blob: 1ee7f2972efa2a7afcb2d556c1c0f6a2d767bd34 [file] [log] [blame]
Misha Brukmanc501f552004-03-01 17:47:27 +00001<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
2 "http://www.w3.org/TR/html4/strict.dtd">
Misha Brukman76307852003-11-08 01:05:38 +00003<html>
4<head>
5 <title>LLVM Assembly Language Reference Manual</title>
Reid Spencercb84e432004-08-26 20:44:00 +00006 <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
7 <meta name="author" content="Chris Lattner">
8 <meta name="description"
9 content="LLVM Assembly Language Reference Manual.">
Misha Brukman76307852003-11-08 01:05:38 +000010 <link rel="stylesheet" href="llvm.css" type="text/css">
11</head>
Chris Lattner757528b0b2004-05-23 21:06:01 +000012
Misha Brukman76307852003-11-08 01:05:38 +000013<body>
Chris Lattner757528b0b2004-05-23 21:06:01 +000014
Chris Lattner48b383b02003-11-25 01:02:51 +000015<div class="doc_title"> LLVM Language Reference Manual </div>
Chris Lattner2f7c9632001-06-06 20:29:01 +000016<ol>
Misha Brukman76307852003-11-08 01:05:38 +000017 <li><a href="#abstract">Abstract</a></li>
18 <li><a href="#introduction">Introduction</a></li>
19 <li><a href="#identifiers">Identifiers</a></li>
Chris Lattner6af02f32004-12-09 16:11:40 +000020 <li><a href="#highlevel">High Level Structure</a>
21 <ol>
22 <li><a href="#modulestructure">Module Structure</a></li>
Chris Lattnerd79749a2004-12-09 16:36:40 +000023 <li><a href="#linkage">Linkage Types</a></li>
Chris Lattner0132aff2005-05-06 22:57:40 +000024 <li><a href="#callingconv">Calling Conventions</a></li>
Chris Lattner6af02f32004-12-09 16:11:40 +000025 <li><a href="#globalvars">Global Variables</a></li>
Chris Lattner91c15c42006-01-23 23:23:47 +000026 <li><a href="#functionstructure">Functions</a></li>
Anton Korobeynikov546ea7e2007-04-29 18:02:48 +000027 <li><a href="#aliasstructure">Aliases</a>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +000028 <li><a href="#paramattrs">Parameter Attributes</a></li>
Gordon Henriksen71183b62007-12-10 03:18:06 +000029 <li><a href="#gc">Garbage Collector Names</a></li>
Chris Lattner91c15c42006-01-23 23:23:47 +000030 <li><a href="#moduleasm">Module-Level Inline Assembly</a></li>
Reid Spencer50c723a2007-02-19 23:54:10 +000031 <li><a href="#datalayout">Data Layout</a></li>
Chris Lattner6af02f32004-12-09 16:11:40 +000032 </ol>
33 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +000034 <li><a href="#typesystem">Type System</a>
35 <ol>
Chris Lattner7824d182008-01-04 04:32:38 +000036 <li><a href="#t_classifications">Type Classifications</a></li>
Robert Bocchino820bc75b2006-02-17 21:18:08 +000037 <li><a href="#t_primitive">Primitive Types</a>
Chris Lattner48b383b02003-11-25 01:02:51 +000038 <ol>
Chris Lattner7824d182008-01-04 04:32:38 +000039 <li><a href="#t_floating">Floating Point Types</a></li>
40 <li><a href="#t_void">Void Type</a></li>
41 <li><a href="#t_label">Label Type</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +000042 </ol>
43 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +000044 <li><a href="#t_derived">Derived Types</a>
45 <ol>
Chris Lattner9a2e3cb2007-12-18 06:18:21 +000046 <li><a href="#t_integer">Integer Type</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +000047 <li><a href="#t_array">Array Type</a></li>
Misha Brukman76307852003-11-08 01:05:38 +000048 <li><a href="#t_function">Function Type</a></li>
49 <li><a href="#t_pointer">Pointer Type</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +000050 <li><a href="#t_struct">Structure Type</a></li>
Andrew Lenharth8df88e22006-12-08 17:13:00 +000051 <li><a href="#t_pstruct">Packed Structure Type</a></li>
Reid Spencer404a3252007-02-15 03:07:05 +000052 <li><a href="#t_vector">Vector Type</a></li>
Chris Lattner37b6b092005-04-25 17:34:15 +000053 <li><a href="#t_opaque">Opaque Type</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +000054 </ol>
55 </li>
56 </ol>
57 </li>
Chris Lattner6af02f32004-12-09 16:11:40 +000058 <li><a href="#constants">Constants</a>
Chris Lattner74d3f822004-12-09 17:30:23 +000059 <ol>
60 <li><a href="#simpleconstants">Simple Constants</a>
61 <li><a href="#aggregateconstants">Aggregate Constants</a>
62 <li><a href="#globalconstants">Global Variable and Function Addresses</a>
63 <li><a href="#undefvalues">Undefined Values</a>
64 <li><a href="#constantexprs">Constant Expressions</a>
65 </ol>
Chris Lattner48b383b02003-11-25 01:02:51 +000066 </li>
Chris Lattner98f013c2006-01-25 23:47:57 +000067 <li><a href="#othervalues">Other Values</a>
68 <ol>
69 <li><a href="#inlineasm">Inline Assembler Expressions</a>
70 </ol>
71 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +000072 <li><a href="#instref">Instruction Reference</a>
73 <ol>
74 <li><a href="#terminators">Terminator Instructions</a>
75 <ol>
Chris Lattner48b383b02003-11-25 01:02:51 +000076 <li><a href="#i_ret">'<tt>ret</tt>' Instruction</a></li>
77 <li><a href="#i_br">'<tt>br</tt>' Instruction</a></li>
Misha Brukman76307852003-11-08 01:05:38 +000078 <li><a href="#i_switch">'<tt>switch</tt>' Instruction</a></li>
79 <li><a href="#i_invoke">'<tt>invoke</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +000080 <li><a href="#i_unwind">'<tt>unwind</tt>' Instruction</a></li>
Chris Lattner08b7d5b2004-10-16 18:04:13 +000081 <li><a href="#i_unreachable">'<tt>unreachable</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +000082 </ol>
83 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +000084 <li><a href="#binaryops">Binary Operations</a>
85 <ol>
Chris Lattner48b383b02003-11-25 01:02:51 +000086 <li><a href="#i_add">'<tt>add</tt>' Instruction</a></li>
87 <li><a href="#i_sub">'<tt>sub</tt>' Instruction</a></li>
88 <li><a href="#i_mul">'<tt>mul</tt>' Instruction</a></li>
Reid Spencer7e80b0b2006-10-26 06:15:43 +000089 <li><a href="#i_udiv">'<tt>udiv</tt>' Instruction</a></li>
90 <li><a href="#i_sdiv">'<tt>sdiv</tt>' Instruction</a></li>
91 <li><a href="#i_fdiv">'<tt>fdiv</tt>' Instruction</a></li>
Reid Spencer7eb55b32006-11-02 01:53:59 +000092 <li><a href="#i_urem">'<tt>urem</tt>' Instruction</a></li>
93 <li><a href="#i_srem">'<tt>srem</tt>' Instruction</a></li>
94 <li><a href="#i_frem">'<tt>frem</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +000095 </ol>
96 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +000097 <li><a href="#bitwiseops">Bitwise Binary Operations</a>
98 <ol>
Reid Spencer2ab01932007-02-02 13:57:07 +000099 <li><a href="#i_shl">'<tt>shl</tt>' Instruction</a></li>
100 <li><a href="#i_lshr">'<tt>lshr</tt>' Instruction</a></li>
101 <li><a href="#i_ashr">'<tt>ashr</tt>' Instruction</a></li>
Misha Brukman76307852003-11-08 01:05:38 +0000102 <li><a href="#i_and">'<tt>and</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000103 <li><a href="#i_or">'<tt>or</tt>' Instruction</a></li>
Misha Brukman76307852003-11-08 01:05:38 +0000104 <li><a href="#i_xor">'<tt>xor</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000105 </ol>
106 </li>
Chris Lattnerce83bff2006-04-08 23:07:04 +0000107 <li><a href="#vectorops">Vector Operations</a>
108 <ol>
109 <li><a href="#i_extractelement">'<tt>extractelement</tt>' Instruction</a></li>
110 <li><a href="#i_insertelement">'<tt>insertelement</tt>' Instruction</a></li>
111 <li><a href="#i_shufflevector">'<tt>shufflevector</tt>' Instruction</a></li>
Chris Lattnerce83bff2006-04-08 23:07:04 +0000112 </ol>
113 </li>
Dan Gohmanb9d66602008-05-12 23:51:09 +0000114 <li><a href="#aggregateops">Aggregate Operations</a>
115 <ol>
116 <li><a href="#i_extractvalue">'<tt>extractvalue</tt>' Instruction</a></li>
117 <li><a href="#i_insertvalue">'<tt>insertvalue</tt>' Instruction</a></li>
118 </ol>
119 </li>
Chris Lattner6ab66722006-08-15 00:45:58 +0000120 <li><a href="#memoryops">Memory Access and Addressing Operations</a>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000121 <ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000122 <li><a href="#i_malloc">'<tt>malloc</tt>' Instruction</a></li>
123 <li><a href="#i_free">'<tt>free</tt>' Instruction</a></li>
124 <li><a href="#i_alloca">'<tt>alloca</tt>' Instruction</a></li>
Robert Bocchino820bc75b2006-02-17 21:18:08 +0000125 <li><a href="#i_load">'<tt>load</tt>' Instruction</a></li>
126 <li><a href="#i_store">'<tt>store</tt>' Instruction</a></li>
127 <li><a href="#i_getelementptr">'<tt>getelementptr</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000128 </ol>
129 </li>
Reid Spencer97c5fa42006-11-08 01:18:52 +0000130 <li><a href="#convertops">Conversion Operations</a>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000131 <ol>
132 <li><a href="#i_trunc">'<tt>trunc .. to</tt>' Instruction</a></li>
133 <li><a href="#i_zext">'<tt>zext .. to</tt>' Instruction</a></li>
134 <li><a href="#i_sext">'<tt>sext .. to</tt>' Instruction</a></li>
135 <li><a href="#i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a></li>
136 <li><a href="#i_fpext">'<tt>fpext .. to</tt>' Instruction</a></li>
Reid Spencer51b07252006-11-09 23:03:26 +0000137 <li><a href="#i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a></li>
138 <li><a href="#i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a></li>
139 <li><a href="#i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a></li>
140 <li><a href="#i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a></li>
Reid Spencerb7344ff2006-11-11 21:00:47 +0000141 <li><a href="#i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a></li>
142 <li><a href="#i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a></li>
Reid Spencer5b950642006-11-11 23:08:07 +0000143 <li><a href="#i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a></li>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000144 </ol>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000145 <li><a href="#otherops">Other Operations</a>
146 <ol>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000147 <li><a href="#i_icmp">'<tt>icmp</tt>' Instruction</a></li>
148 <li><a href="#i_fcmp">'<tt>fcmp</tt>' Instruction</a></li>
Nate Begemand2195702008-05-12 19:01:56 +0000149 <li><a href="#i_vicmp">'<tt>vicmp</tt>' Instruction</a></li>
150 <li><a href="#i_vfcmp">'<tt>vfcmp</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000151 <li><a href="#i_phi">'<tt>phi</tt>' Instruction</a></li>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000152 <li><a href="#i_select">'<tt>select</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000153 <li><a href="#i_call">'<tt>call</tt>' Instruction</a></li>
Chris Lattner33337472006-01-13 23:26:01 +0000154 <li><a href="#i_va_arg">'<tt>va_arg</tt>' Instruction</a></li>
Devang Pateld6cff512008-03-10 20:49:15 +0000155 <li><a href="#i_getresult">'<tt>getresult</tt>' Instruction</a></li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000156 </ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000157 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000158 </ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000159 </li>
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000160 <li><a href="#intrinsics">Intrinsic Functions</a>
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000161 <ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000162 <li><a href="#int_varargs">Variable Argument Handling Intrinsics</a>
163 <ol>
Reid Spencer96a5f022007-04-04 02:42:35 +0000164 <li><a href="#int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a></li>
165 <li><a href="#int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a></li>
166 <li><a href="#int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000167 </ol>
168 </li>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000169 <li><a href="#int_gc">Accurate Garbage Collection Intrinsics</a>
170 <ol>
Reid Spencer96a5f022007-04-04 02:42:35 +0000171 <li><a href="#int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a></li>
172 <li><a href="#int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a></li>
173 <li><a href="#int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a></li>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000174 </ol>
175 </li>
Chris Lattner3649c3a2004-02-14 04:08:35 +0000176 <li><a href="#int_codegen">Code Generator Intrinsics</a>
177 <ol>
Reid Spencer96a5f022007-04-04 02:42:35 +0000178 <li><a href="#int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a></li>
179 <li><a href="#int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a></li>
180 <li><a href="#int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a></li>
181 <li><a href="#int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a></li>
182 <li><a href="#int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a></li>
183 <li><a href="#int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a></li>
184 <li><a href="#int_readcyclecounter"><tt>llvm.readcyclecounter</tt>' Intrinsic</a></li>
John Criswellaa1c3c12004-04-09 16:43:20 +0000185 </ol>
186 </li>
Chris Lattnerfee11462004-02-12 17:01:32 +0000187 <li><a href="#int_libc">Standard C Library Intrinsics</a>
188 <ol>
Reid Spencer96a5f022007-04-04 02:42:35 +0000189 <li><a href="#int_memcpy">'<tt>llvm.memcpy.*</tt>' Intrinsic</a></li>
190 <li><a href="#int_memmove">'<tt>llvm.memmove.*</tt>' Intrinsic</a></li>
191 <li><a href="#int_memset">'<tt>llvm.memset.*</tt>' Intrinsic</a></li>
192 <li><a href="#int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a></li>
193 <li><a href="#int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a></li>
Dan Gohmanb6324c12007-10-15 20:30:11 +0000194 <li><a href="#int_sin">'<tt>llvm.sin.*</tt>' Intrinsic</a></li>
195 <li><a href="#int_cos">'<tt>llvm.cos.*</tt>' Intrinsic</a></li>
196 <li><a href="#int_pow">'<tt>llvm.pow.*</tt>' Intrinsic</a></li>
Chris Lattnerfee11462004-02-12 17:01:32 +0000197 </ol>
198 </li>
Nate Begeman0f223bb2006-01-13 23:26:38 +0000199 <li><a href="#int_manip">Bit Manipulation Intrinsics</a>
Andrew Lenharth1d463522005-05-03 18:01:48 +0000200 <ol>
Reid Spencer96a5f022007-04-04 02:42:35 +0000201 <li><a href="#int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a></li>
Chris Lattnerb748c672006-01-16 22:34:14 +0000202 <li><a href="#int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic </a></li>
203 <li><a href="#int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic </a></li>
204 <li><a href="#int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic </a></li>
Reid Spencer5bf54c82007-04-11 23:23:49 +0000205 <li><a href="#int_part_select">'<tt>llvm.part.select.*</tt>' Intrinsic </a></li>
206 <li><a href="#int_part_set">'<tt>llvm.part.set.*</tt>' Intrinsic </a></li>
Andrew Lenharth1d463522005-05-03 18:01:48 +0000207 </ol>
208 </li>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000209 <li><a href="#int_debugger">Debugger intrinsics</a></li>
Jim Laskey2211f492007-03-14 19:31:19 +0000210 <li><a href="#int_eh">Exception Handling intrinsics</a></li>
Duncan Sands86e01192007-09-11 14:10:23 +0000211 <li><a href="#int_trampoline">Trampoline Intrinsic</a>
Duncan Sands644f9172007-07-27 12:58:54 +0000212 <ol>
213 <li><a href="#int_it">'<tt>llvm.init.trampoline</tt>' Intrinsic</a></li>
Duncan Sands644f9172007-07-27 12:58:54 +0000214 </ol>
215 </li>
Andrew Lenharth9b254ee2008-02-16 01:24:58 +0000216 <li><a href="#int_atomics">Atomic intrinsics</a>
217 <ol>
Andrew Lenharth95528942008-02-21 06:45:13 +0000218 <li><a href="#int_memory_barrier"><tt>llvm.memory_barrier</tt></a></li>
Mon P Wang6a490372008-06-25 08:15:39 +0000219 <li><a href="#int_atomic_cmp_swap"><tt>llvm.atomic.cmp.swap</tt></a></li>
Andrew Lenharth95528942008-02-21 06:45:13 +0000220 <li><a href="#int_atomic_swap"><tt>llvm.atomic.swap</tt></a></li>
Mon P Wang6a490372008-06-25 08:15:39 +0000221 <li><a href="#int_atomic_load_add"><tt>llvm.atomic.load.add</tt></a></li>
222 <li><a href="#int_atomic_load_sub"><tt>llvm.atomic.load.sub</tt></a></li>
223 <li><a href="#int_atomic_load_and"><tt>llvm.atomic.load.and</tt></a></li>
224 <li><a href="#int_atomic_load_nand"><tt>llvm.atomic.load.nand</tt></a></li>
225 <li><a href="#int_atomic_load_or"><tt>llvm.atomic.load.or</tt></a></li>
226 <li><a href="#int_atomic_load_xor"><tt>llvm.atomic.load.xor</tt></a></li>
227 <li><a href="#int_atomic_load_max"><tt>llvm.atomic.load.max</tt></a></li>
228 <li><a href="#int_atomic_load_min"><tt>llvm.atomic.load.min</tt></a></li>
229 <li><a href="#int_atomic_load_umax"><tt>llvm.atomic.load.umax</tt></a></li>
230 <li><a href="#int_atomic_load_umin"><tt>llvm.atomic.load.umin</tt></a></li>
Andrew Lenharth9b254ee2008-02-16 01:24:58 +0000231 </ol>
232 </li>
Reid Spencer5b2cb0f2007-07-20 19:59:11 +0000233 <li><a href="#int_general">General intrinsics</a>
Tanya Lattnercb1b9602007-06-15 20:50:54 +0000234 <ol>
Reid Spencer5b2cb0f2007-07-20 19:59:11 +0000235 <li><a href="#int_var_annotation">
Tanya Lattner08abc812007-09-22 00:01:26 +0000236 <tt>llvm.var.annotation</tt>' Intrinsic</a></li>
Tanya Lattner293c0372007-09-21 22:59:12 +0000237 <li><a href="#int_annotation">
Tanya Lattner08abc812007-09-22 00:01:26 +0000238 <tt>llvm.annotation.*</tt>' Intrinsic</a></li>
Anton Korobeynikov06cbb652008-01-15 22:31:34 +0000239 <li><a href="#int_trap">
240 <tt>llvm.trap</tt>' Intrinsic</a></li>
Tanya Lattner293c0372007-09-21 22:59:12 +0000241 </ol>
Tanya Lattnercb1b9602007-06-15 20:50:54 +0000242 </li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000243 </ol>
244 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000245</ol>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000246
247<div class="doc_author">
248 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>
249 and <a href="mailto:vadve@cs.uiuc.edu">Vikram Adve</a></p>
Misha Brukman76307852003-11-08 01:05:38 +0000250</div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000251
Chris Lattner2f7c9632001-06-06 20:29:01 +0000252<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000253<div class="doc_section"> <a name="abstract">Abstract </a></div>
254<!-- *********************************************************************** -->
Chris Lattner757528b0b2004-05-23 21:06:01 +0000255
Misha Brukman76307852003-11-08 01:05:38 +0000256<div class="doc_text">
Chris Lattner48b383b02003-11-25 01:02:51 +0000257<p>This document is a reference manual for the LLVM assembly language.
258LLVM is an SSA based representation that provides type safety,
259low-level operations, flexibility, and the capability of representing
260'all' high-level languages cleanly. It is the common code
261representation used throughout all phases of the LLVM compilation
262strategy.</p>
Misha Brukman76307852003-11-08 01:05:38 +0000263</div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000264
Chris Lattner2f7c9632001-06-06 20:29:01 +0000265<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000266<div class="doc_section"> <a name="introduction">Introduction</a> </div>
267<!-- *********************************************************************** -->
Chris Lattner757528b0b2004-05-23 21:06:01 +0000268
Misha Brukman76307852003-11-08 01:05:38 +0000269<div class="doc_text">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000270
Chris Lattner48b383b02003-11-25 01:02:51 +0000271<p>The LLVM code representation is designed to be used in three
Gabor Greifa54634a2007-07-06 22:07:22 +0000272different forms: as an in-memory compiler IR, as an on-disk bitcode
Chris Lattner48b383b02003-11-25 01:02:51 +0000273representation (suitable for fast loading by a Just-In-Time compiler),
274and as a human readable assembly language representation. This allows
275LLVM to provide a powerful intermediate representation for efficient
276compiler transformations and analysis, while providing a natural means
277to debug and visualize the transformations. The three different forms
278of LLVM are all equivalent. This document describes the human readable
279representation and notation.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000280
John Criswell4a3327e2005-05-13 22:25:59 +0000281<p>The LLVM representation aims to be light-weight and low-level
Chris Lattner48b383b02003-11-25 01:02:51 +0000282while being expressive, typed, and extensible at the same time. It
283aims to be a "universal IR" of sorts, by being at a low enough level
284that high-level ideas may be cleanly mapped to it (similar to how
285microprocessors are "universal IR's", allowing many source languages to
286be mapped to them). By providing type information, LLVM can be used as
287the target of optimizations: for example, through pointer analysis, it
288can be proven that a C automatic variable is never accessed outside of
289the current function... allowing it to be promoted to a simple SSA
290value instead of a memory location.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000291
Misha Brukman76307852003-11-08 01:05:38 +0000292</div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000293
Chris Lattner2f7c9632001-06-06 20:29:01 +0000294<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000295<div class="doc_subsubsection"> <a name="wellformed">Well-Formedness</a> </div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000296
Misha Brukman76307852003-11-08 01:05:38 +0000297<div class="doc_text">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000298
Chris Lattner48b383b02003-11-25 01:02:51 +0000299<p>It is important to note that this document describes 'well formed'
300LLVM assembly language. There is a difference between what the parser
301accepts and what is considered 'well formed'. For example, the
302following instruction is syntactically okay, but not well formed:</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000303
Bill Wendling3716c5d2007-05-29 09:04:49 +0000304<div class="doc_code">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000305<pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000306%x = <a href="#i_add">add</a> i32 1, %x
Chris Lattner757528b0b2004-05-23 21:06:01 +0000307</pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000308</div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000309
Chris Lattner48b383b02003-11-25 01:02:51 +0000310<p>...because the definition of <tt>%x</tt> does not dominate all of
311its uses. The LLVM infrastructure provides a verification pass that may
312be used to verify that an LLVM module is well formed. This pass is
John Criswell4a3327e2005-05-13 22:25:59 +0000313automatically run by the parser after parsing input assembly and by
Gabor Greifa54634a2007-07-06 22:07:22 +0000314the optimizer before it outputs bitcode. The violations pointed out
Chris Lattner48b383b02003-11-25 01:02:51 +0000315by the verifier pass indicate bugs in transformation passes or input to
316the parser.</p>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000317</div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000318
Chris Lattner87a3dbe2007-10-03 17:34:29 +0000319<!-- Describe the typesetting conventions here. -->
Chris Lattner757528b0b2004-05-23 21:06:01 +0000320
Chris Lattner2f7c9632001-06-06 20:29:01 +0000321<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000322<div class="doc_section"> <a name="identifiers">Identifiers</a> </div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000323<!-- *********************************************************************** -->
Chris Lattner757528b0b2004-05-23 21:06:01 +0000324
Misha Brukman76307852003-11-08 01:05:38 +0000325<div class="doc_text">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000326
Reid Spencerb23b65f2007-08-07 14:34:28 +0000327 <p>LLVM identifiers come in two basic types: global and local. Global
328 identifiers (functions, global variables) begin with the @ character. Local
329 identifiers (register names, types) begin with the % character. Additionally,
330 there are three different formats for identifiers, for different purposes:
Chris Lattner757528b0b2004-05-23 21:06:01 +0000331
Chris Lattner2f7c9632001-06-06 20:29:01 +0000332<ol>
Reid Spencerb23b65f2007-08-07 14:34:28 +0000333 <li>Named values are represented as a string of characters with their prefix.
334 For example, %foo, @DivisionByZero, %a.really.long.identifier. The actual
335 regular expression used is '<tt>[%@][a-zA-Z$._][a-zA-Z$._0-9]*</tt>'.
Chris Lattnerd79749a2004-12-09 16:36:40 +0000336 Identifiers which require other characters in their names can be surrounded
Reid Spencerb23b65f2007-08-07 14:34:28 +0000337 with quotes. In this way, anything except a <tt>&quot;</tt> character can
338 be used in a named value.</li>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000339
Reid Spencerb23b65f2007-08-07 14:34:28 +0000340 <li>Unnamed values are represented as an unsigned numeric value with their
341 prefix. For example, %12, @2, %44.</li>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000342
Reid Spencer8f08d802004-12-09 18:02:53 +0000343 <li>Constants, which are described in a <a href="#constants">section about
344 constants</a>, below.</li>
Misha Brukman76307852003-11-08 01:05:38 +0000345</ol>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000346
Reid Spencerb23b65f2007-08-07 14:34:28 +0000347<p>LLVM requires that values start with a prefix for two reasons: Compilers
Chris Lattnerd79749a2004-12-09 16:36:40 +0000348don't need to worry about name clashes with reserved words, and the set of
349reserved words may be expanded in the future without penalty. Additionally,
350unnamed identifiers allow a compiler to quickly come up with a temporary
351variable without having to avoid symbol table conflicts.</p>
352
Chris Lattner48b383b02003-11-25 01:02:51 +0000353<p>Reserved words in LLVM are very similar to reserved words in other
Reid Spencer5b950642006-11-11 23:08:07 +0000354languages. There are keywords for different opcodes
355('<tt><a href="#i_add">add</a></tt>',
356 '<tt><a href="#i_bitcast">bitcast</a></tt>',
357 '<tt><a href="#i_ret">ret</a></tt>', etc...), for primitive type names ('<tt><a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000358href="#t_void">void</a></tt>', '<tt><a href="#t_primitive">i32</a></tt>', etc...),
Chris Lattnerd79749a2004-12-09 16:36:40 +0000359and others. These reserved words cannot conflict with variable names, because
Reid Spencerb23b65f2007-08-07 14:34:28 +0000360none of them start with a prefix character ('%' or '@').</p>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000361
362<p>Here is an example of LLVM code to multiply the integer variable
363'<tt>%X</tt>' by 8:</p>
364
Misha Brukman76307852003-11-08 01:05:38 +0000365<p>The easy way:</p>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000366
Bill Wendling3716c5d2007-05-29 09:04:49 +0000367<div class="doc_code">
Chris Lattnerd79749a2004-12-09 16:36:40 +0000368<pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000369%result = <a href="#i_mul">mul</a> i32 %X, 8
Chris Lattnerd79749a2004-12-09 16:36:40 +0000370</pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000371</div>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000372
Misha Brukman76307852003-11-08 01:05:38 +0000373<p>After strength reduction:</p>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000374
Bill Wendling3716c5d2007-05-29 09:04:49 +0000375<div class="doc_code">
Chris Lattnerd79749a2004-12-09 16:36:40 +0000376<pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000377%result = <a href="#i_shl">shl</a> i32 %X, i8 3
Chris Lattnerd79749a2004-12-09 16:36:40 +0000378</pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000379</div>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000380
Misha Brukman76307852003-11-08 01:05:38 +0000381<p>And the hard way:</p>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000382
Bill Wendling3716c5d2007-05-29 09:04:49 +0000383<div class="doc_code">
Chris Lattnerd79749a2004-12-09 16:36:40 +0000384<pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000385<a href="#i_add">add</a> i32 %X, %X <i>; yields {i32}:%0</i>
386<a href="#i_add">add</a> i32 %0, %0 <i>; yields {i32}:%1</i>
387%result = <a href="#i_add">add</a> i32 %1, %1
Chris Lattnerd79749a2004-12-09 16:36:40 +0000388</pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000389</div>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000390
Chris Lattner48b383b02003-11-25 01:02:51 +0000391<p>This last way of multiplying <tt>%X</tt> by 8 illustrates several
392important lexical features of LLVM:</p>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000393
Chris Lattner2f7c9632001-06-06 20:29:01 +0000394<ol>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000395
396 <li>Comments are delimited with a '<tt>;</tt>' and go until the end of
397 line.</li>
398
399 <li>Unnamed temporaries are created when the result of a computation is not
400 assigned to a named value.</li>
401
Misha Brukman76307852003-11-08 01:05:38 +0000402 <li>Unnamed temporaries are numbered sequentially</li>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000403
Misha Brukman76307852003-11-08 01:05:38 +0000404</ol>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000405
John Criswell02fdc6f2005-05-12 16:52:32 +0000406<p>...and it also shows a convention that we follow in this document. When
Chris Lattnerd79749a2004-12-09 16:36:40 +0000407demonstrating instructions, we will follow an instruction with a comment that
408defines the type and name of value produced. Comments are shown in italic
409text.</p>
410
Misha Brukman76307852003-11-08 01:05:38 +0000411</div>
Chris Lattner6af02f32004-12-09 16:11:40 +0000412
413<!-- *********************************************************************** -->
414<div class="doc_section"> <a name="highlevel">High Level Structure</a> </div>
415<!-- *********************************************************************** -->
416
417<!-- ======================================================================= -->
418<div class="doc_subsection"> <a name="modulestructure">Module Structure</a>
419</div>
420
421<div class="doc_text">
422
423<p>LLVM programs are composed of "Module"s, each of which is a
424translation unit of the input programs. Each module consists of
425functions, global variables, and symbol table entries. Modules may be
426combined together with the LLVM linker, which merges function (and
427global variable) definitions, resolves forward declarations, and merges
428symbol table entries. Here is an example of the "hello world" module:</p>
429
Bill Wendling3716c5d2007-05-29 09:04:49 +0000430<div class="doc_code">
Chris Lattner6af02f32004-12-09 16:11:40 +0000431<pre><i>; Declare the string constant as a global constant...</i>
Chris Lattner2b0bf4f2007-06-12 17:00:26 +0000432<a href="#identifiers">@.LC0</a> = <a href="#linkage_internal">internal</a> <a
433 href="#globalvars">constant</a> <a href="#t_array">[13 x i8]</a> c"hello world\0A\00" <i>; [13 x i8]*</i>
Chris Lattner6af02f32004-12-09 16:11:40 +0000434
435<i>; External declaration of the puts function</i>
Chris Lattner2b0bf4f2007-06-12 17:00:26 +0000436<a href="#functionstructure">declare</a> i32 @puts(i8 *) <i>; i32(i8 *)* </i>
Chris Lattner6af02f32004-12-09 16:11:40 +0000437
438<i>; Definition of main function</i>
Chris Lattner2b0bf4f2007-06-12 17:00:26 +0000439define i32 @main() { <i>; i32()* </i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000440 <i>; Convert [13x i8 ]* to i8 *...</i>
Chris Lattner6af02f32004-12-09 16:11:40 +0000441 %cast210 = <a
Chris Lattner2150cde2007-06-12 17:01:15 +0000442 href="#i_getelementptr">getelementptr</a> [13 x i8 ]* @.LC0, i64 0, i64 0 <i>; i8 *</i>
Chris Lattner6af02f32004-12-09 16:11:40 +0000443
444 <i>; Call puts function to write out the string to stdout...</i>
445 <a
Chris Lattner2b0bf4f2007-06-12 17:00:26 +0000446 href="#i_call">call</a> i32 @puts(i8 * %cast210) <i>; i32</i>
Chris Lattner6af02f32004-12-09 16:11:40 +0000447 <a
Bill Wendling3716c5d2007-05-29 09:04:49 +0000448 href="#i_ret">ret</a> i32 0<br>}<br>
449</pre>
450</div>
Chris Lattner6af02f32004-12-09 16:11:40 +0000451
452<p>This example is made up of a <a href="#globalvars">global variable</a>
453named "<tt>.LC0</tt>", an external declaration of the "<tt>puts</tt>"
454function, and a <a href="#functionstructure">function definition</a>
455for "<tt>main</tt>".</p>
456
Chris Lattnerd79749a2004-12-09 16:36:40 +0000457<p>In general, a module is made up of a list of global values,
458where both functions and global variables are global values. Global values are
459represented by a pointer to a memory location (in this case, a pointer to an
460array of char, and a pointer to a function), and have one of the following <a
461href="#linkage">linkage types</a>.</p>
Chris Lattner6af02f32004-12-09 16:11:40 +0000462
Chris Lattnerd79749a2004-12-09 16:36:40 +0000463</div>
464
465<!-- ======================================================================= -->
466<div class="doc_subsection">
467 <a name="linkage">Linkage Types</a>
468</div>
469
470<div class="doc_text">
471
472<p>
473All Global Variables and Functions have one of the following types of linkage:
474</p>
Chris Lattner6af02f32004-12-09 16:11:40 +0000475
476<dl>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000477
Dale Johannesen4188aad2008-05-23 23:13:41 +0000478 <dt><tt><b><a name="linkage_internal">internal</a></b></tt>: </dt>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000479
480 <dd>Global values with internal linkage are only directly accessible by
481 objects in the current module. In particular, linking code into a module with
482 an internal global value may cause the internal to be renamed as necessary to
483 avoid collisions. Because the symbol is internal to the module, all
484 references can be updated. This corresponds to the notion of the
Chris Lattnere20b4702007-01-14 06:51:48 +0000485 '<tt>static</tt>' keyword in C.
Chris Lattner6af02f32004-12-09 16:11:40 +0000486 </dd>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000487
Chris Lattner6af02f32004-12-09 16:11:40 +0000488 <dt><tt><b><a name="linkage_linkonce">linkonce</a></b></tt>: </dt>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000489
Chris Lattnere20b4702007-01-14 06:51:48 +0000490 <dd>Globals with "<tt>linkonce</tt>" linkage are merged with other globals of
491 the same name when linkage occurs. This is typically used to implement
492 inline functions, templates, or other code which must be generated in each
493 translation unit that uses it. Unreferenced <tt>linkonce</tt> globals are
494 allowed to be discarded.
Chris Lattner6af02f32004-12-09 16:11:40 +0000495 </dd>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000496
Dale Johannesen4188aad2008-05-23 23:13:41 +0000497 <dt><tt><b><a name="linkage_common">common</a></b></tt>: </dt>
498
499 <dd>"<tt>common</tt>" linkage is exactly the same as <tt>linkonce</tt>
500 linkage, except that unreferenced <tt>common</tt> globals may not be
501 discarded. This is used for globals that may be emitted in multiple
502 translation units, but that are not guaranteed to be emitted into every
503 translation unit that uses them. One example of this is tentative
504 definitions in C, such as "<tt>int X;</tt>" at global scope.
505 </dd>
506
Chris Lattner6af02f32004-12-09 16:11:40 +0000507 <dt><tt><b><a name="linkage_weak">weak</a></b></tt>: </dt>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000508
Dale Johannesen4188aad2008-05-23 23:13:41 +0000509 <dd>"<tt>weak</tt>" linkage is the same as <tt>common</tt> linkage, except
510 that some targets may choose to emit different assembly sequences for them
511 for target-dependent reasons. This is used for globals that are declared
512 "weak" in C source code.
Chris Lattner6af02f32004-12-09 16:11:40 +0000513 </dd>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000514
Chris Lattner6af02f32004-12-09 16:11:40 +0000515 <dt><tt><b><a name="linkage_appending">appending</a></b></tt>: </dt>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000516
517 <dd>"<tt>appending</tt>" linkage may only be applied to global variables of
518 pointer to array type. When two global variables with appending linkage are
519 linked together, the two global arrays are appended together. This is the
520 LLVM, typesafe, equivalent of having the system linker append together
521 "sections" with identical names when .o files are linked.
Chris Lattner6af02f32004-12-09 16:11:40 +0000522 </dd>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000523
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000524 <dt><tt><b><a name="linkage_externweak">extern_weak</a></b></tt>: </dt>
525 <dd>The semantics of this linkage follow the ELF model: the symbol is weak
526 until linked, if not linked, the symbol becomes null instead of being an
527 undefined reference.
528 </dd>
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000529
Chris Lattner6af02f32004-12-09 16:11:40 +0000530 <dt><tt><b><a name="linkage_external">externally visible</a></b></tt>:</dt>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000531
532 <dd>If none of the above identifiers are used, the global is externally
533 visible, meaning that it participates in linkage and can be used to resolve
534 external symbol references.
Chris Lattner6af02f32004-12-09 16:11:40 +0000535 </dd>
Reid Spencer7972c472007-04-11 23:49:50 +0000536</dl>
Anton Korobeynikovd61d39e2006-09-14 18:23:27 +0000537
Anton Korobeynikovd61d39e2006-09-14 18:23:27 +0000538 <p>
539 The next two types of linkage are targeted for Microsoft Windows platform
540 only. They are designed to support importing (exporting) symbols from (to)
541 DLLs.
542 </p>
543
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000544 <dl>
Anton Korobeynikovd61d39e2006-09-14 18:23:27 +0000545 <dt><tt><b><a name="linkage_dllimport">dllimport</a></b></tt>: </dt>
546
547 <dd>"<tt>dllimport</tt>" linkage causes the compiler to reference a function
548 or variable via a global pointer to a pointer that is set up by the DLL
549 exporting the symbol. On Microsoft Windows targets, the pointer name is
550 formed by combining <code>_imp__</code> and the function or variable name.
551 </dd>
552
553 <dt><tt><b><a name="linkage_dllexport">dllexport</a></b></tt>: </dt>
554
555 <dd>"<tt>dllexport</tt>" linkage causes the compiler to provide a global
556 pointer to a pointer in a DLL, so that it can be referenced with the
557 <tt>dllimport</tt> attribute. On Microsoft Windows targets, the pointer
558 name is formed by combining <code>_imp__</code> and the function or variable
559 name.
560 </dd>
561
Chris Lattner6af02f32004-12-09 16:11:40 +0000562</dl>
563
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000564<p><a name="linkage_external"></a>For example, since the "<tt>.LC0</tt>"
Chris Lattner6af02f32004-12-09 16:11:40 +0000565variable is defined to be internal, if another module defined a "<tt>.LC0</tt>"
566variable and was linked with this one, one of the two would be renamed,
567preventing a collision. Since "<tt>main</tt>" and "<tt>puts</tt>" are
568external (i.e., lacking any linkage declarations), they are accessible
Reid Spencer92c671e2007-01-05 00:59:10 +0000569outside of the current module.</p>
570<p>It is illegal for a function <i>declaration</i>
571to have any linkage type other than "externally visible", <tt>dllimport</tt>,
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000572or <tt>extern_weak</tt>.</p>
Anton Korobeynikova97b6942007-04-25 14:27:10 +0000573<p>Aliases can have only <tt>external</tt>, <tt>internal</tt> and <tt>weak</tt>
574linkages.
Chris Lattner6af02f32004-12-09 16:11:40 +0000575</div>
576
577<!-- ======================================================================= -->
578<div class="doc_subsection">
Chris Lattner0132aff2005-05-06 22:57:40 +0000579 <a name="callingconv">Calling Conventions</a>
580</div>
581
582<div class="doc_text">
583
584<p>LLVM <a href="#functionstructure">functions</a>, <a href="#i_call">calls</a>
585and <a href="#i_invoke">invokes</a> can all have an optional calling convention
586specified for the call. The calling convention of any pair of dynamic
587caller/callee must match, or the behavior of the program is undefined. The
588following calling conventions are supported by LLVM, and more may be added in
589the future:</p>
590
591<dl>
592 <dt><b>"<tt>ccc</tt>" - The C calling convention</b>:</dt>
593
594 <dd>This calling convention (the default if no other calling convention is
595 specified) matches the target C calling conventions. This calling convention
John Criswell02fdc6f2005-05-12 16:52:32 +0000596 supports varargs function calls and tolerates some mismatch in the declared
Reid Spencer72ba4992006-12-31 21:30:18 +0000597 prototype and implemented declaration of the function (as does normal C).
Chris Lattner0132aff2005-05-06 22:57:40 +0000598 </dd>
599
600 <dt><b>"<tt>fastcc</tt>" - The fast calling convention</b>:</dt>
601
602 <dd>This calling convention attempts to make calls as fast as possible
603 (e.g. by passing things in registers). This calling convention allows the
604 target to use whatever tricks it wants to produce fast code for the target,
Chris Lattnerc792eb32005-05-06 23:08:23 +0000605 without having to conform to an externally specified ABI. Implementations of
Arnold Schwaighofer2c6b8882008-05-14 09:17:12 +0000606 this convention should allow arbitrary
607 <a href="CodeGenerator.html#tailcallopt">tail call optimization</a> to be
608 supported. This calling convention does not support varargs and requires the
609 prototype of all callees to exactly match the prototype of the function
610 definition.
Chris Lattner0132aff2005-05-06 22:57:40 +0000611 </dd>
612
613 <dt><b>"<tt>coldcc</tt>" - The cold calling convention</b>:</dt>
614
615 <dd>This calling convention attempts to make code in the caller as efficient
616 as possible under the assumption that the call is not commonly executed. As
617 such, these calls often preserve all registers so that the call does not break
618 any live ranges in the caller side. This calling convention does not support
619 varargs and requires the prototype of all callees to exactly match the
620 prototype of the function definition.
621 </dd>
622
Chris Lattner573f64e2005-05-07 01:46:40 +0000623 <dt><b>"<tt>cc &lt;<em>n</em>&gt;</tt>" - Numbered convention</b>:</dt>
Chris Lattner0132aff2005-05-06 22:57:40 +0000624
625 <dd>Any calling convention may be specified by number, allowing
626 target-specific calling conventions to be used. Target specific calling
627 conventions start at 64.
628 </dd>
Chris Lattner573f64e2005-05-07 01:46:40 +0000629</dl>
Chris Lattner0132aff2005-05-06 22:57:40 +0000630
631<p>More calling conventions can be added/defined on an as-needed basis, to
632support pascal conventions or any other well-known target-independent
633convention.</p>
634
635</div>
636
637<!-- ======================================================================= -->
638<div class="doc_subsection">
Anton Korobeynikovc7f9f3d2007-01-23 12:35:46 +0000639 <a name="visibility">Visibility Styles</a>
640</div>
641
642<div class="doc_text">
643
644<p>
645All Global Variables and Functions have one of the following visibility styles:
646</p>
647
648<dl>
649 <dt><b>"<tt>default</tt>" - Default style</b>:</dt>
650
651 <dd>On ELF, default visibility means that the declaration is visible to other
652 modules and, in shared libraries, means that the declared entity may be
653 overridden. On Darwin, default visibility means that the declaration is
654 visible to other modules. Default visibility corresponds to "external
655 linkage" in the language.
656 </dd>
657
658 <dt><b>"<tt>hidden</tt>" - Hidden style</b>:</dt>
659
660 <dd>Two declarations of an object with hidden visibility refer to the same
661 object if they are in the same shared object. Usually, hidden visibility
662 indicates that the symbol will not be placed into the dynamic symbol table,
663 so no other module (executable or shared library) can reference it
664 directly.
665 </dd>
666
Anton Korobeynikov39f3cff2007-04-29 18:35:00 +0000667 <dt><b>"<tt>protected</tt>" - Protected style</b>:</dt>
668
669 <dd>On ELF, protected visibility indicates that the symbol will be placed in
670 the dynamic symbol table, but that references within the defining module will
671 bind to the local symbol. That is, the symbol cannot be overridden by another
672 module.
673 </dd>
Anton Korobeynikovc7f9f3d2007-01-23 12:35:46 +0000674</dl>
675
676</div>
677
678<!-- ======================================================================= -->
679<div class="doc_subsection">
Chris Lattner6af02f32004-12-09 16:11:40 +0000680 <a name="globalvars">Global Variables</a>
681</div>
682
683<div class="doc_text">
684
Chris Lattner5d5aede2005-02-12 19:30:21 +0000685<p>Global variables define regions of memory allocated at compilation time
Chris Lattner662c8722005-11-12 00:45:07 +0000686instead of run-time. Global variables may optionally be initialized, may have
Anton Korobeynikova97b6942007-04-25 14:27:10 +0000687an explicit section to be placed in, and may have an optional explicit alignment
688specified. A variable may be defined as "thread_local", which means that it
689will not be shared by threads (each thread will have a separated copy of the
690variable). A variable may be defined as a global "constant," which indicates
691that the contents of the variable will <b>never</b> be modified (enabling better
Chris Lattner5d5aede2005-02-12 19:30:21 +0000692optimization, allowing the global data to be placed in the read-only section of
693an executable, etc). Note that variables that need runtime initialization
John Criswell4c0cf7f2005-10-24 16:17:18 +0000694cannot be marked "constant" as there is a store to the variable.</p>
Chris Lattner5d5aede2005-02-12 19:30:21 +0000695
696<p>
697LLVM explicitly allows <em>declarations</em> of global variables to be marked
698constant, even if the final definition of the global is not. This capability
699can be used to enable slightly better optimization of the program, but requires
700the language definition to guarantee that optimizations based on the
701'constantness' are valid for the translation units that do not include the
702definition.
