blob: 7dc743f71392ee88a5e92f4e55efe528655262dc [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>
26 <li><a href="#functionstructure">Function Structure</a></li>
27 </ol>
28 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +000029 <li><a href="#typesystem">Type System</a>
30 <ol>
Chris Lattner48b383b02003-11-25 01:02:51 +000031 <li><a href="#t_primitive">Primitive Types</a>
32 <ol>
Misha Brukman76307852003-11-08 01:05:38 +000033 <li><a href="#t_classifications">Type Classifications</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +000034 </ol>
35 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +000036 <li><a href="#t_derived">Derived Types</a>
37 <ol>
Chris Lattner48b383b02003-11-25 01:02:51 +000038 <li><a href="#t_array">Array Type</a></li>
Misha Brukman76307852003-11-08 01:05:38 +000039 <li><a href="#t_function">Function Type</a></li>
40 <li><a href="#t_pointer">Pointer Type</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +000041 <li><a href="#t_struct">Structure Type</a></li>
Chris Lattnerc8cb6952004-08-12 19:12:28 +000042 <li><a href="#t_packed">Packed Type</a></li>
Chris Lattner37b6b092005-04-25 17:34:15 +000043 <li><a href="#t_opaque">Opaque Type</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +000044 </ol>
45 </li>
46 </ol>
47 </li>
Chris Lattner6af02f32004-12-09 16:11:40 +000048 <li><a href="#constants">Constants</a>
Chris Lattner74d3f822004-12-09 17:30:23 +000049 <ol>
50 <li><a href="#simpleconstants">Simple Constants</a>
51 <li><a href="#aggregateconstants">Aggregate Constants</a>
52 <li><a href="#globalconstants">Global Variable and Function Addresses</a>
53 <li><a href="#undefvalues">Undefined Values</a>
54 <li><a href="#constantexprs">Constant Expressions</a>
55 </ol>
Chris Lattner48b383b02003-11-25 01:02:51 +000056 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +000057 <li><a href="#instref">Instruction Reference</a>
58 <ol>
59 <li><a href="#terminators">Terminator Instructions</a>
60 <ol>
Chris Lattner48b383b02003-11-25 01:02:51 +000061 <li><a href="#i_ret">'<tt>ret</tt>' Instruction</a></li>
62 <li><a href="#i_br">'<tt>br</tt>' Instruction</a></li>
Misha Brukman76307852003-11-08 01:05:38 +000063 <li><a href="#i_switch">'<tt>switch</tt>' Instruction</a></li>
64 <li><a href="#i_invoke">'<tt>invoke</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +000065 <li><a href="#i_unwind">'<tt>unwind</tt>' Instruction</a></li>
Chris Lattner08b7d5b2004-10-16 18:04:13 +000066 <li><a href="#i_unreachable">'<tt>unreachable</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +000067 </ol>
68 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +000069 <li><a href="#binaryops">Binary Operations</a>
70 <ol>
Chris Lattner48b383b02003-11-25 01:02:51 +000071 <li><a href="#i_add">'<tt>add</tt>' Instruction</a></li>
72 <li><a href="#i_sub">'<tt>sub</tt>' Instruction</a></li>
73 <li><a href="#i_mul">'<tt>mul</tt>' Instruction</a></li>
74 <li><a href="#i_div">'<tt>div</tt>' Instruction</a></li>
75 <li><a href="#i_rem">'<tt>rem</tt>' Instruction</a></li>
Misha Brukman76307852003-11-08 01:05:38 +000076 <li><a href="#i_setcc">'<tt>set<i>cc</i></tt>' Instructions</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +000077 </ol>
78 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +000079 <li><a href="#bitwiseops">Bitwise Binary Operations</a>
80 <ol>
Misha Brukman76307852003-11-08 01:05:38 +000081 <li><a href="#i_and">'<tt>and</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +000082 <li><a href="#i_or">'<tt>or</tt>' Instruction</a></li>
Misha Brukman76307852003-11-08 01:05:38 +000083 <li><a href="#i_xor">'<tt>xor</tt>' Instruction</a></li>
84 <li><a href="#i_shl">'<tt>shl</tt>' Instruction</a></li>
85 <li><a href="#i_shr">'<tt>shr</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +000086 </ol>
87 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +000088 <li><a href="#memoryops">Memory Access Operations</a>
89 <ol>
Chris Lattner48b383b02003-11-25 01:02:51 +000090 <li><a href="#i_malloc">'<tt>malloc</tt>' Instruction</a></li>
91 <li><a href="#i_free">'<tt>free</tt>' Instruction</a></li>
92 <li><a href="#i_alloca">'<tt>alloca</tt>' Instruction</a></li>
93 <li><a href="#i_load">'<tt>load</tt>' Instruction</a></li>
94 <li><a href="#i_store">'<tt>store</tt>' Instruction</a></li>
95 <li><a href="#i_getelementptr">'<tt>getelementptr</tt>' Instruction</a></li>
96 </ol>
97 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +000098 <li><a href="#otherops">Other Operations</a>
99 <ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000100 <li><a href="#i_phi">'<tt>phi</tt>' Instruction</a></li>
Misha Brukman76307852003-11-08 01:05:38 +0000101 <li><a href="#i_cast">'<tt>cast .. to</tt>' Instruction</a></li>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000102 <li><a href="#i_select">'<tt>select</tt>' Instruction</a></li>
Robert Bocchinob5f210f2006-01-05 17:37:02 +0000103 <li><a href="#i_extractelement">'<tt>extractelement</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000104 <li><a href="#i_call">'<tt>call</tt>' Instruction</a></li>
Chris Lattner33337472006-01-13 23:26:01 +0000105 <li><a href="#i_va_arg">'<tt>va_arg</tt>' Instruction</a></li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000106 </ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000107 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000108 </ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000109 </li>
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000110 <li><a href="#intrinsics">Intrinsic Functions</a>
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000111 <ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000112 <li><a href="#int_varargs">Variable Argument Handling Intrinsics</a>
113 <ol>
114 <li><a href="#i_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a></li>
115 <li><a href="#i_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a></li>
116 <li><a href="#i_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a></li>
117 </ol>
118 </li>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000119 <li><a href="#int_gc">Accurate Garbage Collection Intrinsics</a>
120 <ol>
121 <li><a href="#i_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a></li>
122 <li><a href="#i_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a></li>
123 <li><a href="#i_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a></li>
124 </ol>
125 </li>
Chris Lattner3649c3a2004-02-14 04:08:35 +0000126 <li><a href="#int_codegen">Code Generator Intrinsics</a>
127 <ol>
Nate Begeman611ecc02006-01-13 23:02:51 +0000128 <li><a href="#i_bswap_i16">'<tt>llvm.bswap.i16</tt>' Intrinsic</a></li>
129 <li><a href="#i_bswap_i32">'<tt>llvm.bswap.i32</tt>' Intrinsic</a></li>
130 <li><a href="#i_bswap_i64">'<tt>llvm.bswap.i64</tt>' Intrinsic</a></li>
Chris Lattner3649c3a2004-02-14 04:08:35 +0000131 <li><a href="#i_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a></li>
132 <li><a href="#i_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a></li>
Chris Lattner2f0f0012006-01-13 02:03:13 +0000133 <li><a href="#i_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a></li>
134 <li><a href="#i_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a></li>
Chris Lattnerc8a2c222005-02-28 19:24:19 +0000135 <li><a href="#i_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a></li>
Andrew Lenharthb4427912005-03-28 20:05:49 +0000136 <li><a href="#i_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a></li>
Andrew Lenharth01aa5632005-11-11 16:47:30 +0000137 <li><a href="#i_readcyclecounter"><tt>llvm.readcyclecounter</tt>' Intrinsic</a></li>
John Criswellaa1c3c12004-04-09 16:43:20 +0000138 </ol>
139 </li>
140 <li><a href="#int_os">Operating System Intrinsics</a>
141 <ol>
Chris Lattner3b4f4372004-06-11 02:28:03 +0000142 <li><a href="#i_readport">'<tt>llvm.readport</tt>' Intrinsic</a></li>
143 <li><a href="#i_writeport">'<tt>llvm.writeport</tt>' Intrinsic</a></li>
John Criswella4501222004-04-12 15:02:16 +0000144 <li><a href="#i_readio">'<tt>llvm.readio</tt>' Intrinsic</a></li>
145 <li><a href="#i_writeio">'<tt>llvm.writeio</tt>' Intrinsic</a></li>
Chris Lattner3649c3a2004-02-14 04:08:35 +0000146 </ol>
Chris Lattnerfee11462004-02-12 17:01:32 +0000147 <li><a href="#int_libc">Standard C Library Intrinsics</a>
148 <ol>
149 <li><a href="#i_memcpy">'<tt>llvm.memcpy</tt>' Intrinsic</a></li>
Chris Lattnerf30152e2004-02-12 18:10:10 +0000150 <li><a href="#i_memmove">'<tt>llvm.memmove</tt>' Intrinsic</a></li>
Chris Lattner3649c3a2004-02-14 04:08:35 +0000151 <li><a href="#i_memset">'<tt>llvm.memset</tt>' Intrinsic</a></li>
Alkis Evlogimenos9d740622004-06-12 19:19:14 +0000152 <li><a href="#i_isunordered">'<tt>llvm.isunordered</tt>' Intrinsic</a></li>
Chris Lattner8a8f2e52005-07-21 01:29:16 +0000153 <li><a href="#i_sqrt">'<tt>llvm.sqrt</tt>' Intrinsic</a></li>
154
Chris Lattnerfee11462004-02-12 17:01:32 +0000155 </ol>
156 </li>
Andrew Lenharth1d463522005-05-03 18:01:48 +0000157 <li><a href="#int_count">Bit counting Intrinsics</a>
158 <ol>
159 <li><a href="#int_ctpop">'<tt>llvm.ctpop</tt>' Intrinsic </a></li>
Andrew Lenharth1d463522005-05-03 18:01:48 +0000160 <li><a href="#int_ctlz">'<tt>llvm.ctlz</tt>' Intrinsic </a></li>
Chris Lattnerefa20fa2005-05-15 19:39:26 +0000161 <li><a href="#int_cttz">'<tt>llvm.cttz</tt>' Intrinsic </a></li>
Andrew Lenharth1d463522005-05-03 18:01:48 +0000162 </ol>
163 </li>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000164 <li><a href="#int_debugger">Debugger intrinsics</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000165 </ol>
166 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000167</ol>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000168
169<div class="doc_author">
170 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>
171 and <a href="mailto:vadve@cs.uiuc.edu">Vikram Adve</a></p>
Misha Brukman76307852003-11-08 01:05:38 +0000172</div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000173
Chris Lattner2f7c9632001-06-06 20:29:01 +0000174<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000175<div class="doc_section"> <a name="abstract">Abstract </a></div>
176<!-- *********************************************************************** -->
Chris Lattner757528b0b2004-05-23 21:06:01 +0000177
Misha Brukman76307852003-11-08 01:05:38 +0000178<div class="doc_text">
Chris Lattner48b383b02003-11-25 01:02:51 +0000179<p>This document is a reference manual for the LLVM assembly language.
180LLVM is an SSA based representation that provides type safety,
181low-level operations, flexibility, and the capability of representing
182'all' high-level languages cleanly. It is the common code
183representation used throughout all phases of the LLVM compilation
184strategy.</p>
Misha Brukman76307852003-11-08 01:05:38 +0000185</div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000186
Chris Lattner2f7c9632001-06-06 20:29:01 +0000187<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000188<div class="doc_section"> <a name="introduction">Introduction</a> </div>
189<!-- *********************************************************************** -->
Chris Lattner757528b0b2004-05-23 21:06:01 +0000190
Misha Brukman76307852003-11-08 01:05:38 +0000191<div class="doc_text">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000192
Chris Lattner48b383b02003-11-25 01:02:51 +0000193<p>The LLVM code representation is designed to be used in three
194different forms: as an in-memory compiler IR, as an on-disk bytecode
195representation (suitable for fast loading by a Just-In-Time compiler),
196and as a human readable assembly language representation. This allows
197LLVM to provide a powerful intermediate representation for efficient
198compiler transformations and analysis, while providing a natural means
199to debug and visualize the transformations. The three different forms
200of LLVM are all equivalent. This document describes the human readable
201representation and notation.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000202
John Criswell4a3327e2005-05-13 22:25:59 +0000203<p>The LLVM representation aims to be light-weight and low-level
Chris Lattner48b383b02003-11-25 01:02:51 +0000204while being expressive, typed, and extensible at the same time. It
205aims to be a "universal IR" of sorts, by being at a low enough level
206that high-level ideas may be cleanly mapped to it (similar to how
207microprocessors are "universal IR's", allowing many source languages to
208be mapped to them). By providing type information, LLVM can be used as
209the target of optimizations: for example, through pointer analysis, it
210can be proven that a C automatic variable is never accessed outside of
211the current function... allowing it to be promoted to a simple SSA
212value instead of a memory location.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000213
Misha Brukman76307852003-11-08 01:05:38 +0000214</div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000215
Chris Lattner2f7c9632001-06-06 20:29:01 +0000216<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000217<div class="doc_subsubsection"> <a name="wellformed">Well-Formedness</a> </div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000218
Misha Brukman76307852003-11-08 01:05:38 +0000219<div class="doc_text">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000220
Chris Lattner48b383b02003-11-25 01:02:51 +0000221<p>It is important to note that this document describes 'well formed'
222LLVM assembly language. There is a difference between what the parser
223accepts and what is considered 'well formed'. For example, the
224following instruction is syntactically okay, but not well formed:</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000225
226<pre>
227 %x = <a href="#i_add">add</a> int 1, %x
228</pre>
229
Chris Lattner48b383b02003-11-25 01:02:51 +0000230<p>...because the definition of <tt>%x</tt> does not dominate all of
231its uses. The LLVM infrastructure provides a verification pass that may
232be used to verify that an LLVM module is well formed. This pass is
John Criswell4a3327e2005-05-13 22:25:59 +0000233automatically run by the parser after parsing input assembly and by
Chris Lattner48b383b02003-11-25 01:02:51 +0000234the optimizer before it outputs bytecode. The violations pointed out
235by the verifier pass indicate bugs in transformation passes or input to
236the parser.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000237
Chris Lattner48b383b02003-11-25 01:02:51 +0000238<!-- Describe the typesetting conventions here. --> </div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000239
Chris Lattner2f7c9632001-06-06 20:29:01 +0000240<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000241<div class="doc_section"> <a name="identifiers">Identifiers</a> </div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000242<!-- *********************************************************************** -->
Chris Lattner757528b0b2004-05-23 21:06:01 +0000243
Misha Brukman76307852003-11-08 01:05:38 +0000244<div class="doc_text">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000245
Chris Lattner48b383b02003-11-25 01:02:51 +0000246<p>LLVM uses three different forms of identifiers, for different
247purposes:</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000248
Chris Lattner2f7c9632001-06-06 20:29:01 +0000249<ol>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000250 <li>Named values are represented as a string of characters with a '%' prefix.
251 For example, %foo, %DivisionByZero, %a.really.long.identifier. The actual
252 regular expression used is '<tt>%[a-zA-Z$._][a-zA-Z$._0-9]*</tt>'.
253 Identifiers which require other characters in their names can be surrounded
254 with quotes. In this way, anything except a <tt>"</tt> character can be used
255 in a name.</li>
256
257 <li>Unnamed values are represented as an unsigned numeric value with a '%'
258 prefix. For example, %12, %2, %44.</li>
259
Reid Spencer8f08d802004-12-09 18:02:53 +0000260 <li>Constants, which are described in a <a href="#constants">section about
261 constants</a>, below.</li>
Misha Brukman76307852003-11-08 01:05:38 +0000262</ol>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000263
264<p>LLVM requires that values start with a '%' sign for two reasons: Compilers
265don't need to worry about name clashes with reserved words, and the set of
266reserved words may be expanded in the future without penalty. Additionally,
267unnamed identifiers allow a compiler to quickly come up with a temporary
268variable without having to avoid symbol table conflicts.</p>
269
Chris Lattner48b383b02003-11-25 01:02:51 +0000270<p>Reserved words in LLVM are very similar to reserved words in other
271languages. There are keywords for different opcodes ('<tt><a
Chris Lattnerd79749a2004-12-09 16:36:40 +0000272href="#i_add">add</a></tt>', '<tt><a href="#i_cast">cast</a></tt>', '<tt><a
273href="#i_ret">ret</a></tt>', etc...), for primitive type names ('<tt><a
274href="#t_void">void</a></tt>', '<tt><a href="#t_uint">uint</a></tt>', etc...),
275and others. These reserved words cannot conflict with variable names, because
276none of them start with a '%' character.</p>
277
278<p>Here is an example of LLVM code to multiply the integer variable
279'<tt>%X</tt>' by 8:</p>
280
Misha Brukman76307852003-11-08 01:05:38 +0000281<p>The easy way:</p>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000282
283<pre>
284 %result = <a href="#i_mul">mul</a> uint %X, 8
285</pre>
286
Misha Brukman76307852003-11-08 01:05:38 +0000287<p>After strength reduction:</p>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000288
289<pre>
290 %result = <a href="#i_shl">shl</a> uint %X, ubyte 3
291</pre>
292
Misha Brukman76307852003-11-08 01:05:38 +0000293<p>And the hard way:</p>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000294
295<pre>
296 <a href="#i_add">add</a> uint %X, %X <i>; yields {uint}:%0</i>
297 <a href="#i_add">add</a> uint %0, %0 <i>; yields {uint}:%1</i>
298 %result = <a href="#i_add">add</a> uint %1, %1
299</pre>
300
Chris Lattner48b383b02003-11-25 01:02:51 +0000301<p>This last way of multiplying <tt>%X</tt> by 8 illustrates several
302important lexical features of LLVM:</p>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000303
Chris Lattner2f7c9632001-06-06 20:29:01 +0000304<ol>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000305
306 <li>Comments are delimited with a '<tt>;</tt>' and go until the end of
307 line.</li>
308
309 <li>Unnamed temporaries are created when the result of a computation is not
310 assigned to a named value.</li>
311
Misha Brukman76307852003-11-08 01:05:38 +0000312 <li>Unnamed temporaries are numbered sequentially</li>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000313
Misha Brukman76307852003-11-08 01:05:38 +0000314</ol>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000315
John Criswell02fdc6f2005-05-12 16:52:32 +0000316<p>...and it also shows a convention that we follow in this document. When
Chris Lattnerd79749a2004-12-09 16:36:40 +0000317demonstrating instructions, we will follow an instruction with a comment that
318defines the type and name of value produced. Comments are shown in italic
319text.</p>
320
Misha Brukman76307852003-11-08 01:05:38 +0000321</div>
Chris Lattner6af02f32004-12-09 16:11:40 +0000322
323<!-- *********************************************************************** -->
324<div class="doc_section"> <a name="highlevel">High Level Structure</a> </div>
325<!-- *********************************************************************** -->
326
327<!-- ======================================================================= -->
328<div class="doc_subsection"> <a name="modulestructure">Module Structure</a>
329</div>
330
331<div class="doc_text">
332
333<p>LLVM programs are composed of "Module"s, each of which is a
334translation unit of the input programs. Each module consists of
335functions, global variables, and symbol table entries. Modules may be
336combined together with the LLVM linker, which merges function (and
337global variable) definitions, resolves forward declarations, and merges
338symbol table entries. Here is an example of the "hello world" module:</p>
339
340<pre><i>; Declare the string constant as a global constant...</i>
341<a href="#identifiers">%.LC0</a> = <a href="#linkage_internal">internal</a> <a
342 href="#globalvars">constant</a> <a href="#t_array">[13 x sbyte]</a> c"hello world\0A\00" <i>; [13 x sbyte]*</i>
343
344<i>; External declaration of the puts function</i>
345<a href="#functionstructure">declare</a> int %puts(sbyte*) <i>; int(sbyte*)* </i>
346
347<i>; Definition of main function</i>
348int %main() { <i>; int()* </i>
349 <i>; Convert [13x sbyte]* to sbyte *...</i>
350 %cast210 = <a
351 href="#i_getelementptr">getelementptr</a> [13 x sbyte]* %.LC0, long 0, long 0 <i>; sbyte*</i>
352
353 <i>; Call puts function to write out the string to stdout...</i>
354 <a
355 href="#i_call">call</a> int %puts(sbyte* %cast210) <i>; int</i>
356 <a
357 href="#i_ret">ret</a> int 0<br>}<br></pre>
358
359<p>This example is made up of a <a href="#globalvars">global variable</a>
360named "<tt>.LC0</tt>", an external declaration of the "<tt>puts</tt>"
361function, and a <a href="#functionstructure">function definition</a>
362for "<tt>main</tt>".</p>
363
Chris Lattnerd79749a2004-12-09 16:36:40 +0000364<p>In general, a module is made up of a list of global values,
365where both functions and global variables are global values. Global values are
366represented by a pointer to a memory location (in this case, a pointer to an
367array of char, and a pointer to a function), and have one of the following <a
368href="#linkage">linkage types</a>.</p>
Chris Lattner6af02f32004-12-09 16:11:40 +0000369
Chris Lattnerd79749a2004-12-09 16:36:40 +0000370</div>
371
372<!-- ======================================================================= -->
373<div class="doc_subsection">
374 <a name="linkage">Linkage Types</a>
375</div>
376
377<div class="doc_text">
378
379<p>
380All Global Variables and Functions have one of the following types of linkage:
381</p>
Chris Lattner6af02f32004-12-09 16:11:40 +0000382
383<dl>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000384
Chris Lattner6af02f32004-12-09 16:11:40 +0000385 <dt><tt><b><a name="linkage_internal">internal</a></b></tt> </dt>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000386
387 <dd>Global values with internal linkage are only directly accessible by
388 objects in the current module. In particular, linking code into a module with
389 an internal global value may cause the internal to be renamed as necessary to
390 avoid collisions. Because the symbol is internal to the module, all
391 references can be updated. This corresponds to the notion of the
392 '<tt>static</tt>' keyword in C, or the idea of "anonymous namespaces" in C++.
Chris Lattner6af02f32004-12-09 16:11:40 +0000393 </dd>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000394
Chris Lattner6af02f32004-12-09 16:11:40 +0000395 <dt><tt><b><a name="linkage_linkonce">linkonce</a></b></tt>: </dt>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000396
397 <dd>"<tt>linkonce</tt>" linkage is similar to <tt>internal</tt> linkage, with
398 the twist that linking together two modules defining the same
399 <tt>linkonce</tt> globals will cause one of the globals to be discarded. This
400 is typically used to implement inline functions. Unreferenced
401 <tt>linkonce</tt> globals are allowed to be discarded.
Chris Lattner6af02f32004-12-09 16:11:40 +0000402 </dd>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000403
Chris Lattner6af02f32004-12-09 16:11:40 +0000404 <dt><tt><b><a name="linkage_weak">weak</a></b></tt>: </dt>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000405
406 <dd>"<tt>weak</tt>" linkage is exactly the same as <tt>linkonce</tt> linkage,
407 except that unreferenced <tt>weak</tt> globals may not be discarded. This is
408 used to implement constructs in C such as "<tt>int X;</tt>" at global scope.
Chris Lattner6af02f32004-12-09 16:11:40 +0000409 </dd>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000410
Chris Lattner6af02f32004-12-09 16:11:40 +0000411 <dt><tt><b><a name="linkage_appending">appending</a></b></tt>: </dt>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000412
413 <dd>"<tt>appending</tt>" linkage may only be applied to global variables of
414 pointer to array type. When two global variables with appending linkage are
415 linked together, the two global arrays are appended together. This is the
416 LLVM, typesafe, equivalent of having the system linker append together
417 "sections" with identical names when .o files are linked.
Chris Lattner6af02f32004-12-09 16:11:40 +0000418 </dd>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000419
Chris Lattner6af02f32004-12-09 16:11:40 +0000420 <dt><tt><b><a name="linkage_external">externally visible</a></b></tt>:</dt>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000421
422 <dd>If none of the above identifiers are used, the global is externally
423 visible, meaning that it participates in linkage and can be used to resolve
424 external symbol references.
