Gitiles
Code Review Sign In
gerrit-public.fairphone.software / toolchain / llvm-project / beccb74eec1c126cc6914d4ca0166e6f1a65c70f / . / llvm / docs / LangRef.html
blob: 2fb27f6384a519c45c37d2ed327b8503708b5215 [file] [log] [blame]
Misha Brukmanc501f552004-03-01 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 Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3<html>
4<head>
5 <title>LLVM Assembly Language Reference Manual</title>
Reid Spencercb84e432004-08-26 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 Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]10 <link rel="stylesheet" href="llvm.css" type="text/css">
11</head>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]12
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]13<body>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]14
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]15<div class="doc_title"> LLVM Language Reference Manual </div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]16<ol>
Misha Brukman76307852003-11-08 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 Lattner6af02f32004-12-09 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 Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]23 <li><a href="#linkage">Linkage Types</a></li>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]24 <li><a href="#callingconv">Calling Conventions</a></li>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]25 <li><a href="#globalvars">Global Variables</a></li>
Chris Lattner91c15c42006-01-23 23:23:47 +0000[diff] [blame]26 <li><a href="#functionstructure">Functions</a></li>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]27 <li><a href="#paramattrs">Parameter Attributes</a></li>
Chris Lattner91c15c42006-01-23 23:23:47 +0000[diff] [blame]28 <li><a href="#moduleasm">Module-Level Inline Assembly</a></li>
Reid Spencer50c723a2007-02-19 23:54:10 +0000[diff] [blame]29 <li><a href="#datalayout">Data Layout</a></li>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]30 </ol>
31 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]32 <li><a href="#typesystem">Type System</a>
33 <ol>
Robert Bocchino820bc75b2006-02-17 21:18:08 +0000[diff] [blame]34 <li><a href="#t_primitive">Primitive Types</a>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]35 <ol>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]36 <li><a href="#t_classifications">Type Classifications</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]37 </ol>
38 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]39 <li><a href="#t_derived">Derived Types</a>
40 <ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]41 <li><a href="#t_array">Array Type</a></li>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]42 <li><a href="#t_function">Function Type</a></li>
43 <li><a href="#t_pointer">Pointer Type</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]44 <li><a href="#t_struct">Structure Type</a></li>
Andrew Lenharth8df88e22006-12-08 17:13:00 +0000[diff] [blame]45 <li><a href="#t_pstruct">Packed Structure Type</a></li>
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]46 <li><a href="#t_vector">Vector Type</a></li>
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]47 <li><a href="#t_opaque">Opaque Type</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]48 </ol>
49 </li>
50 </ol>
51 </li>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]52 <li><a href="#constants">Constants</a>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]53 <ol>
54 <li><a href="#simpleconstants">Simple Constants</a>
55 <li><a href="#aggregateconstants">Aggregate Constants</a>
56 <li><a href="#globalconstants">Global Variable and Function Addresses</a>
57 <li><a href="#undefvalues">Undefined Values</a>
58 <li><a href="#constantexprs">Constant Expressions</a>
59 </ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]60 </li>
Chris Lattner98f013c2006-01-25 23:47:57 +0000[diff] [blame]61 <li><a href="#othervalues">Other Values</a>
62 <ol>
63 <li><a href="#inlineasm">Inline Assembler Expressions</a>
64 </ol>
65 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]66 <li><a href="#instref">Instruction Reference</a>
67 <ol>
68 <li><a href="#terminators">Terminator Instructions</a>
69 <ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]70 <li><a href="#i_ret">'<tt>ret</tt>' Instruction</a></li>
71 <li><a href="#i_br">'<tt>br</tt>' Instruction</a></li>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]72 <li><a href="#i_switch">'<tt>switch</tt>' Instruction</a></li>
73 <li><a href="#i_invoke">'<tt>invoke</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]74 <li><a href="#i_unwind">'<tt>unwind</tt>' Instruction</a></li>
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]75 <li><a href="#i_unreachable">'<tt>unreachable</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]76 </ol>
77 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]78 <li><a href="#binaryops">Binary Operations</a>
79 <ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]80 <li><a href="#i_add">'<tt>add</tt>' Instruction</a></li>
81 <li><a href="#i_sub">'<tt>sub</tt>' Instruction</a></li>
82 <li><a href="#i_mul">'<tt>mul</tt>' Instruction</a></li>
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame]83 <li><a href="#i_udiv">'<tt>udiv</tt>' Instruction</a></li>
84 <li><a href="#i_sdiv">'<tt>sdiv</tt>' Instruction</a></li>
85 <li><a href="#i_fdiv">'<tt>fdiv</tt>' Instruction</a></li>
Reid Spencer7eb55b32006-11-02 01:53:59 +0000[diff] [blame]86 <li><a href="#i_urem">'<tt>urem</tt>' Instruction</a></li>
87 <li><a href="#i_srem">'<tt>srem</tt>' Instruction</a></li>
88 <li><a href="#i_frem">'<tt>frem</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]89 </ol>
90 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]91 <li><a href="#bitwiseops">Bitwise Binary Operations</a>
92 <ol>
Reid Spencer2ab01932007-02-02 13:57:07 +0000[diff] [blame]93 <li><a href="#i_shl">'<tt>shl</tt>' Instruction</a></li>
94 <li><a href="#i_lshr">'<tt>lshr</tt>' Instruction</a></li>
95 <li><a href="#i_ashr">'<tt>ashr</tt>' Instruction</a></li>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]96 <li><a href="#i_and">'<tt>and</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]97 <li><a href="#i_or">'<tt>or</tt>' Instruction</a></li>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]98 <li><a href="#i_xor">'<tt>xor</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]99 </ol>
100 </li>
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]101 <li><a href="#vectorops">Vector Operations</a>
102 <ol>
103 <li><a href="#i_extractelement">'<tt>extractelement</tt>' Instruction</a></li>
104 <li><a href="#i_insertelement">'<tt>insertelement</tt>' Instruction</a></li>
105 <li><a href="#i_shufflevector">'<tt>shufflevector</tt>' Instruction</a></li>
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]106 </ol>
107 </li>
Chris Lattner6ab66722006-08-15 00:45:58 +0000[diff] [blame]108 <li><a href="#memoryops">Memory Access and Addressing Operations</a>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]109 <ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]110 <li><a href="#i_malloc">'<tt>malloc</tt>' Instruction</a></li>
111 <li><a href="#i_free">'<tt>free</tt>' Instruction</a></li>
112 <li><a href="#i_alloca">'<tt>alloca</tt>' Instruction</a></li>
Robert Bocchino820bc75b2006-02-17 21:18:08 +0000[diff] [blame]113 <li><a href="#i_load">'<tt>load</tt>' Instruction</a></li>
114 <li><a href="#i_store">'<tt>store</tt>' Instruction</a></li>
115 <li><a href="#i_getelementptr">'<tt>getelementptr</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]116 </ol>
117 </li>
Reid Spencer97c5fa42006-11-08 01:18:52 +0000[diff] [blame]118 <li><a href="#convertops">Conversion Operations</a>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]119 <ol>
120 <li><a href="#i_trunc">'<tt>trunc .. to</tt>' Instruction</a></li>
121 <li><a href="#i_zext">'<tt>zext .. to</tt>' Instruction</a></li>
122 <li><a href="#i_sext">'<tt>sext .. to</tt>' Instruction</a></li>
123 <li><a href="#i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a></li>
124 <li><a href="#i_fpext">'<tt>fpext .. to</tt>' Instruction</a></li>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]125 <li><a href="#i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a></li>
126 <li><a href="#i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a></li>
127 <li><a href="#i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a></li>
128 <li><a href="#i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a></li>
Reid Spencerb7344ff2006-11-11 21:00:47 +0000[diff] [blame]129 <li><a href="#i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a></li>
130 <li><a href="#i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a></li>
Reid Spencer5b950642006-11-11 23:08:07 +0000[diff] [blame]131 <li><a href="#i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a></li>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]132 </ol>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]133 <li><a href="#otherops">Other Operations</a>
134 <ol>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]135 <li><a href="#i_icmp">'<tt>icmp</tt>' Instruction</a></li>
136 <li><a href="#i_fcmp">'<tt>fcmp</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]137 <li><a href="#i_phi">'<tt>phi</tt>' Instruction</a></li>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]138 <li><a href="#i_select">'<tt>select</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]139 <li><a href="#i_call">'<tt>call</tt>' Instruction</a></li>
Chris Lattner33337472006-01-13 23:26:01 +0000[diff] [blame]140 <li><a href="#i_va_arg">'<tt>va_arg</tt>' Instruction</a></li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]141 </ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]142 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]143 </ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]144 </li>
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]145 <li><a href="#intrinsics">Intrinsic Functions</a>
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]146 <ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]147 <li><a href="#int_varargs">Variable Argument Handling Intrinsics</a>
148 <ol>
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]149 <li><a href="#int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a></li>
150 <li><a href="#int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a></li>
151 <li><a href="#int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]152 </ol>
153 </li>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]154 <li><a href="#int_gc">Accurate Garbage Collection Intrinsics</a>
155 <ol>
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]156 <li><a href="#int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a></li>
157 <li><a href="#int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a></li>
158 <li><a href="#int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a></li>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]159 </ol>
160 </li>
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]161 <li><a href="#int_codegen">Code Generator Intrinsics</a>
162 <ol>
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]163 <li><a href="#int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a></li>
164 <li><a href="#int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a></li>
165 <li><a href="#int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a></li>
166 <li><a href="#int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a></li>
167 <li><a href="#int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a></li>
168 <li><a href="#int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a></li>
169 <li><a href="#int_readcyclecounter"><tt>llvm.readcyclecounter</tt>' Intrinsic</a></li>
John Criswellaa1c3c12004-04-09 16:43:20 +0000[diff] [blame]170 </ol>
171 </li>
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]172 <li><a href="#int_libc">Standard C Library Intrinsics</a>
173 <ol>
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]174 <li><a href="#int_memcpy">'<tt>llvm.memcpy.*</tt>' Intrinsic</a></li>
175 <li><a href="#int_memmove">'<tt>llvm.memmove.*</tt>' Intrinsic</a></li>
176 <li><a href="#int_memset">'<tt>llvm.memset.*</tt>' Intrinsic</a></li>
177 <li><a href="#int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a></li>
178 <li><a href="#int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a></li>
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]179 </ol>
180 </li>
Nate Begeman0f223bb2006-01-13 23:26:38 +0000[diff] [blame]181 <li><a href="#int_manip">Bit Manipulation Intrinsics</a>
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]182 <ol>
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]183 <li><a href="#int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a></li>
Chris Lattnerb748c672006-01-16 22:34:14 +0000[diff] [blame]184 <li><a href="#int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic </a></li>
185 <li><a href="#int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic </a></li>
186 <li><a href="#int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic </a></li>
Reid Spencer5bf54c82007-04-11 23:23:49 +0000[diff] [blame]187 <li><a href="#int_part_select">'<tt>llvm.part.select.*</tt>' Intrinsic </a></li>
188 <li><a href="#int_part_set">'<tt>llvm.part.set.*</tt>' Intrinsic </a></li>
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]189 </ol>
190 </li>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]191 <li><a href="#int_debugger">Debugger intrinsics</a></li>
Jim Laskey2211f492007-03-14 19:31:19 +0000[diff] [blame]192 <li><a href="#int_eh">Exception Handling intrinsics</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]193 </ol>
194 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]195</ol>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]196
197<div class="doc_author">
198 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>
199 and <a href="mailto:vadve@cs.uiuc.edu">Vikram Adve</a></p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]200</div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]201
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]202<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]203<div class="doc_section"> <a name="abstract">Abstract </a></div>
204<!-- *********************************************************************** -->
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]205
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]206<div class="doc_text">
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]207<p>This document is a reference manual for the LLVM assembly language.
208LLVM is an SSA based representation that provides type safety,
209low-level operations, flexibility, and the capability of representing
210'all' high-level languages cleanly. It is the common code
211representation used throughout all phases of the LLVM compilation
212strategy.</p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]213</div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]214
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]215<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]216<div class="doc_section"> <a name="introduction">Introduction</a> </div>
217<!-- *********************************************************************** -->
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]218
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]219<div class="doc_text">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]220
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]221<p>The LLVM code representation is designed to be used in three
222different forms: as an in-memory compiler IR, as an on-disk bytecode
223representation (suitable for fast loading by a Just-In-Time compiler),
224and as a human readable assembly language representation. This allows
225LLVM to provide a powerful intermediate representation for efficient
226compiler transformations and analysis, while providing a natural means
227to debug and visualize the transformations. The three different forms
228of LLVM are all equivalent. This document describes the human readable
229representation and notation.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]230
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]231<p>The LLVM representation aims to be light-weight and low-level
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]232while being expressive, typed, and extensible at the same time. It
233aims to be a "universal IR" of sorts, by being at a low enough level
234that high-level ideas may be cleanly mapped to it (similar to how
235microprocessors are "universal IR's", allowing many source languages to
236be mapped to them). By providing type information, LLVM can be used as
237the target of optimizations: for example, through pointer analysis, it
238can be proven that a C automatic variable is never accessed outside of
239the current function... allowing it to be promoted to a simple SSA
240value instead of a memory location.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]241
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]242</div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]243
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]244<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]245<div class="doc_subsubsection"> <a name="wellformed">Well-Formedness</a> </div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]246
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]247<div class="doc_text">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]248
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]249<p>It is important to note that this document describes 'well formed'
250LLVM assembly language. There is a difference between what the parser
251accepts and what is considered 'well formed'. For example, the
252following instruction is syntactically okay, but not well formed:</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]253
254<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]255 %x = <a href="#i_add">add</a> i32 1, %x
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]256</pre>
257
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]258<p>...because the definition of <tt>%x</tt> does not dominate all of
259its uses. The LLVM infrastructure provides a verification pass that may
260be used to verify that an LLVM module is well formed. This pass is
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]261automatically run by the parser after parsing input assembly and by
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]262the optimizer before it outputs bytecode. The violations pointed out
263by the verifier pass indicate bugs in transformation passes or input to
264the parser.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]265
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]266<!-- Describe the typesetting conventions here. --> </div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]267
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]268<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]269<div class="doc_section"> <a name="identifiers">Identifiers</a> </div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]270<!-- *********************************************************************** -->
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]271
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]272<div class="doc_text">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]273
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]274<p>LLVM uses three different forms of identifiers, for different
275purposes:</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]276
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]277<ol>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]278 <li>Named values are represented as a string of characters with a '%' prefix.
279 For example, %foo, %DivisionByZero, %a.really.long.identifier. The actual
280 regular expression used is '<tt>%[a-zA-Z$._][a-zA-Z$._0-9]*</tt>'.
281 Identifiers which require other characters in their names can be surrounded
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]282 with quotes. In this way, anything except a <tt>&quot;</tt> character can be used
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]283 in a name.</li>
284
285 <li>Unnamed values are represented as an unsigned numeric value with a '%'
286 prefix. For example, %12, %2, %44.</li>
287
Reid Spencer8f08d802004-12-09 18:02:53 +0000[diff] [blame]288 <li>Constants, which are described in a <a href="#constants">section about
289 constants</a>, below.</li>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]290</ol>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]291
292<p>LLVM requires that values start with a '%' sign for two reasons: Compilers
293don't need to worry about name clashes with reserved words, and the set of
294reserved words may be expanded in the future without penalty. Additionally,
295unnamed identifiers allow a compiler to quickly come up with a temporary
296variable without having to avoid symbol table conflicts.</p>
297
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]298<p>Reserved words in LLVM are very similar to reserved words in other
Reid Spencer5b950642006-11-11 23:08:07 +0000[diff] [blame]299languages. There are keywords for different opcodes
300('<tt><a href="#i_add">add</a></tt>',
301 '<tt><a href="#i_bitcast">bitcast</a></tt>',
302 '<tt><a href="#i_ret">ret</a></tt>', etc...), for primitive type names ('<tt><a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]303href="#t_void">void</a></tt>', '<tt><a href="#t_primitive">i32</a></tt>', etc...),
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]304and others. These reserved words cannot conflict with variable names, because
305none of them start with a '%' character.</p>
306
307<p>Here is an example of LLVM code to multiply the integer variable
308'<tt>%X</tt>' by 8:</p>
309
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]310<p>The easy way:</p>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]311
312<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]313 %result = <a href="#i_mul">mul</a> i32 %X, 8
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]314</pre>
315
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]316<p>After strength reduction:</p>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]317
318<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]319 %result = <a href="#i_shl">shl</a> i32 %X, i8 3
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]320</pre>
321
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]322<p>And the hard way:</p>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]323
324<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]325 <a href="#i_add">add</a> i32 %X, %X <i>; yields {i32}:%0</i>
326 <a href="#i_add">add</a> i32 %0, %0 <i>; yields {i32}:%1</i>
327 %result = <a href="#i_add">add</a> i32 %1, %1
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]328</pre>
329
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]330<p>This last way of multiplying <tt>%X</tt> by 8 illustrates several
331important lexical features of LLVM:</p>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]332
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]333<ol>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]334
335 <li>Comments are delimited with a '<tt>;</tt>' and go until the end of
336 line.</li>
337
338 <li>Unnamed temporaries are created when the result of a computation is not
339 assigned to a named value.</li>
340
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]341 <li>Unnamed temporaries are numbered sequentially</li>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]342
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]343</ol>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]344
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]345<p>...and it also shows a convention that we follow in this document. When
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]346demonstrating instructions, we will follow an instruction with a comment that
347defines the type and name of value produced. Comments are shown in italic
348text.</p>
349
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]350</div>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]351
352<!-- *********************************************************************** -->
353<div class="doc_section"> <a name="highlevel">High Level Structure</a> </div>
354<!-- *********************************************************************** -->
355
356<!-- ======================================================================= -->
357<div class="doc_subsection"> <a name="modulestructure">Module Structure</a>
358</div>
359
360<div class="doc_text">
361
362<p>LLVM programs are composed of "Module"s, each of which is a
363translation unit of the input programs. Each module consists of
364functions, global variables, and symbol table entries. Modules may be
365combined together with the LLVM linker, which merges function (and
366global variable) definitions, resolves forward declarations, and merges
367symbol table entries. Here is an example of the "hello world" module:</p>
368
369<pre><i>; Declare the string constant as a global constant...</i>
370<a href="#identifiers">%.LC0</a> = <a href="#linkage_internal">internal</a> <a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]371 href="#globalvars">constant</a> <a href="#t_array">[13 x i8 ]</a> c"hello world\0A\00" <i>; [13 x i8 ]*</i>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]372
373<i>; External declaration of the puts function</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]374<a href="#functionstructure">declare</a> i32 %puts(i8 *) <i>; i32(i8 *)* </i>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]375
376<i>; Definition of main function</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]377define i32 %main() { <i>; i32()* </i>
378 <i>; Convert [13x i8 ]* to i8 *...</i>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]379 %cast210 = <a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]380 href="#i_getelementptr">getelementptr</a> [13 x i8 ]* %.LC0, i64 0, i64 0 <i>; i8 *</i>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]381
382 <i>; Call puts function to write out the string to stdout...</i>
383 <a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]384 href="#i_call">call</a> i32 %puts(i8 * %cast210) <i>; i32</i>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]385 <a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]386 href="#i_ret">ret</a> i32 0<br>}<br></pre>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]387
388<p>This example is made up of a <a href="#globalvars">global variable</a>
389named "<tt>.LC0</tt>", an external declaration of the "<tt>puts</tt>"
390function, and a <a href="#functionstructure">function definition</a>
391for "<tt>main</tt>".</p>
392
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]393<p>In general, a module is made up of a list of global values,
394where both functions and global variables are global values. Global values are
395represented by a pointer to a memory location (in this case, a pointer to an
396array of char, and a pointer to a function), and have one of the following <a
397href="#linkage">linkage types</a>.</p>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]398
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]399</div>
400
401<!-- ======================================================================= -->
402<div class="doc_subsection">
403 <a name="linkage">Linkage Types</a>
404</div>
405
406<div class="doc_text">
407
408<p>
409All Global Variables and Functions have one of the following types of linkage:
410</p>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]411
412<dl>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]413
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]414 <dt><tt><b><a name="linkage_internal">internal</a></b></tt> </dt>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]415
416 <dd>Global values with internal linkage are only directly accessible by
417 objects in the current module. In particular, linking code into a module with
418 an internal global value may cause the internal to be renamed as necessary to
419 avoid collisions. Because the symbol is internal to the module, all
420 references can be updated. This corresponds to the notion of the
Chris Lattnere20b4702007-01-14 06:51:48 +0000[diff] [blame]421 '<tt>static</tt>' keyword in C.
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]422 </dd>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]423
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]424 <dt><tt><b><a name="linkage_linkonce">linkonce</a></b></tt>: </dt>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]425
Chris Lattnere20b4702007-01-14 06:51:48 +0000[diff] [blame]426 <dd>Globals with "<tt>linkonce</tt>" linkage are merged with other globals of
427 the same name when linkage occurs. This is typically used to implement
428 inline functions, templates, or other code which must be generated in each
429 translation unit that uses it. Unreferenced <tt>linkonce</tt> globals are
430 allowed to be discarded.
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]431 </dd>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]432
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]433 <dt><tt><b><a name="linkage_weak">weak</a></b></tt>: </dt>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]434
435 <dd>"<tt>weak</tt>" linkage is exactly the same as <tt>linkonce</tt> linkage,
436 except that unreferenced <tt>weak</tt> globals may not be discarded. This is
Chris Lattnere20b4702007-01-14 06:51:48 +0000[diff] [blame]437 used for globals that may be emitted in multiple translation units, but that
438 are not guaranteed to be emitted into every translation unit that uses them.
439 One example of this are common globals in C, such as "<tt>int X;</tt>" at
440 global scope.
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]441 </dd>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]442
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]443 <dt><tt><b><a name="linkage_appending">appending</a></b></tt>: </dt>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]444
445 <dd>"<tt>appending</tt>" linkage may only be applied to global variables of
446 pointer to array type. When two global variables with appending linkage are
447 linked together, the two global arrays are appended together. This is the
448 LLVM, typesafe, equivalent of having the system linker append together
449 "sections" with identical names when .o files are linked.
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]450 </dd>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]451
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]452 <dt><tt><b><a name="linkage_externweak">extern_weak</a></b></tt>: </dt>
453 <dd>The semantics of this linkage follow the ELF model: the symbol is weak
454 until linked, if not linked, the symbol becomes null instead of being an
455 undefined reference.
456 </dd>
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]457
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]458 <dt><tt><b><a name="linkage_external">externally visible</a></b></tt>:</dt>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]459
460 <dd>If none of the above identifiers are used, the global is externally
461 visible, meaning that it participates in linkage and can be used to resolve
462 external symbol references.
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]463 </dd>
Reid Spencer7972c472007-04-11 23:49:50 +0000[diff] [blame]464</dl>
Anton Korobeynikovd61d39e2006-09-14 18:23:27 +0000[diff] [blame]465
Anton Korobeynikovd61d39e2006-09-14 18:23:27 +0000[diff] [blame]466 <p>
467 The next two types of linkage are targeted for Microsoft Windows platform
468 only. They are designed to support importing (exporting) symbols from (to)
469 DLLs.
470 </p>
471
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]472 <dl>
Anton Korobeynikovd61d39e2006-09-14 18:23:27 +0000[diff] [blame]473 <dt><tt><b><a name="linkage_dllimport">dllimport</a></b></tt>: </dt>
474
475 <dd>"<tt>dllimport</tt>" linkage causes the compiler to reference a function
476 or variable via a global pointer to a pointer that is set up by the DLL
477 exporting the symbol. On Microsoft Windows targets, the pointer name is
478 formed by combining <code>_imp__</code> and the function or variable name.
479 </dd>
480
481 <dt><tt><b><a name="linkage_dllexport">dllexport</a></b></tt>: </dt>
482
483 <dd>"<tt>dllexport</tt>" linkage causes the compiler to provide a global
484 pointer to a pointer in a DLL, so that it can be referenced with the
485 <tt>dllimport</tt> attribute. On Microsoft Windows targets, the pointer
486 name is formed by combining <code>_imp__</code> and the function or variable
487 name.
488 </dd>
489
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]490</dl>
491
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]492<p><a name="linkage_external"></a>For example, since the "<tt>.LC0</tt>"
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]493variable is defined to be internal, if another module defined a "<tt>.LC0</tt>"
494variable and was linked with this one, one of the two would be renamed,
495preventing a collision. Since "<tt>main</tt>" and "<tt>puts</tt>" are
496external (i.e., lacking any linkage declarations), they are accessible
Reid Spencer92c671e2007-01-05 00:59:10 +0000[diff] [blame]497outside of the current module.</p>
498<p>It is illegal for a function <i>declaration</i>
499to have any linkage type other than "externally visible", <tt>dllimport</tt>,
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]500or <tt>extern_weak</tt>.</p>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]501
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]502</div>
503
504<!-- ======================================================================= -->
505<div class="doc_subsection">
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]506 <a name="callingconv">Calling Conventions</a>
507</div>
508
509<div class="doc_text">
510
511<p>LLVM <a href="#functionstructure">functions</a>, <a href="#i_call">calls</a>
512and <a href="#i_invoke">invokes</a> can all have an optional calling convention
513specified for the call. The calling convention of any pair of dynamic
514caller/callee must match, or the behavior of the program is undefined. The
515following calling conventions are supported by LLVM, and more may be added in
516the future:</p>
517
518<dl>
519 <dt><b>"<tt>ccc</tt>" - The C calling convention</b>:</dt>
520
521 <dd>This calling convention (the default if no other calling convention is
522 specified) matches the target C calling conventions. This calling convention
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]523 supports varargs function calls and tolerates some mismatch in the declared
Reid Spencer72ba4992006-12-31 21:30:18 +0000[diff] [blame]524 prototype and implemented declaration of the function (as does normal C).
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]525 </dd>
526
527 <dt><b>"<tt>fastcc</tt>" - The fast calling convention</b>:</dt>
528
529 <dd>This calling convention attempts to make calls as fast as possible
530 (e.g. by passing things in registers). This calling convention allows the
531 target to use whatever tricks it wants to produce fast code for the target,
Chris Lattnerc792eb32005-05-06 23:08:23 +0000[diff] [blame]532 without having to conform to an externally specified ABI. Implementations of
533 this convention should allow arbitrary tail call optimization to be supported.
534 This calling convention does not support varargs and requires the prototype of
535 all callees to exactly match the prototype of the function definition.
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]536 </dd>
537
538 <dt><b>"<tt>coldcc</tt>" - The cold calling convention</b>:</dt>
539
540 <dd>This calling convention attempts to make code in the caller as efficient
541 as possible under the assumption that the call is not commonly executed. As
542 such, these calls often preserve all registers so that the call does not break
543 any live ranges in the caller side. This calling convention does not support
544 varargs and requires the prototype of all callees to exactly match the
545 prototype of the function definition.
546 </dd>
547
Chris Lattner573f64e2005-05-07 01:46:40 +0000[diff] [blame]548 <dt><b>"<tt>cc &lt;<em>n</em>&gt;</tt>" - Numbered convention</b>:</dt>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]549
550 <dd>Any calling convention may be specified by number, allowing
551 target-specific calling conventions to be used. Target specific calling
552 conventions start at 64.
553 </dd>
Chris Lattner573f64e2005-05-07 01:46:40 +0000[diff] [blame]554</dl>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]555
556<p>More calling conventions can be added/defined on an as-needed basis, to
557support pascal conventions or any other well-known target-independent
558convention.</p>
559
560</div>
561
562<!-- ======================================================================= -->
563<div class="doc_subsection">
Anton Korobeynikovc7f9f3d2007-01-23 12:35:46 +0000[diff] [blame]564 <a name="visibility">Visibility Styles</a>
565</div>
566
567<div class="doc_text">
568
569<p>
570All Global Variables and Functions have one of the following visibility styles:
571</p>
572
573<dl>
574 <dt><b>"<tt>default</tt>" - Default style</b>:</dt>
575
576 <dd>On ELF, default visibility means that the declaration is visible to other
577 modules and, in shared libraries, means that the declared entity may be
578 overridden. On Darwin, default visibility means that the declaration is
579 visible to other modules. Default visibility corresponds to "external
580 linkage" in the language.
