +0000[diff] [blame]1861<div class="doc_subsection">
Bill Wendling1a2630a2009-07-20 02:32:41 +0000[diff] [blame]1862<a name="aggregateconstants"></a> <!-- old anchor -->
1863<a name="complexconstants">Complex Constants</a>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1864</div>
1865
1866<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1867
Chris Lattner97063852009-02-28 18:32:25 +0000[diff] [blame]1868<p>Complex constants are a (potentially recursive) combination of simple
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1869 constants and smaller complex constants.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1870
1871<dl>
1872 <dt><b>Structure constants</b></dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1873 <dd>Structure constants are represented with notation similar to structure
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1874 type definitions (a comma separated list of elements, surrounded by braces
1875 (<tt>{}</tt>)). For example: "<tt>{ i32 4, float 17.0, i32* @G }</tt>",
1876 where "<tt>@G</tt>" is declared as "<tt>@G = external global i32</tt>".
1877 Structure constants must have <a href="#t_struct">structure type</a>, and
1878 the number and types of elements must match those specified by the
1879 type.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1880
1881 <dt><b>Array constants</b></dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1882 <dd>Array constants are represented with notation similar to array type
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1883 definitions (a comma separated list of elements, surrounded by square
1884 brackets (<tt>[]</tt>)). For example: "<tt>[ i32 42, i32 11, i32 74
1885 ]</tt>". Array constants must have <a href="#t_array">array type</a>, and
1886 the number and types of elements must match those specified by the
1887 type.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1888
1889 <dt><b>Vector constants</b></dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1890 <dd>Vector constants are represented with notation similar to vector type
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1891 definitions (a comma separated list of elements, surrounded by
1892 less-than/greater-than's (<tt>&lt;&gt;</tt>)). For example: "<tt>&lt; i32
1893 42, i32 11, i32 74, i32 100 &gt;</tt>". Vector constants must
1894 have <a href="#t_vector">vector type</a>, and the number and types of
1895 elements must match those specified by the type.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1896
1897 <dt><b>Zero initialization</b></dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1898 <dd>The string '<tt>zeroinitializer</tt>' can be used to zero initialize a
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1899 value to zero of <em>any</em> type, including scalar and aggregate types.
1900 This is often used to avoid having to print large zero initializers
1901 (e.g. for large arrays) and is always exactly equivalent to using explicit
1902 zero initializers.</dd>
Nick Lewycky4dcf8102009-04-04 07:22:01 +0000[diff] [blame]1903
1904 <dt><b>Metadata node</b></dt>
Nick Lewyckyf122c7e2009-05-30 16:08:30 +0000[diff] [blame]1905 <dd>A metadata node is a structure-like constant with
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1906 <a href="#t_metadata">metadata type</a>. For example: "<tt>metadata !{
1907 i32 0, metadata !"test" }</tt>". Unlike other constants that are meant to
1908 be interpreted as part of the instruction stream, metadata is a place to
1909 attach additional information such as debug info.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1910</dl>
1911
1912</div>
1913
1914<!-- ======================================================================= -->
1915<div class="doc_subsection">
1916 <a name="globalconstants">Global Variable and Function Addresses</a>
1917</div>
1918
1919<div class="doc_text">
1920
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1921<p>The addresses of <a href="#globalvars">global variables</a>
1922 and <a href="#functionstructure">functions</a> are always implicitly valid
1923 (link-time) constants. These constants are explicitly referenced when
1924 the <a href="#identifiers">identifier for the global</a> is used and always
1925 have <a href="#t_pointer">pointer</a> type. For example, the following is a
1926 legal LLVM file:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1927
1928<div class="doc_code">
1929<pre>
1930@X = global i32 17
1931@Y = global i32 42
1932@Z = global [2 x i32*] [ i32* @X, i32* @Y ]
1933</pre>
1934</div>
1935
1936</div>
1937
1938<!-- ======================================================================= -->
1939<div class="doc_subsection"><a name="undefvalues">Undefined Values</a></div>
1940<div class="doc_text">
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1941
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1942<p>The string '<tt>undef</tt>' is recognized as a type-less constant that has no
1943 specific value. Undefined values may be of any type and be used anywhere a
1944 constant is permitted.</p>
1945
1946<p>Undefined values indicate to the compiler that the program is well defined no
1947 matter what value is used, giving the compiler more freedom to optimize.</p>
1948
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1949</div>
1950
1951<!-- ======================================================================= -->
1952<div class="doc_subsection"><a name="constantexprs">Constant Expressions</a>
1953</div>
1954
1955<div class="doc_text">
1956
1957<p>Constant expressions are used to allow expressions involving other constants
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1958 to be used as constants. Constant expressions may be of
1959 any <a href="#t_firstclass">first class</a> type and may involve any LLVM
1960 operation that does not have side effects (e.g. load and call are not
1961 supported). The following is the syntax for constant expressions:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1962
1963<dl>
1964 <dt><b><tt>trunc ( CST to TYPE )</tt></b></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1965 <dd>Truncate a constant to another type. The bit size of CST must be larger
1966 than the bit size of TYPE. Both types must be integers.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1967
1968 <dt><b><tt>zext ( CST to TYPE )</tt></b></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1969 <dd>Zero extend a constant to another type. The bit size of CST must be
1970 smaller or equal to the bit size of TYPE. Both types must be
1971 integers.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1972
1973 <dt><b><tt>sext ( CST to TYPE )</tt></b></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1974 <dd>Sign extend a constant to another type. The bit size of CST must be
1975 smaller or equal to the bit size of TYPE. Both types must be
1976 integers.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1977
1978 <dt><b><tt>fptrunc ( CST to TYPE )</tt></b></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1979 <dd>Truncate a floating point constant to another floating point type. The
1980 size of CST must be larger than the size of TYPE. Both types must be
1981 floating point.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1982
1983 <dt><b><tt>fpext ( CST to TYPE )</tt></b></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1984 <dd>Floating point extend a constant to another type. The size of CST must be
1985 smaller or equal to the size of TYPE. Both types must be floating
1986 point.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1987
Reid Spencere6adee82007-07-31 14:40:14 +0000[diff] [blame]1988 <dt><b><tt>fptoui ( CST to TYPE )</tt></b></dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1989 <dd>Convert a floating point constant to the corresponding unsigned integer
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1990 constant. TYPE must be a scalar or vector integer type. CST must be of
1991 scalar or vector floating point type. Both CST and TYPE must be scalars,
1992 or vectors of the same number of elements. If the value won't fit in the
1993 integer type, the results are undefined.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1994
1995 <dt><b><tt>fptosi ( CST to TYPE )</tt></b></dt>
1996 <dd>Convert a floating point constant to the corresponding signed integer
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]1997 constant. TYPE must be a scalar or vector integer type. CST must be of
1998 scalar or vector floating point type. Both CST and TYPE must be scalars,
1999 or vectors of the same number of elements. If the value won't fit in the
2000 integer type, the results are undefined.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2001
2002 <dt><b><tt>uitofp ( CST to TYPE )</tt></b></dt>
2003 <dd>Convert an unsigned integer constant to the corresponding floating point
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2004 constant. TYPE must be a scalar or vector floating point type. CST must be
2005 of scalar or vector integer type. Both CST and TYPE must be scalars, or
2006 vectors of the same number of elements. If the value won't fit in the
2007 floating point type, the results are undefined.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2008
2009 <dt><b><tt>sitofp ( CST to TYPE )</tt></b></dt>
2010 <dd>Convert a signed integer constant to the corresponding floating point
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2011 constant. TYPE must be a scalar or vector floating point type. CST must be
2012 of scalar or vector integer type. Both CST and TYPE must be scalars, or
2013 vectors of the same number of elements. If the value won't fit in the
2014 floating point type, the results are undefined.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2015
2016 <dt><b><tt>ptrtoint ( CST to TYPE )</tt></b></dt>
2017 <dd>Convert a pointer typed constant to the corresponding integer constant
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2018 <tt>TYPE</tt> must be an integer type. <tt>CST</tt> must be of pointer
2019 type. The <tt>CST</tt> value is zero extended, truncated, or unchanged to
2020 make it fit in <tt>TYPE</tt>.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2021
2022 <dt><b><tt>inttoptr ( CST to TYPE )</tt></b></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2023 <dd>Convert a integer constant to a pointer constant. TYPE must be a pointer
2024 type. CST must be of integer type. The CST value is zero extended,
2025 truncated, or unchanged to make it fit in a pointer size. This one is
2026 <i>really</i> dangerous!</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2027
2028 <dt><b><tt>bitcast ( CST to TYPE )</tt></b></dt>
Chris Lattner557bc5d2009-02-28 18:27:03 +0000[diff] [blame]2029 <dd>Convert a constant, CST, to another TYPE. The constraints of the operands
2030 are the same as those for the <a href="#i_bitcast">bitcast
2031 instruction</a>.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2032
2033 <dt><b><tt>getelementptr ( CSTPTR, IDX0, IDX1, ... )</tt></b></dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2034 <dd>Perform the <a href="#i_getelementptr">getelementptr operation</a> on
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2035 constants. As with the <a href="#i_getelementptr">getelementptr</a>
2036 instruction, the index list may have zero or more indexes, which are
2037 required to make sense for the type of "CSTPTR".</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2038
2039 <dt><b><tt>select ( COND, VAL1, VAL2 )</tt></b></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2040 <dd>Perform the <a href="#i_select">select operation</a> on constants.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2041
2042 <dt><b><tt>icmp COND ( VAL1, VAL2 )</tt></b></dt>
2043 <dd>Performs the <a href="#i_icmp">icmp operation</a> on constants.</dd>
2044
2045 <dt><b><tt>fcmp COND ( VAL1, VAL2 )</tt></b></dt>
2046 <dd>Performs the <a href="#i_fcmp">fcmp operation</a> on constants.</dd>
2047
2048 <dt><b><tt>extractelement ( VAL, IDX )</tt></b></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2049 <dd>Perform the <a href="#i_extractelement">extractelement operation</a> on
2050 constants.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2051
2052 <dt><b><tt>insertelement ( VAL, ELT, IDX )</tt></b></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2053 <dd>Perform the <a href="#i_insertelement">insertelement operation</a> on
2054 constants.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2055
2056 <dt><b><tt>shufflevector ( VEC1, VEC2, IDXMASK )</tt></b></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2057 <dd>Perform the <a href="#i_shufflevector">shufflevector operation</a> on
2058 constants.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2059
2060 <dt><b><tt>OPCODE ( LHS, RHS )</tt></b></dt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2061 <dd>Perform the specified operation of the LHS and RHS constants. OPCODE may
2062 be any of the <a href="#binaryops">binary</a>
2063 or <a href="#bitwiseops">bitwise binary</a> operations. The constraints
2064 on operands are the same as those for the corresponding instruction
2065 (e.g. no bitwise operations on floating point values are allowed).</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2066</dl>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2067
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2068</div>
2069
Nick Lewycky4dcf8102009-04-04 07:22:01 +0000[diff] [blame]2070<!-- ======================================================================= -->
2071<div class="doc_subsection"><a name="metadata">Embedded Metadata</a>
2072</div>
2073
2074<div class="doc_text">
2075
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2076<p>Embedded metadata provides a way to attach arbitrary data to the instruction
2077 stream without affecting the behaviour of the program. There are two
2078 metadata primitives, strings and nodes. All metadata has the
2079 <tt>metadata</tt> type and is identified in syntax by a preceding exclamation
2080 point ('<tt>!</tt>').</p>
Nick Lewycky4dcf8102009-04-04 07:22:01 +0000[diff] [blame]2081
2082<p>A metadata string is a string surrounded by double quotes. It can contain
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2083 any character by escaping non-printable characters with "\xx" where "xx" is
2084 the two digit hex code. For example: "<tt>!"test\00"</tt>".</p>
Nick Lewycky4dcf8102009-04-04 07:22:01 +0000[diff] [blame]2085
2086<p>Metadata nodes are represented with notation similar to structure constants
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2087 (a comma separated list of elements, surrounded by braces and preceeded by an
2088 exclamation point). For example: "<tt>!{ metadata !"test\00", i32
2089 10}</tt>".</p>
Nick Lewycky4dcf8102009-04-04 07:22:01 +0000[diff] [blame]2090
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2091<p>A metadata node will attempt to track changes to the values it holds. In the
2092 event that a value is deleted, it will be replaced with a typeless
2093 "<tt>null</tt>", such as "<tt>metadata !{null, i32 10}</tt>".</p>
Nick Lewycky117f4382009-05-10 20:57:05 +0000[diff] [blame]2094
Nick Lewycky4dcf8102009-04-04 07:22:01 +0000[diff] [blame]2095<p>Optimizations may rely on metadata to provide additional information about
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2096 the program that isn't available in the instructions, or that isn't easily
2097 computable. Similarly, the code generator may expect a certain metadata
2098 format to be used to express debugging information.</p>
2099
Nick Lewycky4dcf8102009-04-04 07:22:01 +0000[diff] [blame]2100</div>
2101
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2102<!-- *********************************************************************** -->
2103<div class="doc_section"> <a name="othervalues">Other Values</a> </div>
2104<!-- *********************************************************************** -->
2105
2106<!-- ======================================================================= -->
2107<div class="doc_subsection">
2108<a name="inlineasm">Inline Assembler Expressions</a>
2109</div>
2110
2111<div class="doc_text">
2112
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2113<p>LLVM supports inline assembler expressions (as opposed
2114 to <a href="#moduleasm"> Module-Level Inline Assembly</a>) through the use of
2115 a special value. This value represents the inline assembler as a string
2116 (containing the instructions to emit), a list of operand constraints (stored
2117 as a string), and a flag that indicates whether or not the inline asm
2118 expression has side effects. An example inline assembler expression is:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2119
2120<div class="doc_code">
2121<pre>
2122i32 (i32) asm "bswap $0", "=r,r"
2123</pre>
2124</div>
2125
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2126<p>Inline assembler expressions may <b>only</b> be used as the callee operand of
2127 a <a href="#i_call"><tt>call</tt> instruction</a>. Thus, typically we
2128 have:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2129
2130<div class="doc_code">
2131<pre>
2132%X = call i32 asm "<a href="#int_bswap">bswap</a> $0", "=r,r"(i32 %Y)
2133</pre>
2134</div>
2135
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2136<p>Inline asms with side effects not visible in the constraint list must be
2137 marked as having side effects. This is done through the use of the
2138 '<tt>sideeffect</tt>' keyword, like so:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2139
2140<div class="doc_code">
2141<pre>
2142call void asm sideeffect "eieio", ""()
2143</pre>
2144</div>
2145
2146<p>TODO: The format of the asm and constraints string still need to be
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2147 documented here. Constraints on what can be done (e.g. duplication, moving,
2148 etc need to be documented). This is probably best done by reference to
2149 another document that covers inline asm from a holistic perspective.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2150
2151</div>
2152
2153<!-- *********************************************************************** -->
2154<div class="doc_section"> <a name="instref">Instruction Reference</a> </div>
2155<!-- *********************************************************************** -->
2156
2157<div class="doc_text">
2158
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2159<p>The LLVM instruction set consists of several different classifications of
2160 instructions: <a href="#terminators">terminator
2161 instructions</a>, <a href="#binaryops">binary instructions</a>,
2162 <a href="#bitwiseops">bitwise binary instructions</a>,
2163 <a href="#memoryops">memory instructions</a>, and
2164 <a href="#otherops">other instructions</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2165
2166</div>
2167
2168<!-- ======================================================================= -->
2169<div class="doc_subsection"> <a name="terminators">Terminator
2170Instructions</a> </div>
2171
2172<div class="doc_text">
2173
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2174<p>As mentioned <a href="#functionstructure">previously</a>, every basic block
2175 in a program ends with a "Terminator" instruction, which indicates which
2176 block should be executed after the current block is finished. These
2177 terminator instructions typically yield a '<tt>void</tt>' value: they produce
2178 control flow, not values (the one exception being the
2179 '<a href="#i_invoke"><tt>invoke</tt></a>' instruction).</p>
2180
2181<p>There are six different terminator instructions: the
2182 '<a href="#i_ret"><tt>ret</tt></a>' instruction, the
2183 '<a href="#i_br"><tt>br</tt></a>' instruction, the
2184 '<a href="#i_switch"><tt>switch</tt></a>' instruction, the
2185 '<a href="#i_invoke"><tt>invoke</tt></a>' instruction, the
2186 '<a href="#i_unwind"><tt>unwind</tt></a>' instruction, and the
2187 '<a href="#i_unreachable"><tt>unreachable</tt></a>' instruction.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2188
2189</div>
2190
2191<!-- _______________________________________________________________________ -->
2192<div class="doc_subsubsection"> <a name="i_ret">'<tt>ret</tt>'
2193Instruction</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2194
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2195<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2196
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2197<h5>Syntax:</h5>
Dan Gohman3e700032008-10-04 19:00:07 +0000[diff] [blame]2198<pre>
2199 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]2200 ret void <i>; Return from void function</i>
2201</pre>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]2202
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2203<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2204<p>The '<tt>ret</tt>' instruction is used to return control flow (and optionally
2205 a value) from a function back to the caller.</p>
2206
2207<p>There are two forms of the '<tt>ret</tt>' instruction: one that returns a
2208 value and then causes control flow, and one that just causes control flow to
2209 occur.</p>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]2210
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2211<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2212<p>The '<tt>ret</tt>' instruction optionally accepts a single argument, the
2213 return value. The type of the return value must be a
2214 '<a href="#t_firstclass">first class</a>' type.</p>
Dan Gohman3e700032008-10-04 19:00:07 +0000[diff] [blame]2215
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2216<p>A function is not <a href="#wellformed">well formed</a> if it it has a
2217 non-void return type and contains a '<tt>ret</tt>' instruction with no return
2218 value or a return value with a type that does not match its type, or if it
2219 has a void return type and contains a '<tt>ret</tt>' instruction with a
2220 return value.</p>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]2221
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2222<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2223<p>When the '<tt>ret</tt>' instruction is executed, control flow returns back to
2224 the calling function's context. If the caller is a
2225 "<a href="#i_call"><tt>call</tt></a>" instruction, execution continues at the
2226 instruction after the call. If the caller was an
2227 "<a href="#i_invoke"><tt>invoke</tt></a>" instruction, execution continues at
2228 the beginning of the "normal" destination block. If the instruction returns
2229 a value, that value shall set the call or invoke instruction's return
2230 value.</p>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]2231
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2232<h5>Example:</h5>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]2233<pre>
2234 ret i32 5 <i>; Return an integer value of 5</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2235 ret void <i>; Return from a void function</i>
Bill Wendlingd163e2d2009-02-28 22:12:54 +0000[diff] [blame]2236 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]2237</pre>
Dan Gohman60967192009-01-12 23:12:39 +0000[diff] [blame]2238
djge93155c2009-01-24 15:58:40 +0000[diff] [blame]2239<p>Note that the code generator does not yet fully support large
2240 return values. The specific sizes that are currently supported are
2241 dependent on the target. For integers, on 32-bit targets the limit
2242 is often 64 bits, and on 64-bit targets the limit is often 128 bits.
2243 For aggregate types, the current limits are dependent on the element
2244 types; for example targets are often limited to 2 total integer
2245 elements and 2 total floating-point elements.</p>
Dan Gohman60967192009-01-12 23:12:39 +0000[diff] [blame]2246
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2247</div>
2248<!-- _______________________________________________________________________ -->
2249<div class="doc_subsubsection"> <a name="i_br">'<tt>br</tt>' Instruction</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2250
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2251<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2252
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2253<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2254<pre>
2255 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]2256</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2257
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2258<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2259<p>The '<tt>br</tt>' instruction is used to cause control flow to transfer to a
2260 different basic block in the current function. There are two forms of this
2261 instruction, corresponding to a conditional branch and an unconditional
2262 branch.</p>
2263
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2264<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2265<p>The conditional branch form of the '<tt>br</tt>' instruction takes a single
2266 '<tt>i1</tt>' value and two '<tt>label</tt>' values. The unconditional form
2267 of the '<tt>br</tt>' instruction takes a single '<tt>label</tt>' value as a
2268 target.</p>
2269
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2270<h5>Semantics:</h5>
2271<p>Upon execution of a conditional '<tt>br</tt>' instruction, the '<tt>i1</tt>'
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2272 argument is evaluated. If the value is <tt>true</tt>, control flows to the
2273 '<tt>iftrue</tt>' <tt>label</tt> argument. If "cond" is <tt>false</tt>,
2274 control flows to the '<tt>iffalse</tt>' <tt>label</tt> argument.</p>
2275
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2276<h5>Example:</h5>
Bill Wendling6ec40612009-07-20 02:39:26 +0000[diff] [blame]2277<pre>
2278Test:
2279 %cond = <a href="#i_icmp">icmp</a> eq i32 %a, %b
2280 br i1 %cond, label %IfEqual, label %IfUnequal
2281IfEqual:
2282 <a href="#i_ret">ret</a> i32 1
2283IfUnequal:
2284 <a href="#i_ret">ret</a> i32 0
2285</pre>
2286
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2287</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2288
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2289<!-- _______________________________________________________________________ -->
2290<div class="doc_subsubsection">
2291 <a name="i_switch">'<tt>switch</tt>' Instruction</a>
2292</div>
2293
2294<div class="doc_text">
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2295
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2296<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2297<pre>
2298 switch &lt;intty&gt; &lt;value&gt;, label &lt;defaultdest&gt; [ &lt;intty&gt; &lt;val&gt;, label &lt;dest&gt; ... ]
2299</pre>
2300
2301<h5>Overview:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2302<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]2303 several different places. It is a generalization of the '<tt>br</tt>'
2304 instruction, allowing a branch to occur to one of many possible
2305 destinations.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2306
2307<h5>Arguments:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2308<p>The '<tt>switch</tt>' instruction uses three parameters: an integer
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2309 comparison value '<tt>value</tt>', a default '<tt>label</tt>' destination,
2310 and an array of pairs of comparison value constants and '<tt>label</tt>'s.
2311 The table is not allowed to contain duplicate constant entries.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2312
2313<h5>Semantics:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2314<p>The <tt>switch</tt> instruction specifies a table of values and
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2315 destinations. When the '<tt>switch</tt>' instruction is executed, this table
2316 is searched for the given value. If the value is found, control flow is
2317 transfered to the corresponding destination; otherwise, control flow is
2318 transfered to the default destination.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2319
2320<h5>Implementation:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2321<p>Depending on properties of the target machine and the particular
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2322 <tt>switch</tt> instruction, this instruction may be code generated in
2323 different ways. For example, it could be generated as a series of chained
2324 conditional branches or with a lookup table.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2325
2326<h5>Example:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2327<pre>
2328 <i>; Emulate a conditional br instruction</i>
2329 %Val = <a href="#i_zext">zext</a> i1 %value to i32
Dan Gohman01852382009-01-04 23:44:43 +0000[diff] [blame]2330 switch i32 %Val, label %truedest [ i32 0, label %falsedest ]
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2331
2332 <i>; Emulate an unconditional br instruction</i>
2333 switch i32 0, label %dest [ ]
2334
2335 <i>; Implement a jump table:</i>
Dan Gohman01852382009-01-04 23:44:43 +0000[diff] [blame]2336 switch i32 %val, label %otherwise [ i32 0, label %onzero
2337 i32 1, label %onone
2338 i32 2, label %ontwo ]
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2339</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2340
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2341</div>
2342
2343<!-- _______________________________________________________________________ -->
2344<div class="doc_subsubsection">
2345 <a name="i_invoke">'<tt>invoke</tt>' Instruction</a>
2346</div>
2347
2348<div class="doc_text">
2349
2350<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2351<pre>
Devang Pateld0bfcc72008-10-07 17:48:33 +0000[diff] [blame]2352 &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]2353 to label &lt;normal label&gt; unwind label &lt;exception label&gt;
2354</pre>
2355
2356<h5>Overview:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2357<p>The '<tt>invoke</tt>' instruction causes control to transfer to a specified
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2358 function, with the possibility of control flow transfer to either the
2359 '<tt>normal</tt>' label or the '<tt>exception</tt>' label. If the callee
2360 function returns with the "<tt><a href="#i_ret">ret</a></tt>" instruction,
2361 control flow will return to the "normal" label. If the callee (or any
2362 indirect callees) returns with the "<a href="#i_unwind"><tt>unwind</tt></a>"
2363 instruction, control is interrupted and continued at the dynamically nearest
2364 "exception" label.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2365
2366<h5>Arguments:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2367<p>This instruction requires several arguments:</p>
2368
2369<ol>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2370 <li>The optional "cconv" marker indicates which <a href="#callingconv">calling
2371 convention</a> the call should use. If none is specified, the call
2372 defaults to using C calling conventions.</li>
Devang Patelac2fc272008-10-06 18:50:38 +0000[diff] [blame]2373
2374 <li>The optional <a href="#paramattrs">Parameter Attributes</a> list for
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2375 return values. Only '<tt>zeroext</tt>', '<tt>signext</tt>', and
2376 '<tt>inreg</tt>' attributes are valid here.</li>
Devang Patelac2fc272008-10-06 18:50:38 +0000[diff] [blame]2377
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2378 <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]2379 function value being invoked. In most cases, this is a direct function
2380 invocation, but indirect <tt>invoke</tt>s are just as possible, branching
2381 off an arbitrary pointer to function value.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2382
2383 <li>'<tt>function ptr val</tt>': An LLVM value containing a pointer to a
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2384 function to be invoked. </li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2385
2386 <li>'<tt>function args</tt>': argument list whose types match the function
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2387 signature argument types. If the function signature indicates the
2388 function accepts a variable number of arguments, the extra arguments can
2389 be specified.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2390
2391 <li>'<tt>normal label</tt>': the label reached when the called function
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2392 executes a '<tt><a href="#i_ret">ret</a></tt>' instruction. </li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2393
2394 <li>'<tt>exception label</tt>': the label reached when a callee returns with
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2395 the <a href="#i_unwind"><tt>unwind</tt></a> instruction. </li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2396
Devang Pateld0bfcc72008-10-07 17:48:33 +0000[diff] [blame]2397 <li>The optional <a href="#fnattrs">function attributes</a> list. Only
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2398 '<tt>noreturn</tt>', '<tt>nounwind</tt>', '<tt>readonly</tt>' and
2399 '<tt>readnone</tt>' attributes are valid here.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2400</ol>
2401
2402<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2403<p>This instruction is designed to operate as a standard
2404 '<tt><a href="#i_call">call</a></tt>' instruction in most regards. The
2405 primary difference is that it establishes an association with a label, which
2406 is used by the runtime library to unwind the stack.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2407
2408<p>This instruction is used in languages with destructors to ensure that proper
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2409 cleanup is performed in the case of either a <tt>longjmp</tt> or a thrown
2410 exception. Additionally, this is important for implementation of
2411 '<tt>catch</tt>' clauses in high-level languages that support them.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2412
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2413<p>For the purposes of the SSA form, the definition of the value returned by the
2414 '<tt>invoke</tt>' instruction is deemed to occur on the edge from the current
2415 block to the "normal" label. If the callee unwinds then no return value is
2416 available.</p>
Dan Gohman140ba5d2009-05-22 21:47:08 +0000[diff] [blame]2417
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2418<h5>Example:</h5>
2419<pre>
Nick Lewyckya1c11a12008-03-16 07:18:12 +0000[diff] [blame]2420 %retval = invoke i32 @Test(i32 15) to label %Continue
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2421 unwind label %TestCleanup <i>; {i32}:retval set</i>
Nick Lewyckya1c11a12008-03-16 07:18:12 +0000[diff] [blame]2422 %retval = invoke <a href="#callingconv">coldcc</a> i32 %Testfnptr(i32 15) to label %Continue
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2423 unwind label %TestCleanup <i>; {i32}:retval set</i>
2424</pre>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2425
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2426</div>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2427
2428<!-- _______________________________________________________________________ -->
2429
2430<div class="doc_subsubsection"> <a name="i_unwind">'<tt>unwind</tt>'
2431Instruction</a> </div>
2432
2433<div class="doc_text">
2434
2435<h5>Syntax:</h5>
2436<pre>
2437 unwind
2438</pre>
2439
2440<h5>Overview:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2441<p>The '<tt>unwind</tt>' instruction unwinds the stack, continuing control flow
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2442 at the first callee in the dynamic call stack which used
2443 an <a href="#i_invoke"><tt>invoke</tt></a> instruction to perform the call.