Chris Lattner6af02f32004-12-09 16:11:40 +0000425 </dd>
426</dl>
427
Chris Lattner6af02f32004-12-09 16:11:40 +0000428<p><a name="linkage_external">For example, since the "<tt>.LC0</tt>"
429variable is defined to be internal, if another module defined a "<tt>.LC0</tt>"
430variable and was linked with this one, one of the two would be renamed,
431preventing a collision. Since "<tt>main</tt>" and "<tt>puts</tt>" are
432external (i.e., lacking any linkage declarations), they are accessible
433outside of the current module. It is illegal for a function <i>declaration</i>
434to have any linkage type other than "externally visible".</a></p>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000435
Chris Lattner6af02f32004-12-09 16:11:40 +0000436</div>
437
438<!-- ======================================================================= -->
439<div class="doc_subsection">
Chris Lattner0132aff2005-05-06 22:57:40 +0000440 <a name="callingconv">Calling Conventions</a>
441</div>
442
443<div class="doc_text">
444
445<p>LLVM <a href="#functionstructure">functions</a>, <a href="#i_call">calls</a>
446and <a href="#i_invoke">invokes</a> can all have an optional calling convention
447specified for the call. The calling convention of any pair of dynamic
448caller/callee must match, or the behavior of the program is undefined. The
449following calling conventions are supported by LLVM, and more may be added in
450the future:</p>
451
452<dl>
453 <dt><b>"<tt>ccc</tt>" - The C calling convention</b>:</dt>
454
455 <dd>This calling convention (the default if no other calling convention is
456 specified) matches the target C calling conventions. This calling convention
John Criswell02fdc6f2005-05-12 16:52:32 +0000457 supports varargs function calls and tolerates some mismatch in the declared
Chris Lattner0132aff2005-05-06 22:57:40 +0000458 prototype and implemented declaration of the function (as does normal C).
459 </dd>
460
461 <dt><b>"<tt>fastcc</tt>" - The fast calling convention</b>:</dt>
462
463 <dd>This calling convention attempts to make calls as fast as possible
464 (e.g. by passing things in registers). This calling convention allows the
465 target to use whatever tricks it wants to produce fast code for the target,
Chris Lattnerc792eb32005-05-06 23:08:23 +0000466 without having to conform to an externally specified ABI. Implementations of
467 this convention should allow arbitrary tail call optimization to be supported.
468 This calling convention does not support varargs and requires the prototype of
469 all callees to exactly match the prototype of the function definition.
Chris Lattner0132aff2005-05-06 22:57:40 +0000470 </dd>
471
472 <dt><b>"<tt>coldcc</tt>" - The cold calling convention</b>:</dt>
473
474 <dd>This calling convention attempts to make code in the caller as efficient
475 as possible under the assumption that the call is not commonly executed. As
476 such, these calls often preserve all registers so that the call does not break
477 any live ranges in the caller side. This calling convention does not support
478 varargs and requires the prototype of all callees to exactly match the
479 prototype of the function definition.
480 </dd>
481
Chris Lattner573f64e2005-05-07 01:46:40 +0000482 <dt><b>"<tt>cc &lt;<em>n</em>&gt;</tt>" - Numbered convention</b>:</dt>
Chris Lattner0132aff2005-05-06 22:57:40 +0000483
484 <dd>Any calling convention may be specified by number, allowing
485 target-specific calling conventions to be used. Target specific calling
486 conventions start at 64.
487 </dd>
Chris Lattner573f64e2005-05-07 01:46:40 +0000488</dl>
Chris Lattner0132aff2005-05-06 22:57:40 +0000489
490<p>More calling conventions can be added/defined on an as-needed basis, to
491support pascal conventions or any other well-known target-independent
492convention.</p>
493
494</div>
495
496<!-- ======================================================================= -->
497<div class="doc_subsection">
Chris Lattner6af02f32004-12-09 16:11:40 +0000498 <a name="globalvars">Global Variables</a>
499</div>
500
501<div class="doc_text">
502
Chris Lattner5d5aede2005-02-12 19:30:21 +0000503<p>Global variables define regions of memory allocated at compilation time
Chris Lattner662c8722005-11-12 00:45:07 +0000504instead of run-time. Global variables may optionally be initialized, may have
505an explicit section to be placed in, and may
Chris Lattner54611b42005-11-06 08:02:57 +0000506have an optional explicit alignment specified. A
John Criswell4c0cf7f2005-10-24 16:17:18 +0000507variable may be defined as a global "constant," which indicates that the
Chris Lattner5d5aede2005-02-12 19:30:21 +0000508contents of the variable will <b>never</b> be modified (enabling better
509optimization, allowing the global data to be placed in the read-only section of
510an executable, etc). Note that variables that need runtime initialization
John Criswell4c0cf7f2005-10-24 16:17:18 +0000511cannot be marked "constant" as there is a store to the variable.</p>
Chris Lattner5d5aede2005-02-12 19:30:21 +0000512
513<p>
514LLVM explicitly allows <em>declarations</em> of global variables to be marked
515constant, even if the final definition of the global is not. This capability
516can be used to enable slightly better optimization of the program, but requires
517the language definition to guarantee that optimizations based on the
518'constantness' are valid for the translation units that do not include the
519definition.
520</p>
Chris Lattner6af02f32004-12-09 16:11:40 +0000521
522<p>As SSA values, global variables define pointer values that are in
523scope (i.e. they dominate) all basic blocks in the program. Global
524variables always define a pointer to their "content" type because they
525describe a region of memory, and all memory objects in LLVM are
526accessed through pointers.</p>
527
Chris Lattner662c8722005-11-12 00:45:07 +0000528<p>LLVM allows an explicit section to be specified for globals. If the target
529supports it, it will emit globals to the section specified.</p>
530
Chris Lattner54611b42005-11-06 08:02:57 +0000531<p>An explicit alignment may be specified for a global. If not present, or if
532the alignment is set to zero, the alignment of the global is set by the target
533to whatever it feels convenient. If an explicit alignment is specified, the
534global is forced to have at least that much alignment. All alignments must be
535a power of 2.</p>
536
Chris Lattner6af02f32004-12-09 16:11:40 +0000537</div>
538
539
540<!-- ======================================================================= -->
541<div class="doc_subsection">
542 <a name="functionstructure">Functions</a>
543</div>
544
545<div class="doc_text">
546
Chris Lattner0132aff2005-05-06 22:57:40 +0000547<p>LLVM function definitions consist of an optional <a href="#linkage">linkage
548type</a>, an optional <a href="#callingconv">calling convention</a>, a return
Chris Lattner662c8722005-11-12 00:45:07 +0000549type, a function name, a (possibly empty) argument list, an optional section,
550an optional alignment, an opening curly brace,
Chris Lattner0132aff2005-05-06 22:57:40 +0000551a list of basic blocks, and a closing curly brace. LLVM function declarations
552are defined with the "<tt>declare</tt>" keyword, an optional <a
Chris Lattner54611b42005-11-06 08:02:57 +0000553href="#callingconv">calling convention</a>, a return type, a function name,
554a possibly empty list of arguments, and an optional alignment.</p>
Chris Lattner6af02f32004-12-09 16:11:40 +0000555
556<p>A function definition contains a list of basic blocks, forming the CFG for
557the function. Each basic block may optionally start with a label (giving the
558basic block a symbol table entry), contains a list of instructions, and ends
559with a <a href="#terminators">terminator</a> instruction (such as a branch or
560function return).</p>
561
John Criswell02fdc6f2005-05-12 16:52:32 +0000562<p>The first basic block in a program is special in two ways: it is immediately
Chris Lattner6af02f32004-12-09 16:11:40 +0000563executed on entrance to the function, and it is not allowed to have predecessor
564basic blocks (i.e. there can not be any branches to the entry block of a
565function). Because the block can have no predecessors, it also cannot have any
566<a href="#i_phi">PHI nodes</a>.</p>
567
568<p>LLVM functions are identified by their name and type signature. Hence, two
569functions with the same name but different parameter lists or return values are
Chris Lattner455fc8c2005-03-07 22:13:59 +0000570considered different functions, and LLVM will resolve references to each
Chris Lattner6af02f32004-12-09 16:11:40 +0000571appropriately.</p>
572
Chris Lattner662c8722005-11-12 00:45:07 +0000573<p>LLVM allows an explicit section to be specified for functions. If the target
574supports it, it will emit functions to the section specified.</p>
575
Chris Lattner54611b42005-11-06 08:02:57 +0000576<p>An explicit alignment may be specified for a function. If not present, or if
577the alignment is set to zero, the alignment of the function is set by the target
578to whatever it feels convenient. If an explicit alignment is specified, the
579function is forced to have at least that much alignment. All alignments must be
580a power of 2.</p>
581
Chris Lattner6af02f32004-12-09 16:11:40 +0000582</div>
583
584
585
Chris Lattner2f7c9632001-06-06 20:29:01 +0000586<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000587<div class="doc_section"> <a name="typesystem">Type System</a> </div>
588<!-- *********************************************************************** -->
Chris Lattner6af02f32004-12-09 16:11:40 +0000589
Misha Brukman76307852003-11-08 01:05:38 +0000590<div class="doc_text">
Chris Lattner6af02f32004-12-09 16:11:40 +0000591
Misha Brukman76307852003-11-08 01:05:38 +0000592<p>The LLVM type system is one of the most important features of the
Chris Lattner48b383b02003-11-25 01:02:51 +0000593intermediate representation. Being typed enables a number of
594optimizations to be performed on the IR directly, without having to do
595extra analyses on the side before the transformation. A strong type
596system makes it easier to read the generated code and enables novel
597analyses and transformations that are not feasible to perform on normal
598three address code representations.</p>
Chris Lattner6af02f32004-12-09 16:11:40 +0000599
600</div>
601
Chris Lattner2f7c9632001-06-06 20:29:01 +0000602<!-- ======================================================================= -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000603<div class="doc_subsection"> <a name="t_primitive">Primitive Types</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000604<div class="doc_text">
John Criswell417228d2004-04-09 16:48:45 +0000605<p>The primitive types are the fundamental building blocks of the LLVM
Chris Lattner455fc8c2005-03-07 22:13:59 +0000606system. The current set of primitive types is as follows:</p>
Misha Brukmanc501f552004-03-01 17:47:27 +0000607
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000608<table class="layout">
609 <tr class="layout">
610 <td class="left">
611 <table>
Chris Lattner48b383b02003-11-25 01:02:51 +0000612 <tbody>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000613 <tr><th>Type</th><th>Description</th></tr>
614 <tr><td><tt>void</tt></td><td>No value</td></tr>
Misha Brukman36c6bc12005-04-22 18:02:52 +0000615 <tr><td><tt>ubyte</tt></td><td>Unsigned 8-bit value</td></tr>
616 <tr><td><tt>ushort</tt></td><td>Unsigned 16-bit value</td></tr>
617 <tr><td><tt>uint</tt></td><td>Unsigned 32-bit value</td></tr>
618 <tr><td><tt>ulong</tt></td><td>Unsigned 64-bit value</td></tr>
619 <tr><td><tt>float</tt></td><td>32-bit floating point value</td></tr>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000620 <tr><td><tt>label</tt></td><td>Branch destination</td></tr>
Chris Lattner48b383b02003-11-25 01:02:51 +0000621 </tbody>
622 </table>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000623 </td>
624 <td class="right">
625 <table>
Chris Lattner48b383b02003-11-25 01:02:51 +0000626 <tbody>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000627 <tr><th>Type</th><th>Description</th></tr>
628 <tr><td><tt>bool</tt></td><td>True or False value</td></tr>
Misha Brukman36c6bc12005-04-22 18:02:52 +0000629 <tr><td><tt>sbyte</tt></td><td>Signed 8-bit value</td></tr>
630 <tr><td><tt>short</tt></td><td>Signed 16-bit value</td></tr>
631 <tr><td><tt>int</tt></td><td>Signed 32-bit value</td></tr>
632 <tr><td><tt>long</tt></td><td>Signed 64-bit value</td></tr>
633 <tr><td><tt>double</tt></td><td>64-bit floating point value</td></tr>
Chris Lattner48b383b02003-11-25 01:02:51 +0000634 </tbody>
635 </table>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000636 </td>
637 </tr>
Misha Brukman76307852003-11-08 01:05:38 +0000638</table>
Misha Brukman76307852003-11-08 01:05:38 +0000639</div>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000640
Chris Lattner2f7c9632001-06-06 20:29:01 +0000641<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000642<div class="doc_subsubsection"> <a name="t_classifications">Type
643Classifications</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000644<div class="doc_text">
Chris Lattner48b383b02003-11-25 01:02:51 +0000645<p>These different primitive types fall into a few useful
646classifications:</p>
Misha Brukmanc501f552004-03-01 17:47:27 +0000647
648<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner48b383b02003-11-25 01:02:51 +0000649 <tbody>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000650 <tr><th>Classification</th><th>Types</th></tr>
Chris Lattner48b383b02003-11-25 01:02:51 +0000651 <tr>
652 <td><a name="t_signed">signed</a></td>
653 <td><tt>sbyte, short, int, long, float, double</tt></td>
654 </tr>
655 <tr>
656 <td><a name="t_unsigned">unsigned</a></td>
657 <td><tt>ubyte, ushort, uint, ulong</tt></td>
658 </tr>
659 <tr>
660 <td><a name="t_integer">integer</a></td>
661 <td><tt>ubyte, sbyte, ushort, short, uint, int, ulong, long</tt></td>
662 </tr>
663 <tr>
664 <td><a name="t_integral">integral</a></td>
Misha Brukman20f9a622004-08-12 20:16:08 +0000665 <td><tt>bool, ubyte, sbyte, ushort, short, uint, int, ulong, long</tt>
666 </td>
Chris Lattner48b383b02003-11-25 01:02:51 +0000667 </tr>
668 <tr>
669 <td><a name="t_floating">floating point</a></td>
670 <td><tt>float, double</tt></td>
671 </tr>
672 <tr>
673 <td><a name="t_firstclass">first class</a></td>
Misha Brukman20f9a622004-08-12 20:16:08 +0000674 <td><tt>bool, ubyte, sbyte, ushort, short, uint, int, ulong, long,<br>
675 float, double, <a href="#t_pointer">pointer</a>,
676 <a href="#t_packed">packed</a></tt></td>
Chris Lattner48b383b02003-11-25 01:02:51 +0000677 </tr>
678 </tbody>
Misha Brukman76307852003-11-08 01:05:38 +0000679</table>
Misha Brukmanc501f552004-03-01 17:47:27 +0000680
Chris Lattner48b383b02003-11-25 01:02:51 +0000681<p>The <a href="#t_firstclass">first class</a> types are perhaps the
682most important. Values of these types are the only ones which can be
683produced by instructions, passed as arguments, or used as operands to
684instructions. This means that all structures and arrays must be
685manipulated either by pointer or by component.</p>
Misha Brukman76307852003-11-08 01:05:38 +0000686</div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000687
Chris Lattner2f7c9632001-06-06 20:29:01 +0000688<!-- ======================================================================= -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000689<div class="doc_subsection"> <a name="t_derived">Derived Types</a> </div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000690
Misha Brukman76307852003-11-08 01:05:38 +0000691<div class="doc_text">
Chris Lattner74d3f822004-12-09 17:30:23 +0000692
Chris Lattner48b383b02003-11-25 01:02:51 +0000693<p>The real power in LLVM comes from the derived types in the system.
694This is what allows a programmer to represent arrays, functions,
695pointers, and other useful types. Note that these derived types may be
696recursive: For example, it is possible to have a two dimensional array.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000697
Misha Brukman76307852003-11-08 01:05:38 +0000698</div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000699
Chris Lattner2f7c9632001-06-06 20:29:01 +0000700<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000701<div class="doc_subsubsection"> <a name="t_array">Array Type</a> </div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000702
Misha Brukman76307852003-11-08 01:05:38 +0000703<div class="doc_text">
Chris Lattner74d3f822004-12-09 17:30:23 +0000704
Chris Lattner2f7c9632001-06-06 20:29:01 +0000705<h5>Overview:</h5>
Chris Lattner74d3f822004-12-09 17:30:23 +0000706
Misha Brukman76307852003-11-08 01:05:38 +0000707<p>The array type is a very simple derived type that arranges elements
Chris Lattner48b383b02003-11-25 01:02:51 +0000708sequentially in memory. The array type requires a size (number of
709elements) and an underlying data type.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000710
Chris Lattner590645f2002-04-14 06:13:44 +0000711<h5>Syntax:</h5>
Chris Lattner74d3f822004-12-09 17:30:23 +0000712
713<pre>
714 [&lt;# elements&gt; x &lt;elementtype&gt;]
715</pre>
716
John Criswell02fdc6f2005-05-12 16:52:32 +0000717<p>The number of elements is a constant integer value; elementtype may
Chris Lattner48b383b02003-11-25 01:02:51 +0000718be any type with a size.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000719
Chris Lattner590645f2002-04-14 06:13:44 +0000720<h5>Examples:</h5>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000721<table class="layout">
722 <tr class="layout">
723 <td class="left">
724 <tt>[40 x int ]</tt><br/>
725 <tt>[41 x int ]</tt><br/>
726 <tt>[40 x uint]</tt><br/>
727 </td>
728 <td class="left">
729 Array of 40 integer values.<br/>
730 Array of 41 integer values.<br/>
731 Array of 40 unsigned integer values.<br/>
732 </td>
733 </tr>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000734</table>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000735<p>Here are some examples of multidimensional arrays:</p>
736<table class="layout">
737 <tr class="layout">
738 <td class="left">
739 <tt>[3 x [4 x int]]</tt><br/>
740 <tt>[12 x [10 x float]]</tt><br/>
741 <tt>[2 x [3 x [4 x uint]]]</tt><br/>
742 </td>
743 <td class="left">
John Criswell4a3327e2005-05-13 22:25:59 +0000744 3x4 array of integer values.<br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000745 12x10 array of single precision floating point values.<br/>
746 2x3x4 array of unsigned integer values.<br/>
747 </td>
748 </tr>
749</table>
Chris Lattnerc0ad71e2005-06-24 17:22:57 +0000750
John Criswell4c0cf7f2005-10-24 16:17:18 +0000751<p>Note that 'variable sized arrays' can be implemented in LLVM with a zero
752length array. Normally, accesses past the end of an array are undefined in
Chris Lattnerc0ad71e2005-06-24 17:22:57 +0000753LLVM (e.g. it is illegal to access the 5th element of a 3 element array).
754As a special case, however, zero length arrays are recognized to be variable
755length. This allows implementation of 'pascal style arrays' with the LLVM
756type "{ int, [0 x float]}", for example.</p>
757
Misha Brukman76307852003-11-08 01:05:38 +0000758</div>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000759
Chris Lattner2f7c9632001-06-06 20:29:01 +0000760<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000761<div class="doc_subsubsection"> <a name="t_function">Function Type</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000762<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000763<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000764<p>The function type can be thought of as a function signature. It
765consists of a return type and a list of formal parameter types.
John Criswella0d50d22003-11-25 21:45:46 +0000766Function types are usually used to build virtual function tables
Chris Lattner48b383b02003-11-25 01:02:51 +0000767(which are structures of pointers to functions), for indirect function
768calls, and when defining a function.</p>
John Criswella0d50d22003-11-25 21:45:46 +0000769<p>
770The return type of a function type cannot be an aggregate type.
771</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000772<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000773<pre> &lt;returntype&gt; (&lt;parameter list&gt;)<br></pre>
John Criswell4c0cf7f2005-10-24 16:17:18 +0000774<p>...where '<tt>&lt;parameter list&gt;</tt>' is a comma-separated list of type
Misha Brukman20f9a622004-08-12 20:16:08 +0000775specifiers. Optionally, the parameter list may include a type <tt>...</tt>,
Chris Lattner5ed60612003-09-03 00:41:47 +0000776which indicates that the function takes a variable number of arguments.
777Variable argument functions can access their arguments with the <a
Chris Lattner48b383b02003-11-25 01:02:51 +0000778 href="#int_varargs">variable argument handling intrinsic</a> functions.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000779<h5>Examples:</h5>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000780<table class="layout">
781 <tr class="layout">
782 <td class="left">
783 <tt>int (int)</tt> <br/>
784 <tt>float (int, int *) *</tt><br/>
785 <tt>int (sbyte *, ...)</tt><br/>
786 </td>
787 <td class="left">
788 function taking an <tt>int</tt>, returning an <tt>int</tt><br/>
789 <a href="#t_pointer">Pointer</a> to a function that takes an
Misha Brukman20f9a622004-08-12 20:16:08 +0000790 <tt>int</tt> and a <a href="#t_pointer">pointer</a> to <tt>int</tt>,
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000791 returning <tt>float</tt>.<br/>
792 A vararg function that takes at least one <a href="#t_pointer">pointer</a>
793 to <tt>sbyte</tt> (signed char in C), which returns an integer. This is
794 the signature for <tt>printf</tt> in LLVM.<br/>
795 </td>
796 </tr>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000797</table>
Misha Brukmanc501f552004-03-01 17:47:27 +0000798
Misha Brukman76307852003-11-08 01:05:38 +0000799</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000800<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000801<div class="doc_subsubsection"> <a name="t_struct">Structure Type</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000802<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000803<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000804<p>The structure type is used to represent a collection of data members
805together in memory. The packing of the field types is defined to match
806the ABI of the underlying processor. The elements of a structure may
807be any type that has a size.</p>
808<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
809and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
810field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
811instruction.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000812<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000813<pre> { &lt;type list&gt; }<br></pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000814<h5>Examples:</h5>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000815<table class="layout">
816 <tr class="layout">
817 <td class="left">
818 <tt>{ int, int, int }</tt><br/>
819 <tt>{ float, int (int) * }</tt><br/>
820 </td>
821 <td class="left">
822 a triple of three <tt>int</tt> values<br/>
823 A pair, where the first element is a <tt>float</tt> and the second element
824 is a <a href="#t_pointer">pointer</a> to a <a href="#t_function">function</a>
825 that takes an <tt>int</tt>, returning an <tt>int</tt>.<br/>
826 </td>
827 </tr>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000828</table>
Misha Brukman76307852003-11-08 01:05:38 +0000829</div>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000830
Chris Lattner2f7c9632001-06-06 20:29:01 +0000831<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000832<div class="doc_subsubsection"> <a name="t_pointer">Pointer Type</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000833<div class="doc_text">
Chris Lattner590645f2002-04-14 06:13:44 +0000834<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000835<p>As in many languages, the pointer type represents a pointer or
836reference to another object, which must live in memory.</p>
Chris Lattner590645f2002-04-14 06:13:44 +0000837<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000838<pre> &lt;type&gt; *<br></pre>
Chris Lattner590645f2002-04-14 06:13:44 +0000839<h5>Examples:</h5>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000840<table class="layout">
841 <tr class="layout">
842 <td class="left">
843 <tt>[4x int]*</tt><br/>
844 <tt>int (int *) *</tt><br/>
845 </td>
846 <td class="left">
847 A <a href="#t_pointer">pointer</a> to <a href="#t_array">array</a> of
848 four <tt>int</tt> values<br/>
849 A <a href="#t_pointer">pointer</a> to a <a
Chris Lattner16fb0032005-02-19 02:22:14 +0000850 href="#t_function">function</a> that takes an <tt>int*</tt>, returning an
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000851 <tt>int</tt>.<br/>
852 </td>
853 </tr>
Misha Brukman76307852003-11-08 01:05:38 +0000854</table>
Misha Brukman76307852003-11-08 01:05:38 +0000855</div>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000856
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000857<!-- _______________________________________________________________________ -->
858<div class="doc_subsubsection"> <a name="t_packed">Packed Type</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000859<div class="doc_text">
Chris Lattner37b6b092005-04-25 17:34:15 +0000860
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000861<h5>Overview:</h5>
Chris Lattner37b6b092005-04-25 17:34:15 +0000862
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000863<p>A packed type is a simple derived type that represents a vector
864of elements. Packed types are used when multiple primitive data
865are operated in parallel using a single instruction (SIMD).
