Gitiles
Code Review Sign In
gerrit-public.fairphone.software / fp2-dev / platform / external / llvm / 7f70559bc47877bafc6dfa92b7df6b64650445fb / . / docs / LangRef.html
blob: 2e40842dfb1a752a2d1912d615576a5695b24437 [file] [log] [blame]
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
2 "http://www.w3.org/TR/html4/strict.dtd">
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3<html>
4<head>
5 <title>LLVM Assembly Language Reference Manual</title>
Reid Spencer3921c742004-08-26 20:44:00 +0000[diff] [blame]6 <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
7 <meta name="author" content="Chris Lattner">
8 <meta name="description"
9 content="LLVM Assembly Language Reference Manual.">
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]10 <link rel="stylesheet" href="llvm.css" type="text/css">
11</head>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]12
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]13<body>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]14
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]15<div class="doc_title"> LLVM Language Reference Manual </div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]16<ol>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]17 <li><a href="#abstract">Abstract</a></li>
18 <li><a href="#introduction">Introduction</a></li>
19 <li><a href="#identifiers">Identifiers</a></li>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]20 <li><a href="#highlevel">High Level Structure</a>
21 <ol>
22 <li><a href="#modulestructure">Module Structure</a></li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]23 <li><a href="#linkage">Linkage Types</a></li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]24 <li><a href="#callingconv">Calling Conventions</a></li>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]25 <li><a href="#globalvars">Global Variables</a></li>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]26 <li><a href="#functionstructure">Functions</a></li>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]27 <li><a href="#paramattrs">Parameter Attributes</a></li>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]28 <li><a href="#moduleasm">Module-Level Inline Assembly</a></li>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]29 </ol>
30 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]31 <li><a href="#typesystem">Type System</a>
32 <ol>
Robert Bocchino7b81c752006-02-17 21:18:08 +0000[diff] [blame]33 <li><a href="#t_primitive">Primitive Types</a>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]34 <ol>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]35 <li><a href="#t_classifications">Type Classifications</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]36 </ol>
37 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]38 <li><a href="#t_derived">Derived Types</a>
39 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]40 <li><a href="#t_array">Array Type</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]41 <li><a href="#t_function">Function Type</a></li>
42 <li><a href="#t_pointer">Pointer Type</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]43 <li><a href="#t_struct">Structure Type</a></li>
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]44 <li><a href="#t_pstruct">Packed Structure Type</a></li>
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]45 <li><a href="#t_packed">Packed Type</a></li>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]46 <li><a href="#t_opaque">Opaque Type</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]47 </ol>
48 </li>
49 </ol>
50 </li>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]51 <li><a href="#constants">Constants</a>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]52 <ol>
53 <li><a href="#simpleconstants">Simple Constants</a>
54 <li><a href="#aggregateconstants">Aggregate Constants</a>
55 <li><a href="#globalconstants">Global Variable and Function Addresses</a>
56 <li><a href="#undefvalues">Undefined Values</a>
57 <li><a href="#constantexprs">Constant Expressions</a>
58 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]59 </li>
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]60 <li><a href="#othervalues">Other Values</a>
61 <ol>
62 <li><a href="#inlineasm">Inline Assembler Expressions</a>
63 </ol>
64 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]65 <li><a href="#instref">Instruction Reference</a>
66 <ol>
67 <li><a href="#terminators">Terminator Instructions</a>
68 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]69 <li><a href="#i_ret">'<tt>ret</tt>' Instruction</a></li>
70 <li><a href="#i_br">'<tt>br</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]71 <li><a href="#i_switch">'<tt>switch</tt>' Instruction</a></li>
72 <li><a href="#i_invoke">'<tt>invoke</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]73 <li><a href="#i_unwind">'<tt>unwind</tt>' Instruction</a></li>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]74 <li><a href="#i_unreachable">'<tt>unreachable</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]75 </ol>
76 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]77 <li><a href="#binaryops">Binary Operations</a>
78 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]79 <li><a href="#i_add">'<tt>add</tt>' Instruction</a></li>
80 <li><a href="#i_sub">'<tt>sub</tt>' Instruction</a></li>
81 <li><a href="#i_mul">'<tt>mul</tt>' Instruction</a></li>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]82 <li><a href="#i_udiv">'<tt>udiv</tt>' Instruction</a></li>
83 <li><a href="#i_sdiv">'<tt>sdiv</tt>' Instruction</a></li>
84 <li><a href="#i_fdiv">'<tt>fdiv</tt>' Instruction</a></li>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]85 <li><a href="#i_urem">'<tt>urem</tt>' Instruction</a></li>
86 <li><a href="#i_srem">'<tt>srem</tt>' Instruction</a></li>
87 <li><a href="#i_frem">'<tt>frem</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]88 </ol>
89 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]90 <li><a href="#bitwiseops">Bitwise Binary Operations</a>
91 <ol>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]92 <li><a href="#i_and">'<tt>and</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]93 <li><a href="#i_or">'<tt>or</tt>' Instruction</a></li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]94 <li><a href="#i_xor">'<tt>xor</tt>' Instruction</a></li>
95 <li><a href="#i_shl">'<tt>shl</tt>' Instruction</a></li>
Reid Spencer3822ff52006-11-08 06:47:33 +0000[diff] [blame]96 <li><a href="#i_lshr">'<tt>lshr</tt>' Instruction</a></li>
97 <li><a href="#i_ashr">'<tt>ashr</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]98 </ol>
99 </li>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]100 <li><a href="#vectorops">Vector Operations</a>
101 <ol>
102 <li><a href="#i_extractelement">'<tt>extractelement</tt>' Instruction</a></li>
103 <li><a href="#i_insertelement">'<tt>insertelement</tt>' Instruction</a></li>
104 <li><a href="#i_shufflevector">'<tt>shufflevector</tt>' Instruction</a></li>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]105 </ol>
106 </li>
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]107 <li><a href="#memoryops">Memory Access and Addressing Operations</a>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]108 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]109 <li><a href="#i_malloc">'<tt>malloc</tt>' Instruction</a></li>
110 <li><a href="#i_free">'<tt>free</tt>' Instruction</a></li>
111 <li><a href="#i_alloca">'<tt>alloca</tt>' Instruction</a></li>
Robert Bocchino7b81c752006-02-17 21:18:08 +0000[diff] [blame]112 <li><a href="#i_load">'<tt>load</tt>' Instruction</a></li>
113 <li><a href="#i_store">'<tt>store</tt>' Instruction</a></li>
114 <li><a href="#i_getelementptr">'<tt>getelementptr</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]115 </ol>
116 </li>
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]117 <li><a href="#convertops">Conversion Operations</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]118 <ol>
119 <li><a href="#i_trunc">'<tt>trunc .. to</tt>' Instruction</a></li>
120 <li><a href="#i_zext">'<tt>zext .. to</tt>' Instruction</a></li>
121 <li><a href="#i_sext">'<tt>sext .. to</tt>' Instruction</a></li>
122 <li><a href="#i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a></li>
123 <li><a href="#i_fpext">'<tt>fpext .. to</tt>' Instruction</a></li>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]124 <li><a href="#i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a></li>
125 <li><a href="#i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a></li>
126 <li><a href="#i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a></li>
127 <li><a href="#i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a></li>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]128 <li><a href="#i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a></li>
129 <li><a href="#i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a></li>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]130 <li><a href="#i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a></li>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]131 </ol>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]132 <li><a href="#otherops">Other Operations</a>
133 <ol>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]134 <li><a href="#i_icmp">'<tt>icmp</tt>' Instruction</a></li>
135 <li><a href="#i_fcmp">'<tt>fcmp</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]136 <li><a href="#i_phi">'<tt>phi</tt>' Instruction</a></li>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]137 <li><a href="#i_select">'<tt>select</tt>' Instruction</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]138 <li><a href="#i_call">'<tt>call</tt>' Instruction</a></li>
Chris Lattnerfb6977d2006-01-13 23:26:01 +0000[diff] [blame]139 <li><a href="#i_va_arg">'<tt>va_arg</tt>' Instruction</a></li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]140 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]141 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]142 </ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]143 </li>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]144 <li><a href="#intrinsics">Intrinsic Functions</a>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]145 <ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]146 <li><a href="#int_varargs">Variable Argument Handling Intrinsics</a>
147 <ol>
148 <li><a href="#i_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a></li>
149 <li><a href="#i_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a></li>
150 <li><a href="#i_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a></li>
151 </ol>
152 </li>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]153 <li><a href="#int_gc">Accurate Garbage Collection Intrinsics</a>
154 <ol>
155 <li><a href="#i_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a></li>
156 <li><a href="#i_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a></li>
157 <li><a href="#i_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a></li>
158 </ol>
159 </li>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]160 <li><a href="#int_codegen">Code Generator Intrinsics</a>
161 <ol>
162 <li><a href="#i_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a></li>
163 <li><a href="#i_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a></li>
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]164 <li><a href="#i_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a></li>
165 <li><a href="#i_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a></li>
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]166 <li><a href="#i_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a></li>
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]167 <li><a href="#i_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a></li>
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]168 <li><a href="#i_readcyclecounter"><tt>llvm.readcyclecounter</tt>' Intrinsic</a></li>
John Criswell7123e272004-04-09 16:43:20 +0000[diff] [blame]169 </ol>
170 </li>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]171 <li><a href="#int_libc">Standard C Library Intrinsics</a>
172 <ol>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]173 <li><a href="#i_memcpy">'<tt>llvm.memcpy.*</tt>' Intrinsic</a></li>
174 <li><a href="#i_memmove">'<tt>llvm.memmove.*</tt>' Intrinsic</a></li>
175 <li><a href="#i_memset">'<tt>llvm.memset.*</tt>' Intrinsic</a></li>
Chris Lattnerec6cb612006-01-16 22:38:59 +0000[diff] [blame]176 <li><a href="#i_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a></li>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]177 <li><a href="#i_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a></li>
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]178 </ol>
179 </li>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]180 <li><a href="#int_manip">Bit Manipulation Intrinsics</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]181 <ol>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]182 <li><a href="#i_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a></li>
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]183 <li><a href="#int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic </a></li>
184 <li><a href="#int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic </a></li>
185 <li><a href="#int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic </a></li>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]186 </ol>
187 </li>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]188 <li><a href="#int_debugger">Debugger intrinsics</a></li>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]189 </ol>
190 </li>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]191</ol>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]192
193<div class="doc_author">
194 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>
195 and <a href="mailto:vadve@cs.uiuc.edu">Vikram Adve</a></p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]196</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]197
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]198<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]199<div class="doc_section"> <a name="abstract">Abstract </a></div>
200<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]201
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]202<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]203<p>This document is a reference manual for the LLVM assembly language.
204LLVM is an SSA based representation that provides type safety,
205low-level operations, flexibility, and the capability of representing
206'all' high-level languages cleanly. It is the common code
207representation used throughout all phases of the LLVM compilation
208strategy.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]209</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]210
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]211<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]212<div class="doc_section"> <a name="introduction">Introduction</a> </div>
213<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]214
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]215<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]216
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]217<p>The LLVM code representation is designed to be used in three
218different forms: as an in-memory compiler IR, as an on-disk bytecode
219representation (suitable for fast loading by a Just-In-Time compiler),
220and as a human readable assembly language representation. This allows
221LLVM to provide a powerful intermediate representation for efficient
222compiler transformations and analysis, while providing a natural means
223to debug and visualize the transformations. The three different forms
224of LLVM are all equivalent. This document describes the human readable
225representation and notation.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]226
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]227<p>The LLVM representation aims to be light-weight and low-level
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]228while being expressive, typed, and extensible at the same time. It
229aims to be a "universal IR" of sorts, by being at a low enough level
230that high-level ideas may be cleanly mapped to it (similar to how
231microprocessors are "universal IR's", allowing many source languages to
232be mapped to them). By providing type information, LLVM can be used as
233the target of optimizations: for example, through pointer analysis, it
234can be proven that a C automatic variable is never accessed outside of
235the current function... allowing it to be promoted to a simple SSA
236value instead of a memory location.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]237
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]238</div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]239
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]240<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]241<div class="doc_subsubsection"> <a name="wellformed">Well-Formedness</a> </div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]242
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]243<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]244
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]245<p>It is important to note that this document describes 'well formed'
246LLVM assembly language. There is a difference between what the parser
247accepts and what is considered 'well formed'. For example, the
248following instruction is syntactically okay, but not well formed:</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]249
250<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]251 %x = <a href="#i_add">add</a> i32 1, %x
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]252</pre>
253
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]254<p>...because the definition of <tt>%x</tt> does not dominate all of
255its uses. The LLVM infrastructure provides a verification pass that may
256be used to verify that an LLVM module is well formed. This pass is
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]257automatically run by the parser after parsing input assembly and by
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]258the optimizer before it outputs bytecode. The violations pointed out
259by the verifier pass indicate bugs in transformation passes or input to
260the parser.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]261
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]262<!-- Describe the typesetting conventions here. --> </div>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]263
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]264<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]265<div class="doc_section"> <a name="identifiers">Identifiers</a> </div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]266<!-- *********************************************************************** -->
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]267
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]268<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]269
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]270<p>LLVM uses three different forms of identifiers, for different
271purposes:</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]272
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]273<ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]274 <li>Named values are represented as a string of characters with a '%' prefix.
275 For example, %foo, %DivisionByZero, %a.really.long.identifier. The actual
276 regular expression used is '<tt>%[a-zA-Z$._][a-zA-Z$._0-9]*</tt>'.
277 Identifiers which require other characters in their names can be surrounded
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame^]278 with quotes. In this way, anything except a <tt>&quot;</tt> character can be used
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]279 in a name.</li>
280
281 <li>Unnamed values are represented as an unsigned numeric value with a '%'
282 prefix. For example, %12, %2, %44.</li>
283
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]284 <li>Constants, which are described in a <a href="#constants">section about
285 constants</a>, below.</li>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]286</ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]287
288<p>LLVM requires that values start with a '%' sign for two reasons: Compilers
289don't need to worry about name clashes with reserved words, and the set of
290reserved words may be expanded in the future without penalty. Additionally,
291unnamed identifiers allow a compiler to quickly come up with a temporary
292variable without having to avoid symbol table conflicts.</p>
293
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]294<p>Reserved words in LLVM are very similar to reserved words in other
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]295languages. There are keywords for different opcodes
296('<tt><a href="#i_add">add</a></tt>',
297 '<tt><a href="#i_bitcast">bitcast</a></tt>',
298 '<tt><a href="#i_ret">ret</a></tt>', etc...), for primitive type names ('<tt><a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]299href="#t_void">void</a></tt>', '<tt><a href="#t_primitive">i32</a></tt>', etc...),
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]300and others. These reserved words cannot conflict with variable names, because
301none of them start with a '%' character.</p>
302
303<p>Here is an example of LLVM code to multiply the integer variable
304'<tt>%X</tt>' by 8:</p>
305
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]306<p>The easy way:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]307
308<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]309 %result = <a href="#i_mul">mul</a> i32 %X, 8
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]310</pre>
311
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]312<p>After strength reduction:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]313
314<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]315 %result = <a href="#i_shl">shl</a> i32 %X, i8 3
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]316</pre>
317
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]318<p>And the hard way:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]319
320<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]321 <a href="#i_add">add</a> i32 %X, %X <i>; yields {i32}:%0</i>
322 <a href="#i_add">add</a> i32 %0, %0 <i>; yields {i32}:%1</i>
323 %result = <a href="#i_add">add</a> i32 %1, %1
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]324</pre>
325
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]326<p>This last way of multiplying <tt>%X</tt> by 8 illustrates several
327important lexical features of LLVM:</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]328
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]329<ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]330
331 <li>Comments are delimited with a '<tt>;</tt>' and go until the end of
332 line.</li>
333
334 <li>Unnamed temporaries are created when the result of a computation is not
335 assigned to a named value.</li>
336
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]337 <li>Unnamed temporaries are numbered sequentially</li>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]338
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]339</ol>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]340
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]341<p>...and it also shows a convention that we follow in this document. When
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]342demonstrating instructions, we will follow an instruction with a comment that
343defines the type and name of value produced. Comments are shown in italic
344text.</p>
345
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]346</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]347
348<!-- *********************************************************************** -->
349<div class="doc_section"> <a name="highlevel">High Level Structure</a> </div>
350<!-- *********************************************************************** -->
351
352<!-- ======================================================================= -->
353<div class="doc_subsection"> <a name="modulestructure">Module Structure</a>
354</div>
355
356<div class="doc_text">
357
358<p>LLVM programs are composed of "Module"s, each of which is a
359translation unit of the input programs. Each module consists of
360functions, global variables, and symbol table entries. Modules may be
361combined together with the LLVM linker, which merges function (and
362global variable) definitions, resolves forward declarations, and merges
363symbol table entries. Here is an example of the "hello world" module:</p>
364
365<pre><i>; Declare the string constant as a global constant...</i>
366<a href="#identifiers">%.LC0</a> = <a href="#linkage_internal">internal</a> <a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]367 href="#globalvars">constant</a> <a href="#t_array">[13 x i8 ]</a> c"hello world\0A\00" <i>; [13 x i8 ]*</i>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]368
369<i>; External declaration of the puts function</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]370<a href="#functionstructure">declare</a> i32 %puts(i8 *) <i>; i32(i8 *)* </i>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]371
Chris Lattner81c01f02006-06-13 03:05:47 +0000[diff] [blame]372<i>; Global variable / Function body section separator</i>
373implementation
374
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]375<i>; Definition of main function</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]376define i32 %main() { <i>; i32()* </i>
377 <i>; Convert [13x i8 ]* to i8 *...</i>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]378 %cast210 = <a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]379 href="#i_getelementptr">getelementptr</a> [13 x i8 ]* %.LC0, i64 0, i64 0 <i>; i8 *</i>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]380
381 <i>; Call puts function to write out the string to stdout...</i>
382 <a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]383 href="#i_call">call</a> i32 %puts(i8 * %cast210) <i>; i32</i>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]384 <a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]385 href="#i_ret">ret</a> i32 0<br>}<br></pre>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]386
387<p>This example is made up of a <a href="#globalvars">global variable</a>
388named "<tt>.LC0</tt>", an external declaration of the "<tt>puts</tt>"
389function, and a <a href="#functionstructure">function definition</a>
390for "<tt>main</tt>".</p>
391
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]392<p>In general, a module is made up of a list of global values,
393where both functions and global variables are global values. Global values are
394represented by a pointer to a memory location (in this case, a pointer to an
395array of char, and a pointer to a function), and have one of the following <a
396href="#linkage">linkage types</a>.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]397
Chris Lattner81c01f02006-06-13 03:05:47 +0000[diff] [blame]398<p>Due to a limitation in the current LLVM assembly parser (it is limited by
399one-token lookahead), modules are split into two pieces by the "implementation"
400keyword. Global variable prototypes and definitions must occur before the
401keyword, and function definitions must occur after it. Function prototypes may
402occur either before or after it. In the future, the implementation keyword may
403become a noop, if the parser gets smarter.</p>
404
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]405</div>
406
407<!-- ======================================================================= -->
408<div class="doc_subsection">
409 <a name="linkage">Linkage Types</a>
410</div>
411
412<div class="doc_text">
413
414<p>
415All Global Variables and Functions have one of the following types of linkage:
416</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]417
418<dl>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]419
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]420 <dt><tt><b><a name="linkage_internal">internal</a></b></tt> </dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]421
422 <dd>Global values with internal linkage are only directly accessible by
423 objects in the current module. In particular, linking code into a module with
424 an internal global value may cause the internal to be renamed as necessary to
425 avoid collisions. Because the symbol is internal to the module, all
426 references can be updated. This corresponds to the notion of the
427 '<tt>static</tt>' keyword in C, or the idea of "anonymous namespaces" in C++.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]428 </dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]429
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]430 <dt><tt><b><a name="linkage_linkonce">linkonce</a></b></tt>: </dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]431
432 <dd>"<tt>linkonce</tt>" linkage is similar to <tt>internal</tt> linkage, with
433 the twist that linking together two modules defining the same
434 <tt>linkonce</tt> globals will cause one of the globals to be discarded. This
435 is typically used to implement inline functions. Unreferenced
436 <tt>linkonce</tt> globals are allowed to be discarded.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]437 </dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]438
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]439 <dt><tt><b><a name="linkage_weak">weak</a></b></tt>: </dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]440
441 <dd>"<tt>weak</tt>" linkage is exactly the same as <tt>linkonce</tt> linkage,
442 except that unreferenced <tt>weak</tt> globals may not be discarded. This is
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]443 used to implement constructs in C such as "<tt>i32 X;</tt>" at global scope.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]444 </dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]445
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]446 <dt><tt><b><a name="linkage_appending">appending</a></b></tt>: </dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]447
448 <dd>"<tt>appending</tt>" linkage may only be applied to global variables of
449 pointer to array type. When two global variables with appending linkage are
450 linked together, the two global arrays are appended together. This is the
451 LLVM, typesafe, equivalent of having the system linker append together
452 "sections" with identical names when .o files are linked.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]453 </dd>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]454
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame^]455 <dt><tt><b><a name="linkage_externweak">extern_weak</a></b></tt>: </dt>
456 <dd>The semantics of this linkage follow the ELF model: the symbol is weak
457 until linked, if not linked, the symbol becomes null instead of being an
458 undefined reference.
459 </dd>
460</dl>
461
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]462 <dt><tt><b><a name="linkage_external">externally visible</a></b></tt>:</dt>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]463
464 <dd>If none of the above identifiers are used, the global is externally
465 visible, meaning that it participates in linkage and can be used to resolve
466 external symbol references.
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]467 </dd>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]468
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]469 <p>
470 The next two types of linkage are targeted for Microsoft Windows platform
471 only. They are designed to support importing (exporting) symbols from (to)
472 DLLs.
473 </p>
474
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame^]475 <dl>
Anton Korobeynikovb74ed072006-09-14 18:23:27 +0000[diff] [blame]476 <dt><tt><b><a name="linkage_dllimport">dllimport</a></b></tt>: </dt>
477
478 <dd>"<tt>dllimport</tt>" linkage causes the compiler to reference a function
479 or variable via a global pointer to a pointer that is set up by the DLL
480 exporting the symbol. On Microsoft Windows targets, the pointer name is
481 formed by combining <code>_imp__</code> and the function or variable name.
482 </dd>
483
484 <dt><tt><b><a name="linkage_dllexport">dllexport</a></b></tt>: </dt>
485
486 <dd>"<tt>dllexport</tt>" linkage causes the compiler to provide a global
487 pointer to a pointer in a DLL, so that it can be referenced with the
488 <tt>dllimport</tt> attribute. On Microsoft Windows targets, the pointer
489 name is formed by combining <code>_imp__</code> and the function or variable
490 name.
491 </dd>
492
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]493</dl>
494
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame^]495<p><a name="linkage_external"></a>For example, since the "<tt>.LC0</tt>"
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]496variable is defined to be internal, if another module defined a "<tt>.LC0</tt>"
497variable and was linked with this one, one of the two would be renamed,
498preventing a collision. Since "<tt>main</tt>" and "<tt>puts</tt>" are
499external (i.e., lacking any linkage declarations), they are accessible
Reid Spencerac8d2762007-01-05 00:59:10 +0000[diff] [blame]500outside of the current module.</p>
501<p>It is illegal for a function <i>declaration</i>
502to have any linkage type other than "externally visible", <tt>dllimport</tt>,
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame^]503or <tt>extern_weak</tt>.</p>
Chris Lattnere5d947b2004-12-09 16:36:40 +0000[diff] [blame]504
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]505</div>
506
507<!-- ======================================================================= -->
508<div class="doc_subsection">
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]509 <a name="callingconv">Calling Conventions</a>
510</div>
511
512<div class="doc_text">
513
514<p>LLVM <a href="#functionstructure">functions</a>, <a href="#i_call">calls</a>
515and <a href="#i_invoke">invokes</a> can all have an optional calling convention
516specified for the call. The calling convention of any pair of dynamic
517caller/callee must match, or the behavior of the program is undefined. The
518following calling conventions are supported by LLVM, and more may be added in
519the future:</p>
520
521<dl>
522 <dt><b>"<tt>ccc</tt>" - The C calling convention</b>:</dt>
523
524 <dd>This calling convention (the default if no other calling convention is
525 specified) matches the target C calling conventions. This calling convention
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]526 supports varargs function calls and tolerates some mismatch in the declared
Reid Spencerc28d2bc2006-12-31 21:30:18 +0000[diff] [blame]527 prototype and implemented declaration of the function (as does normal C).