703</p>
Chris Lattner6af02f32004-12-09 16:11:40 +0000704
705<p>As SSA values, global variables define pointer values that are in
706scope (i.e. they dominate) all basic blocks in the program. Global
707variables always define a pointer to their "content" type because they
708describe a region of memory, and all memory objects in LLVM are
709accessed through pointers.</p>
710
Christopher Lamb308121c2007-12-11 09:31:00 +0000711<p>A global variable may be declared to reside in a target-specifc numbered
712address space. For targets that support them, address spaces may affect how
713optimizations are performed and/or what target instructions are used to access
Christopher Lamb25f50762007-12-12 08:44:39 +0000714the variable. The default address space is zero. The address space qualifier
715must precede any other attributes.</p>
Christopher Lamb308121c2007-12-11 09:31:00 +0000716
Chris Lattner662c8722005-11-12 00:45:07 +0000717<p>LLVM allows an explicit section to be specified for globals. If the target
718supports it, it will emit globals to the section specified.</p>
719
Chris Lattner54611b42005-11-06 08:02:57 +0000720<p>An explicit alignment may be specified for a global. If not present, or if
721the alignment is set to zero, the alignment of the global is set by the target
722to whatever it feels convenient. If an explicit alignment is specified, the
723global is forced to have at least that much alignment. All alignments must be
724a power of 2.</p>
725
Christopher Lamb308121c2007-12-11 09:31:00 +0000726<p>For example, the following defines a global in a numbered address space with
727an initializer, section, and alignment:</p>
Chris Lattner5760c502007-01-14 00:27:09 +0000728
Bill Wendling3716c5d2007-05-29 09:04:49 +0000729<div class="doc_code">
Chris Lattner5760c502007-01-14 00:27:09 +0000730<pre>
Christopher Lamb308121c2007-12-11 09:31:00 +0000731@G = constant float 1.0 addrspace(5), section "foo", align 4
Chris Lattner5760c502007-01-14 00:27:09 +0000732</pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000733</div>
Chris Lattner5760c502007-01-14 00:27:09 +0000734
Chris Lattner6af02f32004-12-09 16:11:40 +0000735</div>
736
737
738<!-- ======================================================================= -->
739<div class="doc_subsection">
740 <a name="functionstructure">Functions</a>
741</div>
742
743<div class="doc_text">
744
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000745<p>LLVM function definitions consist of the "<tt>define</tt>" keyord,
746an optional <a href="#linkage">linkage type</a>, an optional
Anton Korobeynikovc7f9f3d2007-01-23 12:35:46 +0000747<a href="#visibility">visibility style</a>, an optional
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000748<a href="#callingconv">calling convention</a>, a return type, an optional
749<a href="#paramattrs">parameter attribute</a> for the return type, a function
750name, a (possibly empty) argument list (each with optional
Anton Korobeynikovc7f9f3d2007-01-23 12:35:46 +0000751<a href="#paramattrs">parameter attributes</a>), an optional section, an
Gordon Henriksendc5cafb2007-12-10 03:30:21 +0000752optional alignment, an optional <a href="#gc">garbage collector name</a>, an
Gordon Henriksen71183b62007-12-10 03:18:06 +0000753opening curly brace, a list of basic blocks, and a closing curly brace.
Anton Korobeynikovc7f9f3d2007-01-23 12:35:46 +0000754
755LLVM function declarations consist of the "<tt>declare</tt>" keyword, an
756optional <a href="#linkage">linkage type</a>, an optional
757<a href="#visibility">visibility style</a>, an optional
758<a href="#callingconv">calling convention</a>, a return type, an optional
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000759<a href="#paramattrs">parameter attribute</a> for the return type, a function
Gordon Henriksen71183b62007-12-10 03:18:06 +0000760name, a possibly empty list of arguments, an optional alignment, and an optional
Gordon Henriksendc5cafb2007-12-10 03:30:21 +0000761<a href="#gc">garbage collector name</a>.</p>
Chris Lattner6af02f32004-12-09 16:11:40 +0000762
763<p>A function definition contains a list of basic blocks, forming the CFG for
764the function. Each basic block may optionally start with a label (giving the
765basic block a symbol table entry), contains a list of instructions, and ends
766with a <a href="#terminators">terminator</a> instruction (such as a branch or
767function return).</p>
768
Chris Lattnera59fb102007-06-08 16:52:14 +0000769<p>The first basic block in a function is special in two ways: it is immediately
Chris Lattner6af02f32004-12-09 16:11:40 +0000770executed on entrance to the function, and it is not allowed to have predecessor
771basic blocks (i.e. there can not be any branches to the entry block of a
772function). Because the block can have no predecessors, it also cannot have any
773<a href="#i_phi">PHI nodes</a>.</p>
774
Chris Lattner662c8722005-11-12 00:45:07 +0000775<p>LLVM allows an explicit section to be specified for functions. If the target
776supports it, it will emit functions to the section specified.</p>
777
Chris Lattner54611b42005-11-06 08:02:57 +0000778<p>An explicit alignment may be specified for a function. If not present, or if
779the alignment is set to zero, the alignment of the function is set by the target
780to whatever it feels convenient. If an explicit alignment is specified, the
781function is forced to have at least that much alignment. All alignments must be
782a power of 2.</p>
783
Chris Lattner6af02f32004-12-09 16:11:40 +0000784</div>
785
Anton Korobeynikova97b6942007-04-25 14:27:10 +0000786
787<!-- ======================================================================= -->
788<div class="doc_subsection">
789 <a name="aliasstructure">Aliases</a>
790</div>
791<div class="doc_text">
792 <p>Aliases act as "second name" for the aliasee value (which can be either
Anton Korobeynikov25b2e822008-03-22 08:36:14 +0000793 function, global variable, another alias or bitcast of global value). Aliases
794 may have an optional <a href="#linkage">linkage type</a>, and an
Anton Korobeynikova97b6942007-04-25 14:27:10 +0000795 optional <a href="#visibility">visibility style</a>.</p>
796
797 <h5>Syntax:</h5>
798
Bill Wendling3716c5d2007-05-29 09:04:49 +0000799<div class="doc_code">
Bill Wendling2d8b9a82007-05-29 09:42:13 +0000800<pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000801@&lt;Name&gt; = [Linkage] [Visibility] alias &lt;AliaseeTy&gt; @&lt;Aliasee&gt;
Bill Wendling2d8b9a82007-05-29 09:42:13 +0000802</pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000803</div>
Anton Korobeynikova97b6942007-04-25 14:27:10 +0000804
805</div>
806
807
808
Chris Lattner91c15c42006-01-23 23:23:47 +0000809<!-- ======================================================================= -->
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000810<div class="doc_subsection"><a name="paramattrs">Parameter Attributes</a></div>
811<div class="doc_text">
812 <p>The return type and each parameter of a function type may have a set of
813 <i>parameter attributes</i> associated with them. Parameter attributes are
814 used to communicate additional information about the result or parameters of
Duncan Sandsad0ea2d2007-11-27 13:23:08 +0000815 a function. Parameter attributes are considered to be part of the function,
816 not of the function type, so functions with different parameter attributes
817 can have the same function type.</p>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000818
Reid Spencercf7ebf52007-01-15 18:27:39 +0000819 <p>Parameter attributes are simple keywords that follow the type specified. If
820 multiple parameter attributes are needed, they are space separated. For
Bill Wendling3716c5d2007-05-29 09:04:49 +0000821 example:</p>
822
823<div class="doc_code">
824<pre>
Duncan Sandsad0ea2d2007-11-27 13:23:08 +0000825declare i32 @printf(i8* noalias , ...) nounwind
826declare i32 @atoi(i8*) nounwind readonly
Bill Wendling3716c5d2007-05-29 09:04:49 +0000827</pre>
828</div>
829
Duncan Sandsad0ea2d2007-11-27 13:23:08 +0000830 <p>Note that any attributes for the function result (<tt>nounwind</tt>,
831 <tt>readonly</tt>) come immediately after the argument list.</p>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000832
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000833 <p>Currently, only the following parameter attributes are defined:</p>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000834 <dl>
Reid Spencer314e1cb2007-07-19 23:13:04 +0000835 <dt><tt>zeroext</tt></dt>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000836 <dd>This indicates that the parameter should be zero extended just before
837 a call to this function.</dd>
Chris Lattner5cee13f2008-01-11 06:20:47 +0000838
Reid Spencer314e1cb2007-07-19 23:13:04 +0000839 <dt><tt>signext</tt></dt>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000840 <dd>This indicates that the parameter should be sign extended just before
841 a call to this function.</dd>
Chris Lattner5cee13f2008-01-11 06:20:47 +0000842
Anton Korobeynikove8166852007-01-28 14:30:45 +0000843 <dt><tt>inreg</tt></dt>
844 <dd>This indicates that the parameter should be placed in register (if
Anton Korobeynikove93c6e82007-01-28 15:27:21 +0000845 possible) during assembling function call. Support for this attribute is
846 target-specific</dd>
Chris Lattner5cee13f2008-01-11 06:20:47 +0000847
848 <dt><tt>byval</tt></dt>
Chris Lattner352ab9b2008-01-15 04:34:22 +0000849 <dd>This indicates that the pointer parameter should really be passed by
850 value to the function. The attribute implies that a hidden copy of the
851 pointee is made between the caller and the callee, so the callee is unable
852 to modify the value in the callee. This attribute is only valid on llvm
853 pointer arguments. It is generally used to pass structs and arrays by
854 value, but is also valid on scalars (even though this is silly).</dd>
Chris Lattner5cee13f2008-01-11 06:20:47 +0000855
Anton Korobeynikove8166852007-01-28 14:30:45 +0000856 <dt><tt>sret</tt></dt>
Duncan Sandsfa4b6732008-02-18 04:19:38 +0000857 <dd>This indicates that the pointer parameter specifies the address of a
858 structure that is the return value of the function in the source program.
Duncan Sandsc572c1e2008-03-17 12:17:41 +0000859 Loads and stores to the structure are assumed not to trap.
Duncan Sandsfa4b6732008-02-18 04:19:38 +0000860 May only be applied to the first parameter.</dd>
Chris Lattner5cee13f2008-01-11 06:20:47 +0000861
Zhou Sheng2444a9a2007-06-05 05:28:26 +0000862 <dt><tt>noalias</tt></dt>
Owen Anderson61101282008-02-18 04:09:01 +0000863 <dd>This indicates that the parameter does not alias any global or any other
864 parameter. The caller is responsible for ensuring that this is the case,
865 usually by placing the value in a stack allocation.</dd>
Chris Lattner5cee13f2008-01-11 06:20:47 +0000866
Reid Spencer9d1700e2007-03-22 02:18:56 +0000867 <dt><tt>noreturn</tt></dt>
868 <dd>This function attribute indicates that the function never returns. This
869 indicates to LLVM that every call to this function should be treated as if
870 an <tt>unreachable</tt> instruction immediately followed the call.</dd>
Chris Lattner5cee13f2008-01-11 06:20:47 +0000871
Reid Spencer05dbb9d2007-03-22 02:02:11 +0000872 <dt><tt>nounwind</tt></dt>
Duncan Sandsc572c1e2008-03-17 12:17:41 +0000873 <dd>This function attribute indicates that no exceptions unwind out of the
874 function. Usually this is because the function makes no use of exceptions,
875 but it may also be that the function catches any exceptions thrown when
876 executing it.</dd>
877
Duncan Sands27e91592007-07-27 19:57:41 +0000878 <dt><tt>nest</tt></dt>
Duncan Sands825bde42008-07-08 09:27:25 +0000879 <dd>This indicates that the pointer parameter can be excised using the
Duncan Sands27e91592007-07-27 19:57:41 +0000880 <a href="#int_trampoline">trampoline intrinsics</a>.</dd>
Duncan Sandsa89a1132007-11-22 20:23:04 +0000881 <dt><tt>readonly</tt></dt>
Duncan Sands730a3262007-11-14 21:14:02 +0000882 <dd>This function attribute indicates that the function has no side-effects
Duncan Sandsa89a1132007-11-22 20:23:04 +0000883 except for producing a return value or throwing an exception. The value
884 returned must only depend on the function arguments and/or global variables.
885 It may use values obtained by dereferencing pointers.</dd>
886 <dt><tt>readnone</tt></dt>
887 <dd>A <tt>readnone</tt> function has the same restrictions as a <tt>readonly</tt>
Duncan Sands730a3262007-11-14 21:14:02 +0000888 function, but in addition it is not allowed to dereference any pointer arguments
889 or global variables.
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000890 </dl>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000891
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000892</div>
893
894<!-- ======================================================================= -->
Chris Lattner91c15c42006-01-23 23:23:47 +0000895<div class="doc_subsection">
Gordon Henriksen71183b62007-12-10 03:18:06 +0000896 <a name="gc">Garbage Collector Names</a>
897</div>
898
899<div class="doc_text">
900<p>Each function may specify a garbage collector name, which is simply a
901string.</p>
902
903<div class="doc_code"><pre
904>define void @f() gc "name" { ...</pre></div>
905
906<p>The compiler declares the supported values of <i>name</i>. Specifying a
907collector which will cause the compiler to alter its output in order to support
908the named garbage collection algorithm.</p>
909</div>
910
911<!-- ======================================================================= -->
912<div class="doc_subsection">
Chris Lattner93564892006-04-08 04:40:53 +0000913 <a name="moduleasm">Module-Level Inline Assembly</a>
Chris Lattner91c15c42006-01-23 23:23:47 +0000914</div>
915
916<div class="doc_text">
917<p>
918Modules may contain "module-level inline asm" blocks, which corresponds to the
919GCC "file scope inline asm" blocks. These blocks are internally concatenated by
920LLVM and treated as a single unit, but may be separated in the .ll file if
921desired. The syntax is very simple:
922</p>
923
Bill Wendling3716c5d2007-05-29 09:04:49 +0000924<div class="doc_code">
925<pre>
926module asm "inline asm code goes here"
927module asm "more can go here"
928</pre>
929</div>
Chris Lattner91c15c42006-01-23 23:23:47 +0000930
931<p>The strings can contain any character by escaping non-printable characters.
932 The escape sequence used is simply "\xx" where "xx" is the two digit hex code
933 for the number.
934</p>
935
936<p>
937 The inline asm code is simply printed to the machine code .s file when
938 assembly code is generated.
939</p>
940</div>
Chris Lattner6af02f32004-12-09 16:11:40 +0000941
Reid Spencer50c723a2007-02-19 23:54:10 +0000942<!-- ======================================================================= -->
943<div class="doc_subsection">
944 <a name="datalayout">Data Layout</a>
945</div>
946
947<div class="doc_text">
948<p>A module may specify a target specific data layout string that specifies how
Reid Spencer7972c472007-04-11 23:49:50 +0000949data is to be laid out in memory. The syntax for the data layout is simply:</p>
950<pre> target datalayout = "<i>layout specification</i>"</pre>
951<p>The <i>layout specification</i> consists of a list of specifications
952separated by the minus sign character ('-'). Each specification starts with a
953letter and may include other information after the letter to define some
954aspect of the data layout. The specifications accepted are as follows: </p>
Reid Spencer50c723a2007-02-19 23:54:10 +0000955<dl>
956 <dt><tt>E</tt></dt>
957 <dd>Specifies that the target lays out data in big-endian form. That is, the
958 bits with the most significance have the lowest address location.</dd>
959 <dt><tt>e</tt></dt>
960 <dd>Specifies that hte target lays out data in little-endian form. That is,
961 the bits with the least significance have the lowest address location.</dd>
962 <dt><tt>p:<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
963 <dd>This specifies the <i>size</i> of a pointer and its <i>abi</i> and
964 <i>preferred</i> alignments. All sizes are in bits. Specifying the <i>pref</i>
965 alignment is optional. If omitted, the preceding <tt>:</tt> should be omitted
966 too.</dd>
967 <dt><tt>i<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
968 <dd>This specifies the alignment for an integer type of a given bit
969 <i>size</i>. The value of <i>size</i> must be in the range [1,2^23).</dd>
970 <dt><tt>v<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
971 <dd>This specifies the alignment for a vector type of a given bit
972 <i>size</i>.</dd>
973 <dt><tt>f<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
974 <dd>This specifies the alignment for a floating point type of a given bit
975 <i>size</i>. The value of <i>size</i> must be either 32 (float) or 64
976 (double).</dd>
977 <dt><tt>a<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
978 <dd>This specifies the alignment for an aggregate type of a given bit
979 <i>size</i>.</dd>
980</dl>
981<p>When constructing the data layout for a given target, LLVM starts with a
982default set of specifications which are then (possibly) overriden by the
983specifications in the <tt>datalayout</tt> keyword. The default specifications
984are given in this list:</p>
985<ul>
986 <li><tt>E</tt> - big endian</li>
987 <li><tt>p:32:64:64</tt> - 32-bit pointers with 64-bit alignment</li>
988 <li><tt>i1:8:8</tt> - i1 is 8-bit (byte) aligned</li>
989 <li><tt>i8:8:8</tt> - i8 is 8-bit (byte) aligned</li>
990 <li><tt>i16:16:16</tt> - i16 is 16-bit aligned</li>
991 <li><tt>i32:32:32</tt> - i32 is 32-bit aligned</li>
992 <li><tt>i64:32:64</tt> - i64 has abi alignment of 32-bits but preferred
993 alignment of 64-bits</li>
994 <li><tt>f32:32:32</tt> - float is 32-bit aligned</li>
995 <li><tt>f64:64:64</tt> - double is 64-bit aligned</li>
996 <li><tt>v64:64:64</tt> - 64-bit vector is 64-bit aligned</li>
997 <li><tt>v128:128:128</tt> - 128-bit vector is 128-bit aligned</li>
998 <li><tt>a0:0:1</tt> - aggregates are 8-bit aligned</li>
999</ul>
1000<p>When llvm is determining the alignment for a given type, it uses the
1001following rules:
1002<ol>
1003 <li>If the type sought is an exact match for one of the specifications, that
1004 specification is used.</li>
1005 <li>If no match is found, and the type sought is an integer type, then the
1006 smallest integer type that is larger than the bitwidth of the sought type is
1007 used. If none of the specifications are larger than the bitwidth then the the
1008 largest integer type is used. For example, given the default specifications
1009 above, the i7 type will use the alignment of i8 (next largest) while both
1010 i65 and i256 will use the alignment of i64 (largest specified).</li>
1011 <li>If no match is found, and the type sought is a vector type, then the
1012 largest vector type that is smaller than the sought vector type will be used
1013 as a fall back. This happens because <128 x double> can be implemented in
1014 terms of 64 <2 x double>, for example.</li>
1015</ol>
1016</div>
Chris Lattner6af02f32004-12-09 16:11:40 +00001017
Chris Lattner2f7c9632001-06-06 20:29:01 +00001018<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +00001019<div class="doc_section"> <a name="typesystem">Type System</a> </div>
1020<!-- *********************************************************************** -->
Chris Lattner6af02f32004-12-09 16:11:40 +00001021
Misha Brukman76307852003-11-08 01:05:38 +00001022<div class="doc_text">
Chris Lattner6af02f32004-12-09 16:11:40 +00001023
Misha Brukman76307852003-11-08 01:05:38 +00001024<p>The LLVM type system is one of the most important features of the
Chris Lattner48b383b02003-11-25 01:02:51 +00001025intermediate representation. Being typed enables a number of
1026optimizations to be performed on the IR directly, without having to do
1027extra analyses on the side before the transformation. A strong type
1028system makes it easier to read the generated code and enables novel
1029analyses and transformations that are not feasible to perform on normal
1030three address code representations.</p>
Chris Lattner6af02f32004-12-09 16:11:40 +00001031
1032</div>
1033
Chris Lattner2f7c9632001-06-06 20:29:01 +00001034<!-- ======================================================================= -->
Chris Lattner7824d182008-01-04 04:32:38 +00001035<div class="doc_subsection"> <a name="t_classifications">Type
Chris Lattner48b383b02003-11-25 01:02:51 +00001036Classifications</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +00001037<div class="doc_text">
Chris Lattner7824d182008-01-04 04:32:38 +00001038<p>The types fall into a few useful
Chris Lattner48b383b02003-11-25 01:02:51 +00001039classifications:</p>
Misha Brukmanc501f552004-03-01 17:47:27 +00001040
1041<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner48b383b02003-11-25 01:02:51 +00001042 <tbody>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +00001043 <tr><th>Classification</th><th>Types</th></tr>
Chris Lattner48b383b02003-11-25 01:02:51 +00001044 <tr>
Chris Lattner7824d182008-01-04 04:32:38 +00001045 <td><a href="#t_integer">integer</a></td>
Reid Spencer138249b2007-05-16 18:44:01 +00001046 <td><tt>i1, i2, i3, ... i8, ... i16, ... i32, ... i64, ... </tt></td>
Chris Lattner48b383b02003-11-25 01:02:51 +00001047 </tr>
1048 <tr>
Chris Lattner7824d182008-01-04 04:32:38 +00001049 <td><a href="#t_floating">floating point</a></td>
1050 <td><tt>float, double, x86_fp80, fp128, ppc_fp128</tt></td>
Chris Lattner48b383b02003-11-25 01:02:51 +00001051 </tr>
1052 <tr>
1053 <td><a name="t_firstclass">first class</a></td>
Chris Lattner7824d182008-01-04 04:32:38 +00001054 <td><a href="#t_integer">integer</a>,
1055 <a href="#t_floating">floating point</a>,
1056 <a href="#t_pointer">pointer</a>,
Dan Gohman08783a882008-06-18 18:42:13 +00001057 <a href="#t_vector">vector</a>,
Dan Gohmanb9d66602008-05-12 23:51:09 +00001058 <a href="#t_struct">structure</a>,
1059 <a href="#t_array">array</a>,
Dan Gohmanda52d212008-05-23 22:50:26 +00001060 <a href="#t_label">label</a>.
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00001061 </td>
Chris Lattner48b383b02003-11-25 01:02:51 +00001062 </tr>
Chris Lattner7824d182008-01-04 04:32:38 +00001063 <tr>
1064 <td><a href="#t_primitive">primitive</a></td>
1065 <td><a href="#t_label">label</a>,
1066 <a href="#t_void">void</a>,
Chris Lattner7824d182008-01-04 04:32:38 +00001067 <a href="#t_floating">floating point</a>.</td>
1068 </tr>
1069 <tr>
1070 <td><a href="#t_derived">derived</a></td>
1071 <td><a href="#t_integer">integer</a>,
1072 <a href="#t_array">array</a>,
1073 <a href="#t_function">function</a>,
1074 <a href="#t_pointer">pointer</a>,
1075 <a href="#t_struct">structure</a>,
1076 <a href="#t_pstruct">packed structure</a>,
1077 <a href="#t_vector">vector</a>,
1078 <a href="#t_opaque">opaque</a>.
1079 </tr>
Chris Lattner48b383b02003-11-25 01:02:51 +00001080 </tbody>
Misha Brukman76307852003-11-08 01:05:38 +00001081</table>
Misha Brukmanc501f552004-03-01 17:47:27 +00001082
Chris Lattner48b383b02003-11-25 01:02:51 +00001083<p>The <a href="#t_firstclass">first class</a> types are perhaps the
1084most important. Values of these types are the only ones which can be
1085produced by instructions, passed as arguments, or used as operands to
Dan Gohman34d1c0d2008-05-23 21:53:15 +00001086instructions.</p>
Misha Brukman76307852003-11-08 01:05:38 +00001087</div>
Chris Lattner74d3f822004-12-09 17:30:23 +00001088
Chris Lattner2f7c9632001-06-06 20:29:01 +00001089<!-- ======================================================================= -->
Chris Lattner7824d182008-01-04 04:32:38 +00001090<div class="doc_subsection"> <a name="t_primitive">Primitive Types</a> </div>
Chris Lattner43542b32008-01-04 04:34:14 +00001091
Chris Lattner7824d182008-01-04 04:32:38 +00001092<div class="doc_text">
1093<p>The primitive types are the fundamental building blocks of the LLVM
1094system.</p>
1095
Chris Lattner43542b32008-01-04 04:34:14 +00001096</div>
1097
Chris Lattner7824d182008-01-04 04:32:38 +00001098<!-- _______________________________________________________________________ -->
1099<div class="doc_subsubsection"> <a name="t_floating">Floating Point Types</a> </div>
1100
1101<div class="doc_text">
1102 <table>
1103 <tbody>
1104 <tr><th>Type</th><th>Description</th></tr>
1105 <tr><td><tt>float</tt></td><td>32-bit floating point value</td></tr>
1106 <tr><td><tt>double</tt></td><td>64-bit floating point value</td></tr>
1107 <tr><td><tt>fp128</tt></td><td>128-bit floating point value (112-bit mantissa)</td></tr>
1108 <tr><td><tt>x86_fp80</tt></td><td>80-bit floating point value (X87)</td></tr>
1109 <tr><td><tt>ppc_fp128</tt></td><td>128-bit floating point value (two 64-bits)</td></tr>
1110 </tbody>
1111 </table>
1112</div>
1113
1114<!-- _______________________________________________________________________ -->
1115<div class="doc_subsubsection"> <a name="t_void">Void Type</a> </div>
1116
1117<div class="doc_text">
1118<h5>Overview:</h5>
1119<p>The void type does not represent any value and has no size.</p>
1120
1121<h5>Syntax:</h5>
1122
1123<pre>
1124 void
1125</pre>
1126</div>
1127
1128<!-- _______________________________________________________________________ -->
1129<div class="doc_subsubsection"> <a name="t_label">Label Type</a> </div>
1130
1131<div class="doc_text">
1132<h5>Overview:</h5>
1133<p>The label type represents code labels.</p>
1134
1135<h5>Syntax:</h5>
1136
1137<pre>
1138 label
1139</pre>
1140</div>
1141
1142
1143<!-- ======================================================================= -->
Chris Lattner48b383b02003-11-25 01:02:51 +00001144<div class="doc_subsection"> <a name="t_derived">Derived Types</a> </div>
Chris Lattner74d3f822004-12-09 17:30:23 +00001145
Misha Brukman76307852003-11-08 01:05:38 +00001146<div class="doc_text">
Chris Lattner74d3f822004-12-09 17:30:23 +00001147
Chris Lattner48b383b02003-11-25 01:02:51 +00001148<p>The real power in LLVM comes from the derived types in the system.
1149This is what allows a programmer to represent arrays, functions,
1150pointers, and other useful types. Note that these derived types may be
1151recursive: For example, it is possible to have a two dimensional array.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +00001152
Misha Brukman76307852003-11-08 01:05:38 +00001153</div>
Chris Lattner74d3f822004-12-09 17:30:23 +00001154
Chris Lattner2f7c9632001-06-06 20:29:01 +00001155<!-- _______________________________________________________________________ -->
Reid Spencer138249b2007-05-16 18:44:01 +00001156<div class="doc_subsubsection"> <a name="t_integer">Integer Type</a> </div>
1157
1158<div class="doc_text">
1159
1160<h5>Overview:</h5>
1161<p>The integer type is a very simple derived type that simply specifies an
1162arbitrary bit width for the integer type desired. Any bit width from 1 bit to
11632^23-1 (about 8 million) can be specified.</p>
1164
1165<h5>Syntax:</h5>
1166
1167<pre>
1168 iN
1169</pre>
1170
1171<p>The number of bits the integer will occupy is specified by the <tt>N</tt>
1172value.</p>
1173
1174<h5>Examples:</h5>
1175<table class="layout">
Chris Lattner9a2e3cb2007-12-18 06:18:21 +00001176 <tbody>
1177 <tr>
1178 <td><tt>i1</tt></td>
1179 <td>a single-bit integer.</td>
1180 </tr><tr>
1181 <td><tt>i32</tt></td>
1182 <td>a 32-bit integer.</td>
1183 </tr><tr>
1184 <td><tt>i1942652</tt></td>
1185 <td>a really big integer of over 1 million bits.</td>
Reid Spencer138249b2007-05-16 18:44:01 +00001186 </tr>
Chris Lattner9a2e3cb2007-12-18 06:18:21 +00001187 </tbody>
Reid Spencer138249b2007-05-16 18:44:01 +00001188</table>
Bill Wendling3716c5d2007-05-29 09:04:49 +00001189</div>
Reid Spencer138249b2007-05-16 18:44:01 +00001190
1191<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +00001192<div class="doc_subsubsection"> <a name="t_array">Array Type</a> </div>
Chris Lattner74d3f822004-12-09 17:30:23 +00001193
Misha Brukman76307852003-11-08 01:05:38 +00001194<div class="doc_text">
Chris Lattner74d3f822004-12-09 17:30:23 +00001195
Chris Lattner2f7c9632001-06-06 20:29:01 +00001196<h5>Overview:</h5>
Chris Lattner74d3f822004-12-09 17:30:23 +00001197
Misha Brukman76307852003-11-08 01:05:38 +00001198<p>The array type is a very simple derived type that arranges elements
Chris Lattner48b383b02003-11-25 01:02:51 +00001199sequentially in memory. The array type requires a size (number of
1200elements) and an underlying data type.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +00001201
Chris Lattner590645f2002-04-14 06:13:44 +00001202<h5>Syntax:</h5>
Chris Lattner74d3f822004-12-09 17:30:23 +00001203
1204<pre>
1205 [&lt;# elements&gt; x &lt;elementtype&gt;]
1206</pre>
1207
John Criswell02fdc6f2005-05-12 16:52:32 +00001208<p>The number of elements is a constant integer value; elementtype may
Chris Lattner48b383b02003-11-25 01:02:51 +00001209be any type with a size.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +00001210
Chris Lattner590645f2002-04-14 06:13:44 +00001211<h5>Examples:</h5>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +00001212<table class="layout">
1213 <tr class="layout">
Chris Lattner747359f2007-12-19 05:04:11 +00001214 <td class="left"><tt>[40 x i32]</tt></td>
1215 <td class="left">Array of 40 32-bit integer values.</td>
1216 </tr>
1217 <tr class="layout">
1218 <td class="left"><tt>[41 x i32]</tt></td>
1219 <td class="left">Array of 41 32-bit integer values.</td>
1220 </tr>
1221 <tr class="layout">
1222 <td class="left"><tt>[4 x i8]</tt></td>
1223 <td class="left">Array of 4 8-bit integer values.</td>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +00001224 </tr>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001225</table>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +00001226<p>Here are some examples of multidimensional arrays:</p>
1227<table class="layout">
1228 <tr class="layout">
Chris Lattner747359f2007-12-19 05:04:11 +00001229 <td class="left"><tt>[3 x [4 x i32]]</tt></td>
1230 <td class="left">3x4 array of 32-bit integer values.</td>
1231 </tr>
1232 <tr class="layout">
1233 <td class="left"><tt>[12 x [10 x float]]</tt></td>
1234 <td class="left">12x10 array of single precision floating point values.</td>
1235 </tr>
1236 <tr class="layout">
1237 <td class="left"><tt>[2 x [3 x [4 x i16]]]</tt></td>
1238 <td class="left">2x3x4 array of 16-bit integer values.</td>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +00001239 </tr>
1240</table>
Chris Lattnerc0ad71e2005-06-24 17:22:57 +00001241
John Criswell4c0cf7f2005-10-24 16:17:18 +00001242<p>Note that 'variable sized arrays' can be implemented in LLVM with a zero
1243length array. Normally, accesses past the end of an array are undefined in
Chris Lattnerc0ad71e2005-06-24 17:22:57 +00001244LLVM (e.g. it is illegal to access the 5th element of a 3 element array).
1245As a special case, however, zero length arrays are recognized to be variable
1246length. This allows implementation of 'pascal style arrays' with the LLVM
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00001247type "{ i32, [0 x float]}", for example.</p>
Chris Lattnerc0ad71e2005-06-24 17:22:57 +00001248
Misha Brukman76307852003-11-08 01:05:38 +00001249</div>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +00001250
Chris Lattner2f7c9632001-06-06 20:29:01 +00001251<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +00001252<div class="doc_subsubsection"> <a name="t_function">Function Type</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +00001253<div class="doc_text">
Chris Lattnerda508ac2008-04-23 04:59:35 +00001254
Chris Lattner2f7c9632001-06-06 20:29:01 +00001255<h5>Overview:</h5>
Chris Lattnerda508ac2008-04-23 04:59:35 +00001256
Chris Lattner48b383b02003-11-25 01:02:51 +00001257<p>The function type can be thought of as a function signature. It
Devang Patele3dfc1c2008-03-24 05:35:41 +00001258consists of a return type and a list of formal parameter types. The
Chris Lattnerda508ac2008-04-23 04:59:35 +00001259return type of a function type is a scalar type, a void type, or a struct type.
Devang Patel9c1f8b12008-03-24 20:52:42 +00001260If the return type is a struct type then all struct elements must be of first
Chris Lattnerda508ac2008-04-23 04:59:35 +00001261class types, and the struct must have at least one element.</p>
Devang Pateld6cff512008-03-10 20:49:15 +00001262
Chris Lattner2f7c9632001-06-06 20:29:01 +00001263<h5>Syntax:</h5>
Chris Lattnerda508ac2008-04-23 04:59:35 +00001264
1265<pre>
1266 &lt;returntype list&gt; (&lt;parameter list&gt;)
1267</pre>
1268
John Criswell4c0cf7f2005-10-24 16:17:18 +00001269<p>...where '<tt>&lt;parameter list&gt;</tt>' is a comma-separated list of type
Misha Brukman20f9a622004-08-12 20:16:08 +00001270specifiers. Optionally, the parameter list may include a type <tt>...</tt>,
Chris Lattner5ed60612003-09-03 00:41:47 +00001271which indicates that the function takes a variable number of arguments.
1272Variable argument functions can access their arguments with the <a
Devang Pateld6cff512008-03-10 20:49:15 +00001273 href="#int_varargs">variable argument handling intrinsic</a> functions.
1274'<tt>&lt;returntype list&gt;</tt>' is a comma-separated list of
1275<a href="#t_firstclass">first class</a> type specifiers.</p>
Chris Lattnerda508ac2008-04-23 04:59:35 +00001276
Chris Lattner2f7c9632001-06-06 20:29:01 +00001277<h5>Examples:</h5>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +00001278<table class="layout">
1279 <tr class="layout">
Reid Spencer58c08712006-12-31 07:18:34 +00001280 <td class="left"><tt>i32 (i32)</tt></td>
1281 <td class="left">function taking an <tt>i32</tt>, returning an <tt>i32</tt>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +00001282 </td>
Reid Spencer58c08712006-12-31 07:18:34 +00001283 </tr><tr class="layout">
Reid Spencer314e1cb2007-07-19 23:13:04 +00001284 <td class="left"><tt>float&nbsp;(i16&nbsp;signext,&nbsp;i32&nbsp;*)&nbsp;*
Reid Spencer655dcc62006-12-31 07:20:23 +00001285 </tt></td>
Reid Spencer58c08712006-12-31 07:18:34 +00001286 <td class="left"><a href="#t_pointer">Pointer</a> to a function that takes
1287 an <tt>i16</tt> that should be sign extended and a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00001288 <a href="#t_pointer">pointer</a> to <tt>i32</tt>, returning
Reid Spencer58c08712006-12-31 07:18:34 +00001289 <tt>float</tt>.
1290 </td>
1291 </tr><tr class="layout">
1292 <td class="left"><tt>i32 (i8*, ...)</tt></td>
1293 <td class="left">A vararg function that takes at least one
Reid Spencer3e628eb92007-01-04 16:43:23 +00001294 <a href="#t_pointer">pointer</a> to <tt>i8 </tt> (char in C),
Reid Spencer58c08712006-12-31 07:18:34 +00001295 which returns an integer. This is the signature for <tt>printf</tt> in
1296 LLVM.
Reid Spencerc3c4c4f2004-11-01 08:19:36 +00001297 </td>
Devang Patele3dfc1c2008-03-24 05:35:41 +00001298 </tr><tr class="layout">
1299 <td class="left"><tt>{i32, i32} (i32)</tt></td>
Devang Patel8dec6c22008-03-24 18:10:52 +00001300 <td class="left">A function taking an <tt>i32></tt>, returning two
1301 <tt> i32 </tt> values as an aggregate of type <tt>{ i32, i32 }</tt>
Devang Patele3dfc1c2008-03-24 05:35:41 +00001302 </td>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +00001303 </tr>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001304</table>
Misha Brukmanc501f552004-03-01 17:47:27 +00001305
Misha Brukman76307852003-11-08 01:05:38 +00001306</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001307<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +00001308<div class="doc_subsubsection"> <a name="t_struct">Structure Type</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +00001309<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +00001310<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001311<p>The structure type is used to represent a collection of data members
1312together in memory. The packing of the field types is defined to match
1313the ABI of the underlying processor. The elements of a structure may
1314be any type that has a size.</p>
1315<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
1316and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
1317field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
1318instruction.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001319<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001320<pre> { &lt;type list&gt; }<br></pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001321<h5>Examples:</h5>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +00001322<table class="layout">
1323 <tr class="layout">
Jeff Cohen5819f182007-04-22 01:17:39 +00001324 <td class="left"><tt>{ i32, i32, i32 }</tt></td>
1325 <td class="left">A triple of three <tt>i32</tt> values</td>
1326 </tr><tr class="layout">
1327 <td class="left"><tt>{&nbsp;float,&nbsp;i32&nbsp;(i32)&nbsp;*&nbsp;}</tt></td>
1328 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1329 second element is a <a href="#t_pointer">pointer</a> to a
1330 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1331 an <tt>i32</tt>.</td>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +00001332 </tr>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001333</table>
Misha Brukman76307852003-11-08 01:05:38 +00001334</div>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +00001335
Chris Lattner2f7c9632001-06-06 20:29:01 +00001336<!-- _______________________________________________________________________ -->
Andrew Lenharth8df88e22006-12-08 17:13:00 +00001337<div class="doc_subsubsection"> <a name="t_pstruct">Packed Structure Type</a>
1338</div>
1339<div class="doc_text">
1340<h5>Overview:</h5>
1341<p>The packed structure type is used to represent a collection of data members
1342together in memory. There is no padding between fields. Further, the alignment
1343of a packed structure is 1 byte. The elements of a packed structure may
1344be any type that has a size.</p>
1345<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
1346and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
1347field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
1348instruction.</p>
1349<h5>Syntax:</h5>
1350<pre> &lt; { &lt;type list&gt; } &gt; <br></pre>
1351<h5>Examples:</h5>
1352<table class="layout">
1353 <tr class="layout">
Jeff Cohen5819f182007-04-22 01:17:39 +00001354 <td class="left"><tt>&lt; { i32, i32, i32 } &gt;</tt></td>
1355 <td class="left">A triple of three <tt>i32</tt> values</td>
1356 </tr><tr class="layout">
Chris Lattner747359f2007-12-19 05:04:11 +00001357 <td class="left"><tt>&lt; { float, i32 (i32)* } &gt;</tt></td>
Jeff Cohen5819f182007-04-22 01:17:39 +00001358 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1359 second element is a <a href="#t_pointer">pointer</a> to a
1360 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1361 an <tt>i32</tt>.</td>
Andrew Lenharth8df88e22006-12-08 17:13:00 +00001362 </tr>
1363</table>
1364</div>
1365
1366<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +00001367<div class="doc_subsubsection"> <a name="t_pointer">Pointer Type</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +00001368<div class="doc_text">
Chris Lattner590645f2002-04-14 06:13:44 +00001369<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001370<p>As in many languages, the pointer type represents a pointer or
Christopher Lamb308121c2007-12-11 09:31:00 +00001371reference to another object, which must live in memory. Pointer types may have
1372an optional address space attribute defining the target-specific numbered
1373address space where the pointed-to object resides. The default address space is
1374zero.</p>
Chris Lattner590645f2002-04-14 06:13:44 +00001375<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001376<pre> &lt;type&gt; *<br></pre>
Chris Lattner590645f2002-04-14 06:13:44 +00001377<h5>Examples:</h5>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +00001378<table class="layout">
1379 <tr class="layout">
Chris Lattner747359f2007-12-19 05:04:11 +00001380 <td class="left"><tt>[4x i32]*</tt></td>
1381 <td class="left">A <a href="#t_pointer">pointer</a> to <a
1382 href="#t_array">array</a> of four <tt>i32</tt> values.</td>
1383 </tr>
1384 <tr class="layout">
1385 <td class="left"><tt>i32 (i32 *) *</tt></td>
1386 <td class="left"> A <a href="#t_pointer">pointer</a> to a <a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00001387 href="#t_function">function</a> that takes an <tt>i32*</tt>, returning an
Chris Lattner747359f2007-12-19 05:04:11 +00001388 <tt>i32</tt>.</td>
1389 </tr>
1390 <tr class="layout">
1391 <td class="left"><tt>i32 addrspace(5)*</tt></td>
1392 <td class="left">A <a href="#t_pointer">pointer</a> to an <tt>i32</tt> value
1393 that resides in address space #5.</td>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +00001394 </tr>
Misha Brukman76307852003-11-08 01:05:38 +00001395</table>
Misha Brukman76307852003-11-08 01:05:38 +00001396</div>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +00001397
Chris Lattnerc8cb6952004-08-12 19:12:28 +00001398<!-- _______________________________________________________________________ -->
Reid Spencer404a3252007-02-15 03:07:05 +00001399<div class="doc_subsubsection"> <a name="t_vector">Vector Type</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +00001400<div class="doc_text">
Chris Lattner37b6b092005-04-25 17:34:15 +00001401
Chris Lattnerc8cb6952004-08-12 19:12:28 +00001402<h5>Overview:</h5>
Chris Lattner37b6b092005-04-25 17:34:15 +00001403
Reid Spencer404a3252007-02-15 03:07:05 +00001404<p>A vector type is a simple derived type that represents a vector
1405of elements. Vector types are used when multiple primitive data
Chris Lattnerc8cb6952004-08-12 19:12:28 +00001406are operated in parallel using a single instruction (SIMD).
Reid Spencer404a3252007-02-15 03:07:05 +00001407A vector type requires a size (number of
Chris Lattner330ce692005-11-10 01:44:22 +00001408elements) and an underlying primitive data type. Vectors must have a power
Reid Spencer404a3252007-02-15 03:07:05 +00001409of two length (1, 2, 4, 8, 16 ...). Vector types are
Chris Lattnerc8cb6952004-08-12 19:12:28 +00001410considered <a href="#t_firstclass">first class</a>.</p>
Chris Lattner37b6b092005-04-25 17:34:15 +00001411
Chris Lattnerc8cb6952004-08-12 19:12:28 +00001412<h5>Syntax:</h5>
Chris Lattner37b6b092005-04-25 17:34:15 +00001413
1414<pre>
1415 &lt; &lt;# elements&gt; x &lt;elementtype&gt; &gt;
1416</pre>
1417
John Criswell4a3327e2005-05-13 22:25:59 +00001418<p>The number of elements is a constant integer value; elementtype may
Chris Lattnerc0f423a2007-01-15 01:54:13 +00001419be any integer or floating point type.</p>
Chris Lattner37b6b092005-04-25 17:34:15 +00001420
Chris Lattnerc8cb6952004-08-12 19:12:28 +00001421<h5>Examples:</h5>
Chris Lattner37b6b092005-04-25 17:34:15 +00001422
Reid Spencerc3c4c4f2004-11-01 08:19:36 +00001423<table class="layout">
1424 <tr class="layout">
Chris Lattner747359f2007-12-19 05:04:11 +00001425 <td class="left"><tt>&lt;4 x i32&gt;</tt></td>
1426 <td class="left">Vector of 4 32-bit integer values.</td>
1427 </tr>
1428 <tr class="layout">
1429 <td class="left"><tt>&lt;8 x float&gt;</tt></td>
1430 <td class="left">Vector of 8 32-bit floating-point values.</td>
1431 </tr>
1432 <tr class="layout">
1433 <td class="left"><tt>&lt;2 x i64&gt;</tt></td>
1434 <td class="left">Vector of 2 64-bit integer values.</td>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +00001435 </tr>
1436</table>
Misha Brukman76307852003-11-08 01:05:38 +00001437</div>
1438
Chris Lattner37b6b092005-04-25 17:34:15 +00001439<!-- _______________________________________________________________________ -->
1440<div class="doc_subsubsection"> <a name="t_opaque">Opaque Type</a> </div>
1441<div class="doc_text">
1442
1443<h5>Overview:</h5>
1444
1445<p>Opaque types are used to represent unknown types in the system. This
Gordon Henriksena699c4d2007-10-14 00:34:53 +00001446corresponds (for example) to the C notion of a forward declared structure type.
Chris Lattner37b6b092005-04-25 17:34:15 +00001447In LLVM, opaque types can eventually be resolved to any type (not just a
1448structure type).</p>
1449
1450<h5>Syntax:</h5>
1451
1452<pre>
1453 opaque
1454</pre>
1455
1456<h5>Examples:</h5>
1457
1458<table class="layout">
1459 <tr class="layout">
Chris Lattner747359f2007-12-19 05:04:11 +00001460 <td class="left"><tt>opaque</tt></td>
1461 <td class="left">An opaque type.</td>
Chris Lattner37b6b092005-04-25 17:34:15 +00001462 </tr>
1463</table>
1464</div>
1465
1466
Chris Lattner74d3f822004-12-09 17:30:23 +00001467<!-- *********************************************************************** -->
1468<div class="doc_section"> <a name="constants">Constants</a> </div>
1469<!-- *********************************************************************** -->
1470
1471<div class="doc_text">
1472
1473<p>LLVM has several different basic types of constants. This section describes
1474them all and their syntax.</p>
1475
1476</div>
1477
1478<!-- ======================================================================= -->
Reid Spencer8f08d802004-12-09 18:02:53 +00001479<div class="doc_subsection"><a name="simpleconstants">Simple Constants</a></div>
Chris Lattner74d3f822004-12-09 17:30:23 +00001480
1481<div class="doc_text">
1482
1483<dl>
1484 <dt><b>Boolean constants</b></dt>
1485
1486 <dd>The two strings '<tt>true</tt>' and '<tt>false</tt>' are both valid
Reid Spencer36a15422007-01-12 03:35:51 +00001487 constants of the <tt><a href="#t_primitive">i1</a></tt> type.
Chris Lattner74d3f822004-12-09 17:30:23 +00001488 </dd>
1489
1490 <dt><b>Integer constants</b></dt>
1491
Reid Spencer8f08d802004-12-09 18:02:53 +00001492 <dd>Standard integers (such as '4') are constants of the <a
Reid Spencer3e628eb92007-01-04 16:43:23 +00001493 href="#t_integer">integer</a> type. Negative numbers may be used with
Chris Lattner74d3f822004-12-09 17:30:23 +00001494 integer types.