866A packed type requires a size (number of
Chris Lattner330ce692005-11-10 01:44:22 +0000867elements) and an underlying primitive data type. Vectors must have a power
868of two length (1, 2, 4, 8, 16 ...). Packed types are
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000869considered <a href="#t_firstclass">first class</a>.</p>
Chris Lattner37b6b092005-04-25 17:34:15 +0000870
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000871<h5>Syntax:</h5>
Chris Lattner37b6b092005-04-25 17:34:15 +0000872
873<pre>
874 &lt; &lt;# elements&gt; x &lt;elementtype&gt; &gt;
875</pre>
876
John Criswell4a3327e2005-05-13 22:25:59 +0000877<p>The number of elements is a constant integer value; elementtype may
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000878be any integral or floating point type.</p>
Chris Lattner37b6b092005-04-25 17:34:15 +0000879
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000880<h5>Examples:</h5>
Chris Lattner37b6b092005-04-25 17:34:15 +0000881
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000882<table class="layout">
883 <tr class="layout">
884 <td class="left">
885 <tt>&lt;4 x int&gt;</tt><br/>
886 <tt>&lt;8 x float&gt;</tt><br/>
887 <tt>&lt;2 x uint&gt;</tt><br/>
888 </td>
889 <td class="left">
890 Packed vector of 4 integer values.<br/>
891 Packed vector of 8 floating-point values.<br/>
892 Packed vector of 2 unsigned integer values.<br/>
893 </td>
894 </tr>
895</table>
Misha Brukman76307852003-11-08 01:05:38 +0000896</div>
897
Chris Lattner37b6b092005-04-25 17:34:15 +0000898<!-- _______________________________________________________________________ -->
899<div class="doc_subsubsection"> <a name="t_opaque">Opaque Type</a> </div>
900<div class="doc_text">
901
902<h5>Overview:</h5>
903
904<p>Opaque types are used to represent unknown types in the system. This
905corresponds (for example) to the C notion of a foward declared structure type.
906In LLVM, opaque types can eventually be resolved to any type (not just a
907structure type).</p>
908
909<h5>Syntax:</h5>
910
911<pre>
912 opaque
913</pre>
914
915<h5>Examples:</h5>
916
917<table class="layout">
918 <tr class="layout">
919 <td class="left">
920 <tt>opaque</tt>
921 </td>
922 <td class="left">
923 An opaque type.<br/>
924 </td>
925 </tr>
926</table>
927</div>
928
929
Chris Lattner74d3f822004-12-09 17:30:23 +0000930<!-- *********************************************************************** -->
931<div class="doc_section"> <a name="constants">Constants</a> </div>
932<!-- *********************************************************************** -->
933
934<div class="doc_text">
935
936<p>LLVM has several different basic types of constants. This section describes
937them all and their syntax.</p>
938
939</div>
940
941<!-- ======================================================================= -->
Reid Spencer8f08d802004-12-09 18:02:53 +0000942<div class="doc_subsection"><a name="simpleconstants">Simple Constants</a></div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000943
944<div class="doc_text">
945
946<dl>
947 <dt><b>Boolean constants</b></dt>
948
949 <dd>The two strings '<tt>true</tt>' and '<tt>false</tt>' are both valid
950 constants of the <tt><a href="#t_primitive">bool</a></tt> type.
951 </dd>
952
953 <dt><b>Integer constants</b></dt>
954
Reid Spencer8f08d802004-12-09 18:02:53 +0000955 <dd>Standard integers (such as '4') are constants of the <a
Chris Lattner74d3f822004-12-09 17:30:23 +0000956 href="#t_integer">integer</a> type. Negative numbers may be used with signed
957 integer types.
958 </dd>
959
960 <dt><b>Floating point constants</b></dt>
961
962 <dd>Floating point constants use standard decimal notation (e.g. 123.421),
963 exponential notation (e.g. 1.23421e+2), or a more precise hexadecimal
Chris Lattner74d3f822004-12-09 17:30:23 +0000964 notation (see below). Floating point constants must have a <a
965 href="#t_floating">floating point</a> type. </dd>
966
967 <dt><b>Null pointer constants</b></dt>
968
John Criswelldfe6a862004-12-10 15:51:16 +0000969 <dd>The identifier '<tt>null</tt>' is recognized as a null pointer constant
Chris Lattner74d3f822004-12-09 17:30:23 +0000970 and must be of <a href="#t_pointer">pointer type</a>.</dd>
971
972</dl>
973
John Criswelldfe6a862004-12-10 15:51:16 +0000974<p>The one non-intuitive notation for constants is the optional hexadecimal form
Chris Lattner74d3f822004-12-09 17:30:23 +0000975of floating point constants. For example, the form '<tt>double
9760x432ff973cafa8000</tt>' is equivalent to (but harder to read than) '<tt>double
9774.5e+15</tt>'. The only time hexadecimal floating point constants are required
Reid Spencer8f08d802004-12-09 18:02:53 +0000978(and the only time that they are generated by the disassembler) is when a
979floating point constant must be emitted but it cannot be represented as a
980decimal floating point number. For example, NaN's, infinities, and other
981special values are represented in their IEEE hexadecimal format so that
982assembly and disassembly do not cause any bits to change in the constants.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000983
984</div>
985
986<!-- ======================================================================= -->
987<div class="doc_subsection"><a name="aggregateconstants">Aggregate Constants</a>
988</div>
989
990<div class="doc_text">
Chris Lattner455fc8c2005-03-07 22:13:59 +0000991<p>Aggregate constants arise from aggregation of simple constants
992and smaller aggregate constants.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000993
994<dl>
995 <dt><b>Structure constants</b></dt>
996
997 <dd>Structure constants are represented with notation similar to structure
998 type definitions (a comma separated list of elements, surrounded by braces
Chris Lattner455fc8c2005-03-07 22:13:59 +0000999 (<tt>{}</tt>)). For example: "<tt>{ int 4, float 17.0, int* %G }</tt>",
1000 where "<tt>%G</tt>" is declared as "<tt>%G = external global int</tt>". Structure constants
1001 must have <a href="#t_struct">structure type</a>, and the number and
Chris Lattner74d3f822004-12-09 17:30:23 +00001002 types of elements must match those specified by the type.
1003 </dd>
1004
1005 <dt><b>Array constants</b></dt>
1006
1007 <dd>Array constants are represented with notation similar to array type
1008 definitions (a comma separated list of elements, surrounded by square brackets
John Criswelldfe6a862004-12-10 15:51:16 +00001009 (<tt>[]</tt>)). For example: "<tt>[ int 42, int 11, int 74 ]</tt>". Array
Chris Lattner74d3f822004-12-09 17:30:23 +00001010 constants must have <a href="#t_array">array type</a>, and the number and
1011 types of elements must match those specified by the type.
1012 </dd>
1013
1014 <dt><b>Packed constants</b></dt>
1015
1016 <dd>Packed constants are represented with notation similar to packed type
1017 definitions (a comma separated list of elements, surrounded by
John Criswelldfe6a862004-12-10 15:51:16 +00001018 less-than/greater-than's (<tt>&lt;&gt;</tt>)). For example: "<tt>&lt; int 42,
Chris Lattner74d3f822004-12-09 17:30:23 +00001019 int 11, int 74, int 100 &gt;</tt>". Packed constants must have <a
1020 href="#t_packed">packed type</a>, and the number and types of elements must
1021 match those specified by the type.
1022 </dd>
1023
1024 <dt><b>Zero initialization</b></dt>
1025
1026 <dd>The string '<tt>zeroinitializer</tt>' can be used to zero initialize a
1027 value to zero of <em>any</em> type, including scalar and aggregate types.
1028 This is often used to avoid having to print large zero initializers (e.g. for
John Criswell4c0cf7f2005-10-24 16:17:18 +00001029 large arrays) and is always exactly equivalent to using explicit zero
Chris Lattner74d3f822004-12-09 17:30:23 +00001030 initializers.
1031 </dd>
1032</dl>
1033
1034</div>
1035
1036<!-- ======================================================================= -->
1037<div class="doc_subsection">
1038 <a name="globalconstants">Global Variable and Function Addresses</a>
1039</div>
1040
1041<div class="doc_text">
1042
1043<p>The addresses of <a href="#globalvars">global variables</a> and <a
1044href="#functionstructure">functions</a> are always implicitly valid (link-time)
John Criswelldfe6a862004-12-10 15:51:16 +00001045constants. These constants are explicitly referenced when the <a
1046href="#identifiers">identifier for the global</a> is used and always have <a
Chris Lattner74d3f822004-12-09 17:30:23 +00001047href="#t_pointer">pointer</a> type. For example, the following is a legal LLVM
1048file:</p>
1049
1050<pre>
1051 %X = global int 17
1052 %Y = global int 42
1053 %Z = global [2 x int*] [ int* %X, int* %Y ]
1054</pre>
1055
1056</div>
1057
1058<!-- ======================================================================= -->
Reid Spencer641f5c92004-12-09 18:13:12 +00001059<div class="doc_subsection"><a name="undefvalues">Undefined Values</a></div>
Chris Lattner74d3f822004-12-09 17:30:23 +00001060<div class="doc_text">
Reid Spencer641f5c92004-12-09 18:13:12 +00001061 <p>The string '<tt>undef</tt>' is recognized as a type-less constant that has
John Criswell4a3327e2005-05-13 22:25:59 +00001062 no specific value. Undefined values may be of any type and be used anywhere
Reid Spencer641f5c92004-12-09 18:13:12 +00001063 a constant is permitted.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +00001064
Reid Spencer641f5c92004-12-09 18:13:12 +00001065 <p>Undefined values indicate to the compiler that the program is well defined
1066 no matter what value is used, giving the compiler more freedom to optimize.
1067 </p>
Chris Lattner74d3f822004-12-09 17:30:23 +00001068</div>
1069
1070<!-- ======================================================================= -->
1071<div class="doc_subsection"><a name="constantexprs">Constant Expressions</a>
1072</div>
1073
1074<div class="doc_text">
1075
1076<p>Constant expressions are used to allow expressions involving other constants
1077to be used as constants. Constant expressions may be of any <a
John Criswell4a3327e2005-05-13 22:25:59 +00001078href="#t_firstclass">first class</a> type and may involve any LLVM operation
Chris Lattner74d3f822004-12-09 17:30:23 +00001079that does not have side effects (e.g. load and call are not supported). The
1080following is the syntax for constant expressions:</p>
1081
1082<dl>
1083 <dt><b><tt>cast ( CST to TYPE )</tt></b></dt>
1084
1085 <dd>Cast a constant to another type.</dd>
1086
1087 <dt><b><tt>getelementptr ( CSTPTR, IDX0, IDX1, ... )</tt></b></dt>
1088
1089 <dd>Perform the <a href="#i_getelementptr">getelementptr operation</a> on
1090 constants. As with the <a href="#i_getelementptr">getelementptr</a>
1091 instruction, the index list may have zero or more indexes, which are required
1092 to make sense for the type of "CSTPTR".</dd>
1093
Robert Bocchino7e97a6d2006-01-10 19:31:34 +00001094 <dt><b><tt>select ( COND, VAL1, VAL2 )</tt></b></dt>
1095
1096 <dd>Perform the <a href="#i_select">select operation</a> on
1097 constants.
1098
1099 <dt><b><tt>extractelement ( VAL, IDX )</tt></b></dt>
1100
1101 <dd>Perform the <a href="#i_extractelement">extractelement
1102 operation</a> on constants.
1103
Chris Lattner74d3f822004-12-09 17:30:23 +00001104 <dt><b><tt>OPCODE ( LHS, RHS )</tt></b></dt>
1105
Reid Spencer641f5c92004-12-09 18:13:12 +00001106 <dd>Perform the specified operation of the LHS and RHS constants. OPCODE may
1107 be any of the <a href="#binaryops">binary</a> or <a href="#bitwiseops">bitwise
Chris Lattner74d3f822004-12-09 17:30:23 +00001108 binary</a> operations. The constraints on operands are the same as those for
1109 the corresponding instruction (e.g. no bitwise operations on floating point
John Criswell02fdc6f2005-05-12 16:52:32 +00001110 values are allowed).</dd>
Chris Lattner74d3f822004-12-09 17:30:23 +00001111</dl>
Chris Lattner74d3f822004-12-09 17:30:23 +00001112</div>
Chris Lattnerb1652612004-03-08 16:49:10 +00001113
Chris Lattner2f7c9632001-06-06 20:29:01 +00001114<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +00001115<div class="doc_section"> <a name="instref">Instruction Reference</a> </div>
1116<!-- *********************************************************************** -->
Chris Lattner74d3f822004-12-09 17:30:23 +00001117
Misha Brukman76307852003-11-08 01:05:38 +00001118<div class="doc_text">
Chris Lattner74d3f822004-12-09 17:30:23 +00001119
Chris Lattner48b383b02003-11-25 01:02:51 +00001120<p>The LLVM instruction set consists of several different
1121classifications of instructions: <a href="#terminators">terminator
John Criswell4a3327e2005-05-13 22:25:59 +00001122instructions</a>, <a href="#binaryops">binary instructions</a>,
1123<a href="#bitwiseops">bitwise binary instructions</a>, <a
Chris Lattner48b383b02003-11-25 01:02:51 +00001124 href="#memoryops">memory instructions</a>, and <a href="#otherops">other
1125instructions</a>.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +00001126
Misha Brukman76307852003-11-08 01:05:38 +00001127</div>
Chris Lattner74d3f822004-12-09 17:30:23 +00001128
Chris Lattner2f7c9632001-06-06 20:29:01 +00001129<!-- ======================================================================= -->
Chris Lattner48b383b02003-11-25 01:02:51 +00001130<div class="doc_subsection"> <a name="terminators">Terminator
1131Instructions</a> </div>
Chris Lattner74d3f822004-12-09 17:30:23 +00001132
Misha Brukman76307852003-11-08 01:05:38 +00001133<div class="doc_text">
Chris Lattner74d3f822004-12-09 17:30:23 +00001134
Chris Lattner48b383b02003-11-25 01:02:51 +00001135<p>As mentioned <a href="#functionstructure">previously</a>, every
1136basic block in a program ends with a "Terminator" instruction, which
1137indicates which block should be executed after the current block is
1138finished. These terminator instructions typically yield a '<tt>void</tt>'
1139value: they produce control flow, not values (the one exception being
1140the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction).</p>
John Criswelldfe6a862004-12-10 15:51:16 +00001141<p>There are six different terminator instructions: the '<a
Chris Lattner48b383b02003-11-25 01:02:51 +00001142 href="#i_ret"><tt>ret</tt></a>' instruction, the '<a href="#i_br"><tt>br</tt></a>'
1143instruction, the '<a href="#i_switch"><tt>switch</tt></a>' instruction,
Chris Lattner08b7d5b2004-10-16 18:04:13 +00001144the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction, the '<a
1145 href="#i_unwind"><tt>unwind</tt></a>' instruction, and the '<a
1146 href="#i_unreachable"><tt>unreachable</tt></a>' instruction.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +00001147
Misha Brukman76307852003-11-08 01:05:38 +00001148</div>
Chris Lattner74d3f822004-12-09 17:30:23 +00001149
Chris Lattner2f7c9632001-06-06 20:29:01 +00001150<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +00001151<div class="doc_subsubsection"> <a name="i_ret">'<tt>ret</tt>'
1152Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +00001153<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +00001154<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001155<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 +00001156 ret void <i>; Return from void function</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001157</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001158<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001159<p>The '<tt>ret</tt>' instruction is used to return control flow (and a
John Criswell4a3327e2005-05-13 22:25:59 +00001160value) from a function back to the caller.</p>
John Criswell417228d2004-04-09 16:48:45 +00001161<p>There are two forms of the '<tt>ret</tt>' instruction: one that
Chris Lattner48b383b02003-11-25 01:02:51 +00001162returns a value and then causes control flow, and one that just causes
1163control flow to occur.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001164<h5>Arguments:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001165<p>The '<tt>ret</tt>' instruction may return any '<a
1166 href="#t_firstclass">first class</a>' type. Notice that a function is
1167not <a href="#wellformed">well formed</a> if there exists a '<tt>ret</tt>'
1168instruction inside of the function that returns a value that does not
1169match the return type of the function.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001170<h5>Semantics:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001171<p>When the '<tt>ret</tt>' instruction is executed, control flow
1172returns back to the calling function's context. If the caller is a "<a
John Criswell40db33f2004-06-25 15:16:57 +00001173 href="#i_call"><tt>call</tt></a>" instruction, execution continues at
Chris Lattner48b383b02003-11-25 01:02:51 +00001174the instruction after the call. If the caller was an "<a
1175 href="#i_invoke"><tt>invoke</tt></a>" instruction, execution continues
John Criswell02fdc6f2005-05-12 16:52:32 +00001176at the beginning of the "normal" destination block. If the instruction
Chris Lattner48b383b02003-11-25 01:02:51 +00001177returns a value, that value shall set the call or invoke instruction's
1178return value.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001179<h5>Example:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001180<pre> ret int 5 <i>; Return an integer value of 5</i>
Chris Lattner590645f2002-04-14 06:13:44 +00001181 ret void <i>; Return from a void function</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001182</pre>
Misha Brukman76307852003-11-08 01:05:38 +00001183</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001184<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +00001185<div class="doc_subsubsection"> <a name="i_br">'<tt>br</tt>' Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +00001186<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +00001187<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001188<pre> br bool &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 +00001189</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001190<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001191<p>The '<tt>br</tt>' instruction is used to cause control flow to
1192transfer to a different basic block in the current function. There are
1193two forms of this instruction, corresponding to a conditional branch
1194and an unconditional branch.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001195<h5>Arguments:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001196<p>The conditional branch form of the '<tt>br</tt>' instruction takes a
1197single '<tt>bool</tt>' value and two '<tt>label</tt>' values. The
1198unconditional form of the '<tt>br</tt>' instruction takes a single '<tt>label</tt>'
1199value as a target.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001200<h5>Semantics:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001201<p>Upon execution of a conditional '<tt>br</tt>' instruction, the '<tt>bool</tt>'
1202argument is evaluated. If the value is <tt>true</tt>, control flows
1203to the '<tt>iftrue</tt>' <tt>label</tt> argument. If "cond" is <tt>false</tt>,
1204control flows to the '<tt>iffalse</tt>' <tt>label</tt> argument.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001205<h5>Example:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001206<pre>Test:<br> %cond = <a href="#i_setcc">seteq</a> int %a, %b<br> br bool %cond, label %IfEqual, label %IfUnequal<br>IfEqual:<br> <a
1207 href="#i_ret">ret</a> int 1<br>IfUnequal:<br> <a href="#i_ret">ret</a> int 0<br></pre>
Misha Brukman76307852003-11-08 01:05:38 +00001208</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001209<!-- _______________________________________________________________________ -->
Chris Lattnercf96c6c2004-02-24 04:54:45 +00001210<div class="doc_subsubsection">
1211 <a name="i_switch">'<tt>switch</tt>' Instruction</a>
1212</div>
1213
Misha Brukman76307852003-11-08 01:05:38 +00001214<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +00001215<h5>Syntax:</h5>
Chris Lattnercf96c6c2004-02-24 04:54:45 +00001216
1217<pre>
1218 switch &lt;intty&gt; &lt;value&gt;, label &lt;defaultdest&gt; [ &lt;intty&gt; &lt;val&gt;, label &lt;dest&gt; ... ]
1219</pre>
1220
Chris Lattner2f7c9632001-06-06 20:29:01 +00001221<h5>Overview:</h5>
Chris Lattnercf96c6c2004-02-24 04:54:45 +00001222
1223<p>The '<tt>switch</tt>' instruction is used to transfer control flow to one of
1224several different places. It is a generalization of the '<tt>br</tt>'
Misha Brukman76307852003-11-08 01:05:38 +00001225instruction, allowing a branch to occur to one of many possible
1226destinations.</p>
Chris Lattnercf96c6c2004-02-24 04:54:45 +00001227
1228
Chris Lattner2f7c9632001-06-06 20:29:01 +00001229<h5>Arguments:</h5>
Chris Lattnercf96c6c2004-02-24 04:54:45 +00001230
1231<p>The '<tt>switch</tt>' instruction uses three parameters: an integer
1232comparison value '<tt>value</tt>', a default '<tt>label</tt>' destination, and
1233an array of pairs of comparison value constants and '<tt>label</tt>'s. The
1234table is not allowed to contain duplicate constant entries.</p>
1235
Chris Lattner2f7c9632001-06-06 20:29:01 +00001236<h5>Semantics:</h5>
Chris Lattnercf96c6c2004-02-24 04:54:45 +00001237
Chris Lattner48b383b02003-11-25 01:02:51 +00001238<p>The <tt>switch</tt> instruction specifies a table of values and
1239destinations. When the '<tt>switch</tt>' instruction is executed, this
John Criswellbcbb18c2004-06-25 16:05:06 +00001240table is searched for the given value. If the value is found, control flow is
1241transfered to the corresponding destination; otherwise, control flow is
1242transfered to the default destination.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001243
Chris Lattnercf96c6c2004-02-24 04:54:45 +00001244<h5>Implementation:</h5>
1245
1246<p>Depending on properties of the target machine and the particular
1247<tt>switch</tt> instruction, this instruction may be code generated in different
John Criswellbcbb18c2004-06-25 16:05:06 +00001248ways. For example, it could be generated as a series of chained conditional
1249branches or with a lookup table.</p>
Chris Lattnercf96c6c2004-02-24 04:54:45 +00001250
1251<h5>Example:</h5>
1252
1253<pre>
1254 <i>; Emulate a conditional br instruction</i>
1255 %Val = <a href="#i_cast">cast</a> bool %value to int
1256 switch int %Val, label %truedest [int 0, label %falsedest ]
1257
1258 <i>; Emulate an unconditional br instruction</i>
1259 switch uint 0, label %dest [ ]
1260
1261 <i>; Implement a jump table:</i>
1262 switch uint %val, label %otherwise [ uint 0, label %onzero
1263 uint 1, label %onone
1264 uint 2, label %ontwo ]
Chris Lattner2f7c9632001-06-06 20:29:01 +00001265</pre>
Misha Brukman76307852003-11-08 01:05:38 +00001266</div>
Chris Lattner0132aff2005-05-06 22:57:40 +00001267
Chris Lattner2f7c9632001-06-06 20:29:01 +00001268<!-- _______________________________________________________________________ -->
Chris Lattner0132aff2005-05-06 22:57:40 +00001269<div class="doc_subsubsection">
1270 <a name="i_invoke">'<tt>invoke</tt>' Instruction</a>
1271</div>
1272
Misha Brukman76307852003-11-08 01:05:38 +00001273<div class="doc_text">
Chris Lattner0132aff2005-05-06 22:57:40 +00001274
Chris Lattner2f7c9632001-06-06 20:29:01 +00001275<h5>Syntax:</h5>
Chris Lattner0132aff2005-05-06 22:57:40 +00001276
1277<pre>
1278 &lt;result&gt; = invoke [<a href="#callingconv">cconv</a>] &lt;ptr to function ty&gt; %&lt;function ptr val&gt;(&lt;function args&gt;)
1279 to label &lt;normal label&gt; except label &lt;exception label&gt;
1280</pre>
1281
Chris Lattnera8292f32002-05-06 22:08:29 +00001282<h5>Overview:</h5>
Chris Lattner0132aff2005-05-06 22:57:40 +00001283
1284<p>The '<tt>invoke</tt>' instruction causes control to transfer to a specified
1285function, with the possibility of control flow transfer to either the
John Criswell02fdc6f2005-05-12 16:52:32 +00001286'<tt>normal</tt>' label or the
1287'<tt>exception</tt>' label. If the callee function returns with the
Chris Lattner0132aff2005-05-06 22:57:40 +00001288"<tt><a href="#i_ret">ret</a></tt>" instruction, control flow will return to the
1289"normal" label. If the callee (or any indirect callees) returns with the "<a
John Criswell02fdc6f2005-05-12 16:52:32 +00001290href="#i_unwind"><tt>unwind</tt></a>" instruction, control is interrupted and
1291continued at the dynamically nearest "exception" label.</p>
Chris Lattner0132aff2005-05-06 22:57:40 +00001292
Chris Lattner2f7c9632001-06-06 20:29:01 +00001293<h5>Arguments:</h5>
Chris Lattner0132aff2005-05-06 22:57:40 +00001294
Misha Brukman76307852003-11-08 01:05:38 +00001295<p>This instruction requires several arguments:</p>
Chris Lattner0132aff2005-05-06 22:57:40 +00001296
Chris Lattner2f7c9632001-06-06 20:29:01 +00001297<ol>
Chris Lattner0132aff2005-05-06 22:57:40 +00001298 <li>
John Criswell4a3327e2005-05-13 22:25:59 +00001299 The optional "cconv" marker indicates which <a href="callingconv">calling
Chris Lattner0132aff2005-05-06 22:57:40 +00001300 convention</a> the call should use. If none is specified, the call defaults
1301 to using C calling conventions.