581 </dd>
582
583 <dt><b>"<tt>hidden</tt>" - Hidden style</b>:</dt>
584
585 <dd>Two declarations of an object with hidden visibility refer to the same
586 object if they are in the same shared object. Usually, hidden visibility
587 indicates that the symbol will not be placed into the dynamic symbol table,
588 so no other module (executable or shared library) can reference it
589 directly.
590 </dd>
591
592</dl>
593
594</div>
595
596<!-- ======================================================================= -->
597<div class="doc_subsection">
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]598 <a name="globalvars">Global Variables</a>
599</div>
600
601<div class="doc_text">
602
Chris Lattner5d5aede2005-02-12 19:30:21 +0000[diff] [blame]603<p>Global variables define regions of memory allocated at compilation time
Chris Lattner662c8722005-11-12 00:45:07 +0000[diff] [blame]604instead of run-time. Global variables may optionally be initialized, may have
605an explicit section to be placed in, and may
Lauro Ramos Venancioc451ec62007-04-12 20:34:36 +0000[diff] [blame]606have an optional explicit alignment specified. A variable may be defined as
607"thread_local", which means that it will not be shared by threads (each thread
608will have a separated copy of the variable).
609A variable may be defined as a global "constant," which indicates that the
Chris Lattner5d5aede2005-02-12 19:30:21 +0000[diff] [blame]610contents of the variable will <b>never</b> be modified (enabling better
611optimization, allowing the global data to be placed in the read-only section of
612an executable, etc). Note that variables that need runtime initialization
John Criswell4c0cf7f2005-10-24 16:17:18 +0000[diff] [blame]613cannot be marked "constant" as there is a store to the variable.</p>
Chris Lattner5d5aede2005-02-12 19:30:21 +0000[diff] [blame]614
615<p>
616LLVM explicitly allows <em>declarations</em> of global variables to be marked
617constant, even if the final definition of the global is not. This capability
618can be used to enable slightly better optimization of the program, but requires
619the language definition to guarantee that optimizations based on the
620'constantness' are valid for the translation units that do not include the
621definition.
622</p>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]623
624<p>As SSA values, global variables define pointer values that are in
625scope (i.e. they dominate) all basic blocks in the program. Global
626variables always define a pointer to their "content" type because they
627describe a region of memory, and all memory objects in LLVM are
628accessed through pointers.</p>
629
Chris Lattner662c8722005-11-12 00:45:07 +0000[diff] [blame]630<p>LLVM allows an explicit section to be specified for globals. If the target
631supports it, it will emit globals to the section specified.</p>
632
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]633<p>An explicit alignment may be specified for a global. If not present, or if
634the alignment is set to zero, the alignment of the global is set by the target
635to whatever it feels convenient. If an explicit alignment is specified, the
636global is forced to have at least that much alignment. All alignments must be
637a power of 2.</p>
638
Chris Lattner5760c502007-01-14 00:27:09 +0000[diff] [blame]639<p>For example, the following defines a global with an initializer, section,
640 and alignment:</p>
641
642<pre>
643 %G = constant float 1.0, section "foo", align 4
644</pre>
645
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]646</div>
647
648
649<!-- ======================================================================= -->
650<div class="doc_subsection">
651 <a name="functionstructure">Functions</a>
652</div>
653
654<div class="doc_text">
655
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]656<p>LLVM function definitions consist of the "<tt>define</tt>" keyord,
657an optional <a href="#linkage">linkage type</a>, an optional
Anton Korobeynikovc7f9f3d2007-01-23 12:35:46 +0000[diff] [blame]658<a href="#visibility">visibility style</a>, an optional
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]659<a href="#callingconv">calling convention</a>, a return type, an optional
660<a href="#paramattrs">parameter attribute</a> for the return type, a function
661name, a (possibly empty) argument list (each with optional
Anton Korobeynikovc7f9f3d2007-01-23 12:35:46 +0000[diff] [blame]662<a href="#paramattrs">parameter attributes</a>), an optional section, an
663optional alignment, an opening curly brace, a list of basic blocks, and a
664closing curly brace.
665
666LLVM function declarations consist of the "<tt>declare</tt>" keyword, an
667optional <a href="#linkage">linkage type</a>, an optional
668<a href="#visibility">visibility style</a>, an optional
669<a href="#callingconv">calling convention</a>, a return type, an optional
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]670<a href="#paramattrs">parameter attribute</a> for the return type, a function
671name, a possibly empty list of arguments, and an optional alignment.</p>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]672
673<p>A function definition contains a list of basic blocks, forming the CFG for
674the function. Each basic block may optionally start with a label (giving the
675basic block a symbol table entry), contains a list of instructions, and ends
676with a <a href="#terminators">terminator</a> instruction (such as a branch or
677function return).</p>
678
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]679<p>The first basic block in a program is special in two ways: it is immediately
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]680executed on entrance to the function, and it is not allowed to have predecessor
681basic blocks (i.e. there can not be any branches to the entry block of a
682function). Because the block can have no predecessors, it also cannot have any
683<a href="#i_phi">PHI nodes</a>.</p>
684
685<p>LLVM functions are identified by their name and type signature. Hence, two
686functions with the same name but different parameter lists or return values are
Chris Lattner455fc8c2005-03-07 22:13:59 +0000[diff] [blame]687considered different functions, and LLVM will resolve references to each
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]688appropriately.</p>
689
Chris Lattner662c8722005-11-12 00:45:07 +0000[diff] [blame]690<p>LLVM allows an explicit section to be specified for functions. If the target
691supports it, it will emit functions to the section specified.</p>
692
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]693<p>An explicit alignment may be specified for a function. If not present, or if
694the alignment is set to zero, the alignment of the function is set by the target
695to whatever it feels convenient. If an explicit alignment is specified, the
696function is forced to have at least that much alignment. All alignments must be
697a power of 2.</p>
698
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]699</div>
700
Chris Lattner91c15c42006-01-23 23:23:47 +0000[diff] [blame]701<!-- ======================================================================= -->
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]702<div class="doc_subsection"><a name="paramattrs">Parameter Attributes</a></div>
703<div class="doc_text">
704 <p>The return type and each parameter of a function type may have a set of
705 <i>parameter attributes</i> associated with them. Parameter attributes are
706 used to communicate additional information about the result or parameters of
707 a function. Parameter attributes are considered to be part of the function
708 type so two functions types that differ only by the parameter attributes
709 are different function types.</p>
710
Reid Spencercf7ebf52007-01-15 18:27:39 +0000[diff] [blame]711 <p>Parameter attributes are simple keywords that follow the type specified. If
712 multiple parameter attributes are needed, they are space separated. For
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]713 example:</p><pre>
Reid Spencercf7ebf52007-01-15 18:27:39 +0000[diff] [blame]714 %someFunc = i16 (i8 sext %someParam) zext
715 %someFunc = i16 (i8 zext %someParam) zext</pre>
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]716 <p>Note that the two function types above are unique because the parameter has
Reid Spencercf7ebf52007-01-15 18:27:39 +0000[diff] [blame]717 a different attribute (sext in the first one, zext in the second). Also note
718 that the attribute for the function result (zext) comes immediately after the
719 argument list.</p>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]720
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]721 <p>Currently, only the following parameter attributes are defined:</p>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]722 <dl>
Reid Spencercf7ebf52007-01-15 18:27:39 +0000[diff] [blame]723 <dt><tt>zext</tt></dt>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]724 <dd>This indicates that the parameter should be zero extended just before
725 a call to this function.</dd>
Reid Spencercf7ebf52007-01-15 18:27:39 +0000[diff] [blame]726 <dt><tt>sext</tt></dt>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]727 <dd>This indicates that the parameter should be sign extended just before
728 a call to this function.</dd>
Anton Korobeynikove8166852007-01-28 14:30:45 +0000[diff] [blame]729 <dt><tt>inreg</tt></dt>
730 <dd>This indicates that the parameter should be placed in register (if
Anton Korobeynikove93c6e82007-01-28 15:27:21 +0000[diff] [blame]731 possible) during assembling function call. Support for this attribute is
732 target-specific</dd>
Anton Korobeynikove8166852007-01-28 14:30:45 +0000[diff] [blame]733 <dt><tt>sret</tt></dt>
Anton Korobeynikove93c6e82007-01-28 15:27:21 +0000[diff] [blame]734 <dd>This indicates that the parameter specifies the address of a structure
Reid Spencer05dbb9d2007-03-22 02:02:11 +0000[diff] [blame]735 that is the return value of the function in the source program.</dd>
Reid Spencer9d1700e2007-03-22 02:18:56 +0000[diff] [blame]736 <dt><tt>noreturn</tt></dt>
737 <dd>This function attribute indicates that the function never returns. This
738 indicates to LLVM that every call to this function should be treated as if
739 an <tt>unreachable</tt> instruction immediately followed the call.</dd>
Reid Spencer05dbb9d2007-03-22 02:02:11 +0000[diff] [blame]740 <dt><tt>nounwind</tt></dt>
741 <dd>This function attribute indicates that the function type does not use
742 the unwind instruction and does not allow stack unwinding to propagate
743 through it.</dd>
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]744 </dl>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]745
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]746</div>
747
748<!-- ======================================================================= -->
Chris Lattner91c15c42006-01-23 23:23:47 +0000[diff] [blame]749<div class="doc_subsection">
Chris Lattner93564892006-04-08 04:40:53 +0000[diff] [blame]750 <a name="moduleasm">Module-Level Inline Assembly</a>
Chris Lattner91c15c42006-01-23 23:23:47 +0000[diff] [blame]751</div>
752
753<div class="doc_text">
754<p>
755Modules may contain "module-level inline asm" blocks, which corresponds to the
756GCC "file scope inline asm" blocks. These blocks are internally concatenated by
757LLVM and treated as a single unit, but may be separated in the .ll file if
758desired. The syntax is very simple:
759</p>
760
761<div class="doc_code"><pre>
Chris Lattnera1280ad2006-01-24 00:37:20 +0000[diff] [blame]762 module asm "inline asm code goes here"
763 module asm "more can go here"
Chris Lattner91c15c42006-01-23 23:23:47 +0000[diff] [blame]764</pre></div>
765
766<p>The strings can contain any character by escaping non-printable characters.
767 The escape sequence used is simply "\xx" where "xx" is the two digit hex code
768 for the number.
769</p>
770
771<p>
772 The inline asm code is simply printed to the machine code .s file when
773 assembly code is generated.
774</p>
775</div>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]776
Reid Spencer50c723a2007-02-19 23:54:10 +0000[diff] [blame]777<!-- ======================================================================= -->
778<div class="doc_subsection">
779 <a name="datalayout">Data Layout</a>
780</div>
781
782<div class="doc_text">
783<p>A module may specify a target specific data layout string that specifies how
Reid Spencer7972c472007-04-11 23:49:50 +0000[diff] [blame]784data is to be laid out in memory. The syntax for the data layout is simply:</p>
785<pre> target datalayout = "<i>layout specification</i>"</pre>
786<p>The <i>layout specification</i> consists of a list of specifications
787separated by the minus sign character ('-'). Each specification starts with a
788letter and may include other information after the letter to define some
789aspect of the data layout. The specifications accepted are as follows: </p>
Reid Spencer50c723a2007-02-19 23:54:10 +0000[diff] [blame]790<dl>
791 <dt><tt>E</tt></dt>
792 <dd>Specifies that the target lays out data in big-endian form. That is, the
793 bits with the most significance have the lowest address location.</dd>
794 <dt><tt>e</tt></dt>
795 <dd>Specifies that hte target lays out data in little-endian form. That is,
796 the bits with the least significance have the lowest address location.</dd>
797 <dt><tt>p:<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
798 <dd>This specifies the <i>size</i> of a pointer and its <i>abi</i> and
799 <i>preferred</i> alignments. All sizes are in bits. Specifying the <i>pref</i>
800 alignment is optional. If omitted, the preceding <tt>:</tt> should be omitted
801 too.</dd>
802 <dt><tt>i<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
803 <dd>This specifies the alignment for an integer type of a given bit
804 <i>size</i>. The value of <i>size</i> must be in the range [1,2^23).</dd>
805 <dt><tt>v<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
806 <dd>This specifies the alignment for a vector type of a given bit
807 <i>size</i>.</dd>
808 <dt><tt>f<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
809 <dd>This specifies the alignment for a floating point type of a given bit
810 <i>size</i>. The value of <i>size</i> must be either 32 (float) or 64
811 (double).</dd>
812 <dt><tt>a<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
813 <dd>This specifies the alignment for an aggregate type of a given bit
814 <i>size</i>.</dd>
815</dl>
816<p>When constructing the data layout for a given target, LLVM starts with a
817default set of specifications which are then (possibly) overriden by the
818specifications in the <tt>datalayout</tt> keyword. The default specifications
819are given in this list:</p>
820<ul>
821 <li><tt>E</tt> - big endian</li>
822 <li><tt>p:32:64:64</tt> - 32-bit pointers with 64-bit alignment</li>
823 <li><tt>i1:8:8</tt> - i1 is 8-bit (byte) aligned</li>
824 <li><tt>i8:8:8</tt> - i8 is 8-bit (byte) aligned</li>
825 <li><tt>i16:16:16</tt> - i16 is 16-bit aligned</li>
826 <li><tt>i32:32:32</tt> - i32 is 32-bit aligned</li>
827 <li><tt>i64:32:64</tt> - i64 has abi alignment of 32-bits but preferred
828 alignment of 64-bits</li>
829 <li><tt>f32:32:32</tt> - float is 32-bit aligned</li>
830 <li><tt>f64:64:64</tt> - double is 64-bit aligned</li>
831 <li><tt>v64:64:64</tt> - 64-bit vector is 64-bit aligned</li>
832 <li><tt>v128:128:128</tt> - 128-bit vector is 128-bit aligned</li>
833 <li><tt>a0:0:1</tt> - aggregates are 8-bit aligned</li>
834</ul>
835<p>When llvm is determining the alignment for a given type, it uses the
836following rules:
837<ol>
838 <li>If the type sought is an exact match for one of the specifications, that
839 specification is used.</li>
840 <li>If no match is found, and the type sought is an integer type, then the
841 smallest integer type that is larger than the bitwidth of the sought type is
842 used. If none of the specifications are larger than the bitwidth then the the
843 largest integer type is used. For example, given the default specifications
844 above, the i7 type will use the alignment of i8 (next largest) while both
845 i65 and i256 will use the alignment of i64 (largest specified).</li>
846 <li>If no match is found, and the type sought is a vector type, then the
847 largest vector type that is smaller than the sought vector type will be used
848 as a fall back. This happens because <128 x double> can be implemented in
849 terms of 64 <2 x double>, for example.</li>
850</ol>
851</div>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]852
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]853<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]854<div class="doc_section"> <a name="typesystem">Type System</a> </div>
855<!-- *********************************************************************** -->
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]856
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]857<div class="doc_text">
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]858
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]859<p>The LLVM type system is one of the most important features of the
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]860intermediate representation. Being typed enables a number of
861optimizations to be performed on the IR directly, without having to do
862extra analyses on the side before the transformation. A strong type
863system makes it easier to read the generated code and enables novel
864analyses and transformations that are not feasible to perform on normal
865three address code representations.</p>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]866
867</div>
868
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]869<!-- ======================================================================= -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]870<div class="doc_subsection"> <a name="t_primitive">Primitive Types</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]871<div class="doc_text">
John Criswell417228d2004-04-09 16:48:45 +0000[diff] [blame]872<p>The primitive types are the fundamental building blocks of the LLVM
Chris Lattner455fc8c2005-03-07 22:13:59 +0000[diff] [blame]873system. The current set of primitive types is as follows:</p>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]874
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]875<table class="layout">
876 <tr class="layout">
877 <td class="left">
878 <table>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]879 <tbody>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]880 <tr><th>Type</th><th>Description</th></tr>
Duncan Sands16f122e2007-03-30 12:22:09 +0000[diff] [blame]881 <tr><td><tt><a name="t_void">void</a></tt></td><td>No value</td></tr>
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]882 <tr><td><tt>i8</tt></td><td>8-bit value</td></tr>
883 <tr><td><tt>i32</tt></td><td>32-bit value</td></tr>
Misha Brukman36c6bc12005-04-22 18:02:52 +0000[diff] [blame]884 <tr><td><tt>float</tt></td><td>32-bit floating point value</td></tr>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]885 <tr><td><tt>label</tt></td><td>Branch destination</td></tr>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]886 </tbody>
887 </table>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]888 </td>
889 <td class="right">
890 <table>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]891 <tbody>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]892 <tr><th>Type</th><th>Description</th></tr>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]893 <tr><td><tt>i1</tt></td><td>True or False value</td></tr>
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]894 <tr><td><tt>i16</tt></td><td>16-bit value</td></tr>
895 <tr><td><tt>i64</tt></td><td>64-bit value</td></tr>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]896 <tr><td><tt>double</tt></td><td>64-bit floating point value</td></tr>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]897 </tbody>
898 </table>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]899 </td>
900 </tr>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]901</table>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]902</div>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]903
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]904<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]905<div class="doc_subsubsection"> <a name="t_classifications">Type
906Classifications</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]907<div class="doc_text">
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]908<p>These different primitive types fall into a few useful
909classifications:</p>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]910
911<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]912 <tbody>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]913 <tr><th>Classification</th><th>Types</th></tr>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]914 <tr>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]915 <td><a name="t_integer">integer</a></td>
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]916 <td><tt>i1, i8, i16, i32, i64</tt></td>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]917 </tr>
918 <tr>
919 <td><a name="t_floating">floating point</a></td>
920 <td><tt>float, double</tt></td>
921 </tr>
922 <tr>
923 <td><a name="t_firstclass">first class</a></td>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]924 <td><tt>i1, i8, i16, i32, i64, float, double, <br/>
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]925 <a href="#t_pointer">pointer</a>,<a href="#t_vector">vector</a></tt>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]926 </td>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]927 </tr>
928 </tbody>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]929</table>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]930
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]931<p>The <a href="#t_firstclass">first class</a> types are perhaps the
932most important. Values of these types are the only ones which can be
933produced by instructions, passed as arguments, or used as operands to
934instructions. This means that all structures and arrays must be
935manipulated either by pointer or by component.</p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]936</div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]937
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]938<!-- ======================================================================= -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]939<div class="doc_subsection"> <a name="t_derived">Derived Types</a> </div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]940
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]941<div class="doc_text">
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]942
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]943<p>The real power in LLVM comes from the derived types in the system.
944This is what allows a programmer to represent arrays, functions,
945pointers, and other useful types. Note that these derived types may be
946recursive: For example, it is possible to have a two dimensional array.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]947
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]948</div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]949
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]950<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]951<div class="doc_subsubsection"> <a name="t_array">Array Type</a> </div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]952
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]953<div class="doc_text">
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]954
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]955<h5>Overview:</h5>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]956
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]957<p>The array type is a very simple derived type that arranges elements
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]958sequentially in memory. The array type requires a size (number of
959elements) and an underlying data type.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]960
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]961<h5>Syntax:</h5>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]962
963<pre>
964 [&lt;# elements&gt; x &lt;elementtype&gt;]
965</pre>
966
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]967<p>The number of elements is a constant integer value; elementtype may
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]968be any type with a size.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]969
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]970<h5>Examples:</h5>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]971<table class="layout">
972 <tr class="layout">
973 <td class="left">
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]974 <tt>[40 x i32 ]</tt><br/>
975 <tt>[41 x i32 ]</tt><br/>
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]976 <tt>[40 x i8]</tt><br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]977 </td>
978 <td class="left">
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]979 Array of 40 32-bit integer values.<br/>
980 Array of 41 32-bit integer values.<br/>
981 Array of 40 8-bit integer values.<br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]982 </td>
983 </tr>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]984</table>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]985<p>Here are some examples of multidimensional arrays:</p>
986<table class="layout">
987 <tr class="layout">
988 <td class="left">
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]989 <tt>[3 x [4 x i32]]</tt><br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]990 <tt>[12 x [10 x float]]</tt><br/>
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]991 <tt>[2 x [3 x [4 x i16]]]</tt><br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]992 </td>
993 <td class="left">
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]994 3x4 array of 32-bit integer values.<br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]995 12x10 array of single precision floating point values.<br/>
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]996 2x3x4 array of 16-bit integer values.<br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]997 </td>
998 </tr>
999</table>
Chris Lattnerc0ad71e2005-06-24 17:22:57 +0000[diff] [blame]1000
John Criswell4c0cf7f2005-10-24 16:17:18 +0000[diff] [blame]1001<p>Note that 'variable sized arrays' can be implemented in LLVM with a zero
1002length array. Normally, accesses past the end of an array are undefined in
Chris Lattnerc0ad71e2005-06-24 17:22:57 +0000[diff] [blame]1003LLVM (e.g. it is illegal to access the 5th element of a 3 element array).
1004As a special case, however, zero length arrays are recognized to be variable
1005length. This allows implementation of 'pascal style arrays' with the LLVM
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1006type "{ i32, [0 x float]}", for example.</p>
Chris Lattnerc0ad71e2005-06-24 17:22:57 +0000[diff] [blame]1007
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1008</div>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1009
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1010<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1011<div class="doc_subsubsection"> <a name="t_function">Function Type</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1012<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1013<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1014<p>The function type can be thought of as a function signature. It
1015consists of a return type and a list of formal parameter types.
John Criswella0d50d22003-11-25 21:45:46 +0000[diff] [blame]1016Function types are usually used to build virtual function tables
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1017(which are structures of pointers to functions), for indirect function
1018calls, and when defining a function.</p>
John Criswella0d50d22003-11-25 21:45:46 +0000[diff] [blame]1019<p>
1020The return type of a function type cannot be an aggregate type.
1021</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1022<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1023<pre> &lt;returntype&gt; (&lt;parameter list&gt;)<br></pre>
John Criswell4c0cf7f2005-10-24 16:17:18 +0000[diff] [blame]1024<p>...where '<tt>&lt;parameter list&gt;</tt>' is a comma-separated list of type
Misha Brukman20f9a622004-08-12 20:16:08 +0000[diff] [blame]1025specifiers. Optionally, the parameter list may include a type <tt>...</tt>,
Chris Lattner5ed60612003-09-03 00:41:47 +0000[diff] [blame]1026which indicates that the function takes a variable number of arguments.
1027Variable argument functions can access their arguments with the <a
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1028 href="#int_varargs">variable argument handling intrinsic</a> functions.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1029<h5>Examples:</h5>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1030<table class="layout">
1031 <tr class="layout">
Reid Spencer58c08712006-12-31 07:18:34 +0000[diff] [blame]1032 <td class="left"><tt>i32 (i32)</tt></td>
1033 <td class="left">function taking an <tt>i32</tt>, returning an <tt>i32</tt>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1034 </td>
Reid Spencer58c08712006-12-31 07:18:34 +0000[diff] [blame]1035 </tr><tr class="layout">
Reid Spencere6a338d2007-01-15 18:28:34 +0000[diff] [blame]1036 <td class="left"><tt>float&nbsp;(i16&nbsp;sext,&nbsp;i32&nbsp;*)&nbsp;*
Reid Spencer655dcc62006-12-31 07:20:23 +0000[diff] [blame]1037 </tt></td>
Reid Spencer58c08712006-12-31 07:18:34 +0000[diff] [blame]1038 <td class="left"><a href="#t_pointer">Pointer</a> to a function that takes
1039 an <tt>i16</tt> that should be sign extended and a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1040 <a href="#t_pointer">pointer</a> to <tt>i32</tt>, returning
Reid Spencer58c08712006-12-31 07:18:34 +0000[diff] [blame]1041 <tt>float</tt>.
1042 </td>
1043 </tr><tr class="layout">
1044 <td class="left"><tt>i32 (i8*, ...)</tt></td>
1045 <td class="left">A vararg function that takes at least one
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]1046 <a href="#t_pointer">pointer</a> to <tt>i8 </tt> (char in C),
Reid Spencer58c08712006-12-31 07:18:34 +0000[diff] [blame]1047 which returns an integer. This is the signature for <tt>printf</tt> in
1048 LLVM.
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1049 </td>
1050 </tr>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1051</table>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]1052
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1053</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1054<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1055<div class="doc_subsubsection"> <a name="t_struct">Structure Type</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1056<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1057<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1058<p>The structure type is used to represent a collection of data members
1059together in memory. The packing of the field types is defined to match
1060the ABI of the underlying processor. The elements of a structure may
1061be any type that has a size.</p>
1062<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
1063and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
1064field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
1065instruction.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1066<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1067<pre> { &lt;type list&gt; }<br></pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1068<h5>Examples:</h5>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1069<table class="layout">
1070 <tr class="layout">
Jeff Cohen5819f182007-04-22 01:17:39 +0000[diff] [blame]1071 <td class="left"><tt>{ i32, i32, i32 }</tt></td>
1072 <td class="left">A triple of three <tt>i32</tt> values</td>
1073 </tr><tr class="layout">
1074 <td class="left"><tt>{&nbsp;float,&nbsp;i32&nbsp;(i32)&nbsp;*&nbsp;}</tt></td>
1075 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1076 second element is a <a href="#t_pointer">pointer</a> to a
1077 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1078 an <tt>i32</tt>.</td>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1079 </tr>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1080</table>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1081</div>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1082
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1083<!-- _______________________________________________________________________ -->
Andrew Lenharth8df88e22006-12-08 17:13:00 +0000[diff] [blame]1084<div class="doc_subsubsection"> <a name="t_pstruct">Packed Structure Type</a>
1085</div>
1086<div class="doc_text">
1087<h5>Overview:</h5>
1088<p>The packed structure type is used to represent a collection of data members
1089together in memory. There is no padding between fields. Further, the alignment
1090of a packed structure is 1 byte. The elements of a packed structure may
1091be any type that has a size.</p>
1092<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
1093and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
1094field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
1095instruction.</p>
1096<h5>Syntax:</h5>
1097<pre> &lt; { &lt;type list&gt; } &gt; <br></pre>
1098<h5>Examples:</h5>
1099<table class="layout">
1100 <tr class="layout">
Jeff Cohen5819f182007-04-22 01:17:39 +0000[diff] [blame]1101 <td class="left"><tt>&lt; { i32, i32, i32 } &gt;</tt></td>
1102 <td class="left">A triple of three <tt>i32</tt> values</td>
1103 </tr><tr class="layout">
1104 <td class="left"><tt>&lt;&nbsp;{&nbsp;float,&nbsp;i32&nbsp;(i32)&nbsp;*&nbsp;}&nbsp;&gt;</tt></td>
1105 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1106 second element is a <a href="#t_pointer">pointer</a> to a
1107 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1108 an <tt>i32</tt>.</td>
Andrew Lenharth8df88e22006-12-08 17:13:00 +0000[diff] [blame]1109 </tr>
1110</table>
1111</div>
1112
1113<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1114<div class="doc_subsubsection"> <a name="t_pointer">Pointer Type</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1115<div class="doc_text">
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]1116<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1117<p>As in many languages, the pointer type represents a pointer or
1118reference to another object, which must live in memory.</p>
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]1119<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1120<pre> &lt;type&gt; *<br></pre>
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]1121<h5>Examples:</h5>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1122<table class="layout">
1123 <tr class="layout">
1124 <td class="left">
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1125 <tt>[4x i32]*</tt><br/>
1126 <tt>i32 (i32 *) *</tt><br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1127 </td>
1128 <td class="left">
1129 A <a href="#t_pointer">pointer</a> to <a href="#t_array">array</a> of
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1130 four <tt>i32</tt> values<br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1131 A <a href="#t_pointer">pointer</a> to a <a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1132 href="#t_function">function</a> that takes an <tt>i32*</tt>, returning an
1133 <tt>i32</tt>.<br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1134 </td>
1135 </tr>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1136</table>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1137</div>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1138
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1139<!-- _______________________________________________________________________ -->
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1140<div class="doc_subsubsection"> <a name="t_vector">Vector Type</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1141<div class="doc_text">
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]1142
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1143<h5>Overview:</h5>
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]1144
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1145<p>A vector type is a simple derived type that represents a vector
1146of elements. Vector types are used when multiple primitive data
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1147are operated in parallel using a single instruction (SIMD).
