Gitiles
Code Review Sign In
gerrit-public.fairphone.software / toolchain / llvm-project / db3b93bbb2e22cc2f4437209dd4a96de18e12029 / . / llvm / docs / LangRef.html
blob: e51618de180d177015b2ba0fa7710881b6bad7b2 [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>
149 <li><a href="#i_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a></li>
150 <li><a href="#i_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a></li>
151 <li><a href="#i_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a></li>
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>
156 <li><a href="#i_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a></li>
157 <li><a href="#i_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a></li>
158 <li><a href="#i_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a></li>
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>
163 <li><a href="#i_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a></li>
164 <li><a href="#i_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a></li>
Chris Lattner2f0f0012006-01-13 02:03:13 +0000[diff] [blame]165 <li><a href="#i_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a></li>
166 <li><a href="#i_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a></li>
Chris Lattnerc8a2c222005-02-28 19:24:19 +0000[diff] [blame]167 <li><a href="#i_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a></li>
Andrew Lenharthb4427912005-03-28 20:05:49 +0000[diff] [blame]168 <li><a href="#i_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a></li>
Andrew Lenharth01aa5632005-11-11 16:47:30 +0000[diff] [blame]169 <li><a href="#i_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>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]174 <li><a href="#i_memcpy">'<tt>llvm.memcpy.*</tt>' Intrinsic</a></li>
175 <li><a href="#i_memmove">'<tt>llvm.memmove.*</tt>' Intrinsic</a></li>
176 <li><a href="#i_memset">'<tt>llvm.memset.*</tt>' Intrinsic</a></li>
Chris Lattner069b5bd2006-01-16 22:38:59 +0000[diff] [blame]177 <li><a href="#i_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a></li>
Chris Lattner33b73f92006-09-08 06:34:02 +0000[diff] [blame]178 <li><a href="#i_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>
Nate Begeman0f223bb2006-01-13 23:26:38 +0000[diff] [blame]183 <li><a href="#i_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>
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]187 </ol>
188 </li>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]189 <li><a href="#int_debugger">Debugger intrinsics</a></li>
Jim Laskey2211f492007-03-14 19:31:19 +0000[diff] [blame]190 <li><a href="#int_eh">Exception Handling intrinsics</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]191 </ol>
192 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]193</ol>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]194
195<div class="doc_author">
196 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>
197 and <a href="mailto:vadve@cs.uiuc.edu">Vikram Adve</a></p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]198</div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]199
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]200<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]201<div class="doc_section"> <a name="abstract">Abstract </a></div>
202<!-- *********************************************************************** -->
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]203
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]204<div class="doc_text">
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]205<p>This document is a reference manual for the LLVM assembly language.
206LLVM is an SSA based representation that provides type safety,
207low-level operations, flexibility, and the capability of representing
208'all' high-level languages cleanly. It is the common code
209representation used throughout all phases of the LLVM compilation
210strategy.</p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]211</div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]212
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]213<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]214<div class="doc_section"> <a name="introduction">Introduction</a> </div>
215<!-- *********************************************************************** -->
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]216
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]217<div class="doc_text">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]218
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]219<p>The LLVM code representation is designed to be used in three
220different forms: as an in-memory compiler IR, as an on-disk bytecode
221representation (suitable for fast loading by a Just-In-Time compiler),
222and as a human readable assembly language representation. This allows
223LLVM to provide a powerful intermediate representation for efficient
224compiler transformations and analysis, while providing a natural means
225to debug and visualize the transformations. The three different forms
226of LLVM are all equivalent. This document describes the human readable
227representation and notation.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]228
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]229<p>The LLVM representation aims to be light-weight and low-level
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]230while being expressive, typed, and extensible at the same time. It
231aims to be a "universal IR" of sorts, by being at a low enough level
232that high-level ideas may be cleanly mapped to it (similar to how
233microprocessors are "universal IR's", allowing many source languages to
234be mapped to them). By providing type information, LLVM can be used as
235the target of optimizations: for example, through pointer analysis, it
236can be proven that a C automatic variable is never accessed outside of
237the current function... allowing it to be promoted to a simple SSA
238value instead of a memory location.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]239
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]240</div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]241
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]242<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]243<div class="doc_subsubsection"> <a name="wellformed">Well-Formedness</a> </div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]244
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]245<div class="doc_text">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]246
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]247<p>It is important to note that this document describes 'well formed'
248LLVM assembly language. There is a difference between what the parser
249accepts and what is considered 'well formed'. For example, the
250following instruction is syntactically okay, but not well formed:</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]251
252<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]253 %x = <a href="#i_add">add</a> i32 1, %x
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]254</pre>
255
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]256<p>...because the definition of <tt>%x</tt> does not dominate all of
257its uses. The LLVM infrastructure provides a verification pass that may
258be used to verify that an LLVM module is well formed. This pass is
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]259automatically run by the parser after parsing input assembly and by
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]260the optimizer before it outputs bytecode. The violations pointed out
261by the verifier pass indicate bugs in transformation passes or input to
262the parser.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]263
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]264<!-- Describe the typesetting conventions here. --> </div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]265
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]266<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]267<div class="doc_section"> <a name="identifiers">Identifiers</a> </div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]268<!-- *********************************************************************** -->
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]269
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]270<div class="doc_text">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]271
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]272<p>LLVM uses three different forms of identifiers, for different
273purposes:</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]274
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]275<ol>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]276 <li>Named values are represented as a string of characters with a '%' prefix.
277 For example, %foo, %DivisionByZero, %a.really.long.identifier. The actual
278 regular expression used is '<tt>%[a-zA-Z$._][a-zA-Z$._0-9]*</tt>'.
279 Identifiers which require other characters in their names can be surrounded
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]280 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]281 in a name.</li>
282
283 <li>Unnamed values are represented as an unsigned numeric value with a '%'
284 prefix. For example, %12, %2, %44.</li>
285
Reid Spencer8f08d802004-12-09 18:02:53 +0000[diff] [blame]286 <li>Constants, which are described in a <a href="#constants">section about
287 constants</a>, below.</li>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]288</ol>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]289
290<p>LLVM requires that values start with a '%' sign for two reasons: Compilers
291don't need to worry about name clashes with reserved words, and the set of
292reserved words may be expanded in the future without penalty. Additionally,
293unnamed identifiers allow a compiler to quickly come up with a temporary
294variable without having to avoid symbol table conflicts.</p>
295
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]296<p>Reserved words in LLVM are very similar to reserved words in other
Reid Spencer5b950642006-11-11 23:08:07 +0000[diff] [blame]297languages. There are keywords for different opcodes
298('<tt><a href="#i_add">add</a></tt>',
299 '<tt><a href="#i_bitcast">bitcast</a></tt>',
300 '<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]301href="#t_void">void</a></tt>', '<tt><a href="#t_primitive">i32</a></tt>', etc...),
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]302and others. These reserved words cannot conflict with variable names, because
303none of them start with a '%' character.</p>
304
305<p>Here is an example of LLVM code to multiply the integer variable
306'<tt>%X</tt>' by 8:</p>
307
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]308<p>The easy way:</p>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]309
310<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]311 %result = <a href="#i_mul">mul</a> i32 %X, 8
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]312</pre>
313
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]314<p>After strength reduction:</p>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]315
316<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]317 %result = <a href="#i_shl">shl</a> i32 %X, i8 3
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]318</pre>
319
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]320<p>And the hard way:</p>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]321
322<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]323 <a href="#i_add">add</a> i32 %X, %X <i>; yields {i32}:%0</i>
324 <a href="#i_add">add</a> i32 %0, %0 <i>; yields {i32}:%1</i>
325 %result = <a href="#i_add">add</a> i32 %1, %1
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]326</pre>
327
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]328<p>This last way of multiplying <tt>%X</tt> by 8 illustrates several
329important lexical features of LLVM:</p>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]330
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]331<ol>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]332
333 <li>Comments are delimited with a '<tt>;</tt>' and go until the end of
334 line.</li>
335
336 <li>Unnamed temporaries are created when the result of a computation is not
337 assigned to a named value.</li>
338
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]339 <li>Unnamed temporaries are numbered sequentially</li>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]340
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]341</ol>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]342
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]343<p>...and it also shows a convention that we follow in this document. When
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]344demonstrating instructions, we will follow an instruction with a comment that
345defines the type and name of value produced. Comments are shown in italic
346text.</p>
347
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]348</div>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]349
350<!-- *********************************************************************** -->
351<div class="doc_section"> <a name="highlevel">High Level Structure</a> </div>
352<!-- *********************************************************************** -->
353
354<!-- ======================================================================= -->
355<div class="doc_subsection"> <a name="modulestructure">Module Structure</a>
356</div>
357
358<div class="doc_text">
359
360<p>LLVM programs are composed of "Module"s, each of which is a
361translation unit of the input programs. Each module consists of
362functions, global variables, and symbol table entries. Modules may be
363combined together with the LLVM linker, which merges function (and
364global variable) definitions, resolves forward declarations, and merges
365symbol table entries. Here is an example of the "hello world" module:</p>
366
367<pre><i>; Declare the string constant as a global constant...</i>
368<a href="#identifiers">%.LC0</a> = <a href="#linkage_internal">internal</a> <a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]369 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]370
371<i>; External declaration of the puts function</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]372<a href="#functionstructure">declare</a> i32 %puts(i8 *) <i>; i32(i8 *)* </i>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]373
Chris Lattnerd2d29a02006-06-13 03:05:47 +0000[diff] [blame]374<i>; Global variable / Function body section separator</i>
375implementation
376
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]377<i>; Definition of main function</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]378define i32 %main() { <i>; i32()* </i>
379 <i>; Convert [13x i8 ]* to i8 *...</i>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]380 %cast210 = <a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]381 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]382
383 <i>; Call puts function to write out the string to stdout...</i>
384 <a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]385 href="#i_call">call</a> i32 %puts(i8 * %cast210) <i>; i32</i>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]386 <a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]387 href="#i_ret">ret</a> i32 0<br>}<br></pre>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]388
389<p>This example is made up of a <a href="#globalvars">global variable</a>
390named "<tt>.LC0</tt>", an external declaration of the "<tt>puts</tt>"
391function, and a <a href="#functionstructure">function definition</a>
392for "<tt>main</tt>".</p>
393
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]394<p>In general, a module is made up of a list of global values,
395where both functions and global variables are global values. Global values are
396represented by a pointer to a memory location (in this case, a pointer to an
397array of char, and a pointer to a function), and have one of the following <a
398href="#linkage">linkage types</a>.</p>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]399
Chris Lattnerd2d29a02006-06-13 03:05:47 +0000[diff] [blame]400<p>Due to a limitation in the current LLVM assembly parser (it is limited by
401one-token lookahead), modules are split into two pieces by the "implementation"
402keyword. Global variable prototypes and definitions must occur before the
403keyword, and function definitions must occur after it. Function prototypes may
404occur either before or after it. In the future, the implementation keyword may
405become a noop, if the parser gets smarter.</p>
406
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]407</div>
408
409<!-- ======================================================================= -->
410<div class="doc_subsection">
411 <a name="linkage">Linkage Types</a>
412</div>
413
414<div class="doc_text">
415
416<p>
417All Global Variables and Functions have one of the following types of linkage:
418</p>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]419
420<dl>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]421
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]422 <dt><tt><b><a name="linkage_internal">internal</a></b></tt> </dt>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]423
424 <dd>Global values with internal linkage are only directly accessible by
425 objects in the current module. In particular, linking code into a module with
426 an internal global value may cause the internal to be renamed as necessary to
427 avoid collisions. Because the symbol is internal to the module, all
428 references can be updated. This corresponds to the notion of the
Chris Lattnere20b4702007-01-14 06:51:48 +0000[diff] [blame]429 '<tt>static</tt>' keyword in C.