2444 This is primarily used to implement exception handling.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2445
2446<h5>Semantics:</h5>
Chris Lattner8b094fc2008-04-19 21:01:16 +0000[diff] [blame]2447<p>The '<tt>unwind</tt>' instruction causes execution of the current function to
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2448 immediately halt. The dynamic call stack is then searched for the
2449 first <a href="#i_invoke"><tt>invoke</tt></a> instruction on the call stack.
2450 Once found, execution continues at the "exceptional" destination block
2451 specified by the <tt>invoke</tt> instruction. If there is no <tt>invoke</tt>
2452 instruction in the dynamic call chain, undefined behavior results.</p>
2453
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2454</div>
2455
2456<!-- _______________________________________________________________________ -->
2457
2458<div class="doc_subsubsection"> <a name="i_unreachable">'<tt>unreachable</tt>'
2459Instruction</a> </div>
2460
2461<div class="doc_text">
2462
2463<h5>Syntax:</h5>
2464<pre>
2465 unreachable
2466</pre>
2467
2468<h5>Overview:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2469<p>The '<tt>unreachable</tt>' instruction has no defined semantics. This
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2470 instruction is used to inform the optimizer that a particular portion of the
2471 code is not reachable. This can be used to indicate that the code after a
2472 no-return function cannot be reached, and other facts.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2473
2474<h5>Semantics:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2475<p>The '<tt>unreachable</tt>' instruction has no defined semantics.</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2476
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2477</div>
2478
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2479<!-- ======================================================================= -->
2480<div class="doc_subsection"> <a name="binaryops">Binary Operations</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2481
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2482<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2483
2484<p>Binary operators are used to do most of the computation in a program. They
2485 require two operands of the same type, execute an operation on them, and
2486 produce a single value. The operands might represent multiple data, as is
2487 the case with the <a href="#t_vector">vector</a> data type. The result value
2488 has the same type as its operands.</p>
2489
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2490<p>There are several different binary operators:</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2491
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2492</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2493
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2494<!-- _______________________________________________________________________ -->
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2495<div class="doc_subsubsection">
2496 <a name="i_add">'<tt>add</tt>' Instruction</a>
2497</div>
2498
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2499<div class="doc_text">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2500
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2501<h5>Syntax:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2502<pre>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]2503 &lt;result&gt; = add &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2504</pre>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2505
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2506<h5>Overview:</h5>
2507<p>The '<tt>add</tt>' instruction returns the sum of its two operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2508
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2509<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2510<p>The two arguments to the '<tt>add</tt>' instruction must
2511 be <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of
2512 integer values. Both arguments must have identical types.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2513
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2514<h5>Semantics:</h5>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]2515<p>The value produced is the integer sum of the two operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2516
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2517<p>If the sum has unsigned overflow, the result returned is the mathematical
2518 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]2519
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2520<p>Because LLVM integers use a two's complement representation, this instruction
2521 is appropriate for both signed and unsigned integers.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2522
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2523<h5>Example:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2524<pre>
2525 &lt;result&gt; = add i32 4, %var <i>; yields {i32}:result = 4 + %var</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2526</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2527
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2528</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2529
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2530<!-- _______________________________________________________________________ -->
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2531<div class="doc_subsubsection">
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]2532 <a name="i_fadd">'<tt>fadd</tt>' Instruction</a>
2533</div>
2534
2535<div class="doc_text">
2536
2537<h5>Syntax:</h5>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]2538<pre>
2539 &lt;result&gt; = fadd &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
2540</pre>
2541
2542<h5>Overview:</h5>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]2543<p>The '<tt>fadd</tt>' instruction returns the sum of its two operands.</p>
2544
2545<h5>Arguments:</h5>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]2546<p>The two arguments to the '<tt>fadd</tt>' instruction must be
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2547 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
2548 floating point values. Both arguments must have identical types.</p>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]2549
2550<h5>Semantics:</h5>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]2551<p>The value produced is the floating point sum of the two operands.</p>
2552
2553<h5>Example:</h5>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]2554<pre>
2555 &lt;result&gt; = fadd float 4.0, %var <i>; yields {float}:result = 4.0 + %var</i>
2556</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2557
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]2558</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2559
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]2560<!-- _______________________________________________________________________ -->
2561<div class="doc_subsubsection">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2562 <a name="i_sub">'<tt>sub</tt>' Instruction</a>
2563</div>
2564
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2565<div class="doc_text">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2566
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2567<h5>Syntax:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2568<pre>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]2569 &lt;result&gt; = sub &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2570</pre>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2571
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2572<h5>Overview:</h5>
2573<p>The '<tt>sub</tt>' instruction returns the difference of its two
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2574 operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2575
2576<p>Note that the '<tt>sub</tt>' instruction is used to represent the
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2577 '<tt>neg</tt>' instruction present in most other intermediate
2578 representations.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2579
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2580<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2581<p>The two arguments to the '<tt>sub</tt>' instruction must
2582 be <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of
2583 integer values. Both arguments must have identical types.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2584
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2585<h5>Semantics:</h5>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]2586<p>The value produced is the integer difference of the two operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2587
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]2588<p>If the difference has unsigned overflow, the result returned is the
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2589 mathematical result modulo 2<sup>n</sup>, where n is the bit width of the
2590 result.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2591
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2592<p>Because LLVM integers use a two's complement representation, this instruction
2593 is appropriate for both signed and unsigned integers.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2594
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2595<h5>Example:</h5>
2596<pre>
2597 &lt;result&gt; = sub i32 4, %var <i>; yields {i32}:result = 4 - %var</i>
2598 &lt;result&gt; = sub i32 0, %val <i>; yields {i32}:result = -%var</i>
2599</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2600
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2601</div>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2602
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2603<!-- _______________________________________________________________________ -->
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2604<div class="doc_subsubsection">
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]2605 <a name="i_fsub">'<tt>fsub</tt>' Instruction</a>
2606</div>
2607
2608<div class="doc_text">
2609
2610<h5>Syntax:</h5>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]2611<pre>
2612 &lt;result&gt; = fsub &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
2613</pre>
2614
2615<h5>Overview:</h5>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]2616<p>The '<tt>fsub</tt>' instruction returns the difference of its two
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2617 operands.</p>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]2618
2619<p>Note that the '<tt>fsub</tt>' instruction is used to represent the
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2620 '<tt>fneg</tt>' instruction present in most other intermediate
2621 representations.</p>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]2622
2623<h5>Arguments:</h5>
Bill Wendling1a2630a2009-07-20 02:32:41 +0000[diff] [blame]2624<p>The two arguments to the '<tt>fsub</tt>' instruction must be
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2625 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
2626 floating point values. Both arguments must have identical types.</p>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]2627
2628<h5>Semantics:</h5>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]2629<p>The value produced is the floating point difference of the two operands.</p>
2630
2631<h5>Example:</h5>
2632<pre>
2633 &lt;result&gt; = fsub float 4.0, %var <i>; yields {float}:result = 4.0 - %var</i>
2634 &lt;result&gt; = fsub float -0.0, %val <i>; yields {float}:result = -%var</i>
2635</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2636
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]2637</div>
2638
2639<!-- _______________________________________________________________________ -->
2640<div class="doc_subsubsection">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2641 <a name="i_mul">'<tt>mul</tt>' Instruction</a>
2642</div>
2643
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2644<div class="doc_text">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2645
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2646<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2647<pre>
2648 &lt;result&gt; = mul &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2649</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2650
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2651<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2652<p>The '<tt>mul</tt>' instruction returns the product of its two operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2653
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2654<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2655<p>The two arguments to the '<tt>mul</tt>' instruction must
2656 be <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of
2657 integer values. Both arguments must have identical types.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2658
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2659<h5>Semantics:</h5>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]2660<p>The value produced is the integer product of the two operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2661
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2662<p>If the result of the multiplication has unsigned overflow, the result
2663 returned is the mathematical result modulo 2<sup>n</sup>, where n is the bit
2664 width of the result.</p>
2665
2666<p>Because LLVM integers use a two's complement representation, and the result
2667 is the same width as the operands, this instruction returns the correct
2668 result for both signed and unsigned integers. If a full product
2669 (e.g. <tt>i32</tt>x<tt>i32</tt>-><tt>i64</tt>) is needed, the operands should
2670 be sign-extended or zero-extended as appropriate to the width of the full
2671 product.</p>
2672
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2673<h5>Example:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2674<pre>
2675 &lt;result&gt; = mul i32 4, %var <i>; yields {i32}:result = 4 * %var</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2676</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2677
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2678</div>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2679
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2680<!-- _______________________________________________________________________ -->
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]2681<div class="doc_subsubsection">
2682 <a name="i_fmul">'<tt>fmul</tt>' Instruction</a>
2683</div>
2684
2685<div class="doc_text">
2686
2687<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2688<pre>
2689 &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]2690</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2691
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]2692<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2693<p>The '<tt>fmul</tt>' instruction returns the product of its two operands.</p>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]2694
2695<h5>Arguments:</h5>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]2696<p>The two arguments to the '<tt>fmul</tt>' instruction must be
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2697 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
2698 floating point values. Both arguments must have identical types.</p>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]2699
2700<h5>Semantics:</h5>
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]2701<p>The value produced is the floating point product of the two operands.</p>
2702
2703<h5>Example:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2704<pre>
2705 &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]2706</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2707
Dan Gohman7ce405e2009-06-04 22:49:04 +0000[diff] [blame]2708</div>
2709
2710<!-- _______________________________________________________________________ -->
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2711<div class="doc_subsubsection"> <a name="i_udiv">'<tt>udiv</tt>' Instruction
2712</a></div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2713
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2714<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2715
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2716<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2717<pre>
2718 &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]2719</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2720
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2721<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2722<p>The '<tt>udiv</tt>' instruction returns the quotient of its two operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2723
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2724<h5>Arguments:</h5>
2725<p>The two arguments to the '<tt>udiv</tt>' instruction must be
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2726 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2727 values. Both arguments must have identical types.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2728
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2729<h5>Semantics:</h5>
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]2730<p>The value produced is the unsigned integer quotient of the two operands.</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2731
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]2732<p>Note that unsigned integer division and signed integer division are distinct
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2733 operations; for signed integer division, use '<tt>sdiv</tt>'.</p>
2734
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]2735<p>Division by zero leads to undefined behavior.</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2736
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2737<h5>Example:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2738<pre>
2739 &lt;result&gt; = udiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2740</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2741
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2742</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2743
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2744<!-- _______________________________________________________________________ -->
2745<div class="doc_subsubsection"> <a name="i_sdiv">'<tt>sdiv</tt>' Instruction
2746</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2747
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2748<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2749
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2750<h5>Syntax:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2751<pre>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]2752 &lt;result&gt; = sdiv &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2753</pre>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2754
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2755<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2756<p>The '<tt>sdiv</tt>' instruction returns the quotient of its two operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2757
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2758<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2759<p>The two arguments to the '<tt>sdiv</tt>' instruction must be
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2760 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2761 values. Both arguments must have identical types.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2762
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2763<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2764<p>The value produced is the signed integer quotient of the two operands rounded
2765 towards zero.</p>
2766
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]2767<p>Note that signed integer division and unsigned integer division are distinct
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2768 operations; for unsigned integer division, use '<tt>udiv</tt>'.</p>
2769
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]2770<p>Division by zero leads to undefined behavior. Overflow also leads to
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2771 undefined behavior; this is a rare case, but can occur, for example, by doing
2772 a 32-bit division of -2147483648 by -1.</p>
2773
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2774<h5>Example:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2775<pre>
2776 &lt;result&gt; = sdiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2777</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2778
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2779</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2780
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2781<!-- _______________________________________________________________________ -->
2782<div class="doc_subsubsection"> <a name="i_fdiv">'<tt>fdiv</tt>'
2783Instruction</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2784
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2785<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2786
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2787<h5>Syntax:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2788<pre>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]2789 &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]2790</pre>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2791
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2792<h5>Overview:</h5>
2793<p>The '<tt>fdiv</tt>' instruction returns the quotient of its two operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2794
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2795<h5>Arguments:</h5>
2796<p>The two arguments to the '<tt>fdiv</tt>' instruction must be
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2797 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
2798 floating point values. Both arguments must have identical types.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2799
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2800<h5>Semantics:</h5>
2801<p>The value produced is the floating point quotient of the two operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2802
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2803<h5>Example:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2804<pre>
2805 &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]2806</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2807
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2808</div>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2809
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2810<!-- _______________________________________________________________________ -->
2811<div class="doc_subsubsection"> <a name="i_urem">'<tt>urem</tt>' Instruction</a>
2812</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2813
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2814<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2815
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2816<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2817<pre>
2818 &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]2819</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2820
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2821<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2822<p>The '<tt>urem</tt>' instruction returns the remainder from the unsigned
2823 division of its two arguments.</p>
2824
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2825<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2826<p>The two arguments to the '<tt>urem</tt>' instruction must be
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2827 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2828 values. Both arguments must have identical types.</p>
2829
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2830<h5>Semantics:</h5>
2831<p>This instruction returns the unsigned integer <i>remainder</i> of a division.
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2832 This instruction always performs an unsigned division to get the
2833 remainder.</p>
2834
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]2835<p>Note that unsigned integer remainder and signed integer remainder are
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2836 distinct operations; for signed integer remainder, use '<tt>srem</tt>'.</p>
2837
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]2838<p>Taking the remainder of a division by zero leads to undefined behavior.</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2839
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2840<h5>Example:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2841<pre>
2842 &lt;result&gt; = urem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2843</pre>
2844
2845</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2846
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2847<!-- _______________________________________________________________________ -->
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2848<div class="doc_subsubsection">
2849 <a name="i_srem">'<tt>srem</tt>' Instruction</a>
2850</div>
2851
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2852<div class="doc_text">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2853
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2854<h5>Syntax:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2855<pre>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]2856 &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]2857</pre>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2858
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2859<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2860<p>The '<tt>srem</tt>' instruction returns the remainder from the signed
2861 division of its two operands. This instruction can also take
2862 <a href="#t_vector">vector</a> versions of the values in which case the
2863 elements must be integers.</p>
Chris Lattner08497ce2008-01-04 04:33:49 +0000[diff] [blame]2864
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2865<h5>Arguments:</h5>
2866<p>The two arguments to the '<tt>srem</tt>' instruction must be
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2867 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2868 values. Both arguments must have identical types.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2869
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2870<h5>Semantics:</h5>
2871<p>This instruction returns the <i>remainder</i> of a division (where the result
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2872 has the same sign as the dividend, <tt>op1</tt>), not the <i>modulo</i>
2873 operator (where the result has the same sign as the divisor, <tt>op2</tt>) of
2874 a value. For more information about the difference,
2875 see <a href="http://mathforum.org/dr.math/problems/anne.4.28.99.html">The
2876 Math Forum</a>. For a table of how this is implemented in various languages,
2877 please see <a href="http://en.wikipedia.org/wiki/Modulo_operation">
2878 Wikipedia: modulo operation</a>.</p>
2879
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]2880<p>Note that signed integer remainder and unsigned integer remainder are
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2881 distinct operations; for unsigned integer remainder, use '<tt>urem</tt>'.</p>
2882
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]2883<p>Taking the remainder of a division by zero leads to undefined behavior.
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2884 Overflow also leads to undefined behavior; this is a rare case, but can
2885 occur, for example, by taking the remainder of a 32-bit division of
2886 -2147483648 by -1. (The remainder doesn't actually overflow, but this rule
2887 lets srem be implemented using instructions that return both the result of
2888 the division and the remainder.)</p>
2889
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2890<h5>Example:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2891<pre>
2892 &lt;result&gt; = srem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2893</pre>
2894
2895</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2896
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2897<!-- _______________________________________________________________________ -->
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2898<div class="doc_subsubsection">
2899 <a name="i_frem">'<tt>frem</tt>' Instruction</a> </div>
2900
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2901<div class="doc_text">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2902
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2903<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2904<pre>
2905 &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]2906</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2907
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2908<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2909<p>The '<tt>frem</tt>' instruction returns the remainder from the division of
2910 its two operands.</p>
2911
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2912<h5>Arguments:</h5>
2913<p>The two arguments to the '<tt>frem</tt>' instruction must be
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2914 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a> of
2915 floating point values. Both arguments must have identical types.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2916
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2917<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2918<p>This instruction returns the <i>remainder</i> of a division. The remainder
2919 has the same sign as the dividend.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2920
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2921<h5>Example:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2922<pre>
2923 &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]2924</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2925
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2926</div>
2927
2928<!-- ======================================================================= -->
2929<div class="doc_subsection"> <a name="bitwiseops">Bitwise Binary
2930Operations</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2931
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2932<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2933
2934<p>Bitwise binary operators are used to do various forms of bit-twiddling in a
2935 program. They are generally very efficient instructions and can commonly be
2936 strength reduced from other instructions. They require two operands of the
2937 same type, execute an operation on them, and produce a single value. The
2938 resulting value is the same type as its operands.</p>
2939
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2940</div>
2941
2942<!-- _______________________________________________________________________ -->
2943<div class="doc_subsubsection"> <a name="i_shl">'<tt>shl</tt>'
2944Instruction</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2945
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2946<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2947
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2948<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2949<pre>
2950 &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]2951</pre>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2952
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2953<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2954<p>The '<tt>shl</tt>' instruction returns the first operand shifted to the left
2955 a specified number of bits.</p>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2956
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2957<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2958<p>Both arguments to the '<tt>shl</tt>' instruction must be the
2959 same <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of
2960 integer type. '<tt>op2</tt>' is treated as an unsigned value.</p>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2961
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2962<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2963<p>The value produced is <tt>op1</tt> * 2<sup><tt>op2</tt></sup> mod
2964 2<sup>n</sup>, where <tt>n</tt> is the width of the result. If <tt>op2</tt>
2965 is (statically or dynamically) negative or equal to or larger than the number
2966 of bits in <tt>op1</tt>, the result is undefined. If the arguments are
2967 vectors, each vector element of <tt>op1</tt> is shifted by the corresponding
2968 shift amount in <tt>op2</tt>.</p>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2969
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2970<h5>Example:</h5>
2971<pre>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2972 &lt;result&gt; = shl i32 4, %var <i>; yields {i32}: 4 &lt;&lt; %var</i>
2973 &lt;result&gt; = shl i32 4, 2 <i>; yields {i32}: 16</i>
2974 &lt;result&gt; = shl i32 1, 10 <i>; yields {i32}: 1024</i>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2975 &lt;result&gt; = shl i32 1, 32 <i>; undefined</i>
Mon P Wang9901e732008-12-09 05:46:39 +0000[diff] [blame]2976 &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]2977</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2978
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2979</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2980
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2981<!-- _______________________________________________________________________ -->
2982<div class="doc_subsubsection"> <a name="i_lshr">'<tt>lshr</tt>'
2983Instruction</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2984
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2985<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2986
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2987<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2988<pre>
2989 &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]2990</pre>
2991
2992<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2993<p>The '<tt>lshr</tt>' instruction (logical shift right) returns the first
2994 operand shifted to the right a specified number of bits with zero fill.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2995
2996<h5>Arguments:</h5>
2997<p>Both arguments to the '<tt>lshr</tt>' instruction must be the same
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]2998 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2999 type. '<tt>op2</tt>' is treated as an unsigned value.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3000
3001<h5>Semantics:</h5>
3002<p>This instruction always performs a logical shift right operation. The most
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3003 significant bits of the result will be filled with zero bits after the shift.