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1148A vector type requires a size (number of
Chris Lattner330ce692005-11-10 01:44:22 +0000[diff] [blame]1149elements) and an underlying primitive data type. Vectors must have a power
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1150of two length (1, 2, 4, 8, 16 ...). Vector types are
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1151considered <a href="#t_firstclass">first class</a>.</p>
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]1152
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1153<h5>Syntax:</h5>
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]1154
1155<pre>
1156 &lt; &lt;# elements&gt; x &lt;elementtype&gt; &gt;
1157</pre>
1158
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]1159<p>The number of elements is a constant integer value; elementtype may
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]1160be any integer or floating point type.</p>
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]1161
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1162<h5>Examples:</h5>
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]1163
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1164<table class="layout">
1165 <tr class="layout">
1166 <td class="left">
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1167 <tt>&lt;4 x i32&gt;</tt><br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1168 <tt>&lt;8 x float&gt;</tt><br/>
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]1169 <tt>&lt;2 x i64&gt;</tt><br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1170 </td>
1171 <td class="left">
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1172 Vector of 4 32-bit integer values.<br/>
1173 Vector of 8 floating-point values.<br/>
1174 Vector of 2 64-bit integer values.<br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1175 </td>
1176 </tr>
1177</table>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1178</div>
1179
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]1180<!-- _______________________________________________________________________ -->
1181<div class="doc_subsubsection"> <a name="t_opaque">Opaque Type</a> </div>
1182<div class="doc_text">
1183
1184<h5>Overview:</h5>
1185
1186<p>Opaque types are used to represent unknown types in the system. This
1187corresponds (for example) to the C notion of a foward declared structure type.
1188In LLVM, opaque types can eventually be resolved to any type (not just a
1189structure type).</p>
1190
1191<h5>Syntax:</h5>
1192
1193<pre>
1194 opaque
1195</pre>
1196
1197<h5>Examples:</h5>
1198
1199<table class="layout">
1200 <tr class="layout">
1201 <td class="left">
1202 <tt>opaque</tt>
1203 </td>
1204 <td class="left">
1205 An opaque type.<br/>
1206 </td>
1207 </tr>
1208</table>
1209</div>
1210
1211
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1212<!-- *********************************************************************** -->
1213<div class="doc_section"> <a name="constants">Constants</a> </div>
1214<!-- *********************************************************************** -->
1215
1216<div class="doc_text">
1217
1218<p>LLVM has several different basic types of constants. This section describes
1219them all and their syntax.</p>
1220
1221</div>
1222
1223<!-- ======================================================================= -->
Reid Spencer8f08d802004-12-09 18:02:53 +0000[diff] [blame]1224<div class="doc_subsection"><a name="simpleconstants">Simple Constants</a></div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1225
1226<div class="doc_text">
1227
1228<dl>
1229 <dt><b>Boolean constants</b></dt>
1230
1231 <dd>The two strings '<tt>true</tt>' and '<tt>false</tt>' are both valid
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]1232 constants of the <tt><a href="#t_primitive">i1</a></tt> type.
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1233 </dd>
1234
1235 <dt><b>Integer constants</b></dt>
1236
Reid Spencer8f08d802004-12-09 18:02:53 +0000[diff] [blame]1237 <dd>Standard integers (such as '4') are constants of the <a
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]1238 href="#t_integer">integer</a> type. Negative numbers may be used with
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1239 integer types.
1240 </dd>
1241
1242 <dt><b>Floating point constants</b></dt>
1243
1244 <dd>Floating point constants use standard decimal notation (e.g. 123.421),
1245 exponential notation (e.g. 1.23421e+2), or a more precise hexadecimal
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1246 notation (see below). Floating point constants must have a <a
1247 href="#t_floating">floating point</a> type. </dd>
1248
1249 <dt><b>Null pointer constants</b></dt>
1250
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]1251 <dd>The identifier '<tt>null</tt>' is recognized as a null pointer constant
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1252 and must be of <a href="#t_pointer">pointer type</a>.</dd>
1253
1254</dl>
1255
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]1256<p>The one non-intuitive notation for constants is the optional hexadecimal form
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1257of floating point constants. For example, the form '<tt>double
12580x432ff973cafa8000</tt>' is equivalent to (but harder to read than) '<tt>double
12594.5e+15</tt>'. The only time hexadecimal floating point constants are required
Reid Spencer8f08d802004-12-09 18:02:53 +0000[diff] [blame]1260(and the only time that they are generated by the disassembler) is when a
1261floating point constant must be emitted but it cannot be represented as a
1262decimal floating point number. For example, NaN's, infinities, and other
1263special values are represented in their IEEE hexadecimal format so that
1264assembly and disassembly do not cause any bits to change in the constants.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1265
1266</div>
1267
1268<!-- ======================================================================= -->
1269<div class="doc_subsection"><a name="aggregateconstants">Aggregate Constants</a>
1270</div>
1271
1272<div class="doc_text">
Chris Lattner455fc8c2005-03-07 22:13:59 +0000[diff] [blame]1273<p>Aggregate constants arise from aggregation of simple constants
1274and smaller aggregate constants.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1275
1276<dl>
1277 <dt><b>Structure constants</b></dt>
1278
1279 <dd>Structure constants are represented with notation similar to structure
1280 type definitions (a comma separated list of elements, surrounded by braces
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1281 (<tt>{}</tt>)). For example: "<tt>{ i32 4, float 17.0, i32* %G }</tt>",
1282 where "<tt>%G</tt>" is declared as "<tt>%G = external global i32</tt>". Structure constants
Chris Lattner455fc8c2005-03-07 22:13:59 +0000[diff] [blame]1283 must have <a href="#t_struct">structure type</a>, and the number and
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1284 types of elements must match those specified by the type.
1285 </dd>
1286
1287 <dt><b>Array constants</b></dt>
1288
1289 <dd>Array constants are represented with notation similar to array type
1290 definitions (a comma separated list of elements, surrounded by square brackets
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1291 (<tt>[]</tt>)). For example: "<tt>[ i32 42, i32 11, i32 74 ]</tt>". Array
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1292 constants must have <a href="#t_array">array type</a>, and the number and
1293 types of elements must match those specified by the type.
1294 </dd>
1295
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1296 <dt><b>Vector constants</b></dt>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1297
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1298 <dd>Vector constants are represented with notation similar to vector type
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1299 definitions (a comma separated list of elements, surrounded by
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1300 less-than/greater-than's (<tt>&lt;&gt;</tt>)). For example: "<tt>&lt; i32 42,
Jeff Cohen5819f182007-04-22 01:17:39 +0000[diff] [blame]1301 i32 11, i32 74, i32 100 &gt;</tt>". Vector constants must have <a
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1302 href="#t_vector">vector type</a>, and the number and types of elements must
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1303 match those specified by the type.
1304 </dd>
1305
1306 <dt><b>Zero initialization</b></dt>
1307
1308 <dd>The string '<tt>zeroinitializer</tt>' can be used to zero initialize a
1309 value to zero of <em>any</em> type, including scalar and aggregate types.
1310 This is often used to avoid having to print large zero initializers (e.g. for
John Criswell4c0cf7f2005-10-24 16:17:18 +0000[diff] [blame]1311 large arrays) and is always exactly equivalent to using explicit zero
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1312 initializers.
1313 </dd>
1314</dl>
1315
1316</div>
1317
1318<!-- ======================================================================= -->
1319<div class="doc_subsection">
1320 <a name="globalconstants">Global Variable and Function Addresses</a>
1321</div>
1322
1323<div class="doc_text">
1324
1325<p>The addresses of <a href="#globalvars">global variables</a> and <a
1326href="#functionstructure">functions</a> are always implicitly valid (link-time)
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]1327constants. These constants are explicitly referenced when the <a
1328href="#identifiers">identifier for the global</a> is used and always have <a
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1329href="#t_pointer">pointer</a> type. For example, the following is a legal LLVM
1330file:</p>
1331
1332<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1333 %X = global i32 17
1334 %Y = global i32 42
1335 %Z = global [2 x i32*] [ i32* %X, i32* %Y ]
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1336</pre>
1337
1338</div>
1339
1340<!-- ======================================================================= -->
Reid Spencer641f5c92004-12-09 18:13:12 +0000[diff] [blame]1341<div class="doc_subsection"><a name="undefvalues">Undefined Values</a></div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1342<div class="doc_text">
Reid Spencer641f5c92004-12-09 18:13:12 +0000[diff] [blame]1343 <p>The string '<tt>undef</tt>' is recognized as a type-less constant that has
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]1344 no specific value. Undefined values may be of any type and be used anywhere
Reid Spencer641f5c92004-12-09 18:13:12 +0000[diff] [blame]1345 a constant is permitted.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1346
Reid Spencer641f5c92004-12-09 18:13:12 +0000[diff] [blame]1347 <p>Undefined values indicate to the compiler that the program is well defined
1348 no matter what value is used, giving the compiler more freedom to optimize.
1349 </p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1350</div>
1351
1352<!-- ======================================================================= -->
1353<div class="doc_subsection"><a name="constantexprs">Constant Expressions</a>
1354</div>
1355
1356<div class="doc_text">
1357
1358<p>Constant expressions are used to allow expressions involving other constants
1359to be used as constants. Constant expressions may be of any <a
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]1360href="#t_firstclass">first class</a> type and may involve any LLVM operation
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1361that does not have side effects (e.g. load and call are not supported). The
1362following is the syntax for constant expressions:</p>
1363
1364<dl>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]1365 <dt><b><tt>trunc ( CST to TYPE )</tt></b></dt>
1366 <dd>Truncate a constant to another type. The bit size of CST must be larger
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]1367 than the bit size of TYPE. Both types must be integers.</dd>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1368
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]1369 <dt><b><tt>zext ( CST to TYPE )</tt></b></dt>
1370 <dd>Zero extend a constant to another type. The bit size of CST must be
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]1371 smaller or equal to the bit size of TYPE. Both types must be integers.</dd>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]1372
1373 <dt><b><tt>sext ( CST to TYPE )</tt></b></dt>
1374 <dd>Sign extend a constant to another type. The bit size of CST must be
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]1375 smaller or equal to the bit size of TYPE. Both types must be integers.</dd>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]1376
1377 <dt><b><tt>fptrunc ( CST to TYPE )</tt></b></dt>
1378 <dd>Truncate a floating point constant to another floating point type. The
1379 size of CST must be larger than the size of TYPE. Both types must be
1380 floating point.</dd>
1381
1382 <dt><b><tt>fpext ( CST to TYPE )</tt></b></dt>
1383 <dd>Floating point extend a constant to another type. The size of CST must be
1384 smaller or equal to the size of TYPE. Both types must be floating point.</dd>
1385
1386 <dt><b><tt>fp2uint ( CST to TYPE )</tt></b></dt>
1387 <dd>Convert a floating point constant to the corresponding unsigned integer
1388 constant. TYPE must be an integer type. CST must be floating point. If the
1389 value won't fit in the integer type, the results are undefined.</dd>
1390
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]1391 <dt><b><tt>fptosi ( CST to TYPE )</tt></b></dt>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]1392 <dd>Convert a floating point constant to the corresponding signed integer
1393 constant. TYPE must be an integer type. CST must be floating point. If the
1394 value won't fit in the integer type, the results are undefined.</dd>
1395
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]1396 <dt><b><tt>uitofp ( CST to TYPE )</tt></b></dt>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]1397 <dd>Convert an unsigned integer constant to the corresponding floating point
1398 constant. TYPE must be floating point. CST must be of integer type. If the
Jeff Cohenbeccb742007-04-22 14:56:37 +0000[diff] [blame^]1399 value won't fit in the floating point type, the results are undefined.</dd>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]1400
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]1401 <dt><b><tt>sitofp ( CST to TYPE )</tt></b></dt>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]1402 <dd>Convert a signed integer constant to the corresponding floating point
1403 constant. TYPE must be floating point. CST must be of integer type. If the
Jeff Cohenbeccb742007-04-22 14:56:37 +0000[diff] [blame^]1404 value won't fit in the floating point type, the results are undefined.</dd>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]1405
Reid Spencer5b950642006-11-11 23:08:07 +0000[diff] [blame]1406 <dt><b><tt>ptrtoint ( CST to TYPE )</tt></b></dt>
1407 <dd>Convert a pointer typed constant to the corresponding integer constant
1408 TYPE must be an integer type. CST must be of pointer type. The CST value is
1409 zero extended, truncated, or unchanged to make it fit in TYPE.</dd>
1410
1411 <dt><b><tt>inttoptr ( CST to TYPE )</tt></b></dt>
1412 <dd>Convert a integer constant to a pointer constant. TYPE must be a
1413 pointer type. CST must be of integer type. The CST value is zero extended,
1414 truncated, or unchanged to make it fit in a pointer size. This one is
1415 <i>really</i> dangerous!</dd>
1416
1417 <dt><b><tt>bitcast ( CST to TYPE )</tt></b></dt>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]1418 <dd>Convert a constant, CST, to another TYPE. The size of CST and TYPE must be
1419 identical (same number of bits). The conversion is done as if the CST value
1420 was stored to memory and read back as TYPE. In other words, no bits change
Reid Spencer5b950642006-11-11 23:08:07 +0000[diff] [blame]1421 with this operator, just the type. This can be used for conversion of
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1422 vector types to any other type, as long as they have the same bit width. For
Reid Spencer5b950642006-11-11 23:08:07 +0000[diff] [blame]1423 pointers it is only valid to cast to another pointer type.
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]1424 </dd>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1425
1426 <dt><b><tt>getelementptr ( CSTPTR, IDX0, IDX1, ... )</tt></b></dt>
1427
1428 <dd>Perform the <a href="#i_getelementptr">getelementptr operation</a> on
1429 constants. As with the <a href="#i_getelementptr">getelementptr</a>
1430 instruction, the index list may have zero or more indexes, which are required
1431 to make sense for the type of "CSTPTR".</dd>
1432
Robert Bocchino7e97a6d2006-01-10 19:31:34 +0000[diff] [blame]1433 <dt><b><tt>select ( COND, VAL1, VAL2 )</tt></b></dt>
1434
1435 <dd>Perform the <a href="#i_select">select operation</a> on
Reid Spencer9965ee72006-12-04 19:23:19 +0000[diff] [blame]1436 constants.</dd>
1437
1438 <dt><b><tt>icmp COND ( VAL1, VAL2 )</tt></b></dt>
1439 <dd>Performs the <a href="#i_icmp">icmp operation</a> on constants.</dd>
1440
1441 <dt><b><tt>fcmp COND ( VAL1, VAL2 )</tt></b></dt>
1442 <dd>Performs the <a href="#i_fcmp">fcmp operation</a> on constants.</dd>
Robert Bocchino7e97a6d2006-01-10 19:31:34 +0000[diff] [blame]1443
1444 <dt><b><tt>extractelement ( VAL, IDX )</tt></b></dt>
1445
1446 <dd>Perform the <a href="#i_extractelement">extractelement
1447 operation</a> on constants.
1448
Robert Bocchinof72fdfe2006-01-15 20:48:27 +0000[diff] [blame]1449 <dt><b><tt>insertelement ( VAL, ELT, IDX )</tt></b></dt>
1450
1451 <dd>Perform the <a href="#i_insertelement">insertelement
Reid Spencer9965ee72006-12-04 19:23:19 +0000[diff] [blame]1452 operation</a> on constants.</dd>
Robert Bocchinof72fdfe2006-01-15 20:48:27 +0000[diff] [blame]1453
Chris Lattner016a0e52006-04-08 00:13:41 +0000[diff] [blame]1454
1455 <dt><b><tt>shufflevector ( VEC1, VEC2, IDXMASK )</tt></b></dt>
1456
1457 <dd>Perform the <a href="#i_shufflevector">shufflevector
Reid Spencer9965ee72006-12-04 19:23:19 +0000[diff] [blame]1458 operation</a> on constants.</dd>
Chris Lattner016a0e52006-04-08 00:13:41 +0000[diff] [blame]1459
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1460 <dt><b><tt>OPCODE ( LHS, RHS )</tt></b></dt>
1461
Reid Spencer641f5c92004-12-09 18:13:12 +0000[diff] [blame]1462 <dd>Perform the specified operation of the LHS and RHS constants. OPCODE may
1463 be any of the <a href="#binaryops">binary</a> or <a href="#bitwiseops">bitwise
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1464 binary</a> operations. The constraints on operands are the same as those for
1465 the corresponding instruction (e.g. no bitwise operations on floating point
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]1466 values are allowed).</dd>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1467</dl>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1468</div>
Chris Lattnerb1652612004-03-08 16:49:10 +0000[diff] [blame]1469
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1470<!-- *********************************************************************** -->
Chris Lattner98f013c2006-01-25 23:47:57 +0000[diff] [blame]1471<div class="doc_section"> <a name="othervalues">Other Values</a> </div>
1472<!-- *********************************************************************** -->
1473
1474<!-- ======================================================================= -->
1475<div class="doc_subsection">
1476<a name="inlineasm">Inline Assembler Expressions</a>
1477</div>
1478
1479<div class="doc_text">
1480
1481<p>
1482LLVM supports inline assembler expressions (as opposed to <a href="#moduleasm">
1483Module-Level Inline Assembly</a>) through the use of a special value. This
1484value represents the inline assembler as a string (containing the instructions
1485to emit), a list of operand constraints (stored as a string), and a flag that
1486indicates whether or not the inline asm expression has side effects. An example
1487inline assembler expression is:
1488</p>
1489
1490<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1491 i32 (i32) asm "bswap $0", "=r,r"
Chris Lattner98f013c2006-01-25 23:47:57 +0000[diff] [blame]1492</pre>
1493
1494<p>
1495Inline assembler expressions may <b>only</b> be used as the callee operand of
1496a <a href="#i_call"><tt>call</tt> instruction</a>. Thus, typically we have:
1497</p>
1498
1499<pre>
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]1500 %X = call i32 asm "<a href="#int_bswap">bswap</a> $0", "=r,r"(i32 %Y)
Chris Lattner98f013c2006-01-25 23:47:57 +0000[diff] [blame]1501</pre>
1502
1503<p>
1504Inline asms with side effects not visible in the constraint list must be marked
1505as having side effects. This is done through the use of the
1506'<tt>sideeffect</tt>' keyword, like so:
1507</p>
1508
1509<pre>
1510 call void asm sideeffect "eieio", ""()
1511</pre>
1512
1513<p>TODO: The format of the asm and constraints string still need to be
1514documented here. Constraints on what can be done (e.g. duplication, moving, etc
1515need to be documented).
1516</p>
1517
1518</div>
1519
1520<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1521<div class="doc_section"> <a name="instref">Instruction Reference</a> </div>
1522<!-- *********************************************************************** -->
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1523
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1524<div class="doc_text">
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1525
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1526<p>The LLVM instruction set consists of several different
1527classifications of instructions: <a href="#terminators">terminator
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]1528instructions</a>, <a href="#binaryops">binary instructions</a>,
1529<a href="#bitwiseops">bitwise binary instructions</a>, <a
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1530 href="#memoryops">memory instructions</a>, and <a href="#otherops">other
1531instructions</a>.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1532
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1533</div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1534
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1535<!-- ======================================================================= -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1536<div class="doc_subsection"> <a name="terminators">Terminator
1537Instructions</a> </div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1538
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1539<div class="doc_text">
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1540
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1541<p>As mentioned <a href="#functionstructure">previously</a>, every
1542basic block in a program ends with a "Terminator" instruction, which
1543indicates which block should be executed after the current block is
1544finished. These terminator instructions typically yield a '<tt>void</tt>'
1545value: they produce control flow, not values (the one exception being
1546the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction).</p>
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]1547<p>There are six different terminator instructions: the '<a
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1548 href="#i_ret"><tt>ret</tt></a>' instruction, the '<a href="#i_br"><tt>br</tt></a>'
1549instruction, the '<a href="#i_switch"><tt>switch</tt></a>' instruction,
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1550the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction, the '<a
1551 href="#i_unwind"><tt>unwind</tt></a>' instruction, and the '<a
1552 href="#i_unreachable"><tt>unreachable</tt></a>' instruction.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1553
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1554</div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1555
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1556<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1557<div class="doc_subsubsection"> <a name="i_ret">'<tt>ret</tt>'
1558Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1559<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1560<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1561<pre> ret &lt;type&gt; &lt;value&gt; <i>; Return a value from a non-void function</i>
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]1562 ret void <i>; Return from void function</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1563</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1564<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1565<p>The '<tt>ret</tt>' instruction is used to return control flow (and a
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]1566value) from a function back to the caller.</p>
John Criswell417228d2004-04-09 16:48:45 +0000[diff] [blame]1567<p>There are two forms of the '<tt>ret</tt>' instruction: one that
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1568returns a value and then causes control flow, and one that just causes
1569control flow to occur.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1570<h5>Arguments:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1571<p>The '<tt>ret</tt>' instruction may return any '<a
1572 href="#t_firstclass">first class</a>' type. Notice that a function is
1573not <a href="#wellformed">well formed</a> if there exists a '<tt>ret</tt>'
1574instruction inside of the function that returns a value that does not
1575match the return type of the function.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1576<h5>Semantics:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1577<p>When the '<tt>ret</tt>' instruction is executed, control flow
1578returns back to the calling function's context. If the caller is a "<a
John Criswell40db33f2004-06-25 15:16:57 +0000[diff] [blame]1579 href="#i_call"><tt>call</tt></a>" instruction, execution continues at
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1580the instruction after the call. If the caller was an "<a
1581 href="#i_invoke"><tt>invoke</tt></a>" instruction, execution continues
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]1582at the beginning of the "normal" destination block. If the instruction
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1583returns a value, that value shall set the call or invoke instruction's
1584return value.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1585<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1586<pre> ret i32 5 <i>; Return an integer value of 5</i>
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]1587 ret void <i>; Return from a void function</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1588</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1589</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1590<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1591<div class="doc_subsubsection"> <a name="i_br">'<tt>br</tt>' Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1592<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1593<h5>Syntax:</h5>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]1594<pre> br i1 &lt;cond&gt;, label &lt;iftrue&gt;, label &lt;iffalse&gt;<br> br label &lt;dest&gt; <i>; Unconditional branch</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1595</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1596<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1597<p>The '<tt>br</tt>' instruction is used to cause control flow to
1598transfer to a different basic block in the current function. There are
1599two forms of this instruction, corresponding to a conditional branch
1600and an unconditional branch.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1601<h5>Arguments:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1602<p>The conditional branch form of the '<tt>br</tt>' instruction takes a
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]1603single '<tt>i1</tt>' value and two '<tt>label</tt>' values. The
Reid Spencer50c723a2007-02-19 23:54:10 +0000[diff] [blame]1604unconditional form of the '<tt>br</tt>' instruction takes a single
1605'<tt>label</tt>' value as a target.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1606<h5>Semantics:</h5>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]1607<p>Upon execution of a conditional '<tt>br</tt>' instruction, the '<tt>i1</tt>'
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1608argument is evaluated. If the value is <tt>true</tt>, control flows
1609to the '<tt>iftrue</tt>' <tt>label</tt> argument. If "cond" is <tt>false</tt>,
1610control flows to the '<tt>iffalse</tt>' <tt>label</tt> argument.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1611<h5>Example:</h5>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]1612<pre>Test:<br> %cond = <a href="#i_icmp">icmp</a> eq, i32 %a, %b<br> br i1 %cond, label %IfEqual, label %IfUnequal<br>IfEqual:<br> <a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1613 href="#i_ret">ret</a> i32 1<br>IfUnequal:<br> <a href="#i_ret">ret</a> i32 0<br></pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1614</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1615<!-- _______________________________________________________________________ -->
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1616<div class="doc_subsubsection">
1617 <a name="i_switch">'<tt>switch</tt>' Instruction</a>
1618</div>
1619
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1620<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1621<h5>Syntax:</h5>
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1622
1623<pre>
1624 switch &lt;intty&gt; &lt;value&gt;, label &lt;defaultdest&gt; [ &lt;intty&gt; &lt;val&gt;, label &lt;dest&gt; ... ]
1625</pre>
1626
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1627<h5>Overview:</h5>
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1628
1629<p>The '<tt>switch</tt>' instruction is used to transfer control flow to one of
1630several different places. It is a generalization of the '<tt>br</tt>'
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1631instruction, allowing a branch to occur to one of many possible
1632destinations.</p>
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1633
1634
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1635<h5>Arguments:</h5>
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1636
1637<p>The '<tt>switch</tt>' instruction uses three parameters: an integer
1638comparison value '<tt>value</tt>', a default '<tt>label</tt>' destination, and
1639an array of pairs of comparison value constants and '<tt>label</tt>'s. The
1640table is not allowed to contain duplicate constant entries.</p>
1641
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1642<h5>Semantics:</h5>
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1643
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1644<p>The <tt>switch</tt> instruction specifies a table of values and
1645destinations. When the '<tt>switch</tt>' instruction is executed, this
John Criswellbcbb18c2004-06-25 16:05:06 +0000[diff] [blame]1646table is searched for the given value. If the value is found, control flow is
1647transfered to the corresponding destination; otherwise, control flow is
1648transfered to the default destination.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1649
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1650<h5>Implementation:</h5>
1651
1652<p>Depending on properties of the target machine and the particular
1653<tt>switch</tt> instruction, this instruction may be code generated in different
John Criswellbcbb18c2004-06-25 16:05:06 +0000[diff] [blame]1654ways. For example, it could be generated as a series of chained conditional
1655branches or with a lookup table.</p>
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1656
1657<h5>Example:</h5>
1658
1659<pre>
1660 <i>; Emulate a conditional br instruction</i>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]1661 %Val = <a href="#i_zext">zext</a> i1 %value to i32
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1662 switch i32 %Val, label %truedest [i32 0, label %falsedest ]
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1663
1664 <i>; Emulate an unconditional br instruction</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1665 switch i32 0, label %dest [ ]
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1666
1667 <i>; Implement a jump table:</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1668 switch i32 %val, label %otherwise [ i32 0, label %onzero
1669 i32 1, label %onone
1670 i32 2, label %ontwo ]
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1671</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1672</div>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1673
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1674<!-- _______________________________________________________________________ -->
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1675<div class="doc_subsubsection">
1676 <a name="i_invoke">'<tt>invoke</tt>' Instruction</a>
1677</div>
1678
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1679<div class="doc_text">
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1680
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1681<h5>Syntax:</h5>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1682
1683<pre>
1684 &lt;result&gt; = invoke [<a href="#callingconv">cconv</a>] &lt;ptr to function ty&gt; %&lt;function ptr val&gt;(&lt;function args&gt;)
Chris Lattner6b7a0082006-05-14 18:23:06 +0000[diff] [blame]1685 to label &lt;normal label&gt; unwind label &lt;exception label&gt;
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1686</pre>
1687
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]1688<h5>Overview:</h5>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1689
1690<p>The '<tt>invoke</tt>' instruction causes control to transfer to a specified
1691function, with the possibility of control flow transfer to either the
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]1692'<tt>normal</tt>' label or the
1693'<tt>exception</tt>' label. If the callee function returns with the
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1694"<tt><a href="#i_ret">ret</a></tt>" instruction, control flow will return to the
1695"normal" label. If the callee (or any indirect callees) returns with the "<a
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]1696href="#i_unwind"><tt>unwind</tt></a>" instruction, control is interrupted and
1697continued at the dynamically nearest "exception" label.</p>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1698
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1699<h5>Arguments:</h5>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1700
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1701<p>This instruction requires several arguments:</p>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1702
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1703<ol>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1704 <li>
Duncan Sands16f122e2007-03-30 12:22:09 +0000[diff] [blame]1705 The optional "cconv" marker indicates which <a href="#callingconv">calling
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1706 convention</a> the call should use. If none is specified, the call defaults
1707 to using C calling conventions.