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]528 </dd>
529
Chris Lattner5710ce92006-05-19 21:15:36 +0000[diff] [blame]530 <dt><b>"<tt>csretcc</tt>" - The C struct return calling convention</b>:</dt>
531
532 <dd>This calling convention matches the target C calling conventions, except
533 that functions with this convention are required to take a pointer as their
534 first argument, and the return type of the function must be void. This is
535 used for C functions that return aggregates by-value. In this case, the
536 function has been transformed to take a pointer to the struct as the first
537 argument to the function. For targets where the ABI specifies specific
538 behavior for structure-return calls, the calling convention can be used to
539 distinguish between struct return functions and other functions that take a
540 pointer to a struct as the first argument.
541 </dd>
542
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]543 <dt><b>"<tt>fastcc</tt>" - The fast calling convention</b>:</dt>
544
545 <dd>This calling convention attempts to make calls as fast as possible
546 (e.g. by passing things in registers). This calling convention allows the
547 target to use whatever tricks it wants to produce fast code for the target,
Chris Lattner8cdc5bc2005-05-06 23:08:23 +0000[diff] [blame]548 without having to conform to an externally specified ABI. Implementations of
549 this convention should allow arbitrary tail call optimization to be supported.
550 This calling convention does not support varargs and requires the prototype of
551 all callees to exactly match the prototype of the function definition.
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]552 </dd>
553
554 <dt><b>"<tt>coldcc</tt>" - The cold calling convention</b>:</dt>
555
556 <dd>This calling convention attempts to make code in the caller as efficient
557 as possible under the assumption that the call is not commonly executed. As
558 such, these calls often preserve all registers so that the call does not break
559 any live ranges in the caller side. This calling convention does not support
560 varargs and requires the prototype of all callees to exactly match the
561 prototype of the function definition.
562 </dd>
563
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]564 <dt><b>"<tt>cc &lt;<em>n</em>&gt;</tt>" - Numbered convention</b>:</dt>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]565
566 <dd>Any calling convention may be specified by number, allowing
567 target-specific calling conventions to be used. Target specific calling
568 conventions start at 64.
569 </dd>
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]570</dl>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]571
572<p>More calling conventions can be added/defined on an as-needed basis, to
573support pascal conventions or any other well-known target-independent
574convention.</p>
575
576</div>
577
578<!-- ======================================================================= -->
579<div class="doc_subsection">
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]580 <a name="globalvars">Global Variables</a>
581</div>
582
583<div class="doc_text">
584
Chris Lattner3689a342005-02-12 19:30:21 +0000[diff] [blame]585<p>Global variables define regions of memory allocated at compilation time
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]586instead of run-time. Global variables may optionally be initialized, may have
587an explicit section to be placed in, and may
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]588have an optional explicit alignment specified. A
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]589variable may be defined as a global "constant," which indicates that the
Chris Lattner3689a342005-02-12 19:30:21 +0000[diff] [blame]590contents of the variable will <b>never</b> be modified (enabling better
591optimization, allowing the global data to be placed in the read-only section of
592an executable, etc). Note that variables that need runtime initialization
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]593cannot be marked "constant" as there is a store to the variable.</p>
Chris Lattner3689a342005-02-12 19:30:21 +0000[diff] [blame]594
595<p>
596LLVM explicitly allows <em>declarations</em> of global variables to be marked
597constant, even if the final definition of the global is not. This capability
598can be used to enable slightly better optimization of the program, but requires
599the language definition to guarantee that optimizations based on the
600'constantness' are valid for the translation units that do not include the
601definition.
602</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]603
604<p>As SSA values, global variables define pointer values that are in
605scope (i.e. they dominate) all basic blocks in the program. Global
606variables always define a pointer to their "content" type because they
607describe a region of memory, and all memory objects in LLVM are
608accessed through pointers.</p>
609
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]610<p>LLVM allows an explicit section to be specified for globals. If the target
611supports it, it will emit globals to the section specified.</p>
612
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]613<p>An explicit alignment may be specified for a global. If not present, or if
614the alignment is set to zero, the alignment of the global is set by the target
615to whatever it feels convenient. If an explicit alignment is specified, the
616global is forced to have at least that much alignment. All alignments must be
617a power of 2.</p>
618
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]619</div>
620
621
622<!-- ======================================================================= -->
623<div class="doc_subsection">
624 <a name="functionstructure">Functions</a>
625</div>
626
627<div class="doc_text">
628
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]629<p>LLVM function definitions consist of the "<tt>define</tt>" keyord,
630an optional <a href="#linkage">linkage type</a>, an optional
631<a href="#callingconv">calling convention</a>, a return type, an optional
632<a href="#paramattrs">parameter attribute</a> for the return type, a function
633name, a (possibly empty) argument list (each with optional
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]634<a href="#paramattrs">parameter attributes</a>), an optional section, an
635optional alignment, an opening curly brace, a list of basic blocks, and a
636closing curly brace. LLVM function declarations
637consist of the "<tt>declare</tt>" keyword, an optional <a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]638 href="#callingconv">calling convention</a>, a return type, an optional
639<a href="#paramattrs">parameter attribute</a> for the return type, a function
640name, a possibly empty list of arguments, and an optional alignment.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]641
642<p>A function definition contains a list of basic blocks, forming the CFG for
643the function. Each basic block may optionally start with a label (giving the
644basic block a symbol table entry), contains a list of instructions, and ends
645with a <a href="#terminators">terminator</a> instruction (such as a branch or
646function return).</p>
647
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]648<p>The first basic block in a program is special in two ways: it is immediately
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]649executed on entrance to the function, and it is not allowed to have predecessor
650basic blocks (i.e. there can not be any branches to the entry block of a
651function). Because the block can have no predecessors, it also cannot have any
652<a href="#i_phi">PHI nodes</a>.</p>
653
654<p>LLVM functions are identified by their name and type signature. Hence, two
655functions with the same name but different parameter lists or return values are
Chris Lattnerd4f6b172005-03-07 22:13:59 +0000[diff] [blame]656considered different functions, and LLVM will resolve references to each
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]657appropriately.</p>
658
Chris Lattner88f6c462005-11-12 00:45:07 +0000[diff] [blame]659<p>LLVM allows an explicit section to be specified for functions. If the target
660supports it, it will emit functions to the section specified.</p>
661
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]662<p>An explicit alignment may be specified for a function. If not present, or if
663the alignment is set to zero, the alignment of the function is set by the target
664to whatever it feels convenient. If an explicit alignment is specified, the
665function is forced to have at least that much alignment. All alignments must be
666a power of 2.</p>
667
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]668</div>
669
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]670<!-- ======================================================================= -->
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]671<div class="doc_subsection"><a name="paramattrs">Parameter Attributes</a></div>
672<div class="doc_text">
673 <p>The return type and each parameter of a function type may have a set of
674 <i>parameter attributes</i> associated with them. Parameter attributes are
675 used to communicate additional information about the result or parameters of
676 a function. Parameter attributes are considered to be part of the function
677 type so two functions types that differ only by the parameter attributes
678 are different function types.</p>
679
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]680 <p>Parameter attributes consist of an at sign (@) followed by either a single
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]681 keyword or a comma separate list of keywords enclosed in parentheses. For
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame^]682 example:</p><pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]683 %someFunc = i16 @zext (i8 @(sext) %someParam)
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]684 %someFunc = i16 @zext (i8 @zext %someParam)</pre>
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame^]685 <p>Note that the two function types above are unique because the parameter has
686 a different attribute (@sext in the first one, @zext in the second).</p>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]687
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame^]688 <p>Currently, only the following parameter attributes are defined:</p>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]689 <dl>
690 <dt><tt>@zext</tt></dt>
691 <dd>This indicates that the parameter should be zero extended just before
692 a call to this function.</dd>
693 <dt><tt>@sext</tt></dt>
694 <dd>This indicates that the parameter should be sign extended just before
695 a call to this function.</dd>
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame^]696 </dl>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]697
698 <p>The current motivation for parameter attributes is to enable the sign and
699 zero extend information necessary for the C calling convention to be passed
700 from the front end to LLVM. The <tt>@zext</tt> and <tt>@sext</tt> attributes
701 are used by the code generator to perform the required extension. However,
702 parameter attributes are an orthogonal feature to calling conventions and
703 may be used for other purposes in the future.</p>
704</div>
705
706<!-- ======================================================================= -->
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]707<div class="doc_subsection">
Chris Lattner1eeeb0c2006-04-08 04:40:53 +0000[diff] [blame]708 <a name="moduleasm">Module-Level Inline Assembly</a>
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]709</div>
710
711<div class="doc_text">
712<p>
713Modules may contain "module-level inline asm" blocks, which corresponds to the
714GCC "file scope inline asm" blocks. These blocks are internally concatenated by
715LLVM and treated as a single unit, but may be separated in the .ll file if
716desired. The syntax is very simple:
717</p>
718
719<div class="doc_code"><pre>
Chris Lattner52599e12006-01-24 00:37:20 +0000[diff] [blame]720 module asm "inline asm code goes here"
721 module asm "more can go here"
Chris Lattner4e9aba72006-01-23 23:23:47 +0000[diff] [blame]722</pre></div>
723
724<p>The strings can contain any character by escaping non-printable characters.
725 The escape sequence used is simply "\xx" where "xx" is the two digit hex code
726 for the number.
727</p>
728
729<p>
730 The inline asm code is simply printed to the machine code .s file when
731 assembly code is generated.
732</p>
733</div>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]734
735
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]736<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]737<div class="doc_section"> <a name="typesystem">Type System</a> </div>
738<!-- *********************************************************************** -->
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]739
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]740<div class="doc_text">
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]741
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]742<p>The LLVM type system is one of the most important features of the
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]743intermediate representation. Being typed enables a number of
744optimizations to be performed on the IR directly, without having to do
745extra analyses on the side before the transformation. A strong type
746system makes it easier to read the generated code and enables novel
747analyses and transformations that are not feasible to perform on normal
748three address code representations.</p>
Chris Lattnerfa730212004-12-09 16:11:40 +0000[diff] [blame]749
750</div>
751
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]752<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]753<div class="doc_subsection"> <a name="t_primitive">Primitive Types</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]754<div class="doc_text">
John Criswell4457dc92004-04-09 16:48:45 +0000[diff] [blame]755<p>The primitive types are the fundamental building blocks of the LLVM
Chris Lattnerd4f6b172005-03-07 22:13:59 +0000[diff] [blame]756system. The current set of primitive types is as follows:</p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]757
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]758<table class="layout">
759 <tr class="layout">
760 <td class="left">
761 <table>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]762 <tbody>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]763 <tr><th>Type</th><th>Description</th></tr>
764 <tr><td><tt>void</tt></td><td>No value</td></tr>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]765 <tr><td><tt>i8</tt></td><td>Signless 8-bit value</td></tr>
766 <tr><td><tt>i32</tt></td><td>Signless 32-bit value</td></tr>
Misha Brukmancfa87bc2005-04-22 18:02:52 +0000[diff] [blame]767 <tr><td><tt>float</tt></td><td>32-bit floating point value</td></tr>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]768 <tr><td><tt>label</tt></td><td>Branch destination</td></tr>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]769 </tbody>
770 </table>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]771 </td>
772 <td class="right">
773 <table>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]774 <tbody>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]775 <tr><th>Type</th><th>Description</th></tr>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]776 <tr><td><tt>i1</tt></td><td>True or False value</td></tr>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]777 <tr><td><tt>i16</tt></td><td>Signless 16-bit value</td></tr>
778 <tr><td><tt>i64</tt></td><td>Signless 64-bit value</td></tr>
779 <tr><td><tt>double</tt></td><td>64-bit floating point value</td></tr>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]780 </tbody>
781 </table>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]782 </td>
783 </tr>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]784</table>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]785</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]786
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]787<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]788<div class="doc_subsubsection"> <a name="t_classifications">Type
789Classifications</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]790<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]791<p>These different primitive types fall into a few useful
792classifications:</p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]793
794<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]795 <tbody>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]796 <tr><th>Classification</th><th>Types</th></tr>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]797 <tr>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]798 <td><a name="t_integer">integer</a></td>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]799 <td><tt>i8, i16, i32, i64</tt></td>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]800 </tr>
801 <tr>
802 <td><a name="t_integral">integral</a></td>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]803 <td><tt>i1, i8, i16, i32, i64</tt>
Misha Brukmanc24b7582004-08-12 20:16:08 +0000[diff] [blame]804 </td>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]805 </tr>
806 <tr>
807 <td><a name="t_floating">floating point</a></td>
808 <td><tt>float, double</tt></td>
809 </tr>
810 <tr>
811 <td><a name="t_firstclass">first class</a></td>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]812 <td><tt>i1, i8, i16, i32, i64, float, double, <br/>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]813 <a href="#t_pointer">pointer</a>,<a href="#t_packed">packed</a></tt>
814 </td>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]815 </tr>
816 </tbody>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]817</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]818
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]819<p>The <a href="#t_firstclass">first class</a> types are perhaps the
820most important. Values of these types are the only ones which can be
821produced by instructions, passed as arguments, or used as operands to
822instructions. This means that all structures and arrays must be
823manipulated either by pointer or by component.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]824</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]825
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]826<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]827<div class="doc_subsection"> <a name="t_derived">Derived Types</a> </div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]828
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]829<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]830
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]831<p>The real power in LLVM comes from the derived types in the system.
832This is what allows a programmer to represent arrays, functions,
833pointers, and other useful types. Note that these derived types may be
834recursive: For example, it is possible to have a two dimensional array.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]835
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]836</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]837
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]838<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]839<div class="doc_subsubsection"> <a name="t_array">Array Type</a> </div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]840
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]841<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]842
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]843<h5>Overview:</h5>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]844
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]845<p>The array type is a very simple derived type that arranges elements
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]846sequentially in memory. The array type requires a size (number of
847elements) and an underlying data type.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]848
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]849<h5>Syntax:</h5>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]850
851<pre>
852 [&lt;# elements&gt; x &lt;elementtype&gt;]
853</pre>
854
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]855<p>The number of elements is a constant integer value; elementtype may
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]856be any type with a size.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]857
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]858<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]859<table class="layout">
860 <tr class="layout">
861 <td class="left">
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]862 <tt>[40 x i32 ]</tt><br/>
863 <tt>[41 x i32 ]</tt><br/>
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]864 <tt>[40 x i8]</tt><br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]865 </td>
866 <td class="left">
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]867 Array of 40 32-bit integer values.<br/>
868 Array of 41 32-bit integer values.<br/>
869 Array of 40 8-bit integer values.<br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]870 </td>
871 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]872</table>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]873<p>Here are some examples of multidimensional arrays:</p>
874<table class="layout">
875 <tr class="layout">
876 <td class="left">
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]877 <tt>[3 x [4 x i32]]</tt><br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]878 <tt>[12 x [10 x float]]</tt><br/>
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]879 <tt>[2 x [3 x [4 x i16]]]</tt><br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]880 </td>
881 <td class="left">
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]882 3x4 array of 32-bit integer values.<br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]883 12x10 array of single precision floating point values.<br/>
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]884 2x3x4 array of 16-bit integer values.<br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]885 </td>
886 </tr>
887</table>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]888
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]889<p>Note that 'variable sized arrays' can be implemented in LLVM with a zero
890length array. Normally, accesses past the end of an array are undefined in
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]891LLVM (e.g. it is illegal to access the 5th element of a 3 element array).
892As a special case, however, zero length arrays are recognized to be variable
893length. This allows implementation of 'pascal style arrays' with the LLVM
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]894type "{ i32, [0 x float]}", for example.</p>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]895
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]896</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]897
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]898<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]899<div class="doc_subsubsection"> <a name="t_function">Function Type</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]900<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]901<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]902<p>The function type can be thought of as a function signature. It
903consists of a return type and a list of formal parameter types.
John Criswell009900b2003-11-25 21:45:46 +0000[diff] [blame]904Function types are usually used to build virtual function tables
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]905(which are structures of pointers to functions), for indirect function
906calls, and when defining a function.</p>
John Criswell009900b2003-11-25 21:45:46 +0000[diff] [blame]907<p>
908The return type of a function type cannot be an aggregate type.
909</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]910<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]911<pre> &lt;returntype&gt; (&lt;parameter list&gt;)<br></pre>
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]912<p>...where '<tt>&lt;parameter list&gt;</tt>' is a comma-separated list of type
Misha Brukmanc24b7582004-08-12 20:16:08 +0000[diff] [blame]913specifiers. Optionally, the parameter list may include a type <tt>...</tt>,
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]914which indicates that the function takes a variable number of arguments.
915Variable argument functions can access their arguments with the <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]916 href="#int_varargs">variable argument handling intrinsic</a> functions.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]917<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]918<table class="layout">
919 <tr class="layout">
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]920 <td class="left"><tt>i32 (i32)</tt></td>
921 <td class="left">function taking an <tt>i32</tt>, returning an <tt>i32</tt>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]922 </td>
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]923 </tr><tr class="layout">
Reid Spencerf17a0b72006-12-31 07:20:23 +0000[diff] [blame]924 <td class="left"><tt>float&nbsp;(i16&nbsp;@sext,&nbsp;i32&nbsp;*)&nbsp;*
925 </tt></td>
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]926 <td class="left"><a href="#t_pointer">Pointer</a> to a function that takes
927 an <tt>i16</tt> that should be sign extended and a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]928 <a href="#t_pointer">pointer</a> to <tt>i32</tt>, returning
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]929 <tt>float</tt>.
930 </td>
931 </tr><tr class="layout">
932 <td class="left"><tt>i32 (i8*, ...)</tt></td>
933 <td class="left">A vararg function that takes at least one
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]934 <a href="#t_pointer">pointer</a> to <tt>i8 </tt> (char in C),
Reid Spencer92f82302006-12-31 07:18:34 +0000[diff] [blame]935 which returns an integer. This is the signature for <tt>printf</tt> in
936 LLVM.
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]937 </td>
938 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]939</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]940
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]941</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]942<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]943<div class="doc_subsubsection"> <a name="t_struct">Structure Type</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]944<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]945<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]946<p>The structure type is used to represent a collection of data members
947together in memory. The packing of the field types is defined to match
948the ABI of the underlying processor. The elements of a structure may
949be any type that has a size.</p>
950<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
951and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
952field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
953instruction.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]954<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]955<pre> { &lt;type list&gt; }<br></pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]956<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]957<table class="layout">
958 <tr class="layout">
959 <td class="left">
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]960 <tt>{ i32, i32, i32 }</tt><br/>
961 <tt>{ float, i32 (i32) * }</tt><br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]962 </td>
963 <td class="left">
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]964 a triple of three <tt>i32</tt> values<br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]965 A pair, where the first element is a <tt>float</tt> and the second element
966 is a <a href="#t_pointer">pointer</a> to a <a href="#t_function">function</a>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]967 that takes an <tt>i32</tt>, returning an <tt>i32</tt>.<br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]968 </td>
969 </tr>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]970</table>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]971</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]972
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]973<!-- _______________________________________________________________________ -->
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]974<div class="doc_subsubsection"> <a name="t_pstruct">Packed Structure Type</a>
975</div>
976<div class="doc_text">
977<h5>Overview:</h5>
978<p>The packed structure type is used to represent a collection of data members
979together in memory. There is no padding between fields. Further, the alignment
980of a packed structure is 1 byte. The elements of a packed structure may
981be any type that has a size.</p>
982<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
983and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
984field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
985instruction.</p>
986<h5>Syntax:</h5>
987<pre> &lt; { &lt;type list&gt; } &gt; <br></pre>
988<h5>Examples:</h5>
989<table class="layout">
990 <tr class="layout">
991 <td class="left">
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]992 <tt> &lt; { i32, i32, i32 } &gt; </tt><br/>
993 <tt> &lt; { float, i32 (i32) * } &gt; </tt><br/>
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]994 </td>
995 <td class="left">
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]996 a triple of three <tt>i32</tt> values<br/>
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]997 A pair, where the first element is a <tt>float</tt> and the second element
998 is a <a href="#t_pointer">pointer</a> to a <a href="#t_function">function</a>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]999 that takes an <tt>i32</tt>, returning an <tt>i32</tt>.<br/>
Andrew Lenharth75e10682006-12-08 17:13:00 +0000[diff] [blame]1000 </td>
1001 </tr>
1002</table>
1003</div>
1004
1005<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1006<div class="doc_subsubsection"> <a name="t_pointer">Pointer Type</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1007<div class="doc_text">
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1008<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1009<p>As in many languages, the pointer type represents a pointer or
1010reference to another object, which must live in memory.</p>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1011<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1012<pre> &lt;type&gt; *<br></pre>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1013<h5>Examples:</h5>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1014<table class="layout">
1015 <tr class="layout">
1016 <td class="left">
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1017 <tt>[4x i32]*</tt><br/>
1018 <tt>i32 (i32 *) *</tt><br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1019 </td>
1020 <td class="left">
1021 A <a href="#t_pointer">pointer</a> to <a href="#t_array">array</a> of
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1022 four <tt>i32</tt> values<br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1023 A <a href="#t_pointer">pointer</a> to a <a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1024 href="#t_function">function</a> that takes an <tt>i32*</tt>, returning an
1025 <tt>i32</tt>.<br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1026 </td>
1027 </tr>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1028</table>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1029</div>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1030
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1031<!-- _______________________________________________________________________ -->
1032<div class="doc_subsubsection"> <a name="t_packed">Packed Type</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1033<div class="doc_text">
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1034
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1035<h5>Overview:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1036
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1037<p>A packed type is a simple derived type that represents a vector
1038of elements. Packed types are used when multiple primitive data
1039are operated in parallel using a single instruction (SIMD).
1040A packed type requires a size (number of
Chris Lattnerb8d172f2005-11-10 01:44:22 +0000[diff] [blame]1041elements) and an underlying primitive data type. Vectors must have a power
1042of two length (1, 2, 4, 8, 16 ...). Packed types are
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1043considered <a href="#t_firstclass">first class</a>.</p>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1044
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1045<h5>Syntax:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1046
1047<pre>
1048 &lt; &lt;# elements&gt; x &lt;elementtype&gt; &gt;
1049</pre>
1050
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]1051<p>The number of elements is a constant integer value; elementtype may
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1052be any integral or floating point type.</p>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1053
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1054<h5>Examples:</h5>
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1055
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1056<table class="layout">
1057 <tr class="layout">
1058 <td class="left">
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1059 <tt>&lt;4 x i32&gt;</tt><br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1060 <tt>&lt;8 x float&gt;</tt><br/>
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]1061 <tt>&lt;2 x i64&gt;</tt><br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1062 </td>
1063 <td class="left">
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]1064 Packed vector of 4 32-bit integer values.<br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1065 Packed vector of 8 floating-point values.<br/>
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]1066 Packed vector of 2 64-bit integer values.<br/>
Reid Spencerd3f876c2004-11-01 08:19:36 +0000[diff] [blame]1067 </td>
1068 </tr>
1069</table>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1070</div>
1071
Chris Lattner69c11bb2005-04-25 17:34:15 +0000[diff] [blame]1072<!-- _______________________________________________________________________ -->
1073<div class="doc_subsubsection"> <a name="t_opaque">Opaque Type</a> </div>
1074<div class="doc_text">
1075
1076<h5>Overview:</h5>
1077
1078<p>Opaque types are used to represent unknown types in the system. This
1079corresponds (for example) to the C notion of a foward declared structure type.
1080In LLVM, opaque types can eventually be resolved to any type (not just a
1081structure type).</p>
1082
1083<h5>Syntax:</h5>
1084
1085<pre>
1086 opaque
1087</pre>
1088
1089<h5>Examples:</h5>
1090
1091<table class="layout">
1092 <tr class="layout">
1093 <td class="left">
1094 <tt>opaque</tt>
1095 </td>
1096 <td class="left">
1097 An opaque type.<br/>
1098 </td>
1099 </tr>
1100</table>
1101</div>
1102
1103
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1104<!-- *********************************************************************** -->
1105<div class="doc_section"> <a name="constants">Constants</a> </div>
1106<!-- *********************************************************************** -->
1107
1108<div class="doc_text">
1109
1110<p>LLVM has several different basic types of constants. This section describes
1111them all and their syntax.</p>
1112
1113</div>
1114
1115<!-- ======================================================================= -->
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]1116<div class="doc_subsection"><a name="simpleconstants">Simple Constants</a></div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1117
1118<div class="doc_text">
1119
1120<dl>
1121 <dt><b>Boolean constants</b></dt>
1122
1123 <dd>The two strings '<tt>true</tt>' and '<tt>false</tt>' are both valid
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]1124 constants of the <tt><a href="#t_primitive">i1</a></tt> type.