3004 If <tt>op2</tt> is (statically or dynamically) equal to or larger than the
3005 number of bits in <tt>op1</tt>, the result is undefined. If the arguments are
3006 vectors, each vector element of <tt>op1</tt> is shifted by the corresponding
3007 shift amount in <tt>op2</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3008
3009<h5>Example:</h5>
3010<pre>
3011 &lt;result&gt; = lshr i32 4, 1 <i>; yields {i32}:result = 2</i>
3012 &lt;result&gt; = lshr i32 4, 2 <i>; yields {i32}:result = 1</i>
3013 &lt;result&gt; = lshr i8 4, 3 <i>; yields {i8}:result = 0</i>
3014 &lt;result&gt; = lshr i8 -2, 1 <i>; yields {i8}:result = 0x7FFFFFFF </i>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]3015 &lt;result&gt; = lshr i32 1, 32 <i>; undefined</i>
Mon P Wang9901e732008-12-09 05:46:39 +0000[diff] [blame]3016 &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]3017</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3018
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3019</div>
3020
3021<!-- _______________________________________________________________________ -->
3022<div class="doc_subsubsection"> <a name="i_ashr">'<tt>ashr</tt>'
3023Instruction</a> </div>
3024<div class="doc_text">
3025
3026<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3027<pre>
3028 &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]3029</pre>
3030
3031<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3032<p>The '<tt>ashr</tt>' instruction (arithmetic shift right) returns the first
3033 operand shifted to the right a specified number of bits with sign
3034 extension.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3035
3036<h5>Arguments:</h5>
3037<p>Both arguments to the '<tt>ashr</tt>' instruction must be the same
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3038 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3039 type. '<tt>op2</tt>' is treated as an unsigned value.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3040
3041<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3042<p>This instruction always performs an arithmetic shift right operation, The
3043 most significant bits of the result will be filled with the sign bit
3044 of <tt>op1</tt>. If <tt>op2</tt> is (statically or dynamically) equal to or
3045 larger than the number of bits in <tt>op1</tt>, the result is undefined. If
3046 the arguments are vectors, each vector element of <tt>op1</tt> is shifted by
3047 the corresponding shift amount in <tt>op2</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3048
3049<h5>Example:</h5>
3050<pre>
3051 &lt;result&gt; = ashr i32 4, 1 <i>; yields {i32}:result = 2</i>
3052 &lt;result&gt; = ashr i32 4, 2 <i>; yields {i32}:result = 1</i>
3053 &lt;result&gt; = ashr i8 4, 3 <i>; yields {i8}:result = 0</i>
3054 &lt;result&gt; = ashr i8 -2, 1 <i>; yields {i8}:result = -1</i>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]3055 &lt;result&gt; = ashr i32 1, 32 <i>; undefined</i>
Mon P Wang9901e732008-12-09 05:46:39 +0000[diff] [blame]3056 &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]3057</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3058
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3059</div>
3060
3061<!-- _______________________________________________________________________ -->
3062<div class="doc_subsubsection"> <a name="i_and">'<tt>and</tt>'
3063Instruction</a> </div>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3064
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3065<div class="doc_text">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3066
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3067<h5>Syntax:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3068<pre>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]3069 &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]3070</pre>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3071
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3072<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3073<p>The '<tt>and</tt>' instruction returns the bitwise logical and of its two
3074 operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3075
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3076<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3077<p>The two arguments to the '<tt>and</tt>' instruction must be
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3078 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3079 values. Both arguments must have identical types.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3080
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3081<h5>Semantics:</h5>
3082<p>The truth table used for the '<tt>and</tt>' instruction is:</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3083
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3084<table border="1" cellspacing="0" cellpadding="4">
3085 <tbody>
3086 <tr>
3087 <td>In0</td>
3088 <td>In1</td>
3089 <td>Out</td>
3090 </tr>
3091 <tr>
3092 <td>0</td>
3093 <td>0</td>
3094 <td>0</td>
3095 </tr>
3096 <tr>
3097 <td>0</td>
3098 <td>1</td>
3099 <td>0</td>
3100 </tr>
3101 <tr>
3102 <td>1</td>
3103 <td>0</td>
3104 <td>0</td>
3105 </tr>
3106 <tr>
3107 <td>1</td>
3108 <td>1</td>
3109 <td>1</td>
3110 </tr>
3111 </tbody>
3112</table>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3113
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3114<h5>Example:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3115<pre>
3116 &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]3117 &lt;result&gt; = and i32 15, 40 <i>; yields {i32}:result = 8</i>
3118 &lt;result&gt; = and i32 4, 8 <i>; yields {i32}:result = 0</i>
3119</pre>
3120</div>
3121<!-- _______________________________________________________________________ -->
3122<div class="doc_subsubsection"> <a name="i_or">'<tt>or</tt>' Instruction</a> </div>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3123
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3124<div class="doc_text">
3125
3126<h5>Syntax:</h5>
3127<pre>
3128 &lt;result&gt; = or &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
3129</pre>
3130
3131<h5>Overview:</h5>
3132<p>The '<tt>or</tt>' instruction returns the bitwise logical inclusive or of its
3133 two operands.</p>
3134
3135<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3136<p>The two arguments to the '<tt>or</tt>' instruction must be
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3137 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3138 values. Both arguments must have identical types.</p>
3139
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3140<h5>Semantics:</h5>
3141<p>The truth table used for the '<tt>or</tt>' instruction is:</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3142
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3143<table border="1" cellspacing="0" cellpadding="4">
3144 <tbody>
3145 <tr>
3146 <td>In0</td>
3147 <td>In1</td>
3148 <td>Out</td>
3149 </tr>
3150 <tr>
3151 <td>0</td>
3152 <td>0</td>
3153 <td>0</td>
3154 </tr>
3155 <tr>
3156 <td>0</td>
3157 <td>1</td>
3158 <td>1</td>
3159 </tr>
3160 <tr>
3161 <td>1</td>
3162 <td>0</td>
3163 <td>1</td>
3164 </tr>
3165 <tr>
3166 <td>1</td>
3167 <td>1</td>
3168 <td>1</td>
3169 </tr>
3170 </tbody>
3171</table>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3172
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3173<h5>Example:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3174<pre>
3175 &lt;result&gt; = or i32 4, %var <i>; yields {i32}:result = 4 | %var</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3176 &lt;result&gt; = or i32 15, 40 <i>; yields {i32}:result = 47</i>
3177 &lt;result&gt; = or i32 4, 8 <i>; yields {i32}:result = 12</i>
3178</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3179
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3180</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3181
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3182<!-- _______________________________________________________________________ -->
3183<div class="doc_subsubsection"> <a name="i_xor">'<tt>xor</tt>'
3184Instruction</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3185
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3186<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3187
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3188<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3189<pre>
3190 &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]3191</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3192
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3193<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3194<p>The '<tt>xor</tt>' instruction returns the bitwise logical exclusive or of
3195 its two operands. The <tt>xor</tt> is used to implement the "one's
3196 complement" operation, which is the "~" operator in C.</p>
3197
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3198<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3199<p>The two arguments to the '<tt>xor</tt>' instruction must be
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3200 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
3201 values. Both arguments must have identical types.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3202
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3203<h5>Semantics:</h5>
3204<p>The truth table used for the '<tt>xor</tt>' instruction is:</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3205
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3206<table border="1" cellspacing="0" cellpadding="4">
3207 <tbody>
3208 <tr>
3209 <td>In0</td>
3210 <td>In1</td>
3211 <td>Out</td>
3212 </tr>
3213 <tr>
3214 <td>0</td>
3215 <td>0</td>
3216 <td>0</td>
3217 </tr>
3218 <tr>
3219 <td>0</td>
3220 <td>1</td>
3221 <td>1</td>
3222 </tr>
3223 <tr>
3224 <td>1</td>
3225 <td>0</td>
3226 <td>1</td>
3227 </tr>
3228 <tr>
3229 <td>1</td>
3230 <td>1</td>
3231 <td>0</td>
3232 </tr>
3233 </tbody>
3234</table>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3235
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3236<h5>Example:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3237<pre>
3238 &lt;result&gt; = xor i32 4, %var <i>; yields {i32}:result = 4 ^ %var</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3239 &lt;result&gt; = xor i32 15, 40 <i>; yields {i32}:result = 39</i>
3240 &lt;result&gt; = xor i32 4, 8 <i>; yields {i32}:result = 12</i>
3241 &lt;result&gt; = xor i32 %V, -1 <i>; yields {i32}:result = ~%V</i>
3242</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3243
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3244</div>
3245
3246<!-- ======================================================================= -->
3247<div class="doc_subsection">
3248 <a name="vectorops">Vector Operations</a>
3249</div>
3250
3251<div class="doc_text">
3252
3253<p>LLVM supports several instructions to represent vector operations in a
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3254 target-independent manner. These instructions cover the element-access and
3255 vector-specific operations needed to process vectors effectively. While LLVM
3256 does directly support these vector operations, many sophisticated algorithms
3257 will want to use target-specific intrinsics to take full advantage of a
3258 specific target.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3259
3260</div>
3261
3262<!-- _______________________________________________________________________ -->
3263<div class="doc_subsubsection">
3264 <a name="i_extractelement">'<tt>extractelement</tt>' Instruction</a>
3265</div>
3266
3267<div class="doc_text">
3268
3269<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3270<pre>
3271 &lt;result&gt; = extractelement &lt;n x &lt;ty&gt;&gt; &lt;val&gt;, i32 &lt;idx&gt; <i>; yields &lt;ty&gt;</i>
3272</pre>
3273
3274<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3275<p>The '<tt>extractelement</tt>' instruction extracts a single scalar element
3276 from a vector at a specified index.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3277
3278
3279<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3280<p>The first operand of an '<tt>extractelement</tt>' instruction is a value
3281 of <a href="#t_vector">vector</a> type. The second operand is an index
3282 indicating the position from which to extract the element. The index may be
3283 a variable.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3284
3285<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3286<p>The result is a scalar of the same type as the element type of
3287 <tt>val</tt>. Its value is the value at position <tt>idx</tt> of
3288 <tt>val</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
3289 results are undefined.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3290
3291<h5>Example:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3292<pre>
3293 %result = extractelement &lt;4 x i32&gt; %vec, i32 0 <i>; yields i32</i>
3294</pre>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3295
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3296</div>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3297
3298<!-- _______________________________________________________________________ -->
3299<div class="doc_subsubsection">
3300 <a name="i_insertelement">'<tt>insertelement</tt>' Instruction</a>
3301</div>
3302
3303<div class="doc_text">
3304
3305<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3306<pre>
Dan Gohmanbcc3c502008-05-12 23:38:42 +0000[diff] [blame]3307 &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]3308</pre>
3309
3310<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3311<p>The '<tt>insertelement</tt>' instruction inserts a scalar element into a
3312 vector at a specified index.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3313
3314<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3315<p>The first operand of an '<tt>insertelement</tt>' instruction is a value
3316 of <a href="#t_vector">vector</a> type. The second operand is a scalar value
3317 whose type must equal the element type of the first operand. The third
3318 operand is an index indicating the position at which to insert the value.
3319 The index may be a variable.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3320
3321<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3322<p>The result is a vector of the same type as <tt>val</tt>. Its element values
3323 are those of <tt>val</tt> except at position <tt>idx</tt>, where it gets the
3324 value <tt>elt</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
3325 results are undefined.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3326
3327<h5>Example:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3328<pre>
3329 %result = insertelement &lt;4 x i32&gt; %vec, i32 1, i32 0 <i>; yields &lt;4 x i32&gt;</i>
3330</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3331
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3332</div>
3333
3334<!-- _______________________________________________________________________ -->
3335<div class="doc_subsubsection">
3336 <a name="i_shufflevector">'<tt>shufflevector</tt>' Instruction</a>
3337</div>
3338
3339<div class="doc_text">
3340
3341<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3342<pre>
Mon P Wangbff5d9c2008-11-10 04:46:22 +0000[diff] [blame]3343 &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]3344</pre>
3345
3346<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3347<p>The '<tt>shufflevector</tt>' instruction constructs a permutation of elements
3348 from two input vectors, returning a vector with the same element type as the
3349 input and length that is the same as the shuffle mask.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3350
3351<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3352<p>The first two operands of a '<tt>shufflevector</tt>' instruction are vectors
3353 with types that match each other. The third argument is a shuffle mask whose
3354 element type is always 'i32'. The result of the instruction is a vector
3355 whose length is the same as the shuffle mask and whose element type is the
3356 same as the element type of the first two operands.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3357
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3358<p>The shuffle mask operand is required to be a constant vector with either
3359 constant integer or undef values.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3360
3361<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3362<p>The elements of the two input vectors are numbered from left to right across
3363 both of the vectors. The shuffle mask operand specifies, for each element of
3364 the result vector, which element of the two input vectors the result element
3365 gets. The element selector may be undef (meaning "don't care") and the
3366 second operand may be undef if performing a shuffle from only one vector.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3367
3368<h5>Example:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3369<pre>
3370 %result = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; %v2,
3371 &lt;4 x i32&gt; &lt;i32 0, i32 4, i32 1, i32 5&gt; <i>; yields &lt;4 x i32&gt;</i>
3372 %result = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; undef,
3373 &lt;4 x i32&gt; &lt;i32 0, i32 1, i32 2, i32 3&gt; <i>; yields &lt;4 x i32&gt;</i> - Identity shuffle.
Mon P Wangbff5d9c2008-11-10 04:46:22 +0000[diff] [blame]3374 %result = shufflevector &lt;8 x i32&gt; %v1, &lt;8 x i32&gt; undef,
3375 &lt;4 x i32&gt; &lt;i32 0, i32 1, i32 2, i32 3&gt; <i>; yields &lt;4 x i32&gt;</i>
3376 %result = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; %v2,
3377 &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]3378</pre>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3379
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3380</div>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3381
3382<!-- ======================================================================= -->
3383<div class="doc_subsection">
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]3384 <a name="aggregateops">Aggregate Operations</a>
3385</div>
3386
3387<div class="doc_text">
3388
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3389<p>LLVM supports several instructions for working with aggregate values.</p>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]3390
3391</div>
3392
3393<!-- _______________________________________________________________________ -->
3394<div class="doc_subsubsection">
3395 <a name="i_extractvalue">'<tt>extractvalue</tt>' Instruction</a>
3396</div>
3397
3398<div class="doc_text">
3399
3400<h5>Syntax:</h5>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]3401<pre>
3402 &lt;result&gt; = extractvalue &lt;aggregate type&gt; &lt;val&gt;, &lt;idx&gt;{, &lt;idx&gt;}*
3403</pre>
3404
3405<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3406<p>The '<tt>extractvalue</tt>' instruction extracts the value of a struct field
3407 or array element from an aggregate value.</p>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]3408
3409<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3410<p>The first operand of an '<tt>extractvalue</tt>' instruction is a value
3411 of <a href="#t_struct">struct</a> or <a href="#t_array">array</a> type. The
3412 operands are constant indices to specify which value to extract in a similar
3413 manner as indices in a
3414 '<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.</p>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]3415
3416<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3417<p>The result is the value at the position in the aggregate specified by the
3418 index operands.</p>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]3419
3420<h5>Example:</h5>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]3421<pre>
Dan Gohmane5febe42008-05-31 00:58:22 +0000[diff] [blame]3422 %result = extractvalue {i32, float} %agg, 0 <i>; yields i32</i>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]3423</pre>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]3424
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3425</div>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]3426
3427<!-- _______________________________________________________________________ -->
3428<div class="doc_subsubsection">
3429 <a name="i_insertvalue">'<tt>insertvalue</tt>' Instruction</a>
3430</div>
3431
3432<div class="doc_text">
3433
3434<h5>Syntax:</h5>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]3435<pre>
Dan Gohmane5febe42008-05-31 00:58:22 +0000[diff] [blame]3436 &lt;result&gt; = insertvalue &lt;aggregate type&gt; &lt;val&gt;, &lt;ty&gt; &lt;val&gt;, &lt;idx&gt; <i>; yields &lt;n x &lt;ty&gt;&gt;</i>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]3437</pre>
3438
3439<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3440<p>The '<tt>insertvalue</tt>' instruction inserts a value into a struct field or
3441 array element in an aggregate.</p>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]3442
3443
3444<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3445<p>The first operand of an '<tt>insertvalue</tt>' instruction is a value
3446 of <a href="#t_struct">struct</a> or <a href="#t_array">array</a> type. The
3447 second operand is a first-class value to insert. The following operands are
3448 constant indices indicating the position at which to insert the value in a
3449 similar manner as indices in a
3450 '<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction. The
3451 value to insert must have the same type as the value identified by the
3452 indices.</p>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]3453
3454<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3455<p>The result is an aggregate of the same type as <tt>val</tt>. Its value is
3456 that of <tt>val</tt> except that the value at the position specified by the
3457 indices is that of <tt>elt</tt>.</p>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]3458
3459<h5>Example:</h5>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]3460<pre>
Dan Gohmanb1aab4e2008-06-23 15:26:37 +0000[diff] [blame]3461 %result = insertvalue {i32, float} %agg, i32 1, 0 <i>; yields {i32, float}</i>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]3462</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3463
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]3464</div>
3465
3466
3467<!-- ======================================================================= -->
3468<div class="doc_subsection">
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3469 <a name="memoryops">Memory Access and Addressing Operations</a>
3470</div>
3471
3472<div class="doc_text">
3473
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3474<p>A key design point of an SSA-based representation is how it represents
3475 memory. In LLVM, no memory locations are in SSA form, which makes things
3476 very simple. This section describes how to read, write, allocate, and free
3477 memory in LLVM.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3478
3479</div>
3480
3481<!-- _______________________________________________________________________ -->
3482<div class="doc_subsubsection">
3483 <a name="i_malloc">'<tt>malloc</tt>' Instruction</a>
3484</div>
3485
3486<div class="doc_text">
3487
3488<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3489<pre>
3490 &lt;result&gt; = malloc &lt;type&gt;[, i32 &lt;NumElements&gt;][, align &lt;alignment&gt;] <i>; yields {type*}:result</i>
3491</pre>
3492
3493<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3494<p>The '<tt>malloc</tt>' instruction allocates memory from the system heap and
3495 returns a pointer to it. The object is always allocated in the generic
3496 address space (address space zero).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3497
3498<h5>Arguments:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3499<p>The '<tt>malloc</tt>' instruction allocates
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3500 <tt>sizeof(&lt;type&gt;)*NumElements</tt> bytes of memory from the operating
3501 system and returns a pointer of the appropriate type to the program. If
3502 "NumElements" is specified, it is the number of elements allocated, otherwise
3503 "NumElements" is defaulted to be one. If a constant alignment is specified,
3504 the value result of the allocation is guaranteed to be aligned to at least
3505 that boundary. If not specified, or if zero, the target can choose to align
3506 the allocation on any convenient boundary compatible with the type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3507
3508<p>'<tt>type</tt>' must be a sized type.</p>
3509
3510<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3511<p>Memory is allocated using the system "<tt>malloc</tt>" function, and a
3512 pointer is returned. The result of a zero byte allocation is undefined. The
3513 result is null if there is insufficient memory available.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3514
3515<h5>Example:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3516<pre>
Dan Gohmanf54f50a2009-01-04 23:49:44 +0000[diff] [blame]3517 %array = malloc [4 x i8] <i>; yields {[%4 x i8]*}:array</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3518
3519 %size = <a href="#i_add">add</a> i32 2, 2 <i>; yields {i32}:size = i32 4</i>
3520 %array1 = malloc i8, i32 4 <i>; yields {i8*}:array1</i>
3521 %array2 = malloc [12 x i8], i32 %size <i>; yields {[12 x i8]*}:array2</i>
3522 %array3 = malloc i32, i32 4, align 1024 <i>; yields {i32*}:array3</i>
3523 %array4 = malloc i32, align 1024 <i>; yields {i32*}:array4</i>
3524</pre>
Dan Gohman60967192009-01-12 23:12:39 +0000[diff] [blame]3525
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3526<p>Note that the code generator does not yet respect the alignment value.</p>
Dan Gohman60967192009-01-12 23:12:39 +0000[diff] [blame]3527
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3528</div>
3529
3530<!-- _______________________________________________________________________ -->
3531<div class="doc_subsubsection">
3532 <a name="i_free">'<tt>free</tt>' Instruction</a>
3533</div>
3534
3535<div class="doc_text">
3536
3537<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3538<pre>
Dan Gohmanf54f50a2009-01-04 23:49:44 +0000[diff] [blame]3539 free &lt;type&gt; &lt;value&gt; <i>; yields {void}</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3540</pre>
3541
3542<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3543<p>The '<tt>free</tt>' instruction returns memory back to the unused memory heap
3544 to be reallocated in the future.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3545
3546<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3547<p>'<tt>value</tt>' shall be a pointer value that points to a value that was
3548 allocated with the '<tt><a href="#i_malloc">malloc</a></tt>' instruction.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3549
3550<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3551<p>Access to the memory pointed to by the pointer is no longer defined after
3552 this instruction executes. If the pointer is null, the operation is a
3553 noop.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3554
3555<h5>Example:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3556<pre>
Dan Gohmanf54f50a2009-01-04 23:49:44 +0000[diff] [blame]3557 %array = <a href="#i_malloc">malloc</a> [4 x i8] <i>; yields {[4 x i8]*}:array</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3558 free [4 x i8]* %array
3559</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3560
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3561</div>
3562
3563<!-- _______________________________________________________________________ -->
3564<div class="doc_subsubsection">
3565 <a name="i_alloca">'<tt>alloca</tt>' Instruction</a>
3566</div>
3567
3568<div class="doc_text">
3569
3570<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3571<pre>
3572 &lt;result&gt; = alloca &lt;type&gt;[, i32 &lt;NumElements&gt;][, align &lt;alignment&gt;] <i>; yields {type*}:result</i>
3573</pre>
3574
3575<h5>Overview:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3576<p>The '<tt>alloca</tt>' instruction allocates memory on the stack frame of the
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3577 currently executing function, to be automatically released when this function
3578 returns to its caller. The object is always allocated in the generic address
3579 space (address space zero).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3580
3581<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3582<p>The '<tt>alloca</tt>' instruction
3583 allocates <tt>sizeof(&lt;type&gt;)*NumElements</tt> bytes of memory on the
3584 runtime stack, returning a pointer of the appropriate type to the program.
3585 If "NumElements" is specified, it is the number of elements allocated,
3586 otherwise "NumElements" is defaulted to be one. If a constant alignment is
3587 specified, the value result of the allocation is guaranteed to be aligned to
3588 at least that boundary. If not specified, or if zero, the target can choose
3589 to align the allocation on any convenient boundary compatible with the
3590 type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3591
3592<p>'<tt>type</tt>' may be any sized type.</p>
3593
3594<h5>Semantics:</h5>
Bill Wendling2a454572009-05-08 20:49:29 +0000[diff] [blame]3595<p>Memory is allocated; a pointer is returned. The operation is undefined if
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3596 there is insufficient stack space for the allocation. '<tt>alloca</tt>'d
3597 memory is automatically released when the function returns. The
3598 '<tt>alloca</tt>' instruction is commonly used to represent automatic
3599 variables that must have an address available. When the function returns
3600 (either with the <tt><a href="#i_ret">ret</a></tt>
3601 or <tt><a href="#i_unwind">unwind</a></tt> instructions), the memory is
3602 reclaimed. Allocating zero bytes is legal, but the result is undefined.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3603
3604<h5>Example:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3605<pre>
Dan Gohmanf54f50a2009-01-04 23:49:44 +0000[diff] [blame]3606 %ptr = alloca i32 <i>; yields {i32*}:ptr</i>
3607 %ptr = alloca i32, i32 4 <i>; yields {i32*}:ptr</i>
3608 %ptr = alloca i32, i32 4, align 1024 <i>; yields {i32*}:ptr</i>
3609 %ptr = alloca i32, align 1024 <i>; yields {i32*}:ptr</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3610</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3611
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3612</div>
3613
3614<!-- _______________________________________________________________________ -->
3615<div class="doc_subsubsection"> <a name="i_load">'<tt>load</tt>'
3616Instruction</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3617
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3618<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3619
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3620<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3621<pre>
3622 &lt;result&gt; = load &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;]
3623 &lt;result&gt; = volatile load &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;]
3624</pre>
3625
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3626<h5>Overview:</h5>
3627<p>The '<tt>load</tt>' instruction is used to read from memory.</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3628
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3629<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3630<p>The argument to the '<tt>load</tt>' instruction specifies the memory address
3631 from which to load. The pointer must point to
3632 a <a href="#t_firstclass">first class</a> type. If the <tt>load</tt> is
3633 marked as <tt>volatile</tt>, then the optimizer is not allowed to modify the
3634 number or order of execution of this <tt>load</tt> with other
3635 volatile <tt>load</tt> and <tt><a href="#i_store">store</a></tt>
3636 instructions. </p>
3637
3638<p>The optional constant "align" argument specifies the alignment of the
3639 operation (that is, the alignment of the memory address). A value of 0 or an
3640 omitted "align" argument means that the operation has the preferential
3641 alignment for the target. It is the responsibility of the code emitter to
3642 ensure that the alignment information is correct. Overestimating the
3643 alignment results in an undefined behavior. Underestimating the alignment may
3644 produce less efficient code. An alignment of 1 is always safe.</p>
3645
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3646<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3647<p>The location of memory pointed to is loaded. If the value being loaded is of
3648 scalar type then the number of bytes read does not exceed the minimum number
3649 of bytes needed to hold all bits of the type. For example, loading an
3650 <tt>i24</tt> reads at most three bytes. When loading a value of a type like
3651 <tt>i20</tt> with a size that is not an integral number of bytes, the result
3652 is undefined if the value was not originally written using a store of the
3653 same type.</p>
3654
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3655<h5>Examples:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3656<pre>
3657 %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
3658 <a href="#i_store">store</a> i32 3, i32* %ptr <i>; yields {void}</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3659 %val = load i32* %ptr <i>; yields {i32}:val = i32 3</i>
3660</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3661
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3662</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3663
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3664<!-- _______________________________________________________________________ -->
3665<div class="doc_subsubsection"> <a name="i_store">'<tt>store</tt>'
3666Instruction</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3667
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3668<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3669
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3670<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3671<pre>
3672 store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;] <i>; yields {void}</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3673 volatile store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;] <i>; yields {void}</i>
3674</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3675
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3676<h5>Overview:</h5>
3677<p>The '<tt>store</tt>' instruction is used to write to memory.</p>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3678
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3679<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3680<p>There are two arguments to the '<tt>store</tt>' instruction: a value to store
3681 and an address at which to store it. The type of the
3682 '<tt>&lt;pointer&gt;</tt>' operand must be a pointer to
3683 the <a href="#t_firstclass">first class</a> type of the
3684 '<tt>&lt;value&gt;</tt>' operand. If the <tt>store</tt> is marked
3685 as <tt>volatile</tt>, then the optimizer is not allowed to modify the number
3686 or order of execution of this <tt>store</tt> with other
3687 volatile <tt>load</tt> and <tt><a href="#i_store">store</a></tt>
3688 instructions.</p>
3689
3690<p>The optional constant "align" argument specifies the alignment of the
3691 operation (that is, the alignment of the memory address). A value of 0 or an
3692 omitted "align" argument means that the operation has the preferential
3693 alignment for the target. It is the responsibility of the code emitter to
3694 ensure that the alignment information is correct. Overestimating the
3695 alignment results in an undefined behavior. Underestimating the alignment may
3696 produce less efficient code. An alignment of 1 is always safe.</p>
3697
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3698<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3699<p>The contents of memory are updated to contain '<tt>&lt;value&gt;</tt>' at the
3700 location specified by the '<tt>&lt;pointer&gt;</tt>' operand. If
3701 '<tt>&lt;value&gt;</tt>' is of scalar type then the number of bytes written
3702 does not exceed the minimum number of bytes needed to hold all bits of the
3703 type. For example, storing an <tt>i24</tt> writes at most three bytes. When
3704 writing a value of a type like <tt>i20</tt> with a size that is not an
3705 integral number of bytes, it is unspecified what happens to the extra bits
3706 that do not belong to the type, but they will typically be overwritten.</p>
3707
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3708<h5>Example:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3709<pre>
3710 %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
Bill Wendling63ffa142007-10-22 05:10:05 +0000[diff] [blame]3711 store i32 3, i32* %ptr <i>; yields {void}</i>
3712 %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]3713</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3714
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3715</div>
3716
3717<!-- _______________________________________________________________________ -->
3718<div class="doc_subsubsection">
3719 <a name="i_getelementptr">'<tt>getelementptr</tt>' Instruction</a>
3720</div>
3721
3722<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3723
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3724<h5>Syntax:</h5>
3725<pre>
Matthijs Kooijman32a080e2008-10-13 13:44:15 +0000[diff] [blame]3726 &lt;result&gt; = getelementptr &lt;pty&gt;* &lt;ptrval&gt;{, &lt;ty&gt; &lt;idx&gt;}*
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3727</pre>
3728
3729<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3730<p>The '<tt>getelementptr</tt>' instruction is used to get the address of a
3731 subelement of an aggregate data structure. It performs address calculation
3732 only and does not access memory.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3733
3734<h5>Arguments:</h5>
Matthijs Kooijman32a080e2008-10-13 13:44:15 +0000[diff] [blame]3735<p>The first argument is always a pointer, and forms the basis of the
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3736 calculation. The remaining arguments are indices, that indicate which of the
3737 elements of the aggregate object are indexed. The interpretation of each
3738 index is dependent on the type being indexed into. The first index always
3739 indexes the pointer value given as the first argument, the second index
3740 indexes a value of the type pointed to (not necessarily the value directly
3741 pointed to, since the first index can be non-zero), etc. The first type
3742 indexed into must be a pointer value, subsequent types can be arrays, vectors
3743 and structs. Note that subsequent types being indexed into can never be
3744 pointers, since that would require loading the pointer before continuing
3745 calculation.</p>
Matthijs Kooijman32a080e2008-10-13 13:44:15 +0000[diff] [blame]3746
3747<p>The type of each index argument depends on the type it is indexing into.