1302 </li>
1303 <li>'<tt>ptr to function ty</tt>': shall be the signature of the pointer to
1304 function value being invoked. In most cases, this is a direct function
1305 invocation, but indirect <tt>invoke</tt>s are just as possible, branching off
1306 an arbitrary pointer to function value.
1307 </li>
1308
1309 <li>'<tt>function ptr val</tt>': An LLVM value containing a pointer to a
1310 function to be invoked. </li>
1311
1312 <li>'<tt>function args</tt>': argument list whose types match the function
1313 signature argument types. If the function signature indicates the function
1314 accepts a variable number of arguments, the extra arguments can be
1315 specified. </li>
1316
1317 <li>'<tt>normal label</tt>': the label reached when the called function
1318 executes a '<tt><a href="#i_ret">ret</a></tt>' instruction. </li>
1319
1320 <li>'<tt>exception label</tt>': the label reached when a callee returns with
1321 the <a href="#i_unwind"><tt>unwind</tt></a> instruction. </li>
1322
Chris Lattner2f7c9632001-06-06 20:29:01 +00001323</ol>
Chris Lattner0132aff2005-05-06 22:57:40 +00001324
Chris Lattner2f7c9632001-06-06 20:29:01 +00001325<h5>Semantics:</h5>
Chris Lattner0132aff2005-05-06 22:57:40 +00001326
Misha Brukman76307852003-11-08 01:05:38 +00001327<p>This instruction is designed to operate as a standard '<tt><a
Chris Lattner0132aff2005-05-06 22:57:40 +00001328href="#i_call">call</a></tt>' instruction in most regards. The primary
1329difference is that it establishes an association with a label, which is used by
1330the runtime library to unwind the stack.</p>
1331
1332<p>This instruction is used in languages with destructors to ensure that proper
1333cleanup is performed in the case of either a <tt>longjmp</tt> or a thrown
1334exception. Additionally, this is important for implementation of
1335'<tt>catch</tt>' clauses in high-level languages that support them.</p>
1336
Chris Lattner2f7c9632001-06-06 20:29:01 +00001337<h5>Example:</h5>
Chris Lattner0132aff2005-05-06 22:57:40 +00001338<pre>
1339 %retval = invoke int %Test(int 15) to label %Continue
1340 except label %TestCleanup <i>; {int}:retval set</i>
1341 %retval = invoke <a href="#callingconv">coldcc</a> int %Test(int 15) to label %Continue
1342 except label %TestCleanup <i>; {int}:retval set</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001343</pre>
Misha Brukman76307852003-11-08 01:05:38 +00001344</div>
Chris Lattner08b7d5b2004-10-16 18:04:13 +00001345
1346
Chris Lattner5ed60612003-09-03 00:41:47 +00001347<!-- _______________________________________________________________________ -->
Chris Lattner08b7d5b2004-10-16 18:04:13 +00001348
Chris Lattner48b383b02003-11-25 01:02:51 +00001349<div class="doc_subsubsection"> <a name="i_unwind">'<tt>unwind</tt>'
1350Instruction</a> </div>
Chris Lattner08b7d5b2004-10-16 18:04:13 +00001351
Misha Brukman76307852003-11-08 01:05:38 +00001352<div class="doc_text">
Chris Lattner08b7d5b2004-10-16 18:04:13 +00001353
Chris Lattner5ed60612003-09-03 00:41:47 +00001354<h5>Syntax:</h5>
Chris Lattner08b7d5b2004-10-16 18:04:13 +00001355<pre>
1356 unwind
1357</pre>
1358
Chris Lattner5ed60612003-09-03 00:41:47 +00001359<h5>Overview:</h5>
Chris Lattner08b7d5b2004-10-16 18:04:13 +00001360
1361<p>The '<tt>unwind</tt>' instruction unwinds the stack, continuing control flow
1362at the first callee in the dynamic call stack which used an <a
1363href="#i_invoke"><tt>invoke</tt></a> instruction to perform the call. This is
1364primarily used to implement exception handling.</p>
1365
Chris Lattner5ed60612003-09-03 00:41:47 +00001366<h5>Semantics:</h5>
Chris Lattner08b7d5b2004-10-16 18:04:13 +00001367
1368<p>The '<tt>unwind</tt>' intrinsic causes execution of the current function to
1369immediately halt. The dynamic call stack is then searched for the first <a
1370href="#i_invoke"><tt>invoke</tt></a> instruction on the call stack. Once found,
1371execution continues at the "exceptional" destination block specified by the
1372<tt>invoke</tt> instruction. If there is no <tt>invoke</tt> instruction in the
1373dynamic call chain, undefined behavior results.</p>
Misha Brukman76307852003-11-08 01:05:38 +00001374</div>
Chris Lattner08b7d5b2004-10-16 18:04:13 +00001375
1376<!-- _______________________________________________________________________ -->
1377
1378<div class="doc_subsubsection"> <a name="i_unreachable">'<tt>unreachable</tt>'
1379Instruction</a> </div>
1380
1381<div class="doc_text">
1382
1383<h5>Syntax:</h5>
1384<pre>
1385 unreachable
1386</pre>
1387
1388<h5>Overview:</h5>
1389
1390<p>The '<tt>unreachable</tt>' instruction has no defined semantics. This
1391instruction is used to inform the optimizer that a particular portion of the
1392code is not reachable. This can be used to indicate that the code after a
1393no-return function cannot be reached, and other facts.</p>
1394
1395<h5>Semantics:</h5>
1396
1397<p>The '<tt>unreachable</tt>' instruction has no defined semantics.</p>
1398</div>
1399
1400
1401
Chris Lattner2f7c9632001-06-06 20:29:01 +00001402<!-- ======================================================================= -->
Chris Lattner48b383b02003-11-25 01:02:51 +00001403<div class="doc_subsection"> <a name="binaryops">Binary Operations</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +00001404<div class="doc_text">
Chris Lattner48b383b02003-11-25 01:02:51 +00001405<p>Binary operators are used to do most of the computation in a
1406program. They require two operands, execute an operation on them, and
John Criswelldfe6a862004-12-10 15:51:16 +00001407produce a single value. The operands might represent
Chris Lattnerc8cb6952004-08-12 19:12:28 +00001408multiple data, as is the case with the <a href="#t_packed">packed</a> data type.
1409The result value of a binary operator is not
Chris Lattner48b383b02003-11-25 01:02:51 +00001410necessarily the same type as its operands.</p>
Misha Brukman76307852003-11-08 01:05:38 +00001411<p>There are several different binary operators:</p>
Misha Brukman76307852003-11-08 01:05:38 +00001412</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001413<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +00001414<div class="doc_subsubsection"> <a name="i_add">'<tt>add</tt>'
1415Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +00001416<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +00001417<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001418<pre> &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 +00001419</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001420<h5>Overview:</h5>
Misha Brukman76307852003-11-08 01:05:38 +00001421<p>The '<tt>add</tt>' instruction returns the sum of its two operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001422<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +00001423<p>The two arguments to the '<tt>add</tt>' instruction must be either <a
Chris Lattnerc8cb6952004-08-12 19:12:28 +00001424 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a> values.
1425 This instruction can also take <a href="#t_packed">packed</a> versions of the values.
1426Both arguments must have identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001427<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +00001428<p>The value produced is the integer or floating point sum of the two
1429operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001430<h5>Example:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001431<pre> &lt;result&gt; = add int 4, %var <i>; yields {int}:result = 4 + %var</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001432</pre>
Misha Brukman76307852003-11-08 01:05:38 +00001433</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001434<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +00001435<div class="doc_subsubsection"> <a name="i_sub">'<tt>sub</tt>'
1436Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +00001437<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +00001438<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001439<pre> &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 +00001440</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001441<h5>Overview:</h5>
Misha Brukman76307852003-11-08 01:05:38 +00001442<p>The '<tt>sub</tt>' instruction returns the difference of its two
1443operands.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +00001444<p>Note that the '<tt>sub</tt>' instruction is used to represent the '<tt>neg</tt>'
1445instruction present in most other intermediate representations.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001446<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +00001447<p>The two arguments to the '<tt>sub</tt>' instruction must be either <a
Chris Lattner48b383b02003-11-25 01:02:51 +00001448 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
Chris Lattnerc8cb6952004-08-12 19:12:28 +00001449values.
1450This instruction can also take <a href="#t_packed">packed</a> versions of the values.
1451Both arguments must have identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001452<h5>Semantics:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001453<p>The value produced is the integer or floating point difference of
1454the two operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001455<h5>Example:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001456<pre> &lt;result&gt; = sub int 4, %var <i>; yields {int}:result = 4 - %var</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001457 &lt;result&gt; = sub int 0, %val <i>; yields {int}:result = -%var</i>
1458</pre>
Misha Brukman76307852003-11-08 01:05:38 +00001459</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001460<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +00001461<div class="doc_subsubsection"> <a name="i_mul">'<tt>mul</tt>'
1462Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +00001463<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +00001464<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001465<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 +00001466</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001467<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001468<p>The '<tt>mul</tt>' instruction returns the product of its two
1469operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001470<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +00001471<p>The two arguments to the '<tt>mul</tt>' instruction must be either <a
Chris Lattner48b383b02003-11-25 01:02:51 +00001472 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
Chris Lattnerc8cb6952004-08-12 19:12:28 +00001473values.
1474This instruction can also take <a href="#t_packed">packed</a> versions of the values.
1475Both arguments must have identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001476<h5>Semantics:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001477<p>The value produced is the integer or floating point product of the
Misha Brukman76307852003-11-08 01:05:38 +00001478two operands.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +00001479<p>There is no signed vs unsigned multiplication. The appropriate
1480action is taken based on the type of the operand.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001481<h5>Example:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001482<pre> &lt;result&gt; = mul int 4, %var <i>; yields {int}:result = 4 * %var</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001483</pre>
Misha Brukman76307852003-11-08 01:05:38 +00001484</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001485<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +00001486<div class="doc_subsubsection"> <a name="i_div">'<tt>div</tt>'
1487Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +00001488<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +00001489<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001490<pre> &lt;result&gt; = div &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
1491</pre>
1492<h5>Overview:</h5>
1493<p>The '<tt>div</tt>' instruction returns the quotient of its two
1494operands.</p>
1495<h5>Arguments:</h5>
1496<p>The two arguments to the '<tt>div</tt>' instruction must be either <a
1497 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
Chris Lattnerc8cb6952004-08-12 19:12:28 +00001498values.
1499This instruction can also take <a href="#t_packed">packed</a> versions of the values.
1500Both arguments must have identical types.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +00001501<h5>Semantics:</h5>
1502<p>The value produced is the integer or floating point quotient of the
1503two operands.</p>
1504<h5>Example:</h5>
1505<pre> &lt;result&gt; = div int 4, %var <i>; yields {int}:result = 4 / %var</i>
1506</pre>
1507</div>
1508<!-- _______________________________________________________________________ -->
1509<div class="doc_subsubsection"> <a name="i_rem">'<tt>rem</tt>'
1510Instruction</a> </div>
1511<div class="doc_text">
1512<h5>Syntax:</h5>
1513<pre> &lt;result&gt; = rem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
1514</pre>
1515<h5>Overview:</h5>
1516<p>The '<tt>rem</tt>' instruction returns the remainder from the
1517division of its two operands.</p>
1518<h5>Arguments:</h5>
1519<p>The two arguments to the '<tt>rem</tt>' instruction must be either <a
1520 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
Chris Lattnerc8cb6952004-08-12 19:12:28 +00001521values.
1522This instruction can also take <a href="#t_packed">packed</a> versions of the values.
1523Both arguments must have identical types.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +00001524<h5>Semantics:</h5>
1525<p>This returns the <i>remainder</i> of a division (where the result
1526has the same sign as the divisor), not the <i>modulus</i> (where the
1527result has the same sign as the dividend) of a value. For more
John Criswell4c0cf7f2005-10-24 16:17:18 +00001528information about the difference, see <a
Chris Lattner48b383b02003-11-25 01:02:51 +00001529 href="http://mathforum.org/dr.math/problems/anne.4.28.99.html">The
1530Math Forum</a>.</p>
1531<h5>Example:</h5>
1532<pre> &lt;result&gt; = rem int 4, %var <i>; yields {int}:result = 4 % %var</i>
1533</pre>
1534</div>
1535<!-- _______________________________________________________________________ -->
1536<div class="doc_subsubsection"> <a name="i_setcc">'<tt>set<i>cc</i></tt>'
1537Instructions</a> </div>
1538<div class="doc_text">
1539<h5>Syntax:</h5>
1540<pre> &lt;result&gt; = seteq &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {bool}:result</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001541 &lt;result&gt; = setne &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {bool}:result</i>
1542 &lt;result&gt; = setlt &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {bool}:result</i>
1543 &lt;result&gt; = setgt &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {bool}:result</i>
1544 &lt;result&gt; = setle &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {bool}:result</i>
1545 &lt;result&gt; = setge &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {bool}:result</i>
1546</pre>
Chris Lattner48b383b02003-11-25 01:02:51 +00001547<h5>Overview:</h5>
1548<p>The '<tt>set<i>cc</i></tt>' family of instructions returns a boolean
1549value based on a comparison of their two operands.</p>
1550<h5>Arguments:</h5>
1551<p>The two arguments to the '<tt>set<i>cc</i></tt>' instructions must
1552be of <a href="#t_firstclass">first class</a> type (it is not possible
1553to compare '<tt>label</tt>'s, '<tt>array</tt>'s, '<tt>structure</tt>'
1554or '<tt>void</tt>' values, etc...). Both arguments must have identical
1555types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001556<h5>Semantics:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001557<p>The '<tt>seteq</tt>' instruction yields a <tt>true</tt> '<tt>bool</tt>'
1558value if both operands are equal.<br>
1559The '<tt>setne</tt>' instruction yields a <tt>true</tt> '<tt>bool</tt>'
1560value if both operands are unequal.<br>
1561The '<tt>setlt</tt>' instruction yields a <tt>true</tt> '<tt>bool</tt>'
1562value if the first operand is less than the second operand.<br>
1563The '<tt>setgt</tt>' instruction yields a <tt>true</tt> '<tt>bool</tt>'
1564value if the first operand is greater than the second operand.<br>
1565The '<tt>setle</tt>' instruction yields a <tt>true</tt> '<tt>bool</tt>'
1566value if the first operand is less than or equal to the second operand.<br>
1567The '<tt>setge</tt>' instruction yields a <tt>true</tt> '<tt>bool</tt>'
1568value if the first operand is greater than or equal to the second
1569operand.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001570<h5>Example:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001571<pre> &lt;result&gt; = seteq int 4, 5 <i>; yields {bool}:result = false</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001572 &lt;result&gt; = setne float 4, 5 <i>; yields {bool}:result = true</i>
1573 &lt;result&gt; = setlt uint 4, 5 <i>; yields {bool}:result = true</i>
1574 &lt;result&gt; = setgt sbyte 4, 5 <i>; yields {bool}:result = false</i>
1575 &lt;result&gt; = setle sbyte 4, 5 <i>; yields {bool}:result = true</i>
1576 &lt;result&gt; = setge sbyte 4, 5 <i>; yields {bool}:result = false</i>
1577</pre>
Misha Brukman76307852003-11-08 01:05:38 +00001578</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001579<!-- ======================================================================= -->
Chris Lattner48b383b02003-11-25 01:02:51 +00001580<div class="doc_subsection"> <a name="bitwiseops">Bitwise Binary
1581Operations</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +00001582<div class="doc_text">
Chris Lattner48b383b02003-11-25 01:02:51 +00001583<p>Bitwise binary operators are used to do various forms of
1584bit-twiddling in a program. They are generally very efficient
John Criswelldfe6a862004-12-10 15:51:16 +00001585instructions and can commonly be strength reduced from other
Chris Lattner48b383b02003-11-25 01:02:51 +00001586instructions. They require two operands, execute an operation on them,
1587and produce a single value. The resulting value of the bitwise binary
1588operators is always the same type as its first operand.</p>
Misha Brukman76307852003-11-08 01:05:38 +00001589</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001590<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +00001591<div class="doc_subsubsection"> <a name="i_and">'<tt>and</tt>'
1592Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +00001593<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +00001594<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001595<pre> &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 +00001596</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001597<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001598<p>The '<tt>and</tt>' instruction returns the bitwise logical and of
1599its two operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001600<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +00001601<p>The two arguments to the '<tt>and</tt>' instruction must be <a
Chris Lattner48b383b02003-11-25 01:02:51 +00001602 href="#t_integral">integral</a> values. Both arguments must have
1603identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001604<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +00001605<p>The truth table used for the '<tt>and</tt>' instruction is:</p>
Chris Lattner48b383b02003-11-25 01:02:51 +00001606<p> </p>
Misha Brukmanc501f552004-03-01 17:47:27 +00001607<div style="align: center">
Misha Brukman76307852003-11-08 01:05:38 +00001608<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner48b383b02003-11-25 01:02:51 +00001609 <tbody>
1610 <tr>
1611 <td>In0</td>
1612 <td>In1</td>
1613 <td>Out</td>
1614 </tr>
1615 <tr>
1616 <td>0</td>
1617 <td>0</td>
1618 <td>0</td>
1619 </tr>
1620 <tr>
1621 <td>0</td>
1622 <td>1</td>
1623 <td>0</td>
1624 </tr>
1625 <tr>
1626 <td>1</td>
1627 <td>0</td>
1628 <td>0</td>
1629 </tr>
1630 <tr>
1631 <td>1</td>
1632 <td>1</td>
1633 <td>1</td>
1634 </tr>
1635 </tbody>
1636</table>
Misha Brukmanc501f552004-03-01 17:47:27 +00001637</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001638<h5>Example:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001639<pre> &lt;result&gt; = and int 4, %var <i>; yields {int}:result = 4 &amp; %var</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001640 &lt;result&gt; = and int 15, 40 <i>; yields {int}:result = 8</i>
1641 &lt;result&gt; = and int 4, 8 <i>; yields {int}:result = 0</i>
1642</pre>
Misha Brukman76307852003-11-08 01:05:38 +00001643</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001644<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +00001645<div class="doc_subsubsection"> <a name="i_or">'<tt>or</tt>' Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +00001646<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +00001647<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001648<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 +00001649</pre>
Chris Lattner48b383b02003-11-25 01:02:51 +00001650<h5>Overview:</h5>
1651<p>The '<tt>or</tt>' instruction returns the bitwise logical inclusive
1652or of its two operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001653<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +00001654<p>The two arguments to the '<tt>or</tt>' instruction must be <a
Chris Lattner48b383b02003-11-25 01:02:51 +00001655 href="#t_integral">integral</a> values. Both arguments must have
1656identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001657<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +00001658<p>The truth table used for the '<tt>or</tt>' instruction is:</p>
Chris Lattner48b383b02003-11-25 01:02:51 +00001659<p> </p>
Misha Brukmanc501f552004-03-01 17:47:27 +00001660<div style="align: center">
Chris Lattner48b383b02003-11-25 01:02:51 +00001661<table border="1" cellspacing="0" cellpadding="4">
1662 <tbody>
1663 <tr>
1664 <td>In0</td>
1665 <td>In1</td>
1666 <td>Out</td>
1667 </tr>
1668 <tr>
1669 <td>0</td>
1670 <td>0</td>
1671 <td>0</td>
1672 </tr>
1673 <tr>
1674 <td>0</td>
1675 <td>1</td>
1676 <td>1</td>
1677 </tr>
1678 <tr>
1679 <td>1</td>
1680 <td>0</td>
1681 <td>1</td>
1682 </tr>
1683 <tr>
1684 <td>1</td>
1685 <td>1</td>
1686 <td>1</td>
1687 </tr>
1688 </tbody>
1689</table>
Misha Brukmanc501f552004-03-01 17:47:27 +00001690</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001691<h5>Example:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001692<pre> &lt;result&gt; = or int 4, %var <i>; yields {int}:result = 4 | %var</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001693 &lt;result&gt; = or int 15, 40 <i>; yields {int}:result = 47</i>
1694 &lt;result&gt; = or int 4, 8 <i>; yields {int}:result = 12</i>
1695</pre>
Misha Brukman76307852003-11-08 01:05:38 +00001696</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001697<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +00001698<div class="doc_subsubsection"> <a name="i_xor">'<tt>xor</tt>'
1699Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +00001700<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +00001701<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001702<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 +00001703</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001704<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001705<p>The '<tt>xor</tt>' instruction returns the bitwise logical exclusive
1706or of its two operands. The <tt>xor</tt> is used to implement the
1707"one's complement" operation, which is the "~" operator in C.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001708<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +00001709<p>The two arguments to the '<tt>xor</tt>' instruction must be <a
Chris Lattner48b383b02003-11-25 01:02:51 +00001710 href="#t_integral">integral</a> values. Both arguments must have
1711identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001712<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +00001713<p>The truth table used for the '<tt>xor</tt>' instruction is:</p>
Chris Lattner48b383b02003-11-25 01:02:51 +00001714<p> </p>
Misha Brukmanc501f552004-03-01 17:47:27 +00001715<div style="align: center">
Chris Lattner48b383b02003-11-25 01:02:51 +00001716<table border="1" cellspacing="0" cellpadding="4">
1717 <tbody>
1718 <tr>
1719 <td>In0</td>
1720 <td>In1</td>
1721 <td>Out</td>
1722 </tr>
1723 <tr>
1724 <td>0</td>
1725 <td>0</td>
1726 <td>0</td>
1727 </tr>
1728 <tr>
1729 <td>0</td>
1730 <td>1</td>
1731 <td>1</td>
1732 </tr>
1733 <tr>
1734 <td>1</td>
1735 <td>0</td>
1736 <td>1</td>
1737 </tr>
1738 <tr>
1739 <td>1</td>
1740 <td>1</td>
1741 <td>0</td>
1742 </tr>
1743 </tbody>
1744</table>
Misha Brukmanc501f552004-03-01 17:47:27 +00001745</div>
Chris Lattner48b383b02003-11-25 01:02:51 +00001746<p> </p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001747<h5>Example:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001748<pre> &lt;result&gt; = xor int 4, %var <i>; yields {int}:result = 4 ^ %var</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001749 &lt;result&gt; = xor int 15, 40 <i>; yields {int}:result = 39</i>
1750 &lt;result&gt; = xor int 4, 8 <i>; yields {int}:result = 12</i>
Chris Lattner5ed60612003-09-03 00:41:47 +00001751 &lt;result&gt; = xor int %V, -1 <i>; yields {int}:result = ~%V</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001752</pre>
Misha Brukman76307852003-11-08 01:05:38 +00001753</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001754<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +00001755<div class="doc_subsubsection"> <a name="i_shl">'<tt>shl</tt>'
1756Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +00001757<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +00001758<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001759<pre> &lt;result&gt; = shl &lt;ty&gt; &lt;var1&gt;, ubyte &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001760</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001761<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001762<p>The '<tt>shl</tt>' instruction returns the first operand shifted to
1763the left a specified number of bits.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001764<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +00001765<p>The first argument to the '<tt>shl</tt>' instruction must be an <a
Chris Lattner48b383b02003-11-25 01:02:51 +00001766 href="#t_integer">integer</a> type. The second argument must be an '<tt>ubyte</tt>'
1767type.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001768<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +00001769<p>The value produced is <tt>var1</tt> * 2<sup><tt>var2</tt></sup>.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001770<h5>Example:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001771<pre> &lt;result&gt; = shl int 4, ubyte %var <i>; yields {int}:result = 4 &lt;&lt; %var</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001772 &lt;result&gt; = shl int 4, ubyte 2 <i>; yields {int}:result = 16</i>
1773 &lt;result&gt; = shl int 1, ubyte 10 <i>; yields {int}:result = 1024</i>
1774</pre>
Misha Brukman76307852003-11-08 01:05:38 +00001775</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001776<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +00001777<div class="doc_subsubsection"> <a name="i_shr">'<tt>shr</tt>'
1778Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +00001779<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +00001780<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001781<pre> &lt;result&gt; = shr &lt;ty&gt; &lt;var1&gt;, ubyte &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001782</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001783<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001784<p>The '<tt>shr</tt>' instruction returns the first operand shifted to
1785the right a specified number of bits.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001786<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +00001787<p>The first argument to the '<tt>shr</tt>' instruction must be an <a
Chris Lattner48b383b02003-11-25 01:02:51 +00001788 href="#t_integer">integer</a> type. The second argument must be an '<tt>ubyte</tt>'
1789type.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001790<h5>Semantics:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001791<p>If the first argument is a <a href="#t_signed">signed</a> type, the
1792most significant bit is duplicated in the newly free'd bit positions.