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1125 </dd>
1126
1127 <dt><b>Integer constants</b></dt>
1128
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]1129 <dd>Standard integers (such as '4') are constants of the <a
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]1130 href="#t_integer">integer</a> type. Negative numbers may be used with
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1131 integer types.
1132 </dd>
1133
1134 <dt><b>Floating point constants</b></dt>
1135
1136 <dd>Floating point constants use standard decimal notation (e.g. 123.421),
1137 exponential notation (e.g. 1.23421e+2), or a more precise hexadecimal
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1138 notation (see below). Floating point constants must have a <a
1139 href="#t_floating">floating point</a> type. </dd>
1140
1141 <dt><b>Null pointer constants</b></dt>
1142
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1143 <dd>The identifier '<tt>null</tt>' is recognized as a null pointer constant
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1144 and must be of <a href="#t_pointer">pointer type</a>.</dd>
1145
1146</dl>
1147
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1148<p>The one non-intuitive notation for constants is the optional hexadecimal form
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1149of floating point constants. For example, the form '<tt>double
11500x432ff973cafa8000</tt>' is equivalent to (but harder to read than) '<tt>double
11514.5e+15</tt>'. The only time hexadecimal floating point constants are required
Reid Spencercc16dc32004-12-09 18:02:53 +0000[diff] [blame]1152(and the only time that they are generated by the disassembler) is when a
1153floating point constant must be emitted but it cannot be represented as a
1154decimal floating point number. For example, NaN's, infinities, and other
1155special values are represented in their IEEE hexadecimal format so that
1156assembly and disassembly do not cause any bits to change in the constants.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1157
1158</div>
1159
1160<!-- ======================================================================= -->
1161<div class="doc_subsection"><a name="aggregateconstants">Aggregate Constants</a>
1162</div>
1163
1164<div class="doc_text">
Chris Lattnerd4f6b172005-03-07 22:13:59 +0000[diff] [blame]1165<p>Aggregate constants arise from aggregation of simple constants
1166and smaller aggregate constants.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1167
1168<dl>
1169 <dt><b>Structure constants</b></dt>
1170
1171 <dd>Structure constants are represented with notation similar to structure
1172 type definitions (a comma separated list of elements, surrounded by braces
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1173 (<tt>{}</tt>)). For example: "<tt>{ i32 4, float 17.0, i32* %G }</tt>",
1174 where "<tt>%G</tt>" is declared as "<tt>%G = external global i32</tt>". Structure constants
Chris Lattnerd4f6b172005-03-07 22:13:59 +0000[diff] [blame]1175 must have <a href="#t_struct">structure type</a>, and the number and
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1176 types of elements must match those specified by the type.
1177 </dd>
1178
1179 <dt><b>Array constants</b></dt>
1180
1181 <dd>Array constants are represented with notation similar to array type
1182 definitions (a comma separated list of elements, surrounded by square brackets
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1183 (<tt>[]</tt>)). For example: "<tt>[ i32 42, i32 11, i32 74 ]</tt>". Array
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1184 constants must have <a href="#t_array">array type</a>, and the number and
1185 types of elements must match those specified by the type.
1186 </dd>
1187
1188 <dt><b>Packed constants</b></dt>
1189
1190 <dd>Packed constants are represented with notation similar to packed type
1191 definitions (a comma separated list of elements, surrounded by
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1192 less-than/greater-than's (<tt>&lt;&gt;</tt>)). For example: "<tt>&lt; i32 42,
1193 i32 11, i32 74, i32 100 &gt;</tt>". Packed constants must have <a
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1194 href="#t_packed">packed type</a>, and the number and types of elements must
1195 match those specified by the type.
1196 </dd>
1197
1198 <dt><b>Zero initialization</b></dt>
1199
1200 <dd>The string '<tt>zeroinitializer</tt>' can be used to zero initialize a
1201 value to zero of <em>any</em> type, including scalar and aggregate types.
1202 This is often used to avoid having to print large zero initializers (e.g. for
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]1203 large arrays) and is always exactly equivalent to using explicit zero
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1204 initializers.
1205 </dd>
1206</dl>
1207
1208</div>
1209
1210<!-- ======================================================================= -->
1211<div class="doc_subsection">
1212 <a name="globalconstants">Global Variable and Function Addresses</a>
1213</div>
1214
1215<div class="doc_text">
1216
1217<p>The addresses of <a href="#globalvars">global variables</a> and <a
1218href="#functionstructure">functions</a> are always implicitly valid (link-time)
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1219constants. These constants are explicitly referenced when the <a
1220href="#identifiers">identifier for the global</a> is used and always have <a
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1221href="#t_pointer">pointer</a> type. For example, the following is a legal LLVM
1222file:</p>
1223
1224<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1225 %X = global i32 17
1226 %Y = global i32 42
1227 %Z = global [2 x i32*] [ i32* %X, i32* %Y ]
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1228</pre>
1229
1230</div>
1231
1232<!-- ======================================================================= -->
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]1233<div class="doc_subsection"><a name="undefvalues">Undefined Values</a></div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1234<div class="doc_text">
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]1235 <p>The string '<tt>undef</tt>' is recognized as a type-less constant that has
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]1236 no specific value. Undefined values may be of any type and be used anywhere
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]1237 a constant is permitted.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1238
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]1239 <p>Undefined values indicate to the compiler that the program is well defined
1240 no matter what value is used, giving the compiler more freedom to optimize.
1241 </p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1242</div>
1243
1244<!-- ======================================================================= -->
1245<div class="doc_subsection"><a name="constantexprs">Constant Expressions</a>
1246</div>
1247
1248<div class="doc_text">
1249
1250<p>Constant expressions are used to allow expressions involving other constants
1251to be used as constants. Constant expressions may be of any <a
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]1252href="#t_firstclass">first class</a> type and may involve any LLVM operation
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1253that does not have side effects (e.g. load and call are not supported). The
1254following is the syntax for constant expressions:</p>
1255
1256<dl>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1257 <dt><b><tt>trunc ( CST to TYPE )</tt></b></dt>
1258 <dd>Truncate a constant to another type. The bit size of CST must be larger
1259 than the bit size of TYPE. Both types must be integral.</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1260
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1261 <dt><b><tt>zext ( CST to TYPE )</tt></b></dt>
1262 <dd>Zero extend a constant to another type. The bit size of CST must be
1263 smaller or equal to the bit size of TYPE. Both types must be integral.</dd>
1264
1265 <dt><b><tt>sext ( CST to TYPE )</tt></b></dt>
1266 <dd>Sign extend a constant to another type. The bit size of CST must be
1267 smaller or equal to the bit size of TYPE. Both types must be integral.</dd>
1268
1269 <dt><b><tt>fptrunc ( CST to TYPE )</tt></b></dt>
1270 <dd>Truncate a floating point constant to another floating point type. The
1271 size of CST must be larger than the size of TYPE. Both types must be
1272 floating point.</dd>
1273
1274 <dt><b><tt>fpext ( CST to TYPE )</tt></b></dt>
1275 <dd>Floating point extend a constant to another type. The size of CST must be
1276 smaller or equal to the size of TYPE. Both types must be floating point.</dd>
1277
1278 <dt><b><tt>fp2uint ( CST to TYPE )</tt></b></dt>
1279 <dd>Convert a floating point constant to the corresponding unsigned integer
1280 constant. TYPE must be an integer type. CST must be floating point. If the
1281 value won't fit in the integer type, the results are undefined.</dd>
1282
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]1283 <dt><b><tt>fptosi ( CST to TYPE )</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1284 <dd>Convert a floating point constant to the corresponding signed integer
1285 constant. TYPE must be an integer type. CST must be floating point. If the
1286 value won't fit in the integer type, the results are undefined.</dd>
1287
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]1288 <dt><b><tt>uitofp ( CST to TYPE )</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1289 <dd>Convert an unsigned integer constant to the corresponding floating point
1290 constant. TYPE must be floating point. CST must be of integer type. If the
1291 value won't fit in the floating point type, the results are undefined.</dd>
1292
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]1293 <dt><b><tt>sitofp ( CST to TYPE )</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1294 <dd>Convert a signed integer constant to the corresponding floating point
1295 constant. TYPE must be floating point. CST must be of integer type. If the
1296 value won't fit in the floating point type, the results are undefined.</dd>
1297
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]1298 <dt><b><tt>ptrtoint ( CST to TYPE )</tt></b></dt>
1299 <dd>Convert a pointer typed constant to the corresponding integer constant
1300 TYPE must be an integer type. CST must be of pointer type. The CST value is
1301 zero extended, truncated, or unchanged to make it fit in TYPE.</dd>
1302
1303 <dt><b><tt>inttoptr ( CST to TYPE )</tt></b></dt>
1304 <dd>Convert a integer constant to a pointer constant. TYPE must be a
1305 pointer type. CST must be of integer type. The CST value is zero extended,
1306 truncated, or unchanged to make it fit in a pointer size. This one is
1307 <i>really</i> dangerous!</dd>
1308
1309 <dt><b><tt>bitcast ( CST to TYPE )</tt></b></dt>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1310 <dd>Convert a constant, CST, to another TYPE. The size of CST and TYPE must be
1311 identical (same number of bits). The conversion is done as if the CST value
1312 was stored to memory and read back as TYPE. In other words, no bits change
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]1313 with this operator, just the type. This can be used for conversion of
1314 packed types to any other type, as long as they have the same bit width. For
1315 pointers it is only valid to cast to another pointer type.
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]1316 </dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1317
1318 <dt><b><tt>getelementptr ( CSTPTR, IDX0, IDX1, ... )</tt></b></dt>
1319
1320 <dd>Perform the <a href="#i_getelementptr">getelementptr operation</a> on
1321 constants. As with the <a href="#i_getelementptr">getelementptr</a>
1322 instruction, the index list may have zero or more indexes, which are required
1323 to make sense for the type of "CSTPTR".</dd>
1324
Robert Bocchino9fbe1452006-01-10 19:31:34 +0000[diff] [blame]1325 <dt><b><tt>select ( COND, VAL1, VAL2 )</tt></b></dt>
1326
1327 <dd>Perform the <a href="#i_select">select operation</a> on
Reid Spencer01c42592006-12-04 19:23:19 +0000[diff] [blame]1328 constants.</dd>
1329
1330 <dt><b><tt>icmp COND ( VAL1, VAL2 )</tt></b></dt>
1331 <dd>Performs the <a href="#i_icmp">icmp operation</a> on constants.</dd>
1332
1333 <dt><b><tt>fcmp COND ( VAL1, VAL2 )</tt></b></dt>
1334 <dd>Performs the <a href="#i_fcmp">fcmp operation</a> on constants.</dd>
Robert Bocchino9fbe1452006-01-10 19:31:34 +0000[diff] [blame]1335
1336 <dt><b><tt>extractelement ( VAL, IDX )</tt></b></dt>
1337
1338 <dd>Perform the <a href="#i_extractelement">extractelement
1339 operation</a> on constants.
1340
Robert Bocchino05ccd702006-01-15 20:48:27 +0000[diff] [blame]1341 <dt><b><tt>insertelement ( VAL, ELT, IDX )</tt></b></dt>
1342
1343 <dd>Perform the <a href="#i_insertelement">insertelement
Reid Spencer01c42592006-12-04 19:23:19 +0000[diff] [blame]1344 operation</a> on constants.</dd>
Robert Bocchino05ccd702006-01-15 20:48:27 +0000[diff] [blame]1345
Chris Lattnerc1989542006-04-08 00:13:41 +0000[diff] [blame]1346
1347 <dt><b><tt>shufflevector ( VEC1, VEC2, IDXMASK )</tt></b></dt>
1348
1349 <dd>Perform the <a href="#i_shufflevector">shufflevector
Reid Spencer01c42592006-12-04 19:23:19 +0000[diff] [blame]1350 operation</a> on constants.</dd>
Chris Lattnerc1989542006-04-08 00:13:41 +0000[diff] [blame]1351
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1352 <dt><b><tt>OPCODE ( LHS, RHS )</tt></b></dt>
1353
Reid Spencer2dc45b82004-12-09 18:13:12 +0000[diff] [blame]1354 <dd>Perform the specified operation of the LHS and RHS constants. OPCODE may
1355 be any of the <a href="#binaryops">binary</a> or <a href="#bitwiseops">bitwise
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1356 binary</a> operations. The constraints on operands are the same as those for
1357 the corresponding instruction (e.g. no bitwise operations on floating point
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]1358 values are allowed).</dd>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1359</dl>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1360</div>
Chris Lattner9ee5d222004-03-08 16:49:10 +0000[diff] [blame]1361
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1362<!-- *********************************************************************** -->
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]1363<div class="doc_section"> <a name="othervalues">Other Values</a> </div>
1364<!-- *********************************************************************** -->
1365
1366<!-- ======================================================================= -->
1367<div class="doc_subsection">
1368<a name="inlineasm">Inline Assembler Expressions</a>
1369</div>
1370
1371<div class="doc_text">
1372
1373<p>
1374LLVM supports inline assembler expressions (as opposed to <a href="#moduleasm">
1375Module-Level Inline Assembly</a>) through the use of a special value. This
1376value represents the inline assembler as a string (containing the instructions
1377to emit), a list of operand constraints (stored as a string), and a flag that
1378indicates whether or not the inline asm expression has side effects. An example
1379inline assembler expression is:
1380</p>
1381
1382<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1383 i32 (i32) asm "bswap $0", "=r,r"
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]1384</pre>
1385
1386<p>
1387Inline assembler expressions may <b>only</b> be used as the callee operand of
1388a <a href="#i_call"><tt>call</tt> instruction</a>. Thus, typically we have:
1389</p>
1390
1391<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1392 %X = call i32 asm "<a href="#i_bswap">bswap</a> $0", "=r,r"(i32 %Y)
Chris Lattnere87d6532006-01-25 23:47:57 +0000[diff] [blame]1393</pre>
1394
1395<p>
1396Inline asms with side effects not visible in the constraint list must be marked
1397as having side effects. This is done through the use of the
1398'<tt>sideeffect</tt>' keyword, like so:
1399</p>
1400
1401<pre>
1402 call void asm sideeffect "eieio", ""()
1403</pre>
1404
1405<p>TODO: The format of the asm and constraints string still need to be
1406documented here. Constraints on what can be done (e.g. duplication, moving, etc
1407need to be documented).
1408</p>
1409
1410</div>
1411
1412<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1413<div class="doc_section"> <a name="instref">Instruction Reference</a> </div>
1414<!-- *********************************************************************** -->
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1415
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1416<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1417
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1418<p>The LLVM instruction set consists of several different
1419classifications of instructions: <a href="#terminators">terminator
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]1420instructions</a>, <a href="#binaryops">binary instructions</a>,
1421<a href="#bitwiseops">bitwise binary instructions</a>, <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1422 href="#memoryops">memory instructions</a>, and <a href="#otherops">other
1423instructions</a>.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1424
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1425</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1426
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1427<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1428<div class="doc_subsection"> <a name="terminators">Terminator
1429Instructions</a> </div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1430
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1431<div class="doc_text">
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1432
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1433<p>As mentioned <a href="#functionstructure">previously</a>, every
1434basic block in a program ends with a "Terminator" instruction, which
1435indicates which block should be executed after the current block is
1436finished. These terminator instructions typically yield a '<tt>void</tt>'
1437value: they produce control flow, not values (the one exception being
1438the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction).</p>
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1439<p>There are six different terminator instructions: the '<a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1440 href="#i_ret"><tt>ret</tt></a>' instruction, the '<a href="#i_br"><tt>br</tt></a>'
1441instruction, the '<a href="#i_switch"><tt>switch</tt></a>' instruction,
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1442the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction, the '<a
1443 href="#i_unwind"><tt>unwind</tt></a>' instruction, and the '<a
1444 href="#i_unreachable"><tt>unreachable</tt></a>' instruction.</p>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1445
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1446</div>
Chris Lattnerc3f59762004-12-09 17:30:23 +0000[diff] [blame]1447
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1448<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1449<div class="doc_subsubsection"> <a name="i_ret">'<tt>ret</tt>'
1450Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1451<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1452<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1453<pre> ret &lt;type&gt; &lt;value&gt; <i>; Return a value from a non-void function</i>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1454 ret void <i>; Return from void function</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1455</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1456<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1457<p>The '<tt>ret</tt>' instruction is used to return control flow (and a
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]1458value) from a function back to the caller.</p>
John Criswell4457dc92004-04-09 16:48:45 +0000[diff] [blame]1459<p>There are two forms of the '<tt>ret</tt>' instruction: one that
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1460returns a value and then causes control flow, and one that just causes
1461control flow to occur.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1462<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1463<p>The '<tt>ret</tt>' instruction may return any '<a
1464 href="#t_firstclass">first class</a>' type. Notice that a function is
1465not <a href="#wellformed">well formed</a> if there exists a '<tt>ret</tt>'
1466instruction inside of the function that returns a value that does not
1467match the return type of the function.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1468<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1469<p>When the '<tt>ret</tt>' instruction is executed, control flow
1470returns back to the calling function's context. If the caller is a "<a
John Criswellfa081872004-06-25 15:16:57 +0000[diff] [blame]1471 href="#i_call"><tt>call</tt></a>" instruction, execution continues at
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1472the instruction after the call. If the caller was an "<a
1473 href="#i_invoke"><tt>invoke</tt></a>" instruction, execution continues
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]1474at the beginning of the "normal" destination block. If the instruction
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1475returns a value, that value shall set the call or invoke instruction's
1476return value.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1477<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1478<pre> ret i32 5 <i>; Return an integer value of 5</i>
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]1479 ret void <i>; Return from a void function</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1480</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1481</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1482<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1483<div class="doc_subsubsection"> <a name="i_br">'<tt>br</tt>' Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1484<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1485<h5>Syntax:</h5>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]1486<pre> br i1 &lt;cond&gt;, label &lt;iftrue&gt;, label &lt;iffalse&gt;<br> br label &lt;dest&gt; <i>; Unconditional branch</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1487</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1488<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1489<p>The '<tt>br</tt>' instruction is used to cause control flow to
1490transfer to a different basic block in the current function. There are
1491two forms of this instruction, corresponding to a conditional branch
1492and an unconditional branch.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1493<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1494<p>The conditional branch form of the '<tt>br</tt>' instruction takes a
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]1495single '<tt>i1</tt>' value and two '<tt>label</tt>' values. The
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1496unconditional form of the '<tt>br</tt>' instruction takes a single '<tt>label</tt>'
1497value as a target.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1498<h5>Semantics:</h5>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]1499<p>Upon execution of a conditional '<tt>br</tt>' instruction, the '<tt>i1</tt>'
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1500argument is evaluated. If the value is <tt>true</tt>, control flows
1501to the '<tt>iftrue</tt>' <tt>label</tt> argument. If "cond" is <tt>false</tt>,
1502control flows to the '<tt>iffalse</tt>' <tt>label</tt> argument.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1503<h5>Example:</h5>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]1504<pre>Test:<br> %cond = <a href="#i_icmp">icmp</a> eq, i32 %a, %b<br> br i1 %cond, label %IfEqual, label %IfUnequal<br>IfEqual:<br> <a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1505 href="#i_ret">ret</a> i32 1<br>IfUnequal:<br> <a href="#i_ret">ret</a> i32 0<br></pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1506</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1507<!-- _______________________________________________________________________ -->
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1508<div class="doc_subsubsection">
1509 <a name="i_switch">'<tt>switch</tt>' Instruction</a>
1510</div>
1511
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1512<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1513<h5>Syntax:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1514
1515<pre>
1516 switch &lt;intty&gt; &lt;value&gt;, label &lt;defaultdest&gt; [ &lt;intty&gt; &lt;val&gt;, label &lt;dest&gt; ... ]
1517</pre>
1518
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1519<h5>Overview:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1520
1521<p>The '<tt>switch</tt>' instruction is used to transfer control flow to one of
1522several different places. It is a generalization of the '<tt>br</tt>'
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1523instruction, allowing a branch to occur to one of many possible
1524destinations.</p>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1525
1526
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1527<h5>Arguments:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1528
1529<p>The '<tt>switch</tt>' instruction uses three parameters: an integer
1530comparison value '<tt>value</tt>', a default '<tt>label</tt>' destination, and
1531an array of pairs of comparison value constants and '<tt>label</tt>'s. The
1532table is not allowed to contain duplicate constant entries.</p>
1533
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1534<h5>Semantics:</h5>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1535
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1536<p>The <tt>switch</tt> instruction specifies a table of values and
1537destinations. When the '<tt>switch</tt>' instruction is executed, this
John Criswell84114752004-06-25 16:05:06 +0000[diff] [blame]1538table is searched for the given value. If the value is found, control flow is
1539transfered to the corresponding destination; otherwise, control flow is
1540transfered to the default destination.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1541
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1542<h5>Implementation:</h5>
1543
1544<p>Depending on properties of the target machine and the particular
1545<tt>switch</tt> instruction, this instruction may be code generated in different
John Criswell84114752004-06-25 16:05:06 +0000[diff] [blame]1546ways. For example, it could be generated as a series of chained conditional
1547branches or with a lookup table.</p>
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1548
1549<h5>Example:</h5>
1550
1551<pre>
1552 <i>; Emulate a conditional br instruction</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]1553 %Val = <a href="#i_zext">zext</a> i1 %value to i32
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1554 switch i32 %Val, label %truedest [i32 0, label %falsedest ]
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1555
1556 <i>; Emulate an unconditional br instruction</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1557 switch i32 0, label %dest [ ]
Chris Lattnerc88c17b2004-02-24 04:54:45 +0000[diff] [blame]1558
1559 <i>; Implement a jump table:</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1560 switch i32 %val, label %otherwise [ i32 0, label %onzero
1561 i32 1, label %onone
1562 i32 2, label %ontwo ]
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1563</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1564</div>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1565
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1566<!-- _______________________________________________________________________ -->
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1567<div class="doc_subsubsection">
1568 <a name="i_invoke">'<tt>invoke</tt>' Instruction</a>
1569</div>
1570
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1571<div class="doc_text">
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1572
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1573<h5>Syntax:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1574
1575<pre>
1576 &lt;result&gt; = invoke [<a href="#callingconv">cconv</a>] &lt;ptr to function ty&gt; %&lt;function ptr val&gt;(&lt;function args&gt;)
Chris Lattner76b8a332006-05-14 18:23:06 +0000[diff] [blame]1577 to label &lt;normal label&gt; unwind label &lt;exception label&gt;
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1578</pre>
1579
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]1580<h5>Overview:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1581
1582<p>The '<tt>invoke</tt>' instruction causes control to transfer to a specified
1583function, with the possibility of control flow transfer to either the
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]1584'<tt>normal</tt>' label or the
1585'<tt>exception</tt>' label. If the callee function returns with the
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1586"<tt><a href="#i_ret">ret</a></tt>" instruction, control flow will return to the
1587"normal" label. If the callee (or any indirect callees) returns with the "<a
John Criswelle4c57cc2005-05-12 16:52:32 +0000[diff] [blame]1588href="#i_unwind"><tt>unwind</tt></a>" instruction, control is interrupted and
1589continued at the dynamically nearest "exception" label.</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1590
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1591<h5>Arguments:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1592
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1593<p>This instruction requires several arguments:</p>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1594
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1595<ol>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1596 <li>
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]1597 The optional "cconv" marker indicates which <a href="callingconv">calling
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1598 convention</a> the call should use. If none is specified, the call defaults
1599 to using C calling conventions.
1600 </li>
1601 <li>'<tt>ptr to function ty</tt>': shall be the signature of the pointer to
1602 function value being invoked. In most cases, this is a direct function
1603 invocation, but indirect <tt>invoke</tt>s are just as possible, branching off
1604 an arbitrary pointer to function value.