1495 </dd>
1496
1497 <dt><b>Floating point constants</b></dt>
1498
1499 <dd>Floating point constants use standard decimal notation (e.g. 123.421),
1500 exponential notation (e.g. 1.23421e+2), or a more precise hexadecimal
Chris Lattner1429e6f2008-04-01 18:45:27 +00001501 notation (see below). The assembler requires the exact decimal value of
1502 a floating-point constant. For example, the assembler accepts 1.25 but
1503 rejects 1.3 because 1.3 is a repeating decimal in binary. Floating point
1504 constants must have a <a href="#t_floating">floating point</a> type. </dd>
Chris Lattner74d3f822004-12-09 17:30:23 +00001505
1506 <dt><b>Null pointer constants</b></dt>
1507
John Criswelldfe6a862004-12-10 15:51:16 +00001508 <dd>The identifier '<tt>null</tt>' is recognized as a null pointer constant
Chris Lattner74d3f822004-12-09 17:30:23 +00001509 and must be of <a href="#t_pointer">pointer type</a>.</dd>
1510
1511</dl>
1512
John Criswelldfe6a862004-12-10 15:51:16 +00001513<p>The one non-intuitive notation for constants is the optional hexadecimal form
Chris Lattner74d3f822004-12-09 17:30:23 +00001514of floating point constants. For example, the form '<tt>double
15150x432ff973cafa8000</tt>' is equivalent to (but harder to read than) '<tt>double
15164.5e+15</tt>'. The only time hexadecimal floating point constants are required
Reid Spencer8f08d802004-12-09 18:02:53 +00001517(and the only time that they are generated by the disassembler) is when a
1518floating point constant must be emitted but it cannot be represented as a
1519decimal floating point number. For example, NaN's, infinities, and other
1520special values are represented in their IEEE hexadecimal format so that
1521assembly and disassembly do not cause any bits to change in the constants.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +00001522
1523</div>
1524
1525<!-- ======================================================================= -->
1526<div class="doc_subsection"><a name="aggregateconstants">Aggregate Constants</a>
1527</div>
1528
1529<div class="doc_text">
Chris Lattner455fc8c2005-03-07 22:13:59 +00001530<p>Aggregate constants arise from aggregation of simple constants
1531and smaller aggregate constants.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +00001532
1533<dl>
1534 <dt><b>Structure constants</b></dt>
1535
1536 <dd>Structure constants are represented with notation similar to structure
1537 type definitions (a comma separated list of elements, surrounded by braces
Chris Lattnerbea11172007-12-25 20:34:52 +00001538 (<tt>{}</tt>)). For example: "<tt>{ i32 4, float 17.0, i32* @G }</tt>",
1539 where "<tt>@G</tt>" is declared as "<tt>@G = external global i32</tt>". Structure constants
Chris Lattner455fc8c2005-03-07 22:13:59 +00001540 must have <a href="#t_struct">structure type</a>, and the number and
Chris Lattner74d3f822004-12-09 17:30:23 +00001541 types of elements must match those specified by the type.
1542 </dd>
1543
1544 <dt><b>Array constants</b></dt>
1545
1546 <dd>Array constants are represented with notation similar to array type
1547 definitions (a comma separated list of elements, surrounded by square brackets
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00001548 (<tt>[]</tt>)). For example: "<tt>[ i32 42, i32 11, i32 74 ]</tt>". Array
Chris Lattner74d3f822004-12-09 17:30:23 +00001549 constants must have <a href="#t_array">array type</a>, and the number and
1550 types of elements must match those specified by the type.
1551 </dd>
1552
Reid Spencer404a3252007-02-15 03:07:05 +00001553 <dt><b>Vector constants</b></dt>
Chris Lattner74d3f822004-12-09 17:30:23 +00001554
Reid Spencer404a3252007-02-15 03:07:05 +00001555 <dd>Vector constants are represented with notation similar to vector type
Chris Lattner74d3f822004-12-09 17:30:23 +00001556 definitions (a comma separated list of elements, surrounded by
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00001557 less-than/greater-than's (<tt>&lt;&gt;</tt>)). For example: "<tt>&lt; i32 42,
Jeff Cohen5819f182007-04-22 01:17:39 +00001558 i32 11, i32 74, i32 100 &gt;</tt>". Vector constants must have <a
Reid Spencer404a3252007-02-15 03:07:05 +00001559 href="#t_vector">vector type</a>, and the number and types of elements must
Chris Lattner74d3f822004-12-09 17:30:23 +00001560 match those specified by the type.
1561 </dd>
1562
1563 <dt><b>Zero initialization</b></dt>
1564
1565 <dd>The string '<tt>zeroinitializer</tt>' can be used to zero initialize a
1566 value to zero of <em>any</em> type, including scalar and aggregate types.
1567 This is often used to avoid having to print large zero initializers (e.g. for
John Criswell4c0cf7f2005-10-24 16:17:18 +00001568 large arrays) and is always exactly equivalent to using explicit zero
Chris Lattner74d3f822004-12-09 17:30:23 +00001569 initializers.
1570 </dd>
1571</dl>
1572
1573</div>
1574
1575<!-- ======================================================================= -->
1576<div class="doc_subsection">
1577 <a name="globalconstants">Global Variable and Function Addresses</a>
1578</div>
1579
1580<div class="doc_text">
1581
1582<p>The addresses of <a href="#globalvars">global variables</a> and <a
1583href="#functionstructure">functions</a> are always implicitly valid (link-time)
John Criswelldfe6a862004-12-10 15:51:16 +00001584constants. These constants are explicitly referenced when the <a
1585href="#identifiers">identifier for the global</a> is used and always have <a
Chris Lattner74d3f822004-12-09 17:30:23 +00001586href="#t_pointer">pointer</a> type. For example, the following is a legal LLVM
1587file:</p>
1588
Bill Wendling3716c5d2007-05-29 09:04:49 +00001589<div class="doc_code">
Chris Lattner74d3f822004-12-09 17:30:23 +00001590<pre>
Chris Lattner00538a12007-06-06 18:28:13 +00001591@X = global i32 17
1592@Y = global i32 42
1593@Z = global [2 x i32*] [ i32* @X, i32* @Y ]
Chris Lattner74d3f822004-12-09 17:30:23 +00001594</pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +00001595</div>
Chris Lattner74d3f822004-12-09 17:30:23 +00001596
1597</div>
1598
1599<!-- ======================================================================= -->
Reid Spencer641f5c92004-12-09 18:13:12 +00001600<div class="doc_subsection"><a name="undefvalues">Undefined Values</a></div>
Chris Lattner74d3f822004-12-09 17:30:23 +00001601<div class="doc_text">
Reid Spencer641f5c92004-12-09 18:13:12 +00001602 <p>The string '<tt>undef</tt>' is recognized as a type-less constant that has
John Criswell4a3327e2005-05-13 22:25:59 +00001603 no specific value. Undefined values may be of any type and be used anywhere
Reid Spencer641f5c92004-12-09 18:13:12 +00001604 a constant is permitted.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +00001605
Reid Spencer641f5c92004-12-09 18:13:12 +00001606 <p>Undefined values indicate to the compiler that the program is well defined
1607 no matter what value is used, giving the compiler more freedom to optimize.
1608 </p>
Chris Lattner74d3f822004-12-09 17:30:23 +00001609</div>
1610
1611<!-- ======================================================================= -->
1612<div class="doc_subsection"><a name="constantexprs">Constant Expressions</a>
1613</div>
1614
1615<div class="doc_text">
1616
1617<p>Constant expressions are used to allow expressions involving other constants
1618to be used as constants. Constant expressions may be of any <a
John Criswell4a3327e2005-05-13 22:25:59 +00001619href="#t_firstclass">first class</a> type and may involve any LLVM operation
Chris Lattner74d3f822004-12-09 17:30:23 +00001620that does not have side effects (e.g. load and call are not supported). The
1621following is the syntax for constant expressions:</p>
1622
1623<dl>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00001624 <dt><b><tt>trunc ( CST to TYPE )</tt></b></dt>
1625 <dd>Truncate a constant to another type. The bit size of CST must be larger
Chris Lattnerc0f423a2007-01-15 01:54:13 +00001626 than the bit size of TYPE. Both types must be integers.</dd>
Chris Lattner74d3f822004-12-09 17:30:23 +00001627
Reid Spencer59b6b7d2006-11-08 01:11:31 +00001628 <dt><b><tt>zext ( CST to TYPE )</tt></b></dt>
1629 <dd>Zero extend a constant to another type. The bit size of CST must be
Chris Lattnerc0f423a2007-01-15 01:54:13 +00001630 smaller or equal to the bit size of TYPE. Both types must be integers.</dd>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00001631
1632 <dt><b><tt>sext ( CST to TYPE )</tt></b></dt>
1633 <dd>Sign extend a constant to another type. The bit size of CST must be
Chris Lattnerc0f423a2007-01-15 01:54:13 +00001634 smaller or equal to the bit size of TYPE. Both types must be integers.</dd>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00001635
1636 <dt><b><tt>fptrunc ( CST to TYPE )</tt></b></dt>
1637 <dd>Truncate a floating point constant to another floating point type. The
1638 size of CST must be larger than the size of TYPE. Both types must be
1639 floating point.</dd>
1640
1641 <dt><b><tt>fpext ( CST to TYPE )</tt></b></dt>
1642 <dd>Floating point extend a constant to another type. The size of CST must be
1643 smaller or equal to the size of TYPE. Both types must be floating point.</dd>
1644
Reid Spencer753163d2007-07-31 14:40:14 +00001645 <dt><b><tt>fptoui ( CST to TYPE )</tt></b></dt>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00001646 <dd>Convert a floating point constant to the corresponding unsigned integer
Nate Begemand4d45c22007-11-17 03:58:34 +00001647 constant. TYPE must be a scalar or vector integer type. CST must be of scalar
1648 or vector floating point type. Both CST and TYPE must be scalars, or vectors
1649 of the same number of elements. If the value won't fit in the integer type,
1650 the results are undefined.</dd>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00001651
Reid Spencer51b07252006-11-09 23:03:26 +00001652 <dt><b><tt>fptosi ( CST to TYPE )</tt></b></dt>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00001653 <dd>Convert a floating point constant to the corresponding signed integer
Nate Begemand4d45c22007-11-17 03:58:34 +00001654 constant. TYPE must be a scalar or vector integer type. CST must be of scalar
1655 or vector floating point type. Both CST and TYPE must be scalars, or vectors
1656 of the same number of elements. If the value won't fit in the integer type,
1657 the results are undefined.</dd>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00001658
Reid Spencer51b07252006-11-09 23:03:26 +00001659 <dt><b><tt>uitofp ( CST to TYPE )</tt></b></dt>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00001660 <dd>Convert an unsigned integer constant to the corresponding floating point
Nate Begemand4d45c22007-11-17 03:58:34 +00001661 constant. TYPE must be a scalar or vector floating point type. CST must be of
1662 scalar or vector integer type. Both CST and TYPE must be scalars, or vectors
1663 of the same number of elements. If the value won't fit in the floating point
1664 type, the results are undefined.</dd>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00001665
Reid Spencer51b07252006-11-09 23:03:26 +00001666 <dt><b><tt>sitofp ( CST to TYPE )</tt></b></dt>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00001667 <dd>Convert a signed integer constant to the corresponding floating point
Nate Begemand4d45c22007-11-17 03:58:34 +00001668 constant. TYPE must be a scalar or vector floating point type. CST must be of
1669 scalar or vector integer type. Both CST and TYPE must be scalars, or vectors
1670 of the same number of elements. If the value won't fit in the floating point
1671 type, the results are undefined.</dd>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00001672
Reid Spencer5b950642006-11-11 23:08:07 +00001673 <dt><b><tt>ptrtoint ( CST to TYPE )</tt></b></dt>
1674 <dd>Convert a pointer typed constant to the corresponding integer constant
1675 TYPE must be an integer type. CST must be of pointer type. The CST value is
1676 zero extended, truncated, or unchanged to make it fit in TYPE.</dd>
1677
1678 <dt><b><tt>inttoptr ( CST to TYPE )</tt></b></dt>
1679 <dd>Convert a integer constant to a pointer constant. TYPE must be a
1680 pointer type. CST must be of integer type. The CST value is zero extended,
1681 truncated, or unchanged to make it fit in a pointer size. This one is
1682 <i>really</i> dangerous!</dd>
1683
1684 <dt><b><tt>bitcast ( CST to TYPE )</tt></b></dt>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00001685 <dd>Convert a constant, CST, to another TYPE. The size of CST and TYPE must be
1686 identical (same number of bits). The conversion is done as if the CST value
1687 was stored to memory and read back as TYPE. In other words, no bits change
Reid Spencer5b950642006-11-11 23:08:07 +00001688 with this operator, just the type. This can be used for conversion of
Reid Spencer404a3252007-02-15 03:07:05 +00001689 vector types to any other type, as long as they have the same bit width. For
Reid Spencer5b950642006-11-11 23:08:07 +00001690 pointers it is only valid to cast to another pointer type.
Reid Spencer59b6b7d2006-11-08 01:11:31 +00001691 </dd>
Chris Lattner74d3f822004-12-09 17:30:23 +00001692
1693 <dt><b><tt>getelementptr ( CSTPTR, IDX0, IDX1, ... )</tt></b></dt>
1694
1695 <dd>Perform the <a href="#i_getelementptr">getelementptr operation</a> on
1696 constants. As with the <a href="#i_getelementptr">getelementptr</a>
1697 instruction, the index list may have zero or more indexes, which are required
1698 to make sense for the type of "CSTPTR".</dd>
1699
Robert Bocchino7e97a6d2006-01-10 19:31:34 +00001700 <dt><b><tt>select ( COND, VAL1, VAL2 )</tt></b></dt>
1701
1702 <dd>Perform the <a href="#i_select">select operation</a> on
Reid Spencer9965ee72006-12-04 19:23:19 +00001703 constants.</dd>
1704
1705 <dt><b><tt>icmp COND ( VAL1, VAL2 )</tt></b></dt>
1706 <dd>Performs the <a href="#i_icmp">icmp operation</a> on constants.</dd>
1707
1708 <dt><b><tt>fcmp COND ( VAL1, VAL2 )</tt></b></dt>
1709 <dd>Performs the <a href="#i_fcmp">fcmp operation</a> on constants.</dd>
Robert Bocchino7e97a6d2006-01-10 19:31:34 +00001710
Nate Begemand2195702008-05-12 19:01:56 +00001711 <dt><b><tt>vicmp COND ( VAL1, VAL2 )</tt></b></dt>
1712 <dd>Performs the <a href="#i_vicmp">vicmp operation</a> on constants.</dd>
1713
1714 <dt><b><tt>vfcmp COND ( VAL1, VAL2 )</tt></b></dt>
1715 <dd>Performs the <a href="#i_vfcmp">vfcmp operation</a> on constants.</dd>
1716
Robert Bocchino7e97a6d2006-01-10 19:31:34 +00001717 <dt><b><tt>extractelement ( VAL, IDX )</tt></b></dt>
1718
1719 <dd>Perform the <a href="#i_extractelement">extractelement
1720 operation</a> on constants.
1721
Robert Bocchinof72fdfe2006-01-15 20:48:27 +00001722 <dt><b><tt>insertelement ( VAL, ELT, IDX )</tt></b></dt>
1723
1724 <dd>Perform the <a href="#i_insertelement">insertelement
Reid Spencer9965ee72006-12-04 19:23:19 +00001725 operation</a> on constants.</dd>
Robert Bocchinof72fdfe2006-01-15 20:48:27 +00001726
Chris Lattner016a0e52006-04-08 00:13:41 +00001727
1728 <dt><b><tt>shufflevector ( VEC1, VEC2, IDXMASK )</tt></b></dt>
1729
1730 <dd>Perform the <a href="#i_shufflevector">shufflevector
Reid Spencer9965ee72006-12-04 19:23:19 +00001731 operation</a> on constants.</dd>
Chris Lattner016a0e52006-04-08 00:13:41 +00001732
Chris Lattner74d3f822004-12-09 17:30:23 +00001733 <dt><b><tt>OPCODE ( LHS, RHS )</tt></b></dt>
1734
Reid Spencer641f5c92004-12-09 18:13:12 +00001735 <dd>Perform the specified operation of the LHS and RHS constants. OPCODE may
1736 be any of the <a href="#binaryops">binary</a> or <a href="#bitwiseops">bitwise
Chris Lattner74d3f822004-12-09 17:30:23 +00001737 binary</a> operations. The constraints on operands are the same as those for
1738 the corresponding instruction (e.g. no bitwise operations on floating point
John Criswell02fdc6f2005-05-12 16:52:32 +00001739 values are allowed).</dd>
Chris Lattner74d3f822004-12-09 17:30:23 +00001740</dl>
Chris Lattner74d3f822004-12-09 17:30:23 +00001741</div>
Chris Lattnerb1652612004-03-08 16:49:10 +00001742
Chris Lattner2f7c9632001-06-06 20:29:01 +00001743<!-- *********************************************************************** -->
Chris Lattner98f013c2006-01-25 23:47:57 +00001744<div class="doc_section"> <a name="othervalues">Other Values</a> </div>
1745<!-- *********************************************************************** -->
1746
1747<!-- ======================================================================= -->
1748<div class="doc_subsection">
1749<a name="inlineasm">Inline Assembler Expressions</a>
1750</div>
1751
1752<div class="doc_text">
1753
1754<p>
1755LLVM supports inline assembler expressions (as opposed to <a href="#moduleasm">
1756Module-Level Inline Assembly</a>) through the use of a special value. This
1757value represents the inline assembler as a string (containing the instructions
1758to emit), a list of operand constraints (stored as a string), and a flag that
1759indicates whether or not the inline asm expression has side effects. An example
1760inline assembler expression is:
1761</p>
1762
Bill Wendling3716c5d2007-05-29 09:04:49 +00001763<div class="doc_code">
Chris Lattner98f013c2006-01-25 23:47:57 +00001764<pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +00001765i32 (i32) asm "bswap $0", "=r,r"
Chris Lattner98f013c2006-01-25 23:47:57 +00001766</pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +00001767</div>
Chris Lattner98f013c2006-01-25 23:47:57 +00001768
1769<p>
1770Inline assembler expressions may <b>only</b> be used as the callee operand of
1771a <a href="#i_call"><tt>call</tt> instruction</a>. Thus, typically we have:
1772</p>
1773
Bill Wendling3716c5d2007-05-29 09:04:49 +00001774<div class="doc_code">
Chris Lattner98f013c2006-01-25 23:47:57 +00001775<pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +00001776%X = call i32 asm "<a href="#int_bswap">bswap</a> $0", "=r,r"(i32 %Y)
Chris Lattner98f013c2006-01-25 23:47:57 +00001777</pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +00001778</div>
Chris Lattner98f013c2006-01-25 23:47:57 +00001779
1780<p>
1781Inline asms with side effects not visible in the constraint list must be marked
1782as having side effects. This is done through the use of the
1783'<tt>sideeffect</tt>' keyword, like so:
1784</p>
1785
Bill Wendling3716c5d2007-05-29 09:04:49 +00001786<div class="doc_code">
Chris Lattner98f013c2006-01-25 23:47:57 +00001787<pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +00001788call void asm sideeffect "eieio", ""()
Chris Lattner98f013c2006-01-25 23:47:57 +00001789</pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +00001790</div>
Chris Lattner98f013c2006-01-25 23:47:57 +00001791
1792<p>TODO: The format of the asm and constraints string still need to be
1793documented here. Constraints on what can be done (e.g. duplication, moving, etc
1794need to be documented).
1795</p>
1796
1797</div>
1798
1799<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +00001800<div class="doc_section"> <a name="instref">Instruction Reference</a> </div>
1801<!-- *********************************************************************** -->
Chris Lattner74d3f822004-12-09 17:30:23 +00001802
Misha Brukman76307852003-11-08 01:05:38 +00001803<div class="doc_text">
Chris Lattner74d3f822004-12-09 17:30:23 +00001804
Chris Lattner48b383b02003-11-25 01:02:51 +00001805<p>The LLVM instruction set consists of several different
1806classifications of instructions: <a href="#terminators">terminator
John Criswell4a3327e2005-05-13 22:25:59 +00001807instructions</a>, <a href="#binaryops">binary instructions</a>,
1808<a href="#bitwiseops">bitwise binary instructions</a>, <a
Chris Lattner48b383b02003-11-25 01:02:51 +00001809 href="#memoryops">memory instructions</a>, and <a href="#otherops">other
1810instructions</a>.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +00001811
Misha Brukman76307852003-11-08 01:05:38 +00001812</div>
Chris Lattner74d3f822004-12-09 17:30:23 +00001813
Chris Lattner2f7c9632001-06-06 20:29:01 +00001814<!-- ======================================================================= -->
Chris Lattner48b383b02003-11-25 01:02:51 +00001815<div class="doc_subsection"> <a name="terminators">Terminator
1816Instructions</a> </div>
Chris Lattner74d3f822004-12-09 17:30:23 +00001817
Misha Brukman76307852003-11-08 01:05:38 +00001818<div class="doc_text">
Chris Lattner74d3f822004-12-09 17:30:23 +00001819
Chris Lattner48b383b02003-11-25 01:02:51 +00001820<p>As mentioned <a href="#functionstructure">previously</a>, every
1821basic block in a program ends with a "Terminator" instruction, which
1822indicates which block should be executed after the current block is
1823finished. These terminator instructions typically yield a '<tt>void</tt>'
1824value: they produce control flow, not values (the one exception being
1825the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction).</p>
John Criswelldfe6a862004-12-10 15:51:16 +00001826<p>There are six different terminator instructions: the '<a
Chris Lattner48b383b02003-11-25 01:02:51 +00001827 href="#i_ret"><tt>ret</tt></a>' instruction, the '<a href="#i_br"><tt>br</tt></a>'
1828instruction, the '<a href="#i_switch"><tt>switch</tt></a>' instruction,
Chris Lattner08b7d5b2004-10-16 18:04:13 +00001829the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction, the '<a
1830 href="#i_unwind"><tt>unwind</tt></a>' instruction, and the '<a
1831 href="#i_unreachable"><tt>unreachable</tt></a>' instruction.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +00001832
Misha Brukman76307852003-11-08 01:05:38 +00001833</div>
Chris Lattner74d3f822004-12-09 17:30:23 +00001834
Chris Lattner2f7c9632001-06-06 20:29:01 +00001835<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +00001836<div class="doc_subsubsection"> <a name="i_ret">'<tt>ret</tt>'
1837Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +00001838<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +00001839<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001840<pre> ret &lt;type&gt; &lt;value&gt; <i>; Return a value from a non-void function</i>
Chris Lattner590645f2002-04-14 06:13:44 +00001841 ret void <i>; Return from void function</i>
Devang Pateld6cff512008-03-10 20:49:15 +00001842 ret &lt;type&gt; &lt;value&gt;, &lt;type&gt; &lt;value&gt; <i>; Return two values from a non-void function </i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001843</pre>
Chris Lattnerda508ac2008-04-23 04:59:35 +00001844
Chris Lattner2f7c9632001-06-06 20:29:01 +00001845<h5>Overview:</h5>
Chris Lattnerda508ac2008-04-23 04:59:35 +00001846
Chris Lattner48b383b02003-11-25 01:02:51 +00001847<p>The '<tt>ret</tt>' instruction is used to return control flow (and a
John Criswell4a3327e2005-05-13 22:25:59 +00001848value) from a function back to the caller.</p>
John Criswell417228d2004-04-09 16:48:45 +00001849<p>There are two forms of the '<tt>ret</tt>' instruction: one that
Chris Lattnerda508ac2008-04-23 04:59:35 +00001850returns value(s) and then causes control flow, and one that just causes
Chris Lattner48b383b02003-11-25 01:02:51 +00001851control flow to occur.</p>
Chris Lattnerda508ac2008-04-23 04:59:35 +00001852
Chris Lattner2f7c9632001-06-06 20:29:01 +00001853<h5>Arguments:</h5>
Chris Lattnerda508ac2008-04-23 04:59:35 +00001854
1855<p>The '<tt>ret</tt>' instruction may return zero, one or multiple values.
1856The type of each return value must be a '<a href="#t_firstclass">first
1857class</a>' type. Note that a function is not <a href="#wellformed">well
1858formed</a> if there exists a '<tt>ret</tt>' instruction inside of the
1859function that returns values that do not match the return type of the
1860function.</p>
1861
Chris Lattner2f7c9632001-06-06 20:29:01 +00001862<h5>Semantics:</h5>
Chris Lattnerda508ac2008-04-23 04:59:35 +00001863
Chris Lattner48b383b02003-11-25 01:02:51 +00001864<p>When the '<tt>ret</tt>' instruction is executed, control flow
1865returns back to the calling function's context. If the caller is a "<a
John Criswell40db33f2004-06-25 15:16:57 +00001866 href="#i_call"><tt>call</tt></a>" instruction, execution continues at
Chris Lattner48b383b02003-11-25 01:02:51 +00001867the instruction after the call. If the caller was an "<a
1868 href="#i_invoke"><tt>invoke</tt></a>" instruction, execution continues
John Criswell02fdc6f2005-05-12 16:52:32 +00001869at the beginning of the "normal" destination block. If the instruction
Chris Lattner48b383b02003-11-25 01:02:51 +00001870returns a value, that value shall set the call or invoke instruction's
Devang Pateld6cff512008-03-10 20:49:15 +00001871return value. If the instruction returns multiple values then these
Devang Pateld0f47642008-03-11 05:51:59 +00001872values can only be accessed through a '<a href="#i_getresult"><tt>getresult</tt>
1873</a>' instruction.</p>
Chris Lattnerda508ac2008-04-23 04:59:35 +00001874
Chris Lattner2f7c9632001-06-06 20:29:01 +00001875<h5>Example:</h5>
Chris Lattnerda508ac2008-04-23 04:59:35 +00001876
1877<pre>
1878 ret i32 5 <i>; Return an integer value of 5</i>
Chris Lattner590645f2002-04-14 06:13:44 +00001879 ret void <i>; Return from a void function</i>
Devang Pateld6cff512008-03-10 20:49:15 +00001880 ret i32 4, i8 2 <i>; Return two values 4 and 2 </i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001881</pre>
Misha Brukman76307852003-11-08 01:05:38 +00001882</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001883<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +00001884<div class="doc_subsubsection"> <a name="i_br">'<tt>br</tt>' Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +00001885<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +00001886<h5>Syntax:</h5>
Reid Spencer36a15422007-01-12 03:35:51 +00001887<pre> br i1 &lt;cond&gt;, label &lt;iftrue&gt;, label &lt;iffalse&gt;<br> br label &lt;dest&gt; <i>; Unconditional branch</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001888</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001889<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001890<p>The '<tt>br</tt>' instruction is used to cause control flow to
1891transfer to a different basic block in the current function. There are
1892two forms of this instruction, corresponding to a conditional branch
1893and an unconditional branch.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001894<h5>Arguments:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001895<p>The conditional branch form of the '<tt>br</tt>' instruction takes a
Reid Spencer36a15422007-01-12 03:35:51 +00001896single '<tt>i1</tt>' value and two '<tt>label</tt>' values. The
Reid Spencer50c723a2007-02-19 23:54:10 +00001897unconditional form of the '<tt>br</tt>' instruction takes a single
1898'<tt>label</tt>' value as a target.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001899<h5>Semantics:</h5>
Reid Spencer36a15422007-01-12 03:35:51 +00001900<p>Upon execution of a conditional '<tt>br</tt>' instruction, the '<tt>i1</tt>'
Chris Lattner48b383b02003-11-25 01:02:51 +00001901argument is evaluated. If the value is <tt>true</tt>, control flows
1902to the '<tt>iftrue</tt>' <tt>label</tt> argument. If "cond" is <tt>false</tt>,
1903control flows to the '<tt>iffalse</tt>' <tt>label</tt> argument.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001904<h5>Example:</h5>
Reid Spencer36a15422007-01-12 03:35:51 +00001905<pre>Test:<br> %cond = <a href="#i_icmp">icmp</a> eq, i32 %a, %b<br> br i1 %cond, label %IfEqual, label %IfUnequal<br>IfEqual:<br> <a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00001906 href="#i_ret">ret</a> i32 1<br>IfUnequal:<br> <a href="#i_ret">ret</a> i32 0<br></pre>
Misha Brukman76307852003-11-08 01:05:38 +00001907</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001908<!-- _______________________________________________________________________ -->
Chris Lattnercf96c6c2004-02-24 04:54:45 +00001909<div class="doc_subsubsection">
1910 <a name="i_switch">'<tt>switch</tt>' Instruction</a>
1911</div>
1912
Misha Brukman76307852003-11-08 01:05:38 +00001913<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +00001914<h5>Syntax:</h5>
Chris Lattnercf96c6c2004-02-24 04:54:45 +00001915
1916<pre>
1917 switch &lt;intty&gt; &lt;value&gt;, label &lt;defaultdest&gt; [ &lt;intty&gt; &lt;val&gt;, label &lt;dest&gt; ... ]
1918</pre>
1919
Chris Lattner2f7c9632001-06-06 20:29:01 +00001920<h5>Overview:</h5>
Chris Lattnercf96c6c2004-02-24 04:54:45 +00001921
1922<p>The '<tt>switch</tt>' instruction is used to transfer control flow to one of
1923several different places. It is a generalization of the '<tt>br</tt>'
Misha Brukman76307852003-11-08 01:05:38 +00001924instruction, allowing a branch to occur to one of many possible
1925destinations.</p>
Chris Lattnercf96c6c2004-02-24 04:54:45 +00001926
1927
Chris Lattner2f7c9632001-06-06 20:29:01 +00001928<h5>Arguments:</h5>
Chris Lattnercf96c6c2004-02-24 04:54:45 +00001929
1930<p>The '<tt>switch</tt>' instruction uses three parameters: an integer
1931comparison value '<tt>value</tt>', a default '<tt>label</tt>' destination, and
1932an array of pairs of comparison value constants and '<tt>label</tt>'s. The
1933table is not allowed to contain duplicate constant entries.</p>
1934
Chris Lattner2f7c9632001-06-06 20:29:01 +00001935<h5>Semantics:</h5>
Chris Lattnercf96c6c2004-02-24 04:54:45 +00001936
Chris Lattner48b383b02003-11-25 01:02:51 +00001937<p>The <tt>switch</tt> instruction specifies a table of values and
1938destinations. When the '<tt>switch</tt>' instruction is executed, this
John Criswellbcbb18c2004-06-25 16:05:06 +00001939table is searched for the given value. If the value is found, control flow is
1940transfered to the corresponding destination; otherwise, control flow is
1941transfered to the default destination.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001942
Chris Lattnercf96c6c2004-02-24 04:54:45 +00001943<h5>Implementation:</h5>
1944
1945<p>Depending on properties of the target machine and the particular
1946<tt>switch</tt> instruction, this instruction may be code generated in different
John Criswellbcbb18c2004-06-25 16:05:06 +00001947ways. For example, it could be generated as a series of chained conditional
1948branches or with a lookup table.</p>
Chris Lattnercf96c6c2004-02-24 04:54:45 +00001949
1950<h5>Example:</h5>
1951
1952<pre>
1953 <i>; Emulate a conditional br instruction</i>
Reid Spencer36a15422007-01-12 03:35:51 +00001954 %Val = <a href="#i_zext">zext</a> i1 %value to i32
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00001955 switch i32 %Val, label %truedest [i32 0, label %falsedest ]
Chris Lattnercf96c6c2004-02-24 04:54:45 +00001956
1957 <i>; Emulate an unconditional br instruction</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00001958 switch i32 0, label %dest [ ]
Chris Lattnercf96c6c2004-02-24 04:54:45 +00001959
1960 <i>; Implement a jump table:</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00001961 switch i32 %val, label %otherwise [ i32 0, label %onzero
1962 i32 1, label %onone
1963 i32 2, label %ontwo ]
Chris Lattner2f7c9632001-06-06 20:29:01 +00001964</pre>
Misha Brukman76307852003-11-08 01:05:38 +00001965</div>
Chris Lattner0132aff2005-05-06 22:57:40 +00001966
Chris Lattner2f7c9632001-06-06 20:29:01 +00001967<!-- _______________________________________________________________________ -->
Chris Lattner0132aff2005-05-06 22:57:40 +00001968<div class="doc_subsubsection">
1969 <a name="i_invoke">'<tt>invoke</tt>' Instruction</a>
1970</div>
1971
Misha Brukman76307852003-11-08 01:05:38 +00001972<div class="doc_text">
Chris Lattner0132aff2005-05-06 22:57:40 +00001973
Chris Lattner2f7c9632001-06-06 20:29:01 +00001974<h5>Syntax:</h5>
Chris Lattner0132aff2005-05-06 22:57:40 +00001975
1976<pre>
Nick Lewycky084ab472008-03-16 07:18:12 +00001977 &lt;result&gt; = invoke [<a href="#callingconv">cconv</a>] &lt;ptr to function ty&gt; &lt;function ptr val&gt;(&lt;function args&gt;)
Chris Lattner6b7a0082006-05-14 18:23:06 +00001978 to label &lt;normal label&gt; unwind label &lt;exception label&gt;
Chris Lattner0132aff2005-05-06 22:57:40 +00001979</pre>
1980
Chris Lattnera8292f32002-05-06 22:08:29 +00001981<h5>Overview:</h5>
Chris Lattner0132aff2005-05-06 22:57:40 +00001982
1983<p>The '<tt>invoke</tt>' instruction causes control to transfer to a specified
1984function, with the possibility of control flow transfer to either the
John Criswell02fdc6f2005-05-12 16:52:32 +00001985'<tt>normal</tt>' label or the
1986'<tt>exception</tt>' label. If the callee function returns with the
Chris Lattner0132aff2005-05-06 22:57:40 +00001987"<tt><a href="#i_ret">ret</a></tt>" instruction, control flow will return to the
1988"normal" label. If the callee (or any indirect callees) returns with the "<a
John Criswell02fdc6f2005-05-12 16:52:32 +00001989href="#i_unwind"><tt>unwind</tt></a>" instruction, control is interrupted and
Devang Pateld6cff512008-03-10 20:49:15 +00001990continued at the dynamically nearest "exception" label. If the callee function
Devang Pateld0f47642008-03-11 05:51:59 +00001991returns multiple values then individual return values are only accessible through
1992a '<tt><a href="#i_getresult">getresult</a></tt>' instruction.</p>
Chris Lattner0132aff2005-05-06 22:57:40 +00001993
Chris Lattner2f7c9632001-06-06 20:29:01 +00001994<h5>Arguments:</h5>
Chris Lattner0132aff2005-05-06 22:57:40 +00001995
Misha Brukman76307852003-11-08 01:05:38 +00001996<p>This instruction requires several arguments:</p>
Chris Lattner0132aff2005-05-06 22:57:40 +00001997
Chris Lattner2f7c9632001-06-06 20:29:01 +00001998<ol>
Chris Lattner0132aff2005-05-06 22:57:40 +00001999 <li>
Duncan Sands16f122e2007-03-30 12:22:09 +00002000 The optional "cconv" marker indicates which <a href="#callingconv">calling
Chris Lattner0132aff2005-05-06 22:57:40 +00002001 convention</a> the call should use. If none is specified, the call defaults
2002 to using C calling conventions.
2003 </li>
2004 <li>'<tt>ptr to function ty</tt>': shall be the signature of the pointer to
2005 function value being invoked. In most cases, this is a direct function
2006 invocation, but indirect <tt>invoke</tt>s are just as possible, branching off
2007 an arbitrary pointer to function value.
2008 </li>
2009
2010 <li>'<tt>function ptr val</tt>': An LLVM value containing a pointer to a
2011 function to be invoked. </li>
2012
2013 <li>'<tt>function args</tt>': argument list whose types match the function
2014 signature argument types. If the function signature indicates the function
2015 accepts a variable number of arguments, the extra arguments can be
2016 specified. </li>
2017
2018 <li>'<tt>normal label</tt>': the label reached when the called function
2019 executes a '<tt><a href="#i_ret">ret</a></tt>' instruction. </li>
2020
2021 <li>'<tt>exception label</tt>': the label reached when a callee returns with
2022 the <a href="#i_unwind"><tt>unwind</tt></a> instruction. </li>
2023
Chris Lattner2f7c9632001-06-06 20:29:01 +00002024</ol>
Chris Lattner0132aff2005-05-06 22:57:40 +00002025
Chris Lattner2f7c9632001-06-06 20:29:01 +00002026<h5>Semantics:</h5>
Chris Lattner0132aff2005-05-06 22:57:40 +00002027
Misha Brukman76307852003-11-08 01:05:38 +00002028<p>This instruction is designed to operate as a standard '<tt><a
Chris Lattner0132aff2005-05-06 22:57:40 +00002029href="#i_call">call</a></tt>' instruction in most regards. The primary
2030difference is that it establishes an association with a label, which is used by
2031the runtime library to unwind the stack.</p>
2032
2033<p>This instruction is used in languages with destructors to ensure that proper
2034cleanup is performed in the case of either a <tt>longjmp</tt> or a thrown
2035exception. Additionally, this is important for implementation of
2036'<tt>catch</tt>' clauses in high-level languages that support them.</p>
2037
Chris Lattner2f7c9632001-06-06 20:29:01 +00002038<h5>Example:</h5>
Chris Lattner0132aff2005-05-06 22:57:40 +00002039<pre>
Nick Lewycky084ab472008-03-16 07:18:12 +00002040 %retval = invoke i32 @Test(i32 15) to label %Continue
Jeff Cohen5819f182007-04-22 01:17:39 +00002041 unwind label %TestCleanup <i>; {i32}:retval set</i>
Nick Lewycky084ab472008-03-16 07:18:12 +00002042 %retval = invoke <a href="#callingconv">coldcc</a> i32 %Testfnptr(i32 15) to label %Continue
Jeff Cohen5819f182007-04-22 01:17:39 +00002043 unwind label %TestCleanup <i>; {i32}:retval set</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00002044</pre>
Misha Brukman76307852003-11-08 01:05:38 +00002045</div>
Chris Lattner08b7d5b2004-10-16 18:04:13 +00002046
2047
Chris Lattner5ed60612003-09-03 00:41:47 +00002048<!-- _______________________________________________________________________ -->
Chris Lattner08b7d5b2004-10-16 18:04:13 +00002049
Chris Lattner48b383b02003-11-25 01:02:51 +00002050<div class="doc_subsubsection"> <a name="i_unwind">'<tt>unwind</tt>'
2051Instruction</a> </div>
Chris Lattner08b7d5b2004-10-16 18:04:13 +00002052
Misha Brukman76307852003-11-08 01:05:38 +00002053<div class="doc_text">
Chris Lattner08b7d5b2004-10-16 18:04:13 +00002054
Chris Lattner5ed60612003-09-03 00:41:47 +00002055<h5>Syntax:</h5>
Chris Lattner08b7d5b2004-10-16 18:04:13 +00002056<pre>
2057 unwind
2058</pre>
2059
Chris Lattner5ed60612003-09-03 00:41:47 +00002060<h5>Overview:</h5>
Chris Lattner08b7d5b2004-10-16 18:04:13 +00002061
2062<p>The '<tt>unwind</tt>' instruction unwinds the stack, continuing control flow
2063at the first callee in the dynamic call stack which used an <a
2064href="#i_invoke"><tt>invoke</tt></a> instruction to perform the call. This is
2065primarily used to implement exception handling.</p>
2066
Chris Lattner5ed60612003-09-03 00:41:47 +00002067<h5>Semantics:</h5>
Chris Lattner08b7d5b2004-10-16 18:04:13 +00002068
Chris Lattnerfe8519c2008-04-19 21:01:16 +00002069<p>The '<tt>unwind</tt>' instruction causes execution of the current function to
Chris Lattner08b7d5b2004-10-16 18:04:13 +00002070immediately halt. The dynamic call stack is then searched for the first <a
2071href="#i_invoke"><tt>invoke</tt></a> instruction on the call stack. Once found,
2072execution continues at the "exceptional" destination block specified by the
2073<tt>invoke</tt> instruction. If there is no <tt>invoke</tt> instruction in the
2074dynamic call chain, undefined behavior results.</p>
Misha Brukman76307852003-11-08 01:05:38 +00002075</div>
Chris Lattner08b7d5b2004-10-16 18:04:13 +00002076
2077<!-- _______________________________________________________________________ -->
2078
2079<div class="doc_subsubsection"> <a name="i_unreachable">'<tt>unreachable</tt>'
2080Instruction</a> </div>
2081
2082<div class="doc_text">
2083
2084<h5>Syntax:</h5>
2085<pre>
2086 unreachable
2087</pre>
2088
2089<h5>Overview:</h5>
2090
2091<p>The '<tt>unreachable</tt>' instruction has no defined semantics. This
2092instruction is used to inform the optimizer that a particular portion of the
2093code is not reachable. This can be used to indicate that the code after a
2094no-return function cannot be reached, and other facts.</p>
2095
2096<h5>Semantics:</h5>
2097
2098<p>The '<tt>unreachable</tt>' instruction has no defined semantics.</p>
2099</div>
2100
2101
2102
Chris Lattner2f7c9632001-06-06 20:29:01 +00002103<!-- ======================================================================= -->
Chris Lattner48b383b02003-11-25 01:02:51 +00002104<div class="doc_subsection"> <a name="binaryops">Binary Operations</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +00002105<div class="doc_text">
Chris Lattner48b383b02003-11-25 01:02:51 +00002106<p>Binary operators are used to do most of the computation in a
Chris Lattner81f92972008-04-01 18:47:32 +00002107program. They require two operands of the same type, execute an operation on them, and
John Criswelldfe6a862004-12-10 15:51:16 +00002108produce a single value. The operands might represent
Reid Spencer404a3252007-02-15 03:07:05 +00002109multiple data, as is the case with the <a href="#t_vector">vector</a> data type.