1708 </li>
1709 <li>'<tt>ptr to function ty</tt>': shall be the signature of the pointer to
1710 function value being invoked. In most cases, this is a direct function
1711 invocation, but indirect <tt>invoke</tt>s are just as possible, branching off
1712 an arbitrary pointer to function value.
1713 </li>
1714
1715 <li>'<tt>function ptr val</tt>': An LLVM value containing a pointer to a
1716 function to be invoked. </li>
1717
1718 <li>'<tt>function args</tt>': argument list whose types match the function
1719 signature argument types. If the function signature indicates the function
1720 accepts a variable number of arguments, the extra arguments can be
1721 specified. </li>
1722
1723 <li>'<tt>normal label</tt>': the label reached when the called function
1724 executes a '<tt><a href="#i_ret">ret</a></tt>' instruction. </li>
1725
1726 <li>'<tt>exception label</tt>': the label reached when a callee returns with
1727 the <a href="#i_unwind"><tt>unwind</tt></a> instruction. </li>
1728
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1729</ol>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1730
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1731<h5>Semantics:</h5>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1732
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1733<p>This instruction is designed to operate as a standard '<tt><a
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1734href="#i_call">call</a></tt>' instruction in most regards. The primary
1735difference is that it establishes an association with a label, which is used by
1736the runtime library to unwind the stack.</p>
1737
1738<p>This instruction is used in languages with destructors to ensure that proper
1739cleanup is performed in the case of either a <tt>longjmp</tt> or a thrown
1740exception. Additionally, this is important for implementation of
1741'<tt>catch</tt>' clauses in high-level languages that support them.</p>
1742
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1743<h5>Example:</h5>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1744<pre>
Jeff Cohen5819f182007-04-22 01:17:39 +0000[diff] [blame]1745 %retval = invoke i32 %Test(i32 15) to label %Continue
1746 unwind label %TestCleanup <i>; {i32}:retval set</i>
1747 %retval = invoke <a href="#callingconv">coldcc</a> i32 %Test(i32 15) to label %Continue
1748 unwind label %TestCleanup <i>; {i32}:retval set</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1749</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1750</div>
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1751
1752
Chris Lattner5ed60612003-09-03 00:41:47 +0000[diff] [blame]1753<!-- _______________________________________________________________________ -->
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1754
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1755<div class="doc_subsubsection"> <a name="i_unwind">'<tt>unwind</tt>'
1756Instruction</a> </div>
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1757
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1758<div class="doc_text">
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1759
Chris Lattner5ed60612003-09-03 00:41:47 +0000[diff] [blame]1760<h5>Syntax:</h5>
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1761<pre>
1762 unwind
1763</pre>
1764
Chris Lattner5ed60612003-09-03 00:41:47 +0000[diff] [blame]1765<h5>Overview:</h5>
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1766
1767<p>The '<tt>unwind</tt>' instruction unwinds the stack, continuing control flow
1768at the first callee in the dynamic call stack which used an <a
1769href="#i_invoke"><tt>invoke</tt></a> instruction to perform the call. This is
1770primarily used to implement exception handling.</p>
1771
Chris Lattner5ed60612003-09-03 00:41:47 +0000[diff] [blame]1772<h5>Semantics:</h5>
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1773
1774<p>The '<tt>unwind</tt>' intrinsic causes execution of the current function to
1775immediately halt. The dynamic call stack is then searched for the first <a
1776href="#i_invoke"><tt>invoke</tt></a> instruction on the call stack. Once found,
1777execution continues at the "exceptional" destination block specified by the
1778<tt>invoke</tt> instruction. If there is no <tt>invoke</tt> instruction in the
1779dynamic call chain, undefined behavior results.</p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1780</div>
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1781
1782<!-- _______________________________________________________________________ -->
1783
1784<div class="doc_subsubsection"> <a name="i_unreachable">'<tt>unreachable</tt>'
1785Instruction</a> </div>
1786
1787<div class="doc_text">
1788
1789<h5>Syntax:</h5>
1790<pre>
1791 unreachable
1792</pre>
1793
1794<h5>Overview:</h5>
1795
1796<p>The '<tt>unreachable</tt>' instruction has no defined semantics. This
1797instruction is used to inform the optimizer that a particular portion of the
1798code is not reachable. This can be used to indicate that the code after a
1799no-return function cannot be reached, and other facts.</p>
1800
1801<h5>Semantics:</h5>
1802
1803<p>The '<tt>unreachable</tt>' instruction has no defined semantics.</p>
1804</div>
1805
1806
1807
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1808<!-- ======================================================================= -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1809<div class="doc_subsection"> <a name="binaryops">Binary Operations</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1810<div class="doc_text">
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1811<p>Binary operators are used to do most of the computation in a
1812program. They require two operands, execute an operation on them, and
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]1813produce a single value. The operands might represent
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1814multiple data, as is the case with the <a href="#t_vector">vector</a> data type.
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1815The result value of a binary operator is not
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1816necessarily the same type as its operands.</p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1817<p>There are several different binary operators:</p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1818</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1819<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1820<div class="doc_subsubsection"> <a name="i_add">'<tt>add</tt>'
1821Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1822<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1823<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1824<pre> &lt;result&gt; = add &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1825</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1826<h5>Overview:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1827<p>The '<tt>add</tt>' instruction returns the sum of its two operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1828<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1829<p>The two arguments to the '<tt>add</tt>' instruction must be either <a
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1830 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a> values.
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1831 This instruction can also take <a href="#t_vector">vector</a> versions of the values.
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1832Both arguments must have identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1833<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1834<p>The value produced is the integer or floating point sum of the two
1835operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1836<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1837<pre> &lt;result&gt; = add i32 4, %var <i>; yields {i32}:result = 4 + %var</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1838</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1839</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1840<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1841<div class="doc_subsubsection"> <a name="i_sub">'<tt>sub</tt>'
1842Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1843<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1844<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1845<pre> &lt;result&gt; = sub &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1846</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1847<h5>Overview:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1848<p>The '<tt>sub</tt>' instruction returns the difference of its two
1849operands.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1850<p>Note that the '<tt>sub</tt>' instruction is used to represent the '<tt>neg</tt>'
1851instruction present in most other intermediate representations.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1852<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1853<p>The two arguments to the '<tt>sub</tt>' instruction must be either <a
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1854 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1855values.
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1856This instruction can also take <a href="#t_vector">vector</a> versions of the values.
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1857Both arguments must have identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1858<h5>Semantics:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1859<p>The value produced is the integer or floating point difference of
1860the two operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1861<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1862<pre> &lt;result&gt; = sub i32 4, %var <i>; yields {i32}:result = 4 - %var</i>
1863 &lt;result&gt; = sub i32 0, %val <i>; yields {i32}:result = -%var</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1864</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1865</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1866<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1867<div class="doc_subsubsection"> <a name="i_mul">'<tt>mul</tt>'
1868Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1869<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1870<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1871<pre> &lt;result&gt; = mul &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1872</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1873<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1874<p>The '<tt>mul</tt>' instruction returns the product of its two
1875operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1876<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1877<p>The two arguments to the '<tt>mul</tt>' instruction must be either <a
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1878 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1879values.
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1880This instruction can also take <a href="#t_vector">vector</a> versions of the values.
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1881Both arguments must have identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1882<h5>Semantics:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1883<p>The value produced is the integer or floating point product of the
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1884two operands.</p>
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]1885<p>Because the operands are the same width, the result of an integer
1886multiplication is the same whether the operands should be deemed unsigned or
1887signed.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1888<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1889<pre> &lt;result&gt; = mul i32 4, %var <i>; yields {i32}:result = 4 * %var</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1890</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1891</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1892<!-- _______________________________________________________________________ -->
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame]1893<div class="doc_subsubsection"> <a name="i_udiv">'<tt>udiv</tt>' Instruction
1894</a></div>
1895<div class="doc_text">
1896<h5>Syntax:</h5>
1897<pre> &lt;result&gt; = udiv &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
1898</pre>
1899<h5>Overview:</h5>
1900<p>The '<tt>udiv</tt>' instruction returns the quotient of its two
1901operands.</p>
1902<h5>Arguments:</h5>
1903<p>The two arguments to the '<tt>udiv</tt>' instruction must be
1904<a href="#t_integer">integer</a> values. Both arguments must have identical
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1905types. This instruction can also take <a href="#t_vector">vector</a> versions
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame]1906of the values in which case the elements must be integers.</p>
1907<h5>Semantics:</h5>
1908<p>The value produced is the unsigned integer quotient of the two operands. This
1909instruction always performs an unsigned division operation, regardless of
1910whether the arguments are unsigned or not.</p>
1911<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1912<pre> &lt;result&gt; = udiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame]1913</pre>
1914</div>
1915<!-- _______________________________________________________________________ -->
1916<div class="doc_subsubsection"> <a name="i_sdiv">'<tt>sdiv</tt>' Instruction
1917</a> </div>
1918<div class="doc_text">
1919<h5>Syntax:</h5>
1920<pre> &lt;result&gt; = sdiv &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
1921</pre>
1922<h5>Overview:</h5>
1923<p>The '<tt>sdiv</tt>' instruction returns the quotient of its two
1924operands.</p>
1925<h5>Arguments:</h5>
1926<p>The two arguments to the '<tt>sdiv</tt>' instruction must be
1927<a href="#t_integer">integer</a> values. Both arguments must have identical
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1928types. This instruction can also take <a href="#t_vector">vector</a> versions
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame]1929of the values in which case the elements must be integers.</p>
1930<h5>Semantics:</h5>
1931<p>The value produced is the signed integer quotient of the two operands. This
1932instruction always performs a signed division operation, regardless of whether
1933the arguments are signed or not.</p>
1934<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1935<pre> &lt;result&gt; = sdiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame]1936</pre>
1937</div>
1938<!-- _______________________________________________________________________ -->
1939<div class="doc_subsubsection"> <a name="i_fdiv">'<tt>fdiv</tt>'
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1940Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1941<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1942<h5>Syntax:</h5>
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame]1943<pre> &lt;result&gt; = fdiv &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1944</pre>
1945<h5>Overview:</h5>
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame]1946<p>The '<tt>fdiv</tt>' instruction returns the quotient of its two
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1947operands.</p>
1948<h5>Arguments:</h5>
Jeff Cohen5819f182007-04-22 01:17:39 +0000[diff] [blame]1949<p>The two arguments to the '<tt>fdiv</tt>' instruction must be
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame]1950<a href="#t_floating">floating point</a> values. Both arguments must have
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1951identical types. This instruction can also take <a href="#t_vector">vector</a>
Jeff Cohen5819f182007-04-22 01:17:39 +0000[diff] [blame]1952versions of floating point values.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1953<h5>Semantics:</h5>
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame]1954<p>The value produced is the floating point quotient of the two operands.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1955<h5>Example:</h5>
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame]1956<pre> &lt;result&gt; = fdiv float 4.0, %var <i>; yields {float}:result = 4.0 / %var</i>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1957</pre>
1958</div>
1959<!-- _______________________________________________________________________ -->
Reid Spencer7eb55b32006-11-02 01:53:59 +0000[diff] [blame]1960<div class="doc_subsubsection"> <a name="i_urem">'<tt>urem</tt>' Instruction</a>
1961</div>
1962<div class="doc_text">
1963<h5>Syntax:</h5>
1964<pre> &lt;result&gt; = urem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
1965</pre>
1966<h5>Overview:</h5>
1967<p>The '<tt>urem</tt>' instruction returns the remainder from the
1968unsigned division of its two arguments.</p>
1969<h5>Arguments:</h5>
1970<p>The two arguments to the '<tt>urem</tt>' instruction must be
1971<a href="#t_integer">integer</a> values. Both arguments must have identical
1972types.</p>
1973<h5>Semantics:</h5>
1974<p>This instruction returns the unsigned integer <i>remainder</i> of a division.
1975This instruction always performs an unsigned division to get the remainder,
1976regardless of whether the arguments are unsigned or not.</p>
1977<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1978<pre> &lt;result&gt; = urem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
Reid Spencer7eb55b32006-11-02 01:53:59 +0000[diff] [blame]1979</pre>
1980
1981</div>
1982<!-- _______________________________________________________________________ -->
1983<div class="doc_subsubsection"> <a name="i_srem">'<tt>srem</tt>'
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1984Instruction</a> </div>
1985<div class="doc_text">
1986<h5>Syntax:</h5>
Reid Spencer7eb55b32006-11-02 01:53:59 +0000[diff] [blame]1987<pre> &lt;result&gt; = srem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1988</pre>
1989<h5>Overview:</h5>
Reid Spencer7eb55b32006-11-02 01:53:59 +0000[diff] [blame]1990<p>The '<tt>srem</tt>' instruction returns the remainder from the
1991signed division of its two operands.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1992<h5>Arguments:</h5>
Reid Spencer7eb55b32006-11-02 01:53:59 +0000[diff] [blame]1993<p>The two arguments to the '<tt>srem</tt>' instruction must be
1994<a href="#t_integer">integer</a> values. Both arguments must have identical
1995types.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1996<h5>Semantics:</h5>
Reid Spencer7eb55b32006-11-02 01:53:59 +0000[diff] [blame]1997<p>This instruction returns the <i>remainder</i> of a division (where the result
Reid Spencer806ad6a2007-03-24 22:23:39 +0000[diff] [blame]1998has the same sign as the dividend, <tt>var1</tt>), not the <i>modulo</i>
1999operator (where the result has the same sign as the divisor, <tt>var2</tt>) of
2000a value. For more information about the difference, see <a
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2001 href="http://mathforum.org/dr.math/problems/anne.4.28.99.html">The
Reid Spencer806ad6a2007-03-24 22:23:39 +0000[diff] [blame]2002Math Forum</a>. For a table of how this is implemented in various languages,
Reid Spencerdb3b93b2007-03-24 22:40:44 +0000[diff] [blame]2003please see <a href="http://en.wikipedia.org/wiki/Modulo_operation">
Reid Spencer806ad6a2007-03-24 22:23:39 +0000[diff] [blame]2004Wikipedia: modulo operation</a>.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2005<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2006<pre> &lt;result&gt; = srem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
Reid Spencer7eb55b32006-11-02 01:53:59 +0000[diff] [blame]2007</pre>
2008
2009</div>
2010<!-- _______________________________________________________________________ -->
2011<div class="doc_subsubsection"> <a name="i_frem">'<tt>frem</tt>'
2012Instruction</a> </div>
2013<div class="doc_text">
2014<h5>Syntax:</h5>
2015<pre> &lt;result&gt; = frem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2016</pre>
2017<h5>Overview:</h5>
2018<p>The '<tt>frem</tt>' instruction returns the remainder from the
2019division of its two operands.</p>
2020<h5>Arguments:</h5>
2021<p>The two arguments to the '<tt>frem</tt>' instruction must be
2022<a href="#t_floating">floating point</a> values. Both arguments must have
2023identical types.</p>
2024<h5>Semantics:</h5>
2025<p>This instruction returns the <i>remainder</i> of a division.</p>
2026<h5>Example:</h5>
2027<pre> &lt;result&gt; = frem float 4.0, %var <i>; yields {float}:result = 4.0 % %var</i>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2028</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2029</div>
Robert Bocchino820bc75b2006-02-17 21:18:08 +0000[diff] [blame]2030
Reid Spencer2ab01932007-02-02 13:57:07 +0000[diff] [blame]2031<!-- ======================================================================= -->
2032<div class="doc_subsection"> <a name="bitwiseops">Bitwise Binary
2033Operations</a> </div>
2034<div class="doc_text">
2035<p>Bitwise binary operators are used to do various forms of
2036bit-twiddling in a program. They are generally very efficient
2037instructions and can commonly be strength reduced from other
2038instructions. They require two operands, execute an operation on them,
2039and produce a single value. The resulting value of the bitwise binary
2040operators is always the same type as its first operand.</p>
2041</div>
2042
Reid Spencer04e259b2007-01-31 21:39:12 +0000[diff] [blame]2043<!-- _______________________________________________________________________ -->
2044<div class="doc_subsubsection"> <a name="i_shl">'<tt>shl</tt>'
2045Instruction</a> </div>
2046<div class="doc_text">
2047<h5>Syntax:</h5>
2048<pre> &lt;result&gt; = shl &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2049</pre>
2050<h5>Overview:</h5>
2051<p>The '<tt>shl</tt>' instruction returns the first operand shifted to
2052the left a specified number of bits.</p>
2053<h5>Arguments:</h5>
2054<p>Both arguments to the '<tt>shl</tt>' instruction must be the same <a
2055 href="#t_integer">integer</a> type.</p>
2056<h5>Semantics:</h5>
2057<p>The value produced is <tt>var1</tt> * 2<sup><tt>var2</tt></sup>.</p>
2058<h5>Example:</h5><pre>
2059 &lt;result&gt; = shl i32 4, %var <i>; yields {i32}: 4 &lt;&lt; %var</i>
2060 &lt;result&gt; = shl i32 4, 2 <i>; yields {i32}: 16</i>
2061 &lt;result&gt; = shl i32 1, 10 <i>; yields {i32}: 1024</i>
2062</pre>
2063</div>
2064<!-- _______________________________________________________________________ -->
2065<div class="doc_subsubsection"> <a name="i_lshr">'<tt>lshr</tt>'
2066Instruction</a> </div>
2067<div class="doc_text">
2068<h5>Syntax:</h5>
2069<pre> &lt;result&gt; = lshr &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2070</pre>
2071
2072<h5>Overview:</h5>
2073<p>The '<tt>lshr</tt>' instruction (logical shift right) returns the first
Jeff Cohen5819f182007-04-22 01:17:39 +0000[diff] [blame]2074operand shifted to the right a specified number of bits with zero fill.</p>
Reid Spencer04e259b2007-01-31 21:39:12 +0000[diff] [blame]2075
2076<h5>Arguments:</h5>
2077<p>Both arguments to the '<tt>lshr</tt>' instruction must be the same
2078<a href="#t_integer">integer</a> type.</p>
2079
2080<h5>Semantics:</h5>
2081<p>This instruction always performs a logical shift right operation. The most
2082significant bits of the result will be filled with zero bits after the
2083shift.</p>
2084
2085<h5>Example:</h5>
2086<pre>
2087 &lt;result&gt; = lshr i32 4, 1 <i>; yields {i32}:result = 2</i>
2088 &lt;result&gt; = lshr i32 4, 2 <i>; yields {i32}:result = 1</i>
2089 &lt;result&gt; = lshr i8 4, 3 <i>; yields {i8}:result = 0</i>
2090 &lt;result&gt; = lshr i8 -2, 1 <i>; yields {i8}:result = 0x7FFFFFFF </i>
2091</pre>
2092</div>
2093
Reid Spencer2ab01932007-02-02 13:57:07 +0000[diff] [blame]2094<!-- _______________________________________________________________________ -->
Reid Spencer04e259b2007-01-31 21:39:12 +0000[diff] [blame]2095<div class="doc_subsubsection"> <a name="i_ashr">'<tt>ashr</tt>'
2096Instruction</a> </div>
2097<div class="doc_text">
2098
2099<h5>Syntax:</h5>
2100<pre> &lt;result&gt; = ashr &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2101</pre>
2102
2103<h5>Overview:</h5>
2104<p>The '<tt>ashr</tt>' instruction (arithmetic shift right) returns the first
Jeff Cohen5819f182007-04-22 01:17:39 +0000[diff] [blame]2105operand shifted to the right a specified number of bits with sign extension.</p>
Reid Spencer04e259b2007-01-31 21:39:12 +0000[diff] [blame]2106
2107<h5>Arguments:</h5>
2108<p>Both arguments to the '<tt>ashr</tt>' instruction must be the same
2109<a href="#t_integer">integer</a> type.</p>
2110
2111<h5>Semantics:</h5>
2112<p>This instruction always performs an arithmetic shift right operation,
2113The most significant bits of the result will be filled with the sign bit
2114of <tt>var1</tt>.</p>
2115
2116<h5>Example:</h5>
2117<pre>
2118 &lt;result&gt; = ashr i32 4, 1 <i>; yields {i32}:result = 2</i>
2119 &lt;result&gt; = ashr i32 4, 2 <i>; yields {i32}:result = 1</i>
2120 &lt;result&gt; = ashr i8 4, 3 <i>; yields {i8}:result = 0</i>
2121 &lt;result&gt; = ashr i8 -2, 1 <i>; yields {i8}:result = -1</i>
2122</pre>
2123</div>
2124
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2125<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2126<div class="doc_subsubsection"> <a name="i_and">'<tt>and</tt>'
2127Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2128<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2129<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2130<pre> &lt;result&gt; = and &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2131</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2132<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2133<p>The '<tt>and</tt>' instruction returns the bitwise logical and of
2134its two operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2135<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2136<p>The two arguments to the '<tt>and</tt>' instruction must be <a
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]2137 href="#t_integer">integer</a> values. Both arguments must have
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2138identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2139<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2140<p>The truth table used for the '<tt>and</tt>' instruction is:</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2141<p> </p>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]2142<div style="align: center">
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2143<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2144 <tbody>
2145 <tr>
2146 <td>In0</td>
2147 <td>In1</td>
2148 <td>Out</td>
2149 </tr>
2150 <tr>
2151 <td>0</td>
2152 <td>0</td>
2153 <td>0</td>
2154 </tr>
2155 <tr>
2156 <td>0</td>
2157 <td>1</td>
2158 <td>0</td>
2159 </tr>
2160 <tr>
2161 <td>1</td>
2162 <td>0</td>
2163 <td>0</td>
2164 </tr>
2165 <tr>
2166 <td>1</td>
2167 <td>1</td>
2168 <td>1</td>
2169 </tr>
2170 </tbody>
2171</table>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]2172</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2173<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2174<pre> &lt;result&gt; = and i32 4, %var <i>; yields {i32}:result = 4 &amp; %var</i>
2175 &lt;result&gt; = and i32 15, 40 <i>; yields {i32}:result = 8</i>
2176 &lt;result&gt; = and i32 4, 8 <i>; yields {i32}:result = 0</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2177</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2178</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2179<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2180<div class="doc_subsubsection"> <a name="i_or">'<tt>or</tt>' Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2181<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2182<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2183<pre> &lt;result&gt; = or &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2184</pre>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2185<h5>Overview:</h5>
2186<p>The '<tt>or</tt>' instruction returns the bitwise logical inclusive
2187or of its two operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2188<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2189<p>The two arguments to the '<tt>or</tt>' instruction must be <a
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]2190 href="#t_integer">integer</a> values. Both arguments must have
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2191identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2192<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2193<p>The truth table used for the '<tt>or</tt>' instruction is:</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2194<p> </p>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]2195<div style="align: center">
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2196<table border="1" cellspacing="0" cellpadding="4">
2197 <tbody>
2198 <tr>
2199 <td>In0</td>
2200 <td>In1</td>
2201 <td>Out</td>
2202 </tr>
2203 <tr>
2204 <td>0</td>
2205 <td>0</td>
2206 <td>0</td>
2207 </tr>
2208 <tr>
2209 <td>0</td>
2210 <td>1</td>
2211 <td>1</td>
2212 </tr>
2213 <tr>
2214 <td>1</td>
2215 <td>0</td>
2216 <td>1</td>
2217 </tr>
2218 <tr>
2219 <td>1</td>
2220 <td>1</td>
2221 <td>1</td>
2222 </tr>
2223 </tbody>
2224</table>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]2225</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2226<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2227<pre> &lt;result&gt; = or i32 4, %var <i>; yields {i32}:result = 4 | %var</i>
2228 &lt;result&gt; = or i32 15, 40 <i>; yields {i32}:result = 47</i>
2229 &lt;result&gt; = or i32 4, 8 <i>; yields {i32}:result = 12</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2230</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2231</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2232<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2233<div class="doc_subsubsection"> <a name="i_xor">'<tt>xor</tt>'
2234Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2235<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2236<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2237<pre> &lt;result&gt; = xor &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2238</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2239<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2240<p>The '<tt>xor</tt>' instruction returns the bitwise logical exclusive
2241or of its two operands. The <tt>xor</tt> is used to implement the
2242"one's complement" operation, which is the "~" operator in C.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2243<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2244<p>The two arguments to the '<tt>xor</tt>' instruction must be <a
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]2245 href="#t_integer">integer</a> values. Both arguments must have
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2246identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2247<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2248<p>The truth table used for the '<tt>xor</tt>' instruction is:</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2249<p> </p>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]2250<div style="align: center">
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2251<table border="1" cellspacing="0" cellpadding="4">
2252 <tbody>
2253 <tr>
2254 <td>In0</td>
2255 <td>In1</td>
2256 <td>Out</td>
2257 </tr>
2258 <tr>
2259 <td>0</td>
2260 <td>0</td>
2261 <td>0</td>
2262 </tr>
2263 <tr>
2264 <td>0</td>
2265 <td>1</td>
2266 <td>1</td>
2267 </tr>
2268 <tr>
2269 <td>1</td>
2270 <td>0</td>
2271 <td>1</td>
2272 </tr>
2273 <tr>
2274 <td>1</td>
2275 <td>1</td>
2276 <td>0</td>
2277 </tr>
2278 </tbody>
2279</table>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]2280</div>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2281<p> </p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2282<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2283<pre> &lt;result&gt; = xor i32 4, %var <i>; yields {i32}:result = 4 ^ %var</i>
2284 &lt;result&gt; = xor i32 15, 40 <i>; yields {i32}:result = 39</i>
2285 &lt;result&gt; = xor i32 4, 8 <i>; yields {i32}:result = 12</i>
2286 &lt;result&gt; = xor i32 %V, -1 <i>; yields {i32}:result = ~%V</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2287</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2288</div>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2289
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2290<!-- ======================================================================= -->
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2291<div class="doc_subsection">
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2292 <a name="vectorops">Vector Operations</a>
2293</div>
2294
2295<div class="doc_text">
2296
2297<p>LLVM supports several instructions to represent vector operations in a
Jeff Cohen5819f182007-04-22 01:17:39 +0000[diff] [blame]2298target-independent manner. These instructions cover the element-access and
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2299vector-specific operations needed to process vectors effectively. While LLVM
2300does directly support these vector operations, many sophisticated algorithms
2301will want to use target-specific intrinsics to take full advantage of a specific
2302target.</p>
2303
2304</div>
2305
2306<!-- _______________________________________________________________________ -->
2307<div class="doc_subsubsection">
2308 <a name="i_extractelement">'<tt>extractelement</tt>' Instruction</a>
2309</div>
2310
2311<div class="doc_text">
2312
2313<h5>Syntax:</h5>
2314
2315<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2316 &lt;result&gt; = extractelement &lt;n x &lt;ty&gt;&gt; &lt;val&gt;, i32 &lt;idx&gt; <i>; yields &lt;ty&gt;</i>
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2317</pre>
2318
2319<h5>Overview:</h5>
2320
2321<p>
2322The '<tt>extractelement</tt>' instruction extracts a single scalar
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]2323element from a vector at a specified index.
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2324</p>
2325
2326
2327<h5>Arguments:</h5>
2328
2329<p>
2330The first operand of an '<tt>extractelement</tt>' instruction is a
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]2331value of <a href="#t_vector">vector</a> type. The second operand is
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2332an index indicating the position from which to extract the element.
2333The index may be a variable.</p>
2334
2335<h5>Semantics:</h5>
2336
2337<p>
2338The result is a scalar of the same type as the element type of
2339<tt>val</tt>. Its value is the value at position <tt>idx</tt> of
2340<tt>val</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
2341results are undefined.