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]430 </dd>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]431
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]432 <dt><tt><b><a name="linkage_linkonce">linkonce</a></b></tt>: </dt>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]433
Chris Lattnere20b4702007-01-14 06:51:48 +0000[diff] [blame]434 <dd>Globals with "<tt>linkonce</tt>" linkage are merged with other globals of
435 the same name when linkage occurs. This is typically used to implement
436 inline functions, templates, or other code which must be generated in each
437 translation unit that uses it. Unreferenced <tt>linkonce</tt> globals are
438 allowed to be discarded.
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]439 </dd>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]440
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]441 <dt><tt><b><a name="linkage_weak">weak</a></b></tt>: </dt>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]442
443 <dd>"<tt>weak</tt>" linkage is exactly the same as <tt>linkonce</tt> linkage,
444 except that unreferenced <tt>weak</tt> globals may not be discarded. This is
Chris Lattnere20b4702007-01-14 06:51:48 +0000[diff] [blame]445 used for globals that may be emitted in multiple translation units, but that
446 are not guaranteed to be emitted into every translation unit that uses them.
447 One example of this are common globals in C, such as "<tt>int X;</tt>" at
448 global scope.
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]449 </dd>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]450
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]451 <dt><tt><b><a name="linkage_appending">appending</a></b></tt>: </dt>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]452
453 <dd>"<tt>appending</tt>" linkage may only be applied to global variables of
454 pointer to array type. When two global variables with appending linkage are
455 linked together, the two global arrays are appended together. This is the
456 LLVM, typesafe, equivalent of having the system linker append together
457 "sections" with identical names when .o files are linked.
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]458 </dd>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]459
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]460 <dt><tt><b><a name="linkage_externweak">extern_weak</a></b></tt>: </dt>
461 <dd>The semantics of this linkage follow the ELF model: the symbol is weak
462 until linked, if not linked, the symbol becomes null instead of being an
463 undefined reference.
464 </dd>
465</dl>
466
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]467 <dt><tt><b><a name="linkage_external">externally visible</a></b></tt>:</dt>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]468
469 <dd>If none of the above identifiers are used, the global is externally
470 visible, meaning that it participates in linkage and can be used to resolve
471 external symbol references.
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]472 </dd>
Anton Korobeynikovd61d39e2006-09-14 18:23:27 +0000[diff] [blame]473
Anton Korobeynikovd61d39e2006-09-14 18:23:27 +0000[diff] [blame]474 <p>
475 The next two types of linkage are targeted for Microsoft Windows platform
476 only. They are designed to support importing (exporting) symbols from (to)
477 DLLs.
478 </p>
479
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]480 <dl>
Anton Korobeynikovd61d39e2006-09-14 18:23:27 +0000[diff] [blame]481 <dt><tt><b><a name="linkage_dllimport">dllimport</a></b></tt>: </dt>
482
483 <dd>"<tt>dllimport</tt>" linkage causes the compiler to reference a function
484 or variable via a global pointer to a pointer that is set up by the DLL
485 exporting the symbol. On Microsoft Windows targets, the pointer name is
486 formed by combining <code>_imp__</code> and the function or variable name.
487 </dd>
488
489 <dt><tt><b><a name="linkage_dllexport">dllexport</a></b></tt>: </dt>
490
491 <dd>"<tt>dllexport</tt>" linkage causes the compiler to provide a global
492 pointer to a pointer in a DLL, so that it can be referenced with the
493 <tt>dllimport</tt> attribute. On Microsoft Windows targets, the pointer
494 name is formed by combining <code>_imp__</code> and the function or variable
495 name.
496 </dd>
497
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]498</dl>
499
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]500<p><a name="linkage_external"></a>For example, since the "<tt>.LC0</tt>"
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]501variable is defined to be internal, if another module defined a "<tt>.LC0</tt>"
502variable and was linked with this one, one of the two would be renamed,
503preventing a collision. Since "<tt>main</tt>" and "<tt>puts</tt>" are
504external (i.e., lacking any linkage declarations), they are accessible
Reid Spencer92c671e2007-01-05 00:59:10 +0000[diff] [blame]505outside of the current module.</p>
506<p>It is illegal for a function <i>declaration</i>
507to have any linkage type other than "externally visible", <tt>dllimport</tt>,
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]508or <tt>extern_weak</tt>.</p>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]509
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]510</div>
511
512<!-- ======================================================================= -->
513<div class="doc_subsection">
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]514 <a name="callingconv">Calling Conventions</a>
515</div>
516
517<div class="doc_text">
518
519<p>LLVM <a href="#functionstructure">functions</a>, <a href="#i_call">calls</a>
520and <a href="#i_invoke">invokes</a> can all have an optional calling convention
521specified for the call. The calling convention of any pair of dynamic
522caller/callee must match, or the behavior of the program is undefined. The
523following calling conventions are supported by LLVM, and more may be added in
524the future:</p>
525
526<dl>
527 <dt><b>"<tt>ccc</tt>" - The C calling convention</b>:</dt>
528
529 <dd>This calling convention (the default if no other calling convention is
530 specified) matches the target C calling conventions. This calling convention
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]531 supports varargs function calls and tolerates some mismatch in the declared
Reid Spencer72ba4992006-12-31 21:30:18 +0000[diff] [blame]532 prototype and implemented declaration of the function (as does normal C).
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]533 </dd>
534
535 <dt><b>"<tt>fastcc</tt>" - The fast calling convention</b>:</dt>
536
537 <dd>This calling convention attempts to make calls as fast as possible
538 (e.g. by passing things in registers). This calling convention allows the
539 target to use whatever tricks it wants to produce fast code for the target,
Chris Lattnerc792eb32005-05-06 23:08:23 +0000[diff] [blame]540 without having to conform to an externally specified ABI. Implementations of
541 this convention should allow arbitrary tail call optimization to be supported.
542 This calling convention does not support varargs and requires the prototype of
543 all callees to exactly match the prototype of the function definition.
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]544 </dd>
545
546 <dt><b>"<tt>coldcc</tt>" - The cold calling convention</b>:</dt>
547
548 <dd>This calling convention attempts to make code in the caller as efficient
549 as possible under the assumption that the call is not commonly executed. As
550 such, these calls often preserve all registers so that the call does not break
551 any live ranges in the caller side. This calling convention does not support
552 varargs and requires the prototype of all callees to exactly match the
553 prototype of the function definition.
554 </dd>
555
Chris Lattner573f64e2005-05-07 01:46:40 +0000[diff] [blame]556 <dt><b>"<tt>cc &lt;<em>n</em>&gt;</tt>" - Numbered convention</b>:</dt>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]557
558 <dd>Any calling convention may be specified by number, allowing
559 target-specific calling conventions to be used. Target specific calling
560 conventions start at 64.
561 </dd>
Chris Lattner573f64e2005-05-07 01:46:40 +0000[diff] [blame]562</dl>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]563
564<p>More calling conventions can be added/defined on an as-needed basis, to
565support pascal conventions or any other well-known target-independent
566convention.</p>
567
568</div>
569
570<!-- ======================================================================= -->
571<div class="doc_subsection">
Anton Korobeynikovc7f9f3d2007-01-23 12:35:46 +0000[diff] [blame]572 <a name="visibility">Visibility Styles</a>
573</div>
574
575<div class="doc_text">
576
577<p>
578All Global Variables and Functions have one of the following visibility styles:
579</p>
580
581<dl>
582 <dt><b>"<tt>default</tt>" - Default style</b>:</dt>
583
584 <dd>On ELF, default visibility means that the declaration is visible to other
585 modules and, in shared libraries, means that the declared entity may be
586 overridden. On Darwin, default visibility means that the declaration is
587 visible to other modules. Default visibility corresponds to "external
588 linkage" in the language.
589 </dd>
590
591 <dt><b>"<tt>hidden</tt>" - Hidden style</b>:</dt>
592
593 <dd>Two declarations of an object with hidden visibility refer to the same
594 object if they are in the same shared object. Usually, hidden visibility
595 indicates that the symbol will not be placed into the dynamic symbol table,
596 so no other module (executable or shared library) can reference it
597 directly.
598 </dd>
599
600</dl>
601
602</div>
603
604<!-- ======================================================================= -->
605<div class="doc_subsection">
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]606 <a name="globalvars">Global Variables</a>
607</div>
608
609<div class="doc_text">
610
Chris Lattner5d5aede2005-02-12 19:30:21 +0000[diff] [blame]611<p>Global variables define regions of memory allocated at compilation time
Chris Lattner662c8722005-11-12 00:45:07 +0000[diff] [blame]612instead of run-time. Global variables may optionally be initialized, may have
613an explicit section to be placed in, and may
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]614have an optional explicit alignment specified. A
John Criswell4c0cf7f2005-10-24 16:17:18 +0000[diff] [blame]615variable may be defined as a global "constant," which indicates that the
Chris Lattner5d5aede2005-02-12 19:30:21 +0000[diff] [blame]616contents of the variable will <b>never</b> be modified (enabling better
617optimization, allowing the global data to be placed in the read-only section of
618an executable, etc). Note that variables that need runtime initialization
John Criswell4c0cf7f2005-10-24 16:17:18 +0000[diff] [blame]619cannot be marked "constant" as there is a store to the variable.</p>
Chris Lattner5d5aede2005-02-12 19:30:21 +0000[diff] [blame]620
621<p>
622LLVM explicitly allows <em>declarations</em> of global variables to be marked
623constant, even if the final definition of the global is not. This capability
624can be used to enable slightly better optimization of the program, but requires
625the language definition to guarantee that optimizations based on the
626'constantness' are valid for the translation units that do not include the
627definition.