Chris Lattner81f92972008-04-01 18:47:32 +00002110The result value has the same type as its operands.</p>
Misha Brukman76307852003-11-08 01:05:38 +00002111<p>There are several different binary operators:</p>
Misha Brukman76307852003-11-08 01:05:38 +00002112</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +00002113<!-- _______________________________________________________________________ -->
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002114<div class="doc_subsubsection">
2115 <a name="i_add">'<tt>add</tt>' Instruction</a>
2116</div>
2117
Misha Brukman76307852003-11-08 01:05:38 +00002118<div class="doc_text">
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002119
Chris Lattner2f7c9632001-06-06 20:29:01 +00002120<h5>Syntax:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002121
2122<pre>
2123 &lt;result&gt; = add &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00002124</pre>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002125
Chris Lattner2f7c9632001-06-06 20:29:01 +00002126<h5>Overview:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002127
Misha Brukman76307852003-11-08 01:05:38 +00002128<p>The '<tt>add</tt>' instruction returns the sum of its two operands.</p>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002129
Chris Lattner2f7c9632001-06-06 20:29:01 +00002130<h5>Arguments:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002131
2132<p>The two arguments to the '<tt>add</tt>' instruction must be <a
2133 href="#t_integer">integer</a>, <a href="#t_floating">floating point</a>, or
2134 <a href="#t_vector">vector</a> values. Both arguments must have identical
2135 types.</p>
2136
Chris Lattner2f7c9632001-06-06 20:29:01 +00002137<h5>Semantics:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002138
Misha Brukman76307852003-11-08 01:05:38 +00002139<p>The value produced is the integer or floating point sum of the two
2140operands.</p>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002141
Chris Lattner2f2427e2008-01-28 00:36:27 +00002142<p>If an integer sum has unsigned overflow, the result returned is the
2143mathematical result modulo 2<sup>n</sup>, where n is the bit width of
2144the result.</p>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002145
Chris Lattner2f2427e2008-01-28 00:36:27 +00002146<p>Because LLVM integers use a two's complement representation, this
2147instruction is appropriate for both signed and unsigned integers.</p>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002148
Chris Lattner2f7c9632001-06-06 20:29:01 +00002149<h5>Example:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002150
2151<pre>
2152 &lt;result&gt; = add i32 4, %var <i>; yields {i32}:result = 4 + %var</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00002153</pre>
Misha Brukman76307852003-11-08 01:05:38 +00002154</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +00002155<!-- _______________________________________________________________________ -->
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002156<div class="doc_subsubsection">
2157 <a name="i_sub">'<tt>sub</tt>' Instruction</a>
2158</div>
2159
Misha Brukman76307852003-11-08 01:05:38 +00002160<div class="doc_text">
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002161
Chris Lattner2f7c9632001-06-06 20:29:01 +00002162<h5>Syntax:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002163
2164<pre>
2165 &lt;result&gt; = sub &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00002166</pre>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002167
Chris Lattner2f7c9632001-06-06 20:29:01 +00002168<h5>Overview:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002169
Misha Brukman76307852003-11-08 01:05:38 +00002170<p>The '<tt>sub</tt>' instruction returns the difference of its two
2171operands.</p>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002172
2173<p>Note that the '<tt>sub</tt>' instruction is used to represent the
2174'<tt>neg</tt>' instruction present in most other intermediate
2175representations.</p>
2176
Chris Lattner2f7c9632001-06-06 20:29:01 +00002177<h5>Arguments:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002178
2179<p>The two arguments to the '<tt>sub</tt>' instruction must be <a
2180 href="#t_integer">integer</a>, <a href="#t_floating">floating point</a>,
2181 or <a href="#t_vector">vector</a> values. Both arguments must have identical
2182 types.</p>
2183
Chris Lattner2f7c9632001-06-06 20:29:01 +00002184<h5>Semantics:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002185
Chris Lattner48b383b02003-11-25 01:02:51 +00002186<p>The value produced is the integer or floating point difference of
2187the two operands.</p>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002188
Chris Lattner2f2427e2008-01-28 00:36:27 +00002189<p>If an integer difference has unsigned overflow, the result returned is the
2190mathematical result modulo 2<sup>n</sup>, where n is the bit width of
2191the result.</p>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002192
Chris Lattner2f2427e2008-01-28 00:36:27 +00002193<p>Because LLVM integers use a two's complement representation, this
2194instruction is appropriate for both signed and unsigned integers.</p>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002195
Chris Lattner2f7c9632001-06-06 20:29:01 +00002196<h5>Example:</h5>
Bill Wendling2d8b9a82007-05-29 09:42:13 +00002197<pre>
2198 &lt;result&gt; = sub i32 4, %var <i>; yields {i32}:result = 4 - %var</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00002199 &lt;result&gt; = sub i32 0, %val <i>; yields {i32}:result = -%var</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00002200</pre>
Misha Brukman76307852003-11-08 01:05:38 +00002201</div>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002202
Chris Lattner2f7c9632001-06-06 20:29:01 +00002203<!-- _______________________________________________________________________ -->
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002204<div class="doc_subsubsection">
2205 <a name="i_mul">'<tt>mul</tt>' Instruction</a>
2206</div>
2207
Misha Brukman76307852003-11-08 01:05:38 +00002208<div class="doc_text">
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002209
Chris Lattner2f7c9632001-06-06 20:29:01 +00002210<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00002211<pre> &lt;result&gt; = mul &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00002212</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +00002213<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00002214<p>The '<tt>mul</tt>' instruction returns the product of its two
2215operands.</p>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002216
Chris Lattner2f7c9632001-06-06 20:29:01 +00002217<h5>Arguments:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002218
2219<p>The two arguments to the '<tt>mul</tt>' instruction must be <a
2220href="#t_integer">integer</a>, <a href="#t_floating">floating point</a>,
2221or <a href="#t_vector">vector</a> values. Both arguments must have identical
2222types.</p>
2223
Chris Lattner2f7c9632001-06-06 20:29:01 +00002224<h5>Semantics:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002225
Chris Lattner48b383b02003-11-25 01:02:51 +00002226<p>The value produced is the integer or floating point product of the
Misha Brukman76307852003-11-08 01:05:38 +00002227two operands.</p>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002228
Chris Lattner2f2427e2008-01-28 00:36:27 +00002229<p>If the result of an integer multiplication has unsigned overflow,
2230the result returned is the mathematical result modulo
22312<sup>n</sup>, where n is the bit width of the result.</p>
2232<p>Because LLVM integers use a two's complement representation, and the
2233result is the same width as the operands, this instruction returns the
2234correct result for both signed and unsigned integers. If a full product
2235(e.g. <tt>i32</tt>x<tt>i32</tt>-><tt>i64</tt>) is needed, the operands
2236should be sign-extended or zero-extended as appropriate to the
2237width of the full product.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00002238<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00002239<pre> &lt;result&gt; = mul i32 4, %var <i>; yields {i32}:result = 4 * %var</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00002240</pre>
Misha Brukman76307852003-11-08 01:05:38 +00002241</div>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002242
Chris Lattner2f7c9632001-06-06 20:29:01 +00002243<!-- _______________________________________________________________________ -->
Reid Spencer7e80b0b2006-10-26 06:15:43 +00002244<div class="doc_subsubsection"> <a name="i_udiv">'<tt>udiv</tt>' Instruction
2245</a></div>
2246<div class="doc_text">
2247<h5>Syntax:</h5>
2248<pre> &lt;result&gt; = udiv &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2249</pre>
2250<h5>Overview:</h5>
2251<p>The '<tt>udiv</tt>' instruction returns the quotient of its two
2252operands.</p>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002253
Reid Spencer7e80b0b2006-10-26 06:15:43 +00002254<h5>Arguments:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002255
Reid Spencer7e80b0b2006-10-26 06:15:43 +00002256<p>The two arguments to the '<tt>udiv</tt>' instruction must be
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002257<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2258values. Both arguments must have identical types.</p>
2259
Reid Spencer7e80b0b2006-10-26 06:15:43 +00002260<h5>Semantics:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002261
Chris Lattner2f2427e2008-01-28 00:36:27 +00002262<p>The value produced is the unsigned integer quotient of the two operands.</p>
2263<p>Note that unsigned integer division and signed integer division are distinct
2264operations; for signed integer division, use '<tt>sdiv</tt>'.</p>
2265<p>Division by zero leads to undefined behavior.</p>
Reid Spencer7e80b0b2006-10-26 06:15:43 +00002266<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00002267<pre> &lt;result&gt; = udiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
Reid Spencer7e80b0b2006-10-26 06:15:43 +00002268</pre>
2269</div>
2270<!-- _______________________________________________________________________ -->
2271<div class="doc_subsubsection"> <a name="i_sdiv">'<tt>sdiv</tt>' Instruction
2272</a> </div>
2273<div class="doc_text">
2274<h5>Syntax:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002275<pre>
2276 &lt;result&gt; = sdiv &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Reid Spencer7e80b0b2006-10-26 06:15:43 +00002277</pre>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002278
Reid Spencer7e80b0b2006-10-26 06:15:43 +00002279<h5>Overview:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002280
Reid Spencer7e80b0b2006-10-26 06:15:43 +00002281<p>The '<tt>sdiv</tt>' instruction returns the quotient of its two
2282operands.</p>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002283
Reid Spencer7e80b0b2006-10-26 06:15:43 +00002284<h5>Arguments:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002285
2286<p>The two arguments to the '<tt>sdiv</tt>' instruction must be
2287<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2288values. Both arguments must have identical types.</p>
2289
Reid Spencer7e80b0b2006-10-26 06:15:43 +00002290<h5>Semantics:</h5>
Chris Lattner1429e6f2008-04-01 18:45:27 +00002291<p>The value produced is the signed integer quotient of the two operands rounded towards zero.</p>
Chris Lattner2f2427e2008-01-28 00:36:27 +00002292<p>Note that signed integer division and unsigned integer division are distinct
2293operations; for unsigned integer division, use '<tt>udiv</tt>'.</p>
2294<p>Division by zero leads to undefined behavior. Overflow also leads to
2295undefined behavior; this is a rare case, but can occur, for example,
2296by doing a 32-bit division of -2147483648 by -1.</p>
Reid Spencer7e80b0b2006-10-26 06:15:43 +00002297<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00002298<pre> &lt;result&gt; = sdiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
Reid Spencer7e80b0b2006-10-26 06:15:43 +00002299</pre>
2300</div>
2301<!-- _______________________________________________________________________ -->
2302<div class="doc_subsubsection"> <a name="i_fdiv">'<tt>fdiv</tt>'
Chris Lattner48b383b02003-11-25 01:02:51 +00002303Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +00002304<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +00002305<h5>Syntax:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002306<pre>
2307 &lt;result&gt; = fdiv &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner48b383b02003-11-25 01:02:51 +00002308</pre>
2309<h5>Overview:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002310
Reid Spencer7e80b0b2006-10-26 06:15:43 +00002311<p>The '<tt>fdiv</tt>' instruction returns the quotient of its two
Chris Lattner48b383b02003-11-25 01:02:51 +00002312operands.</p>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002313
Chris Lattner48b383b02003-11-25 01:02:51 +00002314<h5>Arguments:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002315
Jeff Cohen5819f182007-04-22 01:17:39 +00002316<p>The two arguments to the '<tt>fdiv</tt>' instruction must be
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002317<a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a>
2318of floating point values. Both arguments must have identical types.</p>
2319
Chris Lattner48b383b02003-11-25 01:02:51 +00002320<h5>Semantics:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002321
Reid Spencer7e80b0b2006-10-26 06:15:43 +00002322<p>The value produced is the floating point quotient of the two operands.</p>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002323
Chris Lattner48b383b02003-11-25 01:02:51 +00002324<h5>Example:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002325
2326<pre>
2327 &lt;result&gt; = fdiv float 4.0, %var <i>; yields {float}:result = 4.0 / %var</i>
Chris Lattner48b383b02003-11-25 01:02:51 +00002328</pre>
2329</div>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002330
Chris Lattner48b383b02003-11-25 01:02:51 +00002331<!-- _______________________________________________________________________ -->
Reid Spencer7eb55b32006-11-02 01:53:59 +00002332<div class="doc_subsubsection"> <a name="i_urem">'<tt>urem</tt>' Instruction</a>
2333</div>
2334<div class="doc_text">
2335<h5>Syntax:</h5>
2336<pre> &lt;result&gt; = urem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2337</pre>
2338<h5>Overview:</h5>
2339<p>The '<tt>urem</tt>' instruction returns the remainder from the
2340unsigned division of its two arguments.</p>
2341<h5>Arguments:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002342<p>The two arguments to the '<tt>urem</tt>' instruction must be
2343<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2344values. Both arguments must have identical types.</p>
Reid Spencer7eb55b32006-11-02 01:53:59 +00002345<h5>Semantics:</h5>
2346<p>This instruction returns the unsigned integer <i>remainder</i> of a division.
Chris Lattner1429e6f2008-04-01 18:45:27 +00002347This instruction always performs an unsigned division to get the remainder.</p>
Chris Lattner2f2427e2008-01-28 00:36:27 +00002348<p>Note that unsigned integer remainder and signed integer remainder are
2349distinct operations; for signed integer remainder, use '<tt>srem</tt>'.</p>
2350<p>Taking the remainder of a division by zero leads to undefined behavior.</p>
Reid Spencer7eb55b32006-11-02 01:53:59 +00002351<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00002352<pre> &lt;result&gt; = urem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
Reid Spencer7eb55b32006-11-02 01:53:59 +00002353</pre>
2354
2355</div>
2356<!-- _______________________________________________________________________ -->
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002357<div class="doc_subsubsection">
2358 <a name="i_srem">'<tt>srem</tt>' Instruction</a>
2359</div>
2360
Chris Lattner48b383b02003-11-25 01:02:51 +00002361<div class="doc_text">
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002362
Chris Lattner48b383b02003-11-25 01:02:51 +00002363<h5>Syntax:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002364
2365<pre>
2366 &lt;result&gt; = srem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner48b383b02003-11-25 01:02:51 +00002367</pre>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002368
Chris Lattner48b383b02003-11-25 01:02:51 +00002369<h5>Overview:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002370
Reid Spencer7eb55b32006-11-02 01:53:59 +00002371<p>The '<tt>srem</tt>' instruction returns the remainder from the
Dan Gohman08143e32007-11-05 23:35:22 +00002372signed division of its two operands. This instruction can also take
2373<a href="#t_vector">vector</a> versions of the values in which case
2374the elements must be integers.</p>
Chris Lattnerb8f816e2008-01-04 04:33:49 +00002375
Chris Lattner48b383b02003-11-25 01:02:51 +00002376<h5>Arguments:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002377
Reid Spencer7eb55b32006-11-02 01:53:59 +00002378<p>The two arguments to the '<tt>srem</tt>' instruction must be
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002379<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2380values. Both arguments must have identical types.</p>
2381
Chris Lattner48b383b02003-11-25 01:02:51 +00002382<h5>Semantics:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002383
Reid Spencer7eb55b32006-11-02 01:53:59 +00002384<p>This instruction returns the <i>remainder</i> of a division (where the result
Reid Spencer806ad6a2007-03-24 22:23:39 +00002385has the same sign as the dividend, <tt>var1</tt>), not the <i>modulo</i>
2386operator (where the result has the same sign as the divisor, <tt>var2</tt>) of
2387a value. For more information about the difference, see <a
Chris Lattner48b383b02003-11-25 01:02:51 +00002388 href="http://mathforum.org/dr.math/problems/anne.4.28.99.html">The
Reid Spencer806ad6a2007-03-24 22:23:39 +00002389Math Forum</a>. For a table of how this is implemented in various languages,
Reid Spencerdb3b93b2007-03-24 22:40:44 +00002390please see <a href="http://en.wikipedia.org/wiki/Modulo_operation">
Reid Spencer806ad6a2007-03-24 22:23:39 +00002391Wikipedia: modulo operation</a>.</p>
Chris Lattner2f2427e2008-01-28 00:36:27 +00002392<p>Note that signed integer remainder and unsigned integer remainder are
2393distinct operations; for unsigned integer remainder, use '<tt>urem</tt>'.</p>
2394<p>Taking the remainder of a division by zero leads to undefined behavior.
2395Overflow also leads to undefined behavior; this is a rare case, but can occur,
2396for example, by taking the remainder of a 32-bit division of -2147483648 by -1.
2397(The remainder doesn't actually overflow, but this rule lets srem be
2398implemented using instructions that return both the result of the division
2399and the remainder.)</p>
Chris Lattner48b383b02003-11-25 01:02:51 +00002400<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00002401<pre> &lt;result&gt; = srem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
Reid Spencer7eb55b32006-11-02 01:53:59 +00002402</pre>
2403
2404</div>
2405<!-- _______________________________________________________________________ -->
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002406<div class="doc_subsubsection">
2407 <a name="i_frem">'<tt>frem</tt>' Instruction</a> </div>
2408
Reid Spencer7eb55b32006-11-02 01:53:59 +00002409<div class="doc_text">
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002410
Reid Spencer7eb55b32006-11-02 01:53:59 +00002411<h5>Syntax:</h5>
2412<pre> &lt;result&gt; = frem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2413</pre>
2414<h5>Overview:</h5>
2415<p>The '<tt>frem</tt>' instruction returns the remainder from the
2416division of its two operands.</p>
2417<h5>Arguments:</h5>
2418<p>The two arguments to the '<tt>frem</tt>' instruction must be
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002419<a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a>
2420of floating point values. Both arguments must have identical types.</p>
2421
Reid Spencer7eb55b32006-11-02 01:53:59 +00002422<h5>Semantics:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002423
Chris Lattner1429e6f2008-04-01 18:45:27 +00002424<p>This instruction returns the <i>remainder</i> of a division.
2425The remainder has the same sign as the dividend.</p>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002426
Reid Spencer7eb55b32006-11-02 01:53:59 +00002427<h5>Example:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002428
2429<pre>
2430 &lt;result&gt; = frem float 4.0, %var <i>; yields {float}:result = 4.0 % %var</i>
Chris Lattner48b383b02003-11-25 01:02:51 +00002431</pre>
Misha Brukman76307852003-11-08 01:05:38 +00002432</div>
Robert Bocchino820bc75b2006-02-17 21:18:08 +00002433
Reid Spencer2ab01932007-02-02 13:57:07 +00002434<!-- ======================================================================= -->
2435<div class="doc_subsection"> <a name="bitwiseops">Bitwise Binary
2436Operations</a> </div>
2437<div class="doc_text">
2438<p>Bitwise binary operators are used to do various forms of
2439bit-twiddling in a program. They are generally very efficient
2440instructions and can commonly be strength reduced from other
Chris Lattner1429e6f2008-04-01 18:45:27 +00002441instructions. They require two operands of the same type, execute an operation on them,
2442and produce a single value. The resulting value is the same type as its operands.</p>
Reid Spencer2ab01932007-02-02 13:57:07 +00002443</div>
2444
Reid Spencer04e259b2007-01-31 21:39:12 +00002445<!-- _______________________________________________________________________ -->
2446<div class="doc_subsubsection"> <a name="i_shl">'<tt>shl</tt>'
2447Instruction</a> </div>
2448<div class="doc_text">
2449<h5>Syntax:</h5>
2450<pre> &lt;result&gt; = shl &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2451</pre>
Chris Lattnerf0e50112007-10-03 21:01:14 +00002452
Reid Spencer04e259b2007-01-31 21:39:12 +00002453<h5>Overview:</h5>
Chris Lattnerf0e50112007-10-03 21:01:14 +00002454
Reid Spencer04e259b2007-01-31 21:39:12 +00002455<p>The '<tt>shl</tt>' instruction returns the first operand shifted to
2456the left a specified number of bits.</p>
Chris Lattnerf0e50112007-10-03 21:01:14 +00002457
Reid Spencer04e259b2007-01-31 21:39:12 +00002458<h5>Arguments:</h5>
Chris Lattnerf0e50112007-10-03 21:01:14 +00002459
Reid Spencer04e259b2007-01-31 21:39:12 +00002460<p>Both arguments to the '<tt>shl</tt>' instruction must be the same <a
Nate Begemanfecbc8c2008-07-29 15:49:41 +00002461 href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2462type. '<tt>var2</tt>' is treated as an unsigned value.</p>
Chris Lattnerf0e50112007-10-03 21:01:14 +00002463
Reid Spencer04e259b2007-01-31 21:39:12 +00002464<h5>Semantics:</h5>
Chris Lattnerf0e50112007-10-03 21:01:14 +00002465
Chris Lattner1429e6f2008-04-01 18:45:27 +00002466<p>The value produced is <tt>var1</tt> * 2<sup><tt>var2</tt></sup> mod 2<sup>n</sup>,
2467where n is the width of the result. If <tt>var2</tt> is (statically or dynamically) negative or
2468equal to or larger than the number of bits in <tt>var1</tt>, the result is undefined.</p>
Chris Lattnerf0e50112007-10-03 21:01:14 +00002469
Reid Spencer04e259b2007-01-31 21:39:12 +00002470<h5>Example:</h5><pre>
2471 &lt;result&gt; = shl i32 4, %var <i>; yields {i32}: 4 &lt;&lt; %var</i>
2472 &lt;result&gt; = shl i32 4, 2 <i>; yields {i32}: 16</i>
2473 &lt;result&gt; = shl i32 1, 10 <i>; yields {i32}: 1024</i>
Chris Lattnerf0e50112007-10-03 21:01:14 +00002474 &lt;result&gt; = shl i32 1, 32 <i>; undefined</i>
Reid Spencer04e259b2007-01-31 21:39:12 +00002475</pre>
2476</div>
2477<!-- _______________________________________________________________________ -->
2478<div class="doc_subsubsection"> <a name="i_lshr">'<tt>lshr</tt>'
2479Instruction</a> </div>
2480<div class="doc_text">
2481<h5>Syntax:</h5>
2482<pre> &lt;result&gt; = lshr &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2483</pre>
2484
2485<h5>Overview:</h5>
2486<p>The '<tt>lshr</tt>' instruction (logical shift right) returns the first
Jeff Cohen5819f182007-04-22 01:17:39 +00002487operand shifted to the right a specified number of bits with zero fill.</p>
Reid Spencer04e259b2007-01-31 21:39:12 +00002488
2489<h5>Arguments:</h5>
2490<p>Both arguments to the '<tt>lshr</tt>' instruction must be the same
Nate Begemanfecbc8c2008-07-29 15:49:41 +00002491<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2492type. '<tt>var2</tt>' is treated as an unsigned value.</p>
Reid Spencer04e259b2007-01-31 21:39:12 +00002493
2494<h5>Semantics:</h5>
Chris Lattnerf0e50112007-10-03 21:01:14 +00002495
Reid Spencer04e259b2007-01-31 21:39:12 +00002496<p>This instruction always performs a logical shift right operation. The most
2497significant bits of the result will be filled with zero bits after the
Chris Lattnerf0e50112007-10-03 21:01:14 +00002498shift. If <tt>var2</tt> is (statically or dynamically) equal to or larger than
2499the number of bits in <tt>var1</tt>, the result is undefined.</p>
Reid Spencer04e259b2007-01-31 21:39:12 +00002500
2501<h5>Example:</h5>
2502<pre>
2503 &lt;result&gt; = lshr i32 4, 1 <i>; yields {i32}:result = 2</i>
2504 &lt;result&gt; = lshr i32 4, 2 <i>; yields {i32}:result = 1</i>
2505 &lt;result&gt; = lshr i8 4, 3 <i>; yields {i8}:result = 0</i>
2506 &lt;result&gt; = lshr i8 -2, 1 <i>; yields {i8}:result = 0x7FFFFFFF </i>
Chris Lattnerf0e50112007-10-03 21:01:14 +00002507 &lt;result&gt; = lshr i32 1, 32 <i>; undefined</i>
Reid Spencer04e259b2007-01-31 21:39:12 +00002508</pre>
2509</div>
2510
Reid Spencer2ab01932007-02-02 13:57:07 +00002511<!-- _______________________________________________________________________ -->
Reid Spencer04e259b2007-01-31 21:39:12 +00002512<div class="doc_subsubsection"> <a name="i_ashr">'<tt>ashr</tt>'
2513Instruction</a> </div>
2514<div class="doc_text">
2515
2516<h5>Syntax:</h5>
2517<pre> &lt;result&gt; = ashr &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2518</pre>
2519
2520<h5>Overview:</h5>
2521<p>The '<tt>ashr</tt>' instruction (arithmetic shift right) returns the first
Jeff Cohen5819f182007-04-22 01:17:39 +00002522operand shifted to the right a specified number of bits with sign extension.</p>
Reid Spencer04e259b2007-01-31 21:39:12 +00002523
2524<h5>Arguments:</h5>
2525<p>Both arguments to the '<tt>ashr</tt>' instruction must be the same
Nate Begemanfecbc8c2008-07-29 15:49:41 +00002526<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2527type. '<tt>var2</tt>' is treated as an unsigned value.</p>
Reid Spencer04e259b2007-01-31 21:39:12 +00002528
2529<h5>Semantics:</h5>
2530<p>This instruction always performs an arithmetic shift right operation,
2531The most significant bits of the result will be filled with the sign bit
Chris Lattnerf0e50112007-10-03 21:01:14 +00002532of <tt>var1</tt>. If <tt>var2</tt> is (statically or dynamically) equal to or
2533larger than the number of bits in <tt>var1</tt>, the result is undefined.
2534</p>
Reid Spencer04e259b2007-01-31 21:39:12 +00002535
2536<h5>Example:</h5>
2537<pre>
2538 &lt;result&gt; = ashr i32 4, 1 <i>; yields {i32}:result = 2</i>
2539 &lt;result&gt; = ashr i32 4, 2 <i>; yields {i32}:result = 1</i>
2540 &lt;result&gt; = ashr i8 4, 3 <i>; yields {i8}:result = 0</i>
2541 &lt;result&gt; = ashr i8 -2, 1 <i>; yields {i8}:result = -1</i>
Chris Lattnerf0e50112007-10-03 21:01:14 +00002542 &lt;result&gt; = ashr i32 1, 32 <i>; undefined</i>
Reid Spencer04e259b2007-01-31 21:39:12 +00002543</pre>
2544</div>
2545
Chris Lattner2f7c9632001-06-06 20:29:01 +00002546<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +00002547<div class="doc_subsubsection"> <a name="i_and">'<tt>and</tt>'
2548Instruction</a> </div>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002549
Misha Brukman76307852003-11-08 01:05:38 +00002550<div class="doc_text">
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002551
Chris Lattner2f7c9632001-06-06 20:29:01 +00002552<h5>Syntax:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002553
2554<pre>
2555 &lt;result&gt; = and &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00002556</pre>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002557
Chris Lattner2f7c9632001-06-06 20:29:01 +00002558<h5>Overview:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002559
Chris Lattner48b383b02003-11-25 01:02:51 +00002560<p>The '<tt>and</tt>' instruction returns the bitwise logical and of
2561its two operands.</p>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002562
Chris Lattner2f7c9632001-06-06 20:29:01 +00002563<h5>Arguments:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002564
2565<p>The two arguments to the '<tt>and</tt>' instruction must be
2566<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2567values. Both arguments must have identical types.</p>
2568
Chris Lattner2f7c9632001-06-06 20:29:01 +00002569<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +00002570<p>The truth table used for the '<tt>and</tt>' instruction is:</p>
Chris Lattner48b383b02003-11-25 01:02:51 +00002571<p> </p>
Misha Brukmanc501f552004-03-01 17:47:27 +00002572<div style="align: center">
Misha Brukman76307852003-11-08 01:05:38 +00002573<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner48b383b02003-11-25 01:02:51 +00002574 <tbody>
2575 <tr>
2576 <td>In0</td>
2577 <td>In1</td>
2578 <td>Out</td>
2579 </tr>
2580 <tr>
2581 <td>0</td>
2582 <td>0</td>
2583 <td>0</td>
2584 </tr>
2585 <tr>
2586 <td>0</td>
2587 <td>1</td>
2588 <td>0</td>
2589 </tr>
2590 <tr>
2591 <td>1</td>
2592 <td>0</td>
2593 <td>0</td>
2594 </tr>
2595 <tr>
2596 <td>1</td>
2597 <td>1</td>
2598 <td>1</td>
2599 </tr>
2600 </tbody>
2601</table>
Misha Brukmanc501f552004-03-01 17:47:27 +00002602</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +00002603<h5>Example:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002604<pre>
2605 &lt;result&gt; = and i32 4, %var <i>; yields {i32}:result = 4 &amp; %var</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00002606 &lt;result&gt; = and i32 15, 40 <i>; yields {i32}:result = 8</i>
2607 &lt;result&gt; = and i32 4, 8 <i>; yields {i32}:result = 0</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00002608</pre>
Misha Brukman76307852003-11-08 01:05:38 +00002609</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +00002610<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +00002611<div class="doc_subsubsection"> <a name="i_or">'<tt>or</tt>' Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +00002612<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +00002613<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00002614<pre> &lt;result&gt; = or &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00002615</pre>
Chris Lattner48b383b02003-11-25 01:02:51 +00002616<h5>Overview:</h5>
2617<p>The '<tt>or</tt>' instruction returns the bitwise logical inclusive
2618or of its two operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00002619<h5>Arguments:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002620
2621<p>The two arguments to the '<tt>or</tt>' instruction must be
2622<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2623values. Both arguments must have identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00002624<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +00002625<p>The truth table used for the '<tt>or</tt>' instruction is:</p>
Chris Lattner48b383b02003-11-25 01:02:51 +00002626<p> </p>
Misha Brukmanc501f552004-03-01 17:47:27 +00002627<div style="align: center">
Chris Lattner48b383b02003-11-25 01:02:51 +00002628<table border="1" cellspacing="0" cellpadding="4">
2629 <tbody>
2630 <tr>
2631 <td>In0</td>
2632 <td>In1</td>
2633 <td>Out</td>
2634 </tr>
2635 <tr>
2636 <td>0</td>
2637 <td>0</td>
2638 <td>0</td>
2639 </tr>
2640 <tr>
2641 <td>0</td>
2642 <td>1</td>
2643 <td>1</td>
2644 </tr>
2645 <tr>
2646 <td>1</td>
2647 <td>0</td>
2648 <td>1</td>
2649 </tr>
2650 <tr>
2651 <td>1</td>
2652 <td>1</td>
2653 <td>1</td>
2654 </tr>
2655 </tbody>
2656</table>
Misha Brukmanc501f552004-03-01 17:47:27 +00002657</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +00002658<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00002659<pre> &lt;result&gt; = or i32 4, %var <i>; yields {i32}:result = 4 | %var</i>
2660 &lt;result&gt; = or i32 15, 40 <i>; yields {i32}:result = 47</i>
2661 &lt;result&gt; = or i32 4, 8 <i>; yields {i32}:result = 12</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00002662</pre>
Misha Brukman76307852003-11-08 01:05:38 +00002663</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +00002664<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +00002665<div class="doc_subsubsection"> <a name="i_xor">'<tt>xor</tt>'
2666Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +00002667<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +00002668<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00002669<pre> &lt;result&gt; = xor &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00002670</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +00002671<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00002672<p>The '<tt>xor</tt>' instruction returns the bitwise logical exclusive
2673or of its two operands. The <tt>xor</tt> is used to implement the
2674"one's complement" operation, which is the "~" operator in C.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00002675<h5>Arguments:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002676<p>The two arguments to the '<tt>xor</tt>' instruction must be
2677<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2678values. Both arguments must have identical types.</p>
2679
Chris Lattner2f7c9632001-06-06 20:29:01 +00002680<h5>Semantics:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00002681
Misha Brukman76307852003-11-08 01:05:38 +00002682<p>The truth table used for the '<tt>xor</tt>' instruction is:</p>
Chris Lattner48b383b02003-11-25 01:02:51 +00002683<p> </p>
Misha Brukmanc501f552004-03-01 17:47:27 +00002684<div style="align: center">
Chris Lattner48b383b02003-11-25 01:02:51 +00002685<table border="1" cellspacing="0" cellpadding="4">
2686 <tbody>
2687 <tr>
2688 <td>In0</td>
2689 <td>In1</td>
2690 <td>Out</td>
2691 </tr>
2692 <tr>
2693 <td>0</td>
2694 <td>0</td>
2695 <td>0</td>
2696 </tr>
2697 <tr>
2698 <td>0</td>
2699 <td>1</td>
2700 <td>1</td>
2701 </tr>
2702 <tr>
2703 <td>1</td>
2704 <td>0</td>
2705 <td>1</td>
2706 </tr>
2707 <tr>
2708 <td>1</td>
2709 <td>1</td>
2710 <td>0</td>
2711 </tr>
2712 </tbody>
2713</table>
Misha Brukmanc501f552004-03-01 17:47:27 +00002714</div>
Chris Lattner48b383b02003-11-25 01:02:51 +00002715<p> </p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00002716<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00002717<pre> &lt;result&gt; = xor i32 4, %var <i>; yields {i32}:result = 4 ^ %var</i>
2718 &lt;result&gt; = xor i32 15, 40 <i>; yields {i32}:result = 39</i>
2719 &lt;result&gt; = xor i32 4, 8 <i>; yields {i32}:result = 12</i>
2720 &lt;result&gt; = xor i32 %V, -1 <i>; yields {i32}:result = ~%V</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00002721</pre>
Misha Brukman76307852003-11-08 01:05:38 +00002722</div>
Chris Lattner54611b42005-11-06 08:02:57 +00002723
Chris Lattner2f7c9632001-06-06 20:29:01 +00002724<!-- ======================================================================= -->
Chris Lattner54611b42005-11-06 08:02:57 +00002725<div class="doc_subsection">
Chris Lattnerce83bff2006-04-08 23:07:04 +00002726 <a name="vectorops">Vector Operations</a>
2727</div>
2728
2729<div class="doc_text">
2730
2731<p>LLVM supports several instructions to represent vector operations in a
Jeff Cohen5819f182007-04-22 01:17:39 +00002732target-independent manner. These instructions cover the element-access and
Chris Lattnerce83bff2006-04-08 23:07:04 +00002733vector-specific operations needed to process vectors effectively. While LLVM
2734does directly support these vector operations, many sophisticated algorithms
2735will want to use target-specific intrinsics to take full advantage of a specific
2736target.</p>
2737
2738</div>
2739
2740<!-- _______________________________________________________________________ -->
2741<div class="doc_subsubsection">
2742 <a name="i_extractelement">'<tt>extractelement</tt>' Instruction</a>
2743</div>
2744
2745<div class="doc_text">
2746
2747<h5>Syntax:</h5>
2748
2749<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00002750 &lt;result&gt; = extractelement &lt;n x &lt;ty&gt;&gt; &lt;val&gt;, i32 &lt;idx&gt; <i>; yields &lt;ty&gt;</i>
Chris Lattnerce83bff2006-04-08 23:07:04 +00002751</pre>
2752
2753<h5>Overview:</h5>
2754
2755<p>
2756The '<tt>extractelement</tt>' instruction extracts a single scalar
Reid Spencer404a3252007-02-15 03:07:05 +00002757element from a vector at a specified index.
Chris Lattnerce83bff2006-04-08 23:07:04 +00002758</p>
2759
2760
2761<h5>Arguments:</h5>
2762
2763<p>
2764The first operand of an '<tt>extractelement</tt>' instruction is a
Reid Spencer404a3252007-02-15 03:07:05 +00002765value of <a href="#t_vector">vector</a> type. The second operand is
Chris Lattnerce83bff2006-04-08 23:07:04 +00002766an index indicating the position from which to extract the element.
2767The index may be a variable.</p>
2768
2769<h5>Semantics:</h5>
2770
2771<p>
2772The result is a scalar of the same type as the element type of
2773<tt>val</tt>. Its value is the value at position <tt>idx</tt> of
2774<tt>val</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
2775results are undefined.
2776</p>
2777
2778<h5>Example:</h5>
2779
2780<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00002781 %result = extractelement &lt;4 x i32&gt; %vec, i32 0 <i>; yields i32</i>
Chris Lattnerce83bff2006-04-08 23:07:04 +00002782</pre>
2783</div>
2784
2785
2786<!-- _______________________________________________________________________ -->
2787<div class="doc_subsubsection">
2788 <a name="i_insertelement">'<tt>insertelement</tt>' Instruction</a>
2789</div>
2790
2791<div class="doc_text">
2792
2793<h5>Syntax:</h5>
2794
2795<pre>
Dan Gohman43ba0672008-05-12 23:38:42 +00002796 &lt;result&gt; = insertelement &lt;n x &lt;ty&gt;&gt; &lt;val&gt;, &lt;ty&gt; &lt;elt&gt;, i32 &lt;idx&gt; <i>; yields &lt;n x &lt;ty&gt;&gt;</i>
Chris Lattnerce83bff2006-04-08 23:07:04 +00002797</pre>
2798
2799<h5>Overview:</h5>
2800
2801<p>
2802The '<tt>insertelement</tt>' instruction inserts a scalar
Reid Spencer404a3252007-02-15 03:07:05 +00002803element into a vector at a specified index.
Chris Lattnerce83bff2006-04-08 23:07:04 +00002804</p>
2805
2806
2807<h5>Arguments:</h5>
2808
2809<p>
2810The first operand of an '<tt>insertelement</tt>' instruction is a
Reid Spencer404a3252007-02-15 03:07:05 +00002811value of <a href="#t_vector">vector</a> type. The second operand is a
Chris Lattnerce83bff2006-04-08 23:07:04 +00002812scalar value whose type must equal the element type of the first
2813operand. The third operand is an index indicating the position at
2814which to insert the value. The index may be a variable.</p>
2815
2816<h5>Semantics:</h5>
2817
2818<p>
Reid Spencer404a3252007-02-15 03:07:05 +00002819The result is a vector of the same type as <tt>val</tt>. Its
Chris Lattnerce83bff2006-04-08 23:07:04 +00002820element values are those of <tt>val</tt> except at position
2821<tt>idx</tt>, where it gets the value <tt>elt</tt>. If <tt>idx</tt>
2822exceeds the length of <tt>val</tt>, the results are undefined.
2823</p>
2824
2825<h5>Example:</h5>
2826
2827<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00002828 %result = insertelement &lt;4 x i32&gt; %vec, i32 1, i32 0 <i>; yields &lt;4 x i32&gt;</i>
Chris Lattnerce83bff2006-04-08 23:07:04 +00002829</pre>
2830</div>
2831
2832<!-- _______________________________________________________________________ -->
2833<div class="doc_subsubsection">
2834 <a name="i_shufflevector">'<tt>shufflevector</tt>' Instruction</a>
2835</div>
2836
2837<div class="doc_text">
2838
2839<h5>Syntax:</h5>
2840
2841<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00002842 &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;n x i32&gt; &lt;mask&gt; <i>; yields &lt;n x &lt;ty&gt;&gt;</i>
Chris Lattnerce83bff2006-04-08 23:07:04 +00002843</pre>
2844
2845<h5>Overview:</h5>
2846
2847<p>
2848The '<tt>shufflevector</tt>' instruction constructs a permutation of elements
2849from two input vectors, returning a vector of the same type.
2850</p>
2851
2852<h5>Arguments:</h5>
2853
2854<p>
2855The first two operands of a '<tt>shufflevector</tt>' instruction are vectors
2856with types that match each other and types that match the result of the
2857instruction. The third argument is a shuffle mask, which has the same number
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00002858of elements as the other vector type, but whose element type is always 'i32'.
Chris Lattnerce83bff2006-04-08 23:07:04 +00002859</p>
2860
2861<p>
2862The shuffle mask operand is required to be a constant vector with either
2863constant integer or undef values.
2864</p>
2865
2866<h5>Semantics:</h5>
2867
2868<p>
2869The elements of the two input vectors are numbered from left to right across
2870both of the vectors. The shuffle mask operand specifies, for each element of
2871the result vector, which element of the two input registers the result element
2872gets. The element selector may be undef (meaning "don't care") and the second
2873operand may be undef if performing a shuffle from only one vector.
2874</p>
2875
2876<h5>Example:</h5>
2877
2878<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00002879 %result = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; %v2,
Jeff Cohen5819f182007-04-22 01:17:39 +00002880 &lt;4 x i32&gt; &lt;i32 0, i32 4, i32 1, i32 5&gt; <i>; yields &lt;4 x i32&gt;</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00002881 %result = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; undef,
2882 &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.
Chris Lattnerce83bff2006-04-08 23:07:04 +00002883</pre>
2884</div>
2885
Tanya Lattnerb138bbe2006-04-14 19:24:33 +00002886
Chris Lattnerce83bff2006-04-08 23:07:04 +00002887<!-- ======================================================================= -->
2888<div class="doc_subsection">
Dan Gohmanb9d66602008-05-12 23:51:09 +00002889 <a name="aggregateops">Aggregate Operations</a>
2890</div>
2891
2892<div class="doc_text">
2893
2894<p>LLVM supports several instructions for working with aggregate values.
2895</p>
2896
2897</div>
2898
2899<!-- _______________________________________________________________________ -->
2900<div class="doc_subsubsection">
2901 <a name="i_extractvalue">'<tt>extractvalue</tt>' Instruction</a>
2902</div>
2903
2904<div class="doc_text">
2905
2906<h5>Syntax:</h5>
2907
2908<pre>
2909 &lt;result&gt; = extractvalue &lt;aggregate type&gt; &lt;val&gt;, &lt;idx&gt;{, &lt;idx&gt;}*
2910</pre>
2911
2912<h5>Overview:</h5>
2913
2914<p>
Dan Gohman35a835c2008-05-13 18:16:06 +00002915The '<tt>extractvalue</tt>' instruction extracts the value of a struct field
2916or array element from an aggregate value.
Dan Gohmanb9d66602008-05-12 23:51:09 +00002917</p>
2918
2919
2920<h5>Arguments:</h5>
2921
2922<p>
2923The first operand of an '<tt>extractvalue</tt>' instruction is a
2924value of <a href="#t_struct">struct</a> or <a href="#t_array">array</a>
Dan Gohman35a835c2008-05-13 18:16:06 +00002925type. The operands are constant indices to specify which value to extract
Dan Gohman1ecaf452008-05-31 00:58:22 +00002926in a similar manner as indices in a
Dan Gohmanb9d66602008-05-12 23:51:09 +00002927'<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.
2928</p>
2929
2930<h5>Semantics:</h5>
2931
2932<p>
2933The result is the value at the position in the aggregate specified by
2934the index operands.
2935</p>
2936
2937<h5>Example:</h5>
2938
2939<pre>
Dan Gohman1ecaf452008-05-31 00:58:22 +00002940 %result = extractvalue {i32, float} %agg, 0 <i>; yields i32</i>
Dan Gohmanb9d66602008-05-12 23:51:09 +00002941</pre>
2942</div>
2943
2944
2945<!-- _______________________________________________________________________ -->
2946<div class="doc_subsubsection">
2947 <a name="i_insertvalue">'<tt>insertvalue</tt>' Instruction</a>
2948</div>
2949
2950<div class="doc_text">
2951
2952<h5>Syntax:</h5>
2953
2954<pre>
Dan Gohman1ecaf452008-05-31 00:58:22 +00002955 &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 Gohmanb9d66602008-05-12 23:51:09 +00002956</pre>
2957
2958<h5>Overview:</h5>
2959
2960<p>
2961The '<tt>insertvalue</tt>' instruction inserts a value
Dan Gohman35a835c2008-05-13 18:16:06 +00002962into a struct field or array element in an aggregate.
Dan Gohmanb9d66602008-05-12 23:51:09 +00002963</p>
2964
2965
2966<h5>Arguments:</h5>
2967
2968<p>
2969The first operand of an '<tt>insertvalue</tt>' instruction is a
2970value of <a href="#t_struct">struct</a> or <a href="#t_array">array</a> type.
2971The second operand is a first-class value to insert.
Dan Gohman34d1c0d2008-05-23 21:53:15 +00002972The following operands are constant indices
Dan Gohman1ecaf452008-05-31 00:58:22 +00002973indicating the position at which to insert the value in a similar manner as
Dan Gohman35a835c2008-05-13 18:16:06 +00002974indices in a
Dan Gohmanb9d66602008-05-12 23:51:09 +00002975'<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.
2976The value to insert must have the same type as the value identified
Dan Gohman35a835c2008-05-13 18:16:06 +00002977by the indices.
Dan Gohmanb9d66602008-05-12 23:51:09 +00002978
2979<h5>Semantics:</h5>
2980
2981<p>
2982The result is an aggregate of the same type as <tt>val</tt>. Its
2983value is that of <tt>val</tt> except that the value at the position
Dan Gohman35a835c2008-05-13 18:16:06 +00002984specified by the indices is that of <tt>elt</tt>.
Dan Gohmanb9d66602008-05-12 23:51:09 +00002985</p>
2986
2987<h5>Example:</h5>
2988
2989<pre>
Dan Gohman88ce1a52008-06-23 15:26:37 +00002990 %result = insertvalue {i32, float} %agg, i32 1, 0 <i>; yields {i32, float}</i>
Dan Gohmanb9d66602008-05-12 23:51:09 +00002991</pre>
2992</div>
2993
2994
2995<!-- ======================================================================= -->
2996<div class="doc_subsection">
Chris Lattner6ab66722006-08-15 00:45:58 +00002997 <a name="memoryops">Memory Access and Addressing Operations</a>
Chris Lattner54611b42005-11-06 08:02:57 +00002998</div>
2999
Misha Brukman76307852003-11-08 01:05:38 +00003000<div class="doc_text">
Chris Lattner54611b42005-11-06 08:02:57 +00003001
Chris Lattner48b383b02003-11-25 01:02:51 +00003002<p>A key design point of an SSA-based representation is how it
3003represents memory. In LLVM, no memory locations are in SSA form, which
3004makes things very simple. This section describes how to read, write,
John Criswelldfe6a862004-12-10 15:51:16 +00003005allocate, and free memory in LLVM.</p>
Chris Lattner54611b42005-11-06 08:02:57 +00003006
Misha Brukman76307852003-11-08 01:05:38 +00003007</div>
Chris Lattner54611b42005-11-06 08:02:57 +00003008
Chris Lattner2f7c9632001-06-06 20:29:01 +00003009<!-- _______________________________________________________________________ -->
Chris Lattner54611b42005-11-06 08:02:57 +00003010<div class="doc_subsubsection">
3011 <a name="i_malloc">'<tt>malloc</tt>' Instruction</a>
3012</div>
3013
Misha Brukman76307852003-11-08 01:05:38 +00003014<div class="doc_text">
Chris Lattner54611b42005-11-06 08:02:57 +00003015
Chris Lattner2f7c9632001-06-06 20:29:01 +00003016<h5>Syntax:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +00003017
3018<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00003019 &lt;result&gt; = malloc &lt;type&gt;[, i32 &lt;NumElements&gt;][, align &lt;alignment&gt;] <i>; yields {type*}:result</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00003020</pre>
Chris Lattner54611b42005-11-06 08:02:57 +00003021
Chris Lattner2f7c9632001-06-06 20:29:01 +00003022<h5>Overview:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +00003023
Chris Lattner48b383b02003-11-25 01:02:51 +00003024<p>The '<tt>malloc</tt>' instruction allocates memory from the system
Christopher Lamb55c6d4f2007-12-17 01:00:21 +00003025heap and returns a pointer to it. The object is always allocated in the generic
3026address space (address space zero).</p>
Chris Lattner54611b42005-11-06 08:02:57 +00003027
Chris Lattner2f7c9632001-06-06 20:29:01 +00003028<h5>Arguments:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +00003029
3030<p>The '<tt>malloc</tt>' instruction allocates
3031<tt>sizeof(&lt;type&gt;)*NumElements</tt>
John Criswella92e5862004-02-24 16:13:56 +00003032bytes of memory from the operating system and returns a pointer of the
Chris Lattner54611b42005-11-06 08:02:57 +00003033appropriate type to the program. If "NumElements" is specified, it is the
Gabor Greifdd1fc982008-02-09 22:24:34 +00003034number of elements allocated, otherwise "NumElements" is defaulted to be one.
Chris Lattner1f17cce2008-04-02 00:38:26 +00003035If a constant alignment is specified, the value result of the allocation is guaranteed to
Gabor Greifdd1fc982008-02-09 22:24:34 +00003036be aligned to at least that boundary. If not specified, or if zero, the target can
3037choose to align the allocation on any convenient boundary.</p>
Chris Lattner54611b42005-11-06 08:02:57 +00003038
Misha Brukman76307852003-11-08 01:05:38 +00003039<p>'<tt>type</tt>' must be a sized type.</p>
Chris Lattner54611b42005-11-06 08:02:57 +00003040
Chris Lattner2f7c9632001-06-06 20:29:01 +00003041<h5>Semantics:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +00003042
Chris Lattner48b383b02003-11-25 01:02:51 +00003043<p>Memory is allocated using the system "<tt>malloc</tt>" function, and
Chris Lattnerfe8519c2008-04-19 21:01:16 +00003044a pointer is returned. The result of a zero byte allocattion is undefined. The
3045result is null if there is insufficient memory available.</p>
Misha Brukman76307852003-11-08 01:05:38 +00003046
Chris Lattner54611b42005-11-06 08:02:57 +00003047<h5>Example:</h5>
3048
3049<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00003050 %array = malloc [4 x i8 ] <i>; yields {[%4 x i8]*}:array</i>
Chris Lattner54611b42005-11-06 08:02:57 +00003051
Bill Wendling2d8b9a82007-05-29 09:42:13 +00003052 %size = <a href="#i_add">add</a> i32 2, 2 <i>; yields {i32}:size = i32 4</i>
3053 %array1 = malloc i8, i32 4 <i>; yields {i8*}:array1</i>
3054 %array2 = malloc [12 x i8], i32 %size <i>; yields {[12 x i8]*}:array2</i>
3055 %array3 = malloc i32, i32 4, align 1024 <i>; yields {i32*}:array3</i>
3056 %array4 = malloc i32, align 1024 <i>; yields {i32*}:array4</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00003057</pre>
Misha Brukman76307852003-11-08 01:05:38 +00003058</div>
Chris Lattner54611b42005-11-06 08:02:57 +00003059
Chris Lattner2f7c9632001-06-06 20:29:01 +00003060<!-- _______________________________________________________________________ -->
Chris Lattner54611b42005-11-06 08:02:57 +00003061<div class="doc_subsubsection">
3062 <a name="i_free">'<tt>free</tt>' Instruction</a>
3063</div>
3064
Misha Brukman76307852003-11-08 01:05:38 +00003065<div class="doc_text">
Chris Lattner54611b42005-11-06 08:02:57 +00003066
Chris Lattner2f7c9632001-06-06 20:29:01 +00003067<h5>Syntax:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +00003068
3069<pre>
3070 free &lt;type&gt; &lt;value&gt; <i>; yields {void}</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00003071</pre>
Chris Lattner54611b42005-11-06 08:02:57 +00003072
Chris Lattner2f7c9632001-06-06 20:29:01 +00003073<h5>Overview:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +00003074
Chris Lattner48b383b02003-11-25 01:02:51 +00003075<p>The '<tt>free</tt>' instruction returns memory back to the unused
John Criswell4a3327e2005-05-13 22:25:59 +00003076memory heap to be reallocated in the future.</p>
Chris Lattner54611b42005-11-06 08:02:57 +00003077
Chris Lattner2f7c9632001-06-06 20:29:01 +00003078<h5>Arguments:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +00003079
Chris Lattner48b383b02003-11-25 01:02:51 +00003080<p>'<tt>value</tt>' shall be a pointer value that points to a value
3081that was allocated with the '<tt><a href="#i_malloc">malloc</a></tt>'
3082instruction.</p>
Chris Lattner54611b42005-11-06 08:02:57 +00003083
Chris Lattner2f7c9632001-06-06 20:29:01 +00003084<h5>Semantics:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +00003085
John Criswelldfe6a862004-12-10 15:51:16 +00003086<p>Access to the memory pointed to by the pointer is no longer defined
Chris Lattner0f103e12008-04-19 22:41:32 +00003087after this instruction executes. If the pointer is null, the operation
3088is a noop.</p>
Chris Lattner54611b42005-11-06 08:02:57 +00003089
Chris Lattner2f7c9632001-06-06 20:29:01 +00003090<h5>Example:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +00003091
3092<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00003093 %array = <a href="#i_malloc">malloc</a> [4 x i8] <i>; yields {[4 x i8]*}:array</i>
3094 free [4 x i8]* %array
Chris Lattner2f7c9632001-06-06 20:29:01 +00003095</pre>
Misha Brukman76307852003-11-08 01:05:38 +00003096</div>
Chris Lattner54611b42005-11-06 08:02:57 +00003097
Chris Lattner2f7c9632001-06-06 20:29:01 +00003098<!-- _______________________________________________________________________ -->
Chris Lattner54611b42005-11-06 08:02:57 +00003099<div class="doc_subsubsection">
3100 <a name="i_alloca">'<tt>alloca</tt>' Instruction</a>
3101</div>
3102
Misha Brukman76307852003-11-08 01:05:38 +00003103<div class="doc_text">
Chris Lattner54611b42005-11-06 08:02:57 +00003104
Chris Lattner2f7c9632001-06-06 20:29:01 +00003105<h5>Syntax:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +00003106
3107<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00003108 &lt;result&gt; = alloca &lt;type&gt;[, i32 &lt;NumElements&gt;][, align &lt;alignment&gt;] <i>; yields {type*}:result</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00003109</pre>
Chris Lattner54611b42005-11-06 08:02:57 +00003110
Chris Lattner2f7c9632001-06-06 20:29:01 +00003111<h5>Overview:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +00003112
Jeff Cohen5819f182007-04-22 01:17:39 +00003113<p>The '<tt>alloca</tt>' instruction allocates memory on the stack frame of the
3114currently executing function, to be automatically released when this function
Christopher Lamb55c6d4f2007-12-17 01:00:21 +00003115returns to its caller. The object is always allocated in the generic address
3116space (address space zero).</p>
Chris Lattner54611b42005-11-06 08:02:57 +00003117
Chris Lattner2f7c9632001-06-06 20:29:01 +00003118<h5>Arguments:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +00003119
John Criswelldfe6a862004-12-10 15:51:16 +00003120<p>The '<tt>alloca</tt>' instruction allocates <tt>sizeof(&lt;type&gt;)*NumElements</tt>
Chris Lattner48b383b02003-11-25 01:02:51 +00003121bytes of memory on the runtime stack, returning a pointer of the
Gabor Greifdd1fc982008-02-09 22:24:34 +00003122appropriate type to the program. If "NumElements" is specified, it is the
3123number of elements allocated, otherwise "NumElements" is defaulted to be one.