2342</p>
2343
2344<h5>Example:</h5>
2345
2346<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2347 %result = extractelement &lt;4 x i32&gt; %vec, i32 0 <i>; yields i32</i>
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2348</pre>
2349</div>
2350
2351
2352<!-- _______________________________________________________________________ -->
2353<div class="doc_subsubsection">
2354 <a name="i_insertelement">'<tt>insertelement</tt>' Instruction</a>
2355</div>
2356
2357<div class="doc_text">
2358
2359<h5>Syntax:</h5>
2360
2361<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2362 &lt;result&gt; = insertelement &lt;n x &lt;ty&gt;&gt; &lt;val&gt;, &lt;ty&gt; &lt;elt&gt, i32 &lt;idx&gt; <i>; yields &lt;n x &lt;ty&gt;&gt;</i>
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2363</pre>
2364
2365<h5>Overview:</h5>
2366
2367<p>
2368The '<tt>insertelement</tt>' instruction inserts a scalar
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]2369element into a vector at a specified index.
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2370</p>
2371
2372
2373<h5>Arguments:</h5>
2374
2375<p>
2376The first operand of an '<tt>insertelement</tt>' instruction is a
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]2377value of <a href="#t_vector">vector</a> type. The second operand is a
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2378scalar value whose type must equal the element type of the first
2379operand. The third operand is an index indicating the position at
2380which to insert the value. The index may be a variable.</p>
2381
2382<h5>Semantics:</h5>
2383
2384<p>
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]2385The result is a vector of the same type as <tt>val</tt>. Its
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2386element values are those of <tt>val</tt> except at position
2387<tt>idx</tt>, where it gets the value <tt>elt</tt>. If <tt>idx</tt>
2388exceeds the length of <tt>val</tt>, the results are undefined.
2389</p>
2390
2391<h5>Example:</h5>
2392
2393<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2394 %result = insertelement &lt;4 x i32&gt; %vec, i32 1, i32 0 <i>; yields &lt;4 x i32&gt;</i>
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2395</pre>
2396</div>
2397
2398<!-- _______________________________________________________________________ -->
2399<div class="doc_subsubsection">
2400 <a name="i_shufflevector">'<tt>shufflevector</tt>' Instruction</a>
2401</div>
2402
2403<div class="doc_text">
2404
2405<h5>Syntax:</h5>
2406
2407<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2408 &lt;result&gt; = shufflevector &lt;n x &lt;ty&gt;&gt; &lt;v1&gt;, &lt;n x &lt;ty&gt;&gt; &lt;v2&gt;, &lt;n x i32&gt; &lt;mask&gt; <i>; yields &lt;n x &lt;ty&gt;&gt;</i>
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2409</pre>
2410
2411<h5>Overview:</h5>
2412
2413<p>
2414The '<tt>shufflevector</tt>' instruction constructs a permutation of elements
2415from two input vectors, returning a vector of the same type.
2416</p>
2417
2418<h5>Arguments:</h5>
2419
2420<p>
2421The first two operands of a '<tt>shufflevector</tt>' instruction are vectors
2422with types that match each other and types that match the result of the
2423instruction. The third argument is a shuffle mask, which has the same number
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2424of elements as the other vector type, but whose element type is always 'i32'.
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2425</p>
2426
2427<p>
2428The shuffle mask operand is required to be a constant vector with either
2429constant integer or undef values.
2430</p>
2431
2432<h5>Semantics:</h5>
2433
2434<p>
2435The elements of the two input vectors are numbered from left to right across
2436both of the vectors. The shuffle mask operand specifies, for each element of
2437the result vector, which element of the two input registers the result element
2438gets. The element selector may be undef (meaning "don't care") and the second
2439operand may be undef if performing a shuffle from only one vector.
2440</p>
2441
2442<h5>Example:</h5>
2443
2444<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2445 %result = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; %v2,
Jeff Cohen5819f182007-04-22 01:17:39 +0000[diff] [blame]2446 &lt;4 x i32&gt; &lt;i32 0, i32 4, i32 1, i32 5&gt; <i>; yields &lt;4 x i32&gt;</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2447 %result = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; undef,
2448 &lt;4 x i32&gt; &lt;i32 0, i32 1, i32 2, i32 3&gt; <i>; yields &lt;4 x i32&gt;</i> - Identity shuffle.
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2449</pre>
2450</div>
2451
Tanya Lattnerb138bbe2006-04-14 19:24:33 +0000[diff] [blame]2452
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2453<!-- ======================================================================= -->
2454<div class="doc_subsection">
Chris Lattner6ab66722006-08-15 00:45:58 +0000[diff] [blame]2455 <a name="memoryops">Memory Access and Addressing Operations</a>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2456</div>
2457
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2458<div class="doc_text">
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2459
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2460<p>A key design point of an SSA-based representation is how it
2461represents memory. In LLVM, no memory locations are in SSA form, which
2462makes things very simple. This section describes how to read, write,
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]2463allocate, and free memory in LLVM.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2464
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2465</div>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2466
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2467<!-- _______________________________________________________________________ -->
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2468<div class="doc_subsubsection">
2469 <a name="i_malloc">'<tt>malloc</tt>' Instruction</a>
2470</div>
2471
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2472<div class="doc_text">
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2473
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2474<h5>Syntax:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2475
2476<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2477 &lt;result&gt; = malloc &lt;type&gt;[, i32 &lt;NumElements&gt;][, align &lt;alignment&gt;] <i>; yields {type*}:result</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2478</pre>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2479
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2480<h5>Overview:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2481
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2482<p>The '<tt>malloc</tt>' instruction allocates memory from the system
2483heap and returns a pointer to it.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2484
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2485<h5>Arguments:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2486
2487<p>The '<tt>malloc</tt>' instruction allocates
2488<tt>sizeof(&lt;type&gt;)*NumElements</tt>
John Criswella92e5862004-02-24 16:13:56 +0000[diff] [blame]2489bytes of memory from the operating system and returns a pointer of the
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2490appropriate type to the program. If "NumElements" is specified, it is the
2491number of elements allocated. If an alignment is specified, the value result
2492of the allocation is guaranteed to be aligned to at least that boundary. If
2493not specified, or if zero, the target can choose to align the allocation on any
2494convenient boundary.</p>
2495
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2496<p>'<tt>type</tt>' must be a sized type.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2497
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2498<h5>Semantics:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2499
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2500<p>Memory is allocated using the system "<tt>malloc</tt>" function, and
2501a pointer is returned.</p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2502
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2503<h5>Example:</h5>
2504
2505<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2506 %array = malloc [4 x i8 ] <i>; yields {[%4 x i8]*}:array</i>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2507
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2508 %size = <a href="#i_add">add</a> i32 2, 2 <i>; yields {i32}:size = i32 4</i>
2509 %array1 = malloc i8, i32 4 <i>; yields {i8*}:array1</i>
2510 %array2 = malloc [12 x i8], i32 %size <i>; yields {[12 x i8]*}:array2</i>
2511 %array3 = malloc i32, i32 4, align 1024 <i>; yields {i32*}:array3</i>
2512 %array4 = malloc i32, align 1024 <i>; yields {i32*}:array4</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2513</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2514</div>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2515
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2516<!-- _______________________________________________________________________ -->
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2517<div class="doc_subsubsection">
2518 <a name="i_free">'<tt>free</tt>' Instruction</a>
2519</div>
2520
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2521<div class="doc_text">
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2522
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2523<h5>Syntax:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2524
2525<pre>
2526 free &lt;type&gt; &lt;value&gt; <i>; yields {void}</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2527</pre>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2528
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2529<h5>Overview:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2530
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2531<p>The '<tt>free</tt>' instruction returns memory back to the unused
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]2532memory heap to be reallocated in the future.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2533
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2534<h5>Arguments:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2535
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2536<p>'<tt>value</tt>' shall be a pointer value that points to a value
2537that was allocated with the '<tt><a href="#i_malloc">malloc</a></tt>'
2538instruction.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2539
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2540<h5>Semantics:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2541
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]2542<p>Access to the memory pointed to by the pointer is no longer defined
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2543after this instruction executes.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2544
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2545<h5>Example:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2546
2547<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2548 %array = <a href="#i_malloc">malloc</a> [4 x i8] <i>; yields {[4 x i8]*}:array</i>
2549 free [4 x i8]* %array
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2550</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2551</div>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2552
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2553<!-- _______________________________________________________________________ -->
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2554<div class="doc_subsubsection">
2555 <a name="i_alloca">'<tt>alloca</tt>' Instruction</a>
2556</div>
2557
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2558<div class="doc_text">
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2559
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2560<h5>Syntax:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2561
2562<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2563 &lt;result&gt; = alloca &lt;type&gt;[, i32 &lt;NumElements&gt;][, align &lt;alignment&gt;] <i>; yields {type*}:result</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2564</pre>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2565
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2566<h5>Overview:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2567
Jeff Cohen5819f182007-04-22 01:17:39 +0000[diff] [blame]2568<p>The '<tt>alloca</tt>' instruction allocates memory on the stack frame of the
2569currently executing function, to be automatically released when this function
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2570returns to its caller.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2571
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2572<h5>Arguments:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2573
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]2574<p>The '<tt>alloca</tt>' instruction allocates <tt>sizeof(&lt;type&gt;)*NumElements</tt>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2575bytes of memory on the runtime stack, returning a pointer of the
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2576appropriate type to the program. If "NumElements" is specified, it is the
2577number of elements allocated. If an alignment is specified, the value result
2578of the allocation is guaranteed to be aligned to at least that boundary. If
2579not specified, or if zero, the target can choose to align the allocation on any
2580convenient boundary.</p>
2581
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2582<p>'<tt>type</tt>' may be any sized type.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2583
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2584<h5>Semantics:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2585
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]2586<p>Memory is allocated; a pointer is returned. '<tt>alloca</tt>'d
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2587memory is automatically released when the function returns. The '<tt>alloca</tt>'
2588instruction is commonly used to represent automatic variables that must
2589have an address available. When the function returns (either with the <tt><a
John Criswellc932bef2005-05-12 16:55:34 +0000[diff] [blame]2590 href="#i_ret">ret</a></tt> or <tt><a href="#i_unwind">unwind</a></tt>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2591instructions), the memory is reclaimed.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2592
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2593<h5>Example:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2594
2595<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2596 %ptr = alloca i32 <i>; yields {i32*}:ptr</i>
2597 %ptr = alloca i32, i32 4 <i>; yields {i32*}:ptr</i>
2598 %ptr = alloca i32, i32 4, align 1024 <i>; yields {i32*}:ptr</i>
2599 %ptr = alloca i32, align 1024 <i>; yields {i32*}:ptr</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2600</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2601</div>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2602
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2603<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2604<div class="doc_subsubsection"> <a name="i_load">'<tt>load</tt>'
2605Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2606<div class="doc_text">
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2607<h5>Syntax:</h5>
Christopher Lambbff50202007-04-21 08:16:25 +0000[diff] [blame]2608<pre> &lt;result&gt; = load &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;]<br> &lt;result&gt; = volatile load &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;]<br></pre>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2609<h5>Overview:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2610<p>The '<tt>load</tt>' instruction is used to read from memory.</p>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2611<h5>Arguments:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2612<p>The argument to the '<tt>load</tt>' instruction specifies the memory
John Criswell4c0cf7f2005-10-24 16:17:18 +0000[diff] [blame]2613address from which to load. The pointer must point to a <a
Chris Lattner10ee9652004-06-03 22:57:15 +0000[diff] [blame]2614 href="#t_firstclass">first class</a> type. If the <tt>load</tt> is
John Criswell4c0cf7f2005-10-24 16:17:18 +0000[diff] [blame]2615marked as <tt>volatile</tt>, then the optimizer is not allowed to modify
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2616the number or order of execution of this <tt>load</tt> with other
2617volatile <tt>load</tt> and <tt><a href="#i_store">store</a></tt>
2618instructions. </p>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2619<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2620<p>The location of memory pointed to is loaded.</p>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2621<h5>Examples:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2622<pre> %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2623 <a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2624 href="#i_store">store</a> i32 3, i32* %ptr <i>; yields {void}</i>
2625 %val = load i32* %ptr <i>; yields {i32}:val = i32 3</i>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2626</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2627</div>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2628<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2629<div class="doc_subsubsection"> <a name="i_store">'<tt>store</tt>'
2630Instruction</a> </div>
Reid Spencera89fb182006-11-09 21:18:01 +0000[diff] [blame]2631<div class="doc_text">
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2632<h5>Syntax:</h5>
Christopher Lambbff50202007-04-21 08:16:25 +0000[diff] [blame]2633<pre> store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;] <i>; yields {void}</i>
2634 volatile store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;] <i>; yields {void}</i>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2635</pre>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2636<h5>Overview:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2637<p>The '<tt>store</tt>' instruction is used to write to memory.</p>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2638<h5>Arguments:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2639<p>There are two arguments to the '<tt>store</tt>' instruction: a value
Jeff Cohen5819f182007-04-22 01:17:39 +0000[diff] [blame]2640to store and an address at which to store it. The type of the '<tt>&lt;pointer&gt;</tt>'
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2641operand must be a pointer to the type of the '<tt>&lt;value&gt;</tt>'
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]2642operand. If the <tt>store</tt> is marked as <tt>volatile</tt>, then the
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2643optimizer is not allowed to modify the number or order of execution of
2644this <tt>store</tt> with other volatile <tt>load</tt> and <tt><a
2645 href="#i_store">store</a></tt> instructions.</p>
2646<h5>Semantics:</h5>
2647<p>The contents of memory are updated to contain '<tt>&lt;value&gt;</tt>'
2648at the location specified by the '<tt>&lt;pointer&gt;</tt>' operand.</p>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2649<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2650<pre> %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2651 <a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2652 href="#i_store">store</a> i32 3, i32* %ptr <i>; yields {void}</i>
2653 %val = load i32* %ptr <i>; yields {i32}:val = i32 3</i>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2654</pre>
Reid Spencer443460a2006-11-09 21:15:49 +0000[diff] [blame]2655</div>
2656
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2657<!-- _______________________________________________________________________ -->
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2658<div class="doc_subsubsection">
2659 <a name="i_getelementptr">'<tt>getelementptr</tt>' Instruction</a>
2660</div>
2661
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2662<div class="doc_text">
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]2663<h5>Syntax:</h5>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2664<pre>
2665 &lt;result&gt; = getelementptr &lt;ty&gt;* &lt;ptrval&gt;{, &lt;ty&gt; &lt;idx&gt;}*
2666</pre>
2667
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]2668<h5>Overview:</h5>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2669
2670<p>
2671The '<tt>getelementptr</tt>' instruction is used to get the address of a
2672subelement of an aggregate data structure.</p>
2673
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]2674<h5>Arguments:</h5>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2675
Reid Spencercee005c2006-12-04 21:29:24 +0000[diff] [blame]2676<p>This instruction takes a list of integer operands that indicate what
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2677elements of the aggregate object to index to. The actual types of the arguments
2678provided depend on the type of the first pointer argument. The
2679'<tt>getelementptr</tt>' instruction is used to index down through the type
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]2680levels of a structure or to a specific index in an array. When indexing into a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2681structure, only <tt>i32</tt> integer constants are allowed. When indexing
Reid Spencercee005c2006-12-04 21:29:24 +0000[diff] [blame]2682into an array or pointer, only integers of 32 or 64 bits are allowed, and will
2683be sign extended to 64-bit values.</p>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2684
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2685<p>For example, let's consider a C code fragment and how it gets
2686compiled to LLVM:</p>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2687
2688<pre>
2689 struct RT {
2690 char A;
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2691 i32 B[10][20];
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2692 char C;
2693 };
2694 struct ST {
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2695 i32 X;
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2696 double Y;
2697 struct RT Z;
2698 };
2699
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2700 define i32 *foo(struct ST *s) {
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2701 return &amp;s[1].Z.B[5][13];
2702 }
2703</pre>
2704
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2705<p>The LLVM code generated by the GCC frontend is:</p>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2706
2707<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2708 %RT = type { i8 , [10 x [20 x i32]], i8 }
2709 %ST = type { i32, double, %RT }
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2710
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2711 define i32* %foo(%ST* %s) {
Brian Gaeke317ef962004-07-02 21:08:14 +0000[diff] [blame]2712 entry:
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2713 %reg = getelementptr %ST* %s, i32 1, i32 2, i32 1, i32 5, i32 13
2714 ret i32* %reg
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2715 }
2716</pre>
2717
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]2718<h5>Semantics:</h5>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2719
2720<p>The index types specified for the '<tt>getelementptr</tt>' instruction depend
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]2721on the pointer type that is being indexed into. <a href="#t_pointer">Pointer</a>
Reid Spencercee005c2006-12-04 21:29:24 +0000[diff] [blame]2722and <a href="#t_array">array</a> types can use a 32-bit or 64-bit
Reid Spencerc0312692006-12-03 16:53:48 +0000[diff] [blame]2723<a href="#t_integer">integer</a> type but the value will always be sign extended
Jeff Cohen5819f182007-04-22 01:17:39 +0000[diff] [blame]2724to 64-bits. <a href="#t_struct">Structure</a> types require <tt>i32</tt>
Reid Spencerc0312692006-12-03 16:53:48 +0000[diff] [blame]2725<b>constants</b>.</p>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2726
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2727<p>In the example above, the first index is indexing into the '<tt>%ST*</tt>'
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2728type, which is a pointer, yielding a '<tt>%ST</tt>' = '<tt>{ i32, double, %RT
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2729}</tt>' type, a structure. The second index indexes into the third element of
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2730the structure, yielding a '<tt>%RT</tt>' = '<tt>{ i8 , [10 x [20 x i32]],
2731i8 }</tt>' type, another structure. The third index indexes into the second
2732element of the structure, yielding a '<tt>[10 x [20 x i32]]</tt>' type, an
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2733array. The two dimensions of the array are subscripted into, yielding an
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2734'<tt>i32</tt>' type. The '<tt>getelementptr</tt>' instruction returns a pointer
2735to this element, thus computing a value of '<tt>i32*</tt>' type.</p>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2736
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2737<p>Note that it is perfectly legal to index partially through a
2738structure, returning a pointer to an inner element. Because of this,
2739the LLVM code for the given testcase is equivalent to:</p>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2740
2741<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2742 define i32* %foo(%ST* %s) {
2743 %t1 = getelementptr %ST* %s, i32 1 <i>; yields %ST*:%t1</i>
Jeff Cohen5819f182007-04-22 01:17:39 +0000[diff] [blame]2744 %t2 = getelementptr %ST* %t1, i32 0, i32 2 <i>; yields %RT*:%t2</i>
2745 %t3 = getelementptr %RT* %t2, i32 0, i32 1 <i>; yields [10 x [20 x i32]]*:%t3</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2746 %t4 = getelementptr [10 x [20 x i32]]* %t3, i32 0, i32 5 <i>; yields [20 x i32]*:%t4</i>
2747 %t5 = getelementptr [20 x i32]* %t4, i32 0, i32 13 <i>; yields i32*:%t5</i>
2748 ret i32* %t5
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2749 }
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]2750</pre>
Chris Lattnerc0ad71e2005-06-24 17:22:57 +0000[diff] [blame]2751
2752<p>Note that it is undefined to access an array out of bounds: array and
2753pointer indexes must always be within the defined bounds of the array type.
2754The one exception for this rules is zero length arrays. These arrays are
2755defined to be accessible as variable length arrays, which requires access
2756beyond the zero'th element.</p>
2757
Chris Lattner6ab66722006-08-15 00:45:58 +0000[diff] [blame]2758<p>The getelementptr instruction is often confusing. For some more insight
2759into how it works, see <a href="GetElementPtr.html">the getelementptr
2760FAQ</a>.</p>
2761
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]2762<h5>Example:</h5>
Chris Lattnerc0ad71e2005-06-24 17:22:57 +0000[diff] [blame]2763
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2764<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2765 <i>; yields [12 x i8]*:aptr</i>
2766 %aptr = getelementptr {i32, [12 x i8]}* %sptr, i64 0, i32 1
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2767</pre>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2768</div>
Reid Spencer443460a2006-11-09 21:15:49 +0000[diff] [blame]2769
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2770<!-- ======================================================================= -->
Reid Spencer97c5fa42006-11-08 01:18:52 +0000[diff] [blame]2771<div class="doc_subsection"> <a name="convertops">Conversion Operations</a>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2772</div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2773<div class="doc_text">
Reid Spencer97c5fa42006-11-08 01:18:52 +0000[diff] [blame]2774<p>The instructions in this category are the conversion instructions (casting)
2775which all take a single operand and a type. They perform various bit conversions
2776on the operand.</p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2777</div>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]2778
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]2779<!-- _______________________________________________________________________ -->
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]2780<div class="doc_subsubsection">
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2781 <a name="i_trunc">'<tt>trunc .. to</tt>' Instruction</a>
2782</div>
2783<div class="doc_text">
2784
2785<h5>Syntax:</h5>
2786<pre>
2787 &lt;result&gt; = trunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2788</pre>
2789
2790<h5>Overview:</h5>
2791<p>
2792The '<tt>trunc</tt>' instruction truncates its operand to the type <tt>ty2</tt>.
2793</p>
2794
2795<h5>Arguments:</h5>
2796<p>
2797The '<tt>trunc</tt>' instruction takes a <tt>value</tt> to trunc, which must
2798be an <a href="#t_integer">integer</a> type, and a type that specifies the size
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]2799and type of the result, which must be an <a href="#t_integer">integer</a>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]2800type. The bit size of <tt>value</tt> must be larger than the bit size of
2801<tt>ty2</tt>. Equal sized types are not allowed.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2802
2803<h5>Semantics:</h5>
2804<p>
2805The '<tt>trunc</tt>' instruction truncates the high order bits in <tt>value</tt>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]2806and converts the remaining bits to <tt>ty2</tt>. Since the source size must be
2807larger than the destination size, <tt>trunc</tt> cannot be a <i>no-op cast</i>.
2808It will always truncate bits.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2809
2810<h5>Example:</h5>
2811<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2812 %X = trunc i32 257 to i8 <i>; yields i8:1</i>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]2813 %Y = trunc i32 123 to i1 <i>; yields i1:true</i>
2814 %Y = trunc i32 122 to i1 <i>; yields i1:false</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2815</pre>
2816</div>
2817
2818<!-- _______________________________________________________________________ -->
2819<div class="doc_subsubsection">
2820 <a name="i_zext">'<tt>zext .. to</tt>' Instruction</a>
2821</div>
2822<div class="doc_text">
2823
2824<h5>Syntax:</h5>
2825<pre>
2826 &lt;result&gt; = zext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2827</pre>
2828
2829<h5>Overview:</h5>
2830<p>The '<tt>zext</tt>' instruction zero extends its operand to type
2831<tt>ty2</tt>.</p>
2832
2833
2834<h5>Arguments:</h5>
2835<p>The '<tt>zext</tt>' instruction takes a value to cast, which must be of
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]2836<a href="#t_integer">integer</a> type, and a type to cast it to, which must
2837also be of <a href="#t_integer">integer</a> type. The bit size of the
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]2838<tt>value</tt> must be smaller than the bit size of the destination type,
2839<tt>ty2</tt>.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2840
2841<h5>Semantics:</h5>
2842<p>The <tt>zext</tt> fills the high order bits of the <tt>value</tt> with zero
2843bits until it reaches the size of the destination type, <tt>ty2</tt>. When the
2844the operand and the type are the same size, no bit filling is done and the
2845cast is considered a <i>no-op cast</i> because no bits change (only the type
2846changes).</p>
2847
Reid Spencer07c9c682007-01-12 15:46:11 +0000[diff] [blame]2848<p>When zero extending from i1, the result will always be either 0 or 1.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2849
2850<h5>Example:</h5>
2851<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2852 %X = zext i32 257 to i64 <i>; yields i64:257</i>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]2853 %Y = zext i1 true to i32 <i>; yields i32:1</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2854</pre>
2855</div>
2856
2857<!-- _______________________________________________________________________ -->
2858<div class="doc_subsubsection">
2859 <a name="i_sext">'<tt>sext .. to</tt>' Instruction</a>
2860</div>
2861<div class="doc_text">
2862
2863<h5>Syntax:</h5>
2864<pre>
2865 &lt;result&gt; = sext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2866</pre>
2867
2868<h5>Overview:</h5>
2869<p>The '<tt>sext</tt>' sign extends <tt>value</tt> to the type <tt>ty2</tt>.</p>
2870
2871<h5>Arguments:</h5>
2872<p>
2873The '<tt>sext</tt>' instruction takes a value to cast, which must be of
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]2874<a href="#t_integer">integer</a> type, and a type to cast it to, which must
2875also be of <a href="#t_integer">integer</a> type. The bit size of the
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]2876<tt>value</tt> must be smaller than the bit size of the destination type,
2877<tt>ty2</tt>.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2878
2879<h5>Semantics:</h5>
2880<p>
2881The '<tt>sext</tt>' instruction performs a sign extension by copying the sign
2882bit (highest order bit) of the <tt>value</tt> until it reaches the bit size of
2883the type <tt>ty2</tt>. When the the operand and the type are the same size,
2884no bit filling is done and the cast is considered a <i>no-op cast</i> because
2885no bits change (only the type changes).</p>
2886
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]2887<p>When sign extending from i1, the extension always results in -1 or 0.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2888
2889<h5>Example:</h5>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2890<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2891 %X = sext i8 -1 to i16 <i>; yields i16 :65535</i>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]2892 %Y = sext i1 true to i32 <i>; yields i32:-1</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2893</pre>
2894</div>
2895
2896<!-- _______________________________________________________________________ -->
2897<div class="doc_subsubsection">
Reid Spencer2e2740d2006-11-09 21:48:10 +0000[diff] [blame]2898 <a name="i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a>
2899</div>
2900
2901<div class="doc_text">
2902
2903<h5>Syntax:</h5>
2904
2905<pre>
2906 &lt;result&gt; = fptrunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2907</pre>
2908
2909<h5>Overview:</h5>
2910<p>The '<tt>fptrunc</tt>' instruction truncates <tt>value</tt> to type
2911<tt>ty2</tt>.</p>
2912
2913
2914<h5>Arguments:</h5>
2915<p>The '<tt>fptrunc</tt>' instruction takes a <a href="#t_floating">floating
2916 point</a> value to cast and a <a href="#t_floating">floating point</a> type to
2917cast it to. The size of <tt>value</tt> must be larger than the size of
2918<tt>ty2</tt>. This implies that <tt>fptrunc</tt> cannot be used to make a
2919<i>no-op cast</i>.</p>
2920
2921<h5>Semantics:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]2922<p> The '<tt>fptrunc</tt>' instruction truncates a <tt>value</tt> from a larger
2923<a href="#t_floating">floating point</a> type to a smaller
2924<a href="#t_floating">floating point</a> type. If the value cannot fit within
2925the destination type, <tt>ty2</tt>, then the results are undefined.</p>
Reid Spencer2e2740d2006-11-09 21:48:10 +0000[diff] [blame]2926
2927<h5>Example:</h5>
2928<pre>
2929 %X = fptrunc double 123.0 to float <i>; yields float:123.0</i>
2930 %Y = fptrunc double 1.0E+300 to float <i>; yields undefined</i>
2931</pre>
2932</div>
2933
2934<!-- _______________________________________________________________________ -->
2935<div class="doc_subsubsection">
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2936 <a name="i_fpext">'<tt>fpext .. to</tt>' Instruction</a>
2937</div>
2938<div class="doc_text">
2939
2940<h5>Syntax:</h5>
2941<pre>
2942 &lt;result&gt; = fpext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2943</pre>
2944
2945<h5>Overview:</h5>
2946<p>The '<tt>fpext</tt>' extends a floating point <tt>value</tt> to a larger
2947floating point value.</p>
2948
2949<h5>Arguments:</h5>
2950<p>The '<tt>fpext</tt>' instruction takes a
2951<a href="#t_floating">floating point</a> <tt>value</tt> to cast,
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]2952and a <a href="#t_floating">floating point</a> type to cast it to. The source
2953type must be smaller than the destination type.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2954
2955<h5>Semantics:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]2956<p>The '<tt>fpext</tt>' instruction extends the <tt>value</tt> from a smaller
Duncan Sands16f122e2007-03-30 12:22:09 +0000[diff] [blame]2957<a href="#t_floating">floating point</a> type to a larger
2958<a href="#t_floating">floating point</a> type. The <tt>fpext</tt> cannot be
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]2959used to make a <i>no-op cast</i> because it always changes bits. Use
Reid Spencer5b950642006-11-11 23:08:07 +0000[diff] [blame]2960<tt>bitcast</tt> to make a <i>no-op cast</i> for a floating point cast.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2961
2962<h5>Example:</h5>
2963<pre>
2964 %X = fpext float 3.1415 to double <i>; yields double:3.1415</i>
2965 %Y = fpext float 1.0 to float <i>; yields float:1.0 (no-op)</i>
2966</pre>
2967</div>
2968
2969<!-- _______________________________________________________________________ -->
2970<div class="doc_subsubsection">
Reid Spencer2eadb532007-01-21 00:29:26 +0000[diff] [blame]2971 <a name="i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2972</div>
2973<div class="doc_text">
2974
2975<h5>Syntax:</h5>
2976<pre>
2977 &lt;result&gt; = fp2uint &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2978</pre>
2979
2980<h5>Overview:</h5>
2981<p>The '<tt>fp2uint</tt>' converts a floating point <tt>value</tt> to its
2982unsigned integer equivalent of type <tt>ty2</tt>.