628</p>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]629
630<p>As SSA values, global variables define pointer values that are in
631scope (i.e. they dominate) all basic blocks in the program. Global
632variables always define a pointer to their "content" type because they
633describe a region of memory, and all memory objects in LLVM are
634accessed through pointers.</p>
635
Chris Lattner662c8722005-11-12 00:45:07 +0000[diff] [blame]636<p>LLVM allows an explicit section to be specified for globals. If the target
637supports it, it will emit globals to the section specified.</p>
638
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]639<p>An explicit alignment may be specified for a global. If not present, or if
640the alignment is set to zero, the alignment of the global is set by the target
641to whatever it feels convenient. If an explicit alignment is specified, the
642global is forced to have at least that much alignment. All alignments must be
643a power of 2.</p>
644
Chris Lattner5760c502007-01-14 00:27:09 +0000[diff] [blame]645<p>For example, the following defines a global with an initializer, section,
646 and alignment:</p>
647
648<pre>
649 %G = constant float 1.0, section "foo", align 4
650</pre>
651
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]652</div>
653
654
655<!-- ======================================================================= -->
656<div class="doc_subsection">
657 <a name="functionstructure">Functions</a>
658</div>
659
660<div class="doc_text">
661
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]662<p>LLVM function definitions consist of the "<tt>define</tt>" keyord,
663an optional <a href="#linkage">linkage type</a>, an optional
Anton Korobeynikovc7f9f3d2007-01-23 12:35:46 +0000[diff] [blame]664<a href="#visibility">visibility style</a>, an optional
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]665<a href="#callingconv">calling convention</a>, a return type, an optional
666<a href="#paramattrs">parameter attribute</a> for the return type, a function
667name, a (possibly empty) argument list (each with optional
Anton Korobeynikovc7f9f3d2007-01-23 12:35:46 +0000[diff] [blame]668<a href="#paramattrs">parameter attributes</a>), an optional section, an
669optional alignment, an opening curly brace, a list of basic blocks, and a
670closing curly brace.
671
672LLVM function declarations consist of the "<tt>declare</tt>" keyword, an
673optional <a href="#linkage">linkage type</a>, an optional
674<a href="#visibility">visibility style</a>, an optional
675<a href="#callingconv">calling convention</a>, a return type, an optional
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]676<a href="#paramattrs">parameter attribute</a> for the return type, a function
677name, a possibly empty list of arguments, and an optional alignment.</p>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]678
679<p>A function definition contains a list of basic blocks, forming the CFG for
680the function. Each basic block may optionally start with a label (giving the
681basic block a symbol table entry), contains a list of instructions, and ends
682with a <a href="#terminators">terminator</a> instruction (such as a branch or
683function return).</p>
684
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]685<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]686executed on entrance to the function, and it is not allowed to have predecessor
687basic blocks (i.e. there can not be any branches to the entry block of a
688function). Because the block can have no predecessors, it also cannot have any
689<a href="#i_phi">PHI nodes</a>.</p>
690
691<p>LLVM functions are identified by their name and type signature. Hence, two
692functions with the same name but different parameter lists or return values are
Chris Lattner455fc8c2005-03-07 22:13:59 +0000[diff] [blame]693considered different functions, and LLVM will resolve references to each
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]694appropriately.</p>
695
Chris Lattner662c8722005-11-12 00:45:07 +0000[diff] [blame]696<p>LLVM allows an explicit section to be specified for functions. If the target
697supports it, it will emit functions to the section specified.</p>
698
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]699<p>An explicit alignment may be specified for a function. If not present, or if
700the alignment is set to zero, the alignment of the function is set by the target
701to whatever it feels convenient. If an explicit alignment is specified, the
702function is forced to have at least that much alignment. All alignments must be
703a power of 2.</p>
704
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]705</div>
706
Chris Lattner91c15c42006-01-23 23:23:47 +0000[diff] [blame]707<!-- ======================================================================= -->
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]708<div class="doc_subsection"><a name="paramattrs">Parameter Attributes</a></div>
709<div class="doc_text">
710 <p>The return type and each parameter of a function type may have a set of
711 <i>parameter attributes</i> associated with them. Parameter attributes are
712 used to communicate additional information about the result or parameters of
713 a function. Parameter attributes are considered to be part of the function
714 type so two functions types that differ only by the parameter attributes
715 are different function types.</p>
716
Reid Spencercf7ebf52007-01-15 18:27:39 +0000[diff] [blame]717 <p>Parameter attributes are simple keywords that follow the type specified. If
718 multiple parameter attributes are needed, they are space separated. For
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]719 example:</p><pre>
Reid Spencercf7ebf52007-01-15 18:27:39 +0000[diff] [blame]720 %someFunc = i16 (i8 sext %someParam) zext
721 %someFunc = i16 (i8 zext %someParam) zext</pre>
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]722 <p>Note that the two function types above are unique because the parameter has
Reid Spencercf7ebf52007-01-15 18:27:39 +0000[diff] [blame]723 a different attribute (sext in the first one, zext in the second). Also note
724 that the attribute for the function result (zext) comes immediately after the
725 argument list.</p>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]726
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]727 <p>Currently, only the following parameter attributes are defined:</p>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]728 <dl>
Reid Spencercf7ebf52007-01-15 18:27:39 +0000[diff] [blame]729 <dt><tt>zext</tt></dt>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]730 <dd>This indicates that the parameter should be zero extended just before
731 a call to this function.</dd>
Reid Spencercf7ebf52007-01-15 18:27:39 +0000[diff] [blame]732 <dt><tt>sext</tt></dt>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]733 <dd>This indicates that the parameter should be sign extended just before
734 a call to this function.</dd>
Anton Korobeynikove8166852007-01-28 14:30:45 +0000[diff] [blame]735 <dt><tt>inreg</tt></dt>
736 <dd>This indicates that the parameter should be placed in register (if
Anton Korobeynikove93c6e82007-01-28 15:27:21 +0000[diff] [blame]737 possible) during assembling function call. Support for this attribute is
738 target-specific</dd>
Anton Korobeynikove8166852007-01-28 14:30:45 +0000[diff] [blame]739 <dt><tt>sret</tt></dt>
Anton Korobeynikove93c6e82007-01-28 15:27:21 +0000[diff] [blame]740 <dd>This indicates that the parameter specifies the address of a structure
Reid Spencer05dbb9d2007-03-22 02:02:11 +0000[diff] [blame]741 that is the return value of the function in the source program.</dd>
Reid Spencer9d1700e2007-03-22 02:18:56 +0000[diff] [blame]742 <dt><tt>noreturn</tt></dt>
743 <dd>This function attribute indicates that the function never returns. This
744 indicates to LLVM that every call to this function should be treated as if
745 an <tt>unreachable</tt> instruction immediately followed the call.</dd>
Reid Spencer05dbb9d2007-03-22 02:02:11 +0000[diff] [blame]746 <dt><tt>nounwind</tt></dt>
747 <dd>This function attribute indicates that the function type does not use
748 the unwind instruction and does not allow stack unwinding to propagate
749 through it.</dd>
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]750 </dl>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]751
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]752</div>
753
754<!-- ======================================================================= -->
Chris Lattner91c15c42006-01-23 23:23:47 +0000[diff] [blame]755<div class="doc_subsection">
Chris Lattner93564892006-04-08 04:40:53 +0000[diff] [blame]756 <a name="moduleasm">Module-Level Inline Assembly</a>
Chris Lattner91c15c42006-01-23 23:23:47 +0000[diff] [blame]757</div>
758
759<div class="doc_text">
760<p>
761Modules may contain "module-level inline asm" blocks, which corresponds to the
762GCC "file scope inline asm" blocks. These blocks are internally concatenated by
763LLVM and treated as a single unit, but may be separated in the .ll file if
764desired. The syntax is very simple:
765</p>
766
767<div class="doc_code"><pre>
Chris Lattnera1280ad2006-01-24 00:37:20 +0000[diff] [blame]768 module asm "inline asm code goes here"
769 module asm "more can go here"
Chris Lattner91c15c42006-01-23 23:23:47 +0000[diff] [blame]770</pre></div>
771
772<p>The strings can contain any character by escaping non-printable characters.
773 The escape sequence used is simply "\xx" where "xx" is the two digit hex code
774 for the number.
775</p>
776
777<p>
778 The inline asm code is simply printed to the machine code .s file when
779 assembly code is generated.