Chris Lattner1f17cce2008-04-02 00:38:26 +00003124If a constant alignment is specified, the value result of the allocation is guaranteed
Gabor Greifdd1fc982008-02-09 22:24:34 +00003125to be aligned to at least that boundary. If not specified, or if zero, the target
3126can choose to align the allocation on any convenient boundary.</p>
Chris Lattner54611b42005-11-06 08:02:57 +00003127
Misha Brukman76307852003-11-08 01:05:38 +00003128<p>'<tt>type</tt>' may be any sized type.</p>
Chris Lattner54611b42005-11-06 08:02:57 +00003129
Chris Lattner2f7c9632001-06-06 20:29:01 +00003130<h5>Semantics:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +00003131
Chris Lattnerfe8519c2008-04-19 21:01:16 +00003132<p>Memory is allocated; a pointer is returned. The operation is undefiend if
3133there is insufficient stack space for the allocation. '<tt>alloca</tt>'d
Chris Lattner48b383b02003-11-25 01:02:51 +00003134memory is automatically released when the function returns. The '<tt>alloca</tt>'
3135instruction is commonly used to represent automatic variables that must
3136have an address available. When the function returns (either with the <tt><a
John Criswellc932bef2005-05-12 16:55:34 +00003137 href="#i_ret">ret</a></tt> or <tt><a href="#i_unwind">unwind</a></tt>
Chris Lattner1f17cce2008-04-02 00:38:26 +00003138instructions), the memory is reclaimed. Allocating zero bytes
3139is legal, but the result is undefined.</p>
Chris Lattner54611b42005-11-06 08:02:57 +00003140
Chris Lattner2f7c9632001-06-06 20:29:01 +00003141<h5>Example:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +00003142
3143<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00003144 %ptr = alloca i32 <i>; yields {i32*}:ptr</i>
Bill Wendling3716c5d2007-05-29 09:04:49 +00003145 %ptr = alloca i32, i32 4 <i>; yields {i32*}:ptr</i>
3146 %ptr = alloca i32, i32 4, align 1024 <i>; yields {i32*}:ptr</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00003147 %ptr = alloca i32, align 1024 <i>; yields {i32*}:ptr</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00003148</pre>
Misha Brukman76307852003-11-08 01:05:38 +00003149</div>
Chris Lattner54611b42005-11-06 08:02:57 +00003150
Chris Lattner2f7c9632001-06-06 20:29:01 +00003151<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +00003152<div class="doc_subsubsection"> <a name="i_load">'<tt>load</tt>'
3153Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +00003154<div class="doc_text">
Chris Lattner095735d2002-05-06 03:03:22 +00003155<h5>Syntax:</h5>
Christopher Lambbff50202007-04-21 08:16:25 +00003156<pre> &lt;result&gt; = load &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;]<br> &lt;result&gt; = volatile load &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;]<br></pre>
Chris Lattner095735d2002-05-06 03:03:22 +00003157<h5>Overview:</h5>
Misha Brukman76307852003-11-08 01:05:38 +00003158<p>The '<tt>load</tt>' instruction is used to read from memory.</p>
Chris Lattner095735d2002-05-06 03:03:22 +00003159<h5>Arguments:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00003160<p>The argument to the '<tt>load</tt>' instruction specifies the memory
John Criswell4c0cf7f2005-10-24 16:17:18 +00003161address from which to load. The pointer must point to a <a
Chris Lattner10ee9652004-06-03 22:57:15 +00003162 href="#t_firstclass">first class</a> type. If the <tt>load</tt> is
John Criswell4c0cf7f2005-10-24 16:17:18 +00003163marked as <tt>volatile</tt>, then the optimizer is not allowed to modify
Chris Lattner48b383b02003-11-25 01:02:51 +00003164the number or order of execution of this <tt>load</tt> with other
3165volatile <tt>load</tt> and <tt><a href="#i_store">store</a></tt>
3166instructions. </p>
Chris Lattner2a1993f2008-01-06 21:04:43 +00003167<p>
Chris Lattner1f17cce2008-04-02 00:38:26 +00003168The optional constant "align" argument specifies the alignment of the operation
Chris Lattner2a1993f2008-01-06 21:04:43 +00003169(that is, the alignment of the memory address). A value of 0 or an
3170omitted "align" argument means that the operation has the preferential
3171alignment for the target. It is the responsibility of the code emitter
3172to ensure that the alignment information is correct. Overestimating
3173the alignment results in an undefined behavior. Underestimating the
3174alignment may produce less efficient code. An alignment of 1 is always
3175safe.
3176</p>
Chris Lattner095735d2002-05-06 03:03:22 +00003177<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +00003178<p>The location of memory pointed to is loaded.</p>
Chris Lattner095735d2002-05-06 03:03:22 +00003179<h5>Examples:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00003180<pre> %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
Chris Lattner48b383b02003-11-25 01:02:51 +00003181 <a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00003182 href="#i_store">store</a> i32 3, i32* %ptr <i>; yields {void}</i>
3183 %val = load i32* %ptr <i>; yields {i32}:val = i32 3</i>
Chris Lattner095735d2002-05-06 03:03:22 +00003184</pre>
Misha Brukman76307852003-11-08 01:05:38 +00003185</div>
Chris Lattner095735d2002-05-06 03:03:22 +00003186<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +00003187<div class="doc_subsubsection"> <a name="i_store">'<tt>store</tt>'
3188Instruction</a> </div>
Reid Spencera89fb182006-11-09 21:18:01 +00003189<div class="doc_text">
Chris Lattner095735d2002-05-06 03:03:22 +00003190<h5>Syntax:</h5>
Christopher Lambbff50202007-04-21 08:16:25 +00003191<pre> store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;] <i>; yields {void}</i>
3192 volatile store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;] <i>; yields {void}</i>
Chris Lattner095735d2002-05-06 03:03:22 +00003193</pre>
Chris Lattner095735d2002-05-06 03:03:22 +00003194<h5>Overview:</h5>
Misha Brukman76307852003-11-08 01:05:38 +00003195<p>The '<tt>store</tt>' instruction is used to write to memory.</p>
Chris Lattner095735d2002-05-06 03:03:22 +00003196<h5>Arguments:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00003197<p>There are two arguments to the '<tt>store</tt>' instruction: a value
Jeff Cohen5819f182007-04-22 01:17:39 +00003198to store and an address at which to store it. The type of the '<tt>&lt;pointer&gt;</tt>'
Chris Lattner1f17cce2008-04-02 00:38:26 +00003199operand must be a pointer to the <a href="#t_firstclass">first class</a> type
3200of the '<tt>&lt;value&gt;</tt>'
John Criswell4a3327e2005-05-13 22:25:59 +00003201operand. If the <tt>store</tt> is marked as <tt>volatile</tt>, then the
Chris Lattner48b383b02003-11-25 01:02:51 +00003202optimizer is not allowed to modify the number or order of execution of
3203this <tt>store</tt> with other volatile <tt>load</tt> and <tt><a
3204 href="#i_store">store</a></tt> instructions.</p>
Chris Lattner2a1993f2008-01-06 21:04:43 +00003205<p>
Chris Lattner1f17cce2008-04-02 00:38:26 +00003206The optional constant "align" argument specifies the alignment of the operation
Chris Lattner2a1993f2008-01-06 21:04:43 +00003207(that is, the alignment of the memory address). A value of 0 or an
3208omitted "align" argument means that the operation has the preferential
3209alignment for the target. It is the responsibility of the code emitter
3210to ensure that the alignment information is correct. Overestimating
3211the alignment results in an undefined behavior. Underestimating the
3212alignment may produce less efficient code. An alignment of 1 is always
3213safe.
3214</p>
Chris Lattner48b383b02003-11-25 01:02:51 +00003215<h5>Semantics:</h5>
3216<p>The contents of memory are updated to contain '<tt>&lt;value&gt;</tt>'
3217at the location specified by the '<tt>&lt;pointer&gt;</tt>' operand.</p>
Chris Lattner095735d2002-05-06 03:03:22 +00003218<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00003219<pre> %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
Bill Wendling8830ffe2007-10-22 05:10:05 +00003220 store i32 3, i32* %ptr <i>; yields {void}</i>
3221 %val = <a href="#i_load">load</a> i32* %ptr <i>; yields {i32}:val = i32 3</i>
Chris Lattner095735d2002-05-06 03:03:22 +00003222</pre>
Reid Spencer443460a2006-11-09 21:15:49 +00003223</div>
3224
Chris Lattner095735d2002-05-06 03:03:22 +00003225<!-- _______________________________________________________________________ -->
Chris Lattner33fd7022004-04-05 01:30:49 +00003226<div class="doc_subsubsection">
3227 <a name="i_getelementptr">'<tt>getelementptr</tt>' Instruction</a>
3228</div>
3229
Misha Brukman76307852003-11-08 01:05:38 +00003230<div class="doc_text">
Chris Lattner590645f2002-04-14 06:13:44 +00003231<h5>Syntax:</h5>
Chris Lattner33fd7022004-04-05 01:30:49 +00003232<pre>
3233 &lt;result&gt; = getelementptr &lt;ty&gt;* &lt;ptrval&gt;{, &lt;ty&gt; &lt;idx&gt;}*
3234</pre>
3235
Chris Lattner590645f2002-04-14 06:13:44 +00003236<h5>Overview:</h5>
Chris Lattner33fd7022004-04-05 01:30:49 +00003237
3238<p>
3239The '<tt>getelementptr</tt>' instruction is used to get the address of a
3240subelement of an aggregate data structure.</p>
3241
Chris Lattner590645f2002-04-14 06:13:44 +00003242<h5>Arguments:</h5>
Chris Lattner33fd7022004-04-05 01:30:49 +00003243
Reid Spencercee005c2006-12-04 21:29:24 +00003244<p>This instruction takes a list of integer operands that indicate what
Chris Lattner33fd7022004-04-05 01:30:49 +00003245elements of the aggregate object to index to. The actual types of the arguments
3246provided depend on the type of the first pointer argument. The
3247'<tt>getelementptr</tt>' instruction is used to index down through the type
John Criswell88190562005-05-16 16:17:45 +00003248levels of a structure or to a specific index in an array. When indexing into a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00003249structure, only <tt>i32</tt> integer constants are allowed. When indexing
Chris Lattner851b7712008-04-24 05:59:56 +00003250into an array or pointer, only integers of 32 or 64 bits are allowed; 32-bit
3251values will be sign extended to 64-bits if required.</p>
Chris Lattner33fd7022004-04-05 01:30:49 +00003252
Chris Lattner48b383b02003-11-25 01:02:51 +00003253<p>For example, let's consider a C code fragment and how it gets
3254compiled to LLVM:</p>
Chris Lattner33fd7022004-04-05 01:30:49 +00003255
Bill Wendling3716c5d2007-05-29 09:04:49 +00003256<div class="doc_code">
Chris Lattner33fd7022004-04-05 01:30:49 +00003257<pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +00003258struct RT {
3259 char A;
Chris Lattnera446f1b2007-05-29 15:43:56 +00003260 int B[10][20];
Bill Wendling3716c5d2007-05-29 09:04:49 +00003261 char C;
3262};
3263struct ST {
Chris Lattnera446f1b2007-05-29 15:43:56 +00003264 int X;
Bill Wendling3716c5d2007-05-29 09:04:49 +00003265 double Y;
3266 struct RT Z;
3267};
Chris Lattner33fd7022004-04-05 01:30:49 +00003268
Chris Lattnera446f1b2007-05-29 15:43:56 +00003269int *foo(struct ST *s) {
Bill Wendling3716c5d2007-05-29 09:04:49 +00003270 return &amp;s[1].Z.B[5][13];
3271}
Chris Lattner33fd7022004-04-05 01:30:49 +00003272</pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +00003273</div>
Chris Lattner33fd7022004-04-05 01:30:49 +00003274
Misha Brukman76307852003-11-08 01:05:38 +00003275<p>The LLVM code generated by the GCC frontend is:</p>
Chris Lattner33fd7022004-04-05 01:30:49 +00003276
Bill Wendling3716c5d2007-05-29 09:04:49 +00003277<div class="doc_code">
Chris Lattner33fd7022004-04-05 01:30:49 +00003278<pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +00003279%RT = type { i8 , [10 x [20 x i32]], i8 }
3280%ST = type { i32, double, %RT }
Chris Lattner33fd7022004-04-05 01:30:49 +00003281
Bill Wendling3716c5d2007-05-29 09:04:49 +00003282define i32* %foo(%ST* %s) {
3283entry:
3284 %reg = getelementptr %ST* %s, i32 1, i32 2, i32 1, i32 5, i32 13
3285 ret i32* %reg
3286}
Chris Lattner33fd7022004-04-05 01:30:49 +00003287</pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +00003288</div>
Chris Lattner33fd7022004-04-05 01:30:49 +00003289
Chris Lattner590645f2002-04-14 06:13:44 +00003290<h5>Semantics:</h5>
Chris Lattner33fd7022004-04-05 01:30:49 +00003291
3292<p>The index types specified for the '<tt>getelementptr</tt>' instruction depend
John Criswell4a3327e2005-05-13 22:25:59 +00003293on the pointer type that is being indexed into. <a href="#t_pointer">Pointer</a>
Reid Spencercee005c2006-12-04 21:29:24 +00003294and <a href="#t_array">array</a> types can use a 32-bit or 64-bit
Reid Spencerc0312692006-12-03 16:53:48 +00003295<a href="#t_integer">integer</a> type but the value will always be sign extended
Chris Lattner1f17cce2008-04-02 00:38:26 +00003296to 64-bits. <a href="#t_struct">Structure</a> and <a href="#t_pstruct">packed
3297structure</a> types require <tt>i32</tt> <b>constants</b>.</p>
Chris Lattner33fd7022004-04-05 01:30:49 +00003298
Misha Brukman76307852003-11-08 01:05:38 +00003299<p>In the example above, the first index is indexing into the '<tt>%ST*</tt>'
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00003300type, which is a pointer, yielding a '<tt>%ST</tt>' = '<tt>{ i32, double, %RT
Chris Lattner33fd7022004-04-05 01:30:49 +00003301}</tt>' type, a structure. The second index indexes into the third element of
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00003302the structure, yielding a '<tt>%RT</tt>' = '<tt>{ i8 , [10 x [20 x i32]],
3303i8 }</tt>' type, another structure. The third index indexes into the second
3304element of the structure, yielding a '<tt>[10 x [20 x i32]]</tt>' type, an
Chris Lattner33fd7022004-04-05 01:30:49 +00003305array. The two dimensions of the array are subscripted into, yielding an
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00003306'<tt>i32</tt>' type. The '<tt>getelementptr</tt>' instruction returns a pointer
3307to this element, thus computing a value of '<tt>i32*</tt>' type.</p>
Chris Lattner33fd7022004-04-05 01:30:49 +00003308
Chris Lattner48b383b02003-11-25 01:02:51 +00003309<p>Note that it is perfectly legal to index partially through a
3310structure, returning a pointer to an inner element. Because of this,
3311the LLVM code for the given testcase is equivalent to:</p>
Chris Lattner33fd7022004-04-05 01:30:49 +00003312
3313<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00003314 define i32* %foo(%ST* %s) {
3315 %t1 = getelementptr %ST* %s, i32 1 <i>; yields %ST*:%t1</i>
Jeff Cohen5819f182007-04-22 01:17:39 +00003316 %t2 = getelementptr %ST* %t1, i32 0, i32 2 <i>; yields %RT*:%t2</i>
3317 %t3 = getelementptr %RT* %t2, i32 0, i32 1 <i>; yields [10 x [20 x i32]]*:%t3</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00003318 %t4 = getelementptr [10 x [20 x i32]]* %t3, i32 0, i32 5 <i>; yields [20 x i32]*:%t4</i>
3319 %t5 = getelementptr [20 x i32]* %t4, i32 0, i32 13 <i>; yields i32*:%t5</i>
3320 ret i32* %t5
Chris Lattner33fd7022004-04-05 01:30:49 +00003321 }
Chris Lattnera8292f32002-05-06 22:08:29 +00003322</pre>
Chris Lattnerc0ad71e2005-06-24 17:22:57 +00003323
3324<p>Note that it is undefined to access an array out of bounds: array and
3325pointer indexes must always be within the defined bounds of the array type.
Chris Lattner851b7712008-04-24 05:59:56 +00003326The one exception for this rule is zero length arrays. These arrays are
Chris Lattnerc0ad71e2005-06-24 17:22:57 +00003327defined to be accessible as variable length arrays, which requires access
3328beyond the zero'th element.</p>
3329
Chris Lattner6ab66722006-08-15 00:45:58 +00003330<p>The getelementptr instruction is often confusing. For some more insight
3331into how it works, see <a href="GetElementPtr.html">the getelementptr
3332FAQ</a>.</p>
3333
Chris Lattner590645f2002-04-14 06:13:44 +00003334<h5>Example:</h5>
Chris Lattnerc0ad71e2005-06-24 17:22:57 +00003335
Chris Lattner33fd7022004-04-05 01:30:49 +00003336<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00003337 <i>; yields [12 x i8]*:aptr</i>
3338 %aptr = getelementptr {i32, [12 x i8]}* %sptr, i64 0, i32 1
Chris Lattner33fd7022004-04-05 01:30:49 +00003339</pre>
Chris Lattner33fd7022004-04-05 01:30:49 +00003340</div>
Reid Spencer443460a2006-11-09 21:15:49 +00003341
Chris Lattner2f7c9632001-06-06 20:29:01 +00003342<!-- ======================================================================= -->
Reid Spencer97c5fa42006-11-08 01:18:52 +00003343<div class="doc_subsection"> <a name="convertops">Conversion Operations</a>
Misha Brukman76307852003-11-08 01:05:38 +00003344</div>
Misha Brukman76307852003-11-08 01:05:38 +00003345<div class="doc_text">
Reid Spencer97c5fa42006-11-08 01:18:52 +00003346<p>The instructions in this category are the conversion instructions (casting)
3347which all take a single operand and a type. They perform various bit conversions
3348on the operand.</p>
Misha Brukman76307852003-11-08 01:05:38 +00003349</div>
Chris Lattnerb53c28d2004-03-12 05:50:16 +00003350
Chris Lattnera8292f32002-05-06 22:08:29 +00003351<!-- _______________________________________________________________________ -->
Chris Lattnerb53c28d2004-03-12 05:50:16 +00003352<div class="doc_subsubsection">
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003353 <a name="i_trunc">'<tt>trunc .. to</tt>' Instruction</a>
3354</div>
3355<div class="doc_text">
3356
3357<h5>Syntax:</h5>
3358<pre>
3359 &lt;result&gt; = trunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3360</pre>
3361
3362<h5>Overview:</h5>
3363<p>
3364The '<tt>trunc</tt>' instruction truncates its operand to the type <tt>ty2</tt>.
3365</p>
3366
3367<h5>Arguments:</h5>
3368<p>
3369The '<tt>trunc</tt>' instruction takes a <tt>value</tt> to trunc, which must
3370be an <a href="#t_integer">integer</a> type, and a type that specifies the size
Chris Lattnerc0f423a2007-01-15 01:54:13 +00003371and type of the result, which must be an <a href="#t_integer">integer</a>
Reid Spencer51b07252006-11-09 23:03:26 +00003372type. The bit size of <tt>value</tt> must be larger than the bit size of
3373<tt>ty2</tt>. Equal sized types are not allowed.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003374
3375<h5>Semantics:</h5>
3376<p>
3377The '<tt>trunc</tt>' instruction truncates the high order bits in <tt>value</tt>
Reid Spencer51b07252006-11-09 23:03:26 +00003378and converts the remaining bits to <tt>ty2</tt>. Since the source size must be
3379larger than the destination size, <tt>trunc</tt> cannot be a <i>no-op cast</i>.
3380It will always truncate bits.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003381
3382<h5>Example:</h5>
3383<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00003384 %X = trunc i32 257 to i8 <i>; yields i8:1</i>
Reid Spencer36a15422007-01-12 03:35:51 +00003385 %Y = trunc i32 123 to i1 <i>; yields i1:true</i>
3386 %Y = trunc i32 122 to i1 <i>; yields i1:false</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003387</pre>
3388</div>
3389
3390<!-- _______________________________________________________________________ -->
3391<div class="doc_subsubsection">
3392 <a name="i_zext">'<tt>zext .. to</tt>' Instruction</a>
3393</div>
3394<div class="doc_text">
3395
3396<h5>Syntax:</h5>
3397<pre>
3398 &lt;result&gt; = zext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3399</pre>
3400
3401<h5>Overview:</h5>
3402<p>The '<tt>zext</tt>' instruction zero extends its operand to type
3403<tt>ty2</tt>.</p>
3404
3405
3406<h5>Arguments:</h5>
3407<p>The '<tt>zext</tt>' instruction takes a value to cast, which must be of
Chris Lattnerc0f423a2007-01-15 01:54:13 +00003408<a href="#t_integer">integer</a> type, and a type to cast it to, which must
3409also be of <a href="#t_integer">integer</a> type. The bit size of the
Reid Spencer51b07252006-11-09 23:03:26 +00003410<tt>value</tt> must be smaller than the bit size of the destination type,
3411<tt>ty2</tt>.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003412
3413<h5>Semantics:</h5>
3414<p>The <tt>zext</tt> fills the high order bits of the <tt>value</tt> with zero
Chris Lattnerc87f3df2007-05-24 19:13:27 +00003415bits until it reaches the size of the destination type, <tt>ty2</tt>.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003416
Reid Spencer07c9c682007-01-12 15:46:11 +00003417<p>When zero extending from i1, the result will always be either 0 or 1.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003418
3419<h5>Example:</h5>
3420<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00003421 %X = zext i32 257 to i64 <i>; yields i64:257</i>
Reid Spencer36a15422007-01-12 03:35:51 +00003422 %Y = zext i1 true to i32 <i>; yields i32:1</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003423</pre>
3424</div>
3425
3426<!-- _______________________________________________________________________ -->
3427<div class="doc_subsubsection">
3428 <a name="i_sext">'<tt>sext .. to</tt>' Instruction</a>
3429</div>
3430<div class="doc_text">
3431
3432<h5>Syntax:</h5>
3433<pre>
3434 &lt;result&gt; = sext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3435</pre>
3436
3437<h5>Overview:</h5>
3438<p>The '<tt>sext</tt>' sign extends <tt>value</tt> to the type <tt>ty2</tt>.</p>
3439
3440<h5>Arguments:</h5>
3441<p>
3442The '<tt>sext</tt>' instruction takes a value to cast, which must be of
Chris Lattnerc0f423a2007-01-15 01:54:13 +00003443<a href="#t_integer">integer</a> type, and a type to cast it to, which must
3444also be of <a href="#t_integer">integer</a> type. The bit size of the
Reid Spencer51b07252006-11-09 23:03:26 +00003445<tt>value</tt> must be smaller than the bit size of the destination type,
3446<tt>ty2</tt>.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003447
3448<h5>Semantics:</h5>
3449<p>
3450The '<tt>sext</tt>' instruction performs a sign extension by copying the sign
3451bit (highest order bit) of the <tt>value</tt> until it reaches the bit size of
Chris Lattnerc87f3df2007-05-24 19:13:27 +00003452the type <tt>ty2</tt>.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003453
Reid Spencer36a15422007-01-12 03:35:51 +00003454<p>When sign extending from i1, the extension always results in -1 or 0.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003455
3456<h5>Example:</h5>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003457<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00003458 %X = sext i8 -1 to i16 <i>; yields i16 :65535</i>
Reid Spencer36a15422007-01-12 03:35:51 +00003459 %Y = sext i1 true to i32 <i>; yields i32:-1</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003460</pre>
3461</div>
3462
3463<!-- _______________________________________________________________________ -->
3464<div class="doc_subsubsection">
Reid Spencer2e2740d2006-11-09 21:48:10 +00003465 <a name="i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a>
3466</div>
3467
3468<div class="doc_text">
3469
3470<h5>Syntax:</h5>
3471
3472<pre>
3473 &lt;result&gt; = fptrunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3474</pre>
3475
3476<h5>Overview:</h5>
3477<p>The '<tt>fptrunc</tt>' instruction truncates <tt>value</tt> to type
3478<tt>ty2</tt>.</p>
3479
3480
3481<h5>Arguments:</h5>
3482<p>The '<tt>fptrunc</tt>' instruction takes a <a href="#t_floating">floating
3483 point</a> value to cast and a <a href="#t_floating">floating point</a> type to
3484cast it to. The size of <tt>value</tt> must be larger than the size of
3485<tt>ty2</tt>. This implies that <tt>fptrunc</tt> cannot be used to make a
3486<i>no-op cast</i>.</p>
3487
3488<h5>Semantics:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +00003489<p> The '<tt>fptrunc</tt>' instruction truncates a <tt>value</tt> from a larger
3490<a href="#t_floating">floating point</a> type to a smaller
3491<a href="#t_floating">floating point</a> type. If the value cannot fit within
3492the destination type, <tt>ty2</tt>, then the results are undefined.</p>
Reid Spencer2e2740d2006-11-09 21:48:10 +00003493
3494<h5>Example:</h5>
3495<pre>
3496 %X = fptrunc double 123.0 to float <i>; yields float:123.0</i>
3497 %Y = fptrunc double 1.0E+300 to float <i>; yields undefined</i>
3498</pre>
3499</div>
3500
3501<!-- _______________________________________________________________________ -->
3502<div class="doc_subsubsection">
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003503 <a name="i_fpext">'<tt>fpext .. to</tt>' Instruction</a>
3504</div>
3505<div class="doc_text">
3506
3507<h5>Syntax:</h5>
3508<pre>
3509 &lt;result&gt; = fpext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3510</pre>
3511
3512<h5>Overview:</h5>
3513<p>The '<tt>fpext</tt>' extends a floating point <tt>value</tt> to a larger
3514floating point value.</p>
3515
3516<h5>Arguments:</h5>
3517<p>The '<tt>fpext</tt>' instruction takes a
3518<a href="#t_floating">floating point</a> <tt>value</tt> to cast,
Reid Spencer51b07252006-11-09 23:03:26 +00003519and a <a href="#t_floating">floating point</a> type to cast it to. The source
3520type must be smaller than the destination type.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003521
3522<h5>Semantics:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +00003523<p>The '<tt>fpext</tt>' instruction extends the <tt>value</tt> from a smaller
Duncan Sands16f122e2007-03-30 12:22:09 +00003524<a href="#t_floating">floating point</a> type to a larger
3525<a href="#t_floating">floating point</a> type. The <tt>fpext</tt> cannot be
Reid Spencer51b07252006-11-09 23:03:26 +00003526used to make a <i>no-op cast</i> because it always changes bits. Use
Reid Spencer5b950642006-11-11 23:08:07 +00003527<tt>bitcast</tt> to make a <i>no-op cast</i> for a floating point cast.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003528
3529<h5>Example:</h5>
3530<pre>
3531 %X = fpext float 3.1415 to double <i>; yields double:3.1415</i>
3532 %Y = fpext float 1.0 to float <i>; yields float:1.0 (no-op)</i>
3533</pre>
3534</div>
3535
3536<!-- _______________________________________________________________________ -->
3537<div class="doc_subsubsection">
Reid Spencer2eadb532007-01-21 00:29:26 +00003538 <a name="i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003539</div>
3540<div class="doc_text">
3541
3542<h5>Syntax:</h5>
3543<pre>
Reid Spencer753163d2007-07-31 14:40:14 +00003544 &lt;result&gt; = fptoui &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003545</pre>
3546
3547<h5>Overview:</h5>
Reid Spencer753163d2007-07-31 14:40:14 +00003548<p>The '<tt>fptoui</tt>' converts a floating point <tt>value</tt> to its
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003549unsigned integer equivalent of type <tt>ty2</tt>.
3550</p>
3551
3552<h5>Arguments:</h5>
Reid Spencer753163d2007-07-31 14:40:14 +00003553<p>The '<tt>fptoui</tt>' instruction takes a value to cast, which must be a
Nate Begemand4d45c22007-11-17 03:58:34 +00003554scalar or vector <a href="#t_floating">floating point</a> value, and a type
3555to cast it to <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a>
3556type. If <tt>ty</tt> is a vector floating point type, <tt>ty2</tt> must be a
3557vector integer type with the same number of elements as <tt>ty</tt></p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003558
3559<h5>Semantics:</h5>
Reid Spencer753163d2007-07-31 14:40:14 +00003560<p> The '<tt>fptoui</tt>' instruction converts its
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003561<a href="#t_floating">floating point</a> operand into the nearest (rounding
3562towards zero) unsigned integer value. If the value cannot fit in <tt>ty2</tt>,
3563the results are undefined.</p>
3564
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003565<h5>Example:</h5>
3566<pre>
Reid Spencer753163d2007-07-31 14:40:14 +00003567 %X = fptoui double 123.0 to i32 <i>; yields i32:123</i>
Chris Lattner5b95a172007-09-22 03:17:52 +00003568 %Y = fptoui float 1.0E+300 to i1 <i>; yields undefined:1</i>
Reid Spencer753163d2007-07-31 14:40:14 +00003569 %X = fptoui float 1.04E+17 to i8 <i>; yields undefined:1</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003570</pre>
3571</div>
3572
3573<!-- _______________________________________________________________________ -->
3574<div class="doc_subsubsection">
Reid Spencer51b07252006-11-09 23:03:26 +00003575 <a name="i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003576</div>
3577<div class="doc_text">
3578
3579<h5>Syntax:</h5>
3580<pre>
Reid Spencer51b07252006-11-09 23:03:26 +00003581 &lt;result&gt; = fptosi &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003582</pre>
3583
3584<h5>Overview:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +00003585<p>The '<tt>fptosi</tt>' instruction converts
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003586<a href="#t_floating">floating point</a> <tt>value</tt> to type <tt>ty2</tt>.
Chris Lattnerb53c28d2004-03-12 05:50:16 +00003587</p>
3588
Chris Lattnera8292f32002-05-06 22:08:29 +00003589<h5>Arguments:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +00003590<p> The '<tt>fptosi</tt>' instruction takes a value to cast, which must be a
Nate Begemand4d45c22007-11-17 03:58:34 +00003591scalar or vector <a href="#t_floating">floating point</a> value, and a type
3592to cast it to <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a>
3593type. If <tt>ty</tt> is a vector floating point type, <tt>ty2</tt> must be a
3594vector integer type with the same number of elements as <tt>ty</tt></p>
Chris Lattnerb53c28d2004-03-12 05:50:16 +00003595
Chris Lattnera8292f32002-05-06 22:08:29 +00003596<h5>Semantics:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +00003597<p>The '<tt>fptosi</tt>' instruction converts its
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003598<a href="#t_floating">floating point</a> operand into the nearest (rounding
3599towards zero) signed integer value. If the value cannot fit in <tt>ty2</tt>,
3600the results are undefined.</p>
Chris Lattnerb53c28d2004-03-12 05:50:16 +00003601
Chris Lattner70de6632001-07-09 00:26:23 +00003602<h5>Example:</h5>
Chris Lattnerb53c28d2004-03-12 05:50:16 +00003603<pre>
Reid Spencer36a15422007-01-12 03:35:51 +00003604 %X = fptosi double -123.0 to i32 <i>; yields i32:-123</i>
Chris Lattner5b95a172007-09-22 03:17:52 +00003605 %Y = fptosi float 1.0E-247 to i1 <i>; yields undefined:1</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00003606 %X = fptosi float 1.04E+17 to i8 <i>; yields undefined:1</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003607</pre>
3608</div>
3609
3610<!-- _______________________________________________________________________ -->
3611<div class="doc_subsubsection">
Reid Spencer51b07252006-11-09 23:03:26 +00003612 <a name="i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003613</div>
3614<div class="doc_text">
3615
3616<h5>Syntax:</h5>
3617<pre>
Reid Spencer51b07252006-11-09 23:03:26 +00003618 &lt;result&gt; = uitofp &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003619</pre>
3620
3621<h5>Overview:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +00003622<p>The '<tt>uitofp</tt>' instruction regards <tt>value</tt> as an unsigned
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003623integer and converts that value to the <tt>ty2</tt> type.</p>
3624
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003625<h5>Arguments:</h5>
Nate Begemand4d45c22007-11-17 03:58:34 +00003626<p>The '<tt>uitofp</tt>' instruction takes a value to cast, which must be a
3627scalar or vector <a href="#t_integer">integer</a> value, and a type to cast it
3628to <tt>ty2</tt>, which must be an <a href="#t_floating">floating point</a>
3629type. If <tt>ty</tt> is a vector integer type, <tt>ty2</tt> must be a vector
3630floating point type with the same number of elements as <tt>ty</tt></p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003631
3632<h5>Semantics:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +00003633<p>The '<tt>uitofp</tt>' instruction interprets its operand as an unsigned
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003634integer quantity and converts it to the corresponding floating point value. If
Jeff Cohenbeccb742007-04-22 14:56:37 +00003635the value cannot fit in the floating point value, the results are undefined.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003636
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003637<h5>Example:</h5>
3638<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00003639 %X = uitofp i32 257 to float <i>; yields float:257.0</i>
Jeff Cohen222a8a42007-04-29 01:07:00 +00003640 %Y = uitofp i8 -1 to double <i>; yields double:255.0</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003641</pre>
3642</div>
3643
3644<!-- _______________________________________________________________________ -->
3645<div class="doc_subsubsection">
Reid Spencer51b07252006-11-09 23:03:26 +00003646 <a name="i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003647</div>
3648<div class="doc_text">
3649
3650<h5>Syntax:</h5>
3651<pre>
Reid Spencer51b07252006-11-09 23:03:26 +00003652 &lt;result&gt; = sitofp &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003653</pre>
3654
3655<h5>Overview:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +00003656<p>The '<tt>sitofp</tt>' instruction regards <tt>value</tt> as a signed
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003657integer and converts that value to the <tt>ty2</tt> type.</p>
3658
3659<h5>Arguments:</h5>
Nate Begemand4d45c22007-11-17 03:58:34 +00003660<p>The '<tt>sitofp</tt>' instruction takes a value to cast, which must be a
3661scalar or vector <a href="#t_integer">integer</a> value, and a type to cast it
3662to <tt>ty2</tt>, which must be an <a href="#t_floating">floating point</a>
3663type. If <tt>ty</tt> is a vector integer type, <tt>ty2</tt> must be a vector
3664floating point type with the same number of elements as <tt>ty</tt></p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003665
3666<h5>Semantics:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +00003667<p>The '<tt>sitofp</tt>' instruction interprets its operand as a signed
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003668integer quantity and converts it to the corresponding floating point value. If
Jeff Cohenbeccb742007-04-22 14:56:37 +00003669the value cannot fit in the floating point value, the results are undefined.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003670
3671<h5>Example:</h5>
3672<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00003673 %X = sitofp i32 257 to float <i>; yields float:257.0</i>
Jeff Cohen222a8a42007-04-29 01:07:00 +00003674 %Y = sitofp i8 -1 to double <i>; yields double:-1.0</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003675</pre>
3676</div>
3677
3678<!-- _______________________________________________________________________ -->
3679<div class="doc_subsubsection">
Reid Spencerb7344ff2006-11-11 21:00:47 +00003680 <a name="i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a>
3681</div>
3682<div class="doc_text">
3683
3684<h5>Syntax:</h5>
3685<pre>
3686 &lt;result&gt; = ptrtoint &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3687</pre>
3688
3689<h5>Overview:</h5>
3690<p>The '<tt>ptrtoint</tt>' instruction converts the pointer <tt>value</tt> to
3691the integer type <tt>ty2</tt>.</p>
3692
3693<h5>Arguments:</h5>
3694<p>The '<tt>ptrtoint</tt>' instruction takes a <tt>value</tt> to cast, which
Duncan Sands16f122e2007-03-30 12:22:09 +00003695must be a <a href="#t_pointer">pointer</a> value, and a type to cast it to
Reid Spencerb7344ff2006-11-11 21:00:47 +00003696<tt>ty2</tt>, which must be an <a href="#t_integer">integer</a> type.
3697
3698<h5>Semantics:</h5>
3699<p>The '<tt>ptrtoint</tt>' instruction converts <tt>value</tt> to integer type
3700<tt>ty2</tt> by interpreting the pointer value as an integer and either
3701truncating or zero extending that value to the size of the integer type. If
3702<tt>value</tt> is smaller than <tt>ty2</tt> then a zero extension is done. If
3703<tt>value</tt> is larger than <tt>ty2</tt> then a truncation is done. If they
Jeff Cohen222a8a42007-04-29 01:07:00 +00003704are the same size, then nothing is done (<i>no-op cast</i>) other than a type
3705change.</p>
Reid Spencerb7344ff2006-11-11 21:00:47 +00003706
3707<h5>Example:</h5>
3708<pre>
Jeff Cohen222a8a42007-04-29 01:07:00 +00003709 %X = ptrtoint i32* %X to i8 <i>; yields truncation on 32-bit architecture</i>
3710 %Y = ptrtoint i32* %x to i64 <i>; yields zero extension on 32-bit architecture</i>
Reid Spencerb7344ff2006-11-11 21:00:47 +00003711</pre>
3712</div>
3713
3714<!-- _______________________________________________________________________ -->
3715<div class="doc_subsubsection">
3716 <a name="i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a>
3717</div>
3718<div class="doc_text">
3719
3720<h5>Syntax:</h5>
3721<pre>
3722 &lt;result&gt; = inttoptr &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3723</pre>
3724
3725<h5>Overview:</h5>
3726<p>The '<tt>inttoptr</tt>' instruction converts an integer <tt>value</tt> to
3727a pointer type, <tt>ty2</tt>.</p>
3728
3729<h5>Arguments:</h5>
Duncan Sands16f122e2007-03-30 12:22:09 +00003730<p>The '<tt>inttoptr</tt>' instruction takes an <a href="#t_integer">integer</a>
Reid Spencerb7344ff2006-11-11 21:00:47 +00003731value to cast, and a type to cast it to, which must be a
Anton Korobeynikova0554d92007-01-12 19:20:47 +00003732<a href="#t_pointer">pointer</a> type.
Reid Spencerb7344ff2006-11-11 21:00:47 +00003733
3734<h5>Semantics:</h5>
3735<p>The '<tt>inttoptr</tt>' instruction converts <tt>value</tt> to type
3736<tt>ty2</tt> by applying either a zero extension or a truncation depending on
3737the size of the integer <tt>value</tt>. If <tt>value</tt> is larger than the
3738size of a pointer then a truncation is done. If <tt>value</tt> is smaller than
3739the size of a pointer then a zero extension is done. If they are the same size,
3740nothing is done (<i>no-op cast</i>).</p>
3741
3742<h5>Example:</h5>
3743<pre>
Jeff Cohen222a8a42007-04-29 01:07:00 +00003744 %X = inttoptr i32 255 to i32* <i>; yields zero extension on 64-bit architecture</i>
3745 %X = inttoptr i32 255 to i32* <i>; yields no-op on 32-bit architecture</i>
3746 %Y = inttoptr i64 0 to i32* <i>; yields truncation on 32-bit architecture</i>
Reid Spencerb7344ff2006-11-11 21:00:47 +00003747</pre>
3748</div>
3749
3750<!-- _______________________________________________________________________ -->
3751<div class="doc_subsubsection">
Reid Spencer5b950642006-11-11 23:08:07 +00003752 <a name="i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003753</div>
3754<div class="doc_text">
3755
3756<h5>Syntax:</h5>
3757<pre>
Reid Spencer5b950642006-11-11 23:08:07 +00003758 &lt;result&gt; = bitcast &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003759</pre>
3760
3761<h5>Overview:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00003762
Reid Spencer5b950642006-11-11 23:08:07 +00003763<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003764<tt>ty2</tt> without changing any bits.</p>
3765
3766<h5>Arguments:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00003767
Reid Spencer5b950642006-11-11 23:08:07 +00003768<p>The '<tt>bitcast</tt>' instruction takes a value to cast, which must be
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003769a first class value, and a type to cast it to, which must also be a <a
3770 href="#t_firstclass">first class</a> type. The bit sizes of <tt>value</tt>
Reid Spencere3db84c2007-01-09 20:08:58 +00003771and the destination type, <tt>ty2</tt>, must be identical. If the source
Chris Lattnerb54c30f2008-05-20 20:48:21 +00003772type is a pointer, the destination type must also be a pointer. This
3773instruction supports bitwise conversion of vectors to integers and to vectors
3774of other types (as long as they have the same size).</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003775
3776<h5>Semantics:</h5>
Reid Spencer5b950642006-11-11 23:08:07 +00003777<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
Reid Spencerb7344ff2006-11-11 21:00:47 +00003778<tt>ty2</tt>. It is always a <i>no-op cast</i> because no bits change with
3779this conversion. The conversion is done as if the <tt>value</tt> had been
3780stored to memory and read back as type <tt>ty2</tt>. Pointer types may only be
3781converted to other pointer types with this instruction. To convert pointers to
3782other types, use the <a href="#i_inttoptr">inttoptr</a> or
3783<a href="#i_ptrtoint">ptrtoint</a> instructions first.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +00003784
3785<h5>Example:</h5>
3786<pre>
Jeff Cohen222a8a42007-04-29 01:07:00 +00003787 %X = bitcast i8 255 to i8 <i>; yields i8 :-1</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00003788 %Y = bitcast i32* %x to sint* <i>; yields sint*:%x</i>
3789 %Z = bitcast <2xint> %V to i64; <i>; yields i64: %V</i>
Chris Lattner70de6632001-07-09 00:26:23 +00003790</pre>
Misha Brukman76307852003-11-08 01:05:38 +00003791</div>
Chris Lattnerb53c28d2004-03-12 05:50:16 +00003792
Reid Spencer97c5fa42006-11-08 01:18:52 +00003793<!-- ======================================================================= -->
3794<div class="doc_subsection"> <a name="otherops">Other Operations</a> </div>
3795<div class="doc_text">
3796<p>The instructions in this category are the "miscellaneous"
3797instructions, which defy better classification.</p>
3798</div>
Reid Spencerc828a0e2006-11-18 21:50:54 +00003799
3800<!-- _______________________________________________________________________ -->
3801<div class="doc_subsubsection"><a name="i_icmp">'<tt>icmp</tt>' Instruction</a>
3802</div>
3803<div class="doc_text">
3804<h5>Syntax:</h5>
Jeff Cohen222a8a42007-04-29 01:07:00 +00003805<pre> &lt;result&gt; = icmp &lt;cond&gt; &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {i1}:result</i>
Reid Spencerc828a0e2006-11-18 21:50:54 +00003806</pre>
3807<h5>Overview:</h5>
3808<p>The '<tt>icmp</tt>' instruction returns a boolean value based on comparison
Chris Lattner1f17cce2008-04-02 00:38:26 +00003809of its two integer or pointer operands.</p>
Reid Spencerc828a0e2006-11-18 21:50:54 +00003810<h5>Arguments:</h5>
3811<p>The '<tt>icmp</tt>' instruction takes three operands. The first operand is
Jeff Cohen222a8a42007-04-29 01:07:00 +00003812the condition code indicating the kind of comparison to perform. It is not
3813a value, just a keyword. The possible condition code are:
Reid Spencerc828a0e2006-11-18 21:50:54 +00003814<ol>
3815 <li><tt>eq</tt>: equal</li>
3816 <li><tt>ne</tt>: not equal </li>
3817 <li><tt>ugt</tt>: unsigned greater than</li>
3818 <li><tt>uge</tt>: unsigned greater or equal</li>
3819 <li><tt>ult</tt>: unsigned less than</li>
3820 <li><tt>ule</tt>: unsigned less or equal</li>
3821 <li><tt>sgt</tt>: signed greater than</li>
3822 <li><tt>sge</tt>: signed greater or equal</li>
3823 <li><tt>slt</tt>: signed less than</li>
3824 <li><tt>sle</tt>: signed less or equal</li>
3825</ol>
Chris Lattnerc0f423a2007-01-15 01:54:13 +00003826<p>The remaining two arguments must be <a href="#t_integer">integer</a> or
Reid Spencer784ef792007-01-04 05:19:58 +00003827<a href="#t_pointer">pointer</a> typed. They must also be identical types.</p>
Reid Spencerc828a0e2006-11-18 21:50:54 +00003828<h5>Semantics:</h5>
3829<p>The '<tt>icmp</tt>' compares <tt>var1</tt> and <tt>var2</tt> according to
3830the condition code given as <tt>cond</tt>. The comparison performed always
Reid Spencer36a15422007-01-12 03:35:51 +00003831yields a <a href="#t_primitive">i1</a> result, as follows:
Reid Spencerc828a0e2006-11-18 21:50:54 +00003832<ol>
3833 <li><tt>eq</tt>: yields <tt>true</tt> if the operands are equal,
3834 <tt>false</tt> otherwise. No sign interpretation is necessary or performed.