2983</p>
2984
2985<h5>Arguments:</h5>
2986<p>The '<tt>fp2uint</tt>' instruction takes a value to cast, which must be a
2987<a href="#t_floating">floating point</a> value, and a type to cast it to, which
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]2988must be an <a href="#t_integer">integer</a> type.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2989
2990<h5>Semantics:</h5>
2991<p> The '<tt>fp2uint</tt>' instruction converts its
2992<a href="#t_floating">floating point</a> operand into the nearest (rounding
2993towards zero) unsigned integer value. If the value cannot fit in <tt>ty2</tt>,
2994the results are undefined.</p>
2995
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]2996<p>When converting to i1, the conversion is done as a comparison against
2997zero. If the <tt>value</tt> was zero, the i1 result will be <tt>false</tt>.
2998If the <tt>value</tt> was non-zero, the i1 result will be <tt>true</tt>.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2999
3000<h5>Example:</h5>
3001<pre>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]3002 %X = fp2uint double 123.0 to i32 <i>; yields i32:123</i>
3003 %Y = fp2uint float 1.0E+300 to i1 <i>; yields i1:true</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3004 %X = fp2uint float 1.04E+17 to i8 <i>; yields undefined:1</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3005</pre>
3006</div>
3007
3008<!-- _______________________________________________________________________ -->
3009<div class="doc_subsubsection">
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3010 <a name="i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3011</div>
3012<div class="doc_text">
3013
3014<h5>Syntax:</h5>
3015<pre>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3016 &lt;result&gt; = fptosi &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3017</pre>
3018
3019<h5>Overview:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3020<p>The '<tt>fptosi</tt>' instruction converts
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3021<a href="#t_floating">floating point</a> <tt>value</tt> to type <tt>ty2</tt>.
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]3022</p>
3023
3024
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]3025<h5>Arguments:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3026<p> The '<tt>fptosi</tt>' instruction takes a value to cast, which must be a
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3027<a href="#t_floating">floating point</a> value, and a type to cast it to, which
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]3028must also be an <a href="#t_integer">integer</a> type.</p>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]3029
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]3030<h5>Semantics:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3031<p>The '<tt>fptosi</tt>' instruction converts its
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3032<a href="#t_floating">floating point</a> operand into the nearest (rounding
3033towards zero) signed integer value. If the value cannot fit in <tt>ty2</tt>,
3034the results are undefined.</p>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]3035
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]3036<p>When converting to i1, the conversion is done as a comparison against
3037zero. If the <tt>value</tt> was zero, the i1 result will be <tt>false</tt>.
3038If the <tt>value</tt> was non-zero, the i1 result will be <tt>true</tt>.</p>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]3039
Chris Lattner70de6632001-07-09 00:26:23 +0000[diff] [blame]3040<h5>Example:</h5>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]3041<pre>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]3042 %X = fptosi double -123.0 to i32 <i>; yields i32:-123</i>
3043 %Y = fptosi float 1.0E-247 to i1 <i>; yields i1:true</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3044 %X = fptosi float 1.04E+17 to i8 <i>; yields undefined:1</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3045</pre>
3046</div>
3047
3048<!-- _______________________________________________________________________ -->
3049<div class="doc_subsubsection">
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3050 <a name="i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3051</div>
3052<div class="doc_text">
3053
3054<h5>Syntax:</h5>
3055<pre>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3056 &lt;result&gt; = uitofp &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3057</pre>
3058
3059<h5>Overview:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3060<p>The '<tt>uitofp</tt>' instruction regards <tt>value</tt> as an unsigned
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3061integer and converts that value to the <tt>ty2</tt> type.</p>
3062
3063
3064<h5>Arguments:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3065<p>The '<tt>uitofp</tt>' instruction takes a value to cast, which must be an
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]3066<a href="#t_integer">integer</a> value, and a type to cast it to, which must
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3067be a <a href="#t_floating">floating point</a> type.</p>
3068
3069<h5>Semantics:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3070<p>The '<tt>uitofp</tt>' instruction interprets its operand as an unsigned
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3071integer quantity and converts it to the corresponding floating point value. If
Jeff Cohenbeccb742007-04-22 14:56:37 +0000[diff] [blame^]3072the value cannot fit in the floating point value, the results are undefined.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3073
3074
3075<h5>Example:</h5>
3076<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3077 %X = uitofp i32 257 to float <i>; yields float:257.0</i>
3078 %Y = uitofp i8 -1 to double <i>; yields double:255.0</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3079</pre>
3080</div>
3081
3082<!-- _______________________________________________________________________ -->
3083<div class="doc_subsubsection">
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3084 <a name="i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3085</div>
3086<div class="doc_text">
3087
3088<h5>Syntax:</h5>
3089<pre>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3090 &lt;result&gt; = sitofp &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3091</pre>
3092
3093<h5>Overview:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3094<p>The '<tt>sitofp</tt>' instruction regards <tt>value</tt> as a signed
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3095integer and converts that value to the <tt>ty2</tt> type.</p>
3096
3097<h5>Arguments:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3098<p>The '<tt>sitofp</tt>' instruction takes a value to cast, which must be an
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]3099<a href="#t_integer">integer</a> value, and a type to cast it to, which must be
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3100a <a href="#t_floating">floating point</a> type.</p>
3101
3102<h5>Semantics:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3103<p>The '<tt>sitofp</tt>' instruction interprets its operand as a signed
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3104integer quantity and converts it to the corresponding floating point value. If
Jeff Cohenbeccb742007-04-22 14:56:37 +0000[diff] [blame^]3105the value cannot fit in the floating point value, the results are undefined.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3106
3107<h5>Example:</h5>
3108<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3109 %X = sitofp i32 257 to float <i>; yields float:257.0</i>
3110 %Y = sitofp i8 -1 to double <i>; yields double:-1.0</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3111</pre>
3112</div>
3113
3114<!-- _______________________________________________________________________ -->
3115<div class="doc_subsubsection">
Reid Spencerb7344ff2006-11-11 21:00:47 +0000[diff] [blame]3116 <a name="i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a>
3117</div>
3118<div class="doc_text">
3119
3120<h5>Syntax:</h5>
3121<pre>
3122 &lt;result&gt; = ptrtoint &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3123</pre>
3124
3125<h5>Overview:</h5>
3126<p>The '<tt>ptrtoint</tt>' instruction converts the pointer <tt>value</tt> to
3127the integer type <tt>ty2</tt>.</p>
3128
3129<h5>Arguments:</h5>
3130<p>The '<tt>ptrtoint</tt>' instruction takes a <tt>value</tt> to cast, which
Duncan Sands16f122e2007-03-30 12:22:09 +0000[diff] [blame]3131must be a <a href="#t_pointer">pointer</a> value, and a type to cast it to
Reid Spencerb7344ff2006-11-11 21:00:47 +0000[diff] [blame]3132<tt>ty2</tt>, which must be an <a href="#t_integer">integer</a> type.
3133
3134<h5>Semantics:</h5>
3135<p>The '<tt>ptrtoint</tt>' instruction converts <tt>value</tt> to integer type
3136<tt>ty2</tt> by interpreting the pointer value as an integer and either
3137truncating or zero extending that value to the size of the integer type. If
3138<tt>value</tt> is smaller than <tt>ty2</tt> then a zero extension is done. If
3139<tt>value</tt> is larger than <tt>ty2</tt> then a truncation is done. If they
3140are the same size, then nothing is done (<i>no-op cast</i>).</p>
3141
3142<h5>Example:</h5>
3143<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3144 %X = ptrtoint i32* %X to i8 <i>; yields truncation on 32-bit</i>
3145 %Y = ptrtoint i32* %x to i64 <i>; yields zero extend on 32-bit</i>
Reid Spencerb7344ff2006-11-11 21:00:47 +0000[diff] [blame]3146</pre>
3147</div>
3148
3149<!-- _______________________________________________________________________ -->
3150<div class="doc_subsubsection">
3151 <a name="i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a>
3152</div>
3153<div class="doc_text">
3154
3155<h5>Syntax:</h5>
3156<pre>
3157 &lt;result&gt; = inttoptr &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3158</pre>
3159
3160<h5>Overview:</h5>
3161<p>The '<tt>inttoptr</tt>' instruction converts an integer <tt>value</tt> to
3162a pointer type, <tt>ty2</tt>.</p>
3163
3164<h5>Arguments:</h5>
Duncan Sands16f122e2007-03-30 12:22:09 +0000[diff] [blame]3165<p>The '<tt>inttoptr</tt>' instruction takes an <a href="#t_integer">integer</a>
Reid Spencerb7344ff2006-11-11 21:00:47 +0000[diff] [blame]3166value to cast, and a type to cast it to, which must be a
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]3167<a href="#t_pointer">pointer</a> type.
Reid Spencerb7344ff2006-11-11 21:00:47 +0000[diff] [blame]3168
3169<h5>Semantics:</h5>
3170<p>The '<tt>inttoptr</tt>' instruction converts <tt>value</tt> to type
3171<tt>ty2</tt> by applying either a zero extension or a truncation depending on
3172the size of the integer <tt>value</tt>. If <tt>value</tt> is larger than the
3173size of a pointer then a truncation is done. If <tt>value</tt> is smaller than
3174the size of a pointer then a zero extension is done. If they are the same size,
3175nothing is done (<i>no-op cast</i>).</p>
3176
3177<h5>Example:</h5>
3178<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3179 %X = inttoptr i32 255 to i32* <i>; yields zero extend on 64-bit</i>
3180 %X = inttoptr i32 255 to i32* <i>; yields no-op on 32-bit </i>
3181 %Y = inttoptr i16 0 to i32* <i>; yields zero extend on 32-bit</i>
Reid Spencerb7344ff2006-11-11 21:00:47 +0000[diff] [blame]3182</pre>
3183</div>
3184
3185<!-- _______________________________________________________________________ -->
3186<div class="doc_subsubsection">
Reid Spencer5b950642006-11-11 23:08:07 +0000[diff] [blame]3187 <a name="i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3188</div>
3189<div class="doc_text">
3190
3191<h5>Syntax:</h5>
3192<pre>
Reid Spencer5b950642006-11-11 23:08:07 +0000[diff] [blame]3193 &lt;result&gt; = bitcast &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3194</pre>
3195
3196<h5>Overview:</h5>
Reid Spencer5b950642006-11-11 23:08:07 +0000[diff] [blame]3197<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3198<tt>ty2</tt> without changing any bits.</p>
3199
3200<h5>Arguments:</h5>
Reid Spencer5b950642006-11-11 23:08:07 +0000[diff] [blame]3201<p>The '<tt>bitcast</tt>' instruction takes a value to cast, which must be
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3202a first class value, and a type to cast it to, which must also be a <a
3203 href="#t_firstclass">first class</a> type. The bit sizes of <tt>value</tt>
Reid Spencere3db84c2007-01-09 20:08:58 +0000[diff] [blame]3204and the destination type, <tt>ty2</tt>, must be identical. If the source
3205type is a pointer, the destination type must also be a pointer.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3206
3207<h5>Semantics:</h5>
Reid Spencer5b950642006-11-11 23:08:07 +0000[diff] [blame]3208<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
Reid Spencerb7344ff2006-11-11 21:00:47 +0000[diff] [blame]3209<tt>ty2</tt>. It is always a <i>no-op cast</i> because no bits change with
3210this conversion. The conversion is done as if the <tt>value</tt> had been
3211stored to memory and read back as type <tt>ty2</tt>. Pointer types may only be
3212converted to other pointer types with this instruction. To convert pointers to
3213other types, use the <a href="#i_inttoptr">inttoptr</a> or
3214<a href="#i_ptrtoint">ptrtoint</a> instructions first.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3215
3216<h5>Example:</h5>
3217<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3218 %X = bitcast i8 255 to i8 <i>; yields i8 :-1</i>
3219 %Y = bitcast i32* %x to sint* <i>; yields sint*:%x</i>
3220 %Z = bitcast <2xint> %V to i64; <i>; yields i64: %V</i>
Chris Lattner70de6632001-07-09 00:26:23 +0000[diff] [blame]3221</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3222</div>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]3223
Reid Spencer97c5fa42006-11-08 01:18:52 +0000[diff] [blame]3224<!-- ======================================================================= -->
3225<div class="doc_subsection"> <a name="otherops">Other Operations</a> </div>
3226<div class="doc_text">
3227<p>The instructions in this category are the "miscellaneous"
3228instructions, which defy better classification.</p>
3229</div>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3230
3231<!-- _______________________________________________________________________ -->
3232<div class="doc_subsubsection"><a name="i_icmp">'<tt>icmp</tt>' Instruction</a>
3233</div>
3234<div class="doc_text">
3235<h5>Syntax:</h5>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]3236<pre> &lt;result&gt; = icmp &lt;cond&gt; &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt;
3237<i>; yields {i1}:result</i>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3238</pre>
3239<h5>Overview:</h5>
3240<p>The '<tt>icmp</tt>' instruction returns a boolean value based on comparison
3241of its two integer operands.</p>
3242<h5>Arguments:</h5>
3243<p>The '<tt>icmp</tt>' instruction takes three operands. The first operand is
3244the condition code which indicates the kind of comparison to perform. It is not
3245a value, just a keyword. The possibilities for the condition code are:
3246<ol>
3247 <li><tt>eq</tt>: equal</li>
3248 <li><tt>ne</tt>: not equal </li>
3249 <li><tt>ugt</tt>: unsigned greater than</li>
3250 <li><tt>uge</tt>: unsigned greater or equal</li>
3251 <li><tt>ult</tt>: unsigned less than</li>
3252 <li><tt>ule</tt>: unsigned less or equal</li>
3253 <li><tt>sgt</tt>: signed greater than</li>
3254 <li><tt>sge</tt>: signed greater or equal</li>
3255 <li><tt>slt</tt>: signed less than</li>
3256 <li><tt>sle</tt>: signed less or equal</li>
3257</ol>
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]3258<p>The remaining two arguments must be <a href="#t_integer">integer</a> or
Reid Spencer784ef792007-01-04 05:19:58 +0000[diff] [blame]3259<a href="#t_pointer">pointer</a> typed. They must also be identical types.</p>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3260<h5>Semantics:</h5>
3261<p>The '<tt>icmp</tt>' compares <tt>var1</tt> and <tt>var2</tt> according to
3262the condition code given as <tt>cond</tt>. The comparison performed always
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]3263yields a <a href="#t_primitive">i1</a> result, as follows:
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3264<ol>
3265 <li><tt>eq</tt>: yields <tt>true</tt> if the operands are equal,
3266 <tt>false</tt> otherwise. No sign interpretation is necessary or performed.
3267 </li>
3268 <li><tt>ne</tt>: yields <tt>true</tt> if the operands are unequal,
3269 <tt>false</tt> otherwise. No sign interpretation is necessary or performed.
3270 <li><tt>ugt</tt>: interprets the operands as unsigned values and yields
3271 <tt>true</tt> if <tt>var1</tt> is greater than <tt>var2</tt>.</li>
3272 <li><tt>uge</tt>: interprets the operands as unsigned values and yields
3273 <tt>true</tt> if <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
3274 <li><tt>ult</tt>: interprets the operands as unsigned values and yields
3275 <tt>true</tt> if <tt>var1</tt> is less than <tt>var2</tt>.</li>
3276 <li><tt>ule</tt>: interprets the operands as unsigned values and yields
3277 <tt>true</tt> if <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
3278 <li><tt>sgt</tt>: interprets the operands as signed values and yields
3279 <tt>true</tt> if <tt>var1</tt> is greater than <tt>var2</tt>.</li>
3280 <li><tt>sge</tt>: interprets the operands as signed values and yields
3281 <tt>true</tt> if <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
3282 <li><tt>slt</tt>: interprets the operands as signed values and yields
3283 <tt>true</tt> if <tt>var1</tt> is less than <tt>var2</tt>.</li>
3284 <li><tt>sle</tt>: interprets the operands as signed values and yields
3285 <tt>true</tt> if <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3286</ol>
3287<p>If the operands are <a href="#t_pointer">pointer</a> typed, the pointer
3288values are treated as integers and then compared.</p>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3289
3290<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3291<pre> &lt;result&gt; = icmp eq i32 4, 5 <i>; yields: result=false</i>
3292 &lt;result&gt; = icmp ne float* %X, %X <i>; yields: result=false</i>
3293 &lt;result&gt; = icmp ult i16 4, 5 <i>; yields: result=true</i>
3294 &lt;result&gt; = icmp sgt i16 4, 5 <i>; yields: result=false</i>
3295 &lt;result&gt; = icmp ule i16 -4, 5 <i>; yields: result=false</i>
3296 &lt;result&gt; = icmp sge i16 4, 5 <i>; yields: result=false</i>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3297</pre>
3298</div>
3299
3300<!-- _______________________________________________________________________ -->
3301<div class="doc_subsubsection"><a name="i_fcmp">'<tt>fcmp</tt>' Instruction</a>
3302</div>
3303<div class="doc_text">
3304<h5>Syntax:</h5>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]3305<pre> &lt;result&gt; = fcmp &lt;cond&gt; &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt;
3306<i>; yields {i1}:result</i>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3307</pre>
3308<h5>Overview:</h5>
3309<p>The '<tt>fcmp</tt>' instruction returns a boolean value based on comparison
3310of its floating point operands.</p>
3311<h5>Arguments:</h5>
3312<p>The '<tt>fcmp</tt>' instruction takes three operands. The first operand is
3313the condition code which indicates the kind of comparison to perform. It is not
3314a value, just a keyword. The possibilities for the condition code are:
3315<ol>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3316 <li><tt>false</tt>: no comparison, always returns false</li>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3317 <li><tt>oeq</tt>: ordered and equal</li>
3318 <li><tt>ogt</tt>: ordered and greater than </li>
3319 <li><tt>oge</tt>: ordered and greater than or equal</li>
3320 <li><tt>olt</tt>: ordered and less than </li>
3321 <li><tt>ole</tt>: ordered and less than or equal</li>
3322 <li><tt>one</tt>: ordered and not equal</li>
3323 <li><tt>ord</tt>: ordered (no nans)</li>
3324 <li><tt>ueq</tt>: unordered or equal</li>
3325 <li><tt>ugt</tt>: unordered or greater than </li>
3326 <li><tt>uge</tt>: unordered or greater than or equal</li>
3327 <li><tt>ult</tt>: unordered or less than </li>
3328 <li><tt>ule</tt>: unordered or less than or equal</li>
3329 <li><tt>une</tt>: unordered or not equal</li>
3330 <li><tt>uno</tt>: unordered (either nans)</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3331 <li><tt>true</tt>: no comparison, always returns true</li>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3332</ol>
Reid Spencer02e0d1d2006-12-06 07:08:07 +0000[diff] [blame]3333<p>In the preceding, <i>ordered</i> means that neither operand is a QNAN while
3334<i>unordered</i> means that either operand may be a QNAN.</p>
Reid Spencer784ef792007-01-04 05:19:58 +0000[diff] [blame]3335<p>The <tt>val1</tt> and <tt>val2</tt> arguments must be
3336<a href="#t_floating">floating point</a> typed. They must have identical
3337types.</p>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3338<p>In the foregoing, <i>ordered</i> means that neither operand is a QNAN and
3339<i>unordered</i> means that either operand is a QNAN.</p>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3340<h5>Semantics:</h5>
3341<p>The '<tt>fcmp</tt>' compares <tt>var1</tt> and <tt>var2</tt> according to
3342the condition code given as <tt>cond</tt>. The comparison performed always
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]3343yields a <a href="#t_primitive">i1</a> result, as follows:
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3344<ol>
3345 <li><tt>false</tt>: always yields <tt>false</tt>, regardless of operands.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3346 <li><tt>oeq</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3347 <tt>var1</tt> is equal to <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3348 <li><tt>ogt</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3349 <tt>var1</tt> is greather than <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3350 <li><tt>oge</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3351 <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3352 <li><tt>olt</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3353 <tt>var1</tt> is less than <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3354 <li><tt>ole</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3355 <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3356 <li><tt>one</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3357 <tt>var1</tt> is not equal to <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3358 <li><tt>ord</tt>: yields <tt>true</tt> if both operands are not a QNAN.</li>
3359 <li><tt>ueq</tt>: yields <tt>true</tt> if either operand is a QNAN or
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3360 <tt>var1</tt> is equal to <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3361 <li><tt>ugt</tt>: yields <tt>true</tt> if either operand is a QNAN or
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3362 <tt>var1</tt> is greater than <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3363 <li><tt>uge</tt>: yields <tt>true</tt> if either operand is a QNAN or
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3364 <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3365 <li><tt>ult</tt>: yields <tt>true</tt> if either operand is a QNAN or
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3366 <tt>var1</tt> is less than <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3367 <li><tt>ule</tt>: yields <tt>true</tt> if either operand is a QNAN or
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3368 <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3369 <li><tt>une</tt>: yields <tt>true</tt> if either operand is a QNAN or
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3370 <tt>var1</tt> is not equal to <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3371 <li><tt>uno</tt>: yields <tt>true</tt> if either operand is a QNAN.</li>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3372 <li><tt>true</tt>: always yields <tt>true</tt>, regardless of operands.</li>
3373</ol>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3374
3375<h5>Example:</h5>
3376<pre> &lt;result&gt; = fcmp oeq float 4.0, 5.0 <i>; yields: result=false</i>
3377 &lt;result&gt; = icmp one float 4.0, 5.0 <i>; yields: result=true</i>
3378 &lt;result&gt; = icmp olt float 4.0, 5.0 <i>; yields: result=true</i>
3379 &lt;result&gt; = icmp ueq double 1.0, 2.0 <i>; yields: result=false</i>
3380</pre>
3381</div>
3382
Reid Spencer97c5fa42006-11-08 01:18:52 +0000[diff] [blame]3383<!-- _______________________________________________________________________ -->
3384<div class="doc_subsubsection"> <a name="i_phi">'<tt>phi</tt>'
3385Instruction</a> </div>
3386<div class="doc_text">
3387<h5>Syntax:</h5>
3388<pre> &lt;result&gt; = phi &lt;ty&gt; [ &lt;val0&gt;, &lt;label0&gt;], ...<br></pre>
3389<h5>Overview:</h5>
3390<p>The '<tt>phi</tt>' instruction is used to implement the &#966; node in
3391the SSA graph representing the function.</p>
3392<h5>Arguments:</h5>
3393<p>The type of the incoming values are specified with the first type
3394field. After this, the '<tt>phi</tt>' instruction takes a list of pairs
3395as arguments, with one pair for each predecessor basic block of the
3396current block. Only values of <a href="#t_firstclass">first class</a>
3397type may be used as the value arguments to the PHI node. Only labels
3398may be used as the label arguments.</p>
3399<p>There must be no non-phi instructions between the start of a basic
3400block and the PHI instructions: i.e. PHI instructions must be first in
3401a basic block.</p>
3402<h5>Semantics:</h5>
3403<p>At runtime, the '<tt>phi</tt>' instruction logically takes on the
3404value specified by the parameter, depending on which basic block we
3405came from in the last <a href="#terminators">terminator</a> instruction.</p>
3406<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3407<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 Spencer97c5fa42006-11-08 01:18:52 +0000[diff] [blame]3408</div>
3409
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]3410<!-- _______________________________________________________________________ -->
3411<div class="doc_subsubsection">
3412 <a name="i_select">'<tt>select</tt>' Instruction</a>
3413</div>
3414
3415<div class="doc_text">
3416
3417<h5>Syntax:</h5>
3418
3419<pre>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]3420 &lt;result&gt; = select i1 &lt;cond&gt;, &lt;ty&gt; &lt;val1&gt;, &lt;ty&gt; &lt;val2&gt; <i>; yields ty</i>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]3421</pre>
3422
3423<h5>Overview:</h5>
3424
3425<p>
3426The '<tt>select</tt>' instruction is used to choose one value based on a
3427condition, without branching.
3428</p>
3429
3430
3431<h5>Arguments:</h5>
3432
3433<p>
3434The '<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.
3435</p>
3436
3437<h5>Semantics:</h5>
3438
3439<p>
3440If the boolean condition evaluates to true, the instruction returns the first
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]3441value argument; otherwise, it returns the second value argument.
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]3442</p>
3443
3444<h5>Example:</h5>
3445
3446<pre>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]3447 %X = select i1 true, i8 17, i8 42 <i>; yields i8:17</i>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]3448</pre>
3449</div>
3450
Robert Bocchinof72fdfe2006-01-15 20:48:27 +0000[diff] [blame]3451
3452<!-- _______________________________________________________________________ -->
3453<div class="doc_subsubsection">
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3454 <a name="i_call">'<tt>call</tt>' Instruction</a>
3455</div>
3456
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3457<div class="doc_text">
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3458
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]3459<h5>Syntax:</h5>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3460<pre>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]3461 &lt;result&gt; = [tail] call [<a href="#callingconv">cconv</a>] &lt;ty&gt;* &lt;fnptrval&gt;(&lt;param list&gt;)
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3462</pre>
3463
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]3464<h5>Overview:</h5>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3465
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3466<p>The '<tt>call</tt>' instruction represents a simple function call.</p>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3467
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]3468<h5>Arguments:</h5>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3469
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3470<p>This instruction requires several arguments:</p>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3471
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]3472<ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3473 <li>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]3474 <p>The optional "tail" marker indicates whether the callee function accesses
3475 any allocas or varargs in the caller. If the "tail" marker is present, the
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3476 function call is eligible for tail call optimization. Note that calls may
3477 be marked "tail" even if they do not occur before a <a
3478 href="#i_ret"><tt>ret</tt></a> instruction.
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3479 </li>
3480 <li>
Duncan Sands16f122e2007-03-30 12:22:09 +0000[diff] [blame]3481 <p>The optional "cconv" marker indicates which <a href="#callingconv">calling
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]3482 convention</a> the call should use. If none is specified, the call defaults
3483 to using C calling conventions.