780</p>
781</div>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]782
Reid Spencer50c723a2007-02-19 23:54:10 +0000[diff] [blame]783<!-- ======================================================================= -->
784<div class="doc_subsection">
785 <a name="datalayout">Data Layout</a>
786</div>
787
788<div class="doc_text">
789<p>A module may specify a target specific data layout string that specifies how
790data is to be laid out in memory. The syntax for the data layout is simply:<br/>
791<pre> target datalayout = "<i>layout specification</i>"
792</pre>
793The <i>layout specification</i> consists of a list of specifications separated
794by the minus sign character ('-'). Each specification starts with a letter
795and may include other information after the letter to define some aspect of the
796data layout. The specifications accepted are as follows: </p>
797<dl>
798 <dt><tt>E</tt></dt>
799 <dd>Specifies that the target lays out data in big-endian form. That is, the
800 bits with the most significance have the lowest address location.</dd>
801 <dt><tt>e</tt></dt>
802 <dd>Specifies that hte target lays out data in little-endian form. That is,
803 the bits with the least significance have the lowest address location.</dd>
804 <dt><tt>p:<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
805 <dd>This specifies the <i>size</i> of a pointer and its <i>abi</i> and
806 <i>preferred</i> alignments. All sizes are in bits. Specifying the <i>pref</i>
807 alignment is optional. If omitted, the preceding <tt>:</tt> should be omitted
808 too.</dd>
809 <dt><tt>i<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
810 <dd>This specifies the alignment for an integer type of a given bit
811 <i>size</i>. The value of <i>size</i> must be in the range [1,2^23).</dd>
812 <dt><tt>v<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
813 <dd>This specifies the alignment for a vector type of a given bit
814 <i>size</i>.</dd>
815 <dt><tt>f<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
816 <dd>This specifies the alignment for a floating point type of a given bit
817 <i>size</i>. The value of <i>size</i> must be either 32 (float) or 64
818 (double).</dd>
819 <dt><tt>a<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
820 <dd>This specifies the alignment for an aggregate type of a given bit
821 <i>size</i>.</dd>
822</dl>
823<p>When constructing the data layout for a given target, LLVM starts with a
824default set of specifications which are then (possibly) overriden by the
825specifications in the <tt>datalayout</tt> keyword. The default specifications
826are given in this list:</p>
827<ul>
828 <li><tt>E</tt> - big endian</li>
829 <li><tt>p:32:64:64</tt> - 32-bit pointers with 64-bit alignment</li>
830 <li><tt>i1:8:8</tt> - i1 is 8-bit (byte) aligned</li>
831 <li><tt>i8:8:8</tt> - i8 is 8-bit (byte) aligned</li>
832 <li><tt>i16:16:16</tt> - i16 is 16-bit aligned</li>
833 <li><tt>i32:32:32</tt> - i32 is 32-bit aligned</li>
834 <li><tt>i64:32:64</tt> - i64 has abi alignment of 32-bits but preferred
835 alignment of 64-bits</li>
836 <li><tt>f32:32:32</tt> - float is 32-bit aligned</li>
837 <li><tt>f64:64:64</tt> - double is 64-bit aligned</li>
838 <li><tt>v64:64:64</tt> - 64-bit vector is 64-bit aligned</li>
839 <li><tt>v128:128:128</tt> - 128-bit vector is 128-bit aligned</li>
840 <li><tt>a0:0:1</tt> - aggregates are 8-bit aligned</li>
841</ul>
842<p>When llvm is determining the alignment for a given type, it uses the
843following rules:
844<ol>
845 <li>If the type sought is an exact match for one of the specifications, that
846 specification is used.</li>
847 <li>If no match is found, and the type sought is an integer type, then the
848 smallest integer type that is larger than the bitwidth of the sought type is
849 used. If none of the specifications are larger than the bitwidth then the the
850 largest integer type is used. For example, given the default specifications
851 above, the i7 type will use the alignment of i8 (next largest) while both
852 i65 and i256 will use the alignment of i64 (largest specified).</li>
853 <li>If no match is found, and the type sought is a vector type, then the
854 largest vector type that is smaller than the sought vector type will be used
855 as a fall back. This happens because <128 x double> can be implemented in
856 terms of 64 <2 x double>, for example.</li>
857</ol>
858</div>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]859
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]860<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]861<div class="doc_section"> <a name="typesystem">Type System</a> </div>
862<!-- *********************************************************************** -->
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]863
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]864<div class="doc_text">
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]865
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]866<p>The LLVM type system is one of the most important features of the
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]867intermediate representation. Being typed enables a number of
868optimizations to be performed on the IR directly, without having to do
869extra analyses on the side before the transformation. A strong type
870system makes it easier to read the generated code and enables novel
871analyses and transformations that are not feasible to perform on normal
872three address code representations.</p>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]873
874</div>
875
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]876<!-- ======================================================================= -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]877<div class="doc_subsection"> <a name="t_primitive">Primitive Types</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]878<div class="doc_text">
John Criswell417228d2004-04-09 16:48:45 +0000[diff] [blame]879<p>The primitive types are the fundamental building blocks of the LLVM
Chris Lattner455fc8c2005-03-07 22:13:59 +0000[diff] [blame]880system. The current set of primitive types is as follows:</p>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]881
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]882<table class="layout">
883 <tr class="layout">
884 <td class="left">
885 <table>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]886 <tbody>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]887 <tr><th>Type</th><th>Description</th></tr>
888 <tr><td><tt>void</tt></td><td>No value</td></tr>
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]889 <tr><td><tt>i8</tt></td><td>8-bit value</td></tr>
890 <tr><td><tt>i32</tt></td><td>32-bit value</td></tr>
Misha Brukman36c6bc12005-04-22 18:02:52 +0000[diff] [blame]891 <tr><td><tt>float</tt></td><td>32-bit floating point value</td></tr>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]892 <tr><td><tt>label</tt></td><td>Branch destination</td></tr>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]893 </tbody>
894 </table>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]895 </td>
896 <td class="right">
897 <table>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]898 <tbody>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]899 <tr><th>Type</th><th>Description</th></tr>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]900 <tr><td><tt>i1</tt></td><td>True or False value</td></tr>
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]901 <tr><td><tt>i16</tt></td><td>16-bit value</td></tr>
902 <tr><td><tt>i64</tt></td><td>64-bit value</td></tr>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]903 <tr><td><tt>double</tt></td><td>64-bit floating point value</td></tr>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]904 </tbody>
905 </table>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]906 </td>
907 </tr>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]908</table>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]909</div>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]910
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]911<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]912<div class="doc_subsubsection"> <a name="t_classifications">Type
913Classifications</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]914<div class="doc_text">
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]915<p>These different primitive types fall into a few useful
916classifications:</p>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]917
918<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]919 <tbody>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]920 <tr><th>Classification</th><th>Types</th></tr>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]921 <tr>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]922 <td><a name="t_integer">integer</a></td>
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]923 <td><tt>i1, i8, i16, i32, i64</tt></td>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]924 </tr>
925 <tr>
926 <td><a name="t_floating">floating point</a></td>
927 <td><tt>float, double</tt></td>
928 </tr>
929 <tr>
930 <td><a name="t_firstclass">first class</a></td>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]931 <td><tt>i1, i8, i16, i32, i64, float, double, <br/>
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]932 <a href="#t_pointer">pointer</a>,<a href="#t_vector">vector</a></tt>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]933 </td>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]934 </tr>
935 </tbody>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]936</table>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]937
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]938<p>The <a href="#t_firstclass">first class</a> types are perhaps the
939most important. Values of these types are the only ones which can be
940produced by instructions, passed as arguments, or used as operands to
941instructions. This means that all structures and arrays must be
942manipulated either by pointer or by component.</p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]943</div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]944
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]945<!-- ======================================================================= -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]946<div class="doc_subsection"> <a name="t_derived">Derived Types</a> </div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]947
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]948<div class="doc_text">
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]949
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]950<p>The real power in LLVM comes from the derived types in the system.
951This is what allows a programmer to represent arrays, functions,
952pointers, and other useful types. Note that these derived types may be
953recursive: For example, it is possible to have a two dimensional array.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]954
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]955</div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]956
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]957<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]958<div class="doc_subsubsection"> <a name="t_array">Array Type</a> </div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]959
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]960<div class="doc_text">
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]961
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]962<h5>Overview:</h5>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]963
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]964<p>The array type is a very simple derived type that arranges elements
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]965sequentially in memory. The array type requires a size (number of
966elements) and an underlying data type.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]967
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]968<h5>Syntax:</h5>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]969
970<pre>
971 [&lt;# elements&gt; x &lt;elementtype&gt;]
972</pre>
973
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]974<p>The number of elements is a constant integer value; elementtype may
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]975be any type with a size.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]976
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]977<h5>Examples:</h5>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]978<table class="layout">
979 <tr class="layout">
980 <td class="left">
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]981 <tt>[40 x i32 ]</tt><br/>
982 <tt>[41 x i32 ]</tt><br/>
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]983 <tt>[40 x i8]</tt><br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]984 </td>
985 <td class="left">
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]986 Array of 40 32-bit integer values.<br/>
987 Array of 41 32-bit integer values.<br/>
988 Array of 40 8-bit integer values.<br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]989 </td>
990 </tr>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]991</table>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]992<p>Here are some examples of multidimensional arrays:</p>
993<table class="layout">
994 <tr class="layout">
995 <td class="left">
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]996 <tt>[3 x [4 x i32]]</tt><br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]997 <tt>[12 x [10 x float]]</tt><br/>
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]998 <tt>[2 x [3 x [4 x i16]]]</tt><br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]999 </td>
1000 <td class="left">
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]1001 3x4 array of 32-bit integer values.<br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1002 12x10 array of single precision floating point values.<br/>
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]1003 2x3x4 array of 16-bit integer values.<br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1004 </td>
1005 </tr>
1006</table>
Chris Lattnerc0ad71e2005-06-24 17:22:57 +0000[diff] [blame]1007
John Criswell4c0cf7f2005-10-24 16:17:18 +0000[diff] [blame]1008<p>Note that 'variable sized arrays' can be implemented in LLVM with a zero
1009length array. Normally, accesses past the end of an array are undefined in
Chris Lattnerc0ad71e2005-06-24 17:22:57 +0000[diff] [blame]1010LLVM (e.g. it is illegal to access the 5th element of a 3 element array).
1011As a special case, however, zero length arrays are recognized to be variable
1012length. This allows implementation of 'pascal style arrays' with the LLVM
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1013type "{ i32, [0 x float]}", for example.</p>
Chris Lattnerc0ad71e2005-06-24 17:22:57 +0000[diff] [blame]1014
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1015</div>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1016
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1017<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1018<div class="doc_subsubsection"> <a name="t_function">Function Type</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1019<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1020<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1021<p>The function type can be thought of as a function signature. It
1022consists of a return type and a list of formal parameter types.
John Criswella0d50d22003-11-25 21:45:46 +0000[diff] [blame]1023Function types are usually used to build virtual function tables
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1024(which are structures of pointers to functions), for indirect function
1025calls, and when defining a function.</p>
John Criswella0d50d22003-11-25 21:45:46 +0000[diff] [blame]1026<p>
1027The return type of a function type cannot be an aggregate type.
1028</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1029<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1030<pre> &lt;returntype&gt; (&lt;parameter list&gt;)<br></pre>
John Criswell4c0cf7f2005-10-24 16:17:18 +0000[diff] [blame]1031<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]1032specifiers. Optionally, the parameter list may include a type <tt>...</tt>,
Chris Lattner5ed60612003-09-03 00:41:47 +0000[diff] [blame]1033which indicates that the function takes a variable number of arguments.
1034Variable argument functions can access their arguments with the <a
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1035 href="#int_varargs">variable argument handling intrinsic</a> functions.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1036<h5>Examples:</h5>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1037<table class="layout">
1038 <tr class="layout">
Reid Spencer58c08712006-12-31 07:18:34 +0000[diff] [blame]1039 <td class="left"><tt>i32 (i32)</tt></td>
1040 <td class="left">function taking an <tt>i32</tt>, returning an <tt>i32</tt>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1041 </td>
Reid Spencer58c08712006-12-31 07:18:34 +0000[diff] [blame]1042 </tr><tr class="layout">
Reid Spencere6a338d2007-01-15 18:28:34 +0000[diff] [blame]1043 <td class="left"><tt>float&nbsp;(i16&nbsp;sext,&nbsp;i32&nbsp;*)&nbsp;*
Reid Spencer655dcc62006-12-31 07:20:23 +0000[diff] [blame]1044 </tt></td>
Reid Spencer58c08712006-12-31 07:18:34 +0000[diff] [blame]1045 <td class="left"><a href="#t_pointer">Pointer</a> to a function that takes
1046 an <tt>i16</tt> that should be sign extended and a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1047 <a href="#t_pointer">pointer</a> to <tt>i32</tt>, returning
Reid Spencer58c08712006-12-31 07:18:34 +0000[diff] [blame]1048 <tt>float</tt>.