1793If the first argument is unsigned, zero bits shall fill the empty
1794positions.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001795<h5>Example:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001796<pre> &lt;result&gt; = shr int 4, ubyte %var <i>; yields {int}:result = 4 &gt;&gt; %var</i>
Chris Lattner33426d92003-06-18 21:30:51 +00001797 &lt;result&gt; = shr uint 4, ubyte 1 <i>; yields {uint}:result = 2</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001798 &lt;result&gt; = shr int 4, ubyte 2 <i>; yields {int}:result = 1</i>
Chris Lattner33426d92003-06-18 21:30:51 +00001799 &lt;result&gt; = shr sbyte 4, ubyte 3 <i>; yields {sbyte}:result = 0</i>
1800 &lt;result&gt; = shr sbyte -2, ubyte 1 <i>; yields {sbyte}:result = -1</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001801</pre>
Misha Brukman76307852003-11-08 01:05:38 +00001802</div>
Chris Lattner54611b42005-11-06 08:02:57 +00001803
Chris Lattner2f7c9632001-06-06 20:29:01 +00001804<!-- ======================================================================= -->
Chris Lattner54611b42005-11-06 08:02:57 +00001805<div class="doc_subsection">
1806 <a name="memoryops">Memory Access Operations</a>
1807</div>
1808
Misha Brukman76307852003-11-08 01:05:38 +00001809<div class="doc_text">
Chris Lattner54611b42005-11-06 08:02:57 +00001810
Chris Lattner48b383b02003-11-25 01:02:51 +00001811<p>A key design point of an SSA-based representation is how it
1812represents memory. In LLVM, no memory locations are in SSA form, which
1813makes things very simple. This section describes how to read, write,
John Criswelldfe6a862004-12-10 15:51:16 +00001814allocate, and free memory in LLVM.</p>
Chris Lattner54611b42005-11-06 08:02:57 +00001815
Misha Brukman76307852003-11-08 01:05:38 +00001816</div>
Chris Lattner54611b42005-11-06 08:02:57 +00001817
Chris Lattner2f7c9632001-06-06 20:29:01 +00001818<!-- _______________________________________________________________________ -->
Chris Lattner54611b42005-11-06 08:02:57 +00001819<div class="doc_subsubsection">
1820 <a name="i_malloc">'<tt>malloc</tt>' Instruction</a>
1821</div>
1822
Misha Brukman76307852003-11-08 01:05:38 +00001823<div class="doc_text">
Chris Lattner54611b42005-11-06 08:02:57 +00001824
Chris Lattner2f7c9632001-06-06 20:29:01 +00001825<h5>Syntax:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +00001826
1827<pre>
1828 &lt;result&gt; = malloc &lt;type&gt;[, uint &lt;NumElements&gt;][, align &lt;alignment&gt;] <i>; yields {type*}:result</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001829</pre>
Chris Lattner54611b42005-11-06 08:02:57 +00001830
Chris Lattner2f7c9632001-06-06 20:29:01 +00001831<h5>Overview:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +00001832
Chris Lattner48b383b02003-11-25 01:02:51 +00001833<p>The '<tt>malloc</tt>' instruction allocates memory from the system
1834heap and returns a pointer to it.</p>
Chris Lattner54611b42005-11-06 08:02:57 +00001835
Chris Lattner2f7c9632001-06-06 20:29:01 +00001836<h5>Arguments:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +00001837
1838<p>The '<tt>malloc</tt>' instruction allocates
1839<tt>sizeof(&lt;type&gt;)*NumElements</tt>
John Criswella92e5862004-02-24 16:13:56 +00001840bytes of memory from the operating system and returns a pointer of the
Chris Lattner54611b42005-11-06 08:02:57 +00001841appropriate type to the program. If "NumElements" is specified, it is the
1842number of elements allocated. If an alignment is specified, the value result
1843of the allocation is guaranteed to be aligned to at least that boundary. If
1844not specified, or if zero, the target can choose to align the allocation on any
1845convenient boundary.</p>
1846
Misha Brukman76307852003-11-08 01:05:38 +00001847<p>'<tt>type</tt>' must be a sized type.</p>
Chris Lattner54611b42005-11-06 08:02:57 +00001848
Chris Lattner2f7c9632001-06-06 20:29:01 +00001849<h5>Semantics:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +00001850
Chris Lattner48b383b02003-11-25 01:02:51 +00001851<p>Memory is allocated using the system "<tt>malloc</tt>" function, and
1852a pointer is returned.</p>
Misha Brukman76307852003-11-08 01:05:38 +00001853
Chris Lattner54611b42005-11-06 08:02:57 +00001854<h5>Example:</h5>
1855
1856<pre>
1857 %array = malloc [4 x ubyte ] <i>; yields {[%4 x ubyte]*}:array</i>
1858
1859 %size = <a href="#i_add">add</a> uint 2, 2 <i>; yields {uint}:size = uint 4</i>
Chris Lattner590645f2002-04-14 06:13:44 +00001860 %array1 = malloc ubyte, uint 4 <i>; yields {ubyte*}:array1</i>
1861 %array2 = malloc [12 x ubyte], uint %size <i>; yields {[12 x ubyte]*}:array2</i>
Chris Lattner54611b42005-11-06 08:02:57 +00001862 %array3 = malloc int, uint 4, align 1024 <i>; yields {int*}:array3</i>
1863 %array4 = malloc int, align 1024 <i>; yields {int*}:array4</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001864</pre>
Misha Brukman76307852003-11-08 01:05:38 +00001865</div>
Chris Lattner54611b42005-11-06 08:02:57 +00001866
Chris Lattner2f7c9632001-06-06 20:29:01 +00001867<!-- _______________________________________________________________________ -->
Chris Lattner54611b42005-11-06 08:02:57 +00001868<div class="doc_subsubsection">
1869 <a name="i_free">'<tt>free</tt>' Instruction</a>
1870</div>
1871
Misha Brukman76307852003-11-08 01:05:38 +00001872<div class="doc_text">
Chris Lattner54611b42005-11-06 08:02:57 +00001873
Chris Lattner2f7c9632001-06-06 20:29:01 +00001874<h5>Syntax:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +00001875
1876<pre>
1877 free &lt;type&gt; &lt;value&gt; <i>; yields {void}</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001878</pre>
Chris Lattner54611b42005-11-06 08:02:57 +00001879
Chris Lattner2f7c9632001-06-06 20:29:01 +00001880<h5>Overview:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +00001881
Chris Lattner48b383b02003-11-25 01:02:51 +00001882<p>The '<tt>free</tt>' instruction returns memory back to the unused
John Criswell4a3327e2005-05-13 22:25:59 +00001883memory heap to be reallocated in the future.</p>
Chris Lattner54611b42005-11-06 08:02:57 +00001884
Chris Lattner2f7c9632001-06-06 20:29:01 +00001885<h5>Arguments:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +00001886
Chris Lattner48b383b02003-11-25 01:02:51 +00001887<p>'<tt>value</tt>' shall be a pointer value that points to a value
1888that was allocated with the '<tt><a href="#i_malloc">malloc</a></tt>'
1889instruction.</p>
Chris Lattner54611b42005-11-06 08:02:57 +00001890
Chris Lattner2f7c9632001-06-06 20:29:01 +00001891<h5>Semantics:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +00001892
John Criswelldfe6a862004-12-10 15:51:16 +00001893<p>Access to the memory pointed to by the pointer is no longer defined
Chris Lattner48b383b02003-11-25 01:02:51 +00001894after this instruction executes.</p>
Chris Lattner54611b42005-11-06 08:02:57 +00001895
Chris Lattner2f7c9632001-06-06 20:29:01 +00001896<h5>Example:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +00001897
1898<pre>
1899 %array = <a href="#i_malloc">malloc</a> [4 x ubyte] <i>; yields {[4 x ubyte]*}:array</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001900 free [4 x ubyte]* %array
1901</pre>
Misha Brukman76307852003-11-08 01:05:38 +00001902</div>
Chris Lattner54611b42005-11-06 08:02:57 +00001903
Chris Lattner2f7c9632001-06-06 20:29:01 +00001904<!-- _______________________________________________________________________ -->
Chris Lattner54611b42005-11-06 08:02:57 +00001905<div class="doc_subsubsection">
1906 <a name="i_alloca">'<tt>alloca</tt>' Instruction</a>
1907</div>
1908
Misha Brukman76307852003-11-08 01:05:38 +00001909<div class="doc_text">
Chris Lattner54611b42005-11-06 08:02:57 +00001910
Chris Lattner2f7c9632001-06-06 20:29:01 +00001911<h5>Syntax:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +00001912
1913<pre>
1914 &lt;result&gt; = alloca &lt;type&gt;[, uint &lt;NumElements&gt;][, align &lt;alignment&gt;] <i>; yields {type*}:result</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001915</pre>
Chris Lattner54611b42005-11-06 08:02:57 +00001916
Chris Lattner2f7c9632001-06-06 20:29:01 +00001917<h5>Overview:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +00001918
Chris Lattner48b383b02003-11-25 01:02:51 +00001919<p>The '<tt>alloca</tt>' instruction allocates memory on the current
1920stack frame of the procedure that is live until the current function
1921returns to its caller.</p>
Chris Lattner54611b42005-11-06 08:02:57 +00001922
Chris Lattner2f7c9632001-06-06 20:29:01 +00001923<h5>Arguments:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +00001924
John Criswelldfe6a862004-12-10 15:51:16 +00001925<p>The '<tt>alloca</tt>' instruction allocates <tt>sizeof(&lt;type&gt;)*NumElements</tt>
Chris Lattner48b383b02003-11-25 01:02:51 +00001926bytes of memory on the runtime stack, returning a pointer of the
Chris Lattner54611b42005-11-06 08:02:57 +00001927appropriate type to the program. If "NumElements" is specified, it is the
1928number of elements allocated. If an alignment is specified, the value result
1929of the allocation is guaranteed to be aligned to at least that boundary. If
1930not specified, or if zero, the target can choose to align the allocation on any
1931convenient boundary.</p>
1932
Misha Brukman76307852003-11-08 01:05:38 +00001933<p>'<tt>type</tt>' may be any sized type.</p>
Chris Lattner54611b42005-11-06 08:02:57 +00001934
Chris Lattner2f7c9632001-06-06 20:29:01 +00001935<h5>Semantics:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +00001936
John Criswell4a3327e2005-05-13 22:25:59 +00001937<p>Memory is allocated; a pointer is returned. '<tt>alloca</tt>'d
Chris Lattner48b383b02003-11-25 01:02:51 +00001938memory is automatically released when the function returns. The '<tt>alloca</tt>'
1939instruction is commonly used to represent automatic variables that must
1940have an address available. When the function returns (either with the <tt><a
John Criswellc932bef2005-05-12 16:55:34 +00001941 href="#i_ret">ret</a></tt> or <tt><a href="#i_unwind">unwind</a></tt>
Misha Brukman76307852003-11-08 01:05:38 +00001942instructions), the memory is reclaimed.</p>
Chris Lattner54611b42005-11-06 08:02:57 +00001943
Chris Lattner2f7c9632001-06-06 20:29:01 +00001944<h5>Example:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +00001945
1946<pre>
1947 %ptr = alloca int <i>; yields {int*}:ptr</i>
Chris Lattner590645f2002-04-14 06:13:44 +00001948 %ptr = alloca int, uint 4 <i>; yields {int*}:ptr</i>
Chris Lattner54611b42005-11-06 08:02:57 +00001949 %ptr = alloca int, uint 4, align 1024 <i>; yields {int*}:ptr</i>
1950 %ptr = alloca int, align 1024 <i>; yields {int*}:ptr</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +00001951</pre>
Misha Brukman76307852003-11-08 01:05:38 +00001952</div>
Chris Lattner54611b42005-11-06 08:02:57 +00001953
Chris Lattner2f7c9632001-06-06 20:29:01 +00001954<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +00001955<div class="doc_subsubsection"> <a name="i_load">'<tt>load</tt>'
1956Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +00001957<div class="doc_text">
Chris Lattner095735d2002-05-06 03:03:22 +00001958<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001959<pre> &lt;result&gt; = load &lt;ty&gt;* &lt;pointer&gt;<br> &lt;result&gt; = volatile load &lt;ty&gt;* &lt;pointer&gt;<br></pre>
Chris Lattner095735d2002-05-06 03:03:22 +00001960<h5>Overview:</h5>
Misha Brukman76307852003-11-08 01:05:38 +00001961<p>The '<tt>load</tt>' instruction is used to read from memory.</p>
Chris Lattner095735d2002-05-06 03:03:22 +00001962<h5>Arguments:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001963<p>The argument to the '<tt>load</tt>' instruction specifies the memory
John Criswell4c0cf7f2005-10-24 16:17:18 +00001964address from which to load. The pointer must point to a <a
Chris Lattner10ee9652004-06-03 22:57:15 +00001965 href="#t_firstclass">first class</a> type. If the <tt>load</tt> is
John Criswell4c0cf7f2005-10-24 16:17:18 +00001966marked as <tt>volatile</tt>, then the optimizer is not allowed to modify
Chris Lattner48b383b02003-11-25 01:02:51 +00001967the number or order of execution of this <tt>load</tt> with other
1968volatile <tt>load</tt> and <tt><a href="#i_store">store</a></tt>
1969instructions. </p>
Chris Lattner095735d2002-05-06 03:03:22 +00001970<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +00001971<p>The location of memory pointed to is loaded.</p>
Chris Lattner095735d2002-05-06 03:03:22 +00001972<h5>Examples:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001973<pre> %ptr = <a href="#i_alloca">alloca</a> int <i>; yields {int*}:ptr</i>
1974 <a
1975 href="#i_store">store</a> int 3, int* %ptr <i>; yields {void}</i>
Chris Lattner095735d2002-05-06 03:03:22 +00001976 %val = load int* %ptr <i>; yields {int}:val = int 3</i>
1977</pre>
Misha Brukman76307852003-11-08 01:05:38 +00001978</div>
Chris Lattner095735d2002-05-06 03:03:22 +00001979<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +00001980<div class="doc_subsubsection"> <a name="i_store">'<tt>store</tt>'
1981Instruction</a> </div>
Chris Lattner095735d2002-05-06 03:03:22 +00001982<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001983<pre> store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt; <i>; yields {void}</i>
Chris Lattner12d456c2003-09-08 18:27:49 +00001984 volatile store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt; <i>; yields {void}</i>
Chris Lattner095735d2002-05-06 03:03:22 +00001985</pre>
Chris Lattner095735d2002-05-06 03:03:22 +00001986<h5>Overview:</h5>
Misha Brukman76307852003-11-08 01:05:38 +00001987<p>The '<tt>store</tt>' instruction is used to write to memory.</p>
Chris Lattner095735d2002-05-06 03:03:22 +00001988<h5>Arguments:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00001989<p>There are two arguments to the '<tt>store</tt>' instruction: a value
John Criswell4c0cf7f2005-10-24 16:17:18 +00001990to store and an address in which to store it. The type of the '<tt>&lt;pointer&gt;</tt>'
Chris Lattner48b383b02003-11-25 01:02:51 +00001991operand must be a pointer to the type of the '<tt>&lt;value&gt;</tt>'
John Criswell4a3327e2005-05-13 22:25:59 +00001992operand. If the <tt>store</tt> is marked as <tt>volatile</tt>, then the
Chris Lattner48b383b02003-11-25 01:02:51 +00001993optimizer is not allowed to modify the number or order of execution of
1994this <tt>store</tt> with other volatile <tt>load</tt> and <tt><a
1995 href="#i_store">store</a></tt> instructions.</p>
1996<h5>Semantics:</h5>
1997<p>The contents of memory are updated to contain '<tt>&lt;value&gt;</tt>'
1998at the location specified by the '<tt>&lt;pointer&gt;</tt>' operand.</p>
Chris Lattner095735d2002-05-06 03:03:22 +00001999<h5>Example:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00002000<pre> %ptr = <a href="#i_alloca">alloca</a> int <i>; yields {int*}:ptr</i>
2001 <a
2002 href="#i_store">store</a> int 3, int* %ptr <i>; yields {void}</i>
Chris Lattner095735d2002-05-06 03:03:22 +00002003 %val = load int* %ptr <i>; yields {int}:val = int 3</i>
2004</pre>
Chris Lattner095735d2002-05-06 03:03:22 +00002005<!-- _______________________________________________________________________ -->
Chris Lattner33fd7022004-04-05 01:30:49 +00002006<div class="doc_subsubsection">
2007 <a name="i_getelementptr">'<tt>getelementptr</tt>' Instruction</a>
2008</div>
2009
Misha Brukman76307852003-11-08 01:05:38 +00002010<div class="doc_text">
Chris Lattner590645f2002-04-14 06:13:44 +00002011<h5>Syntax:</h5>
Chris Lattner33fd7022004-04-05 01:30:49 +00002012<pre>
2013 &lt;result&gt; = getelementptr &lt;ty&gt;* &lt;ptrval&gt;{, &lt;ty&gt; &lt;idx&gt;}*
2014</pre>
2015
Chris Lattner590645f2002-04-14 06:13:44 +00002016<h5>Overview:</h5>
Chris Lattner33fd7022004-04-05 01:30:49 +00002017
2018<p>
2019The '<tt>getelementptr</tt>' instruction is used to get the address of a
2020subelement of an aggregate data structure.</p>
2021
Chris Lattner590645f2002-04-14 06:13:44 +00002022<h5>Arguments:</h5>
Chris Lattner33fd7022004-04-05 01:30:49 +00002023
2024<p>This instruction takes a list of integer constants that indicate what
2025elements of the aggregate object to index to. The actual types of the arguments
2026provided depend on the type of the first pointer argument. The
2027'<tt>getelementptr</tt>' instruction is used to index down through the type
John Criswell88190562005-05-16 16:17:45 +00002028levels of a structure or to a specific index in an array. When indexing into a
2029structure, only <tt>uint</tt>
John Criswell4a3327e2005-05-13 22:25:59 +00002030integer constants are allowed. When indexing into an array or pointer,
Chris Lattner33fd7022004-04-05 01:30:49 +00002031<tt>int</tt> and <tt>long</tt> indexes are allowed of any sign.</p>
2032
Chris Lattner48b383b02003-11-25 01:02:51 +00002033<p>For example, let's consider a C code fragment and how it gets
2034compiled to LLVM:</p>
Chris Lattner33fd7022004-04-05 01:30:49 +00002035
2036<pre>
2037 struct RT {
2038 char A;
2039 int B[10][20];
2040 char C;
2041 };
2042 struct ST {
2043 int X;
2044 double Y;
2045 struct RT Z;
2046 };
2047
2048 int *foo(struct ST *s) {
2049 return &amp;s[1].Z.B[5][13];
2050 }
2051</pre>
2052
Misha Brukman76307852003-11-08 01:05:38 +00002053<p>The LLVM code generated by the GCC frontend is:</p>
Chris Lattner33fd7022004-04-05 01:30:49 +00002054
2055<pre>
2056 %RT = type { sbyte, [10 x [20 x int]], sbyte }
2057 %ST = type { int, double, %RT }
2058
Brian Gaeke317ef962004-07-02 21:08:14 +00002059 implementation
2060
2061 int* %foo(%ST* %s) {
2062 entry:
2063 %reg = getelementptr %ST* %s, int 1, uint 2, uint 1, int 5, int 13
Chris Lattner33fd7022004-04-05 01:30:49 +00002064 ret int* %reg
2065 }
2066</pre>
2067
Chris Lattner590645f2002-04-14 06:13:44 +00002068<h5>Semantics:</h5>
Chris Lattner33fd7022004-04-05 01:30:49 +00002069
2070<p>The index types specified for the '<tt>getelementptr</tt>' instruction depend
John Criswell4a3327e2005-05-13 22:25:59 +00002071on the pointer type that is being indexed into. <a href="#t_pointer">Pointer</a>
Chris Lattner10ee9652004-06-03 22:57:15 +00002072and <a href="#t_array">array</a> types require <tt>uint</tt>, <tt>int</tt>,
2073<tt>ulong</tt>, or <tt>long</tt> values, and <a href="#t_struct">structure</a>
Chris Lattner33fd7022004-04-05 01:30:49 +00002074types require <tt>uint</tt> <b>constants</b>.</p>
2075
Misha Brukman76307852003-11-08 01:05:38 +00002076<p>In the example above, the first index is indexing into the '<tt>%ST*</tt>'
Chris Lattner33fd7022004-04-05 01:30:49 +00002077type, which is a pointer, yielding a '<tt>%ST</tt>' = '<tt>{ int, double, %RT
2078}</tt>' type, a structure. The second index indexes into the third element of
2079the structure, yielding a '<tt>%RT</tt>' = '<tt>{ sbyte, [10 x [20 x int]],
2080sbyte }</tt>' type, another structure. The third index indexes into the second
2081element of the structure, yielding a '<tt>[10 x [20 x int]]</tt>' type, an
2082array. The two dimensions of the array are subscripted into, yielding an
John Criswell88190562005-05-16 16:17:45 +00002083'<tt>int</tt>' type. The '<tt>getelementptr</tt>' instruction returns a pointer
Chris Lattner33fd7022004-04-05 01:30:49 +00002084to this element, thus computing a value of '<tt>int*</tt>' type.</p>
2085
Chris Lattner48b383b02003-11-25 01:02:51 +00002086<p>Note that it is perfectly legal to index partially through a
2087structure, returning a pointer to an inner element. Because of this,
2088the LLVM code for the given testcase is equivalent to:</p>
Chris Lattner33fd7022004-04-05 01:30:49 +00002089
2090<pre>
Chris Lattner455fc8c2005-03-07 22:13:59 +00002091 int* %foo(%ST* %s) {
Chris Lattner33fd7022004-04-05 01:30:49 +00002092 %t1 = getelementptr %ST* %s, int 1 <i>; yields %ST*:%t1</i>
2093 %t2 = getelementptr %ST* %t1, int 0, uint 2 <i>; yields %RT*:%t2</i>
2094 %t3 = getelementptr %RT* %t2, int 0, uint 1 <i>; yields [10 x [20 x int]]*:%t3</i>
2095 %t4 = getelementptr [10 x [20 x int]]* %t3, int 0, int 5 <i>; yields [20 x int]*:%t4</i>
2096 %t5 = getelementptr [20 x int]* %t4, int 0, int 13 <i>; yields int*:%t5</i>
2097 ret int* %t5
2098 }
Chris Lattnera8292f32002-05-06 22:08:29 +00002099</pre>
Chris Lattnerc0ad71e2005-06-24 17:22:57 +00002100
2101<p>Note that it is undefined to access an array out of bounds: array and
2102pointer indexes must always be within the defined bounds of the array type.