3835 </li>
3836 <li><tt>ne</tt>: yields <tt>true</tt> if the operands are unequal,
3837 <tt>false</tt> otherwise. No sign interpretation is necessary or performed.
3838 <li><tt>ugt</tt>: interprets the operands as unsigned values and yields
3839 <tt>true</tt> if <tt>var1</tt> is greater than <tt>var2</tt>.</li>
3840 <li><tt>uge</tt>: interprets the operands as unsigned values and yields
3841 <tt>true</tt> if <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
3842 <li><tt>ult</tt>: interprets the operands as unsigned values and yields
3843 <tt>true</tt> if <tt>var1</tt> is less than <tt>var2</tt>.</li>
3844 <li><tt>ule</tt>: interprets the operands as unsigned values and yields
3845 <tt>true</tt> if <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
3846 <li><tt>sgt</tt>: interprets the operands as signed values and yields
3847 <tt>true</tt> if <tt>var1</tt> is greater than <tt>var2</tt>.</li>
3848 <li><tt>sge</tt>: interprets the operands as signed values and yields
3849 <tt>true</tt> if <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
3850 <li><tt>slt</tt>: interprets the operands as signed values and yields
3851 <tt>true</tt> if <tt>var1</tt> is less than <tt>var2</tt>.</li>
3852 <li><tt>sle</tt>: interprets the operands as signed values and yields
3853 <tt>true</tt> if <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
Reid Spencerc828a0e2006-11-18 21:50:54 +00003854</ol>
3855<p>If the operands are <a href="#t_pointer">pointer</a> typed, the pointer
Jeff Cohen222a8a42007-04-29 01:07:00 +00003856values are compared as if they were integers.</p>
Reid Spencerc828a0e2006-11-18 21:50:54 +00003857
3858<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00003859<pre> &lt;result&gt; = icmp eq i32 4, 5 <i>; yields: result=false</i>
3860 &lt;result&gt; = icmp ne float* %X, %X <i>; yields: result=false</i>
3861 &lt;result&gt; = icmp ult i16 4, 5 <i>; yields: result=true</i>
3862 &lt;result&gt; = icmp sgt i16 4, 5 <i>; yields: result=false</i>
3863 &lt;result&gt; = icmp ule i16 -4, 5 <i>; yields: result=false</i>
3864 &lt;result&gt; = icmp sge i16 4, 5 <i>; yields: result=false</i>
Reid Spencerc828a0e2006-11-18 21:50:54 +00003865</pre>
3866</div>
3867
3868<!-- _______________________________________________________________________ -->
3869<div class="doc_subsubsection"><a name="i_fcmp">'<tt>fcmp</tt>' Instruction</a>
3870</div>
3871<div class="doc_text">
3872<h5>Syntax:</h5>
Jeff Cohen222a8a42007-04-29 01:07:00 +00003873<pre> &lt;result&gt; = fcmp &lt;cond&gt; &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {i1}:result</i>
Reid Spencerc828a0e2006-11-18 21:50:54 +00003874</pre>
3875<h5>Overview:</h5>
3876<p>The '<tt>fcmp</tt>' instruction returns a boolean value based on comparison
3877of its floating point operands.</p>
3878<h5>Arguments:</h5>
3879<p>The '<tt>fcmp</tt>' instruction takes three operands. The first operand is
Jeff Cohen222a8a42007-04-29 01:07:00 +00003880the condition code indicating the kind of comparison to perform. It is not
3881a value, just a keyword. The possible condition code are:
Reid Spencerc828a0e2006-11-18 21:50:54 +00003882<ol>
Reid Spencerf69acf32006-11-19 03:00:14 +00003883 <li><tt>false</tt>: no comparison, always returns false</li>
Reid Spencerc828a0e2006-11-18 21:50:54 +00003884 <li><tt>oeq</tt>: ordered and equal</li>
3885 <li><tt>ogt</tt>: ordered and greater than </li>
3886 <li><tt>oge</tt>: ordered and greater than or equal</li>
3887 <li><tt>olt</tt>: ordered and less than </li>
3888 <li><tt>ole</tt>: ordered and less than or equal</li>
3889 <li><tt>one</tt>: ordered and not equal</li>
3890 <li><tt>ord</tt>: ordered (no nans)</li>
3891 <li><tt>ueq</tt>: unordered or equal</li>
3892 <li><tt>ugt</tt>: unordered or greater than </li>
3893 <li><tt>uge</tt>: unordered or greater than or equal</li>
3894 <li><tt>ult</tt>: unordered or less than </li>
3895 <li><tt>ule</tt>: unordered or less than or equal</li>
3896 <li><tt>une</tt>: unordered or not equal</li>
3897 <li><tt>uno</tt>: unordered (either nans)</li>
Reid Spencerf69acf32006-11-19 03:00:14 +00003898 <li><tt>true</tt>: no comparison, always returns true</li>
Reid Spencerc828a0e2006-11-18 21:50:54 +00003899</ol>
Jeff Cohen222a8a42007-04-29 01:07:00 +00003900<p><i>Ordered</i> means that neither operand is a QNAN while
Reid Spencer02e0d1d2006-12-06 07:08:07 +00003901<i>unordered</i> means that either operand may be a QNAN.</p>
Reid Spencer784ef792007-01-04 05:19:58 +00003902<p>The <tt>val1</tt> and <tt>val2</tt> arguments must be
3903<a href="#t_floating">floating point</a> typed. They must have identical
3904types.</p>
Reid Spencerc828a0e2006-11-18 21:50:54 +00003905<h5>Semantics:</h5>
Nate Begemand2195702008-05-12 19:01:56 +00003906<p>The '<tt>fcmp</tt>' instruction compares <tt>var1</tt> and <tt>var2</tt>
3907according to the condition code given as <tt>cond</tt>. The comparison performed
3908always yields a <a href="#t_primitive">i1</a> result, as follows:
Reid Spencerc828a0e2006-11-18 21:50:54 +00003909<ol>
3910 <li><tt>false</tt>: always yields <tt>false</tt>, regardless of operands.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +00003911 <li><tt>oeq</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Reid Spencerc828a0e2006-11-18 21:50:54 +00003912 <tt>var1</tt> is equal to <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +00003913 <li><tt>ogt</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Reid Spencerc828a0e2006-11-18 21:50:54 +00003914 <tt>var1</tt> is greather than <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +00003915 <li><tt>oge</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Reid Spencerc828a0e2006-11-18 21:50:54 +00003916 <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +00003917 <li><tt>olt</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Reid Spencerc828a0e2006-11-18 21:50:54 +00003918 <tt>var1</tt> is less than <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +00003919 <li><tt>ole</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Reid Spencerc828a0e2006-11-18 21:50:54 +00003920 <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +00003921 <li><tt>one</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Reid Spencerc828a0e2006-11-18 21:50:54 +00003922 <tt>var1</tt> is not equal to <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +00003923 <li><tt>ord</tt>: yields <tt>true</tt> if both operands are not a QNAN.</li>
3924 <li><tt>ueq</tt>: yields <tt>true</tt> if either operand is a QNAN or
Reid Spencerc828a0e2006-11-18 21:50:54 +00003925 <tt>var1</tt> is equal to <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +00003926 <li><tt>ugt</tt>: yields <tt>true</tt> if either operand is a QNAN or
Reid Spencerc828a0e2006-11-18 21:50:54 +00003927 <tt>var1</tt> is greater than <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +00003928 <li><tt>uge</tt>: yields <tt>true</tt> if either operand is a QNAN or
Reid Spencerc828a0e2006-11-18 21:50:54 +00003929 <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +00003930 <li><tt>ult</tt>: yields <tt>true</tt> if either operand is a QNAN or
Reid Spencerc828a0e2006-11-18 21:50:54 +00003931 <tt>var1</tt> is less than <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +00003932 <li><tt>ule</tt>: yields <tt>true</tt> if either operand is a QNAN or
Reid Spencerc828a0e2006-11-18 21:50:54 +00003933 <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +00003934 <li><tt>une</tt>: yields <tt>true</tt> if either operand is a QNAN or
Reid Spencerc828a0e2006-11-18 21:50:54 +00003935 <tt>var1</tt> is not equal to <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +00003936 <li><tt>uno</tt>: yields <tt>true</tt> if either operand is a QNAN.</li>
Reid Spencerc828a0e2006-11-18 21:50:54 +00003937 <li><tt>true</tt>: always yields <tt>true</tt>, regardless of operands.</li>
3938</ol>
Reid Spencerc828a0e2006-11-18 21:50:54 +00003939
3940<h5>Example:</h5>
3941<pre> &lt;result&gt; = fcmp oeq float 4.0, 5.0 <i>; yields: result=false</i>
3942 &lt;result&gt; = icmp one float 4.0, 5.0 <i>; yields: result=true</i>
3943 &lt;result&gt; = icmp olt float 4.0, 5.0 <i>; yields: result=true</i>
3944 &lt;result&gt; = icmp ueq double 1.0, 2.0 <i>; yields: result=false</i>
3945</pre>
3946</div>
3947
Reid Spencer97c5fa42006-11-08 01:18:52 +00003948<!-- _______________________________________________________________________ -->
Nate Begemand2195702008-05-12 19:01:56 +00003949<div class="doc_subsubsection">
3950 <a name="i_vicmp">'<tt>vicmp</tt>' Instruction</a>
3951</div>
3952<div class="doc_text">
3953<h5>Syntax:</h5>
3954<pre> &lt;result&gt; = vicmp &lt;cond&gt; &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
3955</pre>
3956<h5>Overview:</h5>
3957<p>The '<tt>vicmp</tt>' instruction returns an integer vector value based on
3958element-wise comparison of its two integer vector operands.</p>
3959<h5>Arguments:</h5>
3960<p>The '<tt>vicmp</tt>' instruction takes three operands. The first operand is
3961the condition code indicating the kind of comparison to perform. It is not
3962a value, just a keyword. The possible condition code are:
3963<ol>
3964 <li><tt>eq</tt>: equal</li>
3965 <li><tt>ne</tt>: not equal </li>
3966 <li><tt>ugt</tt>: unsigned greater than</li>
3967 <li><tt>uge</tt>: unsigned greater or equal</li>
3968 <li><tt>ult</tt>: unsigned less than</li>
3969 <li><tt>ule</tt>: unsigned less or equal</li>
3970 <li><tt>sgt</tt>: signed greater than</li>
3971 <li><tt>sge</tt>: signed greater or equal</li>
3972 <li><tt>slt</tt>: signed less than</li>
3973 <li><tt>sle</tt>: signed less or equal</li>
3974</ol>
3975<p>The remaining two arguments must be <a href="#t_vector">vector</a> of
3976<a href="#t_integer">integer</a> typed. They must also be identical types.</p>
3977<h5>Semantics:</h5>
3978<p>The '<tt>vicmp</tt>' instruction compares <tt>var1</tt> and <tt>var2</tt>
3979according to the condition code given as <tt>cond</tt>. The comparison yields a
3980<a href="#t_vector">vector</a> of <a href="#t_integer">integer</a> result, of
3981identical type as the values being compared. The most significant bit in each
3982element is 1 if the element-wise comparison evaluates to true, and is 0
3983otherwise. All other bits of the result are undefined. The condition codes
3984are evaluated identically to the <a href="#i_icmp">'<tt>icmp</tt>'
3985instruction</a>.
3986
3987<h5>Example:</h5>
3988<pre>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00003989 &lt;result&gt; = vicmp eq &lt;2 x i32&gt; &lt; i32 4, i32 0&gt;, &lt; i32 5, i32 0&gt; <i>; yields: result=&lt;2 x i32&gt; &lt; i32 0, i32 -1 &gt;</i>
3990 &lt;result&gt; = vicmp ult &lt;2 x i8 &gt; &lt; i8 1, i8 2&gt;, &lt; i8 2, i8 2 &gt; <i>; yields: result=&lt;2 x i8&gt; &lt; i8 -1, i8 0 &gt;</i>
Nate Begemand2195702008-05-12 19:01:56 +00003991</pre>
3992</div>
3993
3994<!-- _______________________________________________________________________ -->
3995<div class="doc_subsubsection">
3996 <a name="i_vfcmp">'<tt>vfcmp</tt>' Instruction</a>
3997</div>
3998<div class="doc_text">
3999<h5>Syntax:</h5>
4000<pre> &lt;result&gt; = vfcmp &lt;cond&gt; &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt;</pre>
4001<h5>Overview:</h5>
4002<p>The '<tt>vfcmp</tt>' instruction returns an integer vector value based on
4003element-wise comparison of its two floating point vector operands. The output
4004elements have the same width as the input elements.</p>
4005<h5>Arguments:</h5>
4006<p>The '<tt>vfcmp</tt>' instruction takes three operands. The first operand is
4007the condition code indicating the kind of comparison to perform. It is not
4008a value, just a keyword. The possible condition code are:
4009<ol>
4010 <li><tt>false</tt>: no comparison, always returns false</li>
4011 <li><tt>oeq</tt>: ordered and equal</li>
4012 <li><tt>ogt</tt>: ordered and greater than </li>
4013 <li><tt>oge</tt>: ordered and greater than or equal</li>
4014 <li><tt>olt</tt>: ordered and less than </li>
4015 <li><tt>ole</tt>: ordered and less than or equal</li>
4016 <li><tt>one</tt>: ordered and not equal</li>
4017 <li><tt>ord</tt>: ordered (no nans)</li>
4018 <li><tt>ueq</tt>: unordered or equal</li>
4019 <li><tt>ugt</tt>: unordered or greater than </li>
4020 <li><tt>uge</tt>: unordered or greater than or equal</li>
4021 <li><tt>ult</tt>: unordered or less than </li>
4022 <li><tt>ule</tt>: unordered or less than or equal</li>
4023 <li><tt>une</tt>: unordered or not equal</li>
4024 <li><tt>uno</tt>: unordered (either nans)</li>
4025 <li><tt>true</tt>: no comparison, always returns true</li>
4026</ol>
4027<p>The remaining two arguments must be <a href="#t_vector">vector</a> of
4028<a href="#t_floating">floating point</a> typed. They must also be identical
4029types.</p>
4030<h5>Semantics:</h5>
4031<p>The '<tt>vfcmp</tt>' instruction compares <tt>var1</tt> and <tt>var2</tt>
4032according to the condition code given as <tt>cond</tt>. The comparison yields a
4033<a href="#t_vector">vector</a> of <a href="#t_integer">integer</a> result, with
4034an identical number of elements as the values being compared, and each element
4035having identical with to the width of the floating point elements. The most
4036significant bit in each element is 1 if the element-wise comparison evaluates to
4037true, and is 0 otherwise. All other bits of the result are undefined. The
4038condition codes are evaluated identically to the
4039<a href="#i_fcmp">'<tt>fcmp</tt>' instruction</a>.
4040
4041<h5>Example:</h5>
4042<pre>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00004043 &lt;result&gt; = vfcmp oeq &lt;2 x float&gt; &lt; float 4, float 0 &gt;, &lt; float 5, float 0 &gt; <i>; yields: result=&lt;2 x i32&gt; &lt; i32 0, i32 -1 &gt;</i>
4044 &lt;result&gt; = vfcmp ult &lt;2 x double&gt; &lt; double 1, double 2 &gt;, &lt; double 2, double 2&gt; <i>; yields: result=&lt;2 x i64&gt; &lt; i64 -1, i64 0 &gt;</i>
Nate Begemand2195702008-05-12 19:01:56 +00004045</pre>
4046</div>
4047
4048<!-- _______________________________________________________________________ -->
Chris Lattnerb54c30f2008-05-20 20:48:21 +00004049<div class="doc_subsubsection">
4050 <a name="i_phi">'<tt>phi</tt>' Instruction</a>
4051</div>
4052
Reid Spencer97c5fa42006-11-08 01:18:52 +00004053<div class="doc_text">
Chris Lattnerb54c30f2008-05-20 20:48:21 +00004054
Reid Spencer97c5fa42006-11-08 01:18:52 +00004055<h5>Syntax:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00004056
Reid Spencer97c5fa42006-11-08 01:18:52 +00004057<pre> &lt;result&gt; = phi &lt;ty&gt; [ &lt;val0&gt;, &lt;label0&gt;], ...<br></pre>
4058<h5>Overview:</h5>
4059<p>The '<tt>phi</tt>' instruction is used to implement the &#966; node in
4060the SSA graph representing the function.</p>
4061<h5>Arguments:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00004062
Jeff Cohen222a8a42007-04-29 01:07:00 +00004063<p>The type of the incoming values is specified with the first type
Reid Spencer97c5fa42006-11-08 01:18:52 +00004064field. After this, the '<tt>phi</tt>' instruction takes a list of pairs
4065as arguments, with one pair for each predecessor basic block of the
4066current block. Only values of <a href="#t_firstclass">first class</a>
4067type may be used as the value arguments to the PHI node. Only labels
4068may be used as the label arguments.</p>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00004069
Reid Spencer97c5fa42006-11-08 01:18:52 +00004070<p>There must be no non-phi instructions between the start of a basic
4071block and the PHI instructions: i.e. PHI instructions must be first in
4072a basic block.</p>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00004073
Reid Spencer97c5fa42006-11-08 01:18:52 +00004074<h5>Semantics:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00004075
Jeff Cohen222a8a42007-04-29 01:07:00 +00004076<p>At runtime, the '<tt>phi</tt>' instruction logically takes on the value
4077specified by the pair corresponding to the predecessor basic block that executed
4078just prior to the current block.</p>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00004079
Reid Spencer97c5fa42006-11-08 01:18:52 +00004080<h5>Example:</h5>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00004081<pre>
4082Loop: ; Infinite loop that counts from 0 on up...
4083 %indvar = phi i32 [ 0, %LoopHeader ], [ %nextindvar, %Loop ]
4084 %nextindvar = add i32 %indvar, 1
4085 br label %Loop
4086</pre>
Reid Spencer97c5fa42006-11-08 01:18:52 +00004087</div>
4088
Chris Lattnerb53c28d2004-03-12 05:50:16 +00004089<!-- _______________________________________________________________________ -->
4090<div class="doc_subsubsection">
4091 <a name="i_select">'<tt>select</tt>' Instruction</a>
4092</div>
4093
4094<div class="doc_text">
4095
4096<h5>Syntax:</h5>
4097
4098<pre>
Reid Spencer36a15422007-01-12 03:35:51 +00004099 &lt;result&gt; = select i1 &lt;cond&gt;, &lt;ty&gt; &lt;val1&gt;, &lt;ty&gt; &lt;val2&gt; <i>; yields ty</i>
Chris Lattnerb53c28d2004-03-12 05:50:16 +00004100</pre>
4101
4102<h5>Overview:</h5>
4103
4104<p>
4105The '<tt>select</tt>' instruction is used to choose one value based on a
4106condition, without branching.
4107</p>
4108
4109
4110<h5>Arguments:</h5>
4111
4112<p>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00004113The '<tt>select</tt>' instruction requires an 'i1' value indicating the
4114condition, and two values of the same <a href="#t_firstclass">first class</a>
4115type. If the val1/val2 are vectors, the entire vectors are selected, not
4116individual elements.
Chris Lattnerb53c28d2004-03-12 05:50:16 +00004117</p>
4118
4119<h5>Semantics:</h5>
4120
4121<p>
Chris Lattnerb54c30f2008-05-20 20:48:21 +00004122If the i1 condition evaluates is 1, the instruction returns the first
John Criswell88190562005-05-16 16:17:45 +00004123value argument; otherwise, it returns the second value argument.
Chris Lattnerb53c28d2004-03-12 05:50:16 +00004124</p>
4125
4126<h5>Example:</h5>
4127
4128<pre>
Reid Spencer36a15422007-01-12 03:35:51 +00004129 %X = select i1 true, i8 17, i8 42 <i>; yields i8:17</i>
Chris Lattnerb53c28d2004-03-12 05:50:16 +00004130</pre>
4131</div>
4132
Robert Bocchinof72fdfe2006-01-15 20:48:27 +00004133
4134<!-- _______________________________________________________________________ -->
4135<div class="doc_subsubsection">
Chris Lattnere23c1392005-05-06 05:47:36 +00004136 <a name="i_call">'<tt>call</tt>' Instruction</a>
4137</div>
4138
Misha Brukman76307852003-11-08 01:05:38 +00004139<div class="doc_text">
Chris Lattnere23c1392005-05-06 05:47:36 +00004140
Chris Lattner2f7c9632001-06-06 20:29:01 +00004141<h5>Syntax:</h5>
Chris Lattnere23c1392005-05-06 05:47:36 +00004142<pre>
Nick Lewyckya9b13d52007-09-08 13:57:50 +00004143 &lt;result&gt; = [tail] call [<a href="#callingconv">cconv</a>] &lt;ty&gt; [&lt;fnty&gt;*] &lt;fnptrval&gt;(&lt;param list&gt;)
Chris Lattnere23c1392005-05-06 05:47:36 +00004144</pre>
4145
Chris Lattner2f7c9632001-06-06 20:29:01 +00004146<h5>Overview:</h5>
Chris Lattnere23c1392005-05-06 05:47:36 +00004147
Misha Brukman76307852003-11-08 01:05:38 +00004148<p>The '<tt>call</tt>' instruction represents a simple function call.</p>
Chris Lattnere23c1392005-05-06 05:47:36 +00004149
Chris Lattner2f7c9632001-06-06 20:29:01 +00004150<h5>Arguments:</h5>
Chris Lattnere23c1392005-05-06 05:47:36 +00004151
Misha Brukman76307852003-11-08 01:05:38 +00004152<p>This instruction requires several arguments:</p>
Chris Lattnere23c1392005-05-06 05:47:36 +00004153
Chris Lattnera8292f32002-05-06 22:08:29 +00004154<ol>
Chris Lattner48b383b02003-11-25 01:02:51 +00004155 <li>
Chris Lattner0132aff2005-05-06 22:57:40 +00004156 <p>The optional "tail" marker indicates whether the callee function accesses
4157 any allocas or varargs in the caller. If the "tail" marker is present, the
Chris Lattnere23c1392005-05-06 05:47:36 +00004158 function call is eligible for tail call optimization. Note that calls may
4159 be marked "tail" even if they do not occur before a <a
4160 href="#i_ret"><tt>ret</tt></a> instruction.
Chris Lattner48b383b02003-11-25 01:02:51 +00004161 </li>
4162 <li>
Duncan Sands16f122e2007-03-30 12:22:09 +00004163 <p>The optional "cconv" marker indicates which <a href="#callingconv">calling
Chris Lattner0132aff2005-05-06 22:57:40 +00004164 convention</a> the call should use. If none is specified, the call defaults
4165 to using C calling conventions.
4166 </li>
4167 <li>
Nick Lewyckya9b13d52007-09-08 13:57:50 +00004168 <p>'<tt>ty</tt>': the type of the call instruction itself which is also
4169 the type of the return value. Functions that return no value are marked
4170 <tt><a href="#t_void">void</a></tt>.</p>
4171 </li>
4172 <li>
4173 <p>'<tt>fnty</tt>': shall be the signature of the pointer to function
4174 value being invoked. The argument types must match the types implied by
4175 this signature. This type can be omitted if the function is not varargs
4176 and if the function type does not return a pointer to a function.</p>
Chris Lattnere23c1392005-05-06 05:47:36 +00004177 </li>
4178 <li>
4179 <p>'<tt>fnptrval</tt>': An LLVM value containing a pointer to a function to
4180 be invoked. In most cases, this is a direct function invocation, but
4181 indirect <tt>call</tt>s are just as possible, calling an arbitrary pointer
John Criswell88190562005-05-16 16:17:45 +00004182 to function value.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +00004183 </li>
4184 <li>
4185 <p>'<tt>function args</tt>': argument list whose types match the
Reid Spencerd845d162005-05-01 22:22:57 +00004186 function signature argument types. All arguments must be of
4187 <a href="#t_firstclass">first class</a> type. If the function signature
4188 indicates the function accepts a variable number of arguments, the extra
4189 arguments can be specified.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +00004190 </li>
Chris Lattnera8292f32002-05-06 22:08:29 +00004191</ol>
Chris Lattnere23c1392005-05-06 05:47:36 +00004192
Chris Lattner2f7c9632001-06-06 20:29:01 +00004193<h5>Semantics:</h5>
Chris Lattnere23c1392005-05-06 05:47:36 +00004194
Chris Lattner48b383b02003-11-25 01:02:51 +00004195<p>The '<tt>call</tt>' instruction is used to cause control flow to
4196transfer to a specified function, with its incoming arguments bound to
4197the specified values. Upon a '<tt><a href="#i_ret">ret</a></tt>'
4198instruction in the called function, control flow continues with the
4199instruction after the function call, and the return value of the
Chris Lattnerfb7c88d2008-03-21 17:24:17 +00004200function is bound to the result argument. If the callee returns multiple
4201values then the return values of the function are only accessible through
4202the '<tt><a href="#i_getresult">getresult</a></tt>' instruction.</p>
Chris Lattnere23c1392005-05-06 05:47:36 +00004203
Chris Lattner2f7c9632001-06-06 20:29:01 +00004204<h5>Example:</h5>
Chris Lattnere23c1392005-05-06 05:47:36 +00004205
4206<pre>
Nick Lewyckya9b13d52007-09-08 13:57:50 +00004207 %retval = call i32 @test(i32 %argc)
Chris Lattnerfb7c88d2008-03-21 17:24:17 +00004208 call i32 (i8 *, ...)* @printf(i8 * %msg, i32 12, i8 42) <i>; yields i32</i>
4209 %X = tail call i32 @foo() <i>; yields i32</i>
4210 %Y = tail call <a href="#callingconv">fastcc</a> i32 @foo() <i>; yields i32</i>
4211 call void %foo(i8 97 signext)
Devang Pateld6cff512008-03-10 20:49:15 +00004212
4213 %struct.A = type { i32, i8 }
Chris Lattnerfb7c88d2008-03-21 17:24:17 +00004214 %r = call %struct.A @foo() <i>; yields { 32, i8 }</i>
4215 %gr = getresult %struct.A %r, 0 <i>; yields i32</i>
4216 %gr1 = getresult %struct.A %r, 1 <i>; yields i8</i>
Chris Lattnere23c1392005-05-06 05:47:36 +00004217</pre>
4218
Misha Brukman76307852003-11-08 01:05:38 +00004219</div>
Chris Lattner6a4a0492004-09-27 21:51:25 +00004220
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00004221<!-- _______________________________________________________________________ -->
Chris Lattner6a4a0492004-09-27 21:51:25 +00004222<div class="doc_subsubsection">
Chris Lattner33337472006-01-13 23:26:01 +00004223 <a name="i_va_arg">'<tt>va_arg</tt>' Instruction</a>
Chris Lattner6a4a0492004-09-27 21:51:25 +00004224</div>
4225
Misha Brukman76307852003-11-08 01:05:38 +00004226<div class="doc_text">
Chris Lattner6a4a0492004-09-27 21:51:25 +00004227
Chris Lattner26ca62e2003-10-18 05:51:36 +00004228<h5>Syntax:</h5>
Chris Lattner6a4a0492004-09-27 21:51:25 +00004229
4230<pre>
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00004231 &lt;resultval&gt; = va_arg &lt;va_list*&gt; &lt;arglist&gt;, &lt;argty&gt;
Chris Lattner6a4a0492004-09-27 21:51:25 +00004232</pre>
4233
Chris Lattner26ca62e2003-10-18 05:51:36 +00004234<h5>Overview:</h5>
Chris Lattner6a4a0492004-09-27 21:51:25 +00004235
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00004236<p>The '<tt>va_arg</tt>' instruction is used to access arguments passed through
Chris Lattner6a4a0492004-09-27 21:51:25 +00004237the "variable argument" area of a function call. It is used to implement the
4238<tt>va_arg</tt> macro in C.</p>
4239
Chris Lattner26ca62e2003-10-18 05:51:36 +00004240<h5>Arguments:</h5>
Chris Lattner6a4a0492004-09-27 21:51:25 +00004241
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00004242<p>This instruction takes a <tt>va_list*</tt> value and the type of
4243the argument. It returns a value of the specified argument type and
Jeff Cohen222a8a42007-04-29 01:07:00 +00004244increments the <tt>va_list</tt> to point to the next argument. The
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00004245actual type of <tt>va_list</tt> is target specific.</p>
Chris Lattner6a4a0492004-09-27 21:51:25 +00004246
Chris Lattner26ca62e2003-10-18 05:51:36 +00004247<h5>Semantics:</h5>
Chris Lattner6a4a0492004-09-27 21:51:25 +00004248
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00004249<p>The '<tt>va_arg</tt>' instruction loads an argument of the specified
4250type from the specified <tt>va_list</tt> and causes the
4251<tt>va_list</tt> to point to the next argument. For more information,
4252see the variable argument handling <a href="#int_varargs">Intrinsic
4253Functions</a>.</p>
Chris Lattner6a4a0492004-09-27 21:51:25 +00004254
4255<p>It is legal for this instruction to be called in a function which does not
4256take a variable number of arguments, for example, the <tt>vfprintf</tt>
Misha Brukman76307852003-11-08 01:05:38 +00004257function.</p>
Chris Lattner6a4a0492004-09-27 21:51:25 +00004258
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00004259<p><tt>va_arg</tt> is an LLVM instruction instead of an <a
John Criswell88190562005-05-16 16:17:45 +00004260href="#intrinsics">intrinsic function</a> because it takes a type as an
Chris Lattner6a4a0492004-09-27 21:51:25 +00004261argument.</p>
4262
Chris Lattner26ca62e2003-10-18 05:51:36 +00004263<h5>Example:</h5>
Chris Lattner6a4a0492004-09-27 21:51:25 +00004264
4265<p>See the <a href="#int_varargs">variable argument processing</a> section.</p>
4266
Misha Brukman76307852003-11-08 01:05:38 +00004267</div>
Chris Lattner941515c2004-01-06 05:31:32 +00004268
Devang Pateld6cff512008-03-10 20:49:15 +00004269<!-- _______________________________________________________________________ -->
4270<div class="doc_subsubsection">
4271 <a name="i_getresult">'<tt>getresult</tt>' Instruction</a>
4272</div>
4273
4274<div class="doc_text">
4275
4276<h5>Syntax:</h5>
4277<pre>
Chris Lattner141b6132008-03-21 17:20:51 +00004278 &lt;resultval&gt; = getresult &lt;type&gt; &lt;retval&gt;, &lt;index&gt;
Devang Pateld6cff512008-03-10 20:49:15 +00004279</pre>
Chris Lattner141b6132008-03-21 17:20:51 +00004280
Devang Pateld6cff512008-03-10 20:49:15 +00004281<h5>Overview:</h5>
4282
4283<p> The '<tt>getresult</tt>' instruction is used to extract individual values
Chris Lattner141b6132008-03-21 17:20:51 +00004284from a '<tt><a href="#i_call">call</a></tt>'
4285or '<tt><a href="#i_invoke">invoke</a></tt>' instruction that returns multiple
4286results.</p>
Devang Pateld6cff512008-03-10 20:49:15 +00004287
4288<h5>Arguments:</h5>
4289
Chris Lattner141b6132008-03-21 17:20:51 +00004290<p>The '<tt>getresult</tt>' instruction takes a call or invoke value as its
Chris Lattner1a640a62008-04-23 04:06:52 +00004291first argument, or an undef value. The value must have <a
4292href="#t_struct">structure type</a>. The second argument is a constant
4293unsigned index value which must be in range for the number of values returned
4294by the call.</p>
Devang Pateld6cff512008-03-10 20:49:15 +00004295
4296<h5>Semantics:</h5>
4297
Chris Lattner141b6132008-03-21 17:20:51 +00004298<p>The '<tt>getresult</tt>' instruction extracts the element identified by
4299'<tt>index</tt>' from the aggregate value.</p>
Devang Pateld6cff512008-03-10 20:49:15 +00004300
4301<h5>Example:</h5>
4302
4303<pre>
4304 %struct.A = type { i32, i8 }
4305
4306 %r = call %struct.A @foo()
Chris Lattner141b6132008-03-21 17:20:51 +00004307 %gr = getresult %struct.A %r, 0 <i>; yields i32:%gr</i>
4308 %gr1 = getresult %struct.A %r, 1 <i>; yields i8:%gr1</i>
Devang Pateld6cff512008-03-10 20:49:15 +00004309 add i32 %gr, 42
4310 add i8 %gr1, 41
4311</pre>
4312
4313</div>
4314
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00004315<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +00004316<div class="doc_section"> <a name="intrinsics">Intrinsic Functions</a> </div>
4317<!-- *********************************************************************** -->
Chris Lattner941515c2004-01-06 05:31:32 +00004318
Misha Brukman76307852003-11-08 01:05:38 +00004319<div class="doc_text">
Chris Lattnerfee11462004-02-12 17:01:32 +00004320
4321<p>LLVM supports the notion of an "intrinsic function". These functions have
Reid Spencer4eefaab2007-04-01 08:04:23 +00004322well known names and semantics and are required to follow certain restrictions.
4323Overall, these intrinsics represent an extension mechanism for the LLVM
Jeff Cohen222a8a42007-04-29 01:07:00 +00004324language that does not require changing all of the transformations in LLVM when
Gabor Greifa54634a2007-07-06 22:07:22 +00004325adding to the language (or the bitcode reader/writer, the parser, etc...).</p>
Chris Lattnerfee11462004-02-12 17:01:32 +00004326
John Criswell88190562005-05-16 16:17:45 +00004327<p>Intrinsic function names must all start with an "<tt>llvm.</tt>" prefix. This
Jeff Cohen222a8a42007-04-29 01:07:00 +00004328prefix is reserved in LLVM for intrinsic names; thus, function names may not
4329begin with this prefix. Intrinsic functions must always be external functions:
4330you cannot define the body of intrinsic functions. Intrinsic functions may
4331only be used in call or invoke instructions: it is illegal to take the address
4332of an intrinsic function. Additionally, because intrinsic functions are part
4333of the LLVM language, it is required if any are added that they be documented
4334here.</p>
Chris Lattnerfee11462004-02-12 17:01:32 +00004335
Chandler Carruth7132e002007-08-04 01:51:18 +00004336<p>Some intrinsic functions can be overloaded, i.e., the intrinsic represents
4337a family of functions that perform the same operation but on different data
4338types. Because LLVM can represent over 8 million different integer types,
4339overloading is used commonly to allow an intrinsic function to operate on any
4340integer type. One or more of the argument types or the result type can be
4341overloaded to accept any integer type. Argument types may also be defined as
4342exactly matching a previous argument's type or the result type. This allows an
4343intrinsic function which accepts multiple arguments, but needs all of them to
4344be of the same type, to only be overloaded with respect to a single argument or
4345the result.</p>
Chris Lattnerfee11462004-02-12 17:01:32 +00004346
Chandler Carruth7132e002007-08-04 01:51:18 +00004347<p>Overloaded intrinsics will have the names of its overloaded argument types
4348encoded into its function name, each preceded by a period. Only those types
4349which are overloaded result in a name suffix. Arguments whose type is matched
4350against another type do not. For example, the <tt>llvm.ctpop</tt> function can
4351take an integer of any width and returns an integer of exactly the same integer
4352width. This leads to a family of functions such as
4353<tt>i8 @llvm.ctpop.i8(i8 %val)</tt> and <tt>i29 @llvm.ctpop.i29(i29 %val)</tt>.
4354Only one type, the return type, is overloaded, and only one type suffix is
4355required. Because the argument's type is matched against the return type, it
4356does not require its own name suffix.</p>
Reid Spencer4eefaab2007-04-01 08:04:23 +00004357
4358<p>To learn how to add an intrinsic function, please see the
4359<a href="ExtendingLLVM.html">Extending LLVM Guide</a>.
Chris Lattnerfee11462004-02-12 17:01:32 +00004360</p>
4361
Misha Brukman76307852003-11-08 01:05:38 +00004362</div>
Chris Lattner941515c2004-01-06 05:31:32 +00004363
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00004364<!-- ======================================================================= -->
Chris Lattner941515c2004-01-06 05:31:32 +00004365<div class="doc_subsection">
4366 <a name="int_varargs">Variable Argument Handling Intrinsics</a>
4367</div>
4368
Misha Brukman76307852003-11-08 01:05:38 +00004369<div class="doc_text">
Chris Lattner757528b0b2004-05-23 21:06:01 +00004370
Misha Brukman76307852003-11-08 01:05:38 +00004371<p>Variable argument support is defined in LLVM with the <a
Chris Lattner33337472006-01-13 23:26:01 +00004372 href="#i_va_arg"><tt>va_arg</tt></a> instruction and these three
Chris Lattner48b383b02003-11-25 01:02:51 +00004373intrinsic functions. These functions are related to the similarly
4374named macros defined in the <tt>&lt;stdarg.h&gt;</tt> header file.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +00004375
Chris Lattner48b383b02003-11-25 01:02:51 +00004376<p>All of these functions operate on arguments that use a
4377target-specific value type "<tt>va_list</tt>". The LLVM assembly
4378language reference manual does not define what this type is, so all
Jeff Cohen222a8a42007-04-29 01:07:00 +00004379transformations should be prepared to handle these functions regardless of
4380the type used.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +00004381
Chris Lattner30b868d2006-05-15 17:26:46 +00004382<p>This example shows how the <a href="#i_va_arg"><tt>va_arg</tt></a>
Chris Lattner48b383b02003-11-25 01:02:51 +00004383instruction and the variable argument handling intrinsic functions are
4384used.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +00004385
Bill Wendling3716c5d2007-05-29 09:04:49 +00004386<div class="doc_code">
Chris Lattnerfee11462004-02-12 17:01:32 +00004387<pre>
Anton Korobeynikov640bbe02007-03-21 23:58:04 +00004388define i32 @test(i32 %X, ...) {
Chris Lattnerfee11462004-02-12 17:01:32 +00004389 ; Initialize variable argument processing
Jeff Cohen222a8a42007-04-29 01:07:00 +00004390 %ap = alloca i8*
Chris Lattnerdb0790c2007-01-08 07:55:15 +00004391 %ap2 = bitcast i8** %ap to i8*
Anton Korobeynikov640bbe02007-03-21 23:58:04 +00004392 call void @llvm.va_start(i8* %ap2)
Chris Lattnerfee11462004-02-12 17:01:32 +00004393
4394 ; Read a single integer argument
Jeff Cohen222a8a42007-04-29 01:07:00 +00004395 %tmp = va_arg i8** %ap, i32
Chris Lattnerfee11462004-02-12 17:01:32 +00004396
4397 ; Demonstrate usage of llvm.va_copy and llvm.va_end
Jeff Cohen222a8a42007-04-29 01:07:00 +00004398 %aq = alloca i8*
Chris Lattnerdb0790c2007-01-08 07:55:15 +00004399 %aq2 = bitcast i8** %aq to i8*
Jeff Cohen222a8a42007-04-29 01:07:00 +00004400 call void @llvm.va_copy(i8* %aq2, i8* %ap2)
Anton Korobeynikov640bbe02007-03-21 23:58:04 +00004401 call void @llvm.va_end(i8* %aq2)
Chris Lattnerfee11462004-02-12 17:01:32 +00004402
4403 ; Stop processing of arguments.
Anton Korobeynikov640bbe02007-03-21 23:58:04 +00004404 call void @llvm.va_end(i8* %ap2)
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00004405 ret i32 %tmp
Chris Lattnerfee11462004-02-12 17:01:32 +00004406}
Anton Korobeynikov640bbe02007-03-21 23:58:04 +00004407
4408declare void @llvm.va_start(i8*)
4409declare void @llvm.va_copy(i8*, i8*)
4410declare void @llvm.va_end(i8*)
Chris Lattnerfee11462004-02-12 17:01:32 +00004411</pre>
Misha Brukman76307852003-11-08 01:05:38 +00004412</div>
Chris Lattner941515c2004-01-06 05:31:32 +00004413
Bill Wendling3716c5d2007-05-29 09:04:49 +00004414</div>
4415
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00004416<!-- _______________________________________________________________________ -->
Chris Lattner941515c2004-01-06 05:31:32 +00004417<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +00004418 <a name="int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a>
Chris Lattner941515c2004-01-06 05:31:32 +00004419</div>
4420
4421
Misha Brukman76307852003-11-08 01:05:38 +00004422<div class="doc_text">
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00004423<h5>Syntax:</h5>
Chris Lattnerdb0790c2007-01-08 07:55:15 +00004424<pre> declare void %llvm.va_start(i8* &lt;arglist&gt;)<br></pre>
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00004425<h5>Overview:</h5>
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00004426<P>The '<tt>llvm.va_start</tt>' intrinsic initializes
4427<tt>*&lt;arglist&gt;</tt> for subsequent use by <tt><a
4428href="#i_va_arg">va_arg</a></tt>.</p>
4429
4430<h5>Arguments:</h5>
4431
4432<P>The argument is a pointer to a <tt>va_list</tt> element to initialize.</p>
4433
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00004434<h5>Semantics:</h5>
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00004435
4436<P>The '<tt>llvm.va_start</tt>' intrinsic works just like the <tt>va_start</tt>
4437macro available in C. In a target-dependent way, it initializes the
Jeff Cohen222a8a42007-04-29 01:07:00 +00004438<tt>va_list</tt> element to which the argument points, so that the next call to
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00004439<tt>va_arg</tt> will produce the first variable argument passed to the function.
4440Unlike the C <tt>va_start</tt> macro, this intrinsic does not need to know the
Jeff Cohen222a8a42007-04-29 01:07:00 +00004441last argument of the function as the compiler can figure that out.</p>
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00004442
Misha Brukman76307852003-11-08 01:05:38 +00004443</div>
Chris Lattner941515c2004-01-06 05:31:32 +00004444
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00004445<!-- _______________________________________________________________________ -->
Chris Lattner941515c2004-01-06 05:31:32 +00004446<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +00004447 <a name="int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a>
Chris Lattner941515c2004-01-06 05:31:32 +00004448</div>
4449
Misha Brukman76307852003-11-08 01:05:38 +00004450<div class="doc_text">
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00004451<h5>Syntax:</h5>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +00004452<pre> declare void @llvm.va_end(i8* &lt;arglist&gt;)<br></pre>
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00004453<h5>Overview:</h5>
Chris Lattnerdb0790c2007-01-08 07:55:15 +00004454
Jeff Cohen222a8a42007-04-29 01:07:00 +00004455<p>The '<tt>llvm.va_end</tt>' intrinsic destroys <tt>*&lt;arglist&gt;</tt>,
Reid Spencer96a5f022007-04-04 02:42:35 +00004456which has been initialized previously with <tt><a href="#int_va_start">llvm.va_start</a></tt>
Chris Lattner48b383b02003-11-25 01:02:51 +00004457or <tt><a href="#i_va_copy">llvm.va_copy</a></tt>.</p>
Chris Lattnerdb0790c2007-01-08 07:55:15 +00004458
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00004459<h5>Arguments:</h5>
Chris Lattnerdb0790c2007-01-08 07:55:15 +00004460
Jeff Cohen222a8a42007-04-29 01:07:00 +00004461<p>The argument is a pointer to a <tt>va_list</tt> to destroy.</p>
Chris Lattnerdb0790c2007-01-08 07:55:15 +00004462
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00004463<h5>Semantics:</h5>
Chris Lattnerdb0790c2007-01-08 07:55:15 +00004464
Misha Brukman76307852003-11-08 01:05:38 +00004465<p>The '<tt>llvm.va_end</tt>' intrinsic works just like the <tt>va_end</tt>
Jeff Cohen222a8a42007-04-29 01:07:00 +00004466macro available in C. In a target-dependent way, it destroys the
4467<tt>va_list</tt> element to which the argument points. Calls to <a
4468href="#int_va_start"><tt>llvm.va_start</tt></a> and <a href="#int_va_copy">
4469<tt>llvm.va_copy</tt></a> must be matched exactly with calls to
4470<tt>llvm.va_end</tt>.</p>
Chris Lattnerdb0790c2007-01-08 07:55:15 +00004471
Misha Brukman76307852003-11-08 01:05:38 +00004472</div>
Chris Lattner941515c2004-01-06 05:31:32 +00004473
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00004474<!-- _______________________________________________________________________ -->
Chris Lattner941515c2004-01-06 05:31:32 +00004475<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +00004476 <a name="int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a>
Chris Lattner941515c2004-01-06 05:31:32 +00004477</div>
4478
Misha Brukman76307852003-11-08 01:05:38 +00004479<div class="doc_text">
Chris Lattner757528b0b2004-05-23 21:06:01 +00004480
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00004481<h5>Syntax:</h5>
Chris Lattner757528b0b2004-05-23 21:06:01 +00004482
4483<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +00004484 declare void @llvm.va_copy(i8* &lt;destarglist&gt;, i8* &lt;srcarglist&gt;)
Chris Lattner757528b0b2004-05-23 21:06:01 +00004485</pre>
4486
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00004487<h5>Overview:</h5>
Chris Lattner757528b0b2004-05-23 21:06:01 +00004488
Jeff Cohen222a8a42007-04-29 01:07:00 +00004489<p>The '<tt>llvm.va_copy</tt>' intrinsic copies the current argument position
4490from the source argument list to the destination argument list.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +00004491
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00004492<h5>Arguments:</h5>
Chris Lattner757528b0b2004-05-23 21:06:01 +00004493
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00004494<p>The first argument is a pointer to a <tt>va_list</tt> element to initialize.