1049 </td>
1050 </tr><tr class="layout">
1051 <td class="left"><tt>i32 (i8*, ...)</tt></td>
1052 <td class="left">A vararg function that takes at least one
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]1053 <a href="#t_pointer">pointer</a> to <tt>i8 </tt> (char in C),
Reid Spencer58c08712006-12-31 07:18:34 +0000[diff] [blame]1054 which returns an integer. This is the signature for <tt>printf</tt> in
1055 LLVM.
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1056 </td>
1057 </tr>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1058</table>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]1059
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1060</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1061<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1062<div class="doc_subsubsection"> <a name="t_struct">Structure Type</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1063<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1064<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1065<p>The structure type is used to represent a collection of data members
1066together in memory. The packing of the field types is defined to match
1067the ABI of the underlying processor. The elements of a structure may
1068be any type that has a size.</p>
1069<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
1070and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
1071field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
1072instruction.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1073<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1074<pre> { &lt;type list&gt; }<br></pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1075<h5>Examples:</h5>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1076<table class="layout">
1077 <tr class="layout">
1078 <td class="left">
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1079 <tt>{ i32, i32, i32 }</tt><br/>
1080 <tt>{ float, i32 (i32) * }</tt><br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1081 </td>
1082 <td class="left">
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1083 a triple of three <tt>i32</tt> values<br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1084 A pair, where the first element is a <tt>float</tt> and the second element
1085 is a <a href="#t_pointer">pointer</a> to a <a href="#t_function">function</a>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1086 that takes an <tt>i32</tt>, returning an <tt>i32</tt>.<br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1087 </td>
1088 </tr>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1089</table>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1090</div>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1091
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1092<!-- _______________________________________________________________________ -->
Andrew Lenharth8df88e22006-12-08 17:13:00 +0000[diff] [blame]1093<div class="doc_subsubsection"> <a name="t_pstruct">Packed Structure Type</a>
1094</div>
1095<div class="doc_text">
1096<h5>Overview:</h5>
1097<p>The packed structure type is used to represent a collection of data members
1098together in memory. There is no padding between fields. Further, the alignment
1099of a packed structure is 1 byte. The elements of a packed structure may
1100be any type that has a size.</p>
1101<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
1102and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
1103field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
1104instruction.</p>
1105<h5>Syntax:</h5>
1106<pre> &lt; { &lt;type list&gt; } &gt; <br></pre>
1107<h5>Examples:</h5>
1108<table class="layout">
1109 <tr class="layout">
1110 <td class="left">
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1111 <tt> &lt; { i32, i32, i32 } &gt; </tt><br/>
1112 <tt> &lt; { float, i32 (i32) * } &gt; </tt><br/>
Andrew Lenharth8df88e22006-12-08 17:13:00 +0000[diff] [blame]1113 </td>
1114 <td class="left">
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1115 a triple of three <tt>i32</tt> values<br/>
Andrew Lenharth8df88e22006-12-08 17:13:00 +0000[diff] [blame]1116 A pair, where the first element is a <tt>float</tt> and the second element
1117 is a <a href="#t_pointer">pointer</a> to a <a href="#t_function">function</a>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1118 that takes an <tt>i32</tt>, returning an <tt>i32</tt>.<br/>
Andrew Lenharth8df88e22006-12-08 17:13:00 +0000[diff] [blame]1119 </td>
1120 </tr>
1121</table>
1122</div>
1123
1124<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1125<div class="doc_subsubsection"> <a name="t_pointer">Pointer Type</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1126<div class="doc_text">
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]1127<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1128<p>As in many languages, the pointer type represents a pointer or
1129reference to another object, which must live in memory.</p>
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]1130<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1131<pre> &lt;type&gt; *<br></pre>
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]1132<h5>Examples:</h5>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1133<table class="layout">
1134 <tr class="layout">
1135 <td class="left">
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1136 <tt>[4x i32]*</tt><br/>
1137 <tt>i32 (i32 *) *</tt><br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1138 </td>
1139 <td class="left">
1140 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]1141 four <tt>i32</tt> values<br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1142 A <a href="#t_pointer">pointer</a> to a <a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1143 href="#t_function">function</a> that takes an <tt>i32*</tt>, returning an
1144 <tt>i32</tt>.<br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1145 </td>
1146 </tr>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1147</table>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1148</div>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1149
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1150<!-- _______________________________________________________________________ -->
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1151<div class="doc_subsubsection"> <a name="t_vector">Vector Type</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1152<div class="doc_text">
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]1153
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1154<h5>Overview:</h5>
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]1155
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1156<p>A vector type is a simple derived type that represents a vector
1157of elements. Vector types are used when multiple primitive data
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1158are operated in parallel using a single instruction (SIMD).
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1159A vector type requires a size (number of
Chris Lattner330ce692005-11-10 01:44:22 +0000[diff] [blame]1160elements) and an underlying primitive data type. Vectors must have a power
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1161of two length (1, 2, 4, 8, 16 ...). Vector types are
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1162considered <a href="#t_firstclass">first class</a>.</p>
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]1163
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1164<h5>Syntax:</h5>
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]1165
1166<pre>
1167 &lt; &lt;# elements&gt; x &lt;elementtype&gt; &gt;
1168</pre>
1169
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]1170<p>The number of elements is a constant integer value; elementtype may
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]1171be any integer or floating point type.</p>
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]1172
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1173<h5>Examples:</h5>
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]1174
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1175<table class="layout">
1176 <tr class="layout">
1177 <td class="left">
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1178 <tt>&lt;4 x i32&gt;</tt><br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1179 <tt>&lt;8 x float&gt;</tt><br/>
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]1180 <tt>&lt;2 x i64&gt;</tt><br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1181 </td>
1182 <td class="left">
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1183 Vector of 4 32-bit integer values.<br/>
1184 Vector of 8 floating-point values.<br/>
1185 Vector of 2 64-bit integer values.<br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1186 </td>
1187 </tr>
1188</table>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1189</div>
1190
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]1191<!-- _______________________________________________________________________ -->
1192<div class="doc_subsubsection"> <a name="t_opaque">Opaque Type</a> </div>
1193<div class="doc_text">
1194
1195<h5>Overview:</h5>
1196
1197<p>Opaque types are used to represent unknown types in the system. This
1198corresponds (for example) to the C notion of a foward declared structure type.
1199In LLVM, opaque types can eventually be resolved to any type (not just a
1200structure type).</p>
1201
1202<h5>Syntax:</h5>
1203
1204<pre>
1205 opaque
1206</pre>
1207
1208<h5>Examples:</h5>
1209
1210<table class="layout">
1211 <tr class="layout">
1212 <td class="left">
1213 <tt>opaque</tt>
1214 </td>
1215 <td class="left">
1216 An opaque type.<br/>
1217 </td>
1218 </tr>
1219</table>
1220</div>
1221
1222
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1223<!-- *********************************************************************** -->
1224<div class="doc_section"> <a name="constants">Constants</a> </div>
1225<!-- *********************************************************************** -->
1226
1227<div class="doc_text">
1228
1229<p>LLVM has several different basic types of constants. This section describes
1230them all and their syntax.</p>
1231
1232</div>
1233
1234<!-- ======================================================================= -->
Reid Spencer8f08d802004-12-09 18:02:53 +0000[diff] [blame]1235<div class="doc_subsection"><a name="simpleconstants">Simple Constants</a></div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1236
1237<div class="doc_text">
1238
1239<dl>
1240 <dt><b>Boolean constants</b></dt>
1241
1242 <dd>The two strings '<tt>true</tt>' and '<tt>false</tt>' are both valid
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]1243 constants of the <tt><a href="#t_primitive">i1</a></tt> type.
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1244 </dd>
1245
1246 <dt><b>Integer constants</b></dt>
1247
Reid Spencer8f08d802004-12-09 18:02:53 +0000[diff] [blame]1248 <dd>Standard integers (such as '4') are constants of the <a
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]1249 href="#t_integer">integer</a> type. Negative numbers may be used with
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1250 integer types.
1251 </dd>
1252
1253 <dt><b>Floating point constants</b></dt>
1254
1255 <dd>Floating point constants use standard decimal notation (e.g. 123.421),
1256 exponential notation (e.g. 1.23421e+2), or a more precise hexadecimal
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1257 notation (see below). Floating point constants must have a <a
1258 href="#t_floating">floating point</a> type. </dd>
1259
1260 <dt><b>Null pointer constants</b></dt>
1261
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]1262 <dd>The identifier '<tt>null</tt>' is recognized as a null pointer constant
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1263 and must be of <a href="#t_pointer">pointer type</a>.</dd>
1264
1265</dl>
1266
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]1267<p>The one non-intuitive notation for constants is the optional hexadecimal form
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1268of floating point constants. For example, the form '<tt>double
12690x432ff973cafa8000</tt>' is equivalent to (but harder to read than) '<tt>double
12704.5e+15</tt>'. The only time hexadecimal floating point constants are required
Reid Spencer8f08d802004-12-09 18:02:53 +0000[diff] [blame]1271(and the only time that they are generated by the disassembler) is when a
1272floating point constant must be emitted but it cannot be represented as a
1273decimal floating point number. For example, NaN's, infinities, and other
1274special values are represented in their IEEE hexadecimal format so that
1275assembly and disassembly do not cause any bits to change in the constants.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1276
1277</div>
1278
1279<!-- ======================================================================= -->
1280<div class="doc_subsection"><a name="aggregateconstants">Aggregate Constants</a>
1281</div>
1282
1283<div class="doc_text">
Chris Lattner455fc8c2005-03-07 22:13:59 +0000[diff] [blame]1284<p>Aggregate constants arise from aggregation of simple constants
1285and smaller aggregate constants.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1286
1287<dl>
1288 <dt><b>Structure constants</b></dt>
1289
1290 <dd>Structure constants are represented with notation similar to structure
1291 type definitions (a comma separated list of elements, surrounded by braces
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1292 (<tt>{}</tt>)). For example: "<tt>{ i32 4, float 17.0, i32* %G }</tt>",
1293 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]1294 must have <a href="#t_struct">structure type</a>, and the number and
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1295 types of elements must match those specified by the type.
1296 </dd>
1297
1298 <dt><b>Array constants</b></dt>
1299
1300 <dd>Array constants are represented with notation similar to array type
1301 definitions (a comma separated list of elements, surrounded by square brackets
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1302 (<tt>[]</tt>)). For example: "<tt>[ i32 42, i32 11, i32 74 ]</tt>". Array
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1303 constants must have <a href="#t_array">array type</a>, and the number and
1304 types of elements must match those specified by the type.
1305 </dd>
1306
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1307 <dt><b>Vector constants</b></dt>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1308
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1309 <dd>Vector constants are represented with notation similar to vector type
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1310 definitions (a comma separated list of elements, surrounded by
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1311 less-than/greater-than's (<tt>&lt;&gt;</tt>)). For example: "<tt>&lt; i32 42,
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1312 i32 11, i32 74, i32 100 &gt;</tt>". VEctor constants must have <a
1313 href="#t_vector">vector type</a>, and the number and types of elements must
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1314 match those specified by the type.