3484 </li>
3485 <li>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3486 <p>'<tt>ty</tt>': shall be the signature of the pointer to function value
3487 being invoked. The argument types must match the types implied by this
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]3488 signature. This type can be omitted if the function is not varargs and
3489 if the function type does not return a pointer to a function.</p>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3490 </li>
3491 <li>
3492 <p>'<tt>fnptrval</tt>': An LLVM value containing a pointer to a function to
3493 be invoked. In most cases, this is a direct function invocation, but
3494 indirect <tt>call</tt>s are just as possible, calling an arbitrary pointer
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]3495 to function value.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3496 </li>
3497 <li>
3498 <p>'<tt>function args</tt>': argument list whose types match the
Reid Spencerd845d162005-05-01 22:22:57 +0000[diff] [blame]3499 function signature argument types. All arguments must be of
3500 <a href="#t_firstclass">first class</a> type. If the function signature
3501 indicates the function accepts a variable number of arguments, the extra
3502 arguments can be specified.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3503 </li>
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]3504</ol>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3505
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]3506<h5>Semantics:</h5>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3507
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3508<p>The '<tt>call</tt>' instruction is used to cause control flow to
3509transfer to a specified function, with its incoming arguments bound to
3510the specified values. Upon a '<tt><a href="#i_ret">ret</a></tt>'
3511instruction in the called function, control flow continues with the
3512instruction after the function call, and the return value of the
3513function is bound to the result argument. This is a simpler case of
3514the <a href="#i_invoke">invoke</a> instruction.</p>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3515
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]3516<h5>Example:</h5>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3517
3518<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3519 %retval = call i32 %test(i32 %argc)
3520 call i32(i8 *, ...) *%printf(i8 * %msg, i32 12, i8 42);
3521 %X = tail call i32 %foo()
3522 %Y = tail call <a href="#callingconv">fastcc</a> i32 %foo()
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3523</pre>
3524
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3525</div>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3526
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3527<!-- _______________________________________________________________________ -->
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3528<div class="doc_subsubsection">
Chris Lattner33337472006-01-13 23:26:01 +0000[diff] [blame]3529 <a name="i_va_arg">'<tt>va_arg</tt>' Instruction</a>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3530</div>
3531
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3532<div class="doc_text">
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3533
Chris Lattner26ca62e2003-10-18 05:51:36 +0000[diff] [blame]3534<h5>Syntax:</h5>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3535
3536<pre>
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3537 &lt;resultval&gt; = va_arg &lt;va_list*&gt; &lt;arglist&gt;, &lt;argty&gt;
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3538</pre>
3539
Chris Lattner26ca62e2003-10-18 05:51:36 +0000[diff] [blame]3540<h5>Overview:</h5>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3541
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3542<p>The '<tt>va_arg</tt>' instruction is used to access arguments passed through
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3543the "variable argument" area of a function call. It is used to implement the
3544<tt>va_arg</tt> macro in C.</p>
3545
Chris Lattner26ca62e2003-10-18 05:51:36 +0000[diff] [blame]3546<h5>Arguments:</h5>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3547
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3548<p>This instruction takes a <tt>va_list*</tt> value and the type of
3549the argument. It returns a value of the specified argument type and
Jeff Cohendc6bfea2005-11-11 02:15:27 +0000[diff] [blame]3550increments the <tt>va_list</tt> to point to the next argument. Again, the
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3551actual type of <tt>va_list</tt> is target specific.</p>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3552
Chris Lattner26ca62e2003-10-18 05:51:36 +0000[diff] [blame]3553<h5>Semantics:</h5>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3554
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3555<p>The '<tt>va_arg</tt>' instruction loads an argument of the specified
3556type from the specified <tt>va_list</tt> and causes the
3557<tt>va_list</tt> to point to the next argument. For more information,
3558see the variable argument handling <a href="#int_varargs">Intrinsic
3559Functions</a>.</p>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3560
3561<p>It is legal for this instruction to be called in a function which does not
3562take a variable number of arguments, for example, the <tt>vfprintf</tt>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3563function.</p>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3564
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3565<p><tt>va_arg</tt> is an LLVM instruction instead of an <a
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]3566href="#intrinsics">intrinsic function</a> because it takes a type as an
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3567argument.</p>
3568
Chris Lattner26ca62e2003-10-18 05:51:36 +0000[diff] [blame]3569<h5>Example:</h5>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3570
3571<p>See the <a href="#int_varargs">variable argument processing</a> section.</p>
3572
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3573</div>
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3574
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3575<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3576<div class="doc_section"> <a name="intrinsics">Intrinsic Functions</a> </div>
3577<!-- *********************************************************************** -->
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3578
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3579<div class="doc_text">
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3580
3581<p>LLVM supports the notion of an "intrinsic function". These functions have
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]3582well known names and semantics and are required to follow certain restrictions.
3583Overall, these intrinsics represent an extension mechanism for the LLVM
3584language that does not require changing all of the transformations in LLVM to
3585add to the language (or the bytecode reader/writer, the parser,
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3586etc...).</p>
3587
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]3588<p>Intrinsic function names must all start with an "<tt>llvm.</tt>" prefix. This
3589prefix is reserved in LLVM for intrinsic names; thus, functions may not be named
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3590this. Intrinsic functions must always be external functions: you cannot define
3591the body of intrinsic functions. Intrinsic functions may only be used in call
3592or invoke instructions: it is illegal to take the address of an intrinsic
3593function. Additionally, because intrinsic functions are part of the LLVM
3594language, it is required that they all be documented here if any are added.</p>
3595
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]3596<p>Some intrinsic functions can be overloaded. That is, the intrinsic represents
3597a family of functions that perform the same operation but on different data
3598types. This is most frequent with the integer types. Since LLVM can represent
3599over 8 million different integer types, there is a way to declare an intrinsic
3600that can be overloaded based on its arguments. Such intrinsics will have the
3601names of the arbitrary types encoded into the intrinsic function name, each
3602preceded by a period. For example, the <tt>llvm.ctpop</tt> function can take an
3603integer of any width. This leads to a family of functions such as
3604<tt>i32 @llvm.ctpop.i8(i8 %val)</tt> and <tt>i32 @llvm.ctpop.i29(i29 %val)</tt>.
3605</p>
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3606
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]3607
3608<p>To learn how to add an intrinsic function, please see the
3609<a href="ExtendingLLVM.html">Extending LLVM Guide</a>.
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3610</p>
3611
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3612</div>
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3613
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3614<!-- ======================================================================= -->
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3615<div class="doc_subsection">
3616 <a name="int_varargs">Variable Argument Handling Intrinsics</a>
3617</div>
3618
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3619<div class="doc_text">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3620
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3621<p>Variable argument support is defined in LLVM with the <a
Chris Lattner33337472006-01-13 23:26:01 +0000[diff] [blame]3622 href="#i_va_arg"><tt>va_arg</tt></a> instruction and these three
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3623intrinsic functions. These functions are related to the similarly
3624named macros defined in the <tt>&lt;stdarg.h&gt;</tt> header file.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3625
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3626<p>All of these functions operate on arguments that use a
3627target-specific value type "<tt>va_list</tt>". The LLVM assembly
3628language reference manual does not define what this type is, so all
3629transformations should be prepared to handle intrinsics with any type
3630used.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3631
Chris Lattner30b868d2006-05-15 17:26:46 +0000[diff] [blame]3632<p>This example shows how the <a href="#i_va_arg"><tt>va_arg</tt></a>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3633instruction and the variable argument handling intrinsic functions are
3634used.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3635
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3636<pre>
Anton Korobeynikov640bbe02007-03-21 23:58:04 +0000[diff] [blame]3637define i32 @test(i32 %X, ...) {
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3638 ; Initialize variable argument processing
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3639 %ap = alloca i8 *
Chris Lattnerdb0790c2007-01-08 07:55:15 +0000[diff] [blame]3640 %ap2 = bitcast i8** %ap to i8*
Anton Korobeynikov640bbe02007-03-21 23:58:04 +0000[diff] [blame]3641 call void @llvm.va_start(i8* %ap2)
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3642
3643 ; Read a single integer argument
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3644 %tmp = va_arg i8 ** %ap, i32
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3645
3646 ; Demonstrate usage of llvm.va_copy and llvm.va_end
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3647 %aq = alloca i8 *
Chris Lattnerdb0790c2007-01-08 07:55:15 +0000[diff] [blame]3648 %aq2 = bitcast i8** %aq to i8*
Anton Korobeynikov640bbe02007-03-21 23:58:04 +0000[diff] [blame]3649 call void @llvm.va_copy(i8 *%aq2, i8* %ap2)
3650 call void @llvm.va_end(i8* %aq2)
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3651
3652 ; Stop processing of arguments.
Anton Korobeynikov640bbe02007-03-21 23:58:04 +0000[diff] [blame]3653 call void @llvm.va_end(i8* %ap2)
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3654 ret i32 %tmp
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3655}
Anton Korobeynikov640bbe02007-03-21 23:58:04 +0000[diff] [blame]3656
3657declare void @llvm.va_start(i8*)
3658declare void @llvm.va_copy(i8*, i8*)
3659declare void @llvm.va_end(i8*)
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3660</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3661</div>
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3662
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3663<!-- _______________________________________________________________________ -->
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3664<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]3665 <a name="int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a>
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3666</div>
3667
3668
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3669<div class="doc_text">
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3670<h5>Syntax:</h5>
Chris Lattnerdb0790c2007-01-08 07:55:15 +0000[diff] [blame]3671<pre> declare void %llvm.va_start(i8* &lt;arglist&gt;)<br></pre>
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3672<h5>Overview:</h5>
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3673<P>The '<tt>llvm.va_start</tt>' intrinsic initializes
3674<tt>*&lt;arglist&gt;</tt> for subsequent use by <tt><a
3675href="#i_va_arg">va_arg</a></tt>.</p>
3676
3677<h5>Arguments:</h5>
3678
3679<P>The argument is a pointer to a <tt>va_list</tt> element to initialize.</p>
3680
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3681<h5>Semantics:</h5>
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3682
3683<P>The '<tt>llvm.va_start</tt>' intrinsic works just like the <tt>va_start</tt>
3684macro available in C. In a target-dependent way, it initializes the
3685<tt>va_list</tt> element the argument points to, so that the next call to
3686<tt>va_arg</tt> will produce the first variable argument passed to the function.
3687Unlike the C <tt>va_start</tt> macro, this intrinsic does not need to know the
3688last argument of the function, the compiler can figure that out.</p>
3689
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3690</div>
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3691
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3692<!-- _______________________________________________________________________ -->
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3693<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]3694 <a name="int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a>
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3695</div>
3696
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3697<div class="doc_text">
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3698<h5>Syntax:</h5>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]3699<pre> declare void @llvm.va_end(i8* &lt;arglist&gt;)<br></pre>
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3700<h5>Overview:</h5>
Chris Lattnerdb0790c2007-01-08 07:55:15 +0000[diff] [blame]3701
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3702<p>The '<tt>llvm.va_end</tt>' intrinsic destroys <tt>&lt;arglist&gt;</tt>
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]3703which has been initialized previously with <tt><a href="#int_va_start">llvm.va_start</a></tt>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3704or <tt><a href="#i_va_copy">llvm.va_copy</a></tt>.</p>
Chris Lattnerdb0790c2007-01-08 07:55:15 +0000[diff] [blame]3705
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3706<h5>Arguments:</h5>
Chris Lattnerdb0790c2007-01-08 07:55:15 +0000[diff] [blame]3707
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3708<p>The argument is a <tt>va_list</tt> to destroy.</p>
Chris Lattnerdb0790c2007-01-08 07:55:15 +0000[diff] [blame]3709
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3710<h5>Semantics:</h5>
Chris Lattnerdb0790c2007-01-08 07:55:15 +0000[diff] [blame]3711
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3712<p>The '<tt>llvm.va_end</tt>' intrinsic works just like the <tt>va_end</tt>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3713macro available in C. In a target-dependent way, it destroys the <tt>va_list</tt>.
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]3714Calls to <a href="#int_va_start"><tt>llvm.va_start</tt></a> and <a
3715 href="#int_va_copy"><tt>llvm.va_copy</tt></a> must be matched exactly
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3716with calls to <tt>llvm.va_end</tt>.</p>
Chris Lattnerdb0790c2007-01-08 07:55:15 +0000[diff] [blame]3717
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3718</div>
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3719
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3720<!-- _______________________________________________________________________ -->
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3721<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]3722 <a name="int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a>
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3723</div>
3724
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3725<div class="doc_text">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3726
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3727<h5>Syntax:</h5>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3728
3729<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]3730 declare void @llvm.va_copy(i8* &lt;destarglist&gt;, i8* &lt;srcarglist&gt;)
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3731</pre>
3732
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3733<h5>Overview:</h5>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3734
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3735<p>The '<tt>llvm.va_copy</tt>' intrinsic copies the current argument position from
3736the source argument list to the destination argument list.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3737
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3738<h5>Arguments:</h5>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3739
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3740<p>The first argument is a pointer to a <tt>va_list</tt> element to initialize.
Andrew Lenharth5305ea52005-06-22 20:38:11 +0000[diff] [blame]3741The second argument is a pointer to a <tt>va_list</tt> element to copy from.</p>
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3742
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3743
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3744<h5>Semantics:</h5>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3745
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3746<p>The '<tt>llvm.va_copy</tt>' intrinsic works just like the <tt>va_copy</tt> macro
3747available in C. In a target-dependent way, it copies the source
3748<tt>va_list</tt> element into the destination list. This intrinsic is necessary
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]3749because the <tt><a href="#int_va_start">llvm.va_start</a></tt> intrinsic may be
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3750arbitrarily complex and require memory allocation, for example.</p>
3751
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3752</div>
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3753
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3754<!-- ======================================================================= -->
3755<div class="doc_subsection">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3756 <a name="int_gc">Accurate Garbage Collection Intrinsics</a>
3757</div>
3758
3759<div class="doc_text">
3760
3761<p>
3762LLVM support for <a href="GarbageCollection.html">Accurate Garbage
3763Collection</a> requires the implementation and generation of these intrinsics.
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]3764These intrinsics allow identification of <a href="#int_gcroot">GC roots on the
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3765stack</a>, as well as garbage collector implementations that require <a
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]3766href="#int_gcread">read</a> and <a href="#int_gcwrite">write</a> barriers.
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3767Front-ends for type-safe garbage collected languages should generate these
3768intrinsics to make use of the LLVM garbage collectors. For more details, see <a
3769href="GarbageCollection.html">Accurate Garbage Collection with LLVM</a>.
3770</p>
3771</div>
3772
3773<!-- _______________________________________________________________________ -->
3774<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]3775 <a name="int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3776</div>
3777
3778<div class="doc_text">
3779
3780<h5>Syntax:</h5>
3781
3782<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]3783 declare void @llvm.gcroot(&lt;ty&gt;** %ptrloc, &lt;ty2&gt;* %metadata)
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3784</pre>
3785
3786<h5>Overview:</h5>
3787
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]3788<p>The '<tt>llvm.gcroot</tt>' intrinsic declares the existence of a GC root to
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3789the code generator, and allows some metadata to be associated with it.</p>
3790
3791<h5>Arguments:</h5>
3792
3793<p>The first argument specifies the address of a stack object that contains the
3794root pointer. The second pointer (which must be either a constant or a global
3795value address) contains the meta-data to be associated with the root.</p>
3796
3797<h5>Semantics:</h5>
3798
3799<p>At runtime, a call to this intrinsics stores a null pointer into the "ptrloc"
3800location. At compile-time, the code generator generates information to allow
3801the runtime to find the pointer at GC safe points.
3802</p>
3803
3804</div>
3805
3806
3807<!-- _______________________________________________________________________ -->
3808<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]3809 <a name="int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3810</div>
3811
3812<div class="doc_text">
3813
3814<h5>Syntax:</h5>
3815
3816<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]3817 declare i8 * @llvm.gcread(i8 * %ObjPtr, i8 ** %Ptr)
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3818</pre>
3819
3820<h5>Overview:</h5>
3821
3822<p>The '<tt>llvm.gcread</tt>' intrinsic identifies reads of references from heap
3823locations, allowing garbage collector implementations that require read
3824barriers.</p>
3825
3826<h5>Arguments:</h5>
3827
Chris Lattnerf9228072006-03-14 20:02:51 +0000[diff] [blame]3828<p>The second argument is the address to read from, which should be an address
3829allocated from the garbage collector. The first object is a pointer to the
3830start of the referenced object, if needed by the language runtime (otherwise
3831null).</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3832
3833<h5>Semantics:</h5>
3834
3835<p>The '<tt>llvm.gcread</tt>' intrinsic has the same semantics as a load
3836instruction, but may be replaced with substantially more complex code by the
3837garbage collector runtime, as needed.</p>
3838
3839</div>
3840
3841
3842<!-- _______________________________________________________________________ -->
3843<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]3844 <a name="int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3845</div>
3846
3847<div class="doc_text">
3848
3849<h5>Syntax:</h5>
3850
3851<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]3852 declare void @llvm.gcwrite(i8 * %P1, i8 * %Obj, i8 ** %P2)
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3853</pre>
3854
3855<h5>Overview:</h5>
3856
3857<p>The '<tt>llvm.gcwrite</tt>' intrinsic identifies writes of references to heap
3858locations, allowing garbage collector implementations that require write
3859barriers (such as generational or reference counting collectors).</p>
3860
3861<h5>Arguments:</h5>
3862
Chris Lattnerf9228072006-03-14 20:02:51 +0000[diff] [blame]3863<p>The first argument is the reference to store, the second is the start of the
3864object to store it to, and the third is the address of the field of Obj to
3865store to. If the runtime does not require a pointer to the object, Obj may be
3866null.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3867
3868<h5>Semantics:</h5>
3869
3870<p>The '<tt>llvm.gcwrite</tt>' intrinsic has the same semantics as a store
3871instruction, but may be replaced with substantially more complex code by the
3872garbage collector runtime, as needed.</p>
3873
3874</div>
3875
3876
3877
3878<!-- ======================================================================= -->
3879<div class="doc_subsection">
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3880 <a name="int_codegen">Code Generator Intrinsics</a>
3881</div>
3882
3883<div class="doc_text">
3884<p>
3885These intrinsics are provided by LLVM to expose special features that may only
3886be implemented with code generator support.
3887</p>
3888
3889</div>
3890
3891<!-- _______________________________________________________________________ -->
3892<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]3893 <a name="int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a>
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3894</div>
3895
3896<div class="doc_text">
3897
3898<h5>Syntax:</h5>
3899<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]3900 declare i8 *@llvm.returnaddress(i32 &lt;level&gt;)
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3901</pre>
3902
3903<h5>Overview:</h5>
3904
3905<p>
Chris Lattnerc1fb4262006-10-15 20:05:59 +0000[diff] [blame]3906The '<tt>llvm.returnaddress</tt>' intrinsic attempts to compute a
3907target-specific value indicating the return address of the current function
3908or one of its callers.
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3909</p>
3910
3911<h5>Arguments:</h5>
3912
3913<p>
3914The argument to this intrinsic indicates which function to return the address
3915for. Zero indicates the calling function, one indicates its caller, etc. The
3916argument is <b>required</b> to be a constant integer value.
3917</p>
3918
3919<h5>Semantics:</h5>
3920
3921<p>
3922The '<tt>llvm.returnaddress</tt>' intrinsic either returns a pointer indicating
3923the return address of the specified call frame, or zero if it cannot be
3924identified. The value returned by this intrinsic is likely to be incorrect or 0
3925for arguments other than zero, so it should only be used for debugging purposes.
3926</p>
3927
3928<p>
3929Note that calling this intrinsic does not prevent function inlining or other
Chris Lattner2e6eb5f2005-03-07 20:30:51 +0000[diff] [blame]3930aggressive transformations, so the value returned may not be that of the obvious
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3931source-language caller.
3932</p>
3933</div>
3934
3935
3936<!-- _______________________________________________________________________ -->
3937<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]3938 <a name="int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a>
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3939</div>
3940
3941<div class="doc_text">
3942
3943<h5>Syntax:</h5>
3944<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]3945 declare i8 *@llvm.frameaddress(i32 &lt;level&gt;)
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3946</pre>
3947
3948<h5>Overview:</h5>
3949
3950<p>
Chris Lattnerc1fb4262006-10-15 20:05:59 +0000[diff] [blame]3951The '<tt>llvm.frameaddress</tt>' intrinsic attempts to return the
3952target-specific frame pointer value for the specified stack frame.
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3953</p>
3954
3955<h5>Arguments:</h5>
3956
3957<p>
3958The argument to this intrinsic indicates which function to return the frame
3959pointer for. Zero indicates the calling function, one indicates its caller,
3960etc. The argument is <b>required</b> to be a constant integer value.
3961</p>
3962
3963<h5>Semantics:</h5>
3964
3965<p>
3966The '<tt>llvm.frameaddress</tt>' intrinsic either returns a pointer indicating
3967the frame address of the specified call frame, or zero if it cannot be
3968identified. The value returned by this intrinsic is likely to be incorrect or 0
3969for arguments other than zero, so it should only be used for debugging purposes.
3970</p>
3971
3972<p>
3973Note that calling this intrinsic does not prevent function inlining or other
Chris Lattner2e6eb5f2005-03-07 20:30:51 +0000[diff] [blame]3974aggressive transformations, so the value returned may not be that of the obvious
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3975source-language caller.
3976</p>
3977</div>
3978
Chris Lattnerc8a2c222005-02-28 19:24:19 +0000[diff] [blame]3979<!-- _______________________________________________________________________ -->
3980<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]3981 <a name="int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a>
Chris Lattner2f0f0012006-01-13 02:03:13 +0000[diff] [blame]3982</div>
3983
3984<div class="doc_text">
3985
3986<h5>Syntax:</h5>
3987<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]3988 declare i8 *@llvm.stacksave()
Chris Lattner2f0f0012006-01-13 02:03:13 +0000[diff] [blame]3989</pre>
3990
3991<h5>Overview:</h5>
3992
3993<p>
3994The '<tt>llvm.stacksave</tt>' intrinsic is used to remember the current state of
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]3995the function stack, for use with <a href="#int_stackrestore">
Chris Lattner2f0f0012006-01-13 02:03:13 +0000[diff] [blame]3996<tt>llvm.stackrestore</tt></a>. This is useful for implementing language
3997features like scoped automatic variable sized arrays in C99.
3998</p>
3999
4000<h5>Semantics:</h5>
4001
4002<p>
4003This intrinsic returns a opaque pointer value that can be passed to <a
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4004href="#int_stackrestore"><tt>llvm.stackrestore</tt></a>. When an
Chris Lattner2f0f0012006-01-13 02:03:13 +0000[diff] [blame]4005<tt>llvm.stackrestore</tt> intrinsic is executed with a value saved from
4006<tt>llvm.stacksave</tt>, it effectively restores the state of the stack to the
4007state it was in when the <tt>llvm.stacksave</tt> intrinsic executed. In
4008practice, this pops any <a href="#i_alloca">alloca</a> blocks from the stack
4009that were allocated after the <tt>llvm.stacksave</tt> was executed.
4010</p>
4011
4012</div>
4013
4014<!-- _______________________________________________________________________ -->
4015<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4016 <a name="int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a>
Chris Lattner2f0f0012006-01-13 02:03:13 +0000[diff] [blame]4017</div>
4018
4019<div class="doc_text">
4020
4021<h5>Syntax:</h5>
4022<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4023 declare void @llvm.stackrestore(i8 * %ptr)
Chris Lattner2f0f0012006-01-13 02:03:13 +0000[diff] [blame]4024</pre>
4025
4026<h5>Overview:</h5>
4027
4028<p>
4029The '<tt>llvm.stackrestore</tt>' intrinsic is used to restore the state of
4030the function stack to the state it was in when the corresponding <a
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4031href="#int_stacksave"><tt>llvm.stacksave</tt></a> intrinsic executed. This is
Chris Lattner2f0f0012006-01-13 02:03:13 +0000[diff] [blame]4032useful for implementing language features like scoped automatic variable sized
4033arrays in C99.
4034</p>
4035
4036<h5>Semantics:</h5>
4037
4038<p>
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4039See the description for <a href="#int_stacksave"><tt>llvm.stacksave</tt></a>.
Chris Lattner2f0f0012006-01-13 02:03:13 +0000[diff] [blame]4040</p>
4041
4042</div>
4043
4044
4045<!-- _______________________________________________________________________ -->
4046<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4047 <a name="int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a>
Chris Lattnerc8a2c222005-02-28 19:24:19 +0000[diff] [blame]4048</div>
4049
4050<div class="doc_text">
4051
4052<h5>Syntax:</h5>
4053<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4054 declare void @llvm.prefetch(i8 * &lt;address&gt;,
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]4055 i32 &lt;rw&gt;, i32 &lt;locality&gt;)
Chris Lattnerc8a2c222005-02-28 19:24:19 +0000[diff] [blame]4056</pre>
4057
4058<h5>Overview:</h5>
4059
4060
4061<p>
4062The '<tt>llvm.prefetch</tt>' intrinsic is a hint to the code generator to insert
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]4063a prefetch instruction if supported; otherwise, it is a noop. Prefetches have
4064no
4065effect on the behavior of the program but can change its performance
Chris Lattnerff851072005-02-28 19:47:14 +0000[diff] [blame]4066characteristics.
Chris Lattnerc8a2c222005-02-28 19:24:19 +0000[diff] [blame]4067</p>
4068
4069<h5>Arguments:</h5>
4070
4071<p>
4072<tt>address</tt> is the address to be prefetched, <tt>rw</tt> is the specifier
4073determining if the fetch should be for a read (0) or write (1), and
4074<tt>locality</tt> is a temporal locality specifier ranging from (0) - no
Chris Lattnerd3e641c2005-03-07 20:31:38 +0000[diff] [blame]4075locality, to (3) - extremely local keep in cache. The <tt>rw</tt> and
Chris Lattnerc8a2c222005-02-28 19:24:19 +0000[diff] [blame]4076<tt>locality</tt> arguments must be constant integers.
4077</p>
4078
4079<h5>Semantics:</h5>
4080
4081<p>
4082This intrinsic does not modify the behavior of the program. In particular,
4083prefetches cannot trap and do not produce a value. On targets that support this
4084intrinsic, the prefetch can provide hints to the processor cache for better
4085performance.
4086</p>
4087
4088</div>
4089
Andrew Lenharthb4427912005-03-28 20:05:49 +0000[diff] [blame]4090<!-- _______________________________________________________________________ -->
4091<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4092 <a name="int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a>
Andrew Lenharthb4427912005-03-28 20:05:49 +0000[diff] [blame]4093</div>
4094
4095<div class="doc_text">
4096
4097<h5>Syntax:</h5>
4098<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4099 declare void @llvm.pcmarker( i32 &lt;id&gt; )
Andrew Lenharthb4427912005-03-28 20:05:49 +0000[diff] [blame]4100</pre>
4101
4102<h5>Overview:</h5>
4103
4104
4105<p>
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]4106The '<tt>llvm.pcmarker</tt>' intrinsic is a method to export a Program Counter
4107(PC) in a region of
Andrew Lenharthb4427912005-03-28 20:05:49 +0000[diff] [blame]4108code to simulators and other tools. The method is target specific, but it is
4109expected that the marker will use exported symbols to transmit the PC of the marker.
Jeff Cohendc6bfea2005-11-11 02:15:27 +0000[diff] [blame]4110The marker makes no guarantees that it will remain with any specific instruction
Chris Lattnere64d41d2005-11-15 06:07:55 +0000[diff] [blame]4111after optimizations. It is possible that the presence of a marker will inhibit
Chris Lattnerb40261e2006-03-24 07:16:10 +0000[diff] [blame]4112optimizations. The intended use is to be inserted after optimizations to allow
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]4113correlations of simulation runs.
Andrew Lenharthb4427912005-03-28 20:05:49 +0000[diff] [blame]4114</p>
4115
4116<h5>Arguments:</h5>
4117
4118<p>
4119<tt>id</tt> is a numerical id identifying the marker.
4120</p>
4121
4122<h5>Semantics:</h5>
4123
4124<p>
4125This intrinsic does not modify the behavior of the program. Backends that do not
4126support this intrinisic may ignore it.
4127</p>
4128
4129</div>
4130
Andrew Lenharth01aa5632005-11-11 16:47:30 +0000[diff] [blame]4131<!-- _______________________________________________________________________ -->
4132<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4133 <a name="int_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a>
Andrew Lenharth01aa5632005-11-11 16:47:30 +0000[diff] [blame]4134</div>
4135
4136<div class="doc_text">
4137
4138<h5>Syntax:</h5>
4139<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4140 declare i64 @llvm.readcyclecounter( )
Andrew Lenharth01aa5632005-11-11 16:47:30 +0000[diff] [blame]4141</pre>
4142
4143<h5>Overview:</h5>
4144
4145
4146<p>
4147The '<tt>llvm.readcyclecounter</tt>' intrinsic provides access to the cycle
4148counter register (or similar low latency, high accuracy clocks) on those targets
4149that support it. On X86, it should map to RDTSC. On Alpha, it should map to RPCC.