Andrew Lenharth5305ea52005-06-22 20:38:11 +00004495The second argument is a pointer to a <tt>va_list</tt> element to copy from.</p>
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00004496
Chris Lattner757528b0b2004-05-23 21:06:01 +00004497
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00004498<h5>Semantics:</h5>
Chris Lattner757528b0b2004-05-23 21:06:01 +00004499
Jeff Cohen222a8a42007-04-29 01:07:00 +00004500<p>The '<tt>llvm.va_copy</tt>' intrinsic works just like the <tt>va_copy</tt>
4501macro available in C. In a target-dependent way, it copies the source
4502<tt>va_list</tt> element into the destination <tt>va_list</tt> element. This
4503intrinsic is necessary because the <tt><a href="#int_va_start">
4504llvm.va_start</a></tt> intrinsic may be arbitrarily complex and require, for
4505example, memory allocation.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +00004506
Misha Brukman76307852003-11-08 01:05:38 +00004507</div>
Chris Lattner941515c2004-01-06 05:31:32 +00004508
Chris Lattnerfee11462004-02-12 17:01:32 +00004509<!-- ======================================================================= -->
4510<div class="doc_subsection">
Chris Lattner757528b0b2004-05-23 21:06:01 +00004511 <a name="int_gc">Accurate Garbage Collection Intrinsics</a>
4512</div>
4513
4514<div class="doc_text">
4515
4516<p>
4517LLVM support for <a href="GarbageCollection.html">Accurate Garbage
4518Collection</a> requires the implementation and generation of these intrinsics.
Reid Spencer96a5f022007-04-04 02:42:35 +00004519These intrinsics allow identification of <a href="#int_gcroot">GC roots on the
Chris Lattner757528b0b2004-05-23 21:06:01 +00004520stack</a>, as well as garbage collector implementations that require <a
Reid Spencer96a5f022007-04-04 02:42:35 +00004521href="#int_gcread">read</a> and <a href="#int_gcwrite">write</a> barriers.
Chris Lattner757528b0b2004-05-23 21:06:01 +00004522Front-ends for type-safe garbage collected languages should generate these
4523intrinsics to make use of the LLVM garbage collectors. For more details, see <a
4524href="GarbageCollection.html">Accurate Garbage Collection with LLVM</a>.
4525</p>
Christopher Lamb55c6d4f2007-12-17 01:00:21 +00004526
4527<p>The garbage collection intrinsics only operate on objects in the generic
4528 address space (address space zero).</p>
4529
Chris Lattner757528b0b2004-05-23 21:06:01 +00004530</div>
4531
4532<!-- _______________________________________________________________________ -->
4533<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +00004534 <a name="int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a>
Chris Lattner757528b0b2004-05-23 21:06:01 +00004535</div>
4536
4537<div class="doc_text">
4538
4539<h5>Syntax:</h5>
4540
4541<pre>
Chris Lattner12477732007-09-21 17:30:40 +00004542 declare void @llvm.gcroot(i8** %ptrloc, i8* %metadata)
Chris Lattner757528b0b2004-05-23 21:06:01 +00004543</pre>
4544
4545<h5>Overview:</h5>
4546
John Criswelldfe6a862004-12-10 15:51:16 +00004547<p>The '<tt>llvm.gcroot</tt>' intrinsic declares the existence of a GC root to
Chris Lattner757528b0b2004-05-23 21:06:01 +00004548the code generator, and allows some metadata to be associated with it.</p>
4549
4550<h5>Arguments:</h5>
4551
4552<p>The first argument specifies the address of a stack object that contains the
4553root pointer. The second pointer (which must be either a constant or a global
4554value address) contains the meta-data to be associated with the root.</p>
4555
4556<h5>Semantics:</h5>
4557
Chris Lattner851b7712008-04-24 05:59:56 +00004558<p>At runtime, a call to this intrinsic stores a null pointer into the "ptrloc"
Chris Lattner757528b0b2004-05-23 21:06:01 +00004559location. At compile-time, the code generator generates information to allow
Gordon Henriksenfb56bde2007-12-25 02:31:26 +00004560the runtime to find the pointer at GC safe points. The '<tt>llvm.gcroot</tt>'
4561intrinsic may only be used in a function which <a href="#gc">specifies a GC
4562algorithm</a>.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +00004563
4564</div>
4565
4566
4567<!-- _______________________________________________________________________ -->
4568<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +00004569 <a name="int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a>
Chris Lattner757528b0b2004-05-23 21:06:01 +00004570</div>
4571
4572<div class="doc_text">
4573
4574<h5>Syntax:</h5>
4575
4576<pre>
Chris Lattner12477732007-09-21 17:30:40 +00004577 declare i8* @llvm.gcread(i8* %ObjPtr, i8** %Ptr)
Chris Lattner757528b0b2004-05-23 21:06:01 +00004578</pre>
4579
4580<h5>Overview:</h5>
4581
4582<p>The '<tt>llvm.gcread</tt>' intrinsic identifies reads of references from heap
4583locations, allowing garbage collector implementations that require read
4584barriers.</p>
4585
4586<h5>Arguments:</h5>
4587
Chris Lattnerf9228072006-03-14 20:02:51 +00004588<p>The second argument is the address to read from, which should be an address
4589allocated from the garbage collector. The first object is a pointer to the
4590start of the referenced object, if needed by the language runtime (otherwise
4591null).</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +00004592
4593<h5>Semantics:</h5>
4594
4595<p>The '<tt>llvm.gcread</tt>' intrinsic has the same semantics as a load
4596instruction, but may be replaced with substantially more complex code by the
Gordon Henriksenfb56bde2007-12-25 02:31:26 +00004597garbage collector runtime, as needed. The '<tt>llvm.gcread</tt>' intrinsic
4598may only be used in a function which <a href="#gc">specifies a GC
4599algorithm</a>.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +00004600
4601</div>
4602
4603
4604<!-- _______________________________________________________________________ -->
4605<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +00004606 <a name="int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a>
Chris Lattner757528b0b2004-05-23 21:06:01 +00004607</div>
4608
4609<div class="doc_text">
4610
4611<h5>Syntax:</h5>
4612
4613<pre>
Chris Lattner12477732007-09-21 17:30:40 +00004614 declare void @llvm.gcwrite(i8* %P1, i8* %Obj, i8** %P2)
Chris Lattner757528b0b2004-05-23 21:06:01 +00004615</pre>
4616
4617<h5>Overview:</h5>
4618
4619<p>The '<tt>llvm.gcwrite</tt>' intrinsic identifies writes of references to heap
4620locations, allowing garbage collector implementations that require write
4621barriers (such as generational or reference counting collectors).</p>
4622
4623<h5>Arguments:</h5>
4624
Chris Lattnerf9228072006-03-14 20:02:51 +00004625<p>The first argument is the reference to store, the second is the start of the
4626object to store it to, and the third is the address of the field of Obj to
4627store to. If the runtime does not require a pointer to the object, Obj may be
4628null.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +00004629
4630<h5>Semantics:</h5>
4631
4632<p>The '<tt>llvm.gcwrite</tt>' intrinsic has the same semantics as a store
4633instruction, but may be replaced with substantially more complex code by the
Gordon Henriksenfb56bde2007-12-25 02:31:26 +00004634garbage collector runtime, as needed. The '<tt>llvm.gcwrite</tt>' intrinsic
4635may only be used in a function which <a href="#gc">specifies a GC
4636algorithm</a>.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +00004637
4638</div>
4639
4640
4641
4642<!-- ======================================================================= -->
4643<div class="doc_subsection">
Chris Lattner3649c3a2004-02-14 04:08:35 +00004644 <a name="int_codegen">Code Generator Intrinsics</a>
4645</div>
4646
4647<div class="doc_text">
4648<p>
4649These intrinsics are provided by LLVM to expose special features that may only
4650be implemented with code generator support.
4651</p>
4652
4653</div>
4654
4655<!-- _______________________________________________________________________ -->
4656<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +00004657 <a name="int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a>
Chris Lattner3649c3a2004-02-14 04:08:35 +00004658</div>
4659
4660<div class="doc_text">
4661
4662<h5>Syntax:</h5>
4663<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +00004664 declare i8 *@llvm.returnaddress(i32 &lt;level&gt;)
Chris Lattner3649c3a2004-02-14 04:08:35 +00004665</pre>
4666
4667<h5>Overview:</h5>
4668
4669<p>
Chris Lattnerc1fb4262006-10-15 20:05:59 +00004670The '<tt>llvm.returnaddress</tt>' intrinsic attempts to compute a
4671target-specific value indicating the return address of the current function
4672or one of its callers.
Chris Lattner3649c3a2004-02-14 04:08:35 +00004673</p>
4674
4675<h5>Arguments:</h5>
4676
4677<p>
4678The argument to this intrinsic indicates which function to return the address
4679for. Zero indicates the calling function, one indicates its caller, etc. The
4680argument is <b>required</b> to be a constant integer value.
4681</p>
4682
4683<h5>Semantics:</h5>
4684
4685<p>
4686The '<tt>llvm.returnaddress</tt>' intrinsic either returns a pointer indicating
4687the return address of the specified call frame, or zero if it cannot be
4688identified. The value returned by this intrinsic is likely to be incorrect or 0
4689for arguments other than zero, so it should only be used for debugging purposes.
4690</p>
4691
4692<p>
4693Note that calling this intrinsic does not prevent function inlining or other
Chris Lattner2e6eb5f2005-03-07 20:30:51 +00004694aggressive transformations, so the value returned may not be that of the obvious
Chris Lattner3649c3a2004-02-14 04:08:35 +00004695source-language caller.
4696</p>
4697</div>
4698
4699
4700<!-- _______________________________________________________________________ -->
4701<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +00004702 <a name="int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a>
Chris Lattner3649c3a2004-02-14 04:08:35 +00004703</div>
4704
4705<div class="doc_text">
4706
4707<h5>Syntax:</h5>
4708<pre>
Chris Lattner12477732007-09-21 17:30:40 +00004709 declare i8 *@llvm.frameaddress(i32 &lt;level&gt;)
Chris Lattner3649c3a2004-02-14 04:08:35 +00004710</pre>
4711
4712<h5>Overview:</h5>
4713
4714<p>
Chris Lattnerc1fb4262006-10-15 20:05:59 +00004715The '<tt>llvm.frameaddress</tt>' intrinsic attempts to return the
4716target-specific frame pointer value for the specified stack frame.
Chris Lattner3649c3a2004-02-14 04:08:35 +00004717</p>
4718
4719<h5>Arguments:</h5>
4720
4721<p>
4722The argument to this intrinsic indicates which function to return the frame
4723pointer for. Zero indicates the calling function, one indicates its caller,
4724etc. The argument is <b>required</b> to be a constant integer value.
4725</p>
4726
4727<h5>Semantics:</h5>
4728
4729<p>
4730The '<tt>llvm.frameaddress</tt>' intrinsic either returns a pointer indicating
4731the frame address of the specified call frame, or zero if it cannot be
4732identified. The value returned by this intrinsic is likely to be incorrect or 0
4733for arguments other than zero, so it should only be used for debugging purposes.
4734</p>
4735
4736<p>
4737Note that calling this intrinsic does not prevent function inlining or other
Chris Lattner2e6eb5f2005-03-07 20:30:51 +00004738aggressive transformations, so the value returned may not be that of the obvious
Chris Lattner3649c3a2004-02-14 04:08:35 +00004739source-language caller.
4740</p>
4741</div>
4742
Chris Lattnerc8a2c222005-02-28 19:24:19 +00004743<!-- _______________________________________________________________________ -->
4744<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +00004745 <a name="int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a>
Chris Lattner2f0f0012006-01-13 02:03:13 +00004746</div>
4747
4748<div class="doc_text">
4749
4750<h5>Syntax:</h5>
4751<pre>
Chris Lattner12477732007-09-21 17:30:40 +00004752 declare i8 *@llvm.stacksave()
Chris Lattner2f0f0012006-01-13 02:03:13 +00004753</pre>
4754
4755<h5>Overview:</h5>
4756
4757<p>
4758The '<tt>llvm.stacksave</tt>' intrinsic is used to remember the current state of
Reid Spencer96a5f022007-04-04 02:42:35 +00004759the function stack, for use with <a href="#int_stackrestore">
Chris Lattner2f0f0012006-01-13 02:03:13 +00004760<tt>llvm.stackrestore</tt></a>. This is useful for implementing language
4761features like scoped automatic variable sized arrays in C99.
4762</p>
4763
4764<h5>Semantics:</h5>
4765
4766<p>
4767This intrinsic returns a opaque pointer value that can be passed to <a
Reid Spencer96a5f022007-04-04 02:42:35 +00004768href="#int_stackrestore"><tt>llvm.stackrestore</tt></a>. When an
Chris Lattner2f0f0012006-01-13 02:03:13 +00004769<tt>llvm.stackrestore</tt> intrinsic is executed with a value saved from
4770<tt>llvm.stacksave</tt>, it effectively restores the state of the stack to the
4771state it was in when the <tt>llvm.stacksave</tt> intrinsic executed. In
4772practice, this pops any <a href="#i_alloca">alloca</a> blocks from the stack
4773that were allocated after the <tt>llvm.stacksave</tt> was executed.
4774</p>
4775
4776</div>
4777
4778<!-- _______________________________________________________________________ -->
4779<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +00004780 <a name="int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a>
Chris Lattner2f0f0012006-01-13 02:03:13 +00004781</div>
4782
4783<div class="doc_text">
4784
4785<h5>Syntax:</h5>
4786<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +00004787 declare void @llvm.stackrestore(i8 * %ptr)
Chris Lattner2f0f0012006-01-13 02:03:13 +00004788</pre>
4789
4790<h5>Overview:</h5>
4791
4792<p>
4793The '<tt>llvm.stackrestore</tt>' intrinsic is used to restore the state of
4794the function stack to the state it was in when the corresponding <a
Reid Spencer96a5f022007-04-04 02:42:35 +00004795href="#int_stacksave"><tt>llvm.stacksave</tt></a> intrinsic executed. This is
Chris Lattner2f0f0012006-01-13 02:03:13 +00004796useful for implementing language features like scoped automatic variable sized
4797arrays in C99.
4798</p>
4799
4800<h5>Semantics:</h5>
4801
4802<p>
Reid Spencer96a5f022007-04-04 02:42:35 +00004803See the description for <a href="#int_stacksave"><tt>llvm.stacksave</tt></a>.
Chris Lattner2f0f0012006-01-13 02:03:13 +00004804</p>
4805
4806</div>
4807
4808
4809<!-- _______________________________________________________________________ -->
4810<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +00004811 <a name="int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a>
Chris Lattnerc8a2c222005-02-28 19:24:19 +00004812</div>
4813
4814<div class="doc_text">
4815
4816<h5>Syntax:</h5>
4817<pre>
Chris Lattner12477732007-09-21 17:30:40 +00004818 declare void @llvm.prefetch(i8* &lt;address&gt;, i32 &lt;rw&gt;, i32 &lt;locality&gt;)
Chris Lattnerc8a2c222005-02-28 19:24:19 +00004819</pre>
4820
4821<h5>Overview:</h5>
4822
4823
4824<p>
4825The '<tt>llvm.prefetch</tt>' intrinsic is a hint to the code generator to insert
John Criswell88190562005-05-16 16:17:45 +00004826a prefetch instruction if supported; otherwise, it is a noop. Prefetches have
4827no
4828effect on the behavior of the program but can change its performance
Chris Lattnerff851072005-02-28 19:47:14 +00004829characteristics.
Chris Lattnerc8a2c222005-02-28 19:24:19 +00004830</p>
4831
4832<h5>Arguments:</h5>
4833
4834<p>
4835<tt>address</tt> is the address to be prefetched, <tt>rw</tt> is the specifier
4836determining if the fetch should be for a read (0) or write (1), and
4837<tt>locality</tt> is a temporal locality specifier ranging from (0) - no
Chris Lattnerd3e641c2005-03-07 20:31:38 +00004838locality, to (3) - extremely local keep in cache. The <tt>rw</tt> and
Chris Lattnerc8a2c222005-02-28 19:24:19 +00004839<tt>locality</tt> arguments must be constant integers.
4840</p>
4841
4842<h5>Semantics:</h5>
4843
4844<p>
4845This intrinsic does not modify the behavior of the program. In particular,
4846prefetches cannot trap and do not produce a value. On targets that support this
4847intrinsic, the prefetch can provide hints to the processor cache for better
4848performance.
4849</p>
4850
4851</div>
4852
Andrew Lenharthb4427912005-03-28 20:05:49 +00004853<!-- _______________________________________________________________________ -->
4854<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +00004855 <a name="int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a>
Andrew Lenharthb4427912005-03-28 20:05:49 +00004856</div>
4857
4858<div class="doc_text">
4859
4860<h5>Syntax:</h5>
4861<pre>
Chris Lattner12477732007-09-21 17:30:40 +00004862 declare void @llvm.pcmarker(i32 &lt;id&gt;)
Andrew Lenharthb4427912005-03-28 20:05:49 +00004863</pre>
4864
4865<h5>Overview:</h5>
4866
4867
4868<p>
John Criswell88190562005-05-16 16:17:45 +00004869The '<tt>llvm.pcmarker</tt>' intrinsic is a method to export a Program Counter
4870(PC) in a region of
Andrew Lenharthb4427912005-03-28 20:05:49 +00004871code to simulators and other tools. The method is target specific, but it is
4872expected that the marker will use exported symbols to transmit the PC of the marker.
Jeff Cohendc6bfea2005-11-11 02:15:27 +00004873The marker makes no guarantees that it will remain with any specific instruction
Chris Lattnere64d41d2005-11-15 06:07:55 +00004874after optimizations. It is possible that the presence of a marker will inhibit
Chris Lattnerb40261e2006-03-24 07:16:10 +00004875optimizations. The intended use is to be inserted after optimizations to allow
John Criswell88190562005-05-16 16:17:45 +00004876correlations of simulation runs.
Andrew Lenharthb4427912005-03-28 20:05:49 +00004877</p>
4878
4879<h5>Arguments:</h5>
4880
4881<p>
4882<tt>id</tt> is a numerical id identifying the marker.
4883</p>
4884
4885<h5>Semantics:</h5>
4886
4887<p>
4888This intrinsic does not modify the behavior of the program. Backends that do not
4889support this intrinisic may ignore it.
4890</p>
4891
4892</div>
4893
Andrew Lenharth01aa5632005-11-11 16:47:30 +00004894<!-- _______________________________________________________________________ -->
4895<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +00004896 <a name="int_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a>
Andrew Lenharth01aa5632005-11-11 16:47:30 +00004897</div>
4898
4899<div class="doc_text">
4900
4901<h5>Syntax:</h5>
4902<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +00004903 declare i64 @llvm.readcyclecounter( )
Andrew Lenharth01aa5632005-11-11 16:47:30 +00004904</pre>
4905
4906<h5>Overview:</h5>
4907
4908
4909<p>
4910The '<tt>llvm.readcyclecounter</tt>' intrinsic provides access to the cycle
4911counter register (or similar low latency, high accuracy clocks) on those targets
4912that support it. On X86, it should map to RDTSC. On Alpha, it should map to RPCC.
4913As the backing counters overflow quickly (on the order of 9 seconds on alpha), this
4914should only be used for small timings.
4915</p>
4916
4917<h5>Semantics:</h5>
4918
4919<p>
4920When directly supported, reading the cycle counter should not modify any memory.
4921Implementations are allowed to either return a application specific value or a
4922system wide value. On backends without support, this is lowered to a constant 0.
4923</p>
4924
4925</div>
4926
Chris Lattner3649c3a2004-02-14 04:08:35 +00004927<!-- ======================================================================= -->
4928<div class="doc_subsection">
Chris Lattnerfee11462004-02-12 17:01:32 +00004929 <a name="int_libc">Standard C Library Intrinsics</a>
4930</div>
4931
4932<div class="doc_text">
4933<p>
Chris Lattner3649c3a2004-02-14 04:08:35 +00004934LLVM provides intrinsics for a few important standard C library functions.
4935These intrinsics allow source-language front-ends to pass information about the
4936alignment of the pointer arguments to the code generator, providing opportunity
4937for more efficient code generation.
Chris Lattnerfee11462004-02-12 17:01:32 +00004938</p>
4939
4940</div>
4941
4942<!-- _______________________________________________________________________ -->
4943<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +00004944 <a name="int_memcpy">'<tt>llvm.memcpy</tt>' Intrinsic</a>
Chris Lattnerfee11462004-02-12 17:01:32 +00004945</div>
4946
4947<div class="doc_text">
4948
4949<h5>Syntax:</h5>
4950<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +00004951 declare void @llvm.memcpy.i32(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00004952 i32 &lt;len&gt;, i32 &lt;align&gt;)
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +00004953 declare void @llvm.memcpy.i64(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00004954 i64 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattnerfee11462004-02-12 17:01:32 +00004955</pre>
4956
4957<h5>Overview:</h5>
4958
4959<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +00004960The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
Chris Lattnerfee11462004-02-12 17:01:32 +00004961location to the destination location.
4962</p>
4963
4964<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +00004965Note that, unlike the standard libc function, the <tt>llvm.memcpy.*</tt>
4966intrinsics do not return a value, and takes an extra alignment argument.
Chris Lattnerfee11462004-02-12 17:01:32 +00004967</p>
4968
4969<h5>Arguments:</h5>
4970
4971<p>
4972The first argument is a pointer to the destination, the second is a pointer to
Chris Lattner0c8b2592006-03-03 00:07:20 +00004973the source. The third argument is an integer argument
Chris Lattnerfee11462004-02-12 17:01:32 +00004974specifying the number of bytes to copy, and the fourth argument is the alignment
4975of the source and destination locations.
4976</p>
4977
Chris Lattner4c67c482004-02-12 21:18:15 +00004978<p>
4979If the call to this intrinisic has an alignment value that is not 0 or 1, then
Chris Lattner5316e5d2006-03-04 00:02:10 +00004980the caller guarantees that both the source and destination pointers are aligned
4981to that boundary.
Chris Lattner4c67c482004-02-12 21:18:15 +00004982</p>
4983
Chris Lattnerfee11462004-02-12 17:01:32 +00004984<h5>Semantics:</h5>
4985
4986<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +00004987The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
Chris Lattnerfee11462004-02-12 17:01:32 +00004988location to the destination location, which are not allowed to overlap. It
4989copies "len" bytes of memory over. If the argument is known to be aligned to
4990some boundary, this can be specified as the fourth argument, otherwise it should
4991be set to 0 or 1.
4992</p>
4993</div>
4994
4995
Chris Lattnerf30152e2004-02-12 18:10:10 +00004996<!-- _______________________________________________________________________ -->
4997<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +00004998 <a name="int_memmove">'<tt>llvm.memmove</tt>' Intrinsic</a>
Chris Lattnerf30152e2004-02-12 18:10:10 +00004999</div>
5000
5001<div class="doc_text">
5002
5003<h5>Syntax:</h5>
5004<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +00005005 declare void @llvm.memmove.i32(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00005006 i32 &lt;len&gt;, i32 &lt;align&gt;)
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +00005007 declare void @llvm.memmove.i64(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00005008 i64 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattnerf30152e2004-02-12 18:10:10 +00005009</pre>
5010
5011<h5>Overview:</h5>
5012
5013<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +00005014The '<tt>llvm.memmove.*</tt>' intrinsics move a block of memory from the source
5015location to the destination location. It is similar to the
Chris Lattnerec564022008-01-06 19:51:52 +00005016'<tt>llvm.memcpy</tt>' intrinsic but allows the two memory locations to overlap.
Chris Lattnerf30152e2004-02-12 18:10:10 +00005017</p>
5018
5019<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +00005020Note that, unlike the standard libc function, the <tt>llvm.memmove.*</tt>
5021intrinsics do not return a value, and takes an extra alignment argument.
Chris Lattnerf30152e2004-02-12 18:10:10 +00005022</p>
5023
5024<h5>Arguments:</h5>
5025
5026<p>
5027The first argument is a pointer to the destination, the second is a pointer to
Chris Lattner0c8b2592006-03-03 00:07:20 +00005028the source. The third argument is an integer argument
Chris Lattnerf30152e2004-02-12 18:10:10 +00005029specifying the number of bytes to copy, and the fourth argument is the alignment
5030of the source and destination locations.
5031</p>
5032
Chris Lattner4c67c482004-02-12 21:18:15 +00005033<p>
5034If the call to this intrinisic has an alignment value that is not 0 or 1, then
Chris Lattner5316e5d2006-03-04 00:02:10 +00005035the caller guarantees that the source and destination pointers are aligned to
5036that boundary.
Chris Lattner4c67c482004-02-12 21:18:15 +00005037</p>
5038
Chris Lattnerf30152e2004-02-12 18:10:10 +00005039<h5>Semantics:</h5>
5040
5041<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +00005042The '<tt>llvm.memmove.*</tt>' intrinsics copy a block of memory from the source
Chris Lattnerf30152e2004-02-12 18:10:10 +00005043location to the destination location, which may overlap. It
5044copies "len" bytes of memory over. If the argument is known to be aligned to
5045some boundary, this can be specified as the fourth argument, otherwise it should
5046be set to 0 or 1.
5047</p>
5048</div>
5049
Chris Lattner941515c2004-01-06 05:31:32 +00005050
Chris Lattner3649c3a2004-02-14 04:08:35 +00005051<!-- _______________________________________________________________________ -->
5052<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +00005053 <a name="int_memset">'<tt>llvm.memset.*</tt>' Intrinsics</a>
Chris Lattner3649c3a2004-02-14 04:08:35 +00005054</div>
5055
5056<div class="doc_text">
5057
5058<h5>Syntax:</h5>
5059<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +00005060 declare void @llvm.memset.i32(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00005061 i32 &lt;len&gt;, i32 &lt;align&gt;)
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +00005062 declare void @llvm.memset.i64(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00005063 i64 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattner3649c3a2004-02-14 04:08:35 +00005064</pre>
5065
5066<h5>Overview:</h5>
5067
5068<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +00005069The '<tt>llvm.memset.*</tt>' intrinsics fill a block of memory with a particular
Chris Lattner3649c3a2004-02-14 04:08:35 +00005070byte value.
5071</p>
5072
5073<p>
5074Note that, unlike the standard libc function, the <tt>llvm.memset</tt> intrinsic
5075does not return a value, and takes an extra alignment argument.
5076</p>
5077
5078<h5>Arguments:</h5>
5079
5080<p>
5081The first argument is a pointer to the destination to fill, the second is the
Chris Lattner0c8b2592006-03-03 00:07:20 +00005082byte value to fill it with, the third argument is an integer
Chris Lattner3649c3a2004-02-14 04:08:35 +00005083argument specifying the number of bytes to fill, and the fourth argument is the
5084known alignment of destination location.
5085</p>
5086
5087<p>
5088If the call to this intrinisic has an alignment value that is not 0 or 1, then
Chris Lattner5316e5d2006-03-04 00:02:10 +00005089the caller guarantees that the destination pointer is aligned to that boundary.
Chris Lattner3649c3a2004-02-14 04:08:35 +00005090</p>
5091
5092<h5>Semantics:</h5>
5093
5094<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +00005095The '<tt>llvm.memset.*</tt>' intrinsics fill "len" bytes of memory starting at
5096the
Chris Lattner3649c3a2004-02-14 04:08:35 +00005097destination location. If the argument is known to be aligned to some boundary,
5098this can be specified as the fourth argument, otherwise it should be set to 0 or
50991.
5100</p>
5101</div>
5102
5103
Chris Lattner3b4f4372004-06-11 02:28:03 +00005104<!-- _______________________________________________________________________ -->
5105<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +00005106 <a name="int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a>
Chris Lattner8a8f2e52005-07-21 01:29:16 +00005107</div>
5108
5109<div class="doc_text">
5110
5111<h5>Syntax:</h5>
Dale Johannesendd89d272007-10-02 17:47:38 +00005112<p>This is an overloaded intrinsic. You can use <tt>llvm.sqrt</tt> on any
Dan Gohmanb6324c12007-10-15 20:30:11 +00005113floating point or vector of floating point type. Not all targets support all
5114types however.
Chris Lattner8a8f2e52005-07-21 01:29:16 +00005115<pre>
Dale Johannesendd89d272007-10-02 17:47:38 +00005116 declare float @llvm.sqrt.f32(float %Val)
5117 declare double @llvm.sqrt.f64(double %Val)
5118 declare x86_fp80 @llvm.sqrt.f80(x86_fp80 %Val)
5119 declare fp128 @llvm.sqrt.f128(fp128 %Val)
5120 declare ppc_fp128 @llvm.sqrt.ppcf128(ppc_fp128 %Val)
Chris Lattner8a8f2e52005-07-21 01:29:16 +00005121</pre>
5122
5123<h5>Overview:</h5>
5124
5125<p>
Reid Spencerb4f9a6f2006-01-16 21:12:35 +00005126The '<tt>llvm.sqrt</tt>' intrinsics return the sqrt of the specified operand,
Dan Gohmanb6324c12007-10-15 20:30:11 +00005127returning the same value as the libm '<tt>sqrt</tt>' functions would. Unlike
Chris Lattner8a8f2e52005-07-21 01:29:16 +00005128<tt>sqrt</tt> in libm, however, <tt>llvm.sqrt</tt> has undefined behavior for
Chris Lattner00d7cb92008-01-29 07:00:44 +00005129negative numbers other than -0.0 (which allows for better optimization, because
5130there is no need to worry about errno being set). <tt>llvm.sqrt(-0.0)</tt> is
5131defined to return -0.0 like IEEE sqrt.
Chris Lattner8a8f2e52005-07-21 01:29:16 +00005132</p>
5133
5134<h5>Arguments:</h5>
5135
5136<p>
5137The argument and return value are floating point numbers of the same type.
5138</p>
5139
5140<h5>Semantics:</h5>
5141
5142<p>
Dan Gohman33988db2007-07-16 14:37:41 +00005143This function returns the sqrt of the specified operand if it is a nonnegative
Chris Lattner8a8f2e52005-07-21 01:29:16 +00005144floating point number.
5145</p>
5146</div>
5147
Chris Lattner33b73f92006-09-08 06:34:02 +00005148<!-- _______________________________________________________________________ -->
5149<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +00005150 <a name="int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a>
Chris Lattner33b73f92006-09-08 06:34:02 +00005151</div>
5152
5153<div class="doc_text">
5154
5155<h5>Syntax:</h5>
Dale Johannesendd89d272007-10-02 17:47:38 +00005156<p>This is an overloaded intrinsic. You can use <tt>llvm.powi</tt> on any
Dan Gohmanb6324c12007-10-15 20:30:11 +00005157floating point or vector of floating point type. Not all targets support all
5158types however.
Chris Lattner33b73f92006-09-08 06:34:02 +00005159<pre>
Dale Johannesendd89d272007-10-02 17:47:38 +00005160 declare float @llvm.powi.f32(float %Val, i32 %power)
5161 declare double @llvm.powi.f64(double %Val, i32 %power)
5162 declare x86_fp80 @llvm.powi.f80(x86_fp80 %Val, i32 %power)
5163 declare fp128 @llvm.powi.f128(fp128 %Val, i32 %power)
5164 declare ppc_fp128 @llvm.powi.ppcf128(ppc_fp128 %Val, i32 %power)
Chris Lattner33b73f92006-09-08 06:34:02 +00005165</pre>
5166
5167<h5>Overview:</h5>
5168
5169<p>
5170The '<tt>llvm.powi.*</tt>' intrinsics return the first operand raised to the
5171specified (positive or negative) power. The order of evaluation of
Dan Gohmanb6324c12007-10-15 20:30:11 +00005172multiplications is not defined. When a vector of floating point type is
5173used, the second argument remains a scalar integer value.
Chris Lattner33b73f92006-09-08 06:34:02 +00005174</p>
5175
5176<h5>Arguments:</h5>
5177
5178<p>
5179The second argument is an integer power, and the first is a value to raise to
5180that power.
5181</p>
5182
5183<h5>Semantics:</h5>
5184
5185<p>
5186This function returns the first value raised to the second power with an
5187unspecified sequence of rounding operations.</p>
5188</div>
5189
Dan Gohmanb6324c12007-10-15 20:30:11 +00005190<!-- _______________________________________________________________________ -->
5191<div class="doc_subsubsection">
5192 <a name="int_sin">'<tt>llvm.sin.*</tt>' Intrinsic</a>
5193</div>
5194
5195<div class="doc_text">
5196
5197<h5>Syntax:</h5>
5198<p>This is an overloaded intrinsic. You can use <tt>llvm.sin</tt> on any
5199floating point or vector of floating point type. Not all targets support all
5200types however.
5201<pre>
5202 declare float @llvm.sin.f32(float %Val)
5203 declare double @llvm.sin.f64(double %Val)
5204 declare x86_fp80 @llvm.sin.f80(x86_fp80 %Val)
5205 declare fp128 @llvm.sin.f128(fp128 %Val)
5206 declare ppc_fp128 @llvm.sin.ppcf128(ppc_fp128 %Val)
5207</pre>
5208
5209<h5>Overview:</h5>
5210
5211<p>
5212The '<tt>llvm.sin.*</tt>' intrinsics return the sine of the operand.
5213</p>
5214
5215<h5>Arguments:</h5>
5216
5217<p>
5218The argument and return value are floating point numbers of the same type.
5219</p>
5220
5221<h5>Semantics:</h5>
5222
5223<p>
5224This function returns the sine of the specified operand, returning the
5225same values as the libm <tt>sin</tt> functions would, and handles error
Dan Gohmand0806a02007-10-17 18:05:13 +00005226conditions in the same way.</p>
Dan Gohmanb6324c12007-10-15 20:30:11 +00005227</div>
5228
5229<!-- _______________________________________________________________________ -->
5230<div class="doc_subsubsection">
5231 <a name="int_cos">'<tt>llvm.cos.*</tt>' Intrinsic</a>
5232</div>
5233
5234<div class="doc_text">
5235
5236<h5>Syntax:</h5>
5237<p>This is an overloaded intrinsic. You can use <tt>llvm.cos</tt> on any
5238floating point or vector of floating point type. Not all targets support all
5239types however.
5240<pre>
5241 declare float @llvm.cos.f32(float %Val)
5242 declare double @llvm.cos.f64(double %Val)
5243 declare x86_fp80 @llvm.cos.f80(x86_fp80 %Val)
5244 declare fp128 @llvm.cos.f128(fp128 %Val)
5245 declare ppc_fp128 @llvm.cos.ppcf128(ppc_fp128 %Val)
5246</pre>
5247
5248<h5>Overview:</h5>
5249
5250<p>
5251The '<tt>llvm.cos.*</tt>' intrinsics return the cosine of the operand.
5252</p>
5253
5254<h5>Arguments:</h5>
5255
5256<p>
5257The argument and return value are floating point numbers of the same type.
5258</p>
5259
5260<h5>Semantics:</h5>
5261
5262<p>
5263This function returns the cosine of the specified operand, returning the
5264same values as the libm <tt>cos</tt> functions would, and handles error
Dan Gohmand0806a02007-10-17 18:05:13 +00005265conditions in the same way.</p>
Dan Gohmanb6324c12007-10-15 20:30:11 +00005266</div>
5267
5268<!-- _______________________________________________________________________ -->
5269<div class="doc_subsubsection">
5270 <a name="int_pow">'<tt>llvm.pow.*</tt>' Intrinsic</a>
5271</div>
5272
5273<div class="doc_text">
5274
5275<h5>Syntax:</h5>
5276<p>This is an overloaded intrinsic. You can use <tt>llvm.pow</tt> on any
5277floating point or vector of floating point type. Not all targets support all
5278types however.
5279<pre>
5280 declare float @llvm.pow.f32(float %Val, float %Power)
5281 declare double @llvm.pow.f64(double %Val, double %Power)
5282 declare x86_fp80 @llvm.pow.f80(x86_fp80 %Val, x86_fp80 %Power)
5283 declare fp128 @llvm.pow.f128(fp128 %Val, fp128 %Power)
5284 declare ppc_fp128 @llvm.pow.ppcf128(ppc_fp128 %Val, ppc_fp128 Power)
5285</pre>
5286
5287<h5>Overview:</h5>
5288
5289<p>
5290The '<tt>llvm.pow.*</tt>' intrinsics return the first operand raised to the
5291specified (positive or negative) power.
5292</p>
5293
5294<h5>Arguments:</h5>
5295
5296<p>
5297The second argument is a floating point power, and the first is a value to
5298raise to that power.
5299</p>
5300
5301<h5>Semantics:</h5>
5302
5303<p>
5304This function returns the first value raised to the second power,
5305returning the
5306same values as the libm <tt>pow</tt> functions would, and handles error
Dan Gohmand0806a02007-10-17 18:05:13 +00005307conditions in the same way.</p>
Dan Gohmanb6324c12007-10-15 20:30:11 +00005308</div>
5309
Chris Lattner33b73f92006-09-08 06:34:02 +00005310
Andrew Lenharth1d463522005-05-03 18:01:48 +00005311<!-- ======================================================================= -->
5312<div class="doc_subsection">
Nate Begeman0f223bb2006-01-13 23:26:38 +00005313 <a name="int_manip">Bit Manipulation Intrinsics</a>
Andrew Lenharth1d463522005-05-03 18:01:48 +00005314</div>
5315
5316<div class="doc_text">
5317<p>
Nate Begeman0f223bb2006-01-13 23:26:38 +00005318LLVM provides intrinsics for a few important bit manipulation operations.
Andrew Lenharth1d463522005-05-03 18:01:48 +00005319These allow efficient code generation for some algorithms.
5320</p>
5321
5322</div>
5323
5324<!-- _______________________________________________________________________ -->
5325<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +00005326 <a name="int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a>
Nate Begeman0f223bb2006-01-13 23:26:38 +00005327</div>
5328
5329<div class="doc_text">
5330
5331<h5>Syntax:</h5>
Reid Spencer4eefaab2007-04-01 08:04:23 +00005332<p>This is an overloaded intrinsic function. You can use bswap on any integer
Chandler Carruth7132e002007-08-04 01:51:18 +00005333type that is an even number of bytes (i.e. BitWidth % 16 == 0).
Nate Begeman0f223bb2006-01-13 23:26:38 +00005334<pre>
Chandler Carruth7132e002007-08-04 01:51:18 +00005335 declare i16 @llvm.bswap.i16(i16 &lt;id&gt;)
5336 declare i32 @llvm.bswap.i32(i32 &lt;id&gt;)
5337 declare i64 @llvm.bswap.i64(i64 &lt;id&gt;)
Nate Begeman0f223bb2006-01-13 23:26:38 +00005338</pre>
5339
5340<h5>Overview:</h5>
5341
5342<p>
Reid Spencerf361c4f2007-04-02 02:25:19 +00005343The '<tt>llvm.bswap</tt>' family of intrinsics is used to byte swap integer
Reid Spencer4eefaab2007-04-01 08:04:23 +00005344values with an even number of bytes (positive multiple of 16 bits). These are
5345useful for performing operations on data that is not in the target's native
5346byte order.
Nate Begeman0f223bb2006-01-13 23:26:38 +00005347</p>
5348
5349<h5>Semantics:</h5>
5350
5351<p>
Chandler Carruth7132e002007-08-04 01:51:18 +00005352The <tt>llvm.bswap.i16</tt> intrinsic returns an i16 value that has the high
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00005353and low byte of the input i16 swapped. Similarly, the <tt>llvm.bswap.i32</tt>
5354intrinsic returns an i32 value that has the four bytes of the input i32
5355swapped, so that if the input bytes are numbered 0, 1, 2, 3 then the returned
Chandler Carruth7132e002007-08-04 01:51:18 +00005356i32 will have its bytes in 3, 2, 1, 0 order. The <tt>llvm.bswap.i48</tt>,
5357<tt>llvm.bswap.i64</tt> and other intrinsics extend this concept to
Reid Spencer4eefaab2007-04-01 08:04:23 +00005358additional even-byte lengths (6 bytes, 8 bytes and more, respectively).
Nate Begeman0f223bb2006-01-13 23:26:38 +00005359</p>
5360
5361</div>
5362
5363<!-- _______________________________________________________________________ -->
5364<div class="doc_subsubsection">
Reid Spencerb4f9a6f2006-01-16 21:12:35 +00005365 <a name="int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic</a>
Andrew Lenharth1d463522005-05-03 18:01:48 +00005366</div>
5367
5368<div class="doc_text">
5369
5370<h5>Syntax:</h5>
Reid Spencer4eefaab2007-04-01 08:04:23 +00005371<p>This is an overloaded intrinsic. You can use llvm.ctpop on any integer bit
5372width. Not all targets support all bit widths however.
Andrew Lenharth1d463522005-05-03 18:01:48 +00005373<pre>
Chandler Carruth7132e002007-08-04 01:51:18 +00005374 declare i8 @llvm.ctpop.i8 (i8 &lt;src&gt;)
5375 declare i16 @llvm.ctpop.i16(i16 &lt;src&gt;)
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +00005376 declare i32 @llvm.ctpop.i32(i32 &lt;src&gt;)
Chandler Carruth7132e002007-08-04 01:51:18 +00005377 declare i64 @llvm.ctpop.i64(i64 &lt;src&gt;)
5378 declare i256 @llvm.ctpop.i256(i256 &lt;src&gt;)
Andrew Lenharth1d463522005-05-03 18:01:48 +00005379</pre>
5380
5381<h5>Overview:</h5>
5382
5383<p>
Chris Lattner069b5bd2006-01-16 22:38:59 +00005384The '<tt>llvm.ctpop</tt>' family of intrinsics counts the number of bits set in a
5385value.
Andrew Lenharth1d463522005-05-03 18:01:48 +00005386</p>
5387
5388<h5>Arguments:</h5>
5389
5390<p>
Chris Lattner573f64e2005-05-07 01:46:40 +00005391The only argument is the value to be counted. The argument may be of any
Reid Spencer3e628eb92007-01-04 16:43:23 +00005392integer type. The return type must match the argument type.
Andrew Lenharth1d463522005-05-03 18:01:48 +00005393</p>
5394
5395<h5>Semantics:</h5>
5396
5397<p>
5398The '<tt>llvm.ctpop</tt>' intrinsic counts the 1's in a variable.
5399</p>
5400</div>
5401
5402<!-- _______________________________________________________________________ -->
5403<div class="doc_subsubsection">
Chris Lattnerb748c672006-01-16 22:34:14 +00005404 <a name="int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic</a>
Andrew Lenharth1d463522005-05-03 18:01:48 +00005405</div>
5406
5407<div class="doc_text">
5408
5409<h5>Syntax:</h5>
Reid Spencer4eefaab2007-04-01 08:04:23 +00005410<p>This is an overloaded intrinsic. You can use <tt>llvm.ctlz</tt> on any
5411integer bit width. Not all targets support all bit widths however.
Andrew Lenharth1d463522005-05-03 18:01:48 +00005412<pre>
Chandler Carruth7132e002007-08-04 01:51:18 +00005413 declare i8 @llvm.ctlz.i8 (i8 &lt;src&gt;)
5414 declare i16 @llvm.ctlz.i16(i16 &lt;src&gt;)
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +00005415 declare i32 @llvm.ctlz.i32(i32 &lt;src&gt;)
Chandler Carruth7132e002007-08-04 01:51:18 +00005416 declare i64 @llvm.ctlz.i64(i64 &lt;src&gt;)
5417 declare i256 @llvm.ctlz.i256(i256 &lt;src&gt;)
Andrew Lenharth1d463522005-05-03 18:01:48 +00005418</pre>
5419
5420<h5>Overview:</h5>
5421
5422<p>
Reid Spencerb4f9a6f2006-01-16 21:12:35 +00005423The '<tt>llvm.ctlz</tt>' family of intrinsic functions counts the number of
5424leading zeros in a variable.
Andrew Lenharth1d463522005-05-03 18:01:48 +00005425</p>
5426
5427<h5>Arguments:</h5>
5428
5429<p>
Chris Lattner573f64e2005-05-07 01:46:40 +00005430The only argument is the value to be counted. The argument may be of any
Reid Spencer3e628eb92007-01-04 16:43:23 +00005431integer type. The return type must match the argument type.
Andrew Lenharth1d463522005-05-03 18:01:48 +00005432</p>
5433
5434<h5>Semantics:</h5>
5435
5436<p>
Chris Lattnerefa20fa2005-05-15 19:39:26 +00005437The '<tt>llvm.ctlz</tt>' intrinsic counts the leading (most significant) zeros
5438in a variable. If the src == 0 then the result is the size in bits of the type
Reid Spencerb5ebf3d2006-12-31 07:07:53 +00005439of src. For example, <tt>llvm.ctlz(i32 2) = 30</tt>.
Andrew Lenharth1d463522005-05-03 18:01:48 +00005440</p>
5441</div>
Chris Lattner3b4f4372004-06-11 02:28:03 +00005442
5443
Chris Lattnerefa20fa2005-05-15 19:39:26 +00005444
5445<!-- _______________________________________________________________________ -->
5446<div class="doc_subsubsection">
Chris Lattnerb748c672006-01-16 22:34:14 +00005447 <a name="int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic</a>
Chris Lattnerefa20fa2005-05-15 19:39:26 +00005448</div>
5449
5450<div class="doc_text">
5451
5452<h5>Syntax:</h5>
Reid Spencer4eefaab2007-04-01 08:04:23 +00005453<p>This is an overloaded intrinsic. You can use <tt>llvm.cttz</tt> on any
5454integer bit width. Not all targets support all bit widths however.