1315 </dd>
1316
1317 <dt><b>Zero initialization</b></dt>
1318
1319 <dd>The string '<tt>zeroinitializer</tt>' can be used to zero initialize a
1320 value to zero of <em>any</em> type, including scalar and aggregate types.
1321 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]1322 large arrays) and is always exactly equivalent to using explicit zero
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1323 initializers.
1324 </dd>
1325</dl>
1326
1327</div>
1328
1329<!-- ======================================================================= -->
1330<div class="doc_subsection">
1331 <a name="globalconstants">Global Variable and Function Addresses</a>
1332</div>
1333
1334<div class="doc_text">
1335
1336<p>The addresses of <a href="#globalvars">global variables</a> and <a
1337href="#functionstructure">functions</a> are always implicitly valid (link-time)
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]1338constants. These constants are explicitly referenced when the <a
1339href="#identifiers">identifier for the global</a> is used and always have <a
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1340href="#t_pointer">pointer</a> type. For example, the following is a legal LLVM
1341file:</p>
1342
1343<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1344 %X = global i32 17
1345 %Y = global i32 42
1346 %Z = global [2 x i32*] [ i32* %X, i32* %Y ]
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1347</pre>
1348
1349</div>
1350
1351<!-- ======================================================================= -->
Reid Spencer641f5c92004-12-09 18:13:12 +0000[diff] [blame]1352<div class="doc_subsection"><a name="undefvalues">Undefined Values</a></div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1353<div class="doc_text">
Reid Spencer641f5c92004-12-09 18:13:12 +0000[diff] [blame]1354 <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]1355 no specific value. Undefined values may be of any type and be used anywhere
Reid Spencer641f5c92004-12-09 18:13:12 +0000[diff] [blame]1356 a constant is permitted.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1357
Reid Spencer641f5c92004-12-09 18:13:12 +0000[diff] [blame]1358 <p>Undefined values indicate to the compiler that the program is well defined
1359 no matter what value is used, giving the compiler more freedom to optimize.
1360 </p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1361</div>
1362
1363<!-- ======================================================================= -->
1364<div class="doc_subsection"><a name="constantexprs">Constant Expressions</a>
1365</div>
1366
1367<div class="doc_text">
1368
1369<p>Constant expressions are used to allow expressions involving other constants
1370to be used as constants. Constant expressions may be of any <a
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]1371href="#t_firstclass">first class</a> type and may involve any LLVM operation
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1372that does not have side effects (e.g. load and call are not supported). The
1373following is the syntax for constant expressions:</p>
1374
1375<dl>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]1376 <dt><b><tt>trunc ( CST to TYPE )</tt></b></dt>
1377 <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]1378 than the bit size of TYPE. Both types must be integers.</dd>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1379
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]1380 <dt><b><tt>zext ( CST to TYPE )</tt></b></dt>
1381 <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]1382 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]1383
1384 <dt><b><tt>sext ( CST to TYPE )</tt></b></dt>
1385 <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]1386 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]1387
1388 <dt><b><tt>fptrunc ( CST to TYPE )</tt></b></dt>
1389 <dd>Truncate a floating point constant to another floating point type. The
1390 size of CST must be larger than the size of TYPE. Both types must be
1391 floating point.</dd>
1392
1393 <dt><b><tt>fpext ( CST to TYPE )</tt></b></dt>
1394 <dd>Floating point extend a constant to another type. The size of CST must be
1395 smaller or equal to the size of TYPE. Both types must be floating point.</dd>
1396
1397 <dt><b><tt>fp2uint ( CST to TYPE )</tt></b></dt>
1398 <dd>Convert a floating point constant to the corresponding unsigned integer
1399 constant. TYPE must be an integer type. CST must be floating point. If the
1400 value won't fit in the integer type, the results are undefined.</dd>
1401
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]1402 <dt><b><tt>fptosi ( CST to TYPE )</tt></b></dt>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]1403 <dd>Convert a floating point constant to the corresponding signed integer
1404 constant. TYPE must be an integer type. CST must be floating point. If the
1405 value won't fit in the integer type, the results are undefined.</dd>
1406
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]1407 <dt><b><tt>uitofp ( CST to TYPE )</tt></b></dt>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]1408 <dd>Convert an unsigned integer constant to the corresponding floating point
1409 constant. TYPE must be floating point. CST must be of integer type. If the
1410 value won't fit in the floating point type, the results are undefined.</dd>
1411
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]1412 <dt><b><tt>sitofp ( CST to TYPE )</tt></b></dt>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]1413 <dd>Convert a signed integer constant to the corresponding floating point
1414 constant. TYPE must be floating point. CST must be of integer type. If the
1415 value won't fit in the floating point type, the results are undefined.</dd>
1416
Reid Spencer5b950642006-11-11 23:08:07 +0000[diff] [blame]1417 <dt><b><tt>ptrtoint ( CST to TYPE )</tt></b></dt>
1418 <dd>Convert a pointer typed constant to the corresponding integer constant
1419 TYPE must be an integer type. CST must be of pointer type. The CST value is
1420 zero extended, truncated, or unchanged to make it fit in TYPE.</dd>
1421
1422 <dt><b><tt>inttoptr ( CST to TYPE )</tt></b></dt>
1423 <dd>Convert a integer constant to a pointer constant. TYPE must be a
1424 pointer type. CST must be of integer type. The CST value is zero extended,
1425 truncated, or unchanged to make it fit in a pointer size. This one is
1426 <i>really</i> dangerous!</dd>
1427
1428 <dt><b><tt>bitcast ( CST to TYPE )</tt></b></dt>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]1429 <dd>Convert a constant, CST, to another TYPE. The size of CST and TYPE must be
1430 identical (same number of bits). The conversion is done as if the CST value
1431 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]1432 with this operator, just the type. This can be used for conversion of
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1433 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]1434 pointers it is only valid to cast to another pointer type.
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]1435 </dd>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1436
1437 <dt><b><tt>getelementptr ( CSTPTR, IDX0, IDX1, ... )</tt></b></dt>
1438
1439 <dd>Perform the <a href="#i_getelementptr">getelementptr operation</a> on
1440 constants. As with the <a href="#i_getelementptr">getelementptr</a>
1441 instruction, the index list may have zero or more indexes, which are required
1442 to make sense for the type of "CSTPTR".</dd>
1443
Robert Bocchino7e97a6d2006-01-10 19:31:34 +0000[diff] [blame]1444 <dt><b><tt>select ( COND, VAL1, VAL2 )</tt></b></dt>
1445
1446 <dd>Perform the <a href="#i_select">select operation</a> on
Reid Spencer9965ee72006-12-04 19:23:19 +0000[diff] [blame]1447 constants.</dd>
1448
1449 <dt><b><tt>icmp COND ( VAL1, VAL2 )</tt></b></dt>
1450 <dd>Performs the <a href="#i_icmp">icmp operation</a> on constants.</dd>
1451
1452 <dt><b><tt>fcmp COND ( VAL1, VAL2 )</tt></b></dt>
1453 <dd>Performs the <a href="#i_fcmp">fcmp operation</a> on constants.</dd>
Robert Bocchino7e97a6d2006-01-10 19:31:34 +0000[diff] [blame]1454
1455 <dt><b><tt>extractelement ( VAL, IDX )</tt></b></dt>
1456
1457 <dd>Perform the <a href="#i_extractelement">extractelement
1458 operation</a> on constants.
1459
Robert Bocchinof72fdfe2006-01-15 20:48:27 +0000[diff] [blame]1460 <dt><b><tt>insertelement ( VAL, ELT, IDX )</tt></b></dt>
1461
1462 <dd>Perform the <a href="#i_insertelement">insertelement
Reid Spencer9965ee72006-12-04 19:23:19 +0000[diff] [blame]1463 operation</a> on constants.</dd>
Robert Bocchinof72fdfe2006-01-15 20:48:27 +0000[diff] [blame]1464
Chris Lattner016a0e52006-04-08 00:13:41 +0000[diff] [blame]1465
1466 <dt><b><tt>shufflevector ( VEC1, VEC2, IDXMASK )</tt></b></dt>
1467
1468 <dd>Perform the <a href="#i_shufflevector">shufflevector
Reid Spencer9965ee72006-12-04 19:23:19 +0000[diff] [blame]1469 operation</a> on constants.</dd>
Chris Lattner016a0e52006-04-08 00:13:41 +0000[diff] [blame]1470
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1471 <dt><b><tt>OPCODE ( LHS, RHS )</tt></b></dt>
1472
Reid Spencer641f5c92004-12-09 18:13:12 +0000[diff] [blame]1473 <dd>Perform the specified operation of the LHS and RHS constants. OPCODE may
1474 be any of the <a href="#binaryops">binary</a> or <a href="#bitwiseops">bitwise
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1475 binary</a> operations. The constraints on operands are the same as those for
1476 the corresponding instruction (e.g. no bitwise operations on floating point
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]1477 values are allowed).</dd>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1478</dl>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1479</div>
Chris Lattnerb1652612004-03-08 16:49:10 +0000[diff] [blame]1480
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1481<!-- *********************************************************************** -->
Chris Lattner98f013c2006-01-25 23:47:57 +0000[diff] [blame]1482<div class="doc_section"> <a name="othervalues">Other Values</a> </div>
1483<!-- *********************************************************************** -->
1484
1485<!-- ======================================================================= -->
1486<div class="doc_subsection">
1487<a name="inlineasm">Inline Assembler Expressions</a>
1488</div>
1489
1490<div class="doc_text">
1491
1492<p>
1493LLVM supports inline assembler expressions (as opposed to <a href="#moduleasm">
1494Module-Level Inline Assembly</a>) through the use of a special value. This
1495value represents the inline assembler as a string (containing the instructions
1496to emit), a list of operand constraints (stored as a string), and a flag that
1497indicates whether or not the inline asm expression has side effects. An example
1498inline assembler expression is:
1499</p>
1500
1501<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1502 i32 (i32) asm "bswap $0", "=r,r"
Chris Lattner98f013c2006-01-25 23:47:57 +0000[diff] [blame]1503</pre>
1504
1505<p>
1506Inline assembler expressions may <b>only</b> be used as the callee operand of
1507a <a href="#i_call"><tt>call</tt> instruction</a>. Thus, typically we have:
1508</p>
1509
1510<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1511 %X = call i32 asm "<a href="#i_bswap">bswap</a> $0", "=r,r"(i32 %Y)
Chris Lattner98f013c2006-01-25 23:47:57 +0000[diff] [blame]1512</pre>
1513
1514<p>
1515Inline asms with side effects not visible in the constraint list must be marked
1516as having side effects. This is done through the use of the
1517'<tt>sideeffect</tt>' keyword, like so:
1518</p>
1519
1520<pre>
1521 call void asm sideeffect "eieio", ""()
1522</pre>
1523
1524<p>TODO: The format of the asm and constraints string still need to be
1525documented here. Constraints on what can be done (e.g. duplication, moving, etc
1526need to be documented).