2103The one exception for this rules is zero length arrays. These arrays are
2104defined to be accessible as variable length arrays, which requires access
2105beyond the zero'th element.</p>
2106
Chris Lattner590645f2002-04-14 06:13:44 +00002107<h5>Example:</h5>
Chris Lattnerc0ad71e2005-06-24 17:22:57 +00002108
Chris Lattner33fd7022004-04-05 01:30:49 +00002109<pre>
2110 <i>; yields [12 x ubyte]*:aptr</i>
2111 %aptr = getelementptr {int, [12 x ubyte]}* %sptr, long 0, uint 1
2112</pre>
2113
2114</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +00002115<!-- ======================================================================= -->
Chris Lattner48b383b02003-11-25 01:02:51 +00002116<div class="doc_subsection"> <a name="otherops">Other Operations</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +00002117<div class="doc_text">
John Criswell417228d2004-04-09 16:48:45 +00002118<p>The instructions in this category are the "miscellaneous"
Chris Lattner48b383b02003-11-25 01:02:51 +00002119instructions, which defy better classification.</p>
Misha Brukman76307852003-11-08 01:05:38 +00002120</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +00002121<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +00002122<div class="doc_subsubsection"> <a name="i_phi">'<tt>phi</tt>'
2123Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +00002124<div class="doc_text">
Chris Lattner70de6632001-07-09 00:26:23 +00002125<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00002126<pre> &lt;result&gt; = phi &lt;ty&gt; [ &lt;val0&gt;, &lt;label0&gt;], ...<br></pre>
Chris Lattner70de6632001-07-09 00:26:23 +00002127<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00002128<p>The '<tt>phi</tt>' instruction is used to implement the &#966; node in
2129the SSA graph representing the function.</p>
Chris Lattner70de6632001-07-09 00:26:23 +00002130<h5>Arguments:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00002131<p>The type of the incoming values are specified with the first type
2132field. After this, the '<tt>phi</tt>' instruction takes a list of pairs
2133as arguments, with one pair for each predecessor basic block of the
2134current block. Only values of <a href="#t_firstclass">first class</a>
2135type may be used as the value arguments to the PHI node. Only labels
2136may be used as the label arguments.</p>
2137<p>There must be no non-phi instructions between the start of a basic
2138block and the PHI instructions: i.e. PHI instructions must be first in
2139a basic block.</p>
Chris Lattner70de6632001-07-09 00:26:23 +00002140<h5>Semantics:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00002141<p>At runtime, the '<tt>phi</tt>' instruction logically takes on the
2142value specified by the parameter, depending on which basic block we
2143came from in the last <a href="#terminators">terminator</a> instruction.</p>
Chris Lattnera8292f32002-05-06 22:08:29 +00002144<h5>Example:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00002145<pre>Loop: ; Infinite loop that counts from 0 on up...<br> %indvar = phi uint [ 0, %LoopHeader ], [ %nextindvar, %Loop ]<br> %nextindvar = add uint %indvar, 1<br> br label %Loop<br></pre>
Misha Brukman76307852003-11-08 01:05:38 +00002146</div>
Chris Lattnerb53c28d2004-03-12 05:50:16 +00002147
Chris Lattnera8292f32002-05-06 22:08:29 +00002148<!-- _______________________________________________________________________ -->
Chris Lattnerb53c28d2004-03-12 05:50:16 +00002149<div class="doc_subsubsection">
2150 <a name="i_cast">'<tt>cast .. to</tt>' Instruction</a>
2151</div>
2152
Misha Brukman76307852003-11-08 01:05:38 +00002153<div class="doc_text">
Chris Lattnerb53c28d2004-03-12 05:50:16 +00002154
Chris Lattnera8292f32002-05-06 22:08:29 +00002155<h5>Syntax:</h5>
Chris Lattnerb53c28d2004-03-12 05:50:16 +00002156
2157<pre>
2158 &lt;result&gt; = cast &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
Chris Lattnera8292f32002-05-06 22:08:29 +00002159</pre>
Chris Lattnerb53c28d2004-03-12 05:50:16 +00002160
Chris Lattnera8292f32002-05-06 22:08:29 +00002161<h5>Overview:</h5>
Chris Lattnerb53c28d2004-03-12 05:50:16 +00002162
2163<p>
2164The '<tt>cast</tt>' instruction is used as the primitive means to convert
2165integers to floating point, change data type sizes, and break type safety (by
2166casting pointers).
2167</p>
2168
2169
Chris Lattnera8292f32002-05-06 22:08:29 +00002170<h5>Arguments:</h5>
Chris Lattnerb53c28d2004-03-12 05:50:16 +00002171
2172<p>
2173The '<tt>cast</tt>' instruction takes a value to cast, which must be a first
2174class value, and a type to cast it to, which must also be a <a
2175href="#t_firstclass">first class</a> type.
2176</p>
2177
Chris Lattnera8292f32002-05-06 22:08:29 +00002178<h5>Semantics:</h5>
Chris Lattnerb53c28d2004-03-12 05:50:16 +00002179
2180<p>
2181This instruction follows the C rules for explicit casts when determining how the
2182data being cast must change to fit in its new container.
2183</p>
2184
2185<p>
2186When casting to bool, any value that would be considered true in the context of
2187a C '<tt>if</tt>' condition is converted to the boolean '<tt>true</tt>' values,
2188all else are '<tt>false</tt>'.
2189</p>
2190
2191<p>
2192When extending an integral value from a type of one signness to another (for
2193example '<tt>sbyte</tt>' to '<tt>ulong</tt>'), the value is sign-extended if the
2194<b>source</b> value is signed, and zero-extended if the source value is
2195unsigned. <tt>bool</tt> values are always zero extended into either zero or
2196one.
2197</p>
2198
Chris Lattner70de6632001-07-09 00:26:23 +00002199<h5>Example:</h5>
Chris Lattnerb53c28d2004-03-12 05:50:16 +00002200
2201<pre>
2202 %X = cast int 257 to ubyte <i>; yields ubyte:1</i>
Chris Lattnerd8f8ede2002-06-25 18:03:17 +00002203 %Y = cast int 123 to bool <i>; yields bool:true</i>
Chris Lattner70de6632001-07-09 00:26:23 +00002204</pre>
Misha Brukman76307852003-11-08 01:05:38 +00002205</div>
Chris Lattnerb53c28d2004-03-12 05:50:16 +00002206
2207<!-- _______________________________________________________________________ -->
2208<div class="doc_subsubsection">
2209 <a name="i_select">'<tt>select</tt>' Instruction</a>
2210</div>
2211
2212<div class="doc_text">
2213
2214<h5>Syntax:</h5>
2215
2216<pre>
2217 &lt;result&gt; = select bool &lt;cond&gt;, &lt;ty&gt; &lt;val1&gt;, &lt;ty&gt; &lt;val2&gt; <i>; yields ty</i>
2218</pre>
2219
2220<h5>Overview:</h5>
2221
2222<p>
2223The '<tt>select</tt>' instruction is used to choose one value based on a
2224condition, without branching.
2225</p>
2226
2227
2228<h5>Arguments:</h5>
2229
2230<p>
2231The '<tt>select</tt>' instruction requires a boolean value indicating the condition, and two values of the same <a href="#t_firstclass">first class</a> type.
2232</p>
2233
2234<h5>Semantics:</h5>
2235
2236<p>
2237If the boolean condition evaluates to true, the instruction returns the first
John Criswell88190562005-05-16 16:17:45 +00002238value argument; otherwise, it returns the second value argument.
Chris Lattnerb53c28d2004-03-12 05:50:16 +00002239</p>
2240
2241<h5>Example:</h5>
2242
2243<pre>
2244 %X = select bool true, ubyte 17, ubyte 42 <i>; yields ubyte:17</i>
2245</pre>
2246</div>
2247
2248
Robert Bocchinob5f210f2006-01-05 17:37:02 +00002249<!-- _______________________________________________________________________ -->
2250<div class="doc_subsubsection">
2251 <a name="i_extractelement">'<tt>extractelement</tt>' Instruction</a>
2252</div>
Chris Lattnerb53c28d2004-03-12 05:50:16 +00002253
Robert Bocchinob5f210f2006-01-05 17:37:02 +00002254<div class="doc_text">
2255
2256<h5>Syntax:</h5>
2257
2258<pre>
2259 &lt;result&gt; = extractelement &lt;n x &lt;ty&gt;&gt; &lt;val&gt;, uint &lt;idx&gt; <i>; yields &lt;ty&gt;</i>
2260</pre>
2261
2262<h5>Overview:</h5>
2263
2264<p>
2265The '<tt>extractelement</tt>' instruction extracts a single scalar
2266element from a vector at a specified index.
2267</p>
2268
2269
2270<h5>Arguments:</h5>
2271
2272<p>
2273The first operand of an '<tt>extractelement</tt>' instruction is a
2274value of <a href="#t_packed">packed</a> type. The second operand is
2275an index indicating the position from which to extract the element.
2276The index may be a variable.</p>
2277
2278<h5>Semantics:</h5>
2279
2280<p>
2281The result is a scalar of the same type as the element type of
2282<tt>val</tt>. Its value is the value at position <tt>idx</tt> of
2283<tt>val</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
2284results are undefined.
2285</p>
2286
2287<h5>Example:</h5>
2288
2289<pre>
2290 %result = extractelement &lt;4 x int&gt; %vec, uint 0 <i>; yields int</i>
2291</pre>
2292</div>
Chris Lattnerb53c28d2004-03-12 05:50:16 +00002293
2294
Chris Lattner70de6632001-07-09 00:26:23 +00002295<!-- _______________________________________________________________________ -->
Chris Lattnere23c1392005-05-06 05:47:36 +00002296<div class="doc_subsubsection">
2297 <a name="i_call">'<tt>call</tt>' Instruction</a>
2298</div>
2299
Misha Brukman76307852003-11-08 01:05:38 +00002300<div class="doc_text">
Chris Lattnere23c1392005-05-06 05:47:36 +00002301
Chris Lattner2f7c9632001-06-06 20:29:01 +00002302<h5>Syntax:</h5>
Chris Lattnere23c1392005-05-06 05:47:36 +00002303<pre>
Chris Lattner0132aff2005-05-06 22:57:40 +00002304 &lt;result&gt; = [tail] call [<a href="#callingconv">cconv</a>] &lt;ty&gt;* &lt;fnptrval&gt;(&lt;param list&gt;)
Chris Lattnere23c1392005-05-06 05:47:36 +00002305</pre>
2306
Chris Lattner2f7c9632001-06-06 20:29:01 +00002307<h5>Overview:</h5>
Chris Lattnere23c1392005-05-06 05:47:36 +00002308
Misha Brukman76307852003-11-08 01:05:38 +00002309<p>The '<tt>call</tt>' instruction represents a simple function call.</p>
Chris Lattnere23c1392005-05-06 05:47:36 +00002310
Chris Lattner2f7c9632001-06-06 20:29:01 +00002311<h5>Arguments:</h5>
Chris Lattnere23c1392005-05-06 05:47:36 +00002312
Misha Brukman76307852003-11-08 01:05:38 +00002313<p>This instruction requires several arguments:</p>
Chris Lattnere23c1392005-05-06 05:47:36 +00002314
Chris Lattnera8292f32002-05-06 22:08:29 +00002315<ol>
Chris Lattner48b383b02003-11-25 01:02:51 +00002316 <li>
Chris Lattner0132aff2005-05-06 22:57:40 +00002317 <p>The optional "tail" marker indicates whether the callee function accesses
2318 any allocas or varargs in the caller. If the "tail" marker is present, the
Chris Lattnere23c1392005-05-06 05:47:36 +00002319 function call is eligible for tail call optimization. Note that calls may
2320 be marked "tail" even if they do not occur before a <a
2321 href="#i_ret"><tt>ret</tt></a> instruction.
Chris Lattner48b383b02003-11-25 01:02:51 +00002322 </li>
2323 <li>
Chris Lattner0132aff2005-05-06 22:57:40 +00002324 <p>The optional "cconv" marker indicates which <a href="callingconv">calling
2325 convention</a> the call should use. If none is specified, the call defaults
2326 to using C calling conventions.
2327 </li>
2328 <li>
Chris Lattnere23c1392005-05-06 05:47:36 +00002329 <p>'<tt>ty</tt>': shall be the signature of the pointer to function value
2330 being invoked. The argument types must match the types implied by this
John Criswell88190562005-05-16 16:17:45 +00002331 signature. This type can be omitted if the function is not varargs and
2332 if the function type does not return a pointer to a function.</p>
Chris Lattnere23c1392005-05-06 05:47:36 +00002333 </li>
2334 <li>
2335 <p>'<tt>fnptrval</tt>': An LLVM value containing a pointer to a function to
2336 be invoked. In most cases, this is a direct function invocation, but
2337 indirect <tt>call</tt>s are just as possible, calling an arbitrary pointer
John Criswell88190562005-05-16 16:17:45 +00002338 to function value.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +00002339 </li>
2340 <li>
2341 <p>'<tt>function args</tt>': argument list whose types match the
Reid Spencerd845d162005-05-01 22:22:57 +00002342 function signature argument types. All arguments must be of
2343 <a href="#t_firstclass">first class</a> type. If the function signature
2344 indicates the function accepts a variable number of arguments, the extra
2345 arguments can be specified.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +00002346 </li>
Chris Lattnera8292f32002-05-06 22:08:29 +00002347</ol>
Chris Lattnere23c1392005-05-06 05:47:36 +00002348
Chris Lattner2f7c9632001-06-06 20:29:01 +00002349<h5>Semantics:</h5>
Chris Lattnere23c1392005-05-06 05:47:36 +00002350
Chris Lattner48b383b02003-11-25 01:02:51 +00002351<p>The '<tt>call</tt>' instruction is used to cause control flow to
2352transfer to a specified function, with its incoming arguments bound to
2353the specified values. Upon a '<tt><a href="#i_ret">ret</a></tt>'
2354instruction in the called function, control flow continues with the
2355instruction after the function call, and the return value of the
2356function is bound to the result argument. This is a simpler case of
2357the <a href="#i_invoke">invoke</a> instruction.</p>
Chris Lattnere23c1392005-05-06 05:47:36 +00002358
Chris Lattner2f7c9632001-06-06 20:29:01 +00002359<h5>Example:</h5>
Chris Lattnere23c1392005-05-06 05:47:36 +00002360
2361<pre>
2362 %retval = call int %test(int %argc)
2363 call int(sbyte*, ...) *%printf(sbyte* %msg, int 12, sbyte 42);
2364 %X = tail call int %foo()
Chris Lattner0132aff2005-05-06 22:57:40 +00002365 %Y = tail call <a href="#callingconv">fastcc</a> int %foo()
Chris Lattnere23c1392005-05-06 05:47:36 +00002366</pre>
2367
Misha Brukman76307852003-11-08 01:05:38 +00002368</div>
Chris Lattner6a4a0492004-09-27 21:51:25 +00002369
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00002370<!-- _______________________________________________________________________ -->
Chris Lattner6a4a0492004-09-27 21:51:25 +00002371<div class="doc_subsubsection">
Chris Lattner33337472006-01-13 23:26:01 +00002372 <a name="i_va_arg">'<tt>va_arg</tt>' Instruction</a>
Chris Lattner6a4a0492004-09-27 21:51:25 +00002373</div>
2374
Misha Brukman76307852003-11-08 01:05:38 +00002375<div class="doc_text">
Chris Lattner6a4a0492004-09-27 21:51:25 +00002376
Chris Lattner26ca62e2003-10-18 05:51:36 +00002377<h5>Syntax:</h5>
Chris Lattner6a4a0492004-09-27 21:51:25 +00002378
2379<pre>
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00002380 &lt;resultval&gt; = va_arg &lt;va_list*&gt; &lt;arglist&gt;, &lt;argty&gt;
Chris Lattner6a4a0492004-09-27 21:51:25 +00002381</pre>
2382
Chris Lattner26ca62e2003-10-18 05:51:36 +00002383<h5>Overview:</h5>
Chris Lattner6a4a0492004-09-27 21:51:25 +00002384
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00002385<p>The '<tt>va_arg</tt>' instruction is used to access arguments passed through
Chris Lattner6a4a0492004-09-27 21:51:25 +00002386the "variable argument" area of a function call. It is used to implement the
2387<tt>va_arg</tt> macro in C.</p>
2388
Chris Lattner26ca62e2003-10-18 05:51:36 +00002389<h5>Arguments:</h5>
Chris Lattner6a4a0492004-09-27 21:51:25 +00002390
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00002391<p>This instruction takes a <tt>va_list*</tt> value and the type of
2392the argument. It returns a value of the specified argument type and
Jeff Cohendc6bfea2005-11-11 02:15:27 +00002393increments the <tt>va_list</tt> to point to the next argument. Again, the
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00002394actual type of <tt>va_list</tt> is target specific.</p>
Chris Lattner6a4a0492004-09-27 21:51:25 +00002395
Chris Lattner26ca62e2003-10-18 05:51:36 +00002396<h5>Semantics:</h5>
Chris Lattner6a4a0492004-09-27 21:51:25 +00002397
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00002398<p>The '<tt>va_arg</tt>' instruction loads an argument of the specified
2399type from the specified <tt>va_list</tt> and causes the
2400<tt>va_list</tt> to point to the next argument. For more information,
2401see the variable argument handling <a href="#int_varargs">Intrinsic
2402Functions</a>.</p>
Chris Lattner6a4a0492004-09-27 21:51:25 +00002403
2404<p>It is legal for this instruction to be called in a function which does not
2405take a variable number of arguments, for example, the <tt>vfprintf</tt>
Misha Brukman76307852003-11-08 01:05:38 +00002406function.</p>
Chris Lattner6a4a0492004-09-27 21:51:25 +00002407
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00002408<p><tt>va_arg</tt> is an LLVM instruction instead of an <a
John Criswell88190562005-05-16 16:17:45 +00002409href="#intrinsics">intrinsic function</a> because it takes a type as an
Chris Lattner6a4a0492004-09-27 21:51:25 +00002410argument.</p>
2411
Chris Lattner26ca62e2003-10-18 05:51:36 +00002412<h5>Example:</h5>
Chris Lattner6a4a0492004-09-27 21:51:25 +00002413
2414<p>See the <a href="#int_varargs">variable argument processing</a> section.</p>
2415
Misha Brukman76307852003-11-08 01:05:38 +00002416</div>
Chris Lattner941515c2004-01-06 05:31:32 +00002417
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00002418<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +00002419<div class="doc_section"> <a name="intrinsics">Intrinsic Functions</a> </div>
2420<!-- *********************************************************************** -->
Chris Lattner941515c2004-01-06 05:31:32 +00002421
Misha Brukman76307852003-11-08 01:05:38 +00002422<div class="doc_text">
Chris Lattnerfee11462004-02-12 17:01:32 +00002423
2424<p>LLVM supports the notion of an "intrinsic function". These functions have
John Criswell88190562005-05-16 16:17:45 +00002425well known names and semantics and are required to follow certain
Chris Lattnerfee11462004-02-12 17:01:32 +00002426restrictions. Overall, these instructions represent an extension mechanism for
2427the LLVM language that does not require changing all of the transformations in
2428LLVM to add to the language (or the bytecode reader/writer, the parser,
2429etc...).</p>
2430
John Criswell88190562005-05-16 16:17:45 +00002431<p>Intrinsic function names must all start with an "<tt>llvm.</tt>" prefix. This
2432prefix is reserved in LLVM for intrinsic names; thus, functions may not be named
Chris Lattnerfee11462004-02-12 17:01:32 +00002433this. Intrinsic functions must always be external functions: you cannot define
2434the body of intrinsic functions. Intrinsic functions may only be used in call
2435or invoke instructions: it is illegal to take the address of an intrinsic
2436function. Additionally, because intrinsic functions are part of the LLVM
2437language, it is required that they all be documented here if any are added.</p>
2438
2439
John Criswell88190562005-05-16 16:17:45 +00002440<p>To learn how to add an intrinsic function, please see the <a
Chris Lattner90391c12005-05-11 03:35:57 +00002441href="ExtendingLLVM.html">Extending LLVM Guide</a>.
Chris Lattnerfee11462004-02-12 17:01:32 +00002442</p>
2443
Misha Brukman76307852003-11-08 01:05:38 +00002444</div>
Chris Lattner941515c2004-01-06 05:31:32 +00002445
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00002446<!-- ======================================================================= -->
Chris Lattner941515c2004-01-06 05:31:32 +00002447<div class="doc_subsection">
2448 <a name="int_varargs">Variable Argument Handling Intrinsics</a>
2449</div>
2450
Misha Brukman76307852003-11-08 01:05:38 +00002451<div class="doc_text">
Chris Lattner757528b0b2004-05-23 21:06:01 +00002452
Misha Brukman76307852003-11-08 01:05:38 +00002453<p>Variable argument support is defined in LLVM with the <a
Chris Lattner33337472006-01-13 23:26:01 +00002454 href="#i_va_arg"><tt>va_arg</tt></a> instruction and these three
Chris Lattner48b383b02003-11-25 01:02:51 +00002455intrinsic functions. These functions are related to the similarly
2456named macros defined in the <tt>&lt;stdarg.h&gt;</tt> header file.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +00002457
Chris Lattner48b383b02003-11-25 01:02:51 +00002458<p>All of these functions operate on arguments that use a
2459target-specific value type "<tt>va_list</tt>". The LLVM assembly
2460language reference manual does not define what this type is, so all
2461transformations should be prepared to handle intrinsics with any type
2462used.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +00002463
Misha Brukman76307852003-11-08 01:05:38 +00002464<p>This example shows how the <a href="#i_vanext"><tt>vanext</tt></a>
Chris Lattner48b383b02003-11-25 01:02:51 +00002465instruction and the variable argument handling intrinsic functions are
2466used.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +00002467
Chris Lattnerfee11462004-02-12 17:01:32 +00002468<pre>
2469int %test(int %X, ...) {
2470 ; Initialize variable argument processing
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00002471 %ap = alloca sbyte*
2472 call void %<a href="#i_va_start">llvm.va_start</a>(sbyte** %ap)
Chris Lattnerfee11462004-02-12 17:01:32 +00002473
2474 ; Read a single integer argument
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00002475 %tmp = va_arg sbyte** %ap, int
Chris Lattnerfee11462004-02-12 17:01:32 +00002476
2477 ; Demonstrate usage of llvm.va_copy and llvm.va_end
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00002478 %aq = alloca sbyte*
Andrew Lenharth5305ea52005-06-22 20:38:11 +00002479 call void %<a href="#i_va_copy">llvm.va_copy</a>(sbyte** %aq, sbyte** %ap)
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00002480 call void %<a href="#i_va_end">llvm.va_end</a>(sbyte** %aq)
Chris Lattnerfee11462004-02-12 17:01:32 +00002481
2482 ; Stop processing of arguments.
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00002483 call void %<a href="#i_va_end">llvm.va_end</a>(sbyte** %ap)
Chris Lattnerfee11462004-02-12 17:01:32 +00002484 ret int %tmp
2485}
2486</pre>
Misha Brukman76307852003-11-08 01:05:38 +00002487</div>
Chris Lattner941515c2004-01-06 05:31:32 +00002488
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00002489<!-- _______________________________________________________________________ -->
Chris Lattner941515c2004-01-06 05:31:32 +00002490<div class="doc_subsubsection">
2491 <a name="i_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a>
2492</div>
2493
2494
Misha Brukman76307852003-11-08 01:05:38 +00002495<div class="doc_text">
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00002496<h5>Syntax:</h5>
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00002497<pre> declare void %llvm.va_start(&lt;va_list&gt;* &lt;arglist&gt;)<br></pre>
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00002498<h5>Overview:</h5>
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00002499<P>The '<tt>llvm.va_start</tt>' intrinsic initializes
2500<tt>*&lt;arglist&gt;</tt> for subsequent use by <tt><a
2501href="#i_va_arg">va_arg</a></tt>.</p>
2502
2503<h5>Arguments:</h5>
2504
2505<P>The argument is a pointer to a <tt>va_list</tt> element to initialize.</p>
2506
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00002507<h5>Semantics:</h5>
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00002508
2509<P>The '<tt>llvm.va_start</tt>' intrinsic works just like the <tt>va_start</tt>
2510macro available in C. In a target-dependent way, it initializes the
2511<tt>va_list</tt> element the argument points to, so that the next call to
2512<tt>va_arg</tt> will produce the first variable argument passed to the function.