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3748 When indexing into a (packed) structure, only <tt>i32</tt> integer
3749 <b>constants</b> are allowed. When indexing into an array, pointer or
3750 vector, integers of any width are allowed (also non-constants).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3751
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3752<p>For example, let's consider a C code fragment and how it gets compiled to
3753 LLVM:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3754
3755<div class="doc_code">
3756<pre>
3757struct RT {
3758 char A;
3759 int B[10][20];
3760 char C;
3761};
3762struct ST {
3763 int X;
3764 double Y;
3765 struct RT Z;
3766};
3767
3768int *foo(struct ST *s) {
3769 return &amp;s[1].Z.B[5][13];
3770}
3771</pre>
3772</div>
3773
3774<p>The LLVM code generated by the GCC frontend is:</p>
3775
3776<div class="doc_code">
3777<pre>
Chris Lattner5b6dc6e2009-01-11 20:53:49 +0000[diff] [blame]3778%RT = <a href="#namedtypes">type</a> { i8 , [10 x [20 x i32]], i8 }
3779%ST = <a href="#namedtypes">type</a> { i32, double, %RT }
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3780
3781define i32* %foo(%ST* %s) {
3782entry:
3783 %reg = getelementptr %ST* %s, i32 1, i32 2, i32 1, i32 5, i32 13
3784 ret i32* %reg
3785}
3786</pre>
3787</div>
3788
3789<h5>Semantics:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3790<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]3791 type, which is a pointer, yielding a '<tt>%ST</tt>' = '<tt>{ i32, double, %RT
3792 }</tt>' type, a structure. The second index indexes into the third element
3793 of the structure, yielding a '<tt>%RT</tt>' = '<tt>{ i8 , [10 x [20 x i32]],
3794 i8 }</tt>' type, another structure. The third index indexes into the second
3795 element of the structure, yielding a '<tt>[10 x [20 x i32]]</tt>' type, an
3796 array. The two dimensions of the array are subscripted into, yielding an
3797 '<tt>i32</tt>' type. The '<tt>getelementptr</tt>' instruction returns a
3798 pointer to this element, thus computing a value of '<tt>i32*</tt>' type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3799
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3800<p>Note that it is perfectly legal to index partially through a structure,
3801 returning a pointer to an inner element. Because of this, the LLVM code for
3802 the given testcase is equivalent to:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3803
3804<pre>
3805 define i32* %foo(%ST* %s) {
3806 %t1 = getelementptr %ST* %s, i32 1 <i>; yields %ST*:%t1</i>
3807 %t2 = getelementptr %ST* %t1, i32 0, i32 2 <i>; yields %RT*:%t2</i>
3808 %t3 = getelementptr %RT* %t2, i32 0, i32 1 <i>; yields [10 x [20 x i32]]*:%t3</i>
3809 %t4 = getelementptr [10 x [20 x i32]]* %t3, i32 0, i32 5 <i>; yields [20 x i32]*:%t4</i>
3810 %t5 = getelementptr [20 x i32]* %t4, i32 0, i32 13 <i>; yields i32*:%t5</i>
3811 ret i32* %t5
3812 }
3813</pre>
3814
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3815<p>Note that it is undefined to access an array out of bounds: array and pointer
3816 indexes must always be within the defined bounds of the array type when
3817 accessed with an instruction that dereferences the pointer (e.g. a load or
3818 store instruction). The one exception for this rule is zero length arrays.
3819 These arrays are defined to be accessible as variable length arrays, which
3820 requires access beyond the zero'th element.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3821
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3822<p>The getelementptr instruction is often confusing. For some more insight into
3823 how it works, see <a href="GetElementPtr.html">the getelementptr FAQ</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3824
3825<h5>Example:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3826<pre>
3827 <i>; yields [12 x i8]*:aptr</i>
Matthijs Kooijman32a080e2008-10-13 13:44:15 +0000[diff] [blame]3828 %aptr = getelementptr {i32, [12 x i8]}* %saptr, i64 0, i32 1
3829 <i>; yields i8*:vptr</i>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]3830 %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]3831 <i>; yields i8*:eptr</i>
3832 %eptr = getelementptr [12 x i8]* %aptr, i64 0, i32 1
Sanjiv Gupta4f9a0dc2009-04-25 07:27:44 +0000[diff] [blame]3833 <i>; yields i32*:iptr</i>
Sanjiv Gupta1e46c582009-04-24 16:38:13 +0000[diff] [blame]3834 %iptr = getelementptr [10 x i32]* @arr, i16 0, i16 0
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3835</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3836
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3837</div>
3838
3839<!-- ======================================================================= -->
3840<div class="doc_subsection"> <a name="convertops">Conversion Operations</a>
3841</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3842
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3843<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3844
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3845<p>The instructions in this category are the conversion instructions (casting)
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3846 which all take a single operand and a type. They perform various bit
3847 conversions on the operand.</p>
3848
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3849</div>
3850
3851<!-- _______________________________________________________________________ -->
3852<div class="doc_subsubsection">
3853 <a name="i_trunc">'<tt>trunc .. to</tt>' Instruction</a>
3854</div>
3855<div class="doc_text">
3856
3857<h5>Syntax:</h5>
3858<pre>
3859 &lt;result&gt; = trunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3860</pre>
3861
3862<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3863<p>The '<tt>trunc</tt>' instruction truncates its operand to the
3864 type <tt>ty2</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3865
3866<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3867<p>The '<tt>trunc</tt>' instruction takes a <tt>value</tt> to trunc, which must
3868 be an <a href="#t_integer">integer</a> type, and a type that specifies the
3869 size and type of the result, which must be
3870 an <a href="#t_integer">integer</a> type. The bit size of <tt>value</tt> must
3871 be larger than the bit size of <tt>ty2</tt>. Equal sized types are not
3872 allowed.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3873
3874<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3875<p>The '<tt>trunc</tt>' instruction truncates the high order bits
3876 in <tt>value</tt> and converts the remaining bits to <tt>ty2</tt>. Since the
3877 source size must be larger than the destination size, <tt>trunc</tt> cannot
3878 be a <i>no-op cast</i>. It will always truncate bits.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3879
3880<h5>Example:</h5>
3881<pre>
3882 %X = trunc i32 257 to i8 <i>; yields i8:1</i>
3883 %Y = trunc i32 123 to i1 <i>; yields i1:true</i>
3884 %Y = trunc i32 122 to i1 <i>; yields i1:false</i>
3885</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3886
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3887</div>
3888
3889<!-- _______________________________________________________________________ -->
3890<div class="doc_subsubsection">
3891 <a name="i_zext">'<tt>zext .. to</tt>' Instruction</a>
3892</div>
3893<div class="doc_text">
3894
3895<h5>Syntax:</h5>
3896<pre>
3897 &lt;result&gt; = zext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3898</pre>
3899
3900<h5>Overview:</h5>
3901<p>The '<tt>zext</tt>' instruction zero extends its operand to type
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3902 <tt>ty2</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3903
3904
3905<h5>Arguments:</h5>
3906<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]3907 <a href="#t_integer">integer</a> type, and a type to cast it to, which must
3908 also be of <a href="#t_integer">integer</a> type. The bit size of the
3909 <tt>value</tt> must be smaller than the bit size of the destination type,
3910 <tt>ty2</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3911
3912<h5>Semantics:</h5>
3913<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]3914 bits until it reaches the size of the destination type, <tt>ty2</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3915
3916<p>When zero extending from i1, the result will always be either 0 or 1.</p>
3917
3918<h5>Example:</h5>
3919<pre>
3920 %X = zext i32 257 to i64 <i>; yields i64:257</i>
3921 %Y = zext i1 true to i32 <i>; yields i32:1</i>
3922</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3923
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3924</div>
3925
3926<!-- _______________________________________________________________________ -->
3927<div class="doc_subsubsection">
3928 <a name="i_sext">'<tt>sext .. to</tt>' Instruction</a>
3929</div>
3930<div class="doc_text">
3931
3932<h5>Syntax:</h5>
3933<pre>
3934 &lt;result&gt; = sext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3935</pre>
3936
3937<h5>Overview:</h5>
3938<p>The '<tt>sext</tt>' sign extends <tt>value</tt> to the type <tt>ty2</tt>.</p>
3939
3940<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3941<p>The '<tt>sext</tt>' instruction takes a value to cast, which must be of
3942 <a href="#t_integer">integer</a> type, and a type to cast it to, which must
3943 also be of <a href="#t_integer">integer</a> type. The bit size of the
3944 <tt>value</tt> must be smaller than the bit size of the destination type,
3945 <tt>ty2</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3946
3947<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3948<p>The '<tt>sext</tt>' instruction performs a sign extension by copying the sign
3949 bit (highest order bit) of the <tt>value</tt> until it reaches the bit size
3950 of the type <tt>ty2</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3951
3952<p>When sign extending from i1, the extension always results in -1 or 0.</p>
3953
3954<h5>Example:</h5>
3955<pre>
3956 %X = sext i8 -1 to i16 <i>; yields i16 :65535</i>
3957 %Y = sext i1 true to i32 <i>; yields i32:-1</i>
3958</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3959
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3960</div>
3961
3962<!-- _______________________________________________________________________ -->
3963<div class="doc_subsubsection">
3964 <a name="i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a>
3965</div>
3966
3967<div class="doc_text">
3968
3969<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3970<pre>
3971 &lt;result&gt; = fptrunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3972</pre>
3973
3974<h5>Overview:</h5>
3975<p>The '<tt>fptrunc</tt>' instruction truncates <tt>value</tt> to type
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3976 <tt>ty2</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3977
3978<h5>Arguments:</h5>
3979<p>The '<tt>fptrunc</tt>' instruction takes a <a href="#t_floating">floating
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3980 point</a> value to cast and a <a href="#t_floating">floating point</a> type
3981 to cast it to. The size of <tt>value</tt> must be larger than the size of
3982 <tt>ty2</tt>. This implies that <tt>fptrunc</tt> cannot be used to make a
3983 <i>no-op cast</i>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3984
3985<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3986<p>The '<tt>fptrunc</tt>' instruction truncates a <tt>value</tt> from a larger
3987 <a href="#t_floating">floating point</a> type to a smaller
3988 <a href="#t_floating">floating point</a> type. If the value cannot fit
3989 within the destination type, <tt>ty2</tt>, then the results are
3990 undefined.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3991
3992<h5>Example:</h5>
3993<pre>
3994 %X = fptrunc double 123.0 to float <i>; yields float:123.0</i>
3995 %Y = fptrunc double 1.0E+300 to float <i>; yields undefined</i>
3996</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]3997
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3998</div>
3999
4000<!-- _______________________________________________________________________ -->
4001<div class="doc_subsubsection">
4002 <a name="i_fpext">'<tt>fpext .. to</tt>' Instruction</a>
4003</div>
4004<div class="doc_text">
4005
4006<h5>Syntax:</h5>
4007<pre>
4008 &lt;result&gt; = fpext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4009</pre>
4010
4011<h5>Overview:</h5>
4012<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]4013 floating point value.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4014
4015<h5>Arguments:</h5>
4016<p>The '<tt>fpext</tt>' instruction takes a
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4017 <a href="#t_floating">floating point</a> <tt>value</tt> to cast, and
4018 a <a href="#t_floating">floating point</a> type to cast it to. The source
4019 type must be smaller than the destination type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4020
4021<h5>Semantics:</h5>
4022<p>The '<tt>fpext</tt>' instruction extends the <tt>value</tt> from a smaller
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4023 <a href="#t_floating">floating point</a> type to a larger
4024 <a href="#t_floating">floating point</a> type. The <tt>fpext</tt> cannot be
4025 used to make a <i>no-op cast</i> because it always changes bits. Use
4026 <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]4027
4028<h5>Example:</h5>
4029<pre>
4030 %X = fpext float 3.1415 to double <i>; yields double:3.1415</i>
4031 %Y = fpext float 1.0 to float <i>; yields float:1.0 (no-op)</i>
4032</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4033
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4034</div>
4035
4036<!-- _______________________________________________________________________ -->
4037<div class="doc_subsubsection">
4038 <a name="i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a>
4039</div>
4040<div class="doc_text">
4041
4042<h5>Syntax:</h5>
4043<pre>
Reid Spencere6adee82007-07-31 14:40:14 +0000[diff] [blame]4044 &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]4045</pre>
4046
4047<h5>Overview:</h5>
Reid Spencere6adee82007-07-31 14:40:14 +0000[diff] [blame]4048<p>The '<tt>fptoui</tt>' converts a floating point <tt>value</tt> to its
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4049 unsigned integer equivalent of type <tt>ty2</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4050
4051<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4052<p>The '<tt>fptoui</tt>' instruction takes a value to cast, which must be a
4053 scalar or vector <a href="#t_floating">floating point</a> value, and a type
4054 to cast it to <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a>
4055 type. If <tt>ty</tt> is a vector floating point type, <tt>ty2</tt> must be a
4056 vector integer type with the same number of elements as <tt>ty</tt></p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4057
4058<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4059<p>The '<tt>fptoui</tt>' instruction converts its
4060 <a href="#t_floating">floating point</a> operand into the nearest (rounding
4061 towards zero) unsigned integer value. If the value cannot fit
4062 in <tt>ty2</tt>, the results are undefined.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4063
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4064<h5>Example:</h5>
4065<pre>
Reid Spencere6adee82007-07-31 14:40:14 +0000[diff] [blame]4066 %X = fptoui double 123.0 to i32 <i>; yields i32:123</i>
Chris Lattner681f1e82007-09-22 03:17:52 +0000[diff] [blame]4067 %Y = fptoui float 1.0E+300 to i1 <i>; yields undefined:1</i>
Reid Spencere6adee82007-07-31 14:40:14 +0000[diff] [blame]4068 %X = fptoui float 1.04E+17 to i8 <i>; yields undefined:1</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4069</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4070
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4071</div>
4072
4073<!-- _______________________________________________________________________ -->
4074<div class="doc_subsubsection">
4075 <a name="i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a>
4076</div>
4077<div class="doc_text">
4078
4079<h5>Syntax:</h5>
4080<pre>
4081 &lt;result&gt; = fptosi &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4082</pre>
4083
4084<h5>Overview:</h5>
4085<p>The '<tt>fptosi</tt>' instruction converts
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4086 <a href="#t_floating">floating point</a> <tt>value</tt> to
4087 type <tt>ty2</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4088
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4089<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4090<p>The '<tt>fptosi</tt>' instruction takes a value to cast, which must be a
4091 scalar or vector <a href="#t_floating">floating point</a> value, and a type
4092 to cast it to <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a>
4093 type. If <tt>ty</tt> is a vector floating point type, <tt>ty2</tt> must be a
4094 vector integer type with the same number of elements as <tt>ty</tt></p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4095
4096<h5>Semantics:</h5>
4097<p>The '<tt>fptosi</tt>' instruction converts its
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4098 <a href="#t_floating">floating point</a> operand into the nearest (rounding
4099 towards zero) signed integer value. If the value cannot fit in <tt>ty2</tt>,
4100 the results are undefined.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4101
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4102<h5>Example:</h5>
4103<pre>
4104 %X = fptosi double -123.0 to i32 <i>; yields i32:-123</i>
Chris Lattner681f1e82007-09-22 03:17:52 +0000[diff] [blame]4105 %Y = fptosi float 1.0E-247 to i1 <i>; yields undefined:1</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4106 %X = fptosi float 1.04E+17 to i8 <i>; yields undefined:1</i>
4107</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4108
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4109</div>
4110
4111<!-- _______________________________________________________________________ -->
4112<div class="doc_subsubsection">
4113 <a name="i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a>
4114</div>
4115<div class="doc_text">
4116
4117<h5>Syntax:</h5>
4118<pre>
4119 &lt;result&gt; = uitofp &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4120</pre>
4121
4122<h5>Overview:</h5>
4123<p>The '<tt>uitofp</tt>' instruction regards <tt>value</tt> as an unsigned
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4124 integer and converts that value to the <tt>ty2</tt> type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4125
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4126<h5>Arguments:</h5>
Nate Begeman78246ca2007-11-17 03:58:34 +0000[diff] [blame]4127<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]4128 scalar or vector <a href="#t_integer">integer</a> value, and a type to cast
4129 it to <tt>ty2</tt>, which must be an <a href="#t_floating">floating point</a>
4130 type. If <tt>ty</tt> is a vector integer type, <tt>ty2</tt> must be a vector
4131 floating point type with the same number of elements as <tt>ty</tt></p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4132
4133<h5>Semantics:</h5>
4134<p>The '<tt>uitofp</tt>' instruction interprets its operand as an unsigned
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4135 integer quantity and converts it to the corresponding floating point
4136 value. If the value cannot fit in the floating point value, the results are
4137 undefined.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4138
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4139<h5>Example:</h5>
4140<pre>
4141 %X = uitofp i32 257 to float <i>; yields float:257.0</i>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]4142 %Y = uitofp i8 -1 to double <i>; yields double:255.0</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4143</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4144
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4145</div>
4146
4147<!-- _______________________________________________________________________ -->
4148<div class="doc_subsubsection">
4149 <a name="i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a>
4150</div>
4151<div class="doc_text">
4152
4153<h5>Syntax:</h5>
4154<pre>
4155 &lt;result&gt; = sitofp &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4156</pre>
4157
4158<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4159<p>The '<tt>sitofp</tt>' instruction regards <tt>value</tt> as a signed integer
4160 and converts that value to the <tt>ty2</tt> type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4161
4162<h5>Arguments:</h5>
Nate Begeman78246ca2007-11-17 03:58:34 +0000[diff] [blame]4163<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]4164 scalar or vector <a href="#t_integer">integer</a> value, and a type to cast
4165 it to <tt>ty2</tt>, which must be an <a href="#t_floating">floating point</a>
4166 type. If <tt>ty</tt> is a vector integer type, <tt>ty2</tt> must be a vector
4167 floating point type with the same number of elements as <tt>ty</tt></p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4168
4169<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4170<p>The '<tt>sitofp</tt>' instruction interprets its operand as a signed integer
4171 quantity and converts it to the corresponding floating point value. If the
4172 value cannot fit in the floating point value, the results are undefined.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4173
4174<h5>Example:</h5>
4175<pre>
4176 %X = sitofp i32 257 to float <i>; yields float:257.0</i>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]4177 %Y = sitofp i8 -1 to double <i>; yields double:-1.0</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4178</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4179
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4180</div>
4181
4182<!-- _______________________________________________________________________ -->
4183<div class="doc_subsubsection">
4184 <a name="i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a>
4185</div>
4186<div class="doc_text">
4187
4188<h5>Syntax:</h5>
4189<pre>
4190 &lt;result&gt; = ptrtoint &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4191</pre>
4192
4193<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4194<p>The '<tt>ptrtoint</tt>' instruction converts the pointer <tt>value</tt> to
4195 the integer type <tt>ty2</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4196
4197<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4198<p>The '<tt>ptrtoint</tt>' instruction takes a <tt>value</tt> to cast, which
4199 must be a <a href="#t_pointer">pointer</a> value, and a type to cast it to
4200 <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]4201
4202<h5>Semantics:</h5>
4203<p>The '<tt>ptrtoint</tt>' instruction converts <tt>value</tt> to integer type
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4204 <tt>ty2</tt> by interpreting the pointer value as an integer and either
4205 truncating or zero extending that value to the size of the integer type. If
4206 <tt>value</tt> is smaller than <tt>ty2</tt> then a zero extension is done. If
4207 <tt>value</tt> is larger than <tt>ty2</tt> then a truncation is done. If they
4208 are the same size, then nothing is done (<i>no-op cast</i>) other than a type
4209 change.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4210
4211<h5>Example:</h5>
4212<pre>
4213 %X = ptrtoint i32* %X to i8 <i>; yields truncation on 32-bit architecture</i>
4214 %Y = ptrtoint i32* %x to i64 <i>; yields zero extension on 32-bit architecture</i>
4215</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4216
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4217</div>
4218
4219<!-- _______________________________________________________________________ -->
4220<div class="doc_subsubsection">
4221 <a name="i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a>
4222</div>
4223<div class="doc_text">
4224
4225<h5>Syntax:</h5>
4226<pre>
4227 &lt;result&gt; = inttoptr &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4228</pre>
4229
4230<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4231<p>The '<tt>inttoptr</tt>' instruction converts an integer <tt>value</tt> to a
4232 pointer type, <tt>ty2</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4233
4234<h5>Arguments:</h5>
4235<p>The '<tt>inttoptr</tt>' instruction takes an <a href="#t_integer">integer</a>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4236 value to cast, and a type to cast it to, which must be a
4237 <a href="#t_pointer">pointer</a> type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4238
4239<h5>Semantics:</h5>
4240<p>The '<tt>inttoptr</tt>' instruction converts <tt>value</tt> to type
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4241 <tt>ty2</tt> by applying either a zero extension or a truncation depending on
4242 the size of the integer <tt>value</tt>. If <tt>value</tt> is larger than the
4243 size of a pointer then a truncation is done. If <tt>value</tt> is smaller
4244 than the size of a pointer then a zero extension is done. If they are the
4245 same size, nothing is done (<i>no-op cast</i>).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4246
4247<h5>Example:</h5>
4248<pre>
4249 %X = inttoptr i32 255 to i32* <i>; yields zero extension on 64-bit architecture</i>
4250 %X = inttoptr i32 255 to i32* <i>; yields no-op on 32-bit architecture</i>
4251 %Y = inttoptr i64 0 to i32* <i>; yields truncation on 32-bit architecture</i>
4252</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4253
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4254</div>
4255
4256<!-- _______________________________________________________________________ -->
4257<div class="doc_subsubsection">
4258 <a name="i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a>
4259</div>
4260<div class="doc_text">
4261
4262<h5>Syntax:</h5>
4263<pre>
4264 &lt;result&gt; = bitcast &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
4265</pre>
4266
4267<h5>Overview:</h5>
4268<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4269 <tt>ty2</tt> without changing any bits.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4270
4271<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4272<p>The '<tt>bitcast</tt>' instruction takes a value to cast, which must be a
4273 non-aggregate first class value, and a type to cast it to, which must also be
4274 a non-aggregate <a href="#t_firstclass">first class</a> type. The bit sizes
4275 of <tt>value</tt> and the destination type, <tt>ty2</tt>, must be
4276 identical. If the source type is a pointer, the destination type must also be
4277 a pointer. This instruction supports bitwise conversion of vectors to
4278 integers and to vectors of other types (as long as they have the same
4279 size).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4280
4281<h5>Semantics:</h5>
4282<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4283 <tt>ty2</tt>. It is always a <i>no-op cast</i> because no bits change with
4284 this conversion. The conversion is done as if the <tt>value</tt> had been
4285 stored to memory and read back as type <tt>ty2</tt>. Pointer types may only
4286 be converted to other pointer types with this instruction. To convert
4287 pointers to other types, use the <a href="#i_inttoptr">inttoptr</a> or
4288 <a href="#i_ptrtoint">ptrtoint</a> instructions first.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4289
4290<h5>Example:</h5>
4291<pre>
4292 %X = bitcast i8 255 to i8 <i>; yields i8 :-1</i>
4293 %Y = bitcast i32* %x to sint* <i>; yields sint*:%x</i>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]4294 %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]4295</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4296
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4297</div>
4298
4299<!-- ======================================================================= -->
4300<div class="doc_subsection"> <a name="otherops">Other Operations</a> </div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4301
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4302<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4303
4304<p>The instructions in this category are the "miscellaneous" instructions, which
4305 defy better classification.</p>
4306
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4307</div>
4308
4309<!-- _______________________________________________________________________ -->
4310<div class="doc_subsubsection"><a name="i_icmp">'<tt>icmp</tt>' Instruction</a>
4311</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4312
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4313<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4314
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4315<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4316<pre>
4317 &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]4318</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4319
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4320<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4321<p>The '<tt>icmp</tt>' instruction returns a boolean value or a vector of
4322 boolean values based on comparison of its two integer, integer vector, or
4323 pointer operands.</p>
4324
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4325<h5>Arguments:</h5>
4326<p>The '<tt>icmp</tt>' instruction takes three operands. The first operand is
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4327 the condition code indicating the kind of comparison to perform. It is not a
4328 value, just a keyword. The possible condition code are:</p>
4329
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4330<ol>
4331 <li><tt>eq</tt>: equal</li>
4332 <li><tt>ne</tt>: not equal </li>
4333 <li><tt>ugt</tt>: unsigned greater than</li>