1527</p>
1528
1529</div>
1530
1531<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1532<div class="doc_section"> <a name="instref">Instruction Reference</a> </div>
1533<!-- *********************************************************************** -->
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1534
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1535<div class="doc_text">
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1536
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1537<p>The LLVM instruction set consists of several different
1538classifications of instructions: <a href="#terminators">terminator
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]1539instructions</a>, <a href="#binaryops">binary instructions</a>,
1540<a href="#bitwiseops">bitwise binary instructions</a>, <a
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1541 href="#memoryops">memory instructions</a>, and <a href="#otherops">other
1542instructions</a>.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1543
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1544</div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1545
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1546<!-- ======================================================================= -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1547<div class="doc_subsection"> <a name="terminators">Terminator
1548Instructions</a> </div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1549
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1550<div class="doc_text">
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1551
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1552<p>As mentioned <a href="#functionstructure">previously</a>, every
1553basic block in a program ends with a "Terminator" instruction, which
1554indicates which block should be executed after the current block is
1555finished. These terminator instructions typically yield a '<tt>void</tt>'
1556value: they produce control flow, not values (the one exception being
1557the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction).</p>
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]1558<p>There are six different terminator instructions: the '<a
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1559 href="#i_ret"><tt>ret</tt></a>' instruction, the '<a href="#i_br"><tt>br</tt></a>'
1560instruction, the '<a href="#i_switch"><tt>switch</tt></a>' instruction,
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1561the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction, the '<a
1562 href="#i_unwind"><tt>unwind</tt></a>' instruction, and the '<a
1563 href="#i_unreachable"><tt>unreachable</tt></a>' instruction.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1564
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1565</div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1566
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1567<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1568<div class="doc_subsubsection"> <a name="i_ret">'<tt>ret</tt>'
1569Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1570<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1571<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1572<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]1573 ret void <i>; Return from void function</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1574</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1575<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1576<p>The '<tt>ret</tt>' instruction is used to return control flow (and a
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]1577value) from a function back to the caller.</p>
John Criswell417228d2004-04-09 16:48:45 +0000[diff] [blame]1578<p>There are two forms of the '<tt>ret</tt>' instruction: one that
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1579returns a value and then causes control flow, and one that just causes
1580control flow to occur.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1581<h5>Arguments:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1582<p>The '<tt>ret</tt>' instruction may return any '<a
1583 href="#t_firstclass">first class</a>' type. Notice that a function is
1584not <a href="#wellformed">well formed</a> if there exists a '<tt>ret</tt>'
1585instruction inside of the function that returns a value that does not
1586match the return type of the function.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1587<h5>Semantics:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1588<p>When the '<tt>ret</tt>' instruction is executed, control flow
1589returns back to the calling function's context. If the caller is a "<a
John Criswell40db33f2004-06-25 15:16:57 +0000[diff] [blame]1590 href="#i_call"><tt>call</tt></a>" instruction, execution continues at
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1591the instruction after the call. If the caller was an "<a
1592 href="#i_invoke"><tt>invoke</tt></a>" instruction, execution continues
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]1593at the beginning of the "normal" destination block. If the instruction
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1594returns a value, that value shall set the call or invoke instruction's
1595return value.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1596<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1597<pre> ret i32 5 <i>; Return an integer value of 5</i>
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]1598 ret void <i>; Return from a void function</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1599</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1600</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1601<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1602<div class="doc_subsubsection"> <a name="i_br">'<tt>br</tt>' Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1603<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1604<h5>Syntax:</h5>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]1605<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]1606</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1607<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1608<p>The '<tt>br</tt>' instruction is used to cause control flow to
1609transfer to a different basic block in the current function. There are
1610two forms of this instruction, corresponding to a conditional branch
1611and an unconditional branch.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1612<h5>Arguments:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1613<p>The conditional branch form of the '<tt>br</tt>' instruction takes a
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]1614single '<tt>i1</tt>' value and two '<tt>label</tt>' values. The
Reid Spencer50c723a2007-02-19 23:54:10 +0000[diff] [blame]1615unconditional form of the '<tt>br</tt>' instruction takes a single
1616'<tt>label</tt>' value as a target.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1617<h5>Semantics:</h5>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]1618<p>Upon execution of a conditional '<tt>br</tt>' instruction, the '<tt>i1</tt>'
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1619argument is evaluated. If the value is <tt>true</tt>, control flows
1620to the '<tt>iftrue</tt>' <tt>label</tt> argument. If "cond" is <tt>false</tt>,
1621control flows to the '<tt>iffalse</tt>' <tt>label</tt> argument.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1622<h5>Example:</h5>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]1623<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]1624 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]1625</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1626<!-- _______________________________________________________________________ -->
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1627<div class="doc_subsubsection">
1628 <a name="i_switch">'<tt>switch</tt>' Instruction</a>
1629</div>
1630
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1631<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1632<h5>Syntax:</h5>
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1633
1634<pre>
1635 switch &lt;intty&gt; &lt;value&gt;, label &lt;defaultdest&gt; [ &lt;intty&gt; &lt;val&gt;, label &lt;dest&gt; ... ]
1636</pre>
1637
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1638<h5>Overview:</h5>
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1639
1640<p>The '<tt>switch</tt>' instruction is used to transfer control flow to one of
1641several different places. It is a generalization of the '<tt>br</tt>'
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1642instruction, allowing a branch to occur to one of many possible
1643destinations.</p>
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1644
1645
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1646<h5>Arguments:</h5>
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1647
1648<p>The '<tt>switch</tt>' instruction uses three parameters: an integer
1649comparison value '<tt>value</tt>', a default '<tt>label</tt>' destination, and
1650an array of pairs of comparison value constants and '<tt>label</tt>'s. The
1651table is not allowed to contain duplicate constant entries.</p>
1652
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1653<h5>Semantics:</h5>
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1654
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1655<p>The <tt>switch</tt> instruction specifies a table of values and
1656destinations. When the '<tt>switch</tt>' instruction is executed, this
John Criswellbcbb18c2004-06-25 16:05:06 +0000[diff] [blame]1657table is searched for the given value. If the value is found, control flow is
1658transfered to the corresponding destination; otherwise, control flow is
1659transfered to the default destination.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1660
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1661<h5>Implementation:</h5>
1662
1663<p>Depending on properties of the target machine and the particular
1664<tt>switch</tt> instruction, this instruction may be code generated in different
John Criswellbcbb18c2004-06-25 16:05:06 +0000[diff] [blame]1665ways. For example, it could be generated as a series of chained conditional
1666branches or with a lookup table.</p>
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1667
1668<h5>Example:</h5>
1669
1670<pre>
1671 <i>; Emulate a conditional br instruction</i>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]1672 %Val = <a href="#i_zext">zext</a> i1 %value to i32
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1673 switch i32 %Val, label %truedest [i32 0, label %falsedest ]
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1674
1675 <i>; Emulate an unconditional br instruction</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1676 switch i32 0, label %dest [ ]
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1677
1678 <i>; Implement a jump table:</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1679 switch i32 %val, label %otherwise [ i32 0, label %onzero
1680 i32 1, label %onone
1681 i32 2, label %ontwo ]
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1682</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1683</div>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1684
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1685<!-- _______________________________________________________________________ -->
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1686<div class="doc_subsubsection">
1687 <a name="i_invoke">'<tt>invoke</tt>' Instruction</a>
1688</div>
1689
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1690<div class="doc_text">
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1691
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1692<h5>Syntax:</h5>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1693
1694<pre>
1695 &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]1696 to label &lt;normal label&gt; unwind label &lt;exception label&gt;
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1697</pre>
1698
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]1699<h5>Overview:</h5>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1700
1701<p>The '<tt>invoke</tt>' instruction causes control to transfer to a specified
1702function, with the possibility of control flow transfer to either the
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]1703'<tt>normal</tt>' label or the
1704'<tt>exception</tt>' label. If the callee function returns with the
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1705"<tt><a href="#i_ret">ret</a></tt>" instruction, control flow will return to the
1706"normal" label. If the callee (or any indirect callees) returns with the "<a
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]1707href="#i_unwind"><tt>unwind</tt></a>" instruction, control is interrupted and
1708continued at the dynamically nearest "exception" label.</p>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1709
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1710<h5>Arguments:</h5>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1711
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1712<p>This instruction requires several arguments:</p>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1713
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1714<ol>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1715 <li>
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]1716 The optional "cconv" marker indicates which <a href="callingconv">calling
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1717 convention</a> the call should use. If none is specified, the call defaults
1718 to using C calling conventions.
1719 </li>
1720 <li>'<tt>ptr to function ty</tt>': shall be the signature of the pointer to
1721 function value being invoked. In most cases, this is a direct function
1722 invocation, but indirect <tt>invoke</tt>s are just as possible, branching off
1723 an arbitrary pointer to function value.