2513Unlike the C <tt>va_start</tt> macro, this intrinsic does not need to know the
2514last argument of the function, the compiler can figure that out.</p>
2515
Misha Brukman76307852003-11-08 01:05:38 +00002516</div>
Chris Lattner941515c2004-01-06 05:31:32 +00002517
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00002518<!-- _______________________________________________________________________ -->
Chris Lattner941515c2004-01-06 05:31:32 +00002519<div class="doc_subsubsection">
2520 <a name="i_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a>
2521</div>
2522
Misha Brukman76307852003-11-08 01:05:38 +00002523<div class="doc_text">
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00002524<h5>Syntax:</h5>
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00002525<pre> declare void %llvm.va_end(&lt;va_list*&gt; &lt;arglist&gt;)<br></pre>
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00002526<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +00002527<p>The '<tt>llvm.va_end</tt>' intrinsic destroys <tt>&lt;arglist&gt;</tt>
2528which has been initialized previously with <tt><a href="#i_va_start">llvm.va_start</a></tt>
2529or <tt><a href="#i_va_copy">llvm.va_copy</a></tt>.</p>
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00002530<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +00002531<p>The argument is a <tt>va_list</tt> to destroy.</p>
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00002532<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +00002533<p>The '<tt>llvm.va_end</tt>' intrinsic works just like the <tt>va_end</tt>
Chris Lattner48b383b02003-11-25 01:02:51 +00002534macro available in C. In a target-dependent way, it destroys the <tt>va_list</tt>.
2535Calls to <a href="#i_va_start"><tt>llvm.va_start</tt></a> and <a
2536 href="#i_va_copy"><tt>llvm.va_copy</tt></a> must be matched exactly
2537with calls to <tt>llvm.va_end</tt>.</p>
Misha Brukman76307852003-11-08 01:05:38 +00002538</div>
Chris Lattner941515c2004-01-06 05:31:32 +00002539
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00002540<!-- _______________________________________________________________________ -->
Chris Lattner941515c2004-01-06 05:31:32 +00002541<div class="doc_subsubsection">
2542 <a name="i_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a>
2543</div>
2544
Misha Brukman76307852003-11-08 01:05:38 +00002545<div class="doc_text">
Chris Lattner757528b0b2004-05-23 21:06:01 +00002546
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00002547<h5>Syntax:</h5>
Chris Lattner757528b0b2004-05-23 21:06:01 +00002548
2549<pre>
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00002550 declare void %llvm.va_copy(&lt;va_list&gt;* &lt;destarglist&gt;,
Andrew Lenharth5305ea52005-06-22 20:38:11 +00002551 &lt;va_list&gt;* &lt;srcarglist&gt;)
Chris Lattner757528b0b2004-05-23 21:06:01 +00002552</pre>
2553
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00002554<h5>Overview:</h5>
Chris Lattner757528b0b2004-05-23 21:06:01 +00002555
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00002556<p>The '<tt>llvm.va_copy</tt>' intrinsic copies the current argument position from
2557the source argument list to the destination argument list.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +00002558
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00002559<h5>Arguments:</h5>
Chris Lattner757528b0b2004-05-23 21:06:01 +00002560
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00002561<p>The first argument is a pointer to a <tt>va_list</tt> element to initialize.
Andrew Lenharth5305ea52005-06-22 20:38:11 +00002562The second argument is a pointer to a <tt>va_list</tt> element to copy from.</p>
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00002563
Chris Lattner757528b0b2004-05-23 21:06:01 +00002564
Chris Lattnerbd64b4e2003-05-08 04:57:36 +00002565<h5>Semantics:</h5>
Chris Lattner757528b0b2004-05-23 21:06:01 +00002566
Andrew Lenharth5fb787c2005-06-18 18:28:17 +00002567<p>The '<tt>llvm.va_copy</tt>' intrinsic works just like the <tt>va_copy</tt> macro
2568available in C. In a target-dependent way, it copies the source
2569<tt>va_list</tt> element into the destination list. This intrinsic is necessary
2570because the <tt><a href="i_va_begin">llvm.va_begin</a></tt> intrinsic may be
Chris Lattner757528b0b2004-05-23 21:06:01 +00002571arbitrarily complex and require memory allocation, for example.</p>
2572
Misha Brukman76307852003-11-08 01:05:38 +00002573</div>
Chris Lattner941515c2004-01-06 05:31:32 +00002574
Chris Lattnerfee11462004-02-12 17:01:32 +00002575<!-- ======================================================================= -->
2576<div class="doc_subsection">
Chris Lattner757528b0b2004-05-23 21:06:01 +00002577 <a name="int_gc">Accurate Garbage Collection Intrinsics</a>
2578</div>
2579
2580<div class="doc_text">
2581
2582<p>
2583LLVM support for <a href="GarbageCollection.html">Accurate Garbage
2584Collection</a> requires the implementation and generation of these intrinsics.
2585These intrinsics allow identification of <a href="#i_gcroot">GC roots on the
2586stack</a>, as well as garbage collector implementations that require <a
2587href="#i_gcread">read</a> and <a href="#i_gcwrite">write</a> barriers.
2588Front-ends for type-safe garbage collected languages should generate these
2589intrinsics to make use of the LLVM garbage collectors. For more details, see <a
2590href="GarbageCollection.html">Accurate Garbage Collection with LLVM</a>.
2591</p>
2592</div>
2593
2594<!-- _______________________________________________________________________ -->
2595<div class="doc_subsubsection">
2596 <a name="i_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a>
2597</div>
2598
2599<div class="doc_text">
2600
2601<h5>Syntax:</h5>
2602
2603<pre>
Reid Spencer7821d062005-04-26 20:50:44 +00002604 declare void %llvm.gcroot(&lt;ty&gt;** %ptrloc, &lt;ty2&gt;* %metadata)
Chris Lattner757528b0b2004-05-23 21:06:01 +00002605</pre>
2606
2607<h5>Overview:</h5>
2608
John Criswelldfe6a862004-12-10 15:51:16 +00002609<p>The '<tt>llvm.gcroot</tt>' intrinsic declares the existence of a GC root to
Chris Lattner757528b0b2004-05-23 21:06:01 +00002610the code generator, and allows some metadata to be associated with it.</p>
2611
2612<h5>Arguments:</h5>
2613
2614<p>The first argument specifies the address of a stack object that contains the
2615root pointer. The second pointer (which must be either a constant or a global
2616value address) contains the meta-data to be associated with the root.</p>
2617
2618<h5>Semantics:</h5>
2619
2620<p>At runtime, a call to this intrinsics stores a null pointer into the "ptrloc"
2621location. At compile-time, the code generator generates information to allow
2622the runtime to find the pointer at GC safe points.
2623</p>
2624
2625</div>
2626
2627
2628<!-- _______________________________________________________________________ -->
2629<div class="doc_subsubsection">
2630 <a name="i_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a>
2631</div>
2632
2633<div class="doc_text">
2634
2635<h5>Syntax:</h5>
2636
2637<pre>
Reid Spencer7821d062005-04-26 20:50:44 +00002638 declare sbyte* %llvm.gcread(sbyte** %Ptr)
Chris Lattner757528b0b2004-05-23 21:06:01 +00002639</pre>
2640
2641<h5>Overview:</h5>
2642
2643<p>The '<tt>llvm.gcread</tt>' intrinsic identifies reads of references from heap
2644locations, allowing garbage collector implementations that require read
2645barriers.</p>
2646
2647<h5>Arguments:</h5>
2648
2649<p>The argument is the address to read from, which should be an address
2650allocated from the garbage collector.</p>
2651
2652<h5>Semantics:</h5>
2653
2654<p>The '<tt>llvm.gcread</tt>' intrinsic has the same semantics as a load
2655instruction, but may be replaced with substantially more complex code by the
2656garbage collector runtime, as needed.</p>
2657
2658</div>
2659
2660
2661<!-- _______________________________________________________________________ -->
2662<div class="doc_subsubsection">
2663 <a name="i_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a>
2664</div>
2665
2666<div class="doc_text">
2667
2668<h5>Syntax:</h5>
2669
2670<pre>
Reid Spencer7821d062005-04-26 20:50:44 +00002671 declare void %llvm.gcwrite(sbyte* %P1, sbyte** %P2)
Chris Lattner757528b0b2004-05-23 21:06:01 +00002672</pre>
2673
2674<h5>Overview:</h5>
2675
2676<p>The '<tt>llvm.gcwrite</tt>' intrinsic identifies writes of references to heap
2677locations, allowing garbage collector implementations that require write
2678barriers (such as generational or reference counting collectors).</p>
2679
2680<h5>Arguments:</h5>
2681
2682<p>The first argument is the reference to store, and the second is the heap
2683location to store to.</p>
2684
2685<h5>Semantics:</h5>
2686
2687<p>The '<tt>llvm.gcwrite</tt>' intrinsic has the same semantics as a store
2688instruction, but may be replaced with substantially more complex code by the
2689garbage collector runtime, as needed.</p>
2690
2691</div>
2692
2693
2694
2695<!-- ======================================================================= -->
2696<div class="doc_subsection">
Chris Lattner3649c3a2004-02-14 04:08:35 +00002697 <a name="int_codegen">Code Generator Intrinsics</a>
2698</div>
2699
2700<div class="doc_text">
2701<p>
2702These intrinsics are provided by LLVM to expose special features that may only
2703be implemented with code generator support.
2704</p>
2705
2706</div>
2707
2708<!-- _______________________________________________________________________ -->
2709<div class="doc_subsubsection">
Nate Begeman611ecc02006-01-13 23:02:51 +00002710 <a name="i_bswap_i16">'<tt>llvm.bswap.i16</tt>' Intrinsic</a>
2711</div>
2712
2713<div class="doc_text">
2714
2715<h5>Syntax:</h5>
2716<pre>
2717 declare ushort %llvm.bswap.i16( ushort &lt;id&gt; )
2718</pre>
2719
2720<h5>Overview:</h5>
2721
2722<p>
2723The '<tt>llvm.bwsap.i16</tt>' intrinsic is used to byteswap a 16 bit quantity.
2724This is useful for performing operations on data that is not in the target's
2725native byte order.
2726</p>
2727
2728<h5>Semantics:</h5>
2729
2730<p>
2731This intrinsic returns a ushort value that has the two bytes of the input ushort
2732swapped.
2733</p>
2734
2735</div>
2736
2737<!-- _______________________________________________________________________ -->
2738<div class="doc_subsubsection">
2739 <a name="i_bswap_i32">'<tt>llvm.bswap.i32</tt>' Intrinsic</a>
2740</div>
2741
2742<div class="doc_text">
2743
2744<h5>Syntax:</h5>
2745<pre>
2746 declare uint %llvm.bswap.i32( uint &lt;id&gt; )
2747</pre>
2748
2749<h5>Overview:</h5>
2750
2751<p>
2752The '<tt>llvm.bwsap.i32</tt>' intrinsic is used to byteswap a 32 bit quantity.
2753This is useful for performing operations on data that is not in the target's
2754native byte order.
2755</p>
2756
2757<h5>Semantics:</h5>
2758
2759<p>
2760This intrinsic returns a uint value that has the four bytes of the input uint
2761swapped, so that if the input bytes are numbered 0, 1, 2, 3 then the returned
2762uint will have its bytes in 3, 2, 1, 0 order.
2763</p>
2764
2765</div>
2766
2767<!-- _______________________________________________________________________ -->
2768<div class="doc_subsubsection">
2769 <a name="i_bswap_i64">'<tt>llvm.bswap.i64</tt>' Intrinsic</a>
2770</div>
2771
2772<div class="doc_text">
2773
2774<h5>Syntax:</h5>
2775<pre>
2776 declare ulong %llvm.bswap.i64( ulong &lt;id&gt; )
2777</pre>
2778
2779<h5>Overview:</h5>
2780
2781<p>
2782The '<tt>llvm.bwsap.i64</tt>' intrinsic is used to byteswap a 64 bit quantity.
2783This is useful for performing operations on data that is not in the target's
2784native byte order.
2785</p>
2786
2787<h5>Semantics:</h5>
2788
2789<p>
2790See the description for <a href="#i_bswap_i32"><tt>llvm.bswap.i32</tt></a>.
2791</p>
2792
2793</div>
2794
2795<!-- _______________________________________________________________________ -->
2796<div class="doc_subsubsection">
Chris Lattner3649c3a2004-02-14 04:08:35 +00002797 <a name="i_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a>
2798</div>
2799
2800<div class="doc_text">
2801
2802<h5>Syntax:</h5>
2803<pre>
Chris Lattnerb3d430e2006-01-13 01:20:27 +00002804 declare sbyte *%llvm.returnaddress(uint &lt;level&gt;)
Chris Lattner3649c3a2004-02-14 04:08:35 +00002805</pre>
2806
2807<h5>Overview:</h5>
2808
2809<p>
2810The '<tt>llvm.returnaddress</tt>' intrinsic returns a target-specific value
2811indicating the return address of the current function or one of its callers.
2812</p>
2813
2814<h5>Arguments:</h5>
2815
2816<p>
2817The argument to this intrinsic indicates which function to return the address
2818for. Zero indicates the calling function, one indicates its caller, etc. The
2819argument is <b>required</b> to be a constant integer value.
2820</p>
2821
2822<h5>Semantics:</h5>
2823
2824<p>
2825The '<tt>llvm.returnaddress</tt>' intrinsic either returns a pointer indicating
2826the return address of the specified call frame, or zero if it cannot be
2827identified. The value returned by this intrinsic is likely to be incorrect or 0
2828for arguments other than zero, so it should only be used for debugging purposes.
2829</p>
2830
2831<p>
2832Note that calling this intrinsic does not prevent function inlining or other
Chris Lattner2e6eb5f2005-03-07 20:30:51 +00002833aggressive transformations, so the value returned may not be that of the obvious
Chris Lattner3649c3a2004-02-14 04:08:35 +00002834source-language caller.
2835</p>
2836</div>
2837
2838
2839<!-- _______________________________________________________________________ -->
2840<div class="doc_subsubsection">
2841 <a name="i_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a>
2842</div>
2843
2844<div class="doc_text">
2845
2846<h5>Syntax:</h5>
2847<pre>
Chris Lattnerb3d430e2006-01-13 01:20:27 +00002848 declare sbyte *%llvm.frameaddress(uint &lt;level&gt;)
Chris Lattner3649c3a2004-02-14 04:08:35 +00002849</pre>
2850
2851<h5>Overview:</h5>
2852
2853<p>
2854The '<tt>llvm.frameaddress</tt>' intrinsic returns the target-specific frame
2855pointer value for the specified stack frame.
2856</p>
2857
2858<h5>Arguments:</h5>
2859
2860<p>
2861The argument to this intrinsic indicates which function to return the frame
2862pointer for. Zero indicates the calling function, one indicates its caller,
2863etc. The argument is <b>required</b> to be a constant integer value.
2864</p>
2865
2866<h5>Semantics:</h5>
2867
2868<p>
2869The '<tt>llvm.frameaddress</tt>' intrinsic either returns a pointer indicating
2870the frame address of the specified call frame, or zero if it cannot be
2871identified. The value returned by this intrinsic is likely to be incorrect or 0
2872for arguments other than zero, so it should only be used for debugging purposes.
2873</p>
2874
2875<p>
2876Note that calling this intrinsic does not prevent function inlining or other
Chris Lattner2e6eb5f2005-03-07 20:30:51 +00002877aggressive transformations, so the value returned may not be that of the obvious
Chris Lattner3649c3a2004-02-14 04:08:35 +00002878source-language caller.
2879</p>
2880</div>
2881
Chris Lattnerc8a2c222005-02-28 19:24:19 +00002882<!-- _______________________________________________________________________ -->
2883<div class="doc_subsubsection">
Chris Lattner2f0f0012006-01-13 02:03:13 +00002884 <a name="i_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a>
2885</div>
2886
2887<div class="doc_text">
2888
2889<h5>Syntax:</h5>
2890<pre>
2891 declare sbyte *%llvm.stacksave()
2892</pre>
2893
2894<h5>Overview:</h5>
2895
2896<p>
2897The '<tt>llvm.stacksave</tt>' intrinsic is used to remember the current state of
2898the function stack, for use with <a href="#i_stackrestore">
2899<tt>llvm.stackrestore</tt></a>. This is useful for implementing language
2900features like scoped automatic variable sized arrays in C99.
2901</p>
2902
2903<h5>Semantics:</h5>
2904
2905<p>
2906This intrinsic returns a opaque pointer value that can be passed to <a
2907href="#i_stackrestore"><tt>llvm.stackrestore</tt></a>. When an
2908<tt>llvm.stackrestore</tt> intrinsic is executed with a value saved from
2909<tt>llvm.stacksave</tt>, it effectively restores the state of the stack to the
2910state it was in when the <tt>llvm.stacksave</tt> intrinsic executed. In
2911practice, this pops any <a href="#i_alloca">alloca</a> blocks from the stack
2912that were allocated after the <tt>llvm.stacksave</tt> was executed.
2913</p>
2914
2915</div>
2916
2917<!-- _______________________________________________________________________ -->
2918<div class="doc_subsubsection">
2919 <a name="i_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a>
2920</div>
2921
2922<div class="doc_text">
2923
2924<h5>Syntax:</h5>
2925<pre>
2926 declare void %llvm.stackrestore(sbyte* %ptr)
2927</pre>
2928
2929<h5>Overview:</h5>
2930
2931<p>
2932The '<tt>llvm.stackrestore</tt>' intrinsic is used to restore the state of
2933the function stack to the state it was in when the corresponding <a
2934href="#llvm.stacksave"><tt>llvm.stacksave</tt></a> intrinsic executed. This is
2935useful for implementing language features like scoped automatic variable sized
2936arrays in C99.
2937</p>
2938
2939<h5>Semantics:</h5>
2940
2941<p>
2942See the description for <a href="#i_stacksave"><tt>llvm.stacksave</tt></a>.
2943</p>
2944
2945</div>
2946
2947
2948<!-- _______________________________________________________________________ -->
2949<div class="doc_subsubsection">
Chris Lattnerc8a2c222005-02-28 19:24:19 +00002950 <a name="i_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a>
2951</div>
2952
2953<div class="doc_text">
2954
2955<h5>Syntax:</h5>
2956<pre>
Reid Spencer7821d062005-04-26 20:50:44 +00002957 declare void %llvm.prefetch(sbyte * &lt;address&gt;,
2958 uint &lt;rw&gt;, uint &lt;locality&gt;)
Chris Lattnerc8a2c222005-02-28 19:24:19 +00002959</pre>
2960
2961<h5>Overview:</h5>
2962
2963
2964<p>
2965The '<tt>llvm.prefetch</tt>' intrinsic is a hint to the code generator to insert
John Criswell88190562005-05-16 16:17:45 +00002966a prefetch instruction if supported; otherwise, it is a noop. Prefetches have
2967no
2968effect on the behavior of the program but can change its performance
Chris Lattnerff851072005-02-28 19:47:14 +00002969characteristics.
Chris Lattnerc8a2c222005-02-28 19:24:19 +00002970</p>
2971
2972<h5>Arguments:</h5>
2973
2974<p>
2975<tt>address</tt> is the address to be prefetched, <tt>rw</tt> is the specifier
2976determining if the fetch should be for a read (0) or write (1), and
2977<tt>locality</tt> is a temporal locality specifier ranging from (0) - no
Chris Lattnerd3e641c2005-03-07 20:31:38 +00002978locality, to (3) - extremely local keep in cache. The <tt>rw</tt> and
Chris Lattnerc8a2c222005-02-28 19:24:19 +00002979<tt>locality</tt> arguments must be constant integers.
2980</p>
2981
2982<h5>Semantics:</h5>
2983
2984<p>
2985This intrinsic does not modify the behavior of the program. In particular,
2986prefetches cannot trap and do not produce a value. On targets that support this
2987intrinsic, the prefetch can provide hints to the processor cache for better
2988performance.
2989</p>
2990
2991</div>
2992
Andrew Lenharthb4427912005-03-28 20:05:49 +00002993<!-- _______________________________________________________________________ -->
2994<div class="doc_subsubsection">
2995 <a name="i_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a>
2996</div>
2997
2998<div class="doc_text">
2999
3000<h5>Syntax:</h5>
3001<pre>
Reid Spencer7821d062005-04-26 20:50:44 +00003002 declare void %llvm.pcmarker( uint &lt;id&gt; )
Andrew Lenharthb4427912005-03-28 20:05:49 +00003003</pre>
3004
3005<h5>Overview:</h5>
3006
3007
3008<p>
John Criswell88190562005-05-16 16:17:45 +00003009The '<tt>llvm.pcmarker</tt>' intrinsic is a method to export a Program Counter
3010(PC) in a region of
Andrew Lenharthb4427912005-03-28 20:05:49 +00003011code to simulators and other tools. The method is target specific, but it is
3012expected that the marker will use exported symbols to transmit the PC of the marker.
Jeff Cohendc6bfea2005-11-11 02:15:27 +00003013The marker makes no guarantees that it will remain with any specific instruction
Chris Lattnere64d41d2005-11-15 06:07:55 +00003014after optimizations. It is possible that the presence of a marker will inhibit
Andrew Lenharthb4427912005-03-28 20:05:49 +00003015optimizations. The intended use is to be inserted after optmizations to allow
John Criswell88190562005-05-16 16:17:45 +00003016correlations of simulation runs.
Andrew Lenharthb4427912005-03-28 20:05:49 +00003017</p>
3018
3019<h5>Arguments:</h5>
3020
3021<p>
3022<tt>id</tt> is a numerical id identifying the marker.
3023</p>
3024
3025<h5>Semantics:</h5>
3026
3027<p>
3028This intrinsic does not modify the behavior of the program. Backends that do not
3029support this intrinisic may ignore it.
3030</p>
3031
3032</div>
3033
Andrew Lenharth01aa5632005-11-11 16:47:30 +00003034<!-- _______________________________________________________________________ -->
3035<div class="doc_subsubsection">
3036 <a name="i_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a>
3037</div>
3038
3039<div class="doc_text">
3040
3041<h5>Syntax:</h5>
3042<pre>
3043 declare ulong %llvm.readcyclecounter( )
3044</pre>
3045
3046<h5>Overview:</h5>
3047
3048
3049<p>
3050The '<tt>llvm.readcyclecounter</tt>' intrinsic provides access to the cycle
3051counter register (or similar low latency, high accuracy clocks) on those targets
3052that support it. On X86, it should map to RDTSC. On Alpha, it should map to RPCC.
3053As the backing counters overflow quickly (on the order of 9 seconds on alpha), this
3054should only be used for small timings.
3055</p>
3056
3057<h5>Semantics:</h5>
3058
3059<p>
3060When directly supported, reading the cycle counter should not modify any memory.
3061Implementations are allowed to either return a application specific value or a
3062system wide value. On backends without support, this is lowered to a constant 0.
3063</p>
3064
3065</div>
3066
Chris Lattnerc8a2c222005-02-28 19:24:19 +00003067
John Criswellaa1c3c12004-04-09 16:43:20 +00003068<!-- ======================================================================= -->
3069<div class="doc_subsection">
3070 <a name="int_os">Operating System Intrinsics</a>
3071</div>
3072
3073<div class="doc_text">
3074<p>
3075These intrinsics are provided by LLVM to support the implementation of
3076operating system level code.