4334 <li><tt>uge</tt>: unsigned greater or equal</li>
4335 <li><tt>ult</tt>: unsigned less than</li>
4336 <li><tt>ule</tt>: unsigned less or equal</li>
4337 <li><tt>sgt</tt>: signed greater than</li>
4338 <li><tt>sge</tt>: signed greater or equal</li>
4339 <li><tt>slt</tt>: signed less than</li>
4340 <li><tt>sle</tt>: signed less or equal</li>
4341</ol>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4342
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4343<p>The remaining two arguments must be <a href="#t_integer">integer</a> or
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4344 <a href="#t_pointer">pointer</a> or integer <a href="#t_vector">vector</a>
4345 typed. They must also be identical types.</p>
4346
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4347<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4348<p>The '<tt>icmp</tt>' compares <tt>op1</tt> and <tt>op2</tt> according to the
4349 condition code given as <tt>cond</tt>. The comparison performed always yields
4350 either an <a href="#t_primitive"><tt>i1</tt></a> or vector of <tt>i1</tt>
4351 result, as follows:</p>
4352
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4353<ol>
4354 <li><tt>eq</tt>: yields <tt>true</tt> if the operands are equal,
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4355 <tt>false</tt> otherwise. No sign interpretation is necessary or
4356 performed.</li>
4357
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4358 <li><tt>ne</tt>: yields <tt>true</tt> if the operands are unequal,
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4359 <tt>false</tt> otherwise. No sign interpretation is necessary or
4360 performed.</li>
4361
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4362 <li><tt>ugt</tt>: interprets the operands as unsigned values and yields
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4363 <tt>true</tt> if <tt>op1</tt> is greater than <tt>op2</tt>.</li>
4364
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4365 <li><tt>uge</tt>: interprets the operands as unsigned values and yields
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4366 <tt>true</tt> if <tt>op1</tt> is greater than or equal
4367 to <tt>op2</tt>.</li>
4368
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4369 <li><tt>ult</tt>: interprets the operands as unsigned values and yields
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4370 <tt>true</tt> if <tt>op1</tt> is less than <tt>op2</tt>.</li>
4371
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4372 <li><tt>ule</tt>: interprets the operands as unsigned values and yields
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4373 <tt>true</tt> if <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
4374
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4375 <li><tt>sgt</tt>: interprets the operands as signed values and yields
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4376 <tt>true</tt> if <tt>op1</tt> is greater than <tt>op2</tt>.</li>
4377
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4378 <li><tt>sge</tt>: interprets the operands as signed values and yields
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4379 <tt>true</tt> if <tt>op1</tt> is greater than or equal
4380 to <tt>op2</tt>.</li>
4381
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4382 <li><tt>slt</tt>: interprets the operands as signed values and yields
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4383 <tt>true</tt> if <tt>op1</tt> is less than <tt>op2</tt>.</li>
4384
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4385 <li><tt>sle</tt>: interprets the operands as signed values and yields
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4386 <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]4387</ol>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4388
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4389<p>If the operands are <a href="#t_pointer">pointer</a> typed, the pointer
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4390 values are compared as if they were integers.</p>
4391
4392<p>If the operands are integer vectors, then they are compared element by
4393 element. The result is an <tt>i1</tt> vector with the same number of elements
4394 as the values being compared. Otherwise, the result is an <tt>i1</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4395
4396<h5>Example:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4397<pre>
4398 &lt;result&gt; = icmp eq i32 4, 5 <i>; yields: result=false</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4399 &lt;result&gt; = icmp ne float* %X, %X <i>; yields: result=false</i>
4400 &lt;result&gt; = icmp ult i16 4, 5 <i>; yields: result=true</i>
4401 &lt;result&gt; = icmp sgt i16 4, 5 <i>; yields: result=false</i>
4402 &lt;result&gt; = icmp ule i16 -4, 5 <i>; yields: result=false</i>
4403 &lt;result&gt; = icmp sge i16 4, 5 <i>; yields: result=false</i>
4404</pre>
Dan Gohmana53eb382009-01-22 01:39:38 +0000[diff] [blame]4405
4406<p>Note that the code generator does not yet support vector types with
4407 the <tt>icmp</tt> instruction.</p>
4408
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4409</div>
4410
4411<!-- _______________________________________________________________________ -->
4412<div class="doc_subsubsection"><a name="i_fcmp">'<tt>fcmp</tt>' Instruction</a>
4413</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4414
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4415<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4416
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4417<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4418<pre>
4419 &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]4420</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4421
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4422<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4423<p>The '<tt>fcmp</tt>' instruction returns a boolean value or vector of boolean
4424 values based on comparison of its operands.</p>
4425
4426<p>If the operands are floating point scalars, then the result type is a boolean
4427(<a href="#t_primitive"><tt>i1</tt></a>).</p>
4428
4429<p>If the operands are floating point vectors, then the result type is a vector
4430 of boolean with the same number of elements as the operands being
4431 compared.</p>
4432
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4433<h5>Arguments:</h5>
4434<p>The '<tt>fcmp</tt>' instruction takes three operands. The first operand is
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4435 the condition code indicating the kind of comparison to perform. It is not a
4436 value, just a keyword. The possible condition code are:</p>
4437
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4438<ol>
4439 <li><tt>false</tt>: no comparison, always returns false</li>
4440 <li><tt>oeq</tt>: ordered and equal</li>
4441 <li><tt>ogt</tt>: ordered and greater than </li>
4442 <li><tt>oge</tt>: ordered and greater than or equal</li>
4443 <li><tt>olt</tt>: ordered and less than </li>
4444 <li><tt>ole</tt>: ordered and less than or equal</li>
4445 <li><tt>one</tt>: ordered and not equal</li>
4446 <li><tt>ord</tt>: ordered (no nans)</li>
4447 <li><tt>ueq</tt>: unordered or equal</li>
4448 <li><tt>ugt</tt>: unordered or greater than </li>
4449 <li><tt>uge</tt>: unordered or greater than or equal</li>
4450 <li><tt>ult</tt>: unordered or less than </li>
4451 <li><tt>ule</tt>: unordered or less than or equal</li>
4452 <li><tt>une</tt>: unordered or not equal</li>
4453 <li><tt>uno</tt>: unordered (either nans)</li>
4454 <li><tt>true</tt>: no comparison, always returns true</li>
4455</ol>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4456
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4457<p><i>Ordered</i> means that neither operand is a QNAN while
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4458 <i>unordered</i> means that either operand may be a QNAN.</p>
4459
4460<p>Each of <tt>val1</tt> and <tt>val2</tt> arguments must be either
4461 a <a href="#t_floating">floating point</a> type or
4462 a <a href="#t_vector">vector</a> of floating point type. They must have
4463 identical types.</p>
4464
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4465<h5>Semantics:</h5>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]4466<p>The '<tt>fcmp</tt>' instruction compares <tt>op1</tt> and <tt>op2</tt>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4467 according to the condition code given as <tt>cond</tt>. If the operands are
4468 vectors, then the vectors are compared element by element. Each comparison
4469 performed always yields an <a href="#t_primitive">i1</a> result, as
4470 follows:</p>
4471
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4472<ol>
4473 <li><tt>false</tt>: always yields <tt>false</tt>, regardless of operands.</li>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4474
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4475 <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]4476 <tt>op1</tt> is equal to <tt>op2</tt>.</li>
4477
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4478 <li><tt>ogt</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4479 <tt>op1</tt> is greather than <tt>op2</tt>.</li>
4480
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4481 <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]4482 <tt>op1</tt> is greater than or equal to <tt>op2</tt>.</li>
4483
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4484 <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]4485 <tt>op1</tt> is less than <tt>op2</tt>.</li>
4486
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4487 <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]4488 <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
4489
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4490 <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]4491 <tt>op1</tt> is not equal to <tt>op2</tt>.</li>
4492
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4493 <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]4494
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4495 <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]4496 <tt>op1</tt> is equal to <tt>op2</tt>.</li>
4497
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4498 <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]4499 <tt>op1</tt> is greater than <tt>op2</tt>.</li>
4500
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4501 <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]4502 <tt>op1</tt> is greater than or equal to <tt>op2</tt>.</li>
4503
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4504 <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]4505 <tt>op1</tt> is less than <tt>op2</tt>.</li>
4506
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4507 <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]4508 <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
4509
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4510 <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]4511 <tt>op1</tt> is not equal to <tt>op2</tt>.</li>
4512
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4513 <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]4514
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4515 <li><tt>true</tt>: always yields <tt>true</tt>, regardless of operands.</li>
4516</ol>
4517
4518<h5>Example:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4519<pre>
4520 &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]4521 &lt;result&gt; = fcmp one float 4.0, 5.0 <i>; yields: result=true</i>
4522 &lt;result&gt; = fcmp olt float 4.0, 5.0 <i>; yields: result=true</i>
4523 &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]4524</pre>
Dan Gohmana53eb382009-01-22 01:39:38 +0000[diff] [blame]4525
4526<p>Note that the code generator does not yet support vector types with
4527 the <tt>fcmp</tt> instruction.</p>
4528
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4529</div>
4530
4531<!-- _______________________________________________________________________ -->
Nate Begeman646fa482008-05-12 19:01:56 +0000[diff] [blame]4532<div class="doc_subsubsection">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4533 <a name="i_phi">'<tt>phi</tt>' Instruction</a>
4534</div>
4535
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4536<div class="doc_text">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4537
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4538<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4539<pre>
4540 &lt;result&gt; = phi &lt;ty&gt; [ &lt;val0&gt;, &lt;label0&gt;], ...
4541</pre>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4542
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4543<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4544<p>The '<tt>phi</tt>' instruction is used to implement the &#966; node in the
4545 SSA graph representing the function.</p>
4546
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4547<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4548<p>The type of the incoming values is specified with the first type field. After
4549 this, the '<tt>phi</tt>' instruction takes a list of pairs as arguments, with
4550 one pair for each predecessor basic block of the current block. Only values
4551 of <a href="#t_firstclass">first class</a> type may be used as the value
4552 arguments to the PHI node. Only labels may be used as the label
4553 arguments.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4554
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4555<p>There must be no non-phi instructions between the start of a basic block and
4556 the PHI instructions: i.e. PHI instructions must be first in a basic
4557 block.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4558
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4559<p>For the purposes of the SSA form, the use of each incoming value is deemed to
4560 occur on the edge from the corresponding predecessor block to the current
4561 block (but after any definition of an '<tt>invoke</tt>' instruction's return
4562 value on the same edge).</p>
Jay Foad8e2fd2c2009-06-03 10:20:10 +0000[diff] [blame]4563
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4564<h5>Semantics:</h5>
4565<p>At runtime, the '<tt>phi</tt>' instruction logically takes on the value
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4566 specified by the pair corresponding to the predecessor basic block that
4567 executed just prior to the current block.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4568
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4569<h5>Example:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4570<pre>
4571Loop: ; Infinite loop that counts from 0 on up...
4572 %indvar = phi i32 [ 0, %LoopHeader ], [ %nextindvar, %Loop ]
4573 %nextindvar = add i32 %indvar, 1
4574 br label %Loop
4575</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4576
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4577</div>
4578
4579<!-- _______________________________________________________________________ -->
4580<div class="doc_subsubsection">
4581 <a name="i_select">'<tt>select</tt>' Instruction</a>
4582</div>
4583
4584<div class="doc_text">
4585
4586<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4587<pre>
Dan Gohmanb60ca3c2008-09-09 01:02:47 +0000[diff] [blame]4588 &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>
4589
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]4590 <i>selty</i> is either i1 or {&lt;N x i1&gt;}
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4591</pre>
4592
4593<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4594<p>The '<tt>select</tt>' instruction is used to choose one value based on a
4595 condition, without branching.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4596
4597
4598<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4599<p>The '<tt>select</tt>' instruction requires an 'i1' value or a vector of 'i1'
4600 values indicating the condition, and two values of the
4601 same <a href="#t_firstclass">first class</a> type. If the val1/val2 are
4602 vectors and the condition is a scalar, then entire vectors are selected, not
4603 individual elements.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4604
4605<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4606<p>If the condition is an i1 and it evaluates to 1, the instruction returns the
4607 first value argument; otherwise, it returns the second value argument.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4608
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4609<p>If the condition is a vector of i1, then the value arguments must be vectors
4610 of the same size, and the selection is done element by element.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4611
4612<h5>Example:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4613<pre>
4614 %X = select i1 true, i8 17, i8 42 <i>; yields i8:17</i>
4615</pre>
Dan Gohmana53eb382009-01-22 01:39:38 +0000[diff] [blame]4616
4617<p>Note that the code generator does not yet support conditions
4618 with vector type.</p>
4619
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4620</div>
4621
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4622<!-- _______________________________________________________________________ -->
4623<div class="doc_subsubsection">
4624 <a name="i_call">'<tt>call</tt>' Instruction</a>
4625</div>
4626
4627<div class="doc_text">
4628
4629<h5>Syntax:</h5>
4630<pre>
Devang Pateld0bfcc72008-10-07 17:48:33 +0000[diff] [blame]4631 &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]4632</pre>
4633
4634<h5>Overview:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4635<p>The '<tt>call</tt>' instruction represents a simple function call.</p>
4636
4637<h5>Arguments:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4638<p>This instruction requires several arguments:</p>
4639
4640<ol>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4641 <li>The optional "tail" marker indicates whether the callee function accesses
4642 any allocas or varargs in the caller. If the "tail" marker is present,
4643 the function call is eligible for tail call optimization. Note that calls
4644 may be marked "tail" even if they do not occur before
4645 a <a href="#i_ret"><tt>ret</tt></a> instruction.</li>
Devang Patelac2fc272008-10-06 18:50:38 +0000[diff] [blame]4646
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4647 <li>The optional "cconv" marker indicates which <a href="#callingconv">calling
4648 convention</a> the call should use. If none is specified, the call
4649 defaults to using C calling conventions.</li>
Devang Patelac2fc272008-10-06 18:50:38 +0000[diff] [blame]4650
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4651 <li>The optional <a href="#paramattrs">Parameter Attributes</a> list for
4652 return values. Only '<tt>zeroext</tt>', '<tt>signext</tt>', and
4653 '<tt>inreg</tt>' attributes are valid here.</li>
4654
4655 <li>'<tt>ty</tt>': the type of the call instruction itself which is also the
4656 type of the return value. Functions that return no value are marked
4657 <tt><a href="#t_void">void</a></tt>.</li>
4658
4659 <li>'<tt>fnty</tt>': shall be the signature of the pointer to function value
4660 being invoked. The argument types must match the types implied by this
4661 signature. This type can be omitted if the function is not varargs and if
4662 the function type does not return a pointer to a function.</li>
4663
4664 <li>'<tt>fnptrval</tt>': An LLVM value containing a pointer to a function to
4665 be invoked. In most cases, this is a direct function invocation, but
4666 indirect <tt>call</tt>s are just as possible, calling an arbitrary pointer
4667 to function value.</li>
4668
4669 <li>'<tt>function args</tt>': argument list whose types match the function
4670 signature argument types. All arguments must be of
4671 <a href="#t_firstclass">first class</a> type. If the function signature
4672 indicates the function accepts a variable number of arguments, the extra
4673 arguments can be specified.</li>
4674
4675 <li>The optional <a href="#fnattrs">function attributes</a> list. Only
4676 '<tt>noreturn</tt>', '<tt>nounwind</tt>', '<tt>readonly</tt>' and
4677 '<tt>readnone</tt>' attributes are valid here.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4678</ol>
4679
4680<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4681<p>The '<tt>call</tt>' instruction is used to cause control flow to transfer to
4682 a specified function, with its incoming arguments bound to the specified
4683 values. Upon a '<tt><a href="#i_ret">ret</a></tt>' instruction in the called
4684 function, control flow continues with the instruction after the function
4685 call, and the return value of the function is bound to the result
4686 argument.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4687
4688<h5>Example:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4689<pre>
Nick Lewycky93082fc2007-09-08 13:57:50 +0000[diff] [blame]4690 %retval = call i32 @test(i32 %argc)
Chris Lattner5e893ef2008-03-21 17:24:17 +0000[diff] [blame]4691 call i32 (i8 *, ...)* @printf(i8 * %msg, i32 12, i8 42) <i>; yields i32</i>
4692 %X = tail call i32 @foo() <i>; yields i32</i>
4693 %Y = tail call <a href="#callingconv">fastcc</a> i32 @foo() <i>; yields i32</i>
4694 call void %foo(i8 97 signext)
Devang Patela3cc5372008-03-10 20:49:15 +0000[diff] [blame]4695
4696 %struct.A = type { i32, i8 }
Devang Patelac2fc272008-10-06 18:50:38 +0000[diff] [blame]4697 %r = call %struct.A @foo() <i>; yields { 32, i8 }</i>
Dan Gohman3e700032008-10-04 19:00:07 +0000[diff] [blame]4698 %gr = extractvalue %struct.A %r, 0 <i>; yields i32</i>
4699 %gr1 = extractvalue %struct.A %r, 1 <i>; yields i8</i>
Chris Lattnerac454b32008-10-08 06:26:11 +0000[diff] [blame]4700 %Z = call void @foo() noreturn <i>; indicates that %foo never returns normally</i>
Matthijs Kooijman2c4e05a2008-10-07 10:03:45 +0000[diff] [blame]4701 %ZZ = call zeroext i32 @bar() <i>; Return value is %zero extended</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4702</pre>
4703
4704</div>
4705
4706<!-- _______________________________________________________________________ -->
4707<div class="doc_subsubsection">
4708 <a name="i_va_arg">'<tt>va_arg</tt>' Instruction</a>
4709</div>
4710
4711<div class="doc_text">
4712
4713<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4714<pre>
4715 &lt;resultval&gt; = va_arg &lt;va_list*&gt; &lt;arglist&gt;, &lt;argty&gt;
4716</pre>
4717
4718<h5>Overview:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4719<p>The '<tt>va_arg</tt>' instruction is used to access arguments passed through
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4720 the "variable argument" area of a function call. It is used to implement the
4721 <tt>va_arg</tt> macro in C.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4722
4723<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4724<p>This instruction takes a <tt>va_list*</tt> value and the type of the
4725 argument. It returns a value of the specified argument type and increments
4726 the <tt>va_list</tt> to point to the next argument. The actual type
4727 of <tt>va_list</tt> is target specific.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4728
4729<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4730<p>The '<tt>va_arg</tt>' instruction loads an argument of the specified type
4731 from the specified <tt>va_list</tt> and causes the <tt>va_list</tt> to point
4732 to the next argument. For more information, see the variable argument
4733 handling <a href="#int_varargs">Intrinsic Functions</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4734
4735<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]4736 take a variable number of arguments, for example, the <tt>vfprintf</tt>
4737 function.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4738
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4739<p><tt>va_arg</tt> is an LLVM instruction instead of
4740 an <a href="#intrinsics">intrinsic function</a> because it takes a type as an
4741 argument.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4742
4743<h5>Example:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4744<p>See the <a href="#int_varargs">variable argument processing</a> section.</p>
4745
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4746<p>Note that the code generator does not yet fully support va_arg on many
4747 targets. Also, it does not currently support va_arg with aggregate types on
4748 any target.</p>
Dan Gohman60967192009-01-12 23:12:39 +0000[diff] [blame]4749
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4750</div>
4751
4752<!-- *********************************************************************** -->
4753<div class="doc_section"> <a name="intrinsics">Intrinsic Functions</a> </div>
4754<!-- *********************************************************************** -->
4755
4756<div class="doc_text">
4757
4758<p>LLVM supports the notion of an "intrinsic function". These functions have
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4759 well known names and semantics and are required to follow certain
4760 restrictions. Overall, these intrinsics represent an extension mechanism for
4761 the LLVM language that does not require changing all of the transformations
4762 in LLVM when adding to the language (or the bitcode reader/writer, the
4763 parser, etc...).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4764
4765<p>Intrinsic function names must all start with an "<tt>llvm.</tt>" prefix. This
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4766 prefix is reserved in LLVM for intrinsic names; thus, function names may not
4767 begin with this prefix. Intrinsic functions must always be external
4768 functions: you cannot define the body of intrinsic functions. Intrinsic
4769 functions may only be used in call or invoke instructions: it is illegal to
4770 take the address of an intrinsic function. Additionally, because intrinsic
4771 functions are part of the LLVM language, it is required if any are added that
4772 they be documented here.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4773
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4774<p>Some intrinsic functions can be overloaded, i.e., the intrinsic represents a
4775 family of functions that perform the same operation but on different data
4776 types. Because LLVM can represent over 8 million different integer types,
4777 overloading is used commonly to allow an intrinsic function to operate on any
4778 integer type. One or more of the argument types or the result type can be
4779 overloaded to accept any integer type. Argument types may also be defined as
4780 exactly matching a previous argument's type or the result type. This allows
4781 an intrinsic function which accepts multiple arguments, but needs all of them
4782 to be of the same type, to only be overloaded with respect to a single
4783 argument or the result.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4784
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4785<p>Overloaded intrinsics will have the names of its overloaded argument types
4786 encoded into its function name, each preceded by a period. Only those types
4787 which are overloaded result in a name suffix. Arguments whose type is matched
4788 against another type do not. For example, the <tt>llvm.ctpop</tt> function
4789 can take an integer of any width and returns an integer of exactly the same
4790 integer width. This leads to a family of functions such as
4791 <tt>i8 @llvm.ctpop.i8(i8 %val)</tt> and <tt>i29 @llvm.ctpop.i29(i29
4792 %val)</tt>. Only one type, the return type, is overloaded, and only one type
4793 suffix is required. Because the argument's type is matched against the return
4794 type, it does not require its own name suffix.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4795
4796<p>To learn how to add an intrinsic function, please see the
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4797 <a href="ExtendingLLVM.html">Extending LLVM Guide</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4798
4799</div>
4800
4801<!-- ======================================================================= -->
4802<div class="doc_subsection">
4803 <a name="int_varargs">Variable Argument Handling Intrinsics</a>
4804</div>
4805
4806<div class="doc_text">
4807
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4808<p>Variable argument support is defined in LLVM with
4809 the <a href="#i_va_arg"><tt>va_arg</tt></a> instruction and these three
4810 intrinsic functions. These functions are related to the similarly named
4811 macros defined in the <tt>&lt;stdarg.h&gt;</tt> header file.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4812
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4813<p>All of these functions operate on arguments that use a target-specific value
4814 type "<tt>va_list</tt>". The LLVM assembly language reference manual does
4815 not define what this type is, so all transformations should be prepared to
4816 handle these functions regardless of the type used.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4817
4818<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]4819 instruction and the variable argument handling intrinsic functions are
4820 used.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4821
4822<div class="doc_code">
4823<pre>
4824define i32 @test(i32 %X, ...) {
4825 ; Initialize variable argument processing
4826 %ap = alloca i8*
4827 %ap2 = bitcast i8** %ap to i8*
4828 call void @llvm.va_start(i8* %ap2)
4829
4830 ; Read a single integer argument
4831 %tmp = va_arg i8** %ap, i32
4832
4833 ; Demonstrate usage of llvm.va_copy and llvm.va_end
4834 %aq = alloca i8*
4835 %aq2 = bitcast i8** %aq to i8*
4836 call void @llvm.va_copy(i8* %aq2, i8* %ap2)
4837 call void @llvm.va_end(i8* %aq2)
4838
4839 ; Stop processing of arguments.