1605 </li>
1606
1607 <li>'<tt>function ptr val</tt>': An LLVM value containing a pointer to a
1608 function to be invoked. </li>
1609
1610 <li>'<tt>function args</tt>': argument list whose types match the function
1611 signature argument types. If the function signature indicates the function
1612 accepts a variable number of arguments, the extra arguments can be
1613 specified. </li>
1614
1615 <li>'<tt>normal label</tt>': the label reached when the called function
1616 executes a '<tt><a href="#i_ret">ret</a></tt>' instruction. </li>
1617
1618 <li>'<tt>exception label</tt>': the label reached when a callee returns with
1619 the <a href="#i_unwind"><tt>unwind</tt></a> instruction. </li>
1620
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1621</ol>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1622
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1623<h5>Semantics:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1624
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1625<p>This instruction is designed to operate as a standard '<tt><a
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1626href="#i_call">call</a></tt>' instruction in most regards. The primary
1627difference is that it establishes an association with a label, which is used by
1628the runtime library to unwind the stack.</p>
1629
1630<p>This instruction is used in languages with destructors to ensure that proper
1631cleanup is performed in the case of either a <tt>longjmp</tt> or a thrown
1632exception. Additionally, this is important for implementation of
1633'<tt>catch</tt>' clauses in high-level languages that support them.</p>
1634
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1635<h5>Example:</h5>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]1636<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1637 %retval = invoke i32 %Test(i32 15) to label %Continue
1638 unwind label %TestCleanup <i>; {i32}:retval set</i>
1639 %retval = invoke <a href="#callingconv">coldcc</a> i32 %Test(i32 15) to label %Continue
1640 unwind label %TestCleanup <i>; {i32}:retval set</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1641</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1642</div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1643
1644
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]1645<!-- _______________________________________________________________________ -->
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1646
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1647<div class="doc_subsubsection"> <a name="i_unwind">'<tt>unwind</tt>'
1648Instruction</a> </div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1649
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1650<div class="doc_text">
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1651
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]1652<h5>Syntax:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1653<pre>
1654 unwind
1655</pre>
1656
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]1657<h5>Overview:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1658
1659<p>The '<tt>unwind</tt>' instruction unwinds the stack, continuing control flow
1660at the first callee in the dynamic call stack which used an <a
1661href="#i_invoke"><tt>invoke</tt></a> instruction to perform the call. This is
1662primarily used to implement exception handling.</p>
1663
Chris Lattner27f71f22003-09-03 00:41:47 +0000[diff] [blame]1664<h5>Semantics:</h5>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1665
1666<p>The '<tt>unwind</tt>' intrinsic causes execution of the current function to
1667immediately halt. The dynamic call stack is then searched for the first <a
1668href="#i_invoke"><tt>invoke</tt></a> instruction on the call stack. Once found,
1669execution continues at the "exceptional" destination block specified by the
1670<tt>invoke</tt> instruction. If there is no <tt>invoke</tt> instruction in the
1671dynamic call chain, undefined behavior results.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1672</div>
Chris Lattner35eca582004-10-16 18:04:13 +0000[diff] [blame]1673
1674<!-- _______________________________________________________________________ -->
1675
1676<div class="doc_subsubsection"> <a name="i_unreachable">'<tt>unreachable</tt>'
1677Instruction</a> </div>
1678
1679<div class="doc_text">
1680
1681<h5>Syntax:</h5>
1682<pre>
1683 unreachable
1684</pre>
1685
1686<h5>Overview:</h5>
1687
1688<p>The '<tt>unreachable</tt>' instruction has no defined semantics. This
1689instruction is used to inform the optimizer that a particular portion of the
1690code is not reachable. This can be used to indicate that the code after a
1691no-return function cannot be reached, and other facts.</p>
1692
1693<h5>Semantics:</h5>
1694
1695<p>The '<tt>unreachable</tt>' instruction has no defined semantics.</p>
1696</div>
1697
1698
1699
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1700<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1701<div class="doc_subsection"> <a name="binaryops">Binary Operations</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1702<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1703<p>Binary operators are used to do most of the computation in a
1704program. They require two operands, execute an operation on them, and
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1705produce a single value. The operands might represent
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1706multiple data, as is the case with the <a href="#t_packed">packed</a> data type.
1707The result value of a binary operator is not
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1708necessarily the same type as its operands.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1709<p>There are several different binary operators:</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1710</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1711<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1712<div class="doc_subsubsection"> <a name="i_add">'<tt>add</tt>'
1713Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1714<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1715<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1716<pre> &lt;result&gt; = add &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1717</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1718<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1719<p>The '<tt>add</tt>' instruction returns the sum of its two operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1720<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1721<p>The two arguments to the '<tt>add</tt>' instruction must be either <a
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1722 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a> values.
1723 This instruction can also take <a href="#t_packed">packed</a> versions of the values.
1724Both arguments must have identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1725<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1726<p>The value produced is the integer or floating point sum of the two
1727operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1728<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1729<pre> &lt;result&gt; = add i32 4, %var <i>; yields {i32}:result = 4 + %var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1730</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1731</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1732<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1733<div class="doc_subsubsection"> <a name="i_sub">'<tt>sub</tt>'
1734Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1735<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1736<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1737<pre> &lt;result&gt; = sub &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1738</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1739<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1740<p>The '<tt>sub</tt>' instruction returns the difference of its two
1741operands.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1742<p>Note that the '<tt>sub</tt>' instruction is used to represent the '<tt>neg</tt>'
1743instruction present in most other intermediate representations.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1744<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1745<p>The two arguments to the '<tt>sub</tt>' instruction must be either <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1746 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1747values.
1748This instruction can also take <a href="#t_packed">packed</a> versions of the values.
1749Both arguments must have identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1750<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1751<p>The value produced is the integer or floating point difference of
1752the two operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1753<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1754<pre> &lt;result&gt; = sub i32 4, %var <i>; yields {i32}:result = 4 - %var</i>
1755 &lt;result&gt; = sub i32 0, %val <i>; yields {i32}:result = -%var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1756</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1757</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1758<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1759<div class="doc_subsubsection"> <a name="i_mul">'<tt>mul</tt>'
1760Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1761<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1762<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1763<pre> &lt;result&gt; = mul &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1764</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1765<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1766<p>The '<tt>mul</tt>' instruction returns the product of its two
1767operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1768<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1769<p>The two arguments to the '<tt>mul</tt>' instruction must be either <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1770 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
Chris Lattnera58561b2004-08-12 19:12:28 +0000[diff] [blame]1771values.
1772This instruction can also take <a href="#t_packed">packed</a> versions of the values.
1773Both arguments must have identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1774<h5>Semantics:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1775<p>The value produced is the integer or floating point product of the
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1776two operands.</p>
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]1777<p>Because the operands are the same width, the result of an integer
1778multiplication is the same whether the operands should be deemed unsigned or
1779signed.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1780<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1781<pre> &lt;result&gt; = mul i32 4, %var <i>; yields {i32}:result = 4 * %var</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1782</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1783</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1784<!-- _______________________________________________________________________ -->
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]1785<div class="doc_subsubsection"> <a name="i_udiv">'<tt>udiv</tt>' Instruction
1786</a></div>
1787<div class="doc_text">
1788<h5>Syntax:</h5>
1789<pre> &lt;result&gt; = udiv &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
1790</pre>
1791<h5>Overview:</h5>
1792<p>The '<tt>udiv</tt>' instruction returns the quotient of its two
1793operands.</p>
1794<h5>Arguments:</h5>
1795<p>The two arguments to the '<tt>udiv</tt>' instruction must be
1796<a href="#t_integer">integer</a> values. Both arguments must have identical
1797types. This instruction can also take <a href="#t_packed">packed</a> versions
1798of the values in which case the elements must be integers.</p>
1799<h5>Semantics:</h5>
1800<p>The value produced is the unsigned integer quotient of the two operands. This
1801instruction always performs an unsigned division operation, regardless of
1802whether the arguments are unsigned or not.</p>
1803<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1804<pre> &lt;result&gt; = udiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]1805</pre>
1806</div>
1807<!-- _______________________________________________________________________ -->
1808<div class="doc_subsubsection"> <a name="i_sdiv">'<tt>sdiv</tt>' Instruction
1809</a> </div>
1810<div class="doc_text">
1811<h5>Syntax:</h5>
1812<pre> &lt;result&gt; = sdiv &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
1813</pre>
1814<h5>Overview:</h5>
1815<p>The '<tt>sdiv</tt>' instruction returns the quotient of its two
1816operands.</p>
1817<h5>Arguments:</h5>
1818<p>The two arguments to the '<tt>sdiv</tt>' instruction must be
1819<a href="#t_integer">integer</a> values. Both arguments must have identical
1820types. This instruction can also take <a href="#t_packed">packed</a> versions
1821of the values in which case the elements must be integers.</p>
1822<h5>Semantics:</h5>
1823<p>The value produced is the signed integer quotient of the two operands. This
1824instruction always performs a signed division operation, regardless of whether
1825the arguments are signed or not.</p>
1826<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1827<pre> &lt;result&gt; = sdiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]1828</pre>
1829</div>
1830<!-- _______________________________________________________________________ -->
1831<div class="doc_subsubsection"> <a name="i_fdiv">'<tt>fdiv</tt>'
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1832Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1833<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1834<h5>Syntax:</h5>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]1835<pre> &lt;result&gt; = fdiv &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1836</pre>
1837<h5>Overview:</h5>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]1838<p>The '<tt>fdiv</tt>' instruction returns the quotient of its two
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1839operands.</p>
1840<h5>Arguments:</h5>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]1841<p>The two arguments to the '<tt>div</tt>' instruction must be
1842<a href="#t_floating">floating point</a> values. Both arguments must have
1843identical types. This instruction can also take <a href="#t_packed">packed</a>
1844versions of the values in which case the elements must be floating point.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1845<h5>Semantics:</h5>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]1846<p>The value produced is the floating point quotient of the two operands.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1847<h5>Example:</h5>
Reid Spencer1628cec2006-10-26 06:15:43 +0000[diff] [blame]1848<pre> &lt;result&gt; = fdiv float 4.0, %var <i>; yields {float}:result = 4.0 / %var</i>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1849</pre>
1850</div>
1851<!-- _______________________________________________________________________ -->
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]1852<div class="doc_subsubsection"> <a name="i_urem">'<tt>urem</tt>' Instruction</a>
1853</div>
1854<div class="doc_text">
1855<h5>Syntax:</h5>
1856<pre> &lt;result&gt; = urem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
1857</pre>
1858<h5>Overview:</h5>
1859<p>The '<tt>urem</tt>' instruction returns the remainder from the
1860unsigned division of its two arguments.</p>
1861<h5>Arguments:</h5>
1862<p>The two arguments to the '<tt>urem</tt>' instruction must be
1863<a href="#t_integer">integer</a> values. Both arguments must have identical
1864types.</p>
1865<h5>Semantics:</h5>
1866<p>This instruction returns the unsigned integer <i>remainder</i> of a division.
1867This instruction always performs an unsigned division to get the remainder,
1868regardless of whether the arguments are unsigned or not.</p>
1869<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1870<pre> &lt;result&gt; = urem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]1871</pre>
1872
1873</div>
1874<!-- _______________________________________________________________________ -->
1875<div class="doc_subsubsection"> <a name="i_srem">'<tt>srem</tt>'
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1876Instruction</a> </div>
1877<div class="doc_text">
1878<h5>Syntax:</h5>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]1879<pre> &lt;result&gt; = srem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1880</pre>
1881<h5>Overview:</h5>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]1882<p>The '<tt>srem</tt>' instruction returns the remainder from the
1883signed division of its two operands.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1884<h5>Arguments:</h5>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]1885<p>The two arguments to the '<tt>srem</tt>' instruction must be
1886<a href="#t_integer">integer</a> values. Both arguments must have identical
1887types.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1888<h5>Semantics:</h5>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]1889<p>This instruction returns the <i>remainder</i> of a division (where the result
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1890has the same sign as the divisor), not the <i>modulus</i> (where the
1891result has the same sign as the dividend) of a value. For more
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]1892information about the difference, see <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1893 href="http://mathforum.org/dr.math/problems/anne.4.28.99.html">The
1894Math Forum</a>.</p>
1895<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1896<pre> &lt;result&gt; = srem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
Reid Spencer0a783f72006-11-02 01:53:59 +0000[diff] [blame]1897</pre>
1898
1899</div>
1900<!-- _______________________________________________________________________ -->
1901<div class="doc_subsubsection"> <a name="i_frem">'<tt>frem</tt>'
1902Instruction</a> </div>
1903<div class="doc_text">
1904<h5>Syntax:</h5>
1905<pre> &lt;result&gt; = frem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
1906</pre>
1907<h5>Overview:</h5>
1908<p>The '<tt>frem</tt>' instruction returns the remainder from the
1909division of its two operands.</p>
1910<h5>Arguments:</h5>
1911<p>The two arguments to the '<tt>frem</tt>' instruction must be
1912<a href="#t_floating">floating point</a> values. Both arguments must have
1913identical types.</p>
1914<h5>Semantics:</h5>
1915<p>This instruction returns the <i>remainder</i> of a division.</p>
1916<h5>Example:</h5>
1917<pre> &lt;result&gt; = frem float 4.0, %var <i>; yields {float}:result = 4.0 % %var</i>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1918</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1919</div>
Robert Bocchino7b81c752006-02-17 21:18:08 +0000[diff] [blame]1920
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1921<!-- ======================================================================= -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1922<div class="doc_subsection"> <a name="bitwiseops">Bitwise Binary
1923Operations</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1924<div class="doc_text">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1925<p>Bitwise binary operators are used to do various forms of
1926bit-twiddling in a program. They are generally very efficient
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]1927instructions and can commonly be strength reduced from other
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1928instructions. They require two operands, execute an operation on them,
1929and produce a single value. The resulting value of the bitwise binary
1930operators is always the same type as its first operand.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1931</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1932<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1933<div class="doc_subsubsection"> <a name="i_and">'<tt>and</tt>'
1934Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1935<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1936<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1937<pre> &lt;result&gt; = and &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1938</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1939<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1940<p>The '<tt>and</tt>' instruction returns the bitwise logical and of
1941its two operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1942<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1943<p>The two arguments to the '<tt>and</tt>' instruction must be <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1944 href="#t_integral">integral</a> values. Both arguments must have
1945identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1946<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1947<p>The truth table used for the '<tt>and</tt>' instruction is:</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1948<p> </p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1949<div style="align: center">
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1950<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1951 <tbody>
1952 <tr>
1953 <td>In0</td>
1954 <td>In1</td>
1955 <td>Out</td>
1956 </tr>
1957 <tr>
1958 <td>0</td>
1959 <td>0</td>
1960 <td>0</td>
1961 </tr>
1962 <tr>
1963 <td>0</td>
1964 <td>1</td>
1965 <td>0</td>
1966 </tr>
1967 <tr>
1968 <td>1</td>
1969 <td>0</td>
1970 <td>0</td>
1971 </tr>
1972 <tr>
1973 <td>1</td>
1974 <td>1</td>
1975 <td>1</td>
1976 </tr>
1977 </tbody>
1978</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]1979</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1980<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]1981<pre> &lt;result&gt; = and i32 4, %var <i>; yields {i32}:result = 4 &amp; %var</i>
1982 &lt;result&gt; = and i32 15, 40 <i>; yields {i32}:result = 8</i>
1983 &lt;result&gt; = and i32 4, 8 <i>; yields {i32}:result = 0</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1984</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1985</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1986<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1987<div class="doc_subsubsection"> <a name="i_or">'<tt>or</tt>' Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1988<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1989<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1990<pre> &lt;result&gt; = or &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1991</pre>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1992<h5>Overview:</h5>
1993<p>The '<tt>or</tt>' instruction returns the bitwise logical inclusive
1994or of its two operands.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1995<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]1996<p>The two arguments to the '<tt>or</tt>' instruction must be <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]1997 href="#t_integral">integral</a> values. Both arguments must have
1998identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]1999<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2000<p>The truth table used for the '<tt>or</tt>' instruction is:</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2001<p> </p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]2002<div style="align: center">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2003<table border="1" cellspacing="0" cellpadding="4">
2004 <tbody>
2005 <tr>
2006 <td>In0</td>
2007 <td>In1</td>
2008 <td>Out</td>
2009 </tr>
2010 <tr>
2011 <td>0</td>
2012 <td>0</td>
2013 <td>0</td>
2014 </tr>
2015 <tr>
2016 <td>0</td>
2017 <td>1</td>
2018 <td>1</td>
2019 </tr>
2020 <tr>
2021 <td>1</td>
2022 <td>0</td>
2023 <td>1</td>
2024 </tr>
2025 <tr>
2026 <td>1</td>
2027 <td>1</td>
2028 <td>1</td>
2029 </tr>
2030 </tbody>
2031</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]2032</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2033<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2034<pre> &lt;result&gt; = or i32 4, %var <i>; yields {i32}:result = 4 | %var</i>
2035 &lt;result&gt; = or i32 15, 40 <i>; yields {i32}:result = 47</i>
2036 &lt;result&gt; = or i32 4, 8 <i>; yields {i32}:result = 12</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2037</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2038</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2039<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2040<div class="doc_subsubsection"> <a name="i_xor">'<tt>xor</tt>'
2041Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2042<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2043<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2044<pre> &lt;result&gt; = xor &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2045</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2046<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2047<p>The '<tt>xor</tt>' instruction returns the bitwise logical exclusive
2048or of its two operands. The <tt>xor</tt> is used to implement the
2049"one's complement" operation, which is the "~" operator in C.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2050<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2051<p>The two arguments to the '<tt>xor</tt>' instruction must be <a
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2052 href="#t_integral">integral</a> values. Both arguments must have
2053identical types.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2054<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2055<p>The truth table used for the '<tt>xor</tt>' instruction is:</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2056<p> </p>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]2057<div style="align: center">
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2058<table border="1" cellspacing="0" cellpadding="4">
2059 <tbody>
2060 <tr>
2061 <td>In0</td>
2062 <td>In1</td>
2063 <td>Out</td>
2064 </tr>
2065 <tr>
2066 <td>0</td>
2067 <td>0</td>
2068 <td>0</td>
2069 </tr>
2070 <tr>
2071 <td>0</td>
2072 <td>1</td>
2073 <td>1</td>
2074 </tr>
2075 <tr>
2076 <td>1</td>
2077 <td>0</td>
2078 <td>1</td>
2079 </tr>
2080 <tr>
2081 <td>1</td>
2082 <td>1</td>
2083 <td>0</td>
2084 </tr>
2085 </tbody>
2086</table>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]2087</div>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2088<p> </p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2089<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2090<pre> &lt;result&gt; = xor i32 4, %var <i>; yields {i32}:result = 4 ^ %var</i>
2091 &lt;result&gt; = xor i32 15, 40 <i>; yields {i32}:result = 39</i>
2092 &lt;result&gt; = xor i32 4, 8 <i>; yields {i32}:result = 12</i>
2093 &lt;result&gt; = xor i32 %V, -1 <i>; yields {i32}:result = ~%V</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2094</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2095</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2096<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2097<div class="doc_subsubsection"> <a name="i_shl">'<tt>shl</tt>'
2098Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2099<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2100<h5>Syntax:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2101<pre> &lt;result&gt; = shl &lt;ty&gt; &lt;var1&gt;, i8 &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2102</pre>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2103<h5>Overview:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2104<p>The '<tt>shl</tt>' instruction returns the first operand shifted to
2105the left a specified number of bits.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2106<h5>Arguments:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2107<p>The first argument to the '<tt>shl</tt>' instruction must be an <a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2108 href="#t_integer">integer</a> type. The second argument must be an '<tt>i8</tt>'
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2109type.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2110<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2111<p>The value produced is <tt>var1</tt> * 2<sup><tt>var2</tt></sup>.</p>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2112<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2113<pre> &lt;result&gt; = shl i32 4, i8 %var <i>; yields {i32}:result = 4 &lt;&lt; %var</i>
2114 &lt;result&gt; = shl i32 4, i8 2 <i>; yields {i32}:result = 16</i>
2115 &lt;result&gt; = shl i32 1, i8 10 <i>; yields {i32}:result = 1024</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2116</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2117</div>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2118<!-- _______________________________________________________________________ -->
Reid Spencer3822ff52006-11-08 06:47:33 +0000[diff] [blame]2119<div class="doc_subsubsection"> <a name="i_lshr">'<tt>lshr</tt>'
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2120Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2121<div class="doc_text">
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2122<h5>Syntax:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2123<pre> &lt;result&gt; = lshr &lt;ty&gt; &lt;var1&gt;, i8 &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2124</pre>
Reid Spencer3822ff52006-11-08 06:47:33 +0000[diff] [blame]2125
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2126<h5>Overview:</h5>
Reid Spencer3822ff52006-11-08 06:47:33 +0000[diff] [blame]2127<p>The '<tt>lshr</tt>' instruction (logical shift right) returns the first
2128operand shifted to the right a specified number of bits.</p>
2129
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2130<h5>Arguments:</h5>
Reid Spencer3822ff52006-11-08 06:47:33 +0000[diff] [blame]2131<p>The first argument to the '<tt>lshr</tt>' instruction must be an <a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2132 href="#t_integer">integer</a> type. The second argument must be an '<tt>i8</tt>' type.</p>
Reid Spencer3822ff52006-11-08 06:47:33 +0000[diff] [blame]2133
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2134<h5>Semantics:</h5>
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]2135<p>This instruction always performs a logical shift right operation. The
2136<tt>var2</tt> most significant bits will be filled with zero bits after the
2137shift.</p>
Reid Spencer3822ff52006-11-08 06:47:33 +0000[diff] [blame]2138
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2139<h5>Example:</h5>
Reid Spencer3822ff52006-11-08 06:47:33 +0000[diff] [blame]2140<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2141 &lt;result&gt; = lshr i32 4, i8 1 <i>; yields {i32}:result = 2</i>
2142 &lt;result&gt; = lshr i32 4, i8 2 <i>; yields {i32}:result = 1</i>
2143 &lt;result&gt; = lshr i8 4, i8 3 <i>; yields {i8 }:result = 0</i>
2144 &lt;result&gt; = lshr i8 -2, i8 1 <i>; yields {i8 }:result = 0x7FFFFFFF </i>
Reid Spencer3822ff52006-11-08 06:47:33 +0000[diff] [blame]2145</pre>
2146</div>
2147
2148<!-- ======================================================================= -->
2149<div class="doc_subsubsection"> <a name="i_ashr">'<tt>ashr</tt>'
2150Instruction</a> </div>
2151<div class="doc_text">
2152
2153<h5>Syntax:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2154<pre> &lt;result&gt; = ashr &lt;ty&gt; &lt;var1&gt;, i8 &lt;var2&gt; <i>; yields {ty}:result</i>
Reid Spencer3822ff52006-11-08 06:47:33 +0000[diff] [blame]2155</pre>
2156
2157<h5>Overview:</h5>
2158<p>The '<tt>ashr</tt>' instruction (arithmetic shift right) returns the first
2159operand shifted to the right a specified number of bits.</p>
2160
2161<h5>Arguments:</h5>
2162<p>The first argument to the '<tt>ashr</tt>' instruction must be an
2163<a href="#t_integer">integer</a> type. The second argument must be an
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2164'<tt>i8</tt>' type.</p>
Reid Spencer3822ff52006-11-08 06:47:33 +0000[diff] [blame]2165
2166<h5>Semantics:</h5>
2167<p>This instruction always performs an arithmetic shift right operation,
2168regardless of whether the arguments are signed or not. The <tt>var2</tt> most
2169significant bits will be filled with the sign bit of <tt>var1</tt>.</p>
2170
2171<h5>Example:</h5>
2172<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2173 &lt;result&gt; = ashr i32 4, i8 1 <i>; yields {i32}:result = 2</i>
2174 &lt;result&gt; = ashr i32 4, i8 2 <i>; yields {i32}:result = 1</i>
2175 &lt;result&gt; = ashr i8 4, i8 3 <i>; yields {i8}:result = 0</i>
2176 &lt;result&gt; = ashr i8 -2, i8 1 <i>; yields {i8 }:result = -1</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2177</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2178</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2179
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2180<!-- ======================================================================= -->
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2181<div class="doc_subsection">
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2182 <a name="vectorops">Vector Operations</a>
2183</div>
2184
2185<div class="doc_text">
2186
2187<p>LLVM supports several instructions to represent vector operations in a
2188target-independent manner. This instructions cover the element-access and
2189vector-specific operations needed to process vectors effectively. While LLVM
2190does directly support these vector operations, many sophisticated algorithms
2191will want to use target-specific intrinsics to take full advantage of a specific
2192target.</p>
2193
2194</div>
2195
2196<!-- _______________________________________________________________________ -->
2197<div class="doc_subsubsection">
2198 <a name="i_extractelement">'<tt>extractelement</tt>' Instruction</a>
2199</div>
2200
2201<div class="doc_text">
2202
2203<h5>Syntax:</h5>
2204
2205<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2206 &lt;result&gt; = extractelement &lt;n x &lt;ty&gt;&gt; &lt;val&gt;, i32 &lt;idx&gt; <i>; yields &lt;ty&gt;</i>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2207</pre>
2208
2209<h5>Overview:</h5>
2210
2211<p>
2212The '<tt>extractelement</tt>' instruction extracts a single scalar
2213element from a packed vector at a specified index.