4150As the backing counters overflow quickly (on the order of 9 seconds on alpha), this
4151should only be used for small timings.
4152</p>
4153
4154<h5>Semantics:</h5>
4155
4156<p>
4157When directly supported, reading the cycle counter should not modify any memory.
4158Implementations are allowed to either return a application specific value or a
4159system wide value. On backends without support, this is lowered to a constant 0.
4160</p>
4161
4162</div>
4163
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4164<!-- ======================================================================= -->
4165<div class="doc_subsection">
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]4166 <a name="int_libc">Standard C Library Intrinsics</a>
4167</div>
4168
4169<div class="doc_text">
4170<p>
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4171LLVM provides intrinsics for a few important standard C library functions.
4172These intrinsics allow source-language front-ends to pass information about the
4173alignment of the pointer arguments to the code generator, providing opportunity
4174for more efficient code generation.
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]4175</p>
4176
4177</div>
4178
4179<!-- _______________________________________________________________________ -->
4180<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4181 <a name="int_memcpy">'<tt>llvm.memcpy</tt>' Intrinsic</a>
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]4182</div>
4183
4184<div class="doc_text">
4185
4186<h5>Syntax:</h5>
4187<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4188 declare void @llvm.memcpy.i32(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]4189 i32 &lt;len&gt;, i32 &lt;align&gt;)
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4190 declare void @llvm.memcpy.i64(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]4191 i64 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]4192</pre>
4193
4194<h5>Overview:</h5>
4195
4196<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4197The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]4198location to the destination location.
4199</p>
4200
4201<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4202Note that, unlike the standard libc function, the <tt>llvm.memcpy.*</tt>
4203intrinsics do not return a value, and takes an extra alignment argument.
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]4204</p>
4205
4206<h5>Arguments:</h5>
4207
4208<p>
4209The first argument is a pointer to the destination, the second is a pointer to
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4210the source. The third argument is an integer argument
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]4211specifying the number of bytes to copy, and the fourth argument is the alignment
4212of the source and destination locations.
4213</p>
4214
Chris Lattner4c67c482004-02-12 21:18:15 +0000[diff] [blame]4215<p>
4216If the call to this intrinisic has an alignment value that is not 0 or 1, then
Chris Lattner5316e5d2006-03-04 00:02:10 +0000[diff] [blame]4217the caller guarantees that both the source and destination pointers are aligned
4218to that boundary.
Chris Lattner4c67c482004-02-12 21:18:15 +0000[diff] [blame]4219</p>
4220
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]4221<h5>Semantics:</h5>
4222
4223<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4224The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]4225location to the destination location, which are not allowed to overlap. It
4226copies "len" bytes of memory over. If the argument is known to be aligned to
4227some boundary, this can be specified as the fourth argument, otherwise it should
4228be set to 0 or 1.
4229</p>
4230</div>
4231
4232
Chris Lattnerf30152e2004-02-12 18:10:10 +0000[diff] [blame]4233<!-- _______________________________________________________________________ -->
4234<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4235 <a name="int_memmove">'<tt>llvm.memmove</tt>' Intrinsic</a>
Chris Lattnerf30152e2004-02-12 18:10:10 +0000[diff] [blame]4236</div>
4237
4238<div class="doc_text">
4239
4240<h5>Syntax:</h5>
4241<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4242 declare void @llvm.memmove.i32(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]4243 i32 &lt;len&gt;, i32 &lt;align&gt;)
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4244 declare void @llvm.memmove.i64(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]4245 i64 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattnerf30152e2004-02-12 18:10:10 +0000[diff] [blame]4246</pre>
4247
4248<h5>Overview:</h5>
4249
4250<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4251The '<tt>llvm.memmove.*</tt>' intrinsics move a block of memory from the source
4252location to the destination location. It is similar to the
4253'<tt>llvm.memcmp</tt>' intrinsic but allows the two memory locations to overlap.
Chris Lattnerf30152e2004-02-12 18:10:10 +0000[diff] [blame]4254</p>
4255
4256<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4257Note that, unlike the standard libc function, the <tt>llvm.memmove.*</tt>
4258intrinsics do not return a value, and takes an extra alignment argument.
Chris Lattnerf30152e2004-02-12 18:10:10 +0000[diff] [blame]4259</p>
4260
4261<h5>Arguments:</h5>
4262
4263<p>
4264The first argument is a pointer to the destination, the second is a pointer to
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4265the source. The third argument is an integer argument
Chris Lattnerf30152e2004-02-12 18:10:10 +0000[diff] [blame]4266specifying the number of bytes to copy, and the fourth argument is the alignment
4267of the source and destination locations.
4268</p>
4269
Chris Lattner4c67c482004-02-12 21:18:15 +0000[diff] [blame]4270<p>
4271If the call to this intrinisic has an alignment value that is not 0 or 1, then
Chris Lattner5316e5d2006-03-04 00:02:10 +0000[diff] [blame]4272the caller guarantees that the source and destination pointers are aligned to
4273that boundary.
Chris Lattner4c67c482004-02-12 21:18:15 +0000[diff] [blame]4274</p>
4275
Chris Lattnerf30152e2004-02-12 18:10:10 +0000[diff] [blame]4276<h5>Semantics:</h5>
4277
4278<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4279The '<tt>llvm.memmove.*</tt>' intrinsics copy a block of memory from the source
Chris Lattnerf30152e2004-02-12 18:10:10 +0000[diff] [blame]4280location to the destination location, which may overlap. It
4281copies "len" bytes of memory over. If the argument is known to be aligned to
4282some boundary, this can be specified as the fourth argument, otherwise it should
4283be set to 0 or 1.
4284</p>
4285</div>
4286
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]4287
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4288<!-- _______________________________________________________________________ -->
4289<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4290 <a name="int_memset">'<tt>llvm.memset.*</tt>' Intrinsics</a>
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4291</div>
4292
4293<div class="doc_text">
4294
4295<h5>Syntax:</h5>
4296<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4297 declare void @llvm.memset.i32(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]4298 i32 &lt;len&gt;, i32 &lt;align&gt;)
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4299 declare void @llvm.memset.i64(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]4300 i64 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4301</pre>
4302
4303<h5>Overview:</h5>
4304
4305<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4306The '<tt>llvm.memset.*</tt>' intrinsics fill a block of memory with a particular
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4307byte value.
4308</p>
4309
4310<p>
4311Note that, unlike the standard libc function, the <tt>llvm.memset</tt> intrinsic
4312does not return a value, and takes an extra alignment argument.
4313</p>
4314
4315<h5>Arguments:</h5>
4316
4317<p>
4318The first argument is a pointer to the destination to fill, the second is the
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4319byte value to fill it with, the third argument is an integer
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4320argument specifying the number of bytes to fill, and the fourth argument is the
4321known alignment of destination location.
4322</p>
4323
4324<p>
4325If the call to this intrinisic has an alignment value that is not 0 or 1, then
Chris Lattner5316e5d2006-03-04 00:02:10 +0000[diff] [blame]4326the caller guarantees that the destination pointer is aligned to that boundary.
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4327</p>
4328
4329<h5>Semantics:</h5>
4330
4331<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4332The '<tt>llvm.memset.*</tt>' intrinsics fill "len" bytes of memory starting at
4333the
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4334destination location. If the argument is known to be aligned to some boundary,
4335this can be specified as the fourth argument, otherwise it should be set to 0 or
43361.
4337</p>
4338</div>
4339
4340
Chris Lattner3b4f4372004-06-11 02:28:03 +0000[diff] [blame]4341<!-- _______________________________________________________________________ -->
4342<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4343 <a name="int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a>
Chris Lattner8a8f2e52005-07-21 01:29:16 +0000[diff] [blame]4344</div>
4345
4346<div class="doc_text">
4347
4348<h5>Syntax:</h5>
4349<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4350 declare float @llvm.sqrt.f32(float %Val)
4351 declare double @llvm.sqrt.f64(double %Val)
Chris Lattner8a8f2e52005-07-21 01:29:16 +0000[diff] [blame]4352</pre>
4353
4354<h5>Overview:</h5>
4355
4356<p>
Reid Spencerb4f9a6f2006-01-16 21:12:35 +0000[diff] [blame]4357The '<tt>llvm.sqrt</tt>' intrinsics return the sqrt of the specified operand,
Chris Lattner8a8f2e52005-07-21 01:29:16 +0000[diff] [blame]4358returning the same value as the libm '<tt>sqrt</tt>' function would. Unlike
4359<tt>sqrt</tt> in libm, however, <tt>llvm.sqrt</tt> has undefined behavior for
4360negative numbers (which allows for better optimization).
4361</p>
4362
4363<h5>Arguments:</h5>
4364
4365<p>
4366The argument and return value are floating point numbers of the same type.
4367</p>
4368
4369<h5>Semantics:</h5>
4370
4371<p>
4372This function returns the sqrt of the specified operand if it is a positive
4373floating point number.
4374</p>
4375</div>
4376
Chris Lattner33b73f92006-09-08 06:34:02 +0000[diff] [blame]4377<!-- _______________________________________________________________________ -->
4378<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4379 <a name="int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a>
Chris Lattner33b73f92006-09-08 06:34:02 +0000[diff] [blame]4380</div>
4381
4382<div class="doc_text">
4383
4384<h5>Syntax:</h5>
4385<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4386 declare float @llvm.powi.f32(float %Val, i32 %power)
4387 declare double @llvm.powi.f64(double %Val, i32 %power)
Chris Lattner33b73f92006-09-08 06:34:02 +0000[diff] [blame]4388</pre>
4389
4390<h5>Overview:</h5>
4391
4392<p>
4393The '<tt>llvm.powi.*</tt>' intrinsics return the first operand raised to the
4394specified (positive or negative) power. The order of evaluation of
4395multiplications is not defined.
4396</p>
4397
4398<h5>Arguments:</h5>
4399
4400<p>
4401The second argument is an integer power, and the first is a value to raise to
4402that power.
4403</p>
4404
4405<h5>Semantics:</h5>
4406
4407<p>
4408This function returns the first value raised to the second power with an
4409unspecified sequence of rounding operations.</p>
4410</div>
4411
4412
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4413<!-- ======================================================================= -->
4414<div class="doc_subsection">
Nate Begeman0f223bb2006-01-13 23:26:38 +0000[diff] [blame]4415 <a name="int_manip">Bit Manipulation Intrinsics</a>
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4416</div>
4417
4418<div class="doc_text">
4419<p>
Nate Begeman0f223bb2006-01-13 23:26:38 +0000[diff] [blame]4420LLVM provides intrinsics for a few important bit manipulation operations.
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4421These allow efficient code generation for some algorithms.
4422</p>
4423
4424</div>
4425
4426<!-- _______________________________________________________________________ -->
4427<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4428 <a name="int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a>
Nate Begeman0f223bb2006-01-13 23:26:38 +0000[diff] [blame]4429</div>
4430
4431<div class="doc_text">
4432
4433<h5>Syntax:</h5>
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]4434<p>This is an overloaded intrinsic function. You can use bswap on any integer
4435type that is an even number of bytes (i.e. BitWidth % 16 == 0). Note the suffix
4436that includes the type for the result and the operand.
Nate Begeman0f223bb2006-01-13 23:26:38 +0000[diff] [blame]4437<pre>
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]4438 declare i16 @llvm.bswap.i16.i16(i16 &lt;id&gt;)
4439 declare i32 @llvm.bswap.i32.i32(i32 &lt;id&gt;)
Reid Spencer403a1c42007-04-02 00:19:52 +0000[diff] [blame]4440 declare i64 @llvm.bswap.i64.i64(i64 &lt;id&gt;)
Nate Begeman0f223bb2006-01-13 23:26:38 +0000[diff] [blame]4441</pre>
4442
4443<h5>Overview:</h5>
4444
4445<p>
Reid Spencerf361c4f2007-04-02 02:25:19 +0000[diff] [blame]4446The '<tt>llvm.bswap</tt>' family of intrinsics is used to byte swap integer
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]4447values with an even number of bytes (positive multiple of 16 bits). These are
4448useful for performing operations on data that is not in the target's native
4449byte order.
Nate Begeman0f223bb2006-01-13 23:26:38 +0000[diff] [blame]4450</p>
4451
4452<h5>Semantics:</h5>
4453
4454<p>
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]4455The <tt>llvm.bswap.16.i16</tt> intrinsic returns an i16 value that has the high
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]4456and low byte of the input i16 swapped. Similarly, the <tt>llvm.bswap.i32</tt>
4457intrinsic returns an i32 value that has the four bytes of the input i32
4458swapped, so that if the input bytes are numbered 0, 1, 2, 3 then the returned
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]4459i32 will have its bytes in 3, 2, 1, 0 order. The <tt>llvm.bswap.i48.i48</tt>,
4460<tt>llvm.bswap.i64.i64</tt> and other intrinsics extend this concept to
4461additional even-byte lengths (6 bytes, 8 bytes and more, respectively).
Nate Begeman0f223bb2006-01-13 23:26:38 +0000[diff] [blame]4462</p>
4463
4464</div>
4465
4466<!-- _______________________________________________________________________ -->
4467<div class="doc_subsubsection">
Reid Spencerb4f9a6f2006-01-16 21:12:35 +0000[diff] [blame]4468 <a name="int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic</a>
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4469</div>
4470
4471<div class="doc_text">
4472
4473<h5>Syntax:</h5>
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]4474<p>This is an overloaded intrinsic. You can use llvm.ctpop on any integer bit
4475width. Not all targets support all bit widths however.
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4476<pre>
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]4477 declare i32 @llvm.ctpop.i8 (i8 &lt;src&gt;)
4478 declare i32 @llvm.ctpop.i16(i16 &lt;src&gt;)
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4479 declare i32 @llvm.ctpop.i32(i32 &lt;src&gt;)
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]4480 declare i32 @llvm.ctpop.i64(i64 &lt;src&gt;)
4481 declare i32 @llvm.ctpop.i256(i256 &lt;src&gt;)
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4482</pre>
4483
4484<h5>Overview:</h5>
4485
4486<p>
Chris Lattner069b5bd2006-01-16 22:38:59 +0000[diff] [blame]4487The '<tt>llvm.ctpop</tt>' family of intrinsics counts the number of bits set in a
4488value.
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4489</p>
4490
4491<h5>Arguments:</h5>
4492
4493<p>
Chris Lattner573f64e2005-05-07 01:46:40 +0000[diff] [blame]4494The only argument is the value to be counted. The argument may be of any
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]4495integer type. The return type must match the argument type.
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4496</p>
4497
4498<h5>Semantics:</h5>
4499
4500<p>
4501The '<tt>llvm.ctpop</tt>' intrinsic counts the 1's in a variable.
4502</p>
4503</div>
4504
4505<!-- _______________________________________________________________________ -->
4506<div class="doc_subsubsection">
Chris Lattnerb748c672006-01-16 22:34:14 +0000[diff] [blame]4507 <a name="int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic</a>
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4508</div>
4509
4510<div class="doc_text">
4511
4512<h5>Syntax:</h5>
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]4513<p>This is an overloaded intrinsic. You can use <tt>llvm.ctlz</tt> on any
4514integer bit width. Not all targets support all bit widths however.
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4515<pre>
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]4516 declare i32 @llvm.ctlz.i8 (i8 &lt;src&gt;)
4517 declare i32 @llvm.ctlz.i16(i16 &lt;src&gt;)
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4518 declare i32 @llvm.ctlz.i32(i32 &lt;src&gt;)
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]4519 declare i32 @llvm.ctlz.i64(i64 &lt;src&gt;)
4520 declare i32 @llvm.ctlz.i256(i256 &lt;src&gt;)
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4521</pre>
4522
4523<h5>Overview:</h5>
4524
4525<p>
Reid Spencerb4f9a6f2006-01-16 21:12:35 +0000[diff] [blame]4526The '<tt>llvm.ctlz</tt>' family of intrinsic functions counts the number of
4527leading zeros in a variable.
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4528</p>
4529
4530<h5>Arguments:</h5>
4531
4532<p>
Chris Lattner573f64e2005-05-07 01:46:40 +0000[diff] [blame]4533The only argument is the value to be counted. The argument may be of any
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]4534integer type. The return type must match the argument type.
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4535</p>
4536
4537<h5>Semantics:</h5>
4538
4539<p>
Chris Lattnerefa20fa2005-05-15 19:39:26 +0000[diff] [blame]4540The '<tt>llvm.ctlz</tt>' intrinsic counts the leading (most significant) zeros
4541in a variable. If the src == 0 then the result is the size in bits of the type
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]4542of src. For example, <tt>llvm.ctlz(i32 2) = 30</tt>.
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4543</p>
4544</div>
Chris Lattner3b4f4372004-06-11 02:28:03 +0000[diff] [blame]4545
4546
Chris Lattnerefa20fa2005-05-15 19:39:26 +0000[diff] [blame]4547
4548<!-- _______________________________________________________________________ -->
4549<div class="doc_subsubsection">
Chris Lattnerb748c672006-01-16 22:34:14 +0000[diff] [blame]4550 <a name="int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic</a>
Chris Lattnerefa20fa2005-05-15 19:39:26 +0000[diff] [blame]4551</div>
4552
4553<div class="doc_text">
4554
4555<h5>Syntax:</h5>
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]4556<p>This is an overloaded intrinsic. You can use <tt>llvm.cttz</tt> on any
4557integer bit width. Not all targets support all bit widths however.
Chris Lattnerefa20fa2005-05-15 19:39:26 +0000[diff] [blame]4558<pre>
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]4559 declare i32 @llvm.cttz.i8 (i8 &lt;src&gt;)
4560 declare i32 @llvm.cttz.i16(i16 &lt;src&gt;)
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4561 declare i32 @llvm.cttz.i32(i32 &lt;src&gt;)
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]4562 declare i32 @llvm.cttz.i64(i64 &lt;src&gt;)
4563 declare i32 @llvm.cttz.i256(i256 &lt;src&gt;)
Chris Lattnerefa20fa2005-05-15 19:39:26 +0000[diff] [blame]4564</pre>
4565
4566<h5>Overview:</h5>
4567
4568<p>
Reid Spencerb4f9a6f2006-01-16 21:12:35 +0000[diff] [blame]4569The '<tt>llvm.cttz</tt>' family of intrinsic functions counts the number of
4570trailing zeros.
Chris Lattnerefa20fa2005-05-15 19:39:26 +0000[diff] [blame]4571</p>
4572
4573<h5>Arguments:</h5>
4574
4575<p>
4576The only argument is the value to be counted. The argument may be of any
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]4577integer type. The return type must match the argument type.
Chris Lattnerefa20fa2005-05-15 19:39:26 +0000[diff] [blame]4578</p>
4579
4580<h5>Semantics:</h5>
4581
4582<p>
4583The '<tt>llvm.cttz</tt>' intrinsic counts the trailing (least significant) zeros
4584in a variable. If the src == 0 then the result is the size in bits of the type
4585of src. For example, <tt>llvm.cttz(2) = 1</tt>.
4586</p>
4587</div>
4588
Reid Spencer8a5799f2007-04-01 08:27:01 +0000[diff] [blame]4589<!-- _______________________________________________________________________ -->
4590<div class="doc_subsubsection">
Reid Spencerea2945e2007-04-10 02:51:31 +0000[diff] [blame]4591 <a name="int_part_select">'<tt>llvm.part.select.*</tt>' Intrinsic</a>
Reid Spencer8bc7d952007-04-01 19:00:37 +0000[diff] [blame]4592</div>
4593
4594<div class="doc_text">
4595
4596<h5>Syntax:</h5>
Reid Spencerea2945e2007-04-10 02:51:31 +0000[diff] [blame]4597<p>This is an overloaded intrinsic. You can use <tt>llvm.part.select</tt>
Reid Spencer8bc7d952007-04-01 19:00:37 +0000[diff] [blame]4598on any integer bit width.
4599<pre>
Reid Spencerea2945e2007-04-10 02:51:31 +0000[diff] [blame]4600 declare i17 @llvm.part.select.i17.i17 (i17 %val, i32 %loBit, i32 %hiBit)
4601 declare i29 @llvm.part.select.i29.i29 (i29 %val, i32 %loBit, i32 %hiBit)
Reid Spencer8bc7d952007-04-01 19:00:37 +0000[diff] [blame]4602</pre>
4603
4604<h5>Overview:</h5>
Reid Spencerea2945e2007-04-10 02:51:31 +0000[diff] [blame]4605<p>The '<tt>llvm.part.select</tt>' family of intrinsic functions selects a
Reid Spencer8bc7d952007-04-01 19:00:37 +0000[diff] [blame]4606range of bits from an integer value and returns them in the same bit width as
4607the original value.</p>
4608
4609<h5>Arguments:</h5>
4610<p>The first argument, <tt>%val</tt> and the result may be integer types of
4611any bit width but they must have the same bit width. The second and third
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4612arguments must be <tt>i32</tt> type since they specify only a bit index.</p>
Reid Spencer8bc7d952007-04-01 19:00:37 +0000[diff] [blame]4613
4614<h5>Semantics:</h5>
Reid Spencerea2945e2007-04-10 02:51:31 +0000[diff] [blame]4615<p>The operation of the '<tt>llvm.part.select</tt>' intrinsic has two modes
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4616of operation: forwards and reverse. If <tt>%loBit</tt> is greater than
4617<tt>%hiBits</tt> then the intrinsic operates in reverse mode. Otherwise it
4618operates in forward mode.</p>
4619<p>In forward mode, this intrinsic is the equivalent of shifting <tt>%val</tt>
4620right by <tt>%loBit</tt> bits and then ANDing it with a mask with
Reid Spencer8bc7d952007-04-01 19:00:37 +0000[diff] [blame]4621only the <tt>%hiBit - %loBit</tt> bits set, as follows:</p>
4622<ol>
4623 <li>The <tt>%val</tt> is shifted right (LSHR) by the number of bits specified
4624 by <tt>%loBits</tt>. This normalizes the value to the low order bits.</li>
4625 <li>The <tt>%loBits</tt> value is subtracted from the <tt>%hiBits</tt> value
4626 to determine the number of bits to retain.</li>
4627 <li>A mask of the retained bits is created by shifting a -1 value.</li>
4628 <li>The mask is ANDed with <tt>%val</tt> to produce the result.
4629</ol>
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4630<p>In reverse mode, a similar computation is made except that:</p>
4631<ol>
4632 <li>The bits selected wrap around to include both the highest and lowest bits.
Reid Spencerea2945e2007-04-10 02:51:31 +0000[diff] [blame]4633 For example, part.select(i16 X, 4, 7) selects bits from X with a mask of
4634 0x00F0 (forwards case) while part.select(i16 X, 8, 3) selects bits from X
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4635 with a mask of 0xFF0F.</li>
4636 <li>The bits returned in the reverse case are reversed. So, if X has the value
Reid Spencerea2945e2007-04-10 02:51:31 +0000[diff] [blame]4637 0x6ACF and we apply part.select(i16 X, 8, 3) to it, we get back the value
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4638 0x0A6F.</li>
4639</ol>
Reid Spencer8bc7d952007-04-01 19:00:37 +0000[diff] [blame]4640</div>
4641
Reid Spencer5bf54c82007-04-11 23:23:49 +0000[diff] [blame]4642<div class="doc_subsubsection">
4643 <a name="int_part_set">'<tt>llvm.part.set.*</tt>' Intrinsic</a>
4644</div>
4645
4646<div class="doc_text">
4647
4648<h5>Syntax:</h5>
4649<p>This is an overloaded intrinsic. You can use <tt>llvm.part.set</tt>
4650on any integer bit width.
4651<pre>
4652 declare i17 @llvm.part.set.i17.i17.i9 (i17 %val, i9 %repl, i32 %lo, i32 %hi)
4653 declare i29 @llvm.part.set.i29.i29.i9 (i29 %val, i9 %repl, i32 %lo, i32 %hi)
4654</pre>
4655
4656<h5>Overview:</h5>
4657<p>The '<tt>llvm.part.set</tt>' family of intrinsic functions replaces a range
4658of bits in an integer value with another integer value. It returns the integer
4659with the replaced bits.</p>
4660
4661<h5>Arguments:</h5>
4662<p>The first argument, <tt>%val</tt> and the result may be integer types of
4663any bit width but they must have the same bit width. <tt>%val</tt> is the value
4664whose bits will be replaced. The second argument, <tt>%repl</tt> may be an
4665integer of any bit width. The third and fourth arguments must be <tt>i32</tt>
4666type since they specify only a bit index.</p>
4667
4668<h5>Semantics:</h5>
4669<p>The operation of the '<tt>llvm.part.set</tt>' intrinsic has two modes
4670of operation: forwards and reverse. If <tt>%lo</tt> is greater than
4671<tt>%hi</tt> then the intrinsic operates in reverse mode. Otherwise it
4672operates in forward mode.</p>
4673<p>For both modes, the <tt>%repl</tt> value is prepared for use by either
4674truncating it down to the size of the replacement area or zero extending it
4675up to that size.</p>
4676<p>In forward mode, the bits between <tt>%lo</tt> and <tt>%hi</tt> (inclusive)
4677are replaced with corresponding bits from <tt>%repl</tt>. That is the 0th bit
4678in <tt>%repl</tt> replaces the <tt>%lo</tt>th bit in <tt>%val</tt> and etc. up
4679to the <tt>%hi</tt>th bit.
4680<p>In reverse mode, a similar computation is made except that the bits replaced
4681wrap around to include both the highest and lowest bits. For example, if a
468216 bit value is being replaced then <tt>%lo=8</tt> and <tt>%hi=4</tt> would
Reid Spencer14194dc2007-04-11 23:46:06 +0000[diff] [blame]4683cause these bits to be set: <tt>0xFF1F</tt>.</p>
Reid Spencer5bf54c82007-04-11 23:23:49 +0000[diff] [blame]4684<h5>Examples:</h5>
4685<pre>
Reid Spencerc70afc32007-04-12 01:03:03 +0000[diff] [blame]4686 llvm.part.set(0xFFFF, 0, 4, 7) -&gt; 0xFF0F
4687 llvm.part.set(0xFFFF, 0, 7, 4) -&gt; 0x0060
4688 llvm.part.set(0xFFFF, 0, 8, 3) -&gt; 0x00F0
4689 llvm.part.set(0xFFFF, 0, 3, 8) -&gt; 0xFE07
Reid Spencer7972c472007-04-11 23:49:50 +0000[diff] [blame]4690</pre>
Reid Spencer5bf54c82007-04-11 23:23:49 +0000[diff] [blame]4691</div>
4692
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]4693<!-- ======================================================================= -->
4694<div class="doc_subsection">
4695 <a name="int_debugger">Debugger Intrinsics</a>
4696</div>
4697
4698<div class="doc_text">
4699<p>
4700The LLVM debugger intrinsics (which all start with <tt>llvm.dbg.</tt> prefix),
4701are described in the <a
4702href="SourceLevelDebugging.html#format_common_intrinsics">LLVM Source Level
4703Debugging</a> document.
4704</p>
4705</div>
4706
4707
Jim Laskey2211f492007-03-14 19:31:19 +0000[diff] [blame]4708<!-- ======================================================================= -->
4709<div class="doc_subsection">
4710 <a name="int_eh">Exception Handling Intrinsics</a>
4711</div>
4712
4713<div class="doc_text">
4714<p> The LLVM exception handling intrinsics (which all start with
4715<tt>llvm.eh.</tt> prefix), are described in the <a
4716href="ExceptionHandling.html#format_common_intrinsics">LLVM Exception
4717Handling</a> document. </p>
4718</div>
4719
4720
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]4721<!-- *********************************************************************** -->
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]4722<hr>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]4723<address>
4724 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
4725 src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
4726 <a href="http://validator.w3.org/check/referer"><img
4727 src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!" /></a>
4728
4729 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
Reid Spencerca058542006-03-14 05:39:39 +0000[diff] [blame]4730 <a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]4731 Last modified: $Date$
4732</address>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]4733</body>
4734</html>