Chris Lattnerefa20fa2005-05-15 19:39:26 +00005455<pre>
Chandler Carruth7132e002007-08-04 01:51:18 +00005456 declare i8 @llvm.cttz.i8 (i8 &lt;src&gt;)
5457 declare i16 @llvm.cttz.i16(i16 &lt;src&gt;)
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +00005458 declare i32 @llvm.cttz.i32(i32 &lt;src&gt;)
Chandler Carruth7132e002007-08-04 01:51:18 +00005459 declare i64 @llvm.cttz.i64(i64 &lt;src&gt;)
5460 declare i256 @llvm.cttz.i256(i256 &lt;src&gt;)
Chris Lattnerefa20fa2005-05-15 19:39:26 +00005461</pre>
5462
5463<h5>Overview:</h5>
5464
5465<p>
Reid Spencerb4f9a6f2006-01-16 21:12:35 +00005466The '<tt>llvm.cttz</tt>' family of intrinsic functions counts the number of
5467trailing zeros.
Chris Lattnerefa20fa2005-05-15 19:39:26 +00005468</p>
5469
5470<h5>Arguments:</h5>
5471
5472<p>
5473The only argument is the value to be counted. The argument may be of any
Reid Spencer3e628eb92007-01-04 16:43:23 +00005474integer type. The return type must match the argument type.
Chris Lattnerefa20fa2005-05-15 19:39:26 +00005475</p>
5476
5477<h5>Semantics:</h5>
5478
5479<p>
5480The '<tt>llvm.cttz</tt>' intrinsic counts the trailing (least significant) zeros
5481in a variable. If the src == 0 then the result is the size in bits of the type
5482of src. For example, <tt>llvm.cttz(2) = 1</tt>.
5483</p>
5484</div>
5485
Reid Spencer8a5799f2007-04-01 08:27:01 +00005486<!-- _______________________________________________________________________ -->
5487<div class="doc_subsubsection">
Reid Spencerea2945e2007-04-10 02:51:31 +00005488 <a name="int_part_select">'<tt>llvm.part.select.*</tt>' Intrinsic</a>
Reid Spencer8bc7d952007-04-01 19:00:37 +00005489</div>
5490
5491<div class="doc_text">
5492
5493<h5>Syntax:</h5>
Reid Spencerea2945e2007-04-10 02:51:31 +00005494<p>This is an overloaded intrinsic. You can use <tt>llvm.part.select</tt>
Reid Spencer8bc7d952007-04-01 19:00:37 +00005495on any integer bit width.
5496<pre>
Chandler Carruth7132e002007-08-04 01:51:18 +00005497 declare i17 @llvm.part.select.i17 (i17 %val, i32 %loBit, i32 %hiBit)
5498 declare i29 @llvm.part.select.i29 (i29 %val, i32 %loBit, i32 %hiBit)
Reid Spencer8bc7d952007-04-01 19:00:37 +00005499</pre>
5500
5501<h5>Overview:</h5>
Reid Spencerea2945e2007-04-10 02:51:31 +00005502<p>The '<tt>llvm.part.select</tt>' family of intrinsic functions selects a
Reid Spencer8bc7d952007-04-01 19:00:37 +00005503range of bits from an integer value and returns them in the same bit width as
5504the original value.</p>
5505
5506<h5>Arguments:</h5>
5507<p>The first argument, <tt>%val</tt> and the result may be integer types of
5508any bit width but they must have the same bit width. The second and third
Reid Spencer96a5f022007-04-04 02:42:35 +00005509arguments must be <tt>i32</tt> type since they specify only a bit index.</p>
Reid Spencer8bc7d952007-04-01 19:00:37 +00005510
5511<h5>Semantics:</h5>
Reid Spencerea2945e2007-04-10 02:51:31 +00005512<p>The operation of the '<tt>llvm.part.select</tt>' intrinsic has two modes
Reid Spencer96a5f022007-04-04 02:42:35 +00005513of operation: forwards and reverse. If <tt>%loBit</tt> is greater than
5514<tt>%hiBits</tt> then the intrinsic operates in reverse mode. Otherwise it
5515operates in forward mode.</p>
5516<p>In forward mode, this intrinsic is the equivalent of shifting <tt>%val</tt>
5517right by <tt>%loBit</tt> bits and then ANDing it with a mask with
Reid Spencer8bc7d952007-04-01 19:00:37 +00005518only the <tt>%hiBit - %loBit</tt> bits set, as follows:</p>
5519<ol>
5520 <li>The <tt>%val</tt> is shifted right (LSHR) by the number of bits specified
5521 by <tt>%loBits</tt>. This normalizes the value to the low order bits.</li>
5522 <li>The <tt>%loBits</tt> value is subtracted from the <tt>%hiBits</tt> value
5523 to determine the number of bits to retain.</li>
5524 <li>A mask of the retained bits is created by shifting a -1 value.</li>
5525 <li>The mask is ANDed with <tt>%val</tt> to produce the result.
5526</ol>
Reid Spencer70845c02007-05-14 16:14:57 +00005527<p>In reverse mode, a similar computation is made except that the bits are
5528returned in the reverse order. So, for example, if <tt>X</tt> has the value
5529<tt>i16 0x0ACF (101011001111)</tt> and we apply
5530<tt>part.select(i16 X, 8, 3)</tt> to it, we get back the value
5531<tt>i16 0x0026 (000000100110)</tt>.</p>
Reid Spencer8bc7d952007-04-01 19:00:37 +00005532</div>
5533
Reid Spencer5bf54c82007-04-11 23:23:49 +00005534<div class="doc_subsubsection">
5535 <a name="int_part_set">'<tt>llvm.part.set.*</tt>' Intrinsic</a>
5536</div>
5537
5538<div class="doc_text">
5539
5540<h5>Syntax:</h5>
5541<p>This is an overloaded intrinsic. You can use <tt>llvm.part.set</tt>
5542on any integer bit width.
5543<pre>
Chandler Carruth7132e002007-08-04 01:51:18 +00005544 declare i17 @llvm.part.set.i17.i9 (i17 %val, i9 %repl, i32 %lo, i32 %hi)
5545 declare i29 @llvm.part.set.i29.i9 (i29 %val, i9 %repl, i32 %lo, i32 %hi)
Reid Spencer5bf54c82007-04-11 23:23:49 +00005546</pre>
5547
5548<h5>Overview:</h5>
5549<p>The '<tt>llvm.part.set</tt>' family of intrinsic functions replaces a range
5550of bits in an integer value with another integer value. It returns the integer
5551with the replaced bits.</p>
5552
5553<h5>Arguments:</h5>
5554<p>The first argument, <tt>%val</tt> and the result may be integer types of
5555any bit width but they must have the same bit width. <tt>%val</tt> is the value
5556whose bits will be replaced. The second argument, <tt>%repl</tt> may be an
5557integer of any bit width. The third and fourth arguments must be <tt>i32</tt>
5558type since they specify only a bit index.</p>
5559
5560<h5>Semantics:</h5>
5561<p>The operation of the '<tt>llvm.part.set</tt>' intrinsic has two modes
5562of operation: forwards and reverse. If <tt>%lo</tt> is greater than
5563<tt>%hi</tt> then the intrinsic operates in reverse mode. Otherwise it
5564operates in forward mode.</p>
5565<p>For both modes, the <tt>%repl</tt> value is prepared for use by either
5566truncating it down to the size of the replacement area or zero extending it
5567up to that size.</p>
5568<p>In forward mode, the bits between <tt>%lo</tt> and <tt>%hi</tt> (inclusive)
5569are replaced with corresponding bits from <tt>%repl</tt>. That is the 0th bit
5570in <tt>%repl</tt> replaces the <tt>%lo</tt>th bit in <tt>%val</tt> and etc. up
5571to the <tt>%hi</tt>th bit.
Reid Spencer146281c2007-05-14 16:50:20 +00005572<p>In reverse mode, a similar computation is made except that the bits are
5573reversed. That is, the <tt>0</tt>th bit in <tt>%repl</tt> replaces the
5574<tt>%hi</tt> bit in <tt>%val</tt> and etc. down to the <tt>%lo</tt>th bit.
Reid Spencer5bf54c82007-04-11 23:23:49 +00005575<h5>Examples:</h5>
5576<pre>
Reid Spencerc70afc32007-04-12 01:03:03 +00005577 llvm.part.set(0xFFFF, 0, 4, 7) -&gt; 0xFF0F
Reid Spencer146281c2007-05-14 16:50:20 +00005578 llvm.part.set(0xFFFF, 0, 7, 4) -&gt; 0xFF0F
5579 llvm.part.set(0xFFFF, 1, 7, 4) -&gt; 0xFF8F
5580 llvm.part.set(0xFFFF, F, 8, 3) -&gt; 0xFFE7
Reid Spencerc70afc32007-04-12 01:03:03 +00005581 llvm.part.set(0xFFFF, 0, 3, 8) -&gt; 0xFE07
Reid Spencer7972c472007-04-11 23:49:50 +00005582</pre>
Reid Spencer5bf54c82007-04-11 23:23:49 +00005583</div>
5584
Chris Lattner941515c2004-01-06 05:31:32 +00005585<!-- ======================================================================= -->
5586<div class="doc_subsection">
5587 <a name="int_debugger">Debugger Intrinsics</a>
5588</div>
5589
5590<div class="doc_text">
5591<p>
5592The LLVM debugger intrinsics (which all start with <tt>llvm.dbg.</tt> prefix),
5593are described in the <a
5594href="SourceLevelDebugging.html#format_common_intrinsics">LLVM Source Level
5595Debugging</a> document.
5596</p>
5597</div>
5598
5599
Jim Laskey2211f492007-03-14 19:31:19 +00005600<!-- ======================================================================= -->
5601<div class="doc_subsection">
5602 <a name="int_eh">Exception Handling Intrinsics</a>
5603</div>
5604
5605<div class="doc_text">
5606<p> The LLVM exception handling intrinsics (which all start with
5607<tt>llvm.eh.</tt> prefix), are described in the <a
5608href="ExceptionHandling.html#format_common_intrinsics">LLVM Exception
5609Handling</a> document. </p>
5610</div>
5611
Tanya Lattnercb1b9602007-06-15 20:50:54 +00005612<!-- ======================================================================= -->
5613<div class="doc_subsection">
Duncan Sands86e01192007-09-11 14:10:23 +00005614 <a name="int_trampoline">Trampoline Intrinsic</a>
Duncan Sands644f9172007-07-27 12:58:54 +00005615</div>
5616
5617<div class="doc_text">
5618<p>
Duncan Sands86e01192007-09-11 14:10:23 +00005619 This intrinsic makes it possible to excise one parameter, marked with
Duncan Sands644f9172007-07-27 12:58:54 +00005620 the <tt>nest</tt> attribute, from a function. The result is a callable
5621 function pointer lacking the nest parameter - the caller does not need
5622 to provide a value for it. Instead, the value to use is stored in
5623 advance in a "trampoline", a block of memory usually allocated
5624 on the stack, which also contains code to splice the nest value into the
5625 argument list. This is used to implement the GCC nested function address
5626 extension.
5627</p>
5628<p>
5629 For example, if the function is
5630 <tt>i32 f(i8* nest %c, i32 %x, i32 %y)</tt> then the resulting function
Bill Wendling252570f2007-09-22 09:23:55 +00005631 pointer has signature <tt>i32 (i32, i32)*</tt>. It can be created as follows:</p>
Duncan Sands644f9172007-07-27 12:58:54 +00005632<pre>
Duncan Sands86e01192007-09-11 14:10:23 +00005633 %tramp = alloca [10 x i8], align 4 ; size and alignment only correct for X86
5634 %tramp1 = getelementptr [10 x i8]* %tramp, i32 0, i32 0
5635 %p = call i8* @llvm.init.trampoline( i8* %tramp1, i8* bitcast (i32 (i8* nest , i32, i32)* @f to i8*), i8* %nval )
5636 %fp = bitcast i8* %p to i32 (i32, i32)*
Duncan Sands644f9172007-07-27 12:58:54 +00005637</pre>
Bill Wendling252570f2007-09-22 09:23:55 +00005638 <p>The call <tt>%val = call i32 %fp( i32 %x, i32 %y )</tt> is then equivalent
5639 to <tt>%val = call i32 %f( i8* %nval, i32 %x, i32 %y )</tt>.</p>
Duncan Sands644f9172007-07-27 12:58:54 +00005640</div>
5641
5642<!-- _______________________________________________________________________ -->
5643<div class="doc_subsubsection">
5644 <a name="int_it">'<tt>llvm.init.trampoline</tt>' Intrinsic</a>
5645</div>
5646<div class="doc_text">
5647<h5>Syntax:</h5>
5648<pre>
Duncan Sands86e01192007-09-11 14:10:23 +00005649declare i8* @llvm.init.trampoline(i8* &lt;tramp&gt;, i8* &lt;func&gt;, i8* &lt;nval&gt;)
Duncan Sands644f9172007-07-27 12:58:54 +00005650</pre>
5651<h5>Overview:</h5>
5652<p>
Duncan Sands86e01192007-09-11 14:10:23 +00005653 This fills the memory pointed to by <tt>tramp</tt> with code
5654 and returns a function pointer suitable for executing it.
Duncan Sands644f9172007-07-27 12:58:54 +00005655</p>
5656<h5>Arguments:</h5>
5657<p>
5658 The <tt>llvm.init.trampoline</tt> intrinsic takes three arguments, all
5659 pointers. The <tt>tramp</tt> argument must point to a sufficiently large
5660 and sufficiently aligned block of memory; this memory is written to by the
Duncan Sandsf2bcd372007-08-22 23:39:54 +00005661 intrinsic. Note that the size and the alignment are target-specific - LLVM
5662 currently provides no portable way of determining them, so a front-end that
5663 generates this intrinsic needs to have some target-specific knowledge.
5664 The <tt>func</tt> argument must hold a function bitcast to an <tt>i8*</tt>.
Duncan Sands644f9172007-07-27 12:58:54 +00005665</p>
5666<h5>Semantics:</h5>
5667<p>
5668 The block of memory pointed to by <tt>tramp</tt> is filled with target
Duncan Sands86e01192007-09-11 14:10:23 +00005669 dependent code, turning it into a function. A pointer to this function is
5670 returned, but needs to be bitcast to an
Duncan Sands644f9172007-07-27 12:58:54 +00005671 <a href="#int_trampoline">appropriate function pointer type</a>
Duncan Sands86e01192007-09-11 14:10:23 +00005672 before being called. The new function's signature is the same as that of
5673 <tt>func</tt> with any arguments marked with the <tt>nest</tt> attribute
5674 removed. At most one such <tt>nest</tt> argument is allowed, and it must be
5675 of pointer type. Calling the new function is equivalent to calling
5676 <tt>func</tt> with the same argument list, but with <tt>nval</tt> used for the
5677 missing <tt>nest</tt> argument. If, after calling
5678 <tt>llvm.init.trampoline</tt>, the memory pointed to by <tt>tramp</tt> is
5679 modified, then the effect of any later call to the returned function pointer is
5680 undefined.
Duncan Sands644f9172007-07-27 12:58:54 +00005681</p>
5682</div>
5683
5684<!-- ======================================================================= -->
5685<div class="doc_subsection">
Andrew Lenharth9b254ee2008-02-16 01:24:58 +00005686 <a name="int_atomics">Atomic Operations and Synchronization Intrinsics</a>
5687</div>
5688
5689<div class="doc_text">
5690<p>
5691 These intrinsic functions expand the "universal IR" of LLVM to represent
5692 hardware constructs for atomic operations and memory synchronization. This
5693 provides an interface to the hardware, not an interface to the programmer. It
5694 is aimed at a low enough level to allow any programming models or APIs which
5695 need atomic behaviors to map cleanly onto it. It is also modeled primarily on
5696 hardware behavior. Just as hardware provides a "universal IR" for source
5697 languages, it also provides a starting point for developing a "universal"
5698 atomic operation and synchronization IR.
5699</p>
5700<p>
5701 These do <em>not</em> form an API such as high-level threading libraries,
5702 software transaction memory systems, atomic primitives, and intrinsic
5703 functions as found in BSD, GNU libc, atomic_ops, APR, and other system and
5704 application libraries. The hardware interface provided by LLVM should allow
5705 a clean implementation of all of these APIs and parallel programming models.
5706 No one model or paradigm should be selected above others unless the hardware
5707 itself ubiquitously does so.
5708
5709</p>
5710</div>
5711
5712<!-- _______________________________________________________________________ -->
5713<div class="doc_subsubsection">
5714 <a name="int_memory_barrier">'<tt>llvm.memory.barrier</tt>' Intrinsic</a>
5715</div>
5716<div class="doc_text">
5717<h5>Syntax:</h5>
5718<pre>
5719declare void @llvm.memory.barrier( i1 &lt;ll&gt;, i1 &lt;ls&gt;, i1 &lt;sl&gt;, i1 &lt;ss&gt;,
5720i1 &lt;device&gt; )
5721
5722</pre>
5723<h5>Overview:</h5>
5724<p>
5725 The <tt>llvm.memory.barrier</tt> intrinsic guarantees ordering between
5726 specific pairs of memory access types.
5727</p>
5728<h5>Arguments:</h5>
5729<p>
5730 The <tt>llvm.memory.barrier</tt> intrinsic requires five boolean arguments.
5731 The first four arguments enables a specific barrier as listed below. The fith
5732 argument specifies that the barrier applies to io or device or uncached memory.
5733
5734</p>
5735 <ul>
5736 <li><tt>ll</tt>: load-load barrier</li>
5737 <li><tt>ls</tt>: load-store barrier</li>
5738 <li><tt>sl</tt>: store-load barrier</li>
5739 <li><tt>ss</tt>: store-store barrier</li>
5740 <li><tt>device</tt>: barrier applies to device and uncached memory also.
5741 </ul>
5742<h5>Semantics:</h5>
5743<p>
5744 This intrinsic causes the system to enforce some ordering constraints upon
5745 the loads and stores of the program. This barrier does not indicate
5746 <em>when</em> any events will occur, it only enforces an <em>order</em> in
5747 which they occur. For any of the specified pairs of load and store operations
5748 (f.ex. load-load, or store-load), all of the first operations preceding the
5749 barrier will complete before any of the second operations succeeding the
5750 barrier begin. Specifically the semantics for each pairing is as follows:
5751</p>
5752 <ul>
5753 <li><tt>ll</tt>: All loads before the barrier must complete before any load
5754 after the barrier begins.</li>
5755
5756 <li><tt>ls</tt>: All loads before the barrier must complete before any
5757 store after the barrier begins.</li>
5758 <li><tt>ss</tt>: All stores before the barrier must complete before any
5759 store after the barrier begins.</li>
5760 <li><tt>sl</tt>: All stores before the barrier must complete before any
5761 load after the barrier begins.</li>
5762 </ul>
5763<p>
5764 These semantics are applied with a logical "and" behavior when more than one
5765 is enabled in a single memory barrier intrinsic.
5766</p>
5767<p>
5768 Backends may implement stronger barriers than those requested when they do not
5769 support as fine grained a barrier as requested. Some architectures do not
5770 need all types of barriers and on such architectures, these become noops.
5771</p>
5772<h5>Example:</h5>
5773<pre>
5774%ptr = malloc i32
5775 store i32 4, %ptr
5776
5777%result1 = load i32* %ptr <i>; yields {i32}:result1 = 4</i>
5778 call void @llvm.memory.barrier( i1 false, i1 true, i1 false, i1 false )
5779 <i>; guarantee the above finishes</i>
5780 store i32 8, %ptr <i>; before this begins</i>
5781</pre>
5782</div>
5783
Andrew Lenharth95528942008-02-21 06:45:13 +00005784<!-- _______________________________________________________________________ -->
5785<div class="doc_subsubsection">
Mon P Wang6a490372008-06-25 08:15:39 +00005786 <a name="int_atomic_cmp_swap">'<tt>llvm.atomic.cmp.swap.*</tt>' Intrinsic</a>
Andrew Lenharth95528942008-02-21 06:45:13 +00005787</div>
5788<div class="doc_text">
5789<h5>Syntax:</h5>
5790<p>
Mon P Wang6a490372008-06-25 08:15:39 +00005791 This is an overloaded intrinsic. You can use <tt>llvm.atomic.cmp.swap</tt> on any
Andrew Lenharth95528942008-02-21 06:45:13 +00005792 integer bit width. Not all targets support all bit widths however.</p>
5793
5794<pre>
Mon P Wang6a490372008-06-25 08:15:39 +00005795declare i8 @llvm.atomic.cmp.swap.i8( i8* &lt;ptr&gt;, i8 &lt;cmp&gt;, i8 &lt;val&gt; )
5796declare i16 @llvm.atomic.cmp.swap.i16( i16* &lt;ptr&gt;, i16 &lt;cmp&gt;, i16 &lt;val&gt; )
5797declare i32 @llvm.atomic.cmp.swap.i32( i32* &lt;ptr&gt;, i32 &lt;cmp&gt;, i32 &lt;val&gt; )
5798declare i64 @llvm.atomic.cmp.swap.i64( i64* &lt;ptr&gt;, i64 &lt;cmp&gt;, i64 &lt;val&gt; )
Andrew Lenharth95528942008-02-21 06:45:13 +00005799
5800</pre>
5801<h5>Overview:</h5>
5802<p>
5803 This loads a value in memory and compares it to a given value. If they are
5804 equal, it stores a new value into the memory.
5805</p>
5806<h5>Arguments:</h5>
5807<p>
Mon P Wang6a490372008-06-25 08:15:39 +00005808 The <tt>llvm.atomic.cmp.swap</tt> intrinsic takes three arguments. The result as
Andrew Lenharth95528942008-02-21 06:45:13 +00005809 well as both <tt>cmp</tt> and <tt>val</tt> must be integer values with the
5810 same bit width. The <tt>ptr</tt> argument must be a pointer to a value of
5811 this integer type. While any bit width integer may be used, targets may only
5812 lower representations they support in hardware.
5813
5814</p>
5815<h5>Semantics:</h5>
5816<p>
5817 This entire intrinsic must be executed atomically. It first loads the value
5818 in memory pointed to by <tt>ptr</tt> and compares it with the value
5819 <tt>cmp</tt>. If they are equal, <tt>val</tt> is stored into the memory. The
5820 loaded value is yielded in all cases. This provides the equivalent of an
5821 atomic compare-and-swap operation within the SSA framework.
5822</p>
5823<h5>Examples:</h5>
5824
5825<pre>
5826%ptr = malloc i32
5827 store i32 4, %ptr
5828
5829%val1 = add i32 4, 4
Mon P Wang6a490372008-06-25 08:15:39 +00005830%result1 = call i32 @llvm.atomic.cmp.swap.i32( i32* %ptr, i32 4, %val1 )
Andrew Lenharth95528942008-02-21 06:45:13 +00005831 <i>; yields {i32}:result1 = 4</i>
5832%stored1 = icmp eq i32 %result1, 4 <i>; yields {i1}:stored1 = true</i>
5833%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 8</i>
5834
5835%val2 = add i32 1, 1
Mon P Wang6a490372008-06-25 08:15:39 +00005836%result2 = call i32 @llvm.atomic.cmp.swap.i32( i32* %ptr, i32 5, %val2 )
Andrew Lenharth95528942008-02-21 06:45:13 +00005837 <i>; yields {i32}:result2 = 8</i>
5838%stored2 = icmp eq i32 %result2, 5 <i>; yields {i1}:stored2 = false</i>
5839
5840%memval2 = load i32* %ptr <i>; yields {i32}:memval2 = 8</i>
5841</pre>
5842</div>
5843
5844<!-- _______________________________________________________________________ -->
5845<div class="doc_subsubsection">
5846 <a name="int_atomic_swap">'<tt>llvm.atomic.swap.*</tt>' Intrinsic</a>
5847</div>
5848<div class="doc_text">
5849<h5>Syntax:</h5>
5850
5851<p>
5852 This is an overloaded intrinsic. You can use <tt>llvm.atomic.swap</tt> on any
5853 integer bit width. Not all targets support all bit widths however.</p>
5854<pre>
5855declare i8 @llvm.atomic.swap.i8( i8* &lt;ptr&gt;, i8 &lt;val&gt; )
5856declare i16 @llvm.atomic.swap.i16( i16* &lt;ptr&gt;, i16 &lt;val&gt; )
5857declare i32 @llvm.atomic.swap.i32( i32* &lt;ptr&gt;, i32 &lt;val&gt; )
5858declare i64 @llvm.atomic.swap.i64( i64* &lt;ptr&gt;, i64 &lt;val&gt; )
5859
5860</pre>
5861<h5>Overview:</h5>
5862<p>
5863 This intrinsic loads the value stored in memory at <tt>ptr</tt> and yields
5864 the value from memory. It then stores the value in <tt>val</tt> in the memory
5865 at <tt>ptr</tt>.
5866</p>
5867<h5>Arguments:</h5>
5868
5869<p>
Mon P Wang6a490372008-06-25 08:15:39 +00005870 The <tt>llvm.atomic.swap</tt> intrinsic takes two arguments. Both the
Andrew Lenharth95528942008-02-21 06:45:13 +00005871 <tt>val</tt> argument and the result must be integers of the same bit width.
5872 The first argument, <tt>ptr</tt>, must be a pointer to a value of this
5873 integer type. The targets may only lower integer representations they
5874 support.
5875</p>
5876<h5>Semantics:</h5>
5877<p>
5878 This intrinsic loads the value pointed to by <tt>ptr</tt>, yields it, and
5879 stores <tt>val</tt> back into <tt>ptr</tt> atomically. This provides the
5880 equivalent of an atomic swap operation within the SSA framework.
5881
5882</p>
5883<h5>Examples:</h5>
5884<pre>
5885%ptr = malloc i32
5886 store i32 4, %ptr
5887
5888%val1 = add i32 4, 4
5889%result1 = call i32 @llvm.atomic.swap.i32( i32* %ptr, i32 %val1 )
5890 <i>; yields {i32}:result1 = 4</i>
5891%stored1 = icmp eq i32 %result1, 4 <i>; yields {i1}:stored1 = true</i>
5892%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 8</i>
5893
5894%val2 = add i32 1, 1
5895%result2 = call i32 @llvm.atomic.swap.i32( i32* %ptr, i32 %val2 )
5896 <i>; yields {i32}:result2 = 8</i>
5897
5898%stored2 = icmp eq i32 %result2, 8 <i>; yields {i1}:stored2 = true</i>
5899%memval2 = load i32* %ptr <i>; yields {i32}:memval2 = 2</i>
5900</pre>
5901</div>
5902
5903<!-- _______________________________________________________________________ -->
5904<div class="doc_subsubsection">
Mon P Wang6a490372008-06-25 08:15:39 +00005905 <a name="int_atomic_load_add">'<tt>llvm.atomic.load.add.*</tt>' Intrinsic</a>
Andrew Lenharth95528942008-02-21 06:45:13 +00005906
5907</div>
5908<div class="doc_text">
5909<h5>Syntax:</h5>
5910<p>
Mon P Wang6a490372008-06-25 08:15:39 +00005911 This is an overloaded intrinsic. You can use <tt>llvm.atomic.load.add</tt> on any
Andrew Lenharth95528942008-02-21 06:45:13 +00005912 integer bit width. Not all targets support all bit widths however.</p>
5913<pre>
Mon P Wang6a490372008-06-25 08:15:39 +00005914declare i8 @llvm.atomic.load.add.i8.( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
5915declare i16 @llvm.atomic.load.add.i16.( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
5916declare i32 @llvm.atomic.load.add.i32.( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
5917declare i64 @llvm.atomic.load.add.i64.( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Andrew Lenharth95528942008-02-21 06:45:13 +00005918
5919</pre>
5920<h5>Overview:</h5>
5921<p>
5922 This intrinsic adds <tt>delta</tt> to the value stored in memory at
5923 <tt>ptr</tt>. It yields the original value at <tt>ptr</tt>.
5924</p>
5925<h5>Arguments:</h5>
5926<p>
5927
5928 The intrinsic takes two arguments, the first a pointer to an integer value
5929 and the second an integer value. The result is also an integer value. These
5930 integer types can have any bit width, but they must all have the same bit
5931 width. The targets may only lower integer representations they support.
5932</p>
5933<h5>Semantics:</h5>
5934<p>
5935 This intrinsic does a series of operations atomically. It first loads the
5936 value stored at <tt>ptr</tt>. It then adds <tt>delta</tt>, stores the result
5937 to <tt>ptr</tt>. It yields the original value stored at <tt>ptr</tt>.
5938</p>
5939
5940<h5>Examples:</h5>
5941<pre>
5942%ptr = malloc i32
5943 store i32 4, %ptr
Mon P Wang6a490372008-06-25 08:15:39 +00005944%result1 = call i32 @llvm.atomic.load.add.i32( i32* %ptr, i32 4 )
Andrew Lenharth95528942008-02-21 06:45:13 +00005945 <i>; yields {i32}:result1 = 4</i>
Mon P Wang6a490372008-06-25 08:15:39 +00005946%result2 = call i32 @llvm.atomic.load.add.i32( i32* %ptr, i32 2 )
Andrew Lenharth95528942008-02-21 06:45:13 +00005947 <i>; yields {i32}:result2 = 8</i>
Mon P Wang6a490372008-06-25 08:15:39 +00005948%result3 = call i32 @llvm.atomic.load.add.i32( i32* %ptr, i32 5 )
Andrew Lenharth95528942008-02-21 06:45:13 +00005949 <i>; yields {i32}:result3 = 10</i>
Mon P Wang6a490372008-06-25 08:15:39 +00005950%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 15</i>
Andrew Lenharth95528942008-02-21 06:45:13 +00005951</pre>
5952</div>
5953
Mon P Wang6a490372008-06-25 08:15:39 +00005954<!-- _______________________________________________________________________ -->
5955<div class="doc_subsubsection">
5956 <a name="int_atomic_load_sub">'<tt>llvm.atomic.load.sub.*</tt>' Intrinsic</a>
5957
5958</div>
5959<div class="doc_text">
5960<h5>Syntax:</h5>
5961<p>
5962 This is an overloaded intrinsic. You can use <tt>llvm.atomic.load.sub</tt> on
5963 any integer bit width. Not all targets support all bit widths however.</p>
5964<pre>
5965declare i8 @llvm.atomic.load.sub.i8.( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
5966declare i16 @llvm.atomic.load.sub.i16.( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
5967declare i32 @llvm.atomic.load.sub.i32.( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
5968declare i64 @llvm.atomic.load.sub.i64.( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
5969
5970</pre>
5971<h5>Overview:</h5>
5972<p>
5973 This intrinsic subtracts <tt>delta</tt> to the value stored in memory at
5974 <tt>ptr</tt>. It yields the original value at <tt>ptr</tt>.
5975</p>
5976<h5>Arguments:</h5>
5977<p>
5978
5979 The intrinsic takes two arguments, the first a pointer to an integer value
5980 and the second an integer value. The result is also an integer value. These
5981 integer types can have any bit width, but they must all have the same bit
5982 width. The targets may only lower integer representations they support.
5983</p>
5984<h5>Semantics:</h5>
5985<p>
5986 This intrinsic does a series of operations atomically. It first loads the
5987 value stored at <tt>ptr</tt>. It then subtracts <tt>delta</tt>, stores the
5988 result to <tt>ptr</tt>. It yields the original value stored at <tt>ptr</tt>.
5989</p>
5990
5991<h5>Examples:</h5>
5992<pre>
5993%ptr = malloc i32
5994 store i32 8, %ptr
5995%result1 = call i32 @llvm.atomic.load.sub.i32( i32* %ptr, i32 4 )
5996 <i>; yields {i32}:result1 = 8</i>
5997%result2 = call i32 @llvm.atomic.load.sub.i32( i32* %ptr, i32 2 )
5998 <i>; yields {i32}:result2 = 4</i>
5999%result3 = call i32 @llvm.atomic.load.sub.i32( i32* %ptr, i32 5 )
6000 <i>; yields {i32}:result3 = 2</i>
6001%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = -3</i>
6002</pre>
6003</div>
6004
6005<!-- _______________________________________________________________________ -->
6006<div class="doc_subsubsection">
6007 <a name="int_atomic_load_and">'<tt>llvm.atomic.load.and.*</tt>' Intrinsic</a><br>
6008 <a name="int_atomic_load_nand">'<tt>llvm.atomic.load.nand.*</tt>' Intrinsic</a><br>
6009 <a name="int_atomic_load_or">'<tt>llvm.atomic.load.or.*</tt>' Intrinsic</a><br>
6010 <a name="int_atomic_load_xor">'<tt>llvm.atomic.load.xor.*</tt>' Intrinsic</a><br>
6011
6012</div>
6013<div class="doc_text">
6014<h5>Syntax:</h5>
6015<p>
6016 These are overloaded intrinsics. You can use <tt>llvm.atomic.load_and</tt>,
6017 <tt>llvm.atomic.load_nand</tt>, <tt>llvm.atomic.load_or</tt>, and
6018 <tt>llvm.atomic.load_xor</tt> on any integer bit width. Not all targets
6019 support all bit widths however.</p>
6020<pre>
6021declare i8 @llvm.atomic.load.and.i8.( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6022declare i16 @llvm.atomic.load.and.i16.( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6023declare i32 @llvm.atomic.load.and.i32.( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6024declare i64 @llvm.atomic.load.and.i64.( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
6025
6026</pre>
6027
6028<pre>
6029declare i8 @llvm.atomic.load.or.i8.( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6030declare i16 @llvm.atomic.load.or.i16.( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6031declare i32 @llvm.atomic.load.or.i32.( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6032declare i64 @llvm.atomic.load.or.i64.( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
6033
6034</pre>
6035
6036<pre>
6037declare i8 @llvm.atomic.load.nand.i8.( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6038declare i16 @llvm.atomic.load.nand.i16.( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6039declare i32 @llvm.atomic.load.nand.i32.( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6040declare i64 @llvm.atomic.load.nand.i64.( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
6041
6042</pre>
6043
6044<pre>
6045declare i8 @llvm.atomic.load.xor.i8.( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6046declare i16 @llvm.atomic.load.xor.i16.( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6047declare i32 @llvm.atomic.load.xor.i32.( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6048declare i64 @llvm.atomic.load.xor.i64.( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
6049
6050</pre>
6051<h5>Overview:</h5>
6052<p>
6053 These intrinsics bitwise the operation (and, nand, or, xor) <tt>delta</tt> to
6054 the value stored in memory at <tt>ptr</tt>. It yields the original value
6055 at <tt>ptr</tt>.
6056</p>
6057<h5>Arguments:</h5>
6058<p>
6059
6060 These intrinsics take two arguments, the first a pointer to an integer value
6061 and the second an integer value. The result is also an integer value. These
6062 integer types can have any bit width, but they must all have the same bit
6063 width. The targets may only lower integer representations they support.
6064</p>
6065<h5>Semantics:</h5>
6066<p>
6067 These intrinsics does a series of operations atomically. They first load the
6068 value stored at <tt>ptr</tt>. They then do the bitwise operation
6069 <tt>delta</tt>, store the result to <tt>ptr</tt>. They yield the original
6070 value stored at <tt>ptr</tt>.
6071</p>
6072
6073<h5>Examples:</h5>
6074<pre>
6075%ptr = malloc i32
6076 store i32 0x0F0F, %ptr
6077%result0 = call i32 @llvm.atomic.load.nand.i32( i32* %ptr, i32 0xFF )
6078 <i>; yields {i32}:result0 = 0x0F0F</i>
6079%result1 = call i32 @llvm.atomic.load.and.i32( i32* %ptr, i32 0xFF )
6080 <i>; yields {i32}:result1 = 0xFFFFFFF0</i>
6081%result2 = call i32 @llvm.atomic.load.or.i32( i32* %ptr, i32 0F )
6082 <i>; yields {i32}:result2 = 0xF0</i>
6083%result3 = call i32 @llvm.atomic.load.xor.i32( i32* %ptr, i32 0F )
6084 <i>; yields {i32}:result3 = FF</i>
6085%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = F0</i>
6086</pre>
6087</div>
6088
6089
6090<!-- _______________________________________________________________________ -->
6091<div class="doc_subsubsection">
6092 <a name="int_atomic_load_max">'<tt>llvm.atomic.load.max.*</tt>' Intrinsic</a><br>
6093 <a name="int_atomic_load_min">'<tt>llvm.atomic.load.min.*</tt>' Intrinsic</a><br>
6094 <a name="int_atomic_load_umax">'<tt>llvm.atomic.load.umax.*</tt>' Intrinsic</a><br>
6095 <a name="int_atomic_load_umin">'<tt>llvm.atomic.load.umin.*</tt>' Intrinsic</a><br>
6096
6097</div>
6098<div class="doc_text">
6099<h5>Syntax:</h5>
6100<p>
6101 These are overloaded intrinsics. You can use <tt>llvm.atomic.load_max</tt>,
6102 <tt>llvm.atomic.load_min</tt>, <tt>llvm.atomic.load_umax</tt>, and
6103 <tt>llvm.atomic.load_umin</tt> on any integer bit width. Not all targets
6104 support all bit widths however.</p>
6105<pre>
6106declare i8 @llvm.atomic.load.max.i8.( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6107declare i16 @llvm.atomic.load.max.i16.( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6108declare i32 @llvm.atomic.load.max.i32.( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6109declare i64 @llvm.atomic.load.max.i64.( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
6110
6111</pre>
6112
6113<pre>
6114declare i8 @llvm.atomic.load.min.i8.( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6115declare i16 @llvm.atomic.load.min.i16.( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6116declare i32 @llvm.atomic.load.min.i32.( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6117declare i64 @llvm.atomic.load.min.i64.( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
6118
6119</pre>
6120
6121<pre>
6122declare i8 @llvm.atomic.load.umax.i8.( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6123declare i16 @llvm.atomic.load.umax.i16.( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6124declare i32 @llvm.atomic.load.umax.i32.( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6125declare i64 @llvm.atomic.load.umax.i64.( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
6126
6127</pre>
6128
6129<pre>
6130declare i8 @llvm.atomic.load.umin.i8.( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6131declare i16 @llvm.atomic.load.umin.i16.( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6132declare i32 @llvm.atomic.load.umin.i32.( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6133declare i64 @llvm.atomic.load.umin.i64.( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
6134
6135</pre>
6136<h5>Overview:</h5>
6137<p>
6138 These intrinsics takes the signed or unsigned minimum or maximum of
6139 <tt>delta</tt> and the value stored in memory at <tt>ptr</tt>. It yields the
6140 original value at <tt>ptr</tt>.
6141</p>
6142<h5>Arguments:</h5>
6143<p>
6144
6145 These intrinsics take two arguments, the first a pointer to an integer value
6146 and the second an integer value. The result is also an integer value. These
6147 integer types can have any bit width, but they must all have the same bit
6148 width. The targets may only lower integer representations they support.
6149</p>
6150<h5>Semantics:</h5>
6151<p>
6152 These intrinsics does a series of operations atomically. They first load the
6153 value stored at <tt>ptr</tt>. They then do the signed or unsigned min or max
6154 <tt>delta</tt> and the value, store the result to <tt>ptr</tt>. They yield
6155 the original value stored at <tt>ptr</tt>.
6156</p>
6157
6158<h5>Examples:</h5>
6159<pre>
6160%ptr = malloc i32
6161 store i32 7, %ptr
6162%result0 = call i32 @llvm.atomic.load.min.i32( i32* %ptr, i32 -2 )
6163 <i>; yields {i32}:result0 = 7</i>
6164%result1 = call i32 @llvm.atomic.load.max.i32( i32* %ptr, i32 8 )
6165 <i>; yields {i32}:result1 = -2</i>
6166%result2 = call i32 @llvm.atomic.load.umin.i32( i32* %ptr, i32 10 )
6167 <i>; yields {i32}:result2 = 8</i>
6168%result3 = call i32 @llvm.atomic.load.umax.i32( i32* %ptr, i32 30 )
6169 <i>; yields {i32}:result3 = 8</i>
6170%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 30</i>
6171</pre>
6172</div>
Andrew Lenharth9b254ee2008-02-16 01:24:58 +00006173
6174<!-- ======================================================================= -->
6175<div class="doc_subsection">
Tanya Lattnercb1b9602007-06-15 20:50:54 +00006176 <a name="int_general">General Intrinsics</a>
6177</div>
6178
6179<div class="doc_text">
6180<p> This class of intrinsics is designed to be generic and has
6181no specific purpose. </p>
6182</div>
6183
6184<!-- _______________________________________________________________________ -->
6185<div class="doc_subsubsection">
6186 <a name="int_var_annotation">'<tt>llvm.var.annotation</tt>' Intrinsic</a>
6187</div>
6188
6189<div class="doc_text">
6190
6191<h5>Syntax:</h5>
6192<pre>
Tanya Lattnerbed1d4d2007-06-18 23:42:37 +00006193 declare void @llvm.var.annotation(i8* &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
Tanya Lattnercb1b9602007-06-15 20:50:54 +00006194</pre>
6195
6196<h5>Overview:</h5>
6197
6198<p>
6199The '<tt>llvm.var.annotation</tt>' intrinsic
6200</p>
6201
6202<h5>Arguments:</h5>
6203
6204<p>
Tanya Lattnerbed1d4d2007-06-18 23:42:37 +00006205The first argument is a pointer to a value, the second is a pointer to a
6206global string, the third is a pointer to a global string which is the source
6207file name, and the last argument is the line number.
Tanya Lattnercb1b9602007-06-15 20:50:54 +00006208</p>
6209
6210<h5>Semantics:</h5>
6211
6212<p>
Anton Korobeynikov06cbb652008-01-15 22:31:34 +00006213This intrinsic allows annotation of local variables with arbitrary strings.
Tanya Lattnercb1b9602007-06-15 20:50:54 +00006214This can be useful for special purpose optimizations that want to look for these
Anton Korobeynikov06cbb652008-01-15 22:31:34 +00006215annotations. These have no other defined use, they are ignored by code
6216generation and optimization.
6217</p>
Tanya Lattnercb1b9602007-06-15 20:50:54 +00006218</div>
6219
Tanya Lattner293c0372007-09-21 22:59:12 +00006220<!-- _______________________________________________________________________ -->
6221<div class="doc_subsubsection">
Tanya Lattner0186a652007-09-21 23:57:59 +00006222 <a name="int_annotation">'<tt>llvm.annotation.*</tt>' Intrinsic</a>
Tanya Lattner293c0372007-09-21 22:59:12 +00006223</div>
6224
6225<div class="doc_text">
6226
6227<h5>Syntax:</h5>
Tanya Lattner23dbd572007-09-21 23:56:27 +00006228<p>This is an overloaded intrinsic. You can use '<tt>llvm.annotation</tt>' on
6229any integer bit width.
6230</p>
Tanya Lattner293c0372007-09-21 22:59:12 +00006231<pre>
Tanya Lattnercf3e26f2007-09-22 00:03:01 +00006232 declare i8 @llvm.annotation.i8(i8 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
6233 declare i16 @llvm.annotation.i16(i16 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
6234 declare i32 @llvm.annotation.i32(i32 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
6235 declare i64 @llvm.annotation.i64(i64 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
6236 declare i256 @llvm.annotation.i256(i256 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
Tanya Lattner293c0372007-09-21 22:59:12 +00006237</pre>
6238
6239<h5>Overview:</h5>
Tanya Lattner23dbd572007-09-21 23:56:27 +00006240
6241<p>
6242The '<tt>llvm.annotation</tt>' intrinsic.
Tanya Lattner293c0372007-09-21 22:59:12 +00006243</p>
6244
6245<h5>Arguments:</h5>
6246
6247<p>
6248The first argument is an integer value (result of some expression),
6249the second is a pointer to a global string, the third is a pointer to a global
6250string which is the source file name, and the last argument is the line number.
Tanya Lattner23dbd572007-09-21 23:56:27 +00006251It returns the value of the first argument.
Tanya Lattner293c0372007-09-21 22:59:12 +00006252</p>
6253
6254<h5>Semantics:</h5>
6255
6256<p>
6257This intrinsic allows annotations to be put on arbitrary expressions
6258with arbitrary strings. This can be useful for special purpose optimizations
6259that want to look for these annotations. These have no other defined use, they
6260are ignored by code generation and optimization.
6261</div>
Jim Laskey2211f492007-03-14 19:31:19 +00006262
Anton Korobeynikov06cbb652008-01-15 22:31:34 +00006263<!-- _______________________________________________________________________ -->
6264<div class="doc_subsubsection">
6265 <a name="int_trap">'<tt>llvm.trap</tt>' Intrinsic</a>
6266</div>
6267
6268<div class="doc_text">
6269
6270<h5>Syntax:</h5>
6271<pre>
6272 declare void @llvm.trap()
6273</pre>
6274
6275<h5>Overview:</h5>
6276
6277<p>
6278The '<tt>llvm.trap</tt>' intrinsic
6279</p>
6280
6281<h5>Arguments:</h5>
6282
6283<p>
6284None
6285</p>
6286
6287<h5>Semantics:</h5>
6288
6289<p>
6290This intrinsics is lowered to the target dependent trap instruction. If the
6291target does not have a trap instruction, this intrinsic will be lowered to the
6292call of the abort() function.
6293</p>
6294</div>
6295
Chris Lattner2f7c9632001-06-06 20:29:01 +00006296<!-- *********************************************************************** -->
Chris Lattner2f7c9632001-06-06 20:29:01 +00006297<hr>
Misha Brukmanc501f552004-03-01 17:47:27 +00006298<address>
6299 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
6300 src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
6301 <a href="http://validator.w3.org/check/referer"><img
Chris Lattnerb8f816e2008-01-04 04:33:49 +00006302 src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!"></a>
Misha Brukmanc501f552004-03-01 17:47:27 +00006303
6304 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
Reid Spencerca058542006-03-14 05:39:39 +00006305 <a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br>
Misha Brukmanc501f552004-03-01 17:47:27 +00006306 Last modified: $Date$
6307</address>
Chris Lattnerb8f816e2008-01-04 04:33:49 +00006308
Misha Brukman76307852003-11-08 01:05:38 +00006309</body>
6310</html>