2214</p>
2215
2216
2217<h5>Arguments:</h5>
2218
2219<p>
2220The first operand of an '<tt>extractelement</tt>' instruction is a
2221value of <a href="#t_packed">packed</a> type. The second operand is
2222an index indicating the position from which to extract the element.
2223The index may be a variable.</p>
2224
2225<h5>Semantics:</h5>
2226
2227<p>
2228The result is a scalar of the same type as the element type of
2229<tt>val</tt>. Its value is the value at position <tt>idx</tt> of
2230<tt>val</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
2231results are undefined.
2232</p>
2233
2234<h5>Example:</h5>
2235
2236<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2237 %result = extractelement &lt;4 x i32&gt; %vec, i32 0 <i>; yields i32</i>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2238</pre>
2239</div>
2240
2241
2242<!-- _______________________________________________________________________ -->
2243<div class="doc_subsubsection">
2244 <a name="i_insertelement">'<tt>insertelement</tt>' Instruction</a>
2245</div>
2246
2247<div class="doc_text">
2248
2249<h5>Syntax:</h5>
2250
2251<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2252 &lt;result&gt; = insertelement &lt;n x &lt;ty&gt;&gt; &lt;val&gt;, &lt;ty&gt; &lt;elt&gt, i32 &lt;idx&gt; <i>; yields &lt;n x &lt;ty&gt;&gt;</i>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2253</pre>
2254
2255<h5>Overview:</h5>
2256
2257<p>
2258The '<tt>insertelement</tt>' instruction inserts a scalar
2259element into a packed vector at a specified index.
2260</p>
2261
2262
2263<h5>Arguments:</h5>
2264
2265<p>
2266The first operand of an '<tt>insertelement</tt>' instruction is a
2267value of <a href="#t_packed">packed</a> type. The second operand is a
2268scalar value whose type must equal the element type of the first
2269operand. The third operand is an index indicating the position at
2270which to insert the value. The index may be a variable.</p>
2271
2272<h5>Semantics:</h5>
2273
2274<p>
2275The result is a packed vector of the same type as <tt>val</tt>. Its
2276element values are those of <tt>val</tt> except at position
2277<tt>idx</tt>, where it gets the value <tt>elt</tt>. If <tt>idx</tt>
2278exceeds the length of <tt>val</tt>, the results are undefined.
2279</p>
2280
2281<h5>Example:</h5>
2282
2283<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2284 %result = insertelement &lt;4 x i32&gt; %vec, i32 1, i32 0 <i>; yields &lt;4 x i32&gt;</i>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2285</pre>
2286</div>
2287
2288<!-- _______________________________________________________________________ -->
2289<div class="doc_subsubsection">
2290 <a name="i_shufflevector">'<tt>shufflevector</tt>' Instruction</a>
2291</div>
2292
2293<div class="doc_text">
2294
2295<h5>Syntax:</h5>
2296
2297<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2298 &lt;result&gt; = shufflevector &lt;n x &lt;ty&gt;&gt; &lt;v1&gt;, &lt;n x &lt;ty&gt;&gt; &lt;v2&gt;, &lt;n x i32&gt; &lt;mask&gt; <i>; yields &lt;n x &lt;ty&gt;&gt;</i>
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2299</pre>
2300
2301<h5>Overview:</h5>
2302
2303<p>
2304The '<tt>shufflevector</tt>' instruction constructs a permutation of elements
2305from two input vectors, returning a vector of the same type.
2306</p>
2307
2308<h5>Arguments:</h5>
2309
2310<p>
2311The first two operands of a '<tt>shufflevector</tt>' instruction are vectors
2312with types that match each other and types that match the result of the
2313instruction. The third argument is a shuffle mask, which has the same number
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2314of elements as the other vector type, but whose element type is always 'i32'.
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2315</p>
2316
2317<p>
2318The shuffle mask operand is required to be a constant vector with either
2319constant integer or undef values.
2320</p>
2321
2322<h5>Semantics:</h5>
2323
2324<p>
2325The elements of the two input vectors are numbered from left to right across
2326both of the vectors. The shuffle mask operand specifies, for each element of
2327the result vector, which element of the two input registers the result element
2328gets. The element selector may be undef (meaning "don't care") and the second
2329operand may be undef if performing a shuffle from only one vector.
2330</p>
2331
2332<h5>Example:</h5>
2333
2334<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2335 %result = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; %v2,
2336 &lt;4 x i32&gt; &lt;i32 0, i32 4, i32 1, i32 5&gt; <i>; yields &lt;4 x i32&gt;</i>
2337 %result = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; undef,
2338 &lt;4 x i32&gt; &lt;i32 0, i32 1, i32 2, i32 3&gt; <i>; yields &lt;4 x i32&gt;</i> - Identity shuffle.
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2339</pre>
2340</div>
2341
Tanya Lattner09474292006-04-14 19:24:33 +0000[diff] [blame]2342
Chris Lattner3df241e2006-04-08 23:07:04 +0000[diff] [blame]2343<!-- ======================================================================= -->
2344<div class="doc_subsection">
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]2345 <a name="memoryops">Memory Access and Addressing Operations</a>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2346</div>
2347
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2348<div class="doc_text">
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2349
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2350<p>A key design point of an SSA-based representation is how it
2351represents memory. In LLVM, no memory locations are in SSA form, which
2352makes things very simple. This section describes how to read, write,
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]2353allocate, and free memory in LLVM.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2354
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2355</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2356
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2357<!-- _______________________________________________________________________ -->
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2358<div class="doc_subsubsection">
2359 <a name="i_malloc">'<tt>malloc</tt>' Instruction</a>
2360</div>
2361
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2362<div class="doc_text">
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2363
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2364<h5>Syntax:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2365
2366<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2367 &lt;result&gt; = malloc &lt;type&gt;[, i32 &lt;NumElements&gt;][, align &lt;alignment&gt;] <i>; yields {type*}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2368</pre>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2369
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2370<h5>Overview:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2371
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2372<p>The '<tt>malloc</tt>' instruction allocates memory from the system
2373heap and returns a pointer to it.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2374
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2375<h5>Arguments:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2376
2377<p>The '<tt>malloc</tt>' instruction allocates
2378<tt>sizeof(&lt;type&gt;)*NumElements</tt>
John Criswell6e4ca612004-02-24 16:13:56 +0000[diff] [blame]2379bytes of memory from the operating system and returns a pointer of the
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2380appropriate type to the program. If "NumElements" is specified, it is the
2381number of elements allocated. If an alignment is specified, the value result
2382of the allocation is guaranteed to be aligned to at least that boundary. If
2383not specified, or if zero, the target can choose to align the allocation on any
2384convenient boundary.</p>
2385
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2386<p>'<tt>type</tt>' must be a sized type.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2387
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2388<h5>Semantics:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2389
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2390<p>Memory is allocated using the system "<tt>malloc</tt>" function, and
2391a pointer is returned.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2392
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2393<h5>Example:</h5>
2394
2395<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2396 %array = malloc [4 x i8 ] <i>; yields {[%4 x i8]*}:array</i>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2397
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2398 %size = <a href="#i_add">add</a> i32 2, 2 <i>; yields {i32}:size = i32 4</i>
2399 %array1 = malloc i8, i32 4 <i>; yields {i8*}:array1</i>
2400 %array2 = malloc [12 x i8], i32 %size <i>; yields {[12 x i8]*}:array2</i>
2401 %array3 = malloc i32, i32 4, align 1024 <i>; yields {i32*}:array3</i>
2402 %array4 = malloc i32, align 1024 <i>; yields {i32*}:array4</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2403</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2404</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2405
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2406<!-- _______________________________________________________________________ -->
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2407<div class="doc_subsubsection">
2408 <a name="i_free">'<tt>free</tt>' Instruction</a>
2409</div>
2410
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2411<div class="doc_text">
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2412
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2413<h5>Syntax:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2414
2415<pre>
2416 free &lt;type&gt; &lt;value&gt; <i>; yields {void}</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2417</pre>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2418
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2419<h5>Overview:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2420
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2421<p>The '<tt>free</tt>' instruction returns memory back to the unused
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]2422memory heap to be reallocated in the future.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2423
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2424<h5>Arguments:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2425
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2426<p>'<tt>value</tt>' shall be a pointer value that points to a value
2427that was allocated with the '<tt><a href="#i_malloc">malloc</a></tt>'
2428instruction.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2429
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2430<h5>Semantics:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2431
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]2432<p>Access to the memory pointed to by the pointer is no longer defined
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2433after this instruction executes.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2434
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2435<h5>Example:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2436
2437<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2438 %array = <a href="#i_malloc">malloc</a> [4 x i8] <i>; yields {[4 x i8]*}:array</i>
2439 free [4 x i8]* %array
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2440</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2441</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2442
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2443<!-- _______________________________________________________________________ -->
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2444<div class="doc_subsubsection">
2445 <a name="i_alloca">'<tt>alloca</tt>' Instruction</a>
2446</div>
2447
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2448<div class="doc_text">
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2449
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2450<h5>Syntax:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2451
2452<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2453 &lt;result&gt; = alloca &lt;type&gt;[, i32 &lt;NumElements&gt;][, align &lt;alignment&gt;] <i>; yields {type*}:result</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2454</pre>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2455
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2456<h5>Overview:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2457
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2458<p>The '<tt>alloca</tt>' instruction allocates memory on the current
2459stack frame of the procedure that is live until the current function
2460returns to its caller.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2461
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2462<h5>Arguments:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2463
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]2464<p>The '<tt>alloca</tt>' instruction allocates <tt>sizeof(&lt;type&gt;)*NumElements</tt>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2465bytes of memory on the runtime stack, returning a pointer of the
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2466appropriate type to the program. If "NumElements" is specified, it is the
2467number of elements allocated. If an alignment is specified, the value result
2468of the allocation is guaranteed to be aligned to at least that boundary. If
2469not specified, or if zero, the target can choose to align the allocation on any
2470convenient boundary.</p>
2471
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2472<p>'<tt>type</tt>' may be any sized type.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2473
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2474<h5>Semantics:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2475
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]2476<p>Memory is allocated; a pointer is returned. '<tt>alloca</tt>'d
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2477memory is automatically released when the function returns. The '<tt>alloca</tt>'
2478instruction is commonly used to represent automatic variables that must
2479have an address available. When the function returns (either with the <tt><a
John Criswelldae2e932005-05-12 16:55:34 +0000[diff] [blame]2480 href="#i_ret">ret</a></tt> or <tt><a href="#i_unwind">unwind</a></tt>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2481instructions), the memory is reclaimed.</p>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2482
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2483<h5>Example:</h5>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2484
2485<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2486 %ptr = alloca i32 <i>; yields {i32*}:ptr</i>
2487 %ptr = alloca i32, i32 4 <i>; yields {i32*}:ptr</i>
2488 %ptr = alloca i32, i32 4, align 1024 <i>; yields {i32*}:ptr</i>
2489 %ptr = alloca i32, align 1024 <i>; yields {i32*}:ptr</i>
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2490</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2491</div>
Chris Lattner2cbdc452005-11-06 08:02:57 +0000[diff] [blame]2492
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2493<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2494<div class="doc_subsubsection"> <a name="i_load">'<tt>load</tt>'
2495Instruction</a> </div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2496<div class="doc_text">
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2497<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2498<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 Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2499<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2500<p>The '<tt>load</tt>' instruction is used to read from memory.</p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2501<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2502<p>The argument to the '<tt>load</tt>' instruction specifies the memory
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]2503address from which to load. The pointer must point to a <a
Chris Lattnere53e5082004-06-03 22:57:15 +0000[diff] [blame]2504 href="#t_firstclass">first class</a> type. If the <tt>load</tt> is
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]2505marked as <tt>volatile</tt>, then the optimizer is not allowed to modify
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2506the number or order of execution of this <tt>load</tt> with other
2507volatile <tt>load</tt> and <tt><a href="#i_store">store</a></tt>
2508instructions. </p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2509<h5>Semantics:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2510<p>The location of memory pointed to is loaded.</p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2511<h5>Examples:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2512<pre> %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2513 <a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2514 href="#i_store">store</a> i32 3, i32* %ptr <i>; yields {void}</i>
2515 %val = load i32* %ptr <i>; yields {i32}:val = i32 3</i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2516</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2517</div>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2518<!-- _______________________________________________________________________ -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2519<div class="doc_subsubsection"> <a name="i_store">'<tt>store</tt>'
2520Instruction</a> </div>
Reid Spencer035ab572006-11-09 21:18:01 +0000[diff] [blame]2521<div class="doc_text">
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2522<h5>Syntax:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2523<pre> store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt; <i>; yields {void}</i>
Chris Lattnerf0651072003-09-08 18:27:49 +0000[diff] [blame]2524 volatile store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt; <i>; yields {void}</i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2525</pre>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2526<h5>Overview:</h5>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2527<p>The '<tt>store</tt>' instruction is used to write to memory.</p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2528<h5>Arguments:</h5>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2529<p>There are two arguments to the '<tt>store</tt>' instruction: a value
John Criswell0ec250c2005-10-24 16:17:18 +0000[diff] [blame]2530to store and an address in which to store it. The type of the '<tt>&lt;pointer&gt;</tt>'
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2531operand must be a pointer to the type of the '<tt>&lt;value&gt;</tt>'
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]2532operand. If the <tt>store</tt> is marked as <tt>volatile</tt>, then the
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2533optimizer is not allowed to modify the number or order of execution of
2534this <tt>store</tt> with other volatile <tt>load</tt> and <tt><a
2535 href="#i_store">store</a></tt> instructions.</p>
2536<h5>Semantics:</h5>
2537<p>The contents of memory are updated to contain '<tt>&lt;value&gt;</tt>'
2538at the location specified by the '<tt>&lt;pointer&gt;</tt>' operand.</p>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2539<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2540<pre> %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2541 <a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2542 href="#i_store">store</a> i32 3, i32* %ptr <i>; yields {void}</i>
2543 %val = load i32* %ptr <i>; yields {i32}:val = i32 3</i>
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2544</pre>
Reid Spencer47ce1792006-11-09 21:15:49 +0000[diff] [blame]2545</div>
2546
Chris Lattner2b7d3202002-05-06 03:03:22 +0000[diff] [blame]2547<!-- _______________________________________________________________________ -->
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2548<div class="doc_subsubsection">
2549 <a name="i_getelementptr">'<tt>getelementptr</tt>' Instruction</a>
2550</div>
2551
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2552<div class="doc_text">
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2553<h5>Syntax:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2554<pre>
2555 &lt;result&gt; = getelementptr &lt;ty&gt;* &lt;ptrval&gt;{, &lt;ty&gt; &lt;idx&gt;}*
2556</pre>
2557
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2558<h5>Overview:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2559
2560<p>
2561The '<tt>getelementptr</tt>' instruction is used to get the address of a
2562subelement of an aggregate data structure.</p>
2563
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2564<h5>Arguments:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2565
Reid Spencer85f5b5b2006-12-04 21:29:24 +0000[diff] [blame]2566<p>This instruction takes a list of integer operands that indicate what
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2567elements of the aggregate object to index to. The actual types of the arguments
2568provided depend on the type of the first pointer argument. The
2569'<tt>getelementptr</tt>' instruction is used to index down through the type
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]2570levels of a structure or to a specific index in an array. When indexing into a
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2571structure, only <tt>i32</tt> integer constants are allowed. When indexing
Reid Spencer85f5b5b2006-12-04 21:29:24 +0000[diff] [blame]2572into an array or pointer, only integers of 32 or 64 bits are allowed, and will
2573be sign extended to 64-bit values.</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2574
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2575<p>For example, let's consider a C code fragment and how it gets
2576compiled to LLVM:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2577
2578<pre>
2579 struct RT {
2580 char A;
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2581 i32 B[10][20];
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2582 char C;
2583 };
2584 struct ST {
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2585 i32 X;
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2586 double Y;
2587 struct RT Z;
2588 };
2589
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2590 define i32 *foo(struct ST *s) {
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2591 return &amp;s[1].Z.B[5][13];
2592 }
2593</pre>
2594
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2595<p>The LLVM code generated by the GCC frontend is:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2596
2597<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2598 %RT = type { i8 , [10 x [20 x i32]], i8 }
2599 %ST = type { i32, double, %RT }
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2600
Brian Gaeke7283e7c2004-07-02 21:08:14 +0000[diff] [blame]2601 implementation
2602
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2603 define i32* %foo(%ST* %s) {
Brian Gaeke7283e7c2004-07-02 21:08:14 +0000[diff] [blame]2604 entry:
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2605 %reg = getelementptr %ST* %s, i32 1, i32 2, i32 1, i32 5, i32 13
2606 ret i32* %reg
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2607 }
2608</pre>
2609
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2610<h5>Semantics:</h5>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2611
2612<p>The index types specified for the '<tt>getelementptr</tt>' instruction depend
John Criswellc1f786c2005-05-13 22:25:59 +0000[diff] [blame]2613on the pointer type that is being indexed into. <a href="#t_pointer">Pointer</a>
Reid Spencer85f5b5b2006-12-04 21:29:24 +0000[diff] [blame]2614and <a href="#t_array">array</a> types can use a 32-bit or 64-bit
Reid Spencer42ddd842006-12-03 16:53:48 +0000[diff] [blame]2615<a href="#t_integer">integer</a> type but the value will always be sign extended
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2616to 64-bits. <a href="#t_struct">Structure</a> types, require <tt>i32</tt>
Reid Spencer42ddd842006-12-03 16:53:48 +0000[diff] [blame]2617<b>constants</b>.</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2618
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2619<p>In the example above, the first index is indexing into the '<tt>%ST*</tt>'
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2620type, which is a pointer, yielding a '<tt>%ST</tt>' = '<tt>{ i32, double, %RT
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2621}</tt>' type, a structure. The second index indexes into the third element of
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2622the structure, yielding a '<tt>%RT</tt>' = '<tt>{ i8 , [10 x [20 x i32]],
2623i8 }</tt>' type, another structure. The third index indexes into the second
2624element of the structure, yielding a '<tt>[10 x [20 x i32]]</tt>' type, an
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2625array. The two dimensions of the array are subscripted into, yielding an
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2626'<tt>i32</tt>' type. The '<tt>getelementptr</tt>' instruction returns a pointer
2627to this element, thus computing a value of '<tt>i32*</tt>' type.</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2628
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]2629<p>Note that it is perfectly legal to index partially through a
2630structure, returning a pointer to an inner element. Because of this,
2631the LLVM code for the given testcase is equivalent to:</p>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2632
2633<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2634 define i32* %foo(%ST* %s) {
2635 %t1 = getelementptr %ST* %s, i32 1 <i>; yields %ST*:%t1</i>
2636 %t2 = getelementptr %ST* %t1, i32 0, i32 2 <i>; yields %RT*:%t2</i>
2637 %t3 = getelementptr %RT* %t2, i32 0, i32 1 <i>; yields [10 x [20 x i32]]*:%t3</i>
2638 %t4 = getelementptr [10 x [20 x i32]]* %t3, i32 0, i32 5 <i>; yields [20 x i32]*:%t4</i>
2639 %t5 = getelementptr [20 x i32]* %t4, i32 0, i32 13 <i>; yields i32*:%t5</i>
2640 ret i32* %t5
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2641 }
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]2642</pre>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]2643
2644<p>Note that it is undefined to access an array out of bounds: array and
2645pointer indexes must always be within the defined bounds of the array type.
2646The one exception for this rules is zero length arrays. These arrays are
2647defined to be accessible as variable length arrays, which requires access
2648beyond the zero'th element.</p>
2649
Chris Lattner884a9702006-08-15 00:45:58 +0000[diff] [blame]2650<p>The getelementptr instruction is often confusing. For some more insight
2651into how it works, see <a href="GetElementPtr.html">the getelementptr
2652FAQ</a>.</p>
2653
Chris Lattner7faa8832002-04-14 06:13:44 +0000[diff] [blame]2654<h5>Example:</h5>
Chris Lattnere67a9512005-06-24 17:22:57 +0000[diff] [blame]2655
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2656<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2657 <i>; yields [12 x i8]*:aptr</i>
2658 %aptr = getelementptr {i32, [12 x i8]}* %sptr, i64 0, i32 1
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2659</pre>
Chris Lattnerf74d5c72004-04-05 01:30:49 +0000[diff] [blame]2660</div>
Reid Spencer47ce1792006-11-09 21:15:49 +0000[diff] [blame]2661
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]2662<!-- ======================================================================= -->
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]2663<div class="doc_subsection"> <a name="convertops">Conversion Operations</a>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2664</div>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2665<div class="doc_text">
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]2666<p>The instructions in this category are the conversion instructions (casting)
2667which all take a single operand and a type. They perform various bit conversions
2668on the operand.</p>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]2669</div>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]2670
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]2671<!-- _______________________________________________________________________ -->
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]2672<div class="doc_subsubsection">
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2673 <a name="i_trunc">'<tt>trunc .. to</tt>' Instruction</a>
2674</div>
2675<div class="doc_text">
2676
2677<h5>Syntax:</h5>
2678<pre>
2679 &lt;result&gt; = trunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2680</pre>
2681
2682<h5>Overview:</h5>
2683<p>
2684The '<tt>trunc</tt>' instruction truncates its operand to the type <tt>ty2</tt>.
2685</p>
2686
2687<h5>Arguments:</h5>
2688<p>
2689The '<tt>trunc</tt>' instruction takes a <tt>value</tt> to trunc, which must
2690be an <a href="#t_integer">integer</a> type, and a type that specifies the size
2691and type of the result, which must be an <a href="#t_integral">integral</a>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2692type. The bit size of <tt>value</tt> must be larger than the bit size of
2693<tt>ty2</tt>. Equal sized types are not allowed.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2694
2695<h5>Semantics:</h5>
2696<p>
2697The '<tt>trunc</tt>' instruction truncates the high order bits in <tt>value</tt>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2698and converts the remaining bits to <tt>ty2</tt>. Since the source size must be
2699larger than the destination size, <tt>trunc</tt> cannot be a <i>no-op cast</i>.