4840 call void @llvm.va_end(i8* %ap2)
4841 ret i32 %tmp
4842}
4843
4844declare void @llvm.va_start(i8*)
4845declare void @llvm.va_copy(i8*, i8*)
4846declare void @llvm.va_end(i8*)
4847</pre>
4848</div>
4849
4850</div>
4851
4852<!-- _______________________________________________________________________ -->
4853<div class="doc_subsubsection">
4854 <a name="int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a>
4855</div>
4856
4857
4858<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4859
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4860<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4861<pre>
4862 declare void %llvm.va_start(i8* &lt;arglist&gt;)
4863</pre>
4864
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4865<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4866<p>The '<tt>llvm.va_start</tt>' intrinsic initializes <tt>*&lt;arglist&gt;</tt>
4867 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]4868
4869<h5>Arguments:</h5>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]4870<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]4871
4872<h5>Semantics:</h5>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]4873<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]4874 macro available in C. In a target-dependent way, it initializes
4875 the <tt>va_list</tt> element to which the argument points, so that the next
4876 call to <tt>va_arg</tt> will produce the first variable argument passed to
4877 the function. Unlike the C <tt>va_start</tt> macro, this intrinsic does not
4878 need to know the last argument of the function as the compiler can figure
4879 that out.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4880
4881</div>
4882
4883<!-- _______________________________________________________________________ -->
4884<div class="doc_subsubsection">
4885 <a name="int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a>
4886</div>
4887
4888<div class="doc_text">
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4889
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4890<h5>Syntax:</h5>
4891<pre>
4892 declare void @llvm.va_end(i8* &lt;arglist&gt;)
4893</pre>
4894
4895<h5>Overview:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4896<p>The '<tt>llvm.va_end</tt>' intrinsic destroys <tt>*&lt;arglist&gt;</tt>,
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4897 which has been initialized previously
4898 with <tt><a href="#int_va_start">llvm.va_start</a></tt>
4899 or <tt><a href="#i_va_copy">llvm.va_copy</a></tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4900
4901<h5>Arguments:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4902<p>The argument is a pointer to a <tt>va_list</tt> to destroy.</p>
4903
4904<h5>Semantics:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4905<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]4906 macro available in C. In a target-dependent way, it destroys
4907 the <tt>va_list</tt> element to which the argument points. Calls
4908 to <a href="#int_va_start"><tt>llvm.va_start</tt></a>
4909 and <a href="#int_va_copy"> <tt>llvm.va_copy</tt></a> must be matched exactly
4910 with calls to <tt>llvm.va_end</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4911
4912</div>
4913
4914<!-- _______________________________________________________________________ -->
4915<div class="doc_subsubsection">
4916 <a name="int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a>
4917</div>
4918
4919<div class="doc_text">
4920
4921<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4922<pre>
4923 declare void @llvm.va_copy(i8* &lt;destarglist&gt;, i8* &lt;srcarglist&gt;)
4924</pre>
4925
4926<h5>Overview:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4927<p>The '<tt>llvm.va_copy</tt>' intrinsic copies the current argument position
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4928 from the source argument list to the destination argument list.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4929
4930<h5>Arguments:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4931<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]4932 The second argument is a pointer to a <tt>va_list</tt> element to copy
4933 from.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4934
4935<h5>Semantics:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4936<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]4937 macro available in C. In a target-dependent way, it copies the
4938 source <tt>va_list</tt> element into the destination <tt>va_list</tt>
4939 element. This intrinsic is necessary because
4940 the <tt><a href="#int_va_start"> llvm.va_start</a></tt> intrinsic may be
4941 arbitrarily complex and require, for example, memory allocation.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4942
4943</div>
4944
4945<!-- ======================================================================= -->
4946<div class="doc_subsection">
4947 <a name="int_gc">Accurate Garbage Collection Intrinsics</a>
4948</div>
4949
4950<div class="doc_text">
4951
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4952<p>LLVM support for <a href="GarbageCollection.html">Accurate Garbage
Chris Lattner96451482008-08-05 18:29:16 +0000[diff] [blame]4953Collection</a> (GC) requires the implementation and generation of these
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4954intrinsics. These intrinsics allow identification of <a href="#int_gcroot">GC
4955roots on the stack</a>, as well as garbage collector implementations that
4956require <a href="#int_gcread">read</a> and <a href="#int_gcwrite">write</a>
4957barriers. Front-ends for type-safe garbage collected languages should generate
4958these intrinsics to make use of the LLVM garbage collectors. For more details,
4959see <a href="GarbageCollection.html">Accurate Garbage Collection with
4960LLVM</a>.</p>
Christopher Lambcfe00962007-12-17 01:00:21 +0000[diff] [blame]4961
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4962<p>The garbage collection intrinsics only operate on objects in the generic
4963 address space (address space zero).</p>
Christopher Lambcfe00962007-12-17 01:00:21 +0000[diff] [blame]4964
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4965</div>
4966
4967<!-- _______________________________________________________________________ -->
4968<div class="doc_subsubsection">
4969 <a name="int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a>
4970</div>
4971
4972<div class="doc_text">
4973
4974<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4975<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]4976 declare void @llvm.gcroot(i8** %ptrloc, i8* %metadata)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4977</pre>
4978
4979<h5>Overview:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4980<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]4981 the code generator, and allows some metadata to be associated with it.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4982
4983<h5>Arguments:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4984<p>The first argument specifies the address of a stack object that contains the
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]4985 root pointer. The second pointer (which must be either a constant or a
4986 global value address) contains the meta-data to be associated with the
4987 root.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4988
4989<h5>Semantics:</h5>
Chris Lattnera7d94ba2008-04-24 05:59:56 +0000[diff] [blame]4990<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]4991 location. At compile-time, the code generator generates information to allow
4992 the runtime to find the pointer at GC safe points. The '<tt>llvm.gcroot</tt>'
4993 intrinsic may only be used in a function which <a href="#gc">specifies a GC
4994 algorithm</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4995
4996</div>
4997
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4998<!-- _______________________________________________________________________ -->
4999<div class="doc_subsubsection">
5000 <a name="int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a>
5001</div>
5002
5003<div class="doc_text">
5004
5005<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5006<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]5007 declare i8* @llvm.gcread(i8* %ObjPtr, i8** %Ptr)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5008</pre>
5009
5010<h5>Overview:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5011<p>The '<tt>llvm.gcread</tt>' intrinsic identifies reads of references from heap
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5012 locations, allowing garbage collector implementations that require read
5013 barriers.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5014
5015<h5>Arguments:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5016<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]5017 allocated from the garbage collector. The first object is a pointer to the
5018 start of the referenced object, if needed by the language runtime (otherwise
5019 null).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5020
5021<h5>Semantics:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5022<p>The '<tt>llvm.gcread</tt>' intrinsic has the same semantics as a load
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5023 instruction, but may be replaced with substantially more complex code by the
5024 garbage collector runtime, as needed. The '<tt>llvm.gcread</tt>' intrinsic
5025 may only be used in a function which <a href="#gc">specifies a GC
5026 algorithm</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5027
5028</div>
5029
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5030<!-- _______________________________________________________________________ -->
5031<div class="doc_subsubsection">
5032 <a name="int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a>
5033</div>
5034
5035<div class="doc_text">
5036
5037<h5>Syntax:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5038<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]5039 declare void @llvm.gcwrite(i8* %P1, i8* %Obj, i8** %P2)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5040</pre>
5041
5042<h5>Overview:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5043<p>The '<tt>llvm.gcwrite</tt>' intrinsic identifies writes of references to heap
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5044 locations, allowing garbage collector implementations that require write
5045 barriers (such as generational or reference counting collectors).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5046
5047<h5>Arguments:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5048<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]5049 object to store it to, and the third is the address of the field of Obj to
5050 store to. If the runtime does not require a pointer to the object, Obj may
5051 be null.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5052
5053<h5>Semantics:</h5>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5054<p>The '<tt>llvm.gcwrite</tt>' intrinsic has the same semantics as a store
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5055 instruction, but may be replaced with substantially more complex code by the
5056 garbage collector runtime, as needed. The '<tt>llvm.gcwrite</tt>' intrinsic
5057 may only be used in a function which <a href="#gc">specifies a GC
5058 algorithm</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5059
5060</div>
5061
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5062<!-- ======================================================================= -->
5063<div class="doc_subsection">
5064 <a name="int_codegen">Code Generator Intrinsics</a>
5065</div>
5066
5067<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5068
5069<p>These intrinsics are provided by LLVM to expose special features that may
5070 only be implemented with code generator support.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5071
5072</div>
5073
5074<!-- _______________________________________________________________________ -->
5075<div class="doc_subsubsection">
5076 <a name="int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a>
5077</div>
5078
5079<div class="doc_text">
5080
5081<h5>Syntax:</h5>
5082<pre>
5083 declare i8 *@llvm.returnaddress(i32 &lt;level&gt;)
5084</pre>
5085
5086<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5087<p>The '<tt>llvm.returnaddress</tt>' intrinsic attempts to compute a
5088 target-specific value indicating the return address of the current function
5089 or one of its callers.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5090
5091<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5092<p>The argument to this intrinsic indicates which function to return the address
5093 for. Zero indicates the calling function, one indicates its caller, etc.
5094 The argument is <b>required</b> to be a constant integer value.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5095
5096<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5097<p>The '<tt>llvm.returnaddress</tt>' intrinsic either returns a pointer
5098 indicating the return address of the specified call frame, or zero if it
5099 cannot be identified. The value returned by this intrinsic is likely to be
5100 incorrect or 0 for arguments other than zero, so it should only be used for
5101 debugging purposes.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5102
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5103<p>Note that calling this intrinsic does not prevent function inlining or other
5104 aggressive transformations, so the value returned may not be that of the
5105 obvious source-language caller.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5106
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5107</div>
5108
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5109<!-- _______________________________________________________________________ -->
5110<div class="doc_subsubsection">
5111 <a name="int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a>
5112</div>
5113
5114<div class="doc_text">
5115
5116<h5>Syntax:</h5>
5117<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]5118 declare i8 *@llvm.frameaddress(i32 &lt;level&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5119</pre>
5120
5121<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5122<p>The '<tt>llvm.frameaddress</tt>' intrinsic attempts to return the
5123 target-specific frame pointer value for the specified stack frame.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5124
5125<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5126<p>The argument to this intrinsic indicates which function to return the frame
5127 pointer for. Zero indicates the calling function, one indicates its caller,
5128 etc. The argument is <b>required</b> to be a constant integer value.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5129
5130<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5131<p>The '<tt>llvm.frameaddress</tt>' intrinsic either returns a pointer
5132 indicating the frame address of the specified call frame, or zero if it
5133 cannot be identified. The value returned by this intrinsic is likely to be
5134 incorrect or 0 for arguments other than zero, so it should only be used for
5135 debugging purposes.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5136
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5137<p>Note that calling this intrinsic does not prevent function inlining or other
5138 aggressive transformations, so the value returned may not be that of the
5139 obvious source-language caller.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5140
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5141</div>
5142
5143<!-- _______________________________________________________________________ -->
5144<div class="doc_subsubsection">
5145 <a name="int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a>
5146</div>
5147
5148<div class="doc_text">
5149
5150<h5>Syntax:</h5>
5151<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]5152 declare i8 *@llvm.stacksave()
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5153</pre>
5154
5155<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5156<p>The '<tt>llvm.stacksave</tt>' intrinsic is used to remember the current state
5157 of the function stack, for use
5158 with <a href="#int_stackrestore"> <tt>llvm.stackrestore</tt></a>. This is
5159 useful for implementing language features like scoped automatic variable
5160 sized arrays in C99.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5161
5162<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5163<p>This intrinsic returns a opaque pointer value that can be passed
5164 to <a href="#int_stackrestore"><tt>llvm.stackrestore</tt></a>. When
5165 an <tt>llvm.stackrestore</tt> intrinsic is executed with a value saved
5166 from <tt>llvm.stacksave</tt>, it effectively restores the state of the stack
5167 to the state it was in when the <tt>llvm.stacksave</tt> intrinsic executed.
5168 In practice, this pops any <a href="#i_alloca">alloca</a> blocks from the
5169 stack that were allocated after the <tt>llvm.stacksave</tt> was executed.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5170
5171</div>
5172
5173<!-- _______________________________________________________________________ -->
5174<div class="doc_subsubsection">
5175 <a name="int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a>
5176</div>
5177
5178<div class="doc_text">
5179
5180<h5>Syntax:</h5>
5181<pre>
5182 declare void @llvm.stackrestore(i8 * %ptr)
5183</pre>
5184
5185<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5186<p>The '<tt>llvm.stackrestore</tt>' intrinsic is used to restore the state of
5187 the function stack to the state it was in when the
5188 corresponding <a href="#int_stacksave"><tt>llvm.stacksave</tt></a> intrinsic
5189 executed. This is useful for implementing language features like scoped
5190 automatic variable sized arrays in C99.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5191
5192<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5193<p>See the description
5194 for <a href="#int_stacksave"><tt>llvm.stacksave</tt></a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5195
5196</div>
5197
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5198<!-- _______________________________________________________________________ -->
5199<div class="doc_subsubsection">
5200 <a name="int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a>
5201</div>
5202
5203<div class="doc_text">
5204
5205<h5>Syntax:</h5>
5206<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]5207 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]5208</pre>
5209
5210<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5211<p>The '<tt>llvm.prefetch</tt>' intrinsic is a hint to the code generator to
5212 insert a prefetch instruction if supported; otherwise, it is a noop.
5213 Prefetches have no effect on the behavior of the program but can change its
5214 performance characteristics.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5215
5216<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5217<p><tt>address</tt> is the address to be prefetched, <tt>rw</tt> is the
5218 specifier determining if the fetch should be for a read (0) or write (1),
5219 and <tt>locality</tt> is a temporal locality specifier ranging from (0) - no
5220 locality, to (3) - extremely local keep in cache. The <tt>rw</tt>
5221 and <tt>locality</tt> arguments must be constant integers.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5222
5223<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5224<p>This intrinsic does not modify the behavior of the program. In particular,
5225 prefetches cannot trap and do not produce a value. On targets that support
5226 this intrinsic, the prefetch can provide hints to the processor cache for
5227 better performance.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5228
5229</div>
5230
5231<!-- _______________________________________________________________________ -->
5232<div class="doc_subsubsection">
5233 <a name="int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a>
5234</div>
5235
5236<div class="doc_text">
5237
5238<h5>Syntax:</h5>
5239<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]5240 declare void @llvm.pcmarker(i32 &lt;id&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5241</pre>
5242
5243<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5244<p>The '<tt>llvm.pcmarker</tt>' intrinsic is a method to export a Program
5245 Counter (PC) in a region of code to simulators and other tools. The method
5246 is target specific, but it is expected that the marker will use exported
5247 symbols to transmit the PC of the marker. The marker makes no guarantees
5248 that it will remain with any specific instruction after optimizations. It is
5249 possible that the presence of a marker will inhibit optimizations. The
5250 intended use is to be inserted after optimizations to allow correlations of
5251 simulation runs.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5252
5253<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5254<p><tt>id</tt> is a numerical id identifying the marker.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5255
5256<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5257<p>This intrinsic does not modify the behavior of the program. Backends that do
5258 not support this intrinisic may ignore it.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5259
5260</div>
5261
5262<!-- _______________________________________________________________________ -->
5263<div class="doc_subsubsection">
5264 <a name="int_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a>
5265</div>
5266
5267<div class="doc_text">
5268
5269<h5>Syntax:</h5>
5270<pre>
5271 declare i64 @llvm.readcyclecounter( )
5272</pre>
5273
5274<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5275<p>The '<tt>llvm.readcyclecounter</tt>' intrinsic provides access to the cycle
5276 counter register (or similar low latency, high accuracy clocks) on those
5277 targets that support it. On X86, it should map to RDTSC. On Alpha, it
5278 should map to RPCC. As the backing counters overflow quickly (on the order
5279 of 9 seconds on alpha), this should only be used for small timings.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5280
5281<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5282<p>When directly supported, reading the cycle counter should not modify any
5283 memory. Implementations are allowed to either return a application specific
5284 value or a system wide value. On backends without support, this is lowered
5285 to a constant 0.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5286
5287</div>
5288
5289<!-- ======================================================================= -->
5290<div class="doc_subsection">
5291 <a name="int_libc">Standard C Library Intrinsics</a>
5292</div>
5293
5294<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5295
5296<p>LLVM provides intrinsics for a few important standard C library functions.
5297 These intrinsics allow source-language front-ends to pass information about
5298 the alignment of the pointer arguments to the code generator, providing
5299 opportunity for more efficient code generation.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5300
5301</div>
5302
5303<!-- _______________________________________________________________________ -->
5304<div class="doc_subsubsection">
5305 <a name="int_memcpy">'<tt>llvm.memcpy</tt>' Intrinsic</a>
5306</div>
5307
5308<div class="doc_text">
5309
5310<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5311<p>This is an overloaded intrinsic. You can use <tt>llvm.memcpy</tt> on any
5312 integer bit width. Not all targets support all bit widths however.</p>
5313
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5314<pre>
Chris Lattner82c2e432008-11-21 16:42:48 +0000[diff] [blame]5315 declare void @llvm.memcpy.i8(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5316 i8 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattner82c2e432008-11-21 16:42:48 +0000[diff] [blame]5317 declare void @llvm.memcpy.i16(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
5318 i16 &lt;len&gt;, i32 &lt;align&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5319 declare void @llvm.memcpy.i32(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
5320 i32 &lt;len&gt;, i32 &lt;align&gt;)
5321 declare void @llvm.memcpy.i64(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
5322 i64 &lt;len&gt;, i32 &lt;align&gt;)
5323</pre>
5324
5325<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5326<p>The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the
5327 source location to the destination location.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5328
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5329<p>Note that, unlike the standard libc function, the <tt>llvm.memcpy.*</tt>
5330 intrinsics do not return a value, and takes an extra alignment argument.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5331
5332<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5333<p>The first argument is a pointer to the destination, the second is a pointer
5334 to the source. The third argument is an integer argument specifying the
5335 number of bytes to copy, and the fourth argument is the alignment of the
5336 source and destination locations.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5337
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5338<p>If the call to this intrinisic has an alignment value that is not 0 or 1,
5339 then the caller guarantees that both the source and destination pointers are
5340 aligned to that boundary.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5341
5342<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5343<p>The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the
5344 source location to the destination location, which are not allowed to
5345 overlap. It copies "len" bytes of memory over. If the argument is known to
5346 be aligned to some boundary, this can be specified as the fourth argument,
5347 otherwise it should be set to 0 or 1.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5348
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5349</div>
5350
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5351<!-- _______________________________________________________________________ -->
5352<div class="doc_subsubsection">
5353 <a name="int_memmove">'<tt>llvm.memmove</tt>' Intrinsic</a>
5354</div>
5355
5356<div class="doc_text">
5357
5358<h5>Syntax:</h5>
Chris Lattner82c2e432008-11-21 16:42:48 +0000[diff] [blame]5359<p>This is an overloaded intrinsic. You can use llvm.memmove on any integer bit
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5360 width. Not all targets support all bit widths however.</p>
5361
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5362<pre>
Chris Lattner82c2e432008-11-21 16:42:48 +0000[diff] [blame]5363 declare void @llvm.memmove.i8(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5364 i8 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattner82c2e432008-11-21 16:42:48 +0000[diff] [blame]5365 declare void @llvm.memmove.i16(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
5366 i16 &lt;len&gt;, i32 &lt;align&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5367 declare void @llvm.memmove.i32(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
5368 i32 &lt;len&gt;, i32 &lt;align&gt;)
5369 declare void @llvm.memmove.i64(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
5370 i64 &lt;len&gt;, i32 &lt;align&gt;)
5371</pre>
5372
5373<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5374<p>The '<tt>llvm.memmove.*</tt>' intrinsics move a block of memory from the
5375 source location to the destination location. It is similar to the
5376 '<tt>llvm.memcpy</tt>' intrinsic but allows the two memory locations to
5377 overlap.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5378
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5379<p>Note that, unlike the standard libc function, the <tt>llvm.memmove.*</tt>
5380 intrinsics do not return a value, and takes an extra alignment argument.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5381
5382<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5383<p>The first argument is a pointer to the destination, the second is a pointer
5384 to the source. The third argument is an integer argument specifying the
5385 number of bytes to copy, and the fourth argument is the alignment of the
5386 source and destination locations.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5387
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5388<p>If the call to this intrinisic has an alignment value that is not 0 or 1,
5389 then the caller guarantees that the source and destination pointers are
5390 aligned to that boundary.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5391
5392<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5393<p>The '<tt>llvm.memmove.*</tt>' intrinsics copy a block of memory from the
5394 source location to the destination location, which may overlap. It copies
5395 "len" bytes of memory over. If the argument is known to be aligned to some
5396 boundary, this can be specified as the fourth argument, otherwise it should
5397 be set to 0 or 1.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5398
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5399</div>
5400
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5401<!-- _______________________________________________________________________ -->
5402<div class="doc_subsubsection">
5403 <a name="int_memset">'<tt>llvm.memset.*</tt>' Intrinsics</a>
5404</div>
5405
5406<div class="doc_text">
5407
5408<h5>Syntax:</h5>
Chris Lattner82c2e432008-11-21 16:42:48 +0000[diff] [blame]5409<p>This is an overloaded intrinsic. You can use llvm.memset on any integer bit
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5410 width. Not all targets support all bit widths however.</p>
5411
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5412<pre>
Chris Lattner82c2e432008-11-21 16:42:48 +0000[diff] [blame]5413 declare void @llvm.memset.i8(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5414 i8 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattner82c2e432008-11-21 16:42:48 +0000[diff] [blame]5415 declare void @llvm.memset.i16(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
5416 i16 &lt;len&gt;, i32 &lt;align&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5417 declare void @llvm.memset.i32(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
5418 i32 &lt;len&gt;, i32 &lt;align&gt;)
5419 declare void @llvm.memset.i64(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
5420 i64 &lt;len&gt;, i32 &lt;align&gt;)
5421</pre>
5422
5423<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5424<p>The '<tt>llvm.memset.*</tt>' intrinsics fill a block of memory with a
5425 particular byte value.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5426
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5427<p>Note that, unlike the standard libc function, the <tt>llvm.memset</tt>
5428 intrinsic does not return a value, and takes an extra alignment argument.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5429
5430<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5431<p>The first argument is a pointer to the destination to fill, the second is the
5432 byte value to fill it with, the third argument is an integer argument
5433 specifying the number of bytes to fill, and the fourth argument is the known
5434 alignment of destination location.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5435
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5436<p>If the call to this intrinisic has an alignment value that is not 0 or 1,
5437 then the caller guarantees that the destination pointer is aligned to that
5438 boundary.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5439
5440<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5441<p>The '<tt>llvm.memset.*</tt>' intrinsics fill "len" bytes of memory starting
5442 at the destination location. If the argument is known to be aligned to some
5443 boundary, this can be specified as the fourth argument, otherwise it should
5444 be set to 0 or 1.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5445
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5446</div>
5447
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5448<!-- _______________________________________________________________________ -->
5449<div class="doc_subsubsection">
5450 <a name="int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a>
5451</div>
5452
5453<div class="doc_text">
5454
5455<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5456<p>This is an overloaded intrinsic. You can use <tt>llvm.sqrt</tt> on any
5457 floating point or vector of floating point type. Not all targets support all
5458 types however.</p>
5459
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5460<pre>
Dale Johannesenf9adbb62007-10-02 17:47:38 +0000[diff] [blame]5461 declare float @llvm.sqrt.f32(float %Val)
5462 declare double @llvm.sqrt.f64(double %Val)
5463 declare x86_fp80 @llvm.sqrt.f80(x86_fp80 %Val)
5464 declare fp128 @llvm.sqrt.f128(fp128 %Val)
5465 declare ppc_fp128 @llvm.sqrt.ppcf128(ppc_fp128 %Val)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5466</pre>
5467
5468<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5469<p>The '<tt>llvm.sqrt</tt>' intrinsics return the sqrt of the specified operand,
5470 returning the same value as the libm '<tt>sqrt</tt>' functions would.
5471 Unlike <tt>sqrt</tt> in libm, however, <tt>llvm.sqrt</tt> has undefined
5472 behavior for negative numbers other than -0.0 (which allows for better
5473 optimization, because there is no need to worry about errno being
5474 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]5475
5476<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5477<p>The argument and return value are floating point numbers of the same
5478 type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5479
5480<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5481<p>This function returns the sqrt of the specified operand if it is a
5482 nonnegative floating point number.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5483
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5484</div>
5485
5486<!-- _______________________________________________________________________ -->
5487<div class="doc_subsubsection">
5488 <a name="int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a>
5489</div>
5490
5491<div class="doc_text">
5492
5493<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5494<p>This is an overloaded intrinsic. You can use <tt>llvm.powi</tt> on any
5495 floating point or vector of floating point type. Not all targets support all
5496 types however.</p>
5497
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5498<pre>
Dale Johannesenf9adbb62007-10-02 17:47:38 +0000[diff] [blame]5499 declare float @llvm.powi.f32(float %Val, i32 %power)
5500 declare double @llvm.powi.f64(double %Val, i32 %power)
5501 declare x86_fp80 @llvm.powi.f80(x86_fp80 %Val, i32 %power)
5502 declare fp128 @llvm.powi.f128(fp128 %Val, i32 %power)
5503 declare ppc_fp128 @llvm.powi.ppcf128(ppc_fp128 %Val, i32 %power)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5504</pre>
5505
5506<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5507<p>The '<tt>llvm.powi.*</tt>' intrinsics return the first operand raised to the
5508 specified (positive or negative) power. The order of evaluation of
5509 multiplications is not defined. When a vector of floating point type is
5510 used, the second argument remains a scalar integer value.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5511
5512<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5513<p>The second argument is an integer power, and the first is a value to raise to
5514 that power.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5515
5516<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5517<p>This function returns the first value raised to the second power with an
5518 unspecified sequence of rounding operations.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5519
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5520</div>
5521
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5522<!-- _______________________________________________________________________ -->
5523<div class="doc_subsubsection">
5524 <a name="int_sin">'<tt>llvm.sin.*</tt>' Intrinsic</a>
5525</div>
5526
5527<div class="doc_text">
5528
5529<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5530<p>This is an overloaded intrinsic. You can use <tt>llvm.sin</tt> on any
5531 floating point or vector of floating point type. Not all targets support all
5532 types however.</p>
5533
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5534<pre>
5535 declare float @llvm.sin.f32(float %Val)
5536 declare double @llvm.sin.f64(double %Val)
5537 declare x86_fp80 @llvm.sin.f80(x86_fp80 %Val)
5538 declare fp128 @llvm.sin.f128(fp128 %Val)
5539 declare ppc_fp128 @llvm.sin.ppcf128(ppc_fp128 %Val)
5540</pre>
5541
5542<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5543<p>The '<tt>llvm.sin.*</tt>' intrinsics return the sine of the operand.</p>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5544
5545<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5546<p>The argument and return value are floating point numbers of the same
5547 type.</p>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5548
5549<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5550<p>This function returns the sine of the specified operand, returning the same
5551 values as the libm <tt>sin</tt> functions would, and handles error conditions
5552 in the same way.</p>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5553
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5554</div>
5555
5556<!-- _______________________________________________________________________ -->
5557<div class="doc_subsubsection">
5558 <a name="int_cos">'<tt>llvm.cos.*</tt>' Intrinsic</a>
5559</div>
5560
5561<div class="doc_text">
5562
5563<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5564<p>This is an overloaded intrinsic. You can use <tt>llvm.cos</tt> on any
5565 floating point or vector of floating point type. Not all targets support all
5566 types however.</p>
5567
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5568<pre>
5569 declare float @llvm.cos.f32(float %Val)
5570 declare double @llvm.cos.f64(double %Val)
5571 declare x86_fp80 @llvm.cos.f80(x86_fp80 %Val)
5572 declare fp128 @llvm.cos.f128(fp128 %Val)
5573 declare ppc_fp128 @llvm.cos.ppcf128(ppc_fp128 %Val)
5574</pre>
5575
5576<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5577<p>The '<tt>llvm.cos.*</tt>' intrinsics return the cosine of the operand.</p>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5578
5579<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5580<p>The argument and return value are floating point numbers of the same
5581 type.</p>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5582
5583<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5584<p>This function returns the cosine of the specified operand, returning the same
5585 values as the libm <tt>cos</tt> functions would, and handles error conditions
5586 in the same way.</p>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5587
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5588</div>
5589
5590<!-- _______________________________________________________________________ -->
5591<div class="doc_subsubsection">
5592 <a name="int_pow">'<tt>llvm.pow.*</tt>' Intrinsic</a>
5593</div>
5594
5595<div class="doc_text">
5596
5597<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5598<p>This is an overloaded intrinsic. You can use <tt>llvm.pow</tt> on any
5599 floating point or vector of floating point type. Not all targets support all
5600 types however.</p>
5601
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5602<pre>
5603 declare float @llvm.pow.f32(float %Val, float %Power)
5604 declare double @llvm.pow.f64(double %Val, double %Power)
5605 declare x86_fp80 @llvm.pow.f80(x86_fp80 %Val, x86_fp80 %Power)
5606 declare fp128 @llvm.pow.f128(fp128 %Val, fp128 %Power)
5607 declare ppc_fp128 @llvm.pow.ppcf128(ppc_fp128 %Val, ppc_fp128 Power)
5608</pre>
5609
5610<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5611<p>The '<tt>llvm.pow.*</tt>' intrinsics return the first operand raised to the
5612 specified (positive or negative) power.</p>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5613
5614<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5615<p>The second argument is a floating point power, and the first is a value to
5616 raise to that power.</p>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5617
5618<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5619<p>This function returns the first value raised to the second power, returning
5620 the same values as the libm <tt>pow</tt> functions would, and handles error
5621 conditions in the same way.</p>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5622
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5623</div>
5624
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5625<!-- ======================================================================= -->
5626<div class="doc_subsection">
5627 <a name="int_manip">Bit Manipulation Intrinsics</a>
5628</div>
5629
5630<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5631
5632<p>LLVM provides intrinsics for a few important bit manipulation operations.