3077</p>
3078
3079</div>
John Criswella4501222004-04-12 15:02:16 +00003080
John Criswell508b93c2004-04-09 15:23:37 +00003081<!-- _______________________________________________________________________ -->
3082<div class="doc_subsubsection">
3083 <a name="i_readport">'<tt>llvm.readport</tt>' Intrinsic</a>
3084</div>
3085
3086<div class="doc_text">
3087
3088<h5>Syntax:</h5>
3089<pre>
Reid Spencer7821d062005-04-26 20:50:44 +00003090 declare &lt;integer type&gt; %llvm.readport (&lt;integer type&gt; &lt;address&gt;)
John Criswell508b93c2004-04-09 15:23:37 +00003091</pre>
3092
3093<h5>Overview:</h5>
3094
3095<p>
John Criswellaa1c3c12004-04-09 16:43:20 +00003096The '<tt>llvm.readport</tt>' intrinsic reads data from the specified hardware
3097I/O port.
John Criswell508b93c2004-04-09 15:23:37 +00003098</p>
3099
3100<h5>Arguments:</h5>
3101
3102<p>
John Criswellaa1c3c12004-04-09 16:43:20 +00003103The argument to this intrinsic indicates the hardware I/O address from which
3104to read the data. The address is in the hardware I/O address namespace (as
3105opposed to being a memory location for memory mapped I/O).
John Criswell508b93c2004-04-09 15:23:37 +00003106</p>
3107
3108<h5>Semantics:</h5>
3109
3110<p>
John Criswellaa1c3c12004-04-09 16:43:20 +00003111The '<tt>llvm.readport</tt>' intrinsic reads data from the hardware I/O port
3112specified by <i>address</i> and returns the value. The address and return
3113value must be integers, but the size is dependent upon the platform upon which
3114the program is code generated. For example, on x86, the address must be an
Misha Brukman36c6bc12005-04-22 18:02:52 +00003115unsigned 16-bit value, and the return value must be 8, 16, or 32 bits.
John Criswell508b93c2004-04-09 15:23:37 +00003116</p>
3117
3118</div>
3119
3120<!-- _______________________________________________________________________ -->
3121<div class="doc_subsubsection">
3122 <a name="i_writeport">'<tt>llvm.writeport</tt>' Intrinsic</a>
3123</div>
3124
3125<div class="doc_text">
3126
3127<h5>Syntax:</h5>
3128<pre>
Chris Lattner74d3f822004-12-09 17:30:23 +00003129 call void (&lt;integer type&gt;, &lt;integer type&gt;)*
3130 %llvm.writeport (&lt;integer type&gt; &lt;value&gt;,
3131 &lt;integer type&gt; &lt;address&gt;)
John Criswell508b93c2004-04-09 15:23:37 +00003132</pre>
3133
3134<h5>Overview:</h5>
3135
3136<p>
John Criswellaa1c3c12004-04-09 16:43:20 +00003137The '<tt>llvm.writeport</tt>' intrinsic writes data to the specified hardware
3138I/O port.
John Criswell508b93c2004-04-09 15:23:37 +00003139</p>
3140
3141<h5>Arguments:</h5>
3142
3143<p>
John Criswell7a576472004-04-12 16:33:19 +00003144The first argument is the value to write to the I/O port.
John Criswell508b93c2004-04-09 15:23:37 +00003145</p>
3146
3147<p>
John Criswell7a576472004-04-12 16:33:19 +00003148The second argument indicates the hardware I/O address to which data should be
3149written. The address is in the hardware I/O address namespace (as opposed to
3150being a memory location for memory mapped I/O).
John Criswell508b93c2004-04-09 15:23:37 +00003151</p>
3152
3153<h5>Semantics:</h5>
3154
3155<p>
3156The '<tt>llvm.writeport</tt>' intrinsic writes <i>value</i> to the I/O port
3157specified by <i>address</i>. The address and value must be integers, but the
3158size is dependent upon the platform upon which the program is code generated.
Misha Brukman36c6bc12005-04-22 18:02:52 +00003159For example, on x86, the address must be an unsigned 16-bit value, and the
John Criswellaa1c3c12004-04-09 16:43:20 +00003160value written must be 8, 16, or 32 bits in length.
John Criswell508b93c2004-04-09 15:23:37 +00003161</p>
3162
3163</div>
Chris Lattner3649c3a2004-02-14 04:08:35 +00003164
John Criswella4501222004-04-12 15:02:16 +00003165<!-- _______________________________________________________________________ -->
3166<div class="doc_subsubsection">
3167 <a name="i_readio">'<tt>llvm.readio</tt>' Intrinsic</a>
3168</div>
3169
3170<div class="doc_text">
3171
3172<h5>Syntax:</h5>
3173<pre>
Reid Spencer7821d062005-04-26 20:50:44 +00003174 declare &lt;result&gt; %llvm.readio (&lt;ty&gt; * &lt;pointer&gt;)
John Criswella4501222004-04-12 15:02:16 +00003175</pre>
3176
3177<h5>Overview:</h5>
3178
3179<p>
3180The '<tt>llvm.readio</tt>' intrinsic reads data from a memory mapped I/O
3181address.
3182</p>
3183
3184<h5>Arguments:</h5>
3185
3186<p>
John Criswell7a576472004-04-12 16:33:19 +00003187The argument to this intrinsic is a pointer indicating the memory address from
3188which to read the data. The data must be a
3189<a href="#t_firstclass">first class</a> type.
John Criswella4501222004-04-12 15:02:16 +00003190</p>
3191
3192<h5>Semantics:</h5>
3193
3194<p>
3195The '<tt>llvm.readio</tt>' intrinsic reads data from a memory mapped I/O
John Criswell7a576472004-04-12 16:33:19 +00003196location specified by <i>pointer</i> and returns the value. The argument must
3197be a pointer, and the return value must be a
3198<a href="#t_firstclass">first class</a> type. However, certain architectures
Misha Brukman36c6bc12005-04-22 18:02:52 +00003199may not support I/O on all first class types. For example, 32-bit processors
John Criswell7a576472004-04-12 16:33:19 +00003200may only support I/O on data types that are 32 bits or less.
John Criswella4501222004-04-12 15:02:16 +00003201</p>
3202
3203<p>
John Criswell7a576472004-04-12 16:33:19 +00003204This intrinsic enforces an in-order memory model for llvm.readio and
3205llvm.writeio calls on machines that use dynamic scheduling. Dynamically
3206scheduled processors may execute loads and stores out of order, re-ordering at
3207run time accesses to memory mapped I/O registers. Using these intrinsics
3208ensures that accesses to memory mapped I/O registers occur in program order.
John Criswella4501222004-04-12 15:02:16 +00003209</p>
3210
3211</div>
3212
3213<!-- _______________________________________________________________________ -->
3214<div class="doc_subsubsection">
3215 <a name="i_writeio">'<tt>llvm.writeio</tt>' Intrinsic</a>
3216</div>
3217
3218<div class="doc_text">
3219
3220<h5>Syntax:</h5>
3221<pre>
Reid Spencer7821d062005-04-26 20:50:44 +00003222 declare void %llvm.writeio (&lt;ty1&gt; &lt;value&gt;, &lt;ty2&gt; * &lt;pointer&gt;)
John Criswella4501222004-04-12 15:02:16 +00003223</pre>
3224
3225<h5>Overview:</h5>
3226
3227<p>
3228The '<tt>llvm.writeio</tt>' intrinsic writes data to the specified memory
3229mapped I/O address.
3230</p>
3231
3232<h5>Arguments:</h5>
3233
3234<p>
John Criswell7a576472004-04-12 16:33:19 +00003235The first argument is the value to write to the memory mapped I/O location.
3236The second argument is a pointer indicating the memory address to which the
3237data should be written.
John Criswella4501222004-04-12 15:02:16 +00003238</p>
3239
3240<h5>Semantics:</h5>
3241
3242<p>
3243The '<tt>llvm.writeio</tt>' intrinsic writes <i>value</i> to the memory mapped
John Criswell7a576472004-04-12 16:33:19 +00003244I/O address specified by <i>pointer</i>. The value must be a
3245<a href="#t_firstclass">first class</a> type. However, certain architectures
Misha Brukman36c6bc12005-04-22 18:02:52 +00003246may not support I/O on all first class types. For example, 32-bit processors
John Criswell7a576472004-04-12 16:33:19 +00003247may only support I/O on data types that are 32 bits or less.
John Criswella4501222004-04-12 15:02:16 +00003248</p>
3249
3250<p>
John Criswell7a576472004-04-12 16:33:19 +00003251This intrinsic enforces an in-order memory model for llvm.readio and
3252llvm.writeio calls on machines that use dynamic scheduling. Dynamically
3253scheduled processors may execute loads and stores out of order, re-ordering at
3254run time accesses to memory mapped I/O registers. Using these intrinsics
3255ensures that accesses to memory mapped I/O registers occur in program order.
John Criswella4501222004-04-12 15:02:16 +00003256</p>
3257
3258</div>
3259
Chris Lattner3649c3a2004-02-14 04:08:35 +00003260<!-- ======================================================================= -->
3261<div class="doc_subsection">
Chris Lattnerfee11462004-02-12 17:01:32 +00003262 <a name="int_libc">Standard C Library Intrinsics</a>
3263</div>
3264
3265<div class="doc_text">
3266<p>
Chris Lattner3649c3a2004-02-14 04:08:35 +00003267LLVM provides intrinsics for a few important standard C library functions.
3268These intrinsics allow source-language front-ends to pass information about the
3269alignment of the pointer arguments to the code generator, providing opportunity
3270for more efficient code generation.
Chris Lattnerfee11462004-02-12 17:01:32 +00003271</p>
3272
3273</div>
3274
3275<!-- _______________________________________________________________________ -->
3276<div class="doc_subsubsection">
3277 <a name="i_memcpy">'<tt>llvm.memcpy</tt>' Intrinsic</a>
3278</div>
3279
3280<div class="doc_text">
3281
3282<h5>Syntax:</h5>
3283<pre>
Reid Spencercf669d82005-04-26 20:41:16 +00003284 declare void %llvm.memcpy(sbyte* &lt;dest&gt;, sbyte* &lt;src&gt;,
3285 uint &lt;len&gt;, uint &lt;align&gt;)
Chris Lattnerfee11462004-02-12 17:01:32 +00003286</pre>
3287
3288<h5>Overview:</h5>
3289
3290<p>
3291The '<tt>llvm.memcpy</tt>' intrinsic copies a block of memory from the source
3292location to the destination location.
3293</p>
3294
3295<p>
3296Note that, unlike the standard libc function, the <tt>llvm.memcpy</tt> intrinsic
3297does not return a value, and takes an extra alignment argument.
3298</p>
3299
3300<h5>Arguments:</h5>
3301
3302<p>
3303The first argument is a pointer to the destination, the second is a pointer to
3304the source. The third argument is an (arbitrarily sized) integer argument
3305specifying the number of bytes to copy, and the fourth argument is the alignment
3306of the source and destination locations.
3307</p>
3308
Chris Lattner4c67c482004-02-12 21:18:15 +00003309<p>
3310If the call to this intrinisic has an alignment value that is not 0 or 1, then
3311the caller guarantees that the size of the copy is a multiple of the alignment
3312and that both the source and destination pointers are aligned to that boundary.
3313</p>
3314
Chris Lattnerfee11462004-02-12 17:01:32 +00003315<h5>Semantics:</h5>
3316
3317<p>
3318The '<tt>llvm.memcpy</tt>' intrinsic copies a block of memory from the source
3319location to the destination location, which are not allowed to overlap. It
3320copies "len" bytes of memory over. If the argument is known to be aligned to
3321some boundary, this can be specified as the fourth argument, otherwise it should
3322be set to 0 or 1.
3323</p>
3324</div>
3325
3326
Chris Lattnerf30152e2004-02-12 18:10:10 +00003327<!-- _______________________________________________________________________ -->
3328<div class="doc_subsubsection">
3329 <a name="i_memmove">'<tt>llvm.memmove</tt>' Intrinsic</a>
3330</div>
3331
3332<div class="doc_text">
3333
3334<h5>Syntax:</h5>
3335<pre>
Reid Spencercf669d82005-04-26 20:41:16 +00003336 declare void %llvm.memmove(sbyte* &lt;dest&gt;, sbyte* &lt;src&gt;,
3337 uint &lt;len&gt;, uint &lt;align&gt;)
Chris Lattnerf30152e2004-02-12 18:10:10 +00003338</pre>
3339
3340<h5>Overview:</h5>
3341
3342<p>
3343The '<tt>llvm.memmove</tt>' intrinsic moves a block of memory from the source
3344location to the destination location. It is similar to the '<tt>llvm.memcpy</tt>'
3345intrinsic but allows the two memory locations to overlap.
3346</p>
3347
3348<p>
3349Note that, unlike the standard libc function, the <tt>llvm.memmove</tt> intrinsic
3350does not return a value, and takes an extra alignment argument.
3351</p>
3352
3353<h5>Arguments:</h5>
3354
3355<p>
3356The first argument is a pointer to the destination, the second is a pointer to
3357the source. The third argument is an (arbitrarily sized) integer argument
3358specifying the number of bytes to copy, and the fourth argument is the alignment
3359of the source and destination locations.
3360</p>
3361
Chris Lattner4c67c482004-02-12 21:18:15 +00003362<p>
3363If the call to this intrinisic has an alignment value that is not 0 or 1, then
3364the caller guarantees that the size of the copy is a multiple of the alignment
3365and that both the source and destination pointers are aligned to that boundary.
3366</p>
3367
Chris Lattnerf30152e2004-02-12 18:10:10 +00003368<h5>Semantics:</h5>
3369
3370<p>
3371The '<tt>llvm.memmove</tt>' intrinsic copies a block of memory from the source
3372location to the destination location, which may overlap. It
3373copies "len" bytes of memory over. If the argument is known to be aligned to
3374some boundary, this can be specified as the fourth argument, otherwise it should
3375be set to 0 or 1.
3376</p>
3377</div>
3378
Chris Lattner941515c2004-01-06 05:31:32 +00003379
Chris Lattner3649c3a2004-02-14 04:08:35 +00003380<!-- _______________________________________________________________________ -->
3381<div class="doc_subsubsection">
3382 <a name="i_memset">'<tt>llvm.memset</tt>' Intrinsic</a>
3383</div>
3384
3385<div class="doc_text">
3386
3387<h5>Syntax:</h5>
3388<pre>
Reid Spencercf669d82005-04-26 20:41:16 +00003389 declare void %llvm.memset(sbyte* &lt;dest&gt;, ubyte &lt;val&gt;,
3390 uint &lt;len&gt;, uint &lt;align&gt;)
Chris Lattner3649c3a2004-02-14 04:08:35 +00003391</pre>
3392
3393<h5>Overview:</h5>
3394
3395<p>
3396The '<tt>llvm.memset</tt>' intrinsic fills a block of memory with a particular
3397byte value.
3398</p>
3399
3400<p>
3401Note that, unlike the standard libc function, the <tt>llvm.memset</tt> intrinsic
3402does not return a value, and takes an extra alignment argument.
3403</p>
3404
3405<h5>Arguments:</h5>
3406
3407<p>
3408The first argument is a pointer to the destination to fill, the second is the
3409byte value to fill it with, the third argument is an (arbitrarily sized) integer
3410argument specifying the number of bytes to fill, and the fourth argument is the
3411known alignment of destination location.
3412</p>
3413
3414<p>
3415If the call to this intrinisic has an alignment value that is not 0 or 1, then
3416the caller guarantees that the size of the copy is a multiple of the alignment
3417and that the destination pointer is aligned to that boundary.
3418</p>
3419
3420<h5>Semantics:</h5>
3421
3422<p>
3423The '<tt>llvm.memset</tt>' intrinsic fills "len" bytes of memory starting at the
3424destination location. If the argument is known to be aligned to some boundary,
3425this can be specified as the fourth argument, otherwise it should be set to 0 or
34261.
3427</p>
3428</div>
3429
3430
Chris Lattner3b4f4372004-06-11 02:28:03 +00003431<!-- _______________________________________________________________________ -->
3432<div class="doc_subsubsection">
Alkis Evlogimenos0fa39232004-06-13 01:16:15 +00003433 <a name="i_isunordered">'<tt>llvm.isunordered</tt>' Intrinsic</a>
3434</div>
3435
3436<div class="doc_text">
3437
3438<h5>Syntax:</h5>
3439<pre>
Reid Spencer7821d062005-04-26 20:50:44 +00003440 declare bool %llvm.isunordered(&lt;float or double&gt; Val1, &lt;float or double&gt; Val2)
Alkis Evlogimenos0fa39232004-06-13 01:16:15 +00003441</pre>
3442
3443<h5>Overview:</h5>
3444
3445<p>
3446The '<tt>llvm.isunordered</tt>' intrinsic returns true if either or both of the
3447specified floating point values is a NAN.
3448</p>
3449
3450<h5>Arguments:</h5>
3451
3452<p>
3453The arguments are floating point numbers of the same type.
3454</p>
3455
3456<h5>Semantics:</h5>
3457
3458<p>
3459If either or both of the arguments is a SNAN or QNAN, it returns true, otherwise
3460false.
3461</p>
3462</div>
3463
3464
Chris Lattner8a8f2e52005-07-21 01:29:16 +00003465<!-- _______________________________________________________________________ -->
3466<div class="doc_subsubsection">
3467 <a name="i_sqrt">'<tt>llvm.sqrt</tt>' Intrinsic</a>
3468</div>
3469
3470<div class="doc_text">
3471
3472<h5>Syntax:</h5>
3473<pre>
3474 declare &lt;float or double&gt; %llvm.sqrt(&lt;float or double&gt; Val)
3475</pre>
3476
3477<h5>Overview:</h5>
3478
3479<p>
3480The '<tt>llvm.sqrt</tt>' intrinsic returns the sqrt of the specified operand,
3481returning the same value as the libm '<tt>sqrt</tt>' function would. Unlike
3482<tt>sqrt</tt> in libm, however, <tt>llvm.sqrt</tt> has undefined behavior for
3483negative numbers (which allows for better optimization).
3484</p>
3485
3486<h5>Arguments:</h5>
3487
3488<p>
3489The argument and return value are floating point numbers of the same type.
3490</p>
3491
3492<h5>Semantics:</h5>
3493
3494<p>
3495This function returns the sqrt of the specified operand if it is a positive
3496floating point number.
3497</p>
3498</div>
3499
Andrew Lenharth1d463522005-05-03 18:01:48 +00003500<!-- ======================================================================= -->
3501<div class="doc_subsection">
3502 <a name="int_count">Bit Counting Intrinsics</a>
3503</div>
3504
3505<div class="doc_text">
3506<p>
3507LLVM provides intrinsics for a few important bit counting operations.
3508These allow efficient code generation for some algorithms.
3509</p>
3510
3511</div>
3512
3513<!-- _______________________________________________________________________ -->
3514<div class="doc_subsubsection">
3515 <a name="int_ctpop">'<tt>llvm.ctpop</tt>' Intrinsic</a>
3516</div>
3517
3518<div class="doc_text">
3519
3520<h5>Syntax:</h5>
3521<pre>
3522 declare int %llvm.ctpop(int &lt;src&gt;)
Andrew Lenharth1d463522005-05-03 18:01:48 +00003523</pre>
3524
3525<h5>Overview:</h5>
3526
3527<p>
3528The '<tt>llvm.ctpop</tt>' intrinsic counts the number of ones in a variable.
3529</p>
3530
3531<h5>Arguments:</h5>
3532
3533<p>
Chris Lattner573f64e2005-05-07 01:46:40 +00003534The only argument is the value to be counted. The argument may be of any
3535integer type. The return type must match the argument type.
Andrew Lenharth1d463522005-05-03 18:01:48 +00003536</p>
3537
3538<h5>Semantics:</h5>
3539
3540<p>
3541The '<tt>llvm.ctpop</tt>' intrinsic counts the 1's in a variable.
3542</p>
3543</div>
3544
3545<!-- _______________________________________________________________________ -->
3546<div class="doc_subsubsection">
Andrew Lenharth1d463522005-05-03 18:01:48 +00003547 <a name="int_ctlz">'<tt>llvm.ctlz</tt>' Intrinsic</a>
3548</div>
3549
3550<div class="doc_text">
3551
3552<h5>Syntax:</h5>
3553<pre>
3554 declare int %llvm.ctlz(int &lt;src&gt;)
3555
3556</pre>
3557
3558<h5>Overview:</h5>
3559
3560<p>
Chris Lattner573f64e2005-05-07 01:46:40 +00003561The '<tt>llvm.ctlz</tt>' intrinsic counts the number of leading zeros in a
3562variable.
Andrew Lenharth1d463522005-05-03 18:01:48 +00003563</p>
3564
3565<h5>Arguments:</h5>
3566
3567<p>
Chris Lattner573f64e2005-05-07 01:46:40 +00003568The only argument is the value to be counted. The argument may be of any
3569integer type. The return type must match the argument type.
Andrew Lenharth1d463522005-05-03 18:01:48 +00003570</p>
3571
3572<h5>Semantics:</h5>
3573
3574<p>
Chris Lattnerefa20fa2005-05-15 19:39:26 +00003575The '<tt>llvm.ctlz</tt>' intrinsic counts the leading (most significant) zeros
3576in a variable. If the src == 0 then the result is the size in bits of the type
3577of src. For example, <tt>llvm.cttz(int 2) = 30</tt>.
Andrew Lenharth1d463522005-05-03 18:01:48 +00003578</p>
3579</div>
Chris Lattner3b4f4372004-06-11 02:28:03 +00003580
3581
Chris Lattnerefa20fa2005-05-15 19:39:26 +00003582
3583<!-- _______________________________________________________________________ -->
3584<div class="doc_subsubsection">
3585 <a name="int_cttz">'<tt>llvm.cttz</tt>' Intrinsic</a>
3586</div>
3587
3588<div class="doc_text">
3589
3590<h5>Syntax:</h5>
3591<pre>
3592 declare int %llvm.cttz(int &lt;src&gt;)
Chris Lattnerefa20fa2005-05-15 19:39:26 +00003593</pre>
3594
3595<h5>Overview:</h5>
3596
3597<p>
3598The '<tt>llvm.cttz</tt>' intrinsic counts the number of trailing zeros.
3599</p>
3600
3601<h5>Arguments:</h5>
3602
3603<p>
3604The only argument is the value to be counted. The argument may be of any
3605integer type. The return type must match the argument type.
3606</p>
3607
3608<h5>Semantics:</h5>
3609
3610<p>
3611The '<tt>llvm.cttz</tt>' intrinsic counts the trailing (least significant) zeros
3612in a variable. If the src == 0 then the result is the size in bits of the type
3613of src. For example, <tt>llvm.cttz(2) = 1</tt>.
3614</p>
3615</div>
3616
Chris Lattner941515c2004-01-06 05:31:32 +00003617<!-- ======================================================================= -->
3618<div class="doc_subsection">
3619 <a name="int_debugger">Debugger Intrinsics</a>
3620</div>
3621
3622<div class="doc_text">
3623<p>
3624The LLVM debugger intrinsics (which all start with <tt>llvm.dbg.</tt> prefix),
3625are described in the <a
3626href="SourceLevelDebugging.html#format_common_intrinsics">LLVM Source Level
3627Debugging</a> document.
3628</p>
3629</div>
3630
3631
Chris Lattner2f7c9632001-06-06 20:29:01 +00003632<!-- *********************************************************************** -->
Chris Lattner2f7c9632001-06-06 20:29:01 +00003633<hr>
Misha Brukmanc501f552004-03-01 17:47:27 +00003634<address>
3635 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
3636 src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
3637 <a href="http://validator.w3.org/check/referer"><img
3638 src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!" /></a>
3639
3640 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
3641 <a href="http://llvm.cs.uiuc.edu">The LLVM Compiler Infrastructure</a><br>
3642 Last modified: $Date$
3643</address>
Misha Brukman76307852003-11-08 01:05:38 +00003644</body>
3645</html>