2700It will always truncate bits.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2701
2702<h5>Example:</h5>
2703<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2704 %X = trunc i32 257 to i8 <i>; yields i8:1</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]2705 %Y = trunc i32 123 to i1 <i>; yields i1:true</i>
2706 %Y = trunc i32 122 to i1 <i>; yields i1:false</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2707</pre>
2708</div>
2709
2710<!-- _______________________________________________________________________ -->
2711<div class="doc_subsubsection">
2712 <a name="i_zext">'<tt>zext .. to</tt>' Instruction</a>
2713</div>
2714<div class="doc_text">
2715
2716<h5>Syntax:</h5>
2717<pre>
2718 &lt;result&gt; = zext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2719</pre>
2720
2721<h5>Overview:</h5>
2722<p>The '<tt>zext</tt>' instruction zero extends its operand to type
2723<tt>ty2</tt>.</p>
2724
2725
2726<h5>Arguments:</h5>
2727<p>The '<tt>zext</tt>' instruction takes a value to cast, which must be of
2728<a href="#t_integral">integral</a> type, and a type to cast it to, which must
2729also be of <a href="#t_integral">integral</a> type. The bit size of the
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2730<tt>value</tt> must be smaller than the bit size of the destination type,
2731<tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2732
2733<h5>Semantics:</h5>
2734<p>The <tt>zext</tt> fills the high order bits of the <tt>value</tt> with zero
2735bits until it reaches the size of the destination type, <tt>ty2</tt>. When the
2736the operand and the type are the same size, no bit filling is done and the
2737cast is considered a <i>no-op cast</i> because no bits change (only the type
2738changes).</p>
2739
Reid Spencerb5929522007-01-12 15:46:11 +0000[diff] [blame]2740<p>When zero extending from i1, the result will always be either 0 or 1.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2741
2742<h5>Example:</h5>
2743<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2744 %X = zext i32 257 to i64 <i>; yields i64:257</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]2745 %Y = zext i1 true to i32 <i>; yields i32:1</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2746</pre>
2747</div>
2748
2749<!-- _______________________________________________________________________ -->
2750<div class="doc_subsubsection">
2751 <a name="i_sext">'<tt>sext .. to</tt>' Instruction</a>
2752</div>
2753<div class="doc_text">
2754
2755<h5>Syntax:</h5>
2756<pre>
2757 &lt;result&gt; = sext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2758</pre>
2759
2760<h5>Overview:</h5>
2761<p>The '<tt>sext</tt>' sign extends <tt>value</tt> to the type <tt>ty2</tt>.</p>
2762
2763<h5>Arguments:</h5>
2764<p>
2765The '<tt>sext</tt>' instruction takes a value to cast, which must be of
2766<a href="#t_integral">integral</a> type, and a type to cast it to, which must
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2767also be of <a href="#t_integral">integral</a> type. The bit size of the
2768<tt>value</tt> must be smaller than the bit size of the destination type,
2769<tt>ty2</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2770
2771<h5>Semantics:</h5>
2772<p>
2773The '<tt>sext</tt>' instruction performs a sign extension by copying the sign
2774bit (highest order bit) of the <tt>value</tt> until it reaches the bit size of
2775the type <tt>ty2</tt>. When the the operand and the type are the same size,
2776no bit filling is done and the cast is considered a <i>no-op cast</i> because
2777no bits change (only the type changes).</p>
2778
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]2779<p>When sign extending from i1, the extension always results in -1 or 0.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2780
2781<h5>Example:</h5>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2782<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2783 %X = sext i8 -1 to i16 <i>; yields i16 :65535</i>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]2784 %Y = sext i1 true to i32 <i>; yields i32:-1</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2785</pre>
2786</div>
2787
2788<!-- _______________________________________________________________________ -->
2789<div class="doc_subsubsection">
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]2790 <a name="i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a>
2791</div>
2792
2793<div class="doc_text">
2794
2795<h5>Syntax:</h5>
2796
2797<pre>
2798 &lt;result&gt; = fptrunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2799</pre>
2800
2801<h5>Overview:</h5>
2802<p>The '<tt>fptrunc</tt>' instruction truncates <tt>value</tt> to type
2803<tt>ty2</tt>.</p>
2804
2805
2806<h5>Arguments:</h5>
2807<p>The '<tt>fptrunc</tt>' instruction takes a <a href="#t_floating">floating
2808 point</a> value to cast and a <a href="#t_floating">floating point</a> type to
2809cast it to. The size of <tt>value</tt> must be larger than the size of
2810<tt>ty2</tt>. This implies that <tt>fptrunc</tt> cannot be used to make a
2811<i>no-op cast</i>.</p>
2812
2813<h5>Semantics:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2814<p> The '<tt>fptrunc</tt>' instruction truncates a <tt>value</tt> from a larger
2815<a href="#t_floating">floating point</a> type to a smaller
2816<a href="#t_floating">floating point</a> type. If the value cannot fit within
2817the destination type, <tt>ty2</tt>, then the results are undefined.</p>
Reid Spencer3fa91b02006-11-09 21:48:10 +0000[diff] [blame]2818
2819<h5>Example:</h5>
2820<pre>
2821 %X = fptrunc double 123.0 to float <i>; yields float:123.0</i>
2822 %Y = fptrunc double 1.0E+300 to float <i>; yields undefined</i>
2823</pre>
2824</div>
2825
2826<!-- _______________________________________________________________________ -->
2827<div class="doc_subsubsection">
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2828 <a name="i_fpext">'<tt>fpext .. to</tt>' Instruction</a>
2829</div>
2830<div class="doc_text">
2831
2832<h5>Syntax:</h5>
2833<pre>
2834 &lt;result&gt; = fpext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2835</pre>
2836
2837<h5>Overview:</h5>
2838<p>The '<tt>fpext</tt>' extends a floating point <tt>value</tt> to a larger
2839floating point value.</p>
2840
2841<h5>Arguments:</h5>
2842<p>The '<tt>fpext</tt>' instruction takes a
2843<a href="#t_floating">floating point</a> <tt>value</tt> to cast,
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2844and a <a href="#t_floating">floating point</a> type to cast it to. The source
2845type must be smaller than the destination type.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2846
2847<h5>Semantics:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2848<p>The '<tt>fpext</tt>' instruction extends the <tt>value</tt> from a smaller
2849<a href="t_floating">floating point</a> type to a larger
2850<a href="t_floating">floating point</a> type. The <tt>fpext</tt> cannot be
2851used to make a <i>no-op cast</i> because it always changes bits. Use
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]2852<tt>bitcast</tt> to make a <i>no-op cast</i> for a floating point cast.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2853
2854<h5>Example:</h5>
2855<pre>
2856 %X = fpext float 3.1415 to double <i>; yields double:3.1415</i>
2857 %Y = fpext float 1.0 to float <i>; yields float:1.0 (no-op)</i>
2858</pre>
2859</div>
2860
2861<!-- _______________________________________________________________________ -->
2862<div class="doc_subsubsection">
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2863 <a name="i_fp2uint">'<tt>fptoui .. to</tt>' Instruction</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2864</div>
2865<div class="doc_text">
2866
2867<h5>Syntax:</h5>
2868<pre>
2869 &lt;result&gt; = fp2uint &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2870</pre>
2871
2872<h5>Overview:</h5>
2873<p>The '<tt>fp2uint</tt>' converts a floating point <tt>value</tt> to its
2874unsigned integer equivalent of type <tt>ty2</tt>.
2875</p>
2876
2877<h5>Arguments:</h5>
2878<p>The '<tt>fp2uint</tt>' instruction takes a value to cast, which must be a
2879<a href="#t_floating">floating point</a> value, and a type to cast it to, which
2880must be an <a href="#t_integral">integral</a> type.</p>
2881
2882<h5>Semantics:</h5>
2883<p> The '<tt>fp2uint</tt>' instruction converts its
2884<a href="#t_floating">floating point</a> operand into the nearest (rounding
2885towards zero) unsigned integer value. If the value cannot fit in <tt>ty2</tt>,
2886the results are undefined.</p>
2887
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]2888<p>When converting to i1, the conversion is done as a comparison against
2889zero. If the <tt>value</tt> was zero, the i1 result will be <tt>false</tt>.
2890If the <tt>value</tt> was non-zero, the i1 result will be <tt>true</tt>.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2891
2892<h5>Example:</h5>
2893<pre>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]2894 %X = fp2uint double 123.0 to i32 <i>; yields i32:123</i>
2895 %Y = fp2uint float 1.0E+300 to i1 <i>; yields i1:true</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2896 %X = fp2uint float 1.04E+17 to i8 <i>; yields undefined:1</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2897</pre>
2898</div>
2899
2900<!-- _______________________________________________________________________ -->
2901<div class="doc_subsubsection">
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2902 <a name="i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2903</div>
2904<div class="doc_text">
2905
2906<h5>Syntax:</h5>
2907<pre>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2908 &lt;result&gt; = fptosi &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2909</pre>
2910
2911<h5>Overview:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2912<p>The '<tt>fptosi</tt>' instruction converts
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2913<a href="#t_floating">floating point</a> <tt>value</tt> to type <tt>ty2</tt>.
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]2914</p>
2915
2916
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]2917<h5>Arguments:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2918<p> The '<tt>fptosi</tt>' instruction takes a value to cast, which must be a
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2919<a href="#t_floating">floating point</a> value, and a type to cast it to, which
2920must also be an <a href="#t_integral">integral</a> type.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]2921
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]2922<h5>Semantics:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2923<p>The '<tt>fptosi</tt>' instruction converts its
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2924<a href="#t_floating">floating point</a> operand into the nearest (rounding
2925towards zero) signed integer value. If the value cannot fit in <tt>ty2</tt>,
2926the results are undefined.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]2927
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]2928<p>When converting to i1, the conversion is done as a comparison against
2929zero. If the <tt>value</tt> was zero, the i1 result will be <tt>false</tt>.
2930If the <tt>value</tt> was non-zero, the i1 result will be <tt>true</tt>.</p>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]2931
Chris Lattner33ba0d92001-07-09 00:26:23 +0000[diff] [blame]2932<h5>Example:</h5>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]2933<pre>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]2934 %X = fptosi double -123.0 to i32 <i>; yields i32:-123</i>
2935 %Y = fptosi float 1.0E-247 to i1 <i>; yields i1:true</i>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2936 %X = fptosi float 1.04E+17 to i8 <i>; yields undefined:1</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2937</pre>
2938</div>
2939
2940<!-- _______________________________________________________________________ -->
2941<div class="doc_subsubsection">
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2942 <a name="i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2943</div>
2944<div class="doc_text">
2945
2946<h5>Syntax:</h5>
2947<pre>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2948 &lt;result&gt; = uitofp &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2949</pre>
2950
2951<h5>Overview:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2952<p>The '<tt>uitofp</tt>' instruction regards <tt>value</tt> as an unsigned
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2953integer and converts that value to the <tt>ty2</tt> type.</p>
2954
2955
2956<h5>Arguments:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2957<p>The '<tt>uitofp</tt>' instruction takes a value to cast, which must be an
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2958<a href="#t_integral">integral</a> value, and a type to cast it to, which must
2959be a <a href="#t_floating">floating point</a> type.</p>
2960
2961<h5>Semantics:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2962<p>The '<tt>uitofp</tt>' instruction interprets its operand as an unsigned
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2963integer quantity and converts it to the corresponding floating point value. If
2964the value cannot fit in the floating point value, the results are undefined.</p>
2965
2966
2967<h5>Example:</h5>
2968<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]2969 %X = uitofp i32 257 to float <i>; yields float:257.0</i>
2970 %Y = uitofp i8 -1 to double <i>; yields double:255.0</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2971</pre>
2972</div>
2973
2974<!-- _______________________________________________________________________ -->
2975<div class="doc_subsubsection">
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2976 <a name="i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2977</div>
2978<div class="doc_text">
2979
2980<h5>Syntax:</h5>
2981<pre>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2982 &lt;result&gt; = sitofp &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2983</pre>
2984
2985<h5>Overview:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2986<p>The '<tt>sitofp</tt>' instruction regards <tt>value</tt> as a signed
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2987integer and converts that value to the <tt>ty2</tt> type.</p>
2988
2989<h5>Arguments:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2990<p>The '<tt>sitofp</tt>' instruction takes a value to cast, which must be an
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2991<a href="#t_integral">integral</a> value, and a type to cast it to, which must be
2992a <a href="#t_floating">floating point</a> type.</p>
2993
2994<h5>Semantics:</h5>
Reid Spencerd4448792006-11-09 23:03:26 +0000[diff] [blame]2995<p>The '<tt>sitofp</tt>' instruction interprets its operand as a signed
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]2996integer quantity and converts it to the corresponding floating point value. If
2997the value cannot fit in the floating point value, the results are undefined.</p>
2998
2999<h5>Example:</h5>
3000<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3001 %X = sitofp i32 257 to float <i>; yields float:257.0</i>
3002 %Y = sitofp i8 -1 to double <i>; yields double:-1.0</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3003</pre>
3004</div>
3005
3006<!-- _______________________________________________________________________ -->
3007<div class="doc_subsubsection">
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]3008 <a name="i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a>
3009</div>
3010<div class="doc_text">
3011
3012<h5>Syntax:</h5>
3013<pre>
3014 &lt;result&gt; = ptrtoint &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3015</pre>
3016
3017<h5>Overview:</h5>
3018<p>The '<tt>ptrtoint</tt>' instruction converts the pointer <tt>value</tt> to
3019the integer type <tt>ty2</tt>.</p>
3020
3021<h5>Arguments:</h5>
3022<p>The '<tt>ptrtoint</tt>' instruction takes a <tt>value</tt> to cast, which
3023must be a <a href="t_pointer">pointer</a> value, and a type to cast it to
3024<tt>ty2</tt>, which must be an <a href="#t_integer">integer</a> type.
3025
3026<h5>Semantics:</h5>
3027<p>The '<tt>ptrtoint</tt>' instruction converts <tt>value</tt> to integer type
3028<tt>ty2</tt> by interpreting the pointer value as an integer and either
3029truncating or zero extending that value to the size of the integer type. If
3030<tt>value</tt> is smaller than <tt>ty2</tt> then a zero extension is done. If
3031<tt>value</tt> is larger than <tt>ty2</tt> then a truncation is done. If they
3032are the same size, then nothing is done (<i>no-op cast</i>).</p>
3033
3034<h5>Example:</h5>
3035<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3036 %X = ptrtoint i32* %X to i8 <i>; yields truncation on 32-bit</i>
3037 %Y = ptrtoint i32* %x to i64 <i>; yields zero extend on 32-bit</i>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]3038</pre>
3039</div>
3040
3041<!-- _______________________________________________________________________ -->
3042<div class="doc_subsubsection">
3043 <a name="i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a>
3044</div>
3045<div class="doc_text">
3046
3047<h5>Syntax:</h5>
3048<pre>
3049 &lt;result&gt; = inttoptr &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3050</pre>
3051
3052<h5>Overview:</h5>
3053<p>The '<tt>inttoptr</tt>' instruction converts an integer <tt>value</tt> to
3054a pointer type, <tt>ty2</tt>.</p>
3055
3056<h5>Arguments:</h5>
3057<p>The '<tt>inttoptr</tt>' instruction takes an <a href="i_integer">integer</a>
3058value to cast, and a type to cast it to, which must be a
Anton Korobeynikov7f705592007-01-12 19:20:47 +0000[diff] [blame^]3059<a href="#t_pointer">pointer</a> type.
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]3060
3061<h5>Semantics:</h5>
3062<p>The '<tt>inttoptr</tt>' instruction converts <tt>value</tt> to type
3063<tt>ty2</tt> by applying either a zero extension or a truncation depending on
3064the size of the integer <tt>value</tt>. If <tt>value</tt> is larger than the
3065size of a pointer then a truncation is done. If <tt>value</tt> is smaller than
3066the size of a pointer then a zero extension is done. If they are the same size,
3067nothing is done (<i>no-op cast</i>).</p>
3068
3069<h5>Example:</h5>
3070<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3071 %X = inttoptr i32 255 to i32* <i>; yields zero extend on 64-bit</i>
3072 %X = inttoptr i32 255 to i32* <i>; yields no-op on 32-bit </i>
3073 %Y = inttoptr i16 0 to i32* <i>; yields zero extend on 32-bit</i>
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]3074</pre>
3075</div>
3076
3077<!-- _______________________________________________________________________ -->
3078<div class="doc_subsubsection">
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]3079 <a name="i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3080</div>
3081<div class="doc_text">
3082
3083<h5>Syntax:</h5>
3084<pre>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]3085 &lt;result&gt; = bitcast &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3086</pre>
3087
3088<h5>Overview:</h5>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]3089<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3090<tt>ty2</tt> without changing any bits.</p>
3091
3092<h5>Arguments:</h5>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]3093<p>The '<tt>bitcast</tt>' instruction takes a value to cast, which must be
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3094a first class value, and a type to cast it to, which must also be a <a
3095 href="#t_firstclass">first class</a> type. The bit sizes of <tt>value</tt>
Reid Spencer19b569f2007-01-09 20:08:58 +0000[diff] [blame]3096and the destination type, <tt>ty2</tt>, must be identical. If the source
3097type is a pointer, the destination type must also be a pointer.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3098
3099<h5>Semantics:</h5>
Reid Spencer5c0ef472006-11-11 23:08:07 +0000[diff] [blame]3100<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
Reid Spencer72679252006-11-11 21:00:47 +0000[diff] [blame]3101<tt>ty2</tt>. It is always a <i>no-op cast</i> because no bits change with
3102this conversion. The conversion is done as if the <tt>value</tt> had been
3103stored to memory and read back as type <tt>ty2</tt>. Pointer types may only be
3104converted to other pointer types with this instruction. To convert pointers to
3105other types, use the <a href="#i_inttoptr">inttoptr</a> or
3106<a href="#i_ptrtoint">ptrtoint</a> instructions first.</p>
Reid Spencer9dee3ac2006-11-08 01:11:31 +0000[diff] [blame]3107
3108<h5>Example:</h5>
3109<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3110 %X = bitcast i8 255 to i8 <i>; yields i8 :-1</i>
3111 %Y = bitcast i32* %x to sint* <i>; yields sint*:%x</i>
3112 %Z = bitcast <2xint> %V to i64; <i>; yields i64: %V</i>
Chris Lattner33ba0d92001-07-09 00:26:23 +0000[diff] [blame]3113</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3114</div>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3115
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]3116<!-- ======================================================================= -->
3117<div class="doc_subsection"> <a name="otherops">Other Operations</a> </div>
3118<div class="doc_text">
3119<p>The instructions in this category are the "miscellaneous"
3120instructions, which defy better classification.</p>
3121</div>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3122
3123<!-- _______________________________________________________________________ -->
3124<div class="doc_subsubsection"><a name="i_icmp">'<tt>icmp</tt>' Instruction</a>
3125</div>
3126<div class="doc_text">
3127<h5>Syntax:</h5>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3128<pre> &lt;result&gt; = icmp &lt;cond&gt; &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt;
3129<i>; yields {i1}:result</i>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3130</pre>
3131<h5>Overview:</h5>
3132<p>The '<tt>icmp</tt>' instruction returns a boolean value based on comparison
3133of its two integer operands.</p>
3134<h5>Arguments:</h5>
3135<p>The '<tt>icmp</tt>' instruction takes three operands. The first operand is
3136the condition code which indicates the kind of comparison to perform. It is not
3137a value, just a keyword. The possibilities for the condition code are:
3138<ol>
3139 <li><tt>eq</tt>: equal</li>
3140 <li><tt>ne</tt>: not equal </li>
3141 <li><tt>ugt</tt>: unsigned greater than</li>
3142 <li><tt>uge</tt>: unsigned greater or equal</li>
3143 <li><tt>ult</tt>: unsigned less than</li>
3144 <li><tt>ule</tt>: unsigned less or equal</li>
3145 <li><tt>sgt</tt>: signed greater than</li>
3146 <li><tt>sge</tt>: signed greater or equal</li>
3147 <li><tt>slt</tt>: signed less than</li>
3148 <li><tt>sle</tt>: signed less or equal</li>
3149</ol>
Reid Spencer350f8aa2007-01-04 05:19:58 +0000[diff] [blame]3150<p>The remaining two arguments must be <a href="#t_integral">integral</a> or
3151<a href="#t_pointer">pointer</a> typed. They must also be identical types.</p>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3152<h5>Semantics:</h5>
3153<p>The '<tt>icmp</tt>' compares <tt>var1</tt> and <tt>var2</tt> according to
3154the condition code given as <tt>cond</tt>. The comparison performed always
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3155yields a <a href="#t_primitive">i1</a> result, as follows:
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3156<ol>
3157 <li><tt>eq</tt>: yields <tt>true</tt> if the operands are equal,
3158 <tt>false</tt> otherwise. No sign interpretation is necessary or performed.
3159 </li>
3160 <li><tt>ne</tt>: yields <tt>true</tt> if the operands are unequal,
3161 <tt>false</tt> otherwise. No sign interpretation is necessary or performed.
3162 <li><tt>ugt</tt>: interprets the operands as unsigned values and yields
3163 <tt>true</tt> if <tt>var1</tt> is greater than <tt>var2</tt>.</li>
3164 <li><tt>uge</tt>: interprets the operands as unsigned values and yields
3165 <tt>true</tt> if <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
3166 <li><tt>ult</tt>: interprets the operands as unsigned values and yields
3167 <tt>true</tt> if <tt>var1</tt> is less than <tt>var2</tt>.</li>
3168 <li><tt>ule</tt>: interprets the operands as unsigned values and yields
3169 <tt>true</tt> if <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
3170 <li><tt>sgt</tt>: interprets the operands as signed values and yields
3171 <tt>true</tt> if <tt>var1</tt> is greater than <tt>var2</tt>.</li>
3172 <li><tt>sge</tt>: interprets the operands as signed values and yields
3173 <tt>true</tt> if <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
3174 <li><tt>slt</tt>: interprets the operands as signed values and yields
3175 <tt>true</tt> if <tt>var1</tt> is less than <tt>var2</tt>.</li>
3176 <li><tt>sle</tt>: interprets the operands as signed values and yields
3177 <tt>true</tt> if <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3178</ol>
3179<p>If the operands are <a href="#t_pointer">pointer</a> typed, the pointer
3180values are treated as integers and then compared.</p>
3181<p>If the operands are <a href="#t_packed">packed</a> typed, the elements of
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3182the vector are compared in turn and the predicate must hold for all
3183elements.</p>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3184
3185<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3186<pre> &lt;result&gt; = icmp eq i32 4, 5 <i>; yields: result=false</i>
3187 &lt;result&gt; = icmp ne float* %X, %X <i>; yields: result=false</i>
3188 &lt;result&gt; = icmp ult i16 4, 5 <i>; yields: result=true</i>
3189 &lt;result&gt; = icmp sgt i16 4, 5 <i>; yields: result=false</i>
3190 &lt;result&gt; = icmp ule i16 -4, 5 <i>; yields: result=false</i>
3191 &lt;result&gt; = icmp sge i16 4, 5 <i>; yields: result=false</i>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3192</pre>
3193</div>
3194
3195<!-- _______________________________________________________________________ -->
3196<div class="doc_subsubsection"><a name="i_fcmp">'<tt>fcmp</tt>' Instruction</a>
3197</div>
3198<div class="doc_text">
3199<h5>Syntax:</h5>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3200<pre> &lt;result&gt; = fcmp &lt;cond&gt; &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt;
3201<i>; yields {i1}:result</i>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3202</pre>
3203<h5>Overview:</h5>
3204<p>The '<tt>fcmp</tt>' instruction returns a boolean value based on comparison
3205of its floating point operands.</p>
3206<h5>Arguments:</h5>
3207<p>The '<tt>fcmp</tt>' instruction takes three operands. The first operand is
3208the condition code which indicates the kind of comparison to perform. It is not
3209a value, just a keyword. The possibilities for the condition code are:
3210<ol>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3211 <li><tt>false</tt>: no comparison, always returns false</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3212 <li><tt>oeq</tt>: ordered and equal</li>
3213 <li><tt>ogt</tt>: ordered and greater than </li>
3214 <li><tt>oge</tt>: ordered and greater than or equal</li>
3215 <li><tt>olt</tt>: ordered and less than </li>
3216 <li><tt>ole</tt>: ordered and less than or equal</li>
3217 <li><tt>one</tt>: ordered and not equal</li>
3218 <li><tt>ord</tt>: ordered (no nans)</li>
3219 <li><tt>ueq</tt>: unordered or equal</li>
3220 <li><tt>ugt</tt>: unordered or greater than </li>
3221 <li><tt>uge</tt>: unordered or greater than or equal</li>
3222 <li><tt>ult</tt>: unordered or less than </li>
3223 <li><tt>ule</tt>: unordered or less than or equal</li>
3224 <li><tt>une</tt>: unordered or not equal</li>
3225 <li><tt>uno</tt>: unordered (either nans)</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3226 <li><tt>true</tt>: no comparison, always returns true</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3227</ol>
Reid Spencer93a49852006-12-06 07:08:07 +0000[diff] [blame]3228<p>In the preceding, <i>ordered</i> means that neither operand is a QNAN while
3229<i>unordered</i> means that either operand may be a QNAN.</p>
Reid Spencer350f8aa2007-01-04 05:19:58 +0000[diff] [blame]3230<p>The <tt>val1</tt> and <tt>val2</tt> arguments must be
3231<a href="#t_floating">floating point</a> typed. They must have identical
3232types.</p>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3233<p>In the foregoing, <i>ordered</i> means that neither operand is a QNAN and
3234<i>unordered</i> means that either operand is a QNAN.</p>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3235<h5>Semantics:</h5>
3236<p>The '<tt>fcmp</tt>' compares <tt>var1</tt> and <tt>var2</tt> according to
3237the condition code given as <tt>cond</tt>. The comparison performed always
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3238yields a <a href="#t_primitive">i1</a> result, as follows:
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3239<ol>
3240 <li><tt>false</tt>: always yields <tt>false</tt>, regardless of operands.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3241 <li><tt>oeq</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3242 <tt>var1</tt> is equal to <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3243 <li><tt>ogt</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3244 <tt>var1</tt> is greather than <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3245 <li><tt>oge</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3246 <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3247 <li><tt>olt</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3248 <tt>var1</tt> is less than <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3249 <li><tt>ole</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3250 <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3251 <li><tt>one</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3252 <tt>var1</tt> is not equal to <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3253 <li><tt>ord</tt>: yields <tt>true</tt> if both operands are not a QNAN.</li>
3254 <li><tt>ueq</tt>: yields <tt>true</tt> if either operand is a QNAN or
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3255 <tt>var1</tt> is equal to <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3256 <li><tt>ugt</tt>: yields <tt>true</tt> if either operand is a QNAN or
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3257 <tt>var1</tt> is greater than <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3258 <li><tt>uge</tt>: yields <tt>true</tt> if either operand is a QNAN or
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3259 <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3260 <li><tt>ult</tt>: yields <tt>true</tt> if either operand is a QNAN or
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3261 <tt>var1</tt> is less than <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3262 <li><tt>ule</tt>: yields <tt>true</tt> if either operand is a QNAN or
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3263 <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3264 <li><tt>une</tt>: yields <tt>true</tt> if either operand is a QNAN or
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3265 <tt>var1</tt> is not equal to <tt>var2</tt>.</li>
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3266 <li><tt>uno</tt>: yields <tt>true</tt> if either operand is a QNAN.</li>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3267 <li><tt>true</tt>: always yields <tt>true</tt>, regardless of operands.</li>
3268</ol>
3269<p>If the operands are <a href="#t_packed">packed</a> typed, the elements of
3270the vector are compared in turn and the predicate must hold for all elements.