5633 These allow efficient code generation for some algorithms.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5634
5635</div>
5636
5637<!-- _______________________________________________________________________ -->
5638<div class="doc_subsubsection">
5639 <a name="int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a>
5640</div>
5641
5642<div class="doc_text">
5643
5644<h5>Syntax:</h5>
5645<p>This is an overloaded intrinsic function. You can use bswap on any integer
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5646 type that is an even number of bytes (i.e. BitWidth % 16 == 0).</p>
5647
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5648<pre>
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]5649 declare i16 @llvm.bswap.i16(i16 &lt;id&gt;)
5650 declare i32 @llvm.bswap.i32(i32 &lt;id&gt;)
5651 declare i64 @llvm.bswap.i64(i64 &lt;id&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5652</pre>
5653
5654<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5655<p>The '<tt>llvm.bswap</tt>' family of intrinsics is used to byte swap integer
5656 values with an even number of bytes (positive multiple of 16 bits). These
5657 are useful for performing operations on data that is not in the target's
5658 native byte order.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5659
5660<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5661<p>The <tt>llvm.bswap.i16</tt> intrinsic returns an i16 value that has the high
5662 and low byte of the input i16 swapped. Similarly,
5663 the <tt>llvm.bswap.i32</tt> intrinsic returns an i32 value that has the four
5664 bytes of the input i32 swapped, so that if the input bytes are numbered 0, 1,
5665 2, 3 then the returned i32 will have its bytes in 3, 2, 1, 0 order.
5666 The <tt>llvm.bswap.i48</tt>, <tt>llvm.bswap.i64</tt> and other intrinsics
5667 extend this concept to additional even-byte lengths (6 bytes, 8 bytes and
5668 more, respectively).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5669
5670</div>
5671
5672<!-- _______________________________________________________________________ -->
5673<div class="doc_subsubsection">
5674 <a name="int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic</a>
5675</div>
5676
5677<div class="doc_text">
5678
5679<h5>Syntax:</h5>
5680<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]5681 width. Not all targets support all bit widths however.</p>
5682
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5683<pre>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5684 declare i8 @llvm.ctpop.i8(i8 &lt;src&gt;)
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]5685 declare i16 @llvm.ctpop.i16(i16 &lt;src&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5686 declare i32 @llvm.ctpop.i32(i32 &lt;src&gt;)
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]5687 declare i64 @llvm.ctpop.i64(i64 &lt;src&gt;)
5688 declare i256 @llvm.ctpop.i256(i256 &lt;src&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5689</pre>
5690
5691<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5692<p>The '<tt>llvm.ctpop</tt>' family of intrinsics counts the number of bits set
5693 in a value.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5694
5695<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5696<p>The only argument is the value to be counted. The argument may be of any
5697 integer type. The return type must match the argument type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5698
5699<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5700<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]5701
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5702</div>
5703
5704<!-- _______________________________________________________________________ -->
5705<div class="doc_subsubsection">
5706 <a name="int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic</a>
5707</div>
5708
5709<div class="doc_text">
5710
5711<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5712<p>This is an overloaded intrinsic. You can use <tt>llvm.ctlz</tt> on any
5713 integer bit width. Not all targets support all bit widths however.</p>
5714
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5715<pre>
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]5716 declare i8 @llvm.ctlz.i8 (i8 &lt;src&gt;)
5717 declare i16 @llvm.ctlz.i16(i16 &lt;src&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5718 declare i32 @llvm.ctlz.i32(i32 &lt;src&gt;)
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]5719 declare i64 @llvm.ctlz.i64(i64 &lt;src&gt;)
5720 declare i256 @llvm.ctlz.i256(i256 &lt;src&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5721</pre>
5722
5723<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5724<p>The '<tt>llvm.ctlz</tt>' family of intrinsic functions counts the number of
5725 leading zeros in a variable.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5726
5727<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5728<p>The only argument is the value to be counted. The argument may be of any
5729 integer type. The return type must match the argument type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5730
5731<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5732<p>The '<tt>llvm.ctlz</tt>' intrinsic counts the leading (most significant)
5733 zeros in a variable. If the src == 0 then the result is the size in bits of
5734 the type of src. For example, <tt>llvm.ctlz(i32 2) = 30</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5735
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5736</div>
5737
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5738<!-- _______________________________________________________________________ -->
5739<div class="doc_subsubsection">
5740 <a name="int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic</a>
5741</div>
5742
5743<div class="doc_text">
5744
5745<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5746<p>This is an overloaded intrinsic. You can use <tt>llvm.cttz</tt> on any
5747 integer bit width. Not all targets support all bit widths however.</p>
5748
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5749<pre>
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]5750 declare i8 @llvm.cttz.i8 (i8 &lt;src&gt;)
5751 declare i16 @llvm.cttz.i16(i16 &lt;src&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5752 declare i32 @llvm.cttz.i32(i32 &lt;src&gt;)
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]5753 declare i64 @llvm.cttz.i64(i64 &lt;src&gt;)
5754 declare i256 @llvm.cttz.i256(i256 &lt;src&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5755</pre>
5756
5757<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5758<p>The '<tt>llvm.cttz</tt>' family of intrinsic functions counts the number of
5759 trailing zeros.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5760
5761<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5762<p>The only argument is the value to be counted. The argument may be of any
5763 integer type. The return type must match the argument type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5764
5765<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5766<p>The '<tt>llvm.cttz</tt>' intrinsic counts the trailing (least significant)
5767 zeros in a variable. If the src == 0 then the result is the size in bits of
5768 the type of src. For example, <tt>llvm.cttz(2) = 1</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5769
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5770</div>
5771
Bill Wendling3e1258b2009-02-08 04:04:40 +0000[diff] [blame]5772<!-- ======================================================================= -->
5773<div class="doc_subsection">
5774 <a name="int_overflow">Arithmetic with Overflow Intrinsics</a>
5775</div>
5776
5777<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5778
5779<p>LLVM provides intrinsics for some arithmetic with overflow operations.</p>
Bill Wendling3e1258b2009-02-08 04:04:40 +0000[diff] [blame]5780
5781</div>
5782
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5783<!-- _______________________________________________________________________ -->
5784<div class="doc_subsubsection">
Bill Wendling3e1258b2009-02-08 04:04:40 +0000[diff] [blame]5785 <a name="int_sadd_overflow">'<tt>llvm.sadd.with.overflow.*</tt>' Intrinsics</a>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5786</div>
5787
5788<div class="doc_text">
5789
5790<h5>Syntax:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5791<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]5792 on any integer bit width.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5793
5794<pre>
5795 declare {i16, i1} @llvm.sadd.with.overflow.i16(i16 %a, i16 %b)
5796 declare {i32, i1} @llvm.sadd.with.overflow.i32(i32 %a, i32 %b)
5797 declare {i64, i1} @llvm.sadd.with.overflow.i64(i64 %a, i64 %b)
5798</pre>
5799
5800<h5>Overview:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5801<p>The '<tt>llvm.sadd.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5802 a signed addition of the two arguments, and indicate whether an overflow
5803 occurred during the signed summation.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5804
5805<h5>Arguments:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5806<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]5807 be of integer types of any bit width, but they must have the same bit
5808 width. The second element of the result structure must be of
5809 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
5810 undergo signed addition.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5811
5812<h5>Semantics:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5813<p>The '<tt>llvm.sadd.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5814 a signed addition of the two variables. They return a structure &mdash; the
5815 first element of which is the signed summation, and the second element of
5816 which is a bit specifying if the signed summation resulted in an
5817 overflow.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5818
5819<h5>Examples:</h5>
5820<pre>
5821 %res = call {i32, i1} @llvm.sadd.with.overflow.i32(i32 %a, i32 %b)
5822 %sum = extractvalue {i32, i1} %res, 0
5823 %obit = extractvalue {i32, i1} %res, 1
5824 br i1 %obit, label %overflow, label %normal
5825</pre>
5826
5827</div>
5828
5829<!-- _______________________________________________________________________ -->
5830<div class="doc_subsubsection">
Bill Wendling3e1258b2009-02-08 04:04:40 +0000[diff] [blame]5831 <a name="int_uadd_overflow">'<tt>llvm.uadd.with.overflow.*</tt>' Intrinsics</a>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5832</div>
5833
5834<div class="doc_text">
5835
5836<h5>Syntax:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5837<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]5838 on any integer bit width.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5839
5840<pre>
5841 declare {i16, i1} @llvm.uadd.with.overflow.i16(i16 %a, i16 %b)
5842 declare {i32, i1} @llvm.uadd.with.overflow.i32(i32 %a, i32 %b)
5843 declare {i64, i1} @llvm.uadd.with.overflow.i64(i64 %a, i64 %b)
5844</pre>
5845
5846<h5>Overview:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5847<p>The '<tt>llvm.uadd.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5848 an unsigned addition of the two arguments, and indicate whether a carry
5849 occurred during the unsigned summation.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5850
5851<h5>Arguments:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5852<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]5853 be of integer types of any bit width, but they must have the same bit
5854 width. The second element of the result structure must be of
5855 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
5856 undergo unsigned addition.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5857
5858<h5>Semantics:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5859<p>The '<tt>llvm.uadd.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5860 an unsigned addition of the two arguments. They return a structure &mdash;
5861 the first element of which is the sum, and the second element of which is a
5862 bit specifying if the unsigned summation resulted in a carry.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5863
5864<h5>Examples:</h5>
5865<pre>
5866 %res = call {i32, i1} @llvm.uadd.with.overflow.i32(i32 %a, i32 %b)
5867 %sum = extractvalue {i32, i1} %res, 0
5868 %obit = extractvalue {i32, i1} %res, 1
5869 br i1 %obit, label %carry, label %normal
5870</pre>
5871
5872</div>
5873
5874<!-- _______________________________________________________________________ -->
5875<div class="doc_subsubsection">
Bill Wendling3e1258b2009-02-08 04:04:40 +0000[diff] [blame]5876 <a name="int_ssub_overflow">'<tt>llvm.ssub.with.overflow.*</tt>' Intrinsics</a>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5877</div>
5878
5879<div class="doc_text">
5880
5881<h5>Syntax:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5882<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]5883 on any integer bit width.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5884
5885<pre>
5886 declare {i16, i1} @llvm.ssub.with.overflow.i16(i16 %a, i16 %b)
5887 declare {i32, i1} @llvm.ssub.with.overflow.i32(i32 %a, i32 %b)
5888 declare {i64, i1} @llvm.ssub.with.overflow.i64(i64 %a, i64 %b)
5889</pre>
5890
5891<h5>Overview:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5892<p>The '<tt>llvm.ssub.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5893 a signed subtraction of the two arguments, and indicate whether an overflow
5894 occurred during the signed subtraction.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5895
5896<h5>Arguments:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5897<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]5898 be of integer types of any bit width, but they must have the same bit
5899 width. The second element of the result structure must be of
5900 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
5901 undergo signed subtraction.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5902
5903<h5>Semantics:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5904<p>The '<tt>llvm.ssub.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5905 a signed subtraction of the two arguments. They return a structure &mdash;
5906 the first element of which is the subtraction, and the second element of
5907 which is a bit specifying if the signed subtraction resulted in an
5908 overflow.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5909
5910<h5>Examples:</h5>
5911<pre>
5912 %res = call {i32, i1} @llvm.ssub.with.overflow.i32(i32 %a, i32 %b)
5913 %sum = extractvalue {i32, i1} %res, 0
5914 %obit = extractvalue {i32, i1} %res, 1
5915 br i1 %obit, label %overflow, label %normal
5916</pre>
5917
5918</div>
5919
5920<!-- _______________________________________________________________________ -->
5921<div class="doc_subsubsection">
Bill Wendling3e1258b2009-02-08 04:04:40 +0000[diff] [blame]5922 <a name="int_usub_overflow">'<tt>llvm.usub.with.overflow.*</tt>' Intrinsics</a>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5923</div>
5924
5925<div class="doc_text">
5926
5927<h5>Syntax:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5928<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]5929 on any integer bit width.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5930
5931<pre>
5932 declare {i16, i1} @llvm.usub.with.overflow.i16(i16 %a, i16 %b)
5933 declare {i32, i1} @llvm.usub.with.overflow.i32(i32 %a, i32 %b)
5934 declare {i64, i1} @llvm.usub.with.overflow.i64(i64 %a, i64 %b)
5935</pre>
5936
5937<h5>Overview:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5938<p>The '<tt>llvm.usub.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5939 an unsigned subtraction of the two arguments, and indicate whether an
5940 overflow occurred during the unsigned subtraction.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5941
5942<h5>Arguments:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5943<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]5944 be of integer types of any bit width, but they must have the same bit
5945 width. The second element of the result structure must be of
5946 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
5947 undergo unsigned subtraction.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5948
5949<h5>Semantics:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5950<p>The '<tt>llvm.usub.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5951 an unsigned subtraction of the two arguments. They return a structure &mdash;
5952 the first element of which is the subtraction, and the second element of
5953 which is a bit specifying if the unsigned subtraction resulted in an
5954 overflow.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5955
5956<h5>Examples:</h5>
5957<pre>
5958 %res = call {i32, i1} @llvm.usub.with.overflow.i32(i32 %a, i32 %b)
5959 %sum = extractvalue {i32, i1} %res, 0
5960 %obit = extractvalue {i32, i1} %res, 1
5961 br i1 %obit, label %overflow, label %normal
5962</pre>
5963
5964</div>
5965
5966<!-- _______________________________________________________________________ -->
5967<div class="doc_subsubsection">
Bill Wendling3e1258b2009-02-08 04:04:40 +0000[diff] [blame]5968 <a name="int_smul_overflow">'<tt>llvm.smul.with.overflow.*</tt>' Intrinsics</a>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5969</div>
5970
5971<div class="doc_text">
5972
5973<h5>Syntax:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5974<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]5975 on any integer bit width.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5976
5977<pre>
5978 declare {i16, i1} @llvm.smul.with.overflow.i16(i16 %a, i16 %b)
5979 declare {i32, i1} @llvm.smul.with.overflow.i32(i32 %a, i32 %b)
5980 declare {i64, i1} @llvm.smul.with.overflow.i64(i64 %a, i64 %b)
5981</pre>
5982
5983<h5>Overview:</h5>
5984
5985<p>The '<tt>llvm.smul.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5986 a signed multiplication of the two arguments, and indicate whether an
5987 overflow occurred during the signed multiplication.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5988
5989<h5>Arguments:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5990<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]5991 be of integer types of any bit width, but they must have the same bit
5992 width. The second element of the result structure must be of
5993 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
5994 undergo signed multiplication.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5995
5996<h5>Semantics:</h5>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]5997<p>The '<tt>llvm.smul.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]5998 a signed multiplication of the two arguments. They return a structure &mdash;
5999 the first element of which is the multiplication, and the second element of
6000 which is a bit specifying if the signed multiplication resulted in an
6001 overflow.</p>
Bill Wendling3f8cebe2009-02-08 01:40:31 +0000[diff] [blame]6002
6003<h5>Examples:</h5>
6004<pre>
6005 %res = call {i32, i1} @llvm.smul.with.overflow.i32(i32 %a, i32 %b)
6006 %sum = extractvalue {i32, i1} %res, 0
6007 %obit = extractvalue {i32, i1} %res, 1
6008 br i1 %obit, label %overflow, label %normal
6009</pre>
6010
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6011</div>
6012
Bill Wendlingbda98b62009-02-08 23:00:09 +0000[diff] [blame]6013<!-- _______________________________________________________________________ -->
6014<div class="doc_subsubsection">
6015 <a name="int_umul_overflow">'<tt>llvm.umul.with.overflow.*</tt>' Intrinsics</a>
6016</div>
6017
6018<div class="doc_text">
6019
6020<h5>Syntax:</h5>
Bill Wendlingbda98b62009-02-08 23:00:09 +0000[diff] [blame]6021<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]6022 on any integer bit width.</p>
Bill Wendlingbda98b62009-02-08 23:00:09 +0000[diff] [blame]6023
6024<pre>
6025 declare {i16, i1} @llvm.umul.with.overflow.i16(i16 %a, i16 %b)
6026 declare {i32, i1} @llvm.umul.with.overflow.i32(i32 %a, i32 %b)
6027 declare {i64, i1} @llvm.umul.with.overflow.i64(i64 %a, i64 %b)
6028</pre>
6029
6030<h5>Overview:</h5>
Bill Wendlingbda98b62009-02-08 23:00:09 +0000[diff] [blame]6031<p>The '<tt>llvm.umul.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6032 a unsigned multiplication of the two arguments, and indicate whether an
6033 overflow occurred during the unsigned multiplication.</p>
Bill Wendlingbda98b62009-02-08 23:00:09 +0000[diff] [blame]6034
6035<h5>Arguments:</h5>
Bill Wendlingbda98b62009-02-08 23:00:09 +0000[diff] [blame]6036<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]6037 be of integer types of any bit width, but they must have the same bit
6038 width. The second element of the result structure must be of
6039 type <tt>i1</tt>. <tt>%a</tt> and <tt>%b</tt> are the two values that will
6040 undergo unsigned multiplication.</p>
Bill Wendlingbda98b62009-02-08 23:00:09 +0000[diff] [blame]6041
6042<h5>Semantics:</h5>
Bill Wendlingbda98b62009-02-08 23:00:09 +0000[diff] [blame]6043<p>The '<tt>llvm.umul.with.overflow</tt>' family of intrinsic functions perform
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6044 an unsigned multiplication of the two arguments. They return a structure
6045 &mdash; the first element of which is the multiplication, and the second
6046 element of which is a bit specifying if the unsigned multiplication resulted
6047 in an overflow.</p>
Bill Wendlingbda98b62009-02-08 23:00:09 +0000[diff] [blame]6048
6049<h5>Examples:</h5>
6050<pre>
6051 %res = call {i32, i1} @llvm.umul.with.overflow.i32(i32 %a, i32 %b)
6052 %sum = extractvalue {i32, i1} %res, 0
6053 %obit = extractvalue {i32, i1} %res, 1
6054 br i1 %obit, label %overflow, label %normal
6055</pre>
6056
6057</div>
6058
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6059<!-- ======================================================================= -->
6060<div class="doc_subsection">
6061 <a name="int_debugger">Debugger Intrinsics</a>
6062</div>
6063
6064<div class="doc_text">
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6065
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6066<p>The LLVM debugger intrinsics (which all start with <tt>llvm.dbg.</tt>
6067 prefix), are described in
6068 the <a href="SourceLevelDebugging.html#format_common_intrinsics">LLVM Source
6069 Level Debugging</a> document.</p>
6070
6071</div>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6072
6073<!-- ======================================================================= -->
6074<div class="doc_subsection">
6075 <a name="int_eh">Exception Handling Intrinsics</a>
6076</div>
6077
6078<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6079
6080<p>The LLVM exception handling intrinsics (which all start with
6081 <tt>llvm.eh.</tt> prefix), are described in
6082 the <a href="ExceptionHandling.html#format_common_intrinsics">LLVM Exception
6083 Handling</a> document.</p>
6084
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6085</div>
6086
6087<!-- ======================================================================= -->
6088<div class="doc_subsection">
Duncan Sands7407a9f2007-09-11 14:10:23 +0000[diff] [blame]6089 <a name="int_trampoline">Trampoline Intrinsic</a>
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]6090</div>
6091
6092<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6093
6094<p>This intrinsic makes it possible to excise one parameter, marked with
6095 the <tt>nest</tt> attribute, from a function. The result is a callable
6096 function pointer lacking the nest parameter - the caller does not need to
6097 provide a value for it. Instead, the value to use is stored in advance in a
6098 "trampoline", a block of memory usually allocated on the stack, which also
6099 contains code to splice the nest value into the argument list. This is used
6100 to implement the GCC nested function address extension.</p>
6101
6102<p>For example, if the function is
6103 <tt>i32 f(i8* nest %c, i32 %x, i32 %y)</tt> then the resulting function
6104 pointer has signature <tt>i32 (i32, i32)*</tt>. It can be created as
6105 follows:</p>
6106
6107<div class="doc_code">
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]6108<pre>
Duncan Sands7407a9f2007-09-11 14:10:23 +0000[diff] [blame]6109 %tramp = alloca [10 x i8], align 4 ; size and alignment only correct for X86
6110 %tramp1 = getelementptr [10 x i8]* %tramp, i32 0, i32 0
6111 %p = call i8* @llvm.init.trampoline( i8* %tramp1, i8* bitcast (i32 (i8* nest , i32, i32)* @f to i8*), i8* %nval )
6112 %fp = bitcast i8* %p to i32 (i32, i32)*
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]6113</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6114</div>
6115
6116<p>The call <tt>%val = call i32 %fp( i32 %x, i32 %y )</tt> is then equivalent
6117 to <tt>%val = call i32 %f( i8* %nval, i32 %x, i32 %y )</tt>.</p>
6118
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]6119</div>
6120
6121<!-- _______________________________________________________________________ -->
6122<div class="doc_subsubsection">
6123 <a name="int_it">'<tt>llvm.init.trampoline</tt>' Intrinsic</a>
6124</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6125
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]6126<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6127
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]6128<h5>Syntax:</h5>
6129<pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6130 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]6131</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6132
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]6133<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6134<p>This fills the memory pointed to by <tt>tramp</tt> with code and returns a
6135 function pointer suitable for executing it.</p>
6136
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]6137<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6138<p>The <tt>llvm.init.trampoline</tt> intrinsic takes three arguments, all
6139 pointers. The <tt>tramp</tt> argument must point to a sufficiently large and
6140 sufficiently aligned block of memory; this memory is written to by the
6141 intrinsic. Note that the size and the alignment are target-specific - LLVM
6142 currently provides no portable way of determining them, so a front-end that
6143 generates this intrinsic needs to have some target-specific knowledge.
6144 The <tt>func</tt> argument must hold a function bitcast to
6145 an <tt>i8*</tt>.</p>
6146
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]6147<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6148<p>The block of memory pointed to by <tt>tramp</tt> is filled with target
6149 dependent code, turning it into a function. A pointer to this function is
6150 returned, but needs to be bitcast to an <a href="#int_trampoline">appropriate
6151 function pointer type</a> before being called. The new function's signature
6152 is the same as that of <tt>func</tt> with any arguments marked with
6153 the <tt>nest</tt> attribute removed. At most one such <tt>nest</tt> argument
6154 is allowed, and it must be of pointer type. Calling the new function is
6155 equivalent to calling <tt>func</tt> with the same argument list, but
6156 with <tt>nval</tt> used for the missing <tt>nest</tt> argument. If, after
6157 calling <tt>llvm.init.trampoline</tt>, the memory pointed to
6158 by <tt>tramp</tt> is modified, then the effect of any later call to the
6159 returned function pointer is undefined.</p>
6160
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]6161</div>
6162
6163<!-- ======================================================================= -->
6164<div class="doc_subsection">
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]6165 <a name="int_atomics">Atomic Operations and Synchronization Intrinsics</a>
6166</div>
6167
6168<div class="doc_text">
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]6169
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6170<p>These intrinsic functions expand the "universal IR" of LLVM to represent
6171 hardware constructs for atomic operations and memory synchronization. This
6172 provides an interface to the hardware, not an interface to the programmer. It
6173 is aimed at a low enough level to allow any programming models or APIs
6174 (Application Programming Interfaces) which need atomic behaviors to map
6175 cleanly onto it. It is also modeled primarily on hardware behavior. Just as
6176 hardware provides a "universal IR" for source languages, it also provides a
6177 starting point for developing a "universal" atomic operation and
6178 synchronization IR.</p>
6179
6180<p>These do <em>not</em> form an API such as high-level threading libraries,
6181 software transaction memory systems, atomic primitives, and intrinsic
6182 functions as found in BSD, GNU libc, atomic_ops, APR, and other system and
6183 application libraries. The hardware interface provided by LLVM should allow
6184 a clean implementation of all of these APIs and parallel programming models.
6185 No one model or paradigm should be selected above others unless the hardware
6186 itself ubiquitously does so.</p>
6187
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]6188</div>
6189
6190<!-- _______________________________________________________________________ -->
6191<div class="doc_subsubsection">
6192 <a name="int_memory_barrier">'<tt>llvm.memory.barrier</tt>' Intrinsic</a>
6193</div>
6194<div class="doc_text">
6195<h5>Syntax:</h5>
6196<pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6197 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]6198</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6199
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]6200<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6201<p>The <tt>llvm.memory.barrier</tt> intrinsic guarantees ordering between
6202 specific pairs of memory access types.</p>
6203
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]6204<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6205<p>The <tt>llvm.memory.barrier</tt> intrinsic requires five boolean arguments.