Reid Spencerb7f26282006-11-19 03:00:14 +0000[diff] [blame]3271</p>
Reid Spencerf3a70a62006-11-18 21:50:54 +0000[diff] [blame]3272
3273<h5>Example:</h5>
3274<pre> &lt;result&gt; = fcmp oeq float 4.0, 5.0 <i>; yields: result=false</i>
3275 &lt;result&gt; = icmp one float 4.0, 5.0 <i>; yields: result=true</i>
3276 &lt;result&gt; = icmp olt float 4.0, 5.0 <i>; yields: result=true</i>
3277 &lt;result&gt; = icmp ueq double 1.0, 2.0 <i>; yields: result=false</i>
3278</pre>
3279</div>
3280
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]3281<!-- _______________________________________________________________________ -->
3282<div class="doc_subsubsection"> <a name="i_phi">'<tt>phi</tt>'
3283Instruction</a> </div>
3284<div class="doc_text">
3285<h5>Syntax:</h5>
3286<pre> &lt;result&gt; = phi &lt;ty&gt; [ &lt;val0&gt;, &lt;label0&gt;], ...<br></pre>
3287<h5>Overview:</h5>
3288<p>The '<tt>phi</tt>' instruction is used to implement the &#966; node in
3289the SSA graph representing the function.</p>
3290<h5>Arguments:</h5>
3291<p>The type of the incoming values are specified with the first type
3292field. After this, the '<tt>phi</tt>' instruction takes a list of pairs
3293as arguments, with one pair for each predecessor basic block of the
3294current block. Only values of <a href="#t_firstclass">first class</a>
3295type may be used as the value arguments to the PHI node. Only labels
3296may be used as the label arguments.</p>
3297<p>There must be no non-phi instructions between the start of a basic
3298block and the PHI instructions: i.e. PHI instructions must be first in
3299a basic block.</p>
3300<h5>Semantics:</h5>
3301<p>At runtime, the '<tt>phi</tt>' instruction logically takes on the
3302value specified by the parameter, depending on which basic block we
3303came from in the last <a href="#terminators">terminator</a> instruction.</p>
3304<h5>Example:</h5>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3305<pre>Loop: ; Infinite loop that counts from 0 on up...<br> %indvar = phi i32 [ 0, %LoopHeader ], [ %nextindvar, %Loop ]<br> %nextindvar = add i32 %indvar, 1<br> br label %Loop<br></pre>
Reid Spencer2fd21e62006-11-08 01:18:52 +0000[diff] [blame]3306</div>
3307
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3308<!-- _______________________________________________________________________ -->
3309<div class="doc_subsubsection">
3310 <a name="i_select">'<tt>select</tt>' Instruction</a>
3311</div>
3312
3313<div class="doc_text">
3314
3315<h5>Syntax:</h5>
3316
3317<pre>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3318 &lt;result&gt; = select i1 &lt;cond&gt;, &lt;ty&gt; &lt;val1&gt;, &lt;ty&gt; &lt;val2&gt; <i>; yields ty</i>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3319</pre>
3320
3321<h5>Overview:</h5>
3322
3323<p>
3324The '<tt>select</tt>' instruction is used to choose one value based on a
3325condition, without branching.
3326</p>
3327
3328
3329<h5>Arguments:</h5>
3330
3331<p>
3332The '<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.
3333</p>
3334
3335<h5>Semantics:</h5>
3336
3337<p>
3338If the boolean condition evaluates to true, the instruction returns the first
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3339value argument; otherwise, it returns the second value argument.
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3340</p>
3341
3342<h5>Example:</h5>
3343
3344<pre>
Reid Spencerc78f3372007-01-12 03:35:51 +0000[diff] [blame]3345 %X = select i1 true, i8 17, i8 42 <i>; yields i8:17</i>
Chris Lattnercc37aae2004-03-12 05:50:16 +0000[diff] [blame]3346</pre>
3347</div>
3348
Robert Bocchino05ccd702006-01-15 20:48:27 +0000[diff] [blame]3349
3350<!-- _______________________________________________________________________ -->
3351<div class="doc_subsubsection">
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3352 <a name="i_call">'<tt>call</tt>' Instruction</a>
3353</div>
3354
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3355<div class="doc_text">
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3356
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3357<h5>Syntax:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3358<pre>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3359 &lt;result&gt; = [tail] call [<a href="#callingconv">cconv</a>] &lt;ty&gt;* &lt;fnptrval&gt;(&lt;param list&gt;)
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3360</pre>
3361
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3362<h5>Overview:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3363
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3364<p>The '<tt>call</tt>' instruction represents a simple function call.</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3365
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3366<h5>Arguments:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3367
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3368<p>This instruction requires several arguments:</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3369
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]3370<ol>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3371 <li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3372 <p>The optional "tail" marker indicates whether the callee function accesses
3373 any allocas or varargs in the caller. If the "tail" marker is present, the
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3374 function call is eligible for tail call optimization. Note that calls may
3375 be marked "tail" even if they do not occur before a <a
3376 href="#i_ret"><tt>ret</tt></a> instruction.
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3377 </li>
3378 <li>
Chris Lattnerbad10ee2005-05-06 22:57:40 +0000[diff] [blame]3379 <p>The optional "cconv" marker indicates which <a href="callingconv">calling
3380 convention</a> the call should use. If none is specified, the call defaults
3381 to using C calling conventions.
3382 </li>
3383 <li>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3384 <p>'<tt>ty</tt>': shall be the signature of the pointer to function value
3385 being invoked. The argument types must match the types implied by this
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3386 signature. This type can be omitted if the function is not varargs and
3387 if the function type does not return a pointer to a function.</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3388 </li>
3389 <li>
3390 <p>'<tt>fnptrval</tt>': An LLVM value containing a pointer to a function to
3391 be invoked. In most cases, this is a direct function invocation, but
3392 indirect <tt>call</tt>s are just as possible, calling an arbitrary pointer
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3393 to function value.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3394 </li>
3395 <li>
3396 <p>'<tt>function args</tt>': argument list whose types match the
Reid Spencera7e302a2005-05-01 22:22:57 +0000[diff] [blame]3397 function signature argument types. All arguments must be of
3398 <a href="#t_firstclass">first class</a> type. If the function signature
3399 indicates the function accepts a variable number of arguments, the extra
3400 arguments can be specified.</p>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3401 </li>
Chris Lattner6536cfe2002-05-06 22:08:29 +0000[diff] [blame]3402</ol>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3403
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3404<h5>Semantics:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3405
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3406<p>The '<tt>call</tt>' instruction is used to cause control flow to
3407transfer to a specified function, with its incoming arguments bound to
3408the specified values. Upon a '<tt><a href="#i_ret">ret</a></tt>'
3409instruction in the called function, control flow continues with the
3410instruction after the function call, and the return value of the
3411function is bound to the result argument. This is a simpler case of
3412the <a href="#i_invoke">invoke</a> instruction.</p>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3413
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]3414<h5>Example:</h5>
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3415
3416<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3417 %retval = call i32 %test(i32 %argc)
3418 call i32(i8 *, ...) *%printf(i8 * %msg, i32 12, i8 42);
3419 %X = tail call i32 %foo()
3420 %Y = tail call <a href="#callingconv">fastcc</a> i32 %foo()
Chris Lattner2bff5242005-05-06 05:47:36 +0000[diff] [blame]3421</pre>
3422
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3423</div>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3424
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3425<!-- _______________________________________________________________________ -->
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3426<div class="doc_subsubsection">
Chris Lattnerfb6977d2006-01-13 23:26:01 +0000[diff] [blame]3427 <a name="i_va_arg">'<tt>va_arg</tt>' Instruction</a>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3428</div>
3429
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3430<div class="doc_text">
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3431
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]3432<h5>Syntax:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3433
3434<pre>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3435 &lt;resultval&gt; = va_arg &lt;va_list*&gt; &lt;arglist&gt;, &lt;argty&gt;
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3436</pre>
3437
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]3438<h5>Overview:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3439
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3440<p>The '<tt>va_arg</tt>' instruction is used to access arguments passed through
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3441the "variable argument" area of a function call. It is used to implement the
3442<tt>va_arg</tt> macro in C.</p>
3443
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]3444<h5>Arguments:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3445
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3446<p>This instruction takes a <tt>va_list*</tt> value and the type of
3447the argument. It returns a value of the specified argument type and
Jeff Cohen25d4f7e2005-11-11 02:15:27 +0000[diff] [blame]3448increments the <tt>va_list</tt> to point to the next argument. Again, the
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3449actual type of <tt>va_list</tt> is target specific.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3450
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]3451<h5>Semantics:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3452
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3453<p>The '<tt>va_arg</tt>' instruction loads an argument of the specified
3454type from the specified <tt>va_list</tt> and causes the
3455<tt>va_list</tt> to point to the next argument. For more information,
3456see the variable argument handling <a href="#int_varargs">Intrinsic
3457Functions</a>.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3458
3459<p>It is legal for this instruction to be called in a function which does not
3460take a variable number of arguments, for example, the <tt>vfprintf</tt>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3461function.</p>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3462
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3463<p><tt>va_arg</tt> is an LLVM instruction instead of an <a
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3464href="#intrinsics">intrinsic function</a> because it takes a type as an
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3465argument.</p>
3466
Chris Lattner8d1a81d2003-10-18 05:51:36 +0000[diff] [blame]3467<h5>Example:</h5>
Chris Lattnere19d7a72004-09-27 21:51:25 +0000[diff] [blame]3468
3469<p>See the <a href="#int_varargs">variable argument processing</a> section.</p>
3470
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3471</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3472
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3473<!-- *********************************************************************** -->
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3474<div class="doc_section"> <a name="intrinsics">Intrinsic Functions</a> </div>
3475<!-- *********************************************************************** -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3476
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3477<div class="doc_text">
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3478
3479<p>LLVM supports the notion of an "intrinsic function". These functions have
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3480well known names and semantics and are required to follow certain
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3481restrictions. Overall, these instructions represent an extension mechanism for
3482the LLVM language that does not require changing all of the transformations in
3483LLVM to add to the language (or the bytecode reader/writer, the parser,
3484etc...).</p>
3485
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3486<p>Intrinsic function names must all start with an "<tt>llvm.</tt>" prefix. This
3487prefix is reserved in LLVM for intrinsic names; thus, functions may not be named
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3488this. Intrinsic functions must always be external functions: you cannot define
3489the body of intrinsic functions. Intrinsic functions may only be used in call
3490or invoke instructions: it is illegal to take the address of an intrinsic
3491function. Additionally, because intrinsic functions are part of the LLVM
3492language, it is required that they all be documented here if any are added.</p>
3493
3494
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3495<p>To learn how to add an intrinsic function, please see the <a
Chris Lattner590cff32005-05-11 03:35:57 +0000[diff] [blame]3496href="ExtendingLLVM.html">Extending LLVM Guide</a>.
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3497</p>
3498
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3499</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3500
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3501<!-- ======================================================================= -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3502<div class="doc_subsection">
3503 <a name="int_varargs">Variable Argument Handling Intrinsics</a>
3504</div>
3505
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3506<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3507
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3508<p>Variable argument support is defined in LLVM with the <a
Chris Lattnerfb6977d2006-01-13 23:26:01 +0000[diff] [blame]3509 href="#i_va_arg"><tt>va_arg</tt></a> instruction and these three
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3510intrinsic functions. These functions are related to the similarly
3511named macros defined in the <tt>&lt;stdarg.h&gt;</tt> header file.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3512
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3513<p>All of these functions operate on arguments that use a
3514target-specific value type "<tt>va_list</tt>". The LLVM assembly
3515language reference manual does not define what this type is, so all
3516transformations should be prepared to handle intrinsics with any type
3517used.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3518
Chris Lattner374ab302006-05-15 17:26:46 +0000[diff] [blame]3519<p>This example shows how the <a href="#i_va_arg"><tt>va_arg</tt></a>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3520instruction and the variable argument handling intrinsic functions are
3521used.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3522
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3523<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3524define i32 %test(i32 %X, ...) {
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3525 ; Initialize variable argument processing
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3526 %ap = alloca i8 *
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]3527 %ap2 = bitcast i8** %ap to i8*
3528 call void %<a href="#i_va_start">llvm.va_start</a>(i8* %ap2)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3529
3530 ; Read a single integer argument
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3531 %tmp = va_arg i8 ** %ap, i32
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3532
3533 ; Demonstrate usage of llvm.va_copy and llvm.va_end
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3534 %aq = alloca i8 *
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]3535 %aq2 = bitcast i8** %aq to i8*
3536 call void %<a href="#i_va_copy">llvm.va_copy</a>(i8 *%aq2, i8* %ap2)
3537 call void %<a href="#i_va_end">llvm.va_end</a>(i8* %aq2)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3538
3539 ; Stop processing of arguments.
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]3540 call void %<a href="#i_va_end">llvm.va_end</a>(i8* %ap2)
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3541 ret i32 %tmp
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3542}
3543</pre>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3544</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3545
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3546<!-- _______________________________________________________________________ -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3547<div class="doc_subsubsection">
3548 <a name="i_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a>
3549</div>
3550
3551
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3552<div class="doc_text">
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3553<h5>Syntax:</h5>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]3554<pre> declare void %llvm.va_start(i8* &lt;arglist&gt;)<br></pre>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3555<h5>Overview:</h5>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3556<P>The '<tt>llvm.va_start</tt>' intrinsic initializes
3557<tt>*&lt;arglist&gt;</tt> for subsequent use by <tt><a
3558href="#i_va_arg">va_arg</a></tt>.</p>
3559
3560<h5>Arguments:</h5>
3561
3562<P>The argument is a pointer to a <tt>va_list</tt> element to initialize.</p>
3563
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3564<h5>Semantics:</h5>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3565
3566<P>The '<tt>llvm.va_start</tt>' intrinsic works just like the <tt>va_start</tt>
3567macro available in C. In a target-dependent way, it initializes the
3568<tt>va_list</tt> element the argument points to, so that the next call to
3569<tt>va_arg</tt> will produce the first variable argument passed to the function.
3570Unlike the C <tt>va_start</tt> macro, this intrinsic does not need to know the
3571last argument of the function, the compiler can figure that out.</p>
3572
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3573</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3574
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3575<!-- _______________________________________________________________________ -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3576<div class="doc_subsubsection">
3577 <a name="i_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a>
3578</div>
3579
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3580<div class="doc_text">
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3581<h5>Syntax:</h5>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]3582<pre> declare void %llvm.va_end(i8* &lt;arglist&gt;)<br></pre>
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3583<h5>Overview:</h5>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]3584
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3585<p>The '<tt>llvm.va_end</tt>' intrinsic destroys <tt>&lt;arglist&gt;</tt>
3586which has been initialized previously with <tt><a href="#i_va_start">llvm.va_start</a></tt>
3587or <tt><a href="#i_va_copy">llvm.va_copy</a></tt>.</p>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]3588
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3589<h5>Arguments:</h5>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]3590
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3591<p>The argument is a <tt>va_list</tt> to destroy.</p>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]3592
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3593<h5>Semantics:</h5>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]3594
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3595<p>The '<tt>llvm.va_end</tt>' intrinsic works just like the <tt>va_end</tt>
Chris Lattner261efe92003-11-25 01:02:51 +0000[diff] [blame]3596macro available in C. In a target-dependent way, it destroys the <tt>va_list</tt>.
3597Calls to <a href="#i_va_start"><tt>llvm.va_start</tt></a> and <a
3598 href="#i_va_copy"><tt>llvm.va_copy</tt></a> must be matched exactly
3599with calls to <tt>llvm.va_end</tt>.</p>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]3600
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3601</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3602
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3603<!-- _______________________________________________________________________ -->
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3604<div class="doc_subsubsection">
3605 <a name="i_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a>
3606</div>
3607
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3608<div class="doc_text">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3609
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3610<h5>Syntax:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3611
3612<pre>
Chris Lattnerb75137d2007-01-08 07:55:15 +0000[diff] [blame]3613 declare void %llvm.va_copy(i8* &lt;destarglist&gt;, i8* &lt;srcarglist&gt;)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3614</pre>
3615
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3616<h5>Overview:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3617
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3618<p>The '<tt>llvm.va_copy</tt>' intrinsic copies the current argument position from
3619the source argument list to the destination argument list.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3620
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3621<h5>Arguments:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3622
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3623<p>The first argument is a pointer to a <tt>va_list</tt> element to initialize.
Andrew Lenharthd0a4c622005-06-22 20:38:11 +0000[diff] [blame]3624The second argument is a pointer to a <tt>va_list</tt> element to copy from.</p>
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3625
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3626
Chris Lattnerd9ad5b32003-05-08 04:57:36 +0000[diff] [blame]3627<h5>Semantics:</h5>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3628
Andrew Lenharth8bf607a2005-06-18 18:28:17 +0000[diff] [blame]3629<p>The '<tt>llvm.va_copy</tt>' intrinsic works just like the <tt>va_copy</tt> macro
3630available in C. In a target-dependent way, it copies the source
3631<tt>va_list</tt> element into the destination list. This intrinsic is necessary
3632because the <tt><a href="i_va_begin">llvm.va_begin</a></tt> intrinsic may be
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3633arbitrarily complex and require memory allocation, for example.</p>
3634
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]3635</div>
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]3636
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]3637<!-- ======================================================================= -->
3638<div class="doc_subsection">
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3639 <a name="int_gc">Accurate Garbage Collection Intrinsics</a>
3640</div>
3641
3642<div class="doc_text">
3643
3644<p>
3645LLVM support for <a href="GarbageCollection.html">Accurate Garbage
3646Collection</a> requires the implementation and generation of these intrinsics.
3647These intrinsics allow identification of <a href="#i_gcroot">GC roots on the
3648stack</a>, as well as garbage collector implementations that require <a
3649href="#i_gcread">read</a> and <a href="#i_gcwrite">write</a> barriers.
3650Front-ends for type-safe garbage collected languages should generate these
3651intrinsics to make use of the LLVM garbage collectors. For more details, see <a
3652href="GarbageCollection.html">Accurate Garbage Collection with LLVM</a>.
3653</p>
3654</div>
3655
3656<!-- _______________________________________________________________________ -->
3657<div class="doc_subsubsection">
3658 <a name="i_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a>
3659</div>
3660
3661<div class="doc_text">
3662
3663<h5>Syntax:</h5>
3664
3665<pre>
Reid Spencera8d451e2005-04-26 20:50:44 +0000[diff] [blame]3666 declare void %llvm.gcroot(&lt;ty&gt;** %ptrloc, &lt;ty2&gt;* %metadata)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3667</pre>
3668
3669<h5>Overview:</h5>
3670
John Criswell9e2485c2004-12-10 15:51:16 +0000[diff] [blame]3671<p>The '<tt>llvm.gcroot</tt>' intrinsic declares the existence of a GC root to
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3672the code generator, and allows some metadata to be associated with it.</p>
3673
3674<h5>Arguments:</h5>
3675
3676<p>The first argument specifies the address of a stack object that contains the
3677root pointer. The second pointer (which must be either a constant or a global
3678value address) contains the meta-data to be associated with the root.</p>
3679
3680<h5>Semantics:</h5>
3681
3682<p>At runtime, a call to this intrinsics stores a null pointer into the "ptrloc"
3683location. At compile-time, the code generator generates information to allow
3684the runtime to find the pointer at GC safe points.
3685</p>
3686
3687</div>
3688
3689
3690<!-- _______________________________________________________________________ -->
3691<div class="doc_subsubsection">
3692 <a name="i_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a>
3693</div>
3694
3695<div class="doc_text">
3696
3697<h5>Syntax:</h5>
3698
3699<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3700 declare i8 * %llvm.gcread(i8 * %ObjPtr, i8 ** %Ptr)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3701</pre>
3702
3703<h5>Overview:</h5>
3704
3705<p>The '<tt>llvm.gcread</tt>' intrinsic identifies reads of references from heap
3706locations, allowing garbage collector implementations that require read
3707barriers.</p>
3708
3709<h5>Arguments:</h5>
3710
Chris Lattner80626e92006-03-14 20:02:51 +0000[diff] [blame]3711<p>The second argument is the address to read from, which should be an address
3712allocated from the garbage collector. The first object is a pointer to the
3713start of the referenced object, if needed by the language runtime (otherwise
3714null).</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3715
3716<h5>Semantics:</h5>
3717
3718<p>The '<tt>llvm.gcread</tt>' intrinsic has the same semantics as a load
3719instruction, but may be replaced with substantially more complex code by the
3720garbage collector runtime, as needed.</p>
3721
3722</div>
3723
3724
3725<!-- _______________________________________________________________________ -->
3726<div class="doc_subsubsection">
3727 <a name="i_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a>
3728</div>
3729
3730<div class="doc_text">
3731
3732<h5>Syntax:</h5>
3733
3734<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3735 declare void %llvm.gcwrite(i8 * %P1, i8 * %Obj, i8 ** %P2)
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3736</pre>
3737
3738<h5>Overview:</h5>
3739
3740<p>The '<tt>llvm.gcwrite</tt>' intrinsic identifies writes of references to heap
3741locations, allowing garbage collector implementations that require write
3742barriers (such as generational or reference counting collectors).</p>
3743
3744<h5>Arguments:</h5>
3745
Chris Lattner80626e92006-03-14 20:02:51 +0000[diff] [blame]3746<p>The first argument is the reference to store, the second is the start of the
3747object to store it to, and the third is the address of the field of Obj to
3748store to. If the runtime does not require a pointer to the object, Obj may be
3749null.</p>
Chris Lattnerd7923912004-05-23 21:06:01 +0000[diff] [blame]3750
3751<h5>Semantics:</h5>
3752
3753<p>The '<tt>llvm.gcwrite</tt>' intrinsic has the same semantics as a store
3754instruction, but may be replaced with substantially more complex code by the
3755garbage collector runtime, as needed.</p>
3756
3757</div>
3758
3759
3760
3761<!-- ======================================================================= -->
3762<div class="doc_subsection">
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3763 <a name="int_codegen">Code Generator Intrinsics</a>
3764</div>
3765
3766<div class="doc_text">
3767<p>
3768These intrinsics are provided by LLVM to expose special features that may only
3769be implemented with code generator support.
3770</p>
3771
3772</div>
3773
3774<!-- _______________________________________________________________________ -->
3775<div class="doc_subsubsection">
3776 <a name="i_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a>
3777</div>
3778
3779<div class="doc_text">
3780
3781<h5>Syntax:</h5>
3782<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3783 declare i8 *%llvm.returnaddress(i32 &lt;level&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3784</pre>
3785
3786<h5>Overview:</h5>
3787
3788<p>
Chris Lattner32b5d712006-10-15 20:05:59 +0000[diff] [blame]3789The '<tt>llvm.returnaddress</tt>' intrinsic attempts to compute a
3790target-specific value indicating the return address of the current function
3791or one of its callers.
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3792</p>
3793
3794<h5>Arguments:</h5>
3795
3796<p>
3797The argument to this intrinsic indicates which function to return the address
3798for. Zero indicates the calling function, one indicates its caller, etc. The
3799argument is <b>required</b> to be a constant integer value.
3800</p>
3801
3802<h5>Semantics:</h5>
3803
3804<p>
3805The '<tt>llvm.returnaddress</tt>' intrinsic either returns a pointer indicating
3806the return address of the specified call frame, or zero if it cannot be
3807identified. The value returned by this intrinsic is likely to be incorrect or 0
3808for arguments other than zero, so it should only be used for debugging purposes.
3809</p>
3810
3811<p>
3812Note that calling this intrinsic does not prevent function inlining or other
Chris Lattnerb40bb382005-03-07 20:30:51 +0000[diff] [blame]3813aggressive transformations, so the value returned may not be that of the obvious
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3814source-language caller.
3815</p>
3816</div>
3817
3818
3819<!-- _______________________________________________________________________ -->
3820<div class="doc_subsubsection">
3821 <a name="i_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a>
3822</div>
3823
3824<div class="doc_text">
3825
3826<h5>Syntax:</h5>
3827<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3828 declare i8 *%llvm.frameaddress(i32 &lt;level&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3829</pre>
3830
3831<h5>Overview:</h5>
3832
3833<p>
Chris Lattner32b5d712006-10-15 20:05:59 +0000[diff] [blame]3834The '<tt>llvm.frameaddress</tt>' intrinsic attempts to return the
3835target-specific frame pointer value for the specified stack frame.
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3836</p>
3837
3838<h5>Arguments:</h5>
3839
3840<p>
3841The argument to this intrinsic indicates which function to return the frame
3842pointer for. Zero indicates the calling function, one indicates its caller,
3843etc. The argument is <b>required</b> to be a constant integer value.
3844</p>
3845
3846<h5>Semantics:</h5>
3847
3848<p>
3849The '<tt>llvm.frameaddress</tt>' intrinsic either returns a pointer indicating
3850the frame address of the specified call frame, or zero if it cannot be
3851identified. The value returned by this intrinsic is likely to be incorrect or 0
3852for arguments other than zero, so it should only be used for debugging purposes.
3853</p>
3854
3855<p>
3856Note that calling this intrinsic does not prevent function inlining or other
Chris Lattnerb40bb382005-03-07 20:30:51 +0000[diff] [blame]3857aggressive transformations, so the value returned may not be that of the obvious
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]3858source-language caller.
3859</p>
3860</div>
3861
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]3862<!-- _______________________________________________________________________ -->
3863<div class="doc_subsubsection">
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]3864 <a name="i_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a>
3865</div>
3866
3867<div class="doc_text">
3868
3869<h5>Syntax:</h5>
3870<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3871 declare i8 *%llvm.stacksave()
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]3872</pre>
3873
3874<h5>Overview:</h5>
3875
3876<p>
3877The '<tt>llvm.stacksave</tt>' intrinsic is used to remember the current state of
3878the function stack, for use with <a href="#i_stackrestore">
3879<tt>llvm.stackrestore</tt></a>. This is useful for implementing language
3880features like scoped automatic variable sized arrays in C99.
3881</p>
3882
3883<h5>Semantics:</h5>
3884
3885<p>
3886This intrinsic returns a opaque pointer value that can be passed to <a
3887href="#i_stackrestore"><tt>llvm.stackrestore</tt></a>. When an
3888<tt>llvm.stackrestore</tt> intrinsic is executed with a value saved from
3889<tt>llvm.stacksave</tt>, it effectively restores the state of the stack to the
3890state it was in when the <tt>llvm.stacksave</tt> intrinsic executed. In
3891practice, this pops any <a href="#i_alloca">alloca</a> blocks from the stack
3892that were allocated after the <tt>llvm.stacksave</tt> was executed.
3893</p>
3894
3895</div>
3896
3897<!-- _______________________________________________________________________ -->
3898<div class="doc_subsubsection">
3899 <a name="i_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a>
3900</div>
3901
3902<div class="doc_text">
3903
3904<h5>Syntax:</h5>
3905<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3906 declare void %llvm.stackrestore(i8 * %ptr)
Chris Lattner57e1f392006-01-13 02:03:13 +0000[diff] [blame]3907</pre>
3908
3909<h5>Overview:</h5>
3910
3911<p>
3912The '<tt>llvm.stackrestore</tt>' intrinsic is used to restore the state of
3913the function stack to the state it was in when the corresponding <a
3914href="#llvm.stacksave"><tt>llvm.stacksave</tt></a> intrinsic executed. This is
3915useful for implementing language features like scoped automatic variable sized
3916arrays in C99.
3917</p>
3918
3919<h5>Semantics:</h5>
3920
3921<p>
3922See the description for <a href="#i_stacksave"><tt>llvm.stacksave</tt></a>.
3923</p>
3924
3925</div>
3926
3927
3928<!-- _______________________________________________________________________ -->
3929<div class="doc_subsubsection">
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]3930 <a name="i_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a>
3931</div>
3932
3933<div class="doc_text">
3934
3935<h5>Syntax:</h5>
3936<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3937 declare void %llvm.prefetch(i8 * &lt;address&gt;,
3938 i32 &lt;rw&gt;, i32 &lt;locality&gt;)
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]3939</pre>
3940
3941<h5>Overview:</h5>
3942
3943
3944<p>
3945The '<tt>llvm.prefetch</tt>' intrinsic is a hint to the code generator to insert
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3946a prefetch instruction if supported; otherwise, it is a noop. Prefetches have
3947no
3948effect on the behavior of the program but can change its performance
Chris Lattner2a615362005-02-28 19:47:14 +0000[diff] [blame]3949characteristics.
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]3950</p>
3951
3952<h5>Arguments:</h5>
3953
3954<p>
3955<tt>address</tt> is the address to be prefetched, <tt>rw</tt> is the specifier
3956determining if the fetch should be for a read (0) or write (1), and
3957<tt>locality</tt> is a temporal locality specifier ranging from (0) - no
Chris Lattneraeffb4a2005-03-07 20:31:38 +0000[diff] [blame]3958locality, to (3) - extremely local keep in cache. The <tt>rw</tt> and
Chris Lattner9a9d7ac2005-02-28 19:24:19 +0000[diff] [blame]3959<tt>locality</tt> arguments must be constant integers.
3960</p>
3961
3962<h5>Semantics:</h5>
3963
3964<p>
3965This intrinsic does not modify the behavior of the program. In particular,
3966prefetches cannot trap and do not produce a value. On targets that support this
3967intrinsic, the prefetch can provide hints to the processor cache for better
3968performance.
3969</p>
3970
3971</div>
3972
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]3973<!-- _______________________________________________________________________ -->
3974<div class="doc_subsubsection">
3975 <a name="i_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a>
3976</div>
3977
3978<div class="doc_text">
3979
3980<h5>Syntax:</h5>
3981<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]3982 declare void %llvm.pcmarker( i32 &lt;id&gt; )
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]3983</pre>
3984
3985<h5>Overview:</h5>
3986
3987
3988<p>
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3989The '<tt>llvm.pcmarker</tt>' intrinsic is a method to export a Program Counter
3990(PC) in a region of
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]3991code to simulators and other tools. The method is target specific, but it is
3992expected that the marker will use exported symbols to transmit the PC of the marker.
Jeff Cohen25d4f7e2005-11-11 02:15:27 +0000[diff] [blame]3993The marker makes no guarantees that it will remain with any specific instruction
Chris Lattnerd07c3f42005-11-15 06:07:55 +0000[diff] [blame]3994after optimizations. It is possible that the presence of a marker will inhibit
Chris Lattnerb3e7afd2006-03-24 07:16:10 +0000[diff] [blame]3995optimizations. The intended use is to be inserted after optimizations to allow
John Criswellfc6b8952005-05-16 16:17:45 +0000[diff] [blame]3996correlations of simulation runs.
Andrew Lenharth7f4ec3b2005-03-28 20:05:49 +0000[diff] [blame]3997</p>
3998
3999<h5>Arguments:</h5>
4000
4001<p>
4002<tt>id</tt> is a numerical id identifying the marker.
4003</p>
4004
4005<h5>Semantics:</h5>
4006
4007<p>
4008This intrinsic does not modify the behavior of the program. Backends that do not
4009support this intrinisic may ignore it.
4010</p>
4011
4012</div>
4013
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]4014<!-- _______________________________________________________________________ -->
4015<div class="doc_subsubsection">
4016 <a name="i_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a>
4017</div>
4018
4019<div class="doc_text">
4020
4021<h5>Syntax:</h5>
4022<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4023 declare i64 %llvm.readcyclecounter( )
Andrew Lenharth51b8d542005-11-11 16:47:30 +0000[diff] [blame]4024</pre>
4025
4026<h5>Overview:</h5>
4027
4028
4029<p>
4030The '<tt>llvm.readcyclecounter</tt>' intrinsic provides access to the cycle
4031counter register (or similar low latency, high accuracy clocks) on those targets
4032that support it. On X86, it should map to RDTSC. On Alpha, it should map to RPCC.
4033As the backing counters overflow quickly (on the order of 9 seconds on alpha), this
4034should only be used for small timings.
4035</p>
4036
4037<h5>Semantics:</h5>
4038
4039<p>
4040When directly supported, reading the cycle counter should not modify any memory.
4041Implementations are allowed to either return a application specific value or a
4042system wide value. On backends without support, this is lowered to a constant 0.
4043</p>
4044
4045</div>
4046
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4047<!-- ======================================================================= -->
4048<div class="doc_subsection">
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4049 <a name="int_libc">Standard C Library Intrinsics</a>
4050</div>
4051
4052<div class="doc_text">
4053<p>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4054LLVM provides intrinsics for a few important standard C library functions.
4055These intrinsics allow source-language front-ends to pass information about the
4056alignment of the pointer arguments to the code generator, providing opportunity
4057for more efficient code generation.
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4058</p>
4059
4060</div>
4061
4062<!-- _______________________________________________________________________ -->
4063<div class="doc_subsubsection">
4064 <a name="i_memcpy">'<tt>llvm.memcpy</tt>' Intrinsic</a>
4065</div>
4066
4067<div class="doc_text">
4068
4069<h5>Syntax:</h5>
4070<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4071 declare void %llvm.memcpy.i32(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
4072 i32 &lt;len&gt;, i32 &lt;align&gt;)
4073 declare void %llvm.memcpy.i64(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
4074 i64 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4075</pre>
4076
4077<h5>Overview:</h5>
4078
4079<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4080The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4081location to the destination location.
4082</p>
4083
4084<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4085Note that, unlike the standard libc function, the <tt>llvm.memcpy.*</tt>
4086intrinsics do not return a value, and takes an extra alignment argument.
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4087</p>
4088
4089<h5>Arguments:</h5>
4090
4091<p>
4092The first argument is a pointer to the destination, the second is a pointer to
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4093the source. The third argument is an integer argument
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4094specifying the number of bytes to copy, and the fourth argument is the alignment
4095of the source and destination locations.
4096</p>
4097
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]4098<p>
4099If the call to this intrinisic has an alignment value that is not 0 or 1, then
Chris Lattnerf0afc2c2006-03-04 00:02:10 +0000[diff] [blame]4100the caller guarantees that both the source and destination pointers are aligned
4101to that boundary.
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]4102</p>
4103
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4104<h5>Semantics:</h5>
4105
4106<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4107The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
Chris Lattner33aec9e2004-02-12 17:01:32 +0000[diff] [blame]4108location to the destination location, which are not allowed to overlap. It
4109copies "len" bytes of memory over. If the argument is known to be aligned to
4110some boundary, this can be specified as the fourth argument, otherwise it should
4111be set to 0 or 1.
4112</p>
4113</div>
4114
4115
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]4116<!-- _______________________________________________________________________ -->
4117<div class="doc_subsubsection">
4118 <a name="i_memmove">'<tt>llvm.memmove</tt>' Intrinsic</a>
4119</div>
4120
4121<div class="doc_text">
4122
4123<h5>Syntax:</h5>
4124<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4125 declare void %llvm.memmove.i32(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
4126 i32 &lt;len&gt;, i32 &lt;align&gt;)
4127 declare void %llvm.memmove.i64(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
4128 i64 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]4129</pre>
4130
4131<h5>Overview:</h5>
4132
4133<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4134The '<tt>llvm.memmove.*</tt>' intrinsics move a block of memory from the source
4135location to the destination location. It is similar to the
4136'<tt>llvm.memcmp</tt>' intrinsic but allows the two memory locations to overlap.
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]4137</p>
4138
4139<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4140Note that, unlike the standard libc function, the <tt>llvm.memmove.*</tt>
4141intrinsics do not return a value, and takes an extra alignment argument.
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]4142</p>
4143
4144<h5>Arguments:</h5>
4145
4146<p>
4147The first argument is a pointer to the destination, the second is a pointer to
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4148the source. The third argument is an integer argument
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]4149specifying the number of bytes to copy, and the fourth argument is the alignment
4150of the source and destination locations.
4151</p>
4152
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]4153<p>
4154If the call to this intrinisic has an alignment value that is not 0 or 1, then
Chris Lattnerf0afc2c2006-03-04 00:02:10 +0000[diff] [blame]4155the caller guarantees that the source and destination pointers are aligned to
4156that boundary.
Chris Lattner3301ced2004-02-12 21:18:15 +0000[diff] [blame]4157</p>
4158
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]4159<h5>Semantics:</h5>
4160
4161<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4162The '<tt>llvm.memmove.*</tt>' intrinsics copy a block of memory from the source
Chris Lattner0eb51b42004-02-12 18:10:10 +0000[diff] [blame]4163location to the destination location, which may overlap. It
4164copies "len" bytes of memory over. If the argument is known to be aligned to
4165some boundary, this can be specified as the fourth argument, otherwise it should
4166be set to 0 or 1.
4167</p>
4168</div>
4169
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]4170
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4171<!-- _______________________________________________________________________ -->
4172<div class="doc_subsubsection">
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4173 <a name="i_memset">'<tt>llvm.memset.*</tt>' Intrinsics</a>
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4174</div>
4175
4176<div class="doc_text">
4177
4178<h5>Syntax:</h5>
4179<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4180 declare void %llvm.memset.i32(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
4181 i32 &lt;len&gt;, i32 &lt;align&gt;)
4182 declare void %llvm.memset.i64(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
4183 i64 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4184</pre>
4185
4186<h5>Overview:</h5>
4187
4188<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4189The '<tt>llvm.memset.*</tt>' intrinsics fill a block of memory with a particular
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4190byte value.
4191</p>
4192
4193<p>
4194Note that, unlike the standard libc function, the <tt>llvm.memset</tt> intrinsic
4195does not return a value, and takes an extra alignment argument.
4196</p>
4197
4198<h5>Arguments:</h5>
4199
4200<p>
4201The first argument is a pointer to the destination to fill, the second is the
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4202byte value to fill it with, the third argument is an integer
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4203argument specifying the number of bytes to fill, and the fourth argument is the
4204known alignment of destination location.
4205</p>
4206
4207<p>
4208If the call to this intrinisic has an alignment value that is not 0 or 1, then
Chris Lattnerf0afc2c2006-03-04 00:02:10 +0000[diff] [blame]4209the caller guarantees that the destination pointer is aligned to that boundary.
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4210</p>
4211
4212<h5>Semantics:</h5>
4213
4214<p>
Chris Lattner5b310c32006-03-03 00:07:20 +0000[diff] [blame]4215The '<tt>llvm.memset.*</tt>' intrinsics fill "len" bytes of memory starting at
4216the
Chris Lattner10610642004-02-14 04:08:35 +0000[diff] [blame]4217destination location. If the argument is known to be aligned to some boundary,
4218this can be specified as the fourth argument, otherwise it should be set to 0 or
42191.
4220</p>
4221</div>
4222
4223
Chris Lattner32006282004-06-11 02:28:03 +0000[diff] [blame]4224<!-- _______________________________________________________________________ -->
4225<div class="doc_subsubsection">
Chris Lattnerec6cb612006-01-16 22:38:59 +0000[diff] [blame]4226 <a name="i_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a>
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]4227</div>
4228
4229<div class="doc_text">
4230
4231<h5>Syntax:</h5>
4232<pre>
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]4233 declare float %llvm.sqrt.f32(float %Val)
4234 declare double %llvm.sqrt.f64(double %Val)
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]4235</pre>
4236
4237<h5>Overview:</h5>
4238
4239<p>
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]4240The '<tt>llvm.sqrt</tt>' intrinsics return the sqrt of the specified operand,
Chris Lattnera4d74142005-07-21 01:29:16 +0000[diff] [blame]4241returning the same value as the libm '<tt>sqrt</tt>' function would. Unlike
4242<tt>sqrt</tt> in libm, however, <tt>llvm.sqrt</tt> has undefined behavior for
4243negative numbers (which allows for better optimization).
4244</p>
4245
4246<h5>Arguments:</h5>
4247
4248<p>
4249The argument and return value are floating point numbers of the same type.
4250</p>
4251
4252<h5>Semantics:</h5>
4253
4254<p>
4255This function returns the sqrt of the specified operand if it is a positive
4256floating point number.
4257</p>
4258</div>
4259
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]4260<!-- _______________________________________________________________________ -->
4261<div class="doc_subsubsection">
4262 <a name="i_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a>
4263</div>
4264
4265<div class="doc_text">
4266
4267<h5>Syntax:</h5>
4268<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4269 declare float %llvm.powi.f32(float %Val, i32 %power)
4270 declare double %llvm.powi.f64(double %Val, i32 %power)
Chris Lattnerf4d252d2006-09-08 06:34:02 +0000[diff] [blame]4271</pre>
4272
4273<h5>Overview:</h5>
4274
4275<p>
4276The '<tt>llvm.powi.*</tt>' intrinsics return the first operand raised to the
4277specified (positive or negative) power. The order of evaluation of
4278multiplications is not defined.
4279</p>
4280
4281<h5>Arguments:</h5>
4282
4283<p>
4284The second argument is an integer power, and the first is a value to raise to
4285that power.
4286</p>
4287
4288<h5>Semantics:</h5>
4289
4290<p>
4291This function returns the first value raised to the second power with an
4292unspecified sequence of rounding operations.</p>
4293</div>
4294
4295
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4296<!-- ======================================================================= -->
4297<div class="doc_subsection">
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]4298 <a name="int_manip">Bit Manipulation Intrinsics</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4299</div>
4300
4301<div class="doc_text">
4302<p>
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]4303LLVM provides intrinsics for a few important bit manipulation operations.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4304These allow efficient code generation for some algorithms.
4305</p>
4306
4307</div>
4308
4309<!-- _______________________________________________________________________ -->
4310<div class="doc_subsubsection">
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]4311 <a name="i_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a>
4312</div>
4313
4314<div class="doc_text">
4315
4316<h5>Syntax:</h5>
4317<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4318 declare i16 %llvm.bswap.i16(i16 &lt;id&gt;)
4319 declare i32 %llvm.bswap.i32(i32 &lt;id&gt;)
4320 declare i64 %llvm.bswap.i64(i64 &lt;id&gt;)
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]4321</pre>
4322
4323<h5>Overview:</h5>
4324
4325<p>
4326The '<tt>llvm.bwsap</tt>' family of intrinsics is used to byteswap a 16, 32 or
432764 bit quantity. These are useful for performing operations on data that is not
4328in the target's native byte order.
4329</p>
4330
4331<h5>Semantics:</h5>
4332
4333<p>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4334The <tt>llvm.bswap.16</tt> intrinsic returns an i16 value that has the high
4335and low byte of the input i16 swapped. Similarly, the <tt>llvm.bswap.i32</tt>
4336intrinsic returns an i32 value that has the four bytes of the input i32
4337swapped, so that if the input bytes are numbered 0, 1, 2, 3 then the returned
4338i32 will have its bytes in 3, 2, 1, 0 order. The <tt>llvm.bswap.i64</tt>
4339intrinsic extends this concept to 64 bits.
Nate Begeman7e36c472006-01-13 23:26:38 +0000[diff] [blame]4340</p>
4341
4342</div>
4343
4344<!-- _______________________________________________________________________ -->
4345<div class="doc_subsubsection">
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]4346 <a name="int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4347</div>
4348
4349<div class="doc_text">
4350
4351<h5>Syntax:</h5>
4352<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4353 declare i8 %llvm.ctpop.i8 (i8 &lt;src&gt;)
4354 declare i16 %llvm.ctpop.i16(i16 &lt;src&gt;)
4355 declare i32 %llvm.ctpop.i32(i32 &lt;src&gt;)
4356 declare i64 %llvm.ctpop.i64(i64 &lt;src&gt;)
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4357</pre>
4358
4359<h5>Overview:</h5>
4360
4361<p>
Chris Lattnerec6cb612006-01-16 22:38:59 +0000[diff] [blame]4362The '<tt>llvm.ctpop</tt>' family of intrinsics counts the number of bits set in a
4363value.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4364</p>
4365
4366<h5>Arguments:</h5>
4367
4368<p>
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]4369The only argument is the value to be counted. The argument may be of any
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]4370integer type. The return type must match the argument type.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4371</p>
4372
4373<h5>Semantics:</h5>
4374
4375<p>
4376The '<tt>llvm.ctpop</tt>' intrinsic counts the 1's in a variable.
4377</p>
4378</div>
4379
4380<!-- _______________________________________________________________________ -->
4381<div class="doc_subsubsection">
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]4382 <a name="int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic</a>
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4383</div>
4384
4385<div class="doc_text">
4386
4387<h5>Syntax:</h5>
4388<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4389 declare i8 %llvm.ctlz.i8 (i8 &lt;src&gt;)
4390 declare i16 %llvm.ctlz.i16(i16 &lt;src&gt;)
4391 declare i32 %llvm.ctlz.i32(i32 &lt;src&gt;)
4392 declare i64 %llvm.ctlz.i64(i64 &lt;src&gt;)
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4393</pre>
4394
4395<h5>Overview:</h5>
4396
4397<p>
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]4398The '<tt>llvm.ctlz</tt>' family of intrinsic functions counts the number of
4399leading zeros in a variable.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4400</p>
4401
4402<h5>Arguments:</h5>
4403
4404<p>
Chris Lattnercfe6b372005-05-07 01:46:40 +0000[diff] [blame]4405The only argument is the value to be counted. The argument may be of any
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]4406integer type. The return type must match the argument type.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4407</p>
4408
4409<h5>Semantics:</h5>
4410
4411<p>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4412The '<tt>llvm.ctlz</tt>' intrinsic counts the leading (most significant) zeros
4413in a variable. If the src == 0 then the result is the size in bits of the type
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4414of src. For example, <tt>llvm.ctlz(i32 2) = 30</tt>.
Andrew Lenharthec370fd2005-05-03 18:01:48 +0000[diff] [blame]4415</p>
4416</div>
Chris Lattner32006282004-06-11 02:28:03 +0000[diff] [blame]4417
4418
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4419
4420<!-- _______________________________________________________________________ -->
4421<div class="doc_subsubsection">
Chris Lattner8a886be2006-01-16 22:34:14 +0000[diff] [blame]4422 <a name="int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic</a>
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4423</div>
4424
4425<div class="doc_text">
4426
4427<h5>Syntax:</h5>
4428<pre>
Reid Spencerca86e162006-12-31 07:07:53 +0000[diff] [blame]4429 declare i8 %llvm.cttz.i8 (i8 &lt;src&gt;)
4430 declare i16 %llvm.cttz.i16(i16 &lt;src&gt;)
4431 declare i32 %llvm.cttz.i32(i32 &lt;src&gt;)
4432 declare i64 %llvm.cttz.i64(i64 &lt;src&gt;)
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4433</pre>
4434
4435<h5>Overview:</h5>
4436
4437<p>
Reid Spencer0b118202006-01-16 21:12:35 +0000[diff] [blame]4438The '<tt>llvm.cttz</tt>' family of intrinsic functions counts the number of
4439trailing zeros.
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4440</p>
4441
4442<h5>Arguments:</h5>
4443
4444<p>
4445The only argument is the value to be counted. The argument may be of any
Reid Spencera5173382007-01-04 16:43:23 +0000[diff] [blame]4446integer type. The return type must match the argument type.
Chris Lattnereff29ab2005-05-15 19:39:26 +0000[diff] [blame]4447</p>
4448
4449<h5>Semantics:</h5>
4450
4451<p>
4452The '<tt>llvm.cttz</tt>' intrinsic counts the trailing (least significant) zeros
4453in a variable. If the src == 0 then the result is the size in bits of the type
4454of src. For example, <tt>llvm.cttz(2) = 1</tt>.
4455</p>
4456</div>
4457
Chris Lattner8ff75902004-01-06 05:31:32 +0000[diff] [blame]4458<!-- ======================================================================= -->
4459<div class="doc_subsection">
4460 <a name="int_debugger">Debugger Intrinsics</a>
4461</div>
4462
4463<div class="doc_text">
4464<p>
4465The LLVM debugger intrinsics (which all start with <tt>llvm.dbg.</tt> prefix),
4466are described in the <a
4467href="SourceLevelDebugging.html#format_common_intrinsics">LLVM Source Level
4468Debugging</a> document.
4469</p>
4470</div>
4471
4472
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4473<!-- *********************************************************************** -->
Chris Lattner00950542001-06-06 20:29:01 +0000[diff] [blame]4474<hr>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]4475<address>
4476 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
4477 src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
4478 <a href="http://validator.w3.org/check/referer"><img
4479 src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!" /></a>
4480
4481 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
Reid Spencer05fe4b02006-03-14 05:39:39 +0000[diff] [blame]4482 <a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br>
Misha Brukmandaa4cb02004-03-01 17:47:27 +0000[diff] [blame]4483 Last modified: $Date$
4484</address>
Misha Brukman9d0919f2003-11-08 01:05:38 +0000[diff] [blame]4485</body>
4486</html>