6206 The first four arguments enables a specific barrier as listed below. The
6207 fith argument specifies that the barrier applies to io or device or uncached
6208 memory.</p>
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]6209
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6210<ul>
6211 <li><tt>ll</tt>: load-load barrier</li>
6212 <li><tt>ls</tt>: load-store barrier</li>
6213 <li><tt>sl</tt>: store-load barrier</li>
6214 <li><tt>ss</tt>: store-store barrier</li>
6215 <li><tt>device</tt>: barrier applies to device and uncached memory also.</li>
6216</ul>
6217
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]6218<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6219<p>This intrinsic causes the system to enforce some ordering constraints upon
6220 the loads and stores of the program. This barrier does not
6221 indicate <em>when</em> any events will occur, it only enforces
6222 an <em>order</em> in which they occur. For any of the specified pairs of load
6223 and store operations (f.ex. load-load, or store-load), all of the first
6224 operations preceding the barrier will complete before any of the second
6225 operations succeeding the barrier begin. Specifically the semantics for each
6226 pairing is as follows:</p>
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]6227
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6228<ul>
6229 <li><tt>ll</tt>: All loads before the barrier must complete before any load
6230 after the barrier begins.</li>
6231 <li><tt>ls</tt>: All loads before the barrier must complete before any
6232 store after the barrier begins.</li>
6233 <li><tt>ss</tt>: All stores before the barrier must complete before any
6234 store after the barrier begins.</li>
6235 <li><tt>sl</tt>: All stores before the barrier must complete before any
6236 load after the barrier begins.</li>
6237</ul>
6238
6239<p>These semantics are applied with a logical "and" behavior when more than one
6240 is enabled in a single memory barrier intrinsic.</p>
6241
6242<p>Backends may implement stronger barriers than those requested when they do
6243 not support as fine grained a barrier as requested. Some architectures do
6244 not need all types of barriers and on such architectures, these become
6245 noops.</p>
6246
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]6247<h5>Example:</h5>
6248<pre>
6249%ptr = malloc i32
6250 store i32 4, %ptr
6251
6252%result1 = load i32* %ptr <i>; yields {i32}:result1 = 4</i>
6253 call void @llvm.memory.barrier( i1 false, i1 true, i1 false, i1 false )
6254 <i>; guarantee the above finishes</i>
6255 store i32 8, %ptr <i>; before this begins</i>
6256</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6257
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]6258</div>
6259
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6260<!-- _______________________________________________________________________ -->
6261<div class="doc_subsubsection">
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6262 <a name="int_atomic_cmp_swap">'<tt>llvm.atomic.cmp.swap.*</tt>' Intrinsic</a>
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6263</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6264
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6265<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6266
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6267<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6268<p>This is an overloaded intrinsic. You can use <tt>llvm.atomic.cmp.swap</tt> on
6269 any integer bit width and for different address spaces. Not all targets
6270 support all bit widths however.</p>
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6271
6272<pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6273 declare i8 @llvm.atomic.cmp.swap.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;cmp&gt;, i8 &lt;val&gt; )
6274 declare i16 @llvm.atomic.cmp.swap.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;cmp&gt;, i16 &lt;val&gt; )
6275 declare i32 @llvm.atomic.cmp.swap.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;cmp&gt;, i32 &lt;val&gt; )
6276 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]6277</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6278
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6279<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6280<p>This loads a value in memory and compares it to a given value. If they are
6281 equal, it stores a new value into the memory.</p>
6282
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6283<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6284<p>The <tt>llvm.atomic.cmp.swap</tt> intrinsic takes three arguments. The result
6285 as well as both <tt>cmp</tt> and <tt>val</tt> must be integer values with the
6286 same bit width. The <tt>ptr</tt> argument must be a pointer to a value of
6287 this integer type. While any bit width integer may be used, targets may only
6288 lower representations they support in hardware.</p>
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6289
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6290<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6291<p>This entire intrinsic must be executed atomically. It first loads the value
6292 in memory pointed to by <tt>ptr</tt> and compares it with the
6293 value <tt>cmp</tt>. If they are equal, <tt>val</tt> is stored into the
6294 memory. The loaded value is yielded in all cases. This provides the
6295 equivalent of an atomic compare-and-swap operation within the SSA
6296 framework.</p>
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6297
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6298<h5>Examples:</h5>
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6299<pre>
6300%ptr = malloc i32
6301 store i32 4, %ptr
6302
6303%val1 = add i32 4, 4
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6304%result1 = call i32 @llvm.atomic.cmp.swap.i32.p0i32( i32* %ptr, i32 4, %val1 )
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6305 <i>; yields {i32}:result1 = 4</i>
6306%stored1 = icmp eq i32 %result1, 4 <i>; yields {i1}:stored1 = true</i>
6307%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 8</i>
6308
6309%val2 = add i32 1, 1
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6310%result2 = call i32 @llvm.atomic.cmp.swap.i32.p0i32( i32* %ptr, i32 5, %val2 )
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6311 <i>; yields {i32}:result2 = 8</i>
6312%stored2 = icmp eq i32 %result2, 5 <i>; yields {i1}:stored2 = false</i>
6313
6314%memval2 = load i32* %ptr <i>; yields {i32}:memval2 = 8</i>
6315</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6316
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6317</div>
6318
6319<!-- _______________________________________________________________________ -->
6320<div class="doc_subsubsection">
6321 <a name="int_atomic_swap">'<tt>llvm.atomic.swap.*</tt>' Intrinsic</a>
6322</div>
6323<div class="doc_text">
6324<h5>Syntax:</h5>
6325
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6326<p>This is an overloaded intrinsic. You can use <tt>llvm.atomic.swap</tt> on any
6327 integer bit width. Not all targets support all bit widths however.</p>
6328
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6329<pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6330 declare i8 @llvm.atomic.swap.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;val&gt; )
6331 declare i16 @llvm.atomic.swap.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;val&gt; )
6332 declare i32 @llvm.atomic.swap.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;val&gt; )
6333 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]6334</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6335
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6336<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6337<p>This intrinsic loads the value stored in memory at <tt>ptr</tt> and yields
6338 the value from memory. It then stores the value in <tt>val</tt> in the memory
6339 at <tt>ptr</tt>.</p>
6340
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6341<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6342<p>The <tt>llvm.atomic.swap</tt> intrinsic takes two arguments. Both
6343 the <tt>val</tt> argument and the result must be integers of the same bit
6344 width. The first argument, <tt>ptr</tt>, must be a pointer to a value of this
6345 integer type. The targets may only lower integer representations they
6346 support.</p>
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6347
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6348<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6349<p>This intrinsic loads the value pointed to by <tt>ptr</tt>, yields it, and
6350 stores <tt>val</tt> back into <tt>ptr</tt> atomically. This provides the
6351 equivalent of an atomic swap operation within the SSA framework.</p>
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6352
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6353<h5>Examples:</h5>
6354<pre>
6355%ptr = malloc i32
6356 store i32 4, %ptr
6357
6358%val1 = add i32 4, 4
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6359%result1 = call i32 @llvm.atomic.swap.i32.p0i32( i32* %ptr, i32 %val1 )
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6360 <i>; yields {i32}:result1 = 4</i>
6361%stored1 = icmp eq i32 %result1, 4 <i>; yields {i1}:stored1 = true</i>
6362%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 8</i>
6363
6364%val2 = add i32 1, 1
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6365%result2 = call i32 @llvm.atomic.swap.i32.p0i32( i32* %ptr, i32 %val2 )
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6366 <i>; yields {i32}:result2 = 8</i>
6367
6368%stored2 = icmp eq i32 %result2, 8 <i>; yields {i1}:stored2 = true</i>
6369%memval2 = load i32* %ptr <i>; yields {i32}:memval2 = 2</i>
6370</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6371
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6372</div>
6373
6374<!-- _______________________________________________________________________ -->
6375<div class="doc_subsubsection">
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6376 <a name="int_atomic_load_add">'<tt>llvm.atomic.load.add.*</tt>' Intrinsic</a>
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6377
6378</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6379
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6380<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6381
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6382<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6383<p>This is an overloaded intrinsic. You can use <tt>llvm.atomic.load.add</tt> on
6384 any integer bit width. Not all targets support all bit widths however.</p>
6385
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6386<pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6387 declare i8 @llvm.atomic.load.add.i8..p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6388 declare i16 @llvm.atomic.load.add.i16..p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6389 declare i32 @llvm.atomic.load.add.i32..p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6390 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]6391</pre>
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6392
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6393<h5>Overview:</h5>
6394<p>This intrinsic adds <tt>delta</tt> to the value stored in memory
6395 at <tt>ptr</tt>. It yields the original value at <tt>ptr</tt>.</p>
6396
6397<h5>Arguments:</h5>
6398<p>The intrinsic takes two arguments, the first a pointer to an integer value
6399 and the second an integer value. The result is also an integer value. These
6400 integer types can have any bit width, but they must all have the same bit
6401 width. The targets may only lower integer representations they support.</p>
6402
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6403<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6404<p>This intrinsic does a series of operations atomically. It first loads the
6405 value stored at <tt>ptr</tt>. It then adds <tt>delta</tt>, stores the result
6406 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]6407
6408<h5>Examples:</h5>
6409<pre>
6410%ptr = malloc i32
6411 store i32 4, %ptr
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6412%result1 = call i32 @llvm.atomic.load.add.i32.p0i32( i32* %ptr, i32 4 )
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6413 <i>; yields {i32}:result1 = 4</i>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6414%result2 = call i32 @llvm.atomic.load.add.i32.p0i32( i32* %ptr, i32 2 )
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6415 <i>; yields {i32}:result2 = 8</i>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6416%result3 = call i32 @llvm.atomic.load.add.i32.p0i32( i32* %ptr, i32 5 )
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6417 <i>; yields {i32}:result3 = 10</i>
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6418%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 15</i>
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6419</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6420
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6421</div>
6422
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6423<!-- _______________________________________________________________________ -->
6424<div class="doc_subsubsection">
6425 <a name="int_atomic_load_sub">'<tt>llvm.atomic.load.sub.*</tt>' Intrinsic</a>
6426
6427</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6428
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6429<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6430
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6431<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6432<p>This is an overloaded intrinsic. You can use <tt>llvm.atomic.load.sub</tt> on
6433 any integer bit width and for different address spaces. Not all targets
6434 support all bit widths however.</p>
6435
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6436<pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6437 declare i8 @llvm.atomic.load.sub.i8.p0i32( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6438 declare i16 @llvm.atomic.load.sub.i16.p0i32( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6439 declare i32 @llvm.atomic.load.sub.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6440 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]6441</pre>
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6442
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6443<h5>Overview:</h5>
6444<p>This intrinsic subtracts <tt>delta</tt> to the value stored in memory at
6445 <tt>ptr</tt>. It yields the original value at <tt>ptr</tt>.</p>
6446
6447<h5>Arguments:</h5>
6448<p>The intrinsic takes two arguments, the first a pointer to an integer value
6449 and the second an integer value. The result is also an integer value. These
6450 integer types can have any bit width, but they must all have the same bit
6451 width. The targets may only lower integer representations they support.</p>
6452
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6453<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6454<p>This intrinsic does a series of operations atomically. It first loads the
6455 value stored at <tt>ptr</tt>. It then subtracts <tt>delta</tt>, stores the
6456 result to <tt>ptr</tt>. It yields the original value stored
6457 at <tt>ptr</tt>.</p>
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6458
6459<h5>Examples:</h5>
6460<pre>
6461%ptr = malloc i32
6462 store i32 8, %ptr
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6463%result1 = call i32 @llvm.atomic.load.sub.i32.p0i32( i32* %ptr, i32 4 )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6464 <i>; yields {i32}:result1 = 8</i>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6465%result2 = call i32 @llvm.atomic.load.sub.i32.p0i32( i32* %ptr, i32 2 )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6466 <i>; yields {i32}:result2 = 4</i>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6467%result3 = call i32 @llvm.atomic.load.sub.i32.p0i32( i32* %ptr, i32 5 )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6468 <i>; yields {i32}:result3 = 2</i>
6469%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = -3</i>
6470</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6471
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6472</div>
6473
6474<!-- _______________________________________________________________________ -->
6475<div class="doc_subsubsection">
6476 <a name="int_atomic_load_and">'<tt>llvm.atomic.load.and.*</tt>' Intrinsic</a><br>
6477 <a name="int_atomic_load_nand">'<tt>llvm.atomic.load.nand.*</tt>' Intrinsic</a><br>
6478 <a name="int_atomic_load_or">'<tt>llvm.atomic.load.or.*</tt>' Intrinsic</a><br>
6479 <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]6480</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6481
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6482<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6483
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6484<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6485<p>These are overloaded intrinsics. You can
6486 use <tt>llvm.atomic.load_and</tt>, <tt>llvm.atomic.load_nand</tt>,
6487 <tt>llvm.atomic.load_or</tt>, and <tt>llvm.atomic.load_xor</tt> on any integer
6488 bit width and for different address spaces. Not all targets support all bit
6489 widths however.</p>
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6490
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6491<pre>
6492 declare i8 @llvm.atomic.load.and.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6493 declare i16 @llvm.atomic.load.and.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6494 declare i32 @llvm.atomic.load.and.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6495 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]6496</pre>
6497
6498<pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6499 declare i8 @llvm.atomic.load.or.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6500 declare i16 @llvm.atomic.load.or.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6501 declare i32 @llvm.atomic.load.or.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6502 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]6503</pre>
6504
6505<pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6506 declare i8 @llvm.atomic.load.nand.i8.p0i32( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6507 declare i16 @llvm.atomic.load.nand.i16.p0i32( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6508 declare i32 @llvm.atomic.load.nand.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6509 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]6510</pre>
6511
6512<pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6513 declare i8 @llvm.atomic.load.xor.i8.p0i32( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6514 declare i16 @llvm.atomic.load.xor.i16.p0i32( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6515 declare i32 @llvm.atomic.load.xor.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6516 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]6517</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6518
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6519<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6520<p>These intrinsics bitwise the operation (and, nand, or, xor) <tt>delta</tt> to
6521 the value stored in memory at <tt>ptr</tt>. It yields the original value
6522 at <tt>ptr</tt>.</p>
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6523
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6524<h5>Arguments:</h5>
6525<p>These intrinsics take two arguments, the first a pointer to an integer value
6526 and the second an integer value. The result is also an integer value. These
6527 integer types can have any bit width, but they must all have the same bit
6528 width. The targets may only lower integer representations they support.</p>
6529
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6530<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6531<p>These intrinsics does a series of operations atomically. They first load the
6532 value stored at <tt>ptr</tt>. They then do the bitwise
6533 operation <tt>delta</tt>, store the result to <tt>ptr</tt>. They yield the
6534 original value stored at <tt>ptr</tt>.</p>
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6535
6536<h5>Examples:</h5>
6537<pre>
6538%ptr = malloc i32
6539 store i32 0x0F0F, %ptr
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6540%result0 = call i32 @llvm.atomic.load.nand.i32.p0i32( i32* %ptr, i32 0xFF )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6541 <i>; yields {i32}:result0 = 0x0F0F</i>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6542%result1 = call i32 @llvm.atomic.load.and.i32.p0i32( i32* %ptr, i32 0xFF )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6543 <i>; yields {i32}:result1 = 0xFFFFFFF0</i>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6544%result2 = call i32 @llvm.atomic.load.or.i32.p0i32( i32* %ptr, i32 0F )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6545 <i>; yields {i32}:result2 = 0xF0</i>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6546%result3 = call i32 @llvm.atomic.load.xor.i32.p0i32( i32* %ptr, i32 0F )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6547 <i>; yields {i32}:result3 = FF</i>
6548%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = F0</i>
6549</pre>
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6550
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6551</div>
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6552
6553<!-- _______________________________________________________________________ -->
6554<div class="doc_subsubsection">
6555 <a name="int_atomic_load_max">'<tt>llvm.atomic.load.max.*</tt>' Intrinsic</a><br>
6556 <a name="int_atomic_load_min">'<tt>llvm.atomic.load.min.*</tt>' Intrinsic</a><br>
6557 <a name="int_atomic_load_umax">'<tt>llvm.atomic.load.umax.*</tt>' Intrinsic</a><br>
6558 <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]6559</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6560
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6561<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6562
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6563<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6564<p>These are overloaded intrinsics. You can use <tt>llvm.atomic.load_max</tt>,
6565 <tt>llvm.atomic.load_min</tt>, <tt>llvm.atomic.load_umax</tt>, and
6566 <tt>llvm.atomic.load_umin</tt> on any integer bit width and for different
6567 address spaces. Not all targets support all bit widths however.</p>
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6568
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6569<pre>
6570 declare i8 @llvm.atomic.load.max.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6571 declare i16 @llvm.atomic.load.max.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6572 declare i32 @llvm.atomic.load.max.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6573 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]6574</pre>
6575
6576<pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6577 declare i8 @llvm.atomic.load.min.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6578 declare i16 @llvm.atomic.load.min.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6579 declare i32 @llvm.atomic.load.min.i32..p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6580 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]6581</pre>
6582
6583<pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6584 declare i8 @llvm.atomic.load.umax.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6585 declare i16 @llvm.atomic.load.umax.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6586 declare i32 @llvm.atomic.load.umax.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6587 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]6588</pre>
6589
6590<pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6591 declare i8 @llvm.atomic.load.umin.i8..p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6592 declare i16 @llvm.atomic.load.umin.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6593 declare i32 @llvm.atomic.load.umin.i32..p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6594 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]6595</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6596
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6597<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6598<p>These intrinsics takes the signed or unsigned minimum or maximum of
6599 <tt>delta</tt> and the value stored in memory at <tt>ptr</tt>. It yields the
6600 original value at <tt>ptr</tt>.</p>
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6601
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6602<h5>Arguments:</h5>
6603<p>These intrinsics take two arguments, the first a pointer to an integer value
6604 and the second an integer value. The result is also an integer value. These
6605 integer types can have any bit width, but they must all have the same bit
6606 width. The targets may only lower integer representations they support.</p>
6607
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6608<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6609<p>These intrinsics does a series of operations atomically. They first load the
6610 value stored at <tt>ptr</tt>. They then do the signed or unsigned min or
6611 max <tt>delta</tt> and the value, store the result to <tt>ptr</tt>. They
6612 yield the original value stored at <tt>ptr</tt>.</p>
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6613
6614<h5>Examples:</h5>
6615<pre>
6616%ptr = malloc i32
6617 store i32 7, %ptr
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6618%result0 = call i32 @llvm.atomic.load.min.i32.p0i32( i32* %ptr, i32 -2 )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6619 <i>; yields {i32}:result0 = 7</i>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6620%result1 = call i32 @llvm.atomic.load.max.i32.p0i32( i32* %ptr, i32 8 )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6621 <i>; yields {i32}:result1 = -2</i>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6622%result2 = call i32 @llvm.atomic.load.umin.i32.p0i32( i32* %ptr, i32 10 )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6623 <i>; yields {i32}:result2 = 8</i>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6624%result3 = call i32 @llvm.atomic.load.umax.i32.p0i32( i32* %ptr, i32 30 )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6625 <i>; yields {i32}:result3 = 8</i>
6626%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 30</i>
6627</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6628
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6629</div>
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]6630
6631<!-- ======================================================================= -->
6632<div class="doc_subsection">
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6633 <a name="int_general">General Intrinsics</a>
6634</div>
6635
6636<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6637
6638<p>This class of intrinsics is designed to be generic and has no specific
6639 purpose.</p>
6640
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6641</div>
6642
6643<!-- _______________________________________________________________________ -->
6644<div class="doc_subsubsection">
6645 <a name="int_var_annotation">'<tt>llvm.var.annotation</tt>' Intrinsic</a>
6646</div>
6647
6648<div class="doc_text">
6649
6650<h5>Syntax:</h5>
6651<pre>
6652 declare void @llvm.var.annotation(i8* &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
6653</pre>
6654
6655<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6656<p>The '<tt>llvm.var.annotation</tt>' intrinsic.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6657
6658<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6659<p>The first argument is a pointer to a value, the second is a pointer to a
6660 global string, the third is a pointer to a global string which is the source
6661 file name, and the last argument is the line number.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6662
6663<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6664<p>This intrinsic allows annotation of local variables with arbitrary strings.
6665 This can be useful for special purpose optimizations that want to look for
6666 these annotations. These have no other defined use, they are ignored by code
6667 generation and optimization.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6668
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6669</div>
6670
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]6671<!-- _______________________________________________________________________ -->
6672<div class="doc_subsubsection">
Tanya Lattnerc9869b12007-09-21 23:57:59 +0000[diff] [blame]6673 <a name="int_annotation">'<tt>llvm.annotation.*</tt>' Intrinsic</a>
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]6674</div>
6675
6676<div class="doc_text">
6677
6678<h5>Syntax:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6679<p>This is an overloaded intrinsic. You can use '<tt>llvm.annotation</tt>' on
6680 any integer bit width.</p>
6681
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]6682<pre>
Tanya Lattner09161fe2007-09-22 00:03:01 +0000[diff] [blame]6683 declare i8 @llvm.annotation.i8(i8 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
6684 declare i16 @llvm.annotation.i16(i16 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
6685 declare i32 @llvm.annotation.i32(i32 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
6686 declare i64 @llvm.annotation.i64(i64 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
6687 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]6688</pre>
6689
6690<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6691<p>The '<tt>llvm.annotation</tt>' intrinsic.</p>
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]6692
6693<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6694<p>The first argument is an integer value (result of some expression), the
6695 second is a pointer to a global string, the third is a pointer to a global
6696 string which is the source file name, and the last argument is the line
6697 number. It returns the value of the first argument.</p>
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]6698
6699<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6700<p>This intrinsic allows annotations to be put on arbitrary expressions with
6701 arbitrary strings. This can be useful for special purpose optimizations that
6702 want to look for these annotations. These have no other defined use, they
6703 are ignored by code generation and optimization.</p>
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]6704
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]6705</div>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6706
Anton Korobeynikove6e764f2008-01-15 22:31:34 +0000[diff] [blame]6707<!-- _______________________________________________________________________ -->
6708<div class="doc_subsubsection">
6709 <a name="int_trap">'<tt>llvm.trap</tt>' Intrinsic</a>
6710</div>
6711
6712<div class="doc_text">
6713
6714<h5>Syntax:</h5>
6715<pre>
6716 declare void @llvm.trap()
6717</pre>
6718
6719<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6720<p>The '<tt>llvm.trap</tt>' intrinsic.</p>
Anton Korobeynikove6e764f2008-01-15 22:31:34 +0000[diff] [blame]6721
6722<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6723<p>None.</p>
Anton Korobeynikove6e764f2008-01-15 22:31:34 +0000[diff] [blame]6724
6725<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6726<p>This intrinsics is lowered to the target dependent trap instruction. If the
6727 target does not have a trap instruction, this intrinsic will be lowered to
6728 the call of the <tt>abort()</tt> function.</p>
Anton Korobeynikove6e764f2008-01-15 22:31:34 +0000[diff] [blame]6729
Anton Korobeynikove6e764f2008-01-15 22:31:34 +0000[diff] [blame]6730</div>
6731
Bill Wendlinge4164592008-11-19 05:56:17 +0000[diff] [blame]6732<!-- _______________________________________________________________________ -->
6733<div class="doc_subsubsection">
Misha Brukman5dd7f4d2008-11-22 23:55:29 +0000[diff] [blame]6734 <a name="int_stackprotector">'<tt>llvm.stackprotector</tt>' Intrinsic</a>
Bill Wendlinge4164592008-11-19 05:56:17 +0000[diff] [blame]6735</div>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6736
Bill Wendlinge4164592008-11-19 05:56:17 +0000[diff] [blame]6737<div class="doc_text">
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6738
Bill Wendlinge4164592008-11-19 05:56:17 +0000[diff] [blame]6739<h5>Syntax:</h5>
6740<pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6741 declare void @llvm.stackprotector( i8* &lt;guard&gt;, i8** &lt;slot&gt; )
Bill Wendlinge4164592008-11-19 05:56:17 +0000[diff] [blame]6742</pre>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6743
Bill Wendlinge4164592008-11-19 05:56:17 +0000[diff] [blame]6744<h5>Overview:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6745<p>The <tt>llvm.stackprotector</tt> intrinsic takes the <tt>guard</tt> and
6746 stores it onto the stack at <tt>slot</tt>. The stack slot is adjusted to
6747 ensure that it is placed on the stack before local variables.</p>
6748
Bill Wendlinge4164592008-11-19 05:56:17 +0000[diff] [blame]6749<h5>Arguments:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6750<p>The <tt>llvm.stackprotector</tt> intrinsic requires two pointer
6751 arguments. The first argument is the value loaded from the stack
6752 guard <tt>@__stack_chk_guard</tt>. The second variable is an <tt>alloca</tt>
6753 that has enough space to hold the value of the guard.</p>
6754
Bill Wendlinge4164592008-11-19 05:56:17 +0000[diff] [blame]6755<h5>Semantics:</h5>
Bill Wendlingf85859d2009-07-20 02:29:24 +0000[diff] [blame]6756<p>This intrinsic causes the prologue/epilogue inserter to force the position of
6757 the <tt>AllocaInst</tt> stack slot to be before local variables on the
6758 stack. This is to ensure that if a local variable on the stack is
6759 overwritten, it will destroy the value of the guard. When the function exits,
6760 the guard on the stack is checked against the original guard. If they're
6761 different, then the program aborts by calling the <tt>__stack_chk_fail()</tt>
6762 function.</p>
6763
Bill Wendlinge4164592008-11-19 05:56:17 +0000[diff] [blame]6764</div>
6765
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6766<!-- *********************************************************************** -->
6767<hr>
6768<address>
6769 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
Misha Brukman947321d2008-12-11 17:34:48 +0000[diff] [blame]6770 src="http://jigsaw.w3.org/css-validator/images/vcss-blue" alt="Valid CSS"></a>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6771 <a href="http://validator.w3.org/check/referer"><img
Misha Brukman947321d2008-12-11 17:34:48 +0000[diff] [blame]6772 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]6773
6774 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
6775 <a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br>
6776 Last modified: $Date$
6777</address>
Chris Lattner08497ce2008-01-04 04:33:49 +0000[diff] [blame]6778
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6779</body>
6780</html>