1724 </li>
1725
1726 <li>'<tt>function ptr val</tt>': An LLVM value containing a pointer to a
1727 function to be invoked. </li>
1728
1729 <li>'<tt>function args</tt>': argument list whose types match the function
1730 signature argument types. If the function signature indicates the function
1731 accepts a variable number of arguments, the extra arguments can be
1732 specified. </li>
1733
1734 <li>'<tt>normal label</tt>': the label reached when the called function
1735 executes a '<tt><a href="#i_ret">ret</a></tt>' instruction. </li>
1736
1737 <li>'<tt>exception label</tt>': the label reached when a callee returns with
1738 the <a href="#i_unwind"><tt>unwind</tt></a> instruction. </li>
1739
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1740</ol>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1741
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1742<h5>Semantics:</h5>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1743
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1744<p>This instruction is designed to operate as a standard '<tt><a
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1745href="#i_call">call</a></tt>' instruction in most regards. The primary
1746difference is that it establishes an association with a label, which is used by
1747the runtime library to unwind the stack.</p>
1748
1749<p>This instruction is used in languages with destructors to ensure that proper
1750cleanup is performed in the case of either a <tt>longjmp</tt> or a thrown
1751exception. Additionally, this is important for implementation of
1752'<tt>catch</tt>' clauses in high-level languages that support them.</p>
1753
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1754<h5>Example:</h5>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1755<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1756 %retval = invoke i32 %Test(i32 15) to label %Continue
1757 unwind label %TestCleanup <i>; {i32}:retval set</i>
1758 %retval = invoke <a href="#callingconv">coldcc</a> i32 %Test(i32 15) to label %Continue
1759 unwind label %TestCleanup <i>; {i32}:retval set</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1760</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1761</div>
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1762
1763
Chris Lattner5ed60612003-09-03 00:41:47 +0000[diff] [blame]1764<!-- _______________________________________________________________________ -->
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1765
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1766<div class="doc_subsubsection"> <a name="i_unwind">'<tt>unwind</tt>'
1767Instruction</a> </div>
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1768
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1769<div class="doc_text">
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1770
Chris Lattner5ed60612003-09-03 00:41:47 +0000[diff] [blame]1771<h5>Syntax:</h5>
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1772<pre>
1773 unwind
1774</pre>
1775
Chris Lattner5ed60612003-09-03 00:41:47 +0000[diff] [blame]1776<h5>Overview:</h5>
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1777
1778<p>The '<tt>unwind</tt>' instruction unwinds the stack, continuing control flow
1779at the first callee in the dynamic call stack which used an <a
1780href="#i_invoke"><tt>invoke</tt></a> instruction to perform the call. This is
1781primarily used to implement exception handling.</p>
1782
Chris Lattner5ed60612003-09-03 00:41:47 +0000[diff] [blame]1783<h5>Semantics:</h5>
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1784
1785<p>The '<tt>unwind</tt>' intrinsic causes execution of the current function to
1786immediately halt. The dynamic call stack is then searched for the first <a
1787href="#i_invoke"><tt>invoke</tt></a> instruction on the call stack. Once found,
1788execution continues at the "exceptional" destination block specified by the
1789<tt>invoke</tt> instruction. If there is no <tt>invoke</tt> instruction in the
1790dynamic call chain, undefined behavior results.</p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1791</div>
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1792
1793<!-- _______________________________________________________________________ -->
1794
1795<div class="doc_subsubsection"> <a name="i_unreachable">'<tt>unreachable</tt>'
1796Instruction</a> </div>
1797
1798<div class="doc_text">
1799
1800<h5>Syntax:</h5>
1801<pre>
1802 unreachable
1803</pre>
1804
1805<h5>Overview:</h5>
1806
1807<p>The '<tt>unreachable</tt>' instruction has no defined semantics. This
1808instruction is used to inform the optimizer that a particular portion of the
1809code is not reachable. This can be used to indicate that the code after a
1810no-return function cannot be reached, and other facts.</p>
1811
1812<h5>Semantics:</h5>
1813
1814<p>The '<tt>unreachable</tt>' instruction has no defined semantics.</p>
1815</div>
1816
1817
1818
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_subsection"> <a name="binaryops">Binary Operations</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1821<div class="doc_text">
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1822<p>Binary operators are used to do most of the computation in a
1823program. They require two operands, execute an operation on them, and
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]1824produce a single value. The operands might represent
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1825multiple 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]1826The result value of a binary operator is not
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1827necessarily the same type as its operands.</p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1828<p>There are several different binary operators:</p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1829</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1830<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1831<div class="doc_subsubsection"> <a name="i_add">'<tt>add</tt>'
1832Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1833<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1834<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1835<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]1836</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1837<h5>Overview:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1838<p>The '<tt>add</tt>' instruction returns the sum of its two operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1839<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1840<p>The two arguments to the '<tt>add</tt>' instruction must be either <a
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1841 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a> values.
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1842 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]1843Both arguments must have identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1844<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1845<p>The value produced is the integer or floating point sum of the two
1846operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1847<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1848<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]1849</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1850</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1851<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1852<div class="doc_subsubsection"> <a name="i_sub">'<tt>sub</tt>'
1853Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1854<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1855<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1856<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]1857</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1858<h5>Overview:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1859<p>The '<tt>sub</tt>' instruction returns the difference of its two
1860operands.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1861<p>Note that the '<tt>sub</tt>' instruction is used to represent the '<tt>neg</tt>'
1862instruction present in most other intermediate representations.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1863<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1864<p>The two arguments to the '<tt>sub</tt>' instruction must be either <a
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1865 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1866values.
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1867This instruction can also take <a href="#t_vector">vector</a> versions of the values.
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1868Both arguments must have identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1869<h5>Semantics:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1870<p>The value produced is the integer or floating point difference of
1871the two operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1872<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1873<pre> &lt;result&gt; = sub i32 4, %var <i>; yields {i32}:result = 4 - %var</i>
1874 &lt;result&gt; = sub i32 0, %val <i>; yields {i32}:result = -%var</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1875</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1876</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1877<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1878<div class="doc_subsubsection"> <a name="i_mul">'<tt>mul</tt>'
1879Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1880<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1881<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1882<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]1883</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1884<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1885<p>The '<tt>mul</tt>' instruction returns the product of its two
1886operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1887<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1888<p>The two arguments to the '<tt>mul</tt>' instruction must be either <a
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1889 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1890values.
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1891This instruction can also take <a href="#t_vector">vector</a> versions of the values.
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1892Both arguments must have identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1893<h5>Semantics:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1894<p>The value produced is the integer or floating point product of the
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1895two operands.</p>
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]1896<p>Because the operands are the same width, the result of an integer
1897multiplication is the same whether the operands should be deemed unsigned or
1898signed.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1899<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1900<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]1901</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1902</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1903<!-- _______________________________________________________________________ -->
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame]1904<div class="doc_subsubsection"> <a name="i_udiv">'<tt>udiv</tt>' Instruction
1905</a></div>
1906<div class="doc_text">
1907<h5>Syntax:</h5>
1908<pre> &lt;result&gt; = udiv &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
1909</pre>
1910<h5>Overview:</h5>
1911<p>The '<tt>udiv</tt>' instruction returns the quotient of its two
1912operands.</p>
1913<h5>Arguments:</h5>
1914<p>The two arguments to the '<tt>udiv</tt>' instruction must be
1915<a href="#t_integer">integer</a> values. Both arguments must have identical
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1916types. This instruction can also take <a href="#t_vector">vector</a> versions
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame]1917of the values in which case the elements must be integers.</p>
1918<h5>Semantics:</h5>
1919<p>The value produced is the unsigned integer quotient of the two operands. This
1920instruction always performs an unsigned division operation, regardless of
1921whether the arguments are unsigned or not.</p>
1922<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1923<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]1924</pre>
1925</div>
1926<!-- _______________________________________________________________________ -->
1927<div class="doc_subsubsection"> <a name="i_sdiv">'<tt>sdiv</tt>' Instruction
1928</a> </div>
1929<div class="doc_text">
1930<h5>Syntax:</h5>
1931<pre> &lt;result&gt; = sdiv &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
1932</pre>
1933<h5>Overview:</h5>
1934<p>The '<tt>sdiv</tt>' instruction returns the quotient of its two
1935operands.</p>
1936<h5>Arguments:</h5>
1937<p>The two arguments to the '<tt>sdiv</tt>' instruction must be
1938<a href="#t_integer">integer</a> values. Both arguments must have identical
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1939types. This instruction can also take <a href="#t_vector">vector</a> versions
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame]1940of the values in which case the elements must be integers.</p>
1941<h5>Semantics:</h5>
1942<p>The value produced is the signed integer quotient of the two operands. This
1943instruction always performs a signed division operation, regardless of whether
1944the arguments are signed or not.</p>
1945<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1946<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]1947</pre>
1948</div>
1949<!-- _______________________________________________________________________ -->
1950<div class="doc_subsubsection"> <a name="i_fdiv">'<tt>fdiv</tt>'
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1951Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1952<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1953<h5>Syntax:</h5>
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame]1954<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]1955</pre>
1956<h5>Overview:</h5>
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame]1957<p>The '<tt>fdiv</tt>' instruction returns the quotient of its two
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1958operands.</p>
1959<h5>Arguments:</h5>
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame]1960<p>The two arguments to the '<tt>div</tt>' instruction must be
1961<a href="#t_floating">floating point</a> values. Both arguments must have
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1962identical types. This instruction can also take <a href="#t_vector">vector</a>
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame]1963versions of the values in which case the elements must be floating point.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1964<h5>Semantics:</h5>
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame]1965<p>The value produced is the floating point quotient of the two operands.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1966<h5>Example:</h5>
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame]1967<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]1968</pre>
1969</div>
1970<!-- _______________________________________________________________________ -->
Reid Spencer7eb55b32006-11-02 01:53:59 +0000[diff] [blame]1971<div class="doc_subsubsection"> <a name="i_urem">'<tt>urem</tt>' Instruction</a>
1972</div>
1973<div class="doc_text">
1974<h5>Syntax:</h5>
1975<pre> &lt;result&gt; = urem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
1976</pre>
1977<h5>Overview:</h5>
1978<p>The '<tt>urem</tt>' instruction returns the remainder from the
1979unsigned division of its two arguments.</p>
1980<h5>Arguments:</h5>
1981<p>The two arguments to the '<tt>urem</tt>' instruction must be
1982<a href="#t_integer">integer</a> values. Both arguments must have identical
1983types.</p>
1984<h5>Semantics:</h5>
1985<p>This instruction returns the unsigned integer <i>remainder</i> of a division.
1986This instruction always performs an unsigned division to get the remainder,
1987regardless of whether the arguments are unsigned or not.</p>
1988<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1989<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]1990</pre>
1991
1992</div>
1993<!-- _______________________________________________________________________ -->
1994<div class="doc_subsubsection"> <a name="i_srem">'<tt>srem</tt>'
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1995Instruction</a> </div>
1996<div class="doc_text">
1997<h5>Syntax:</h5>
Reid Spencer7eb55b32006-11-02 01:53:59 +0000[diff] [blame]1998<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]1999</pre>
2000<h5>Overview:</h5>
Reid Spencer7eb55b32006-11-02 01:53:59 +0000[diff] [blame]2001<p>The '<tt>srem</tt>' instruction returns the remainder from the
2002signed division of its two operands.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2003<h5>Arguments:</h5>
Reid Spencer7eb55b32006-11-02 01:53:59 +0000[diff] [blame]2004<p>The two arguments to the '<tt>srem</tt>' instruction must be
2005<a href="#t_integer">integer</a> values. Both arguments must have identical
2006types.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2007<h5>Semantics:</h5>
Reid Spencer7eb55b32006-11-02 01:53:59 +0000[diff] [blame]2008<p>This instruction returns the <i>remainder</i> of a division (where the result
Reid Spencer806ad6a2007-03-24 22:23:39 +0000[diff] [blame]2009has the same sign as the dividend, <tt>var1</tt>), not the <i>modulo</i>
2010operator (where the result has the same sign as the divisor, <tt>var2</tt>) of
2011a value. For more information about the difference, see <a
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2012 href="http://mathforum.org/dr.math/problems/anne.4.28.99.html">The
Reid Spencer806ad6a2007-03-24 22:23:39 +0000[diff] [blame]2013Math Forum</a>. For a table of how this is implemented in various languages,
Reid Spencerdb3b93b2007-03-24 22:40:44 +0000[diff] [blame^]2014please see <a href="http://en.wikipedia.org/wiki/Modulo_operation">
Reid Spencer806ad6a2007-03-24 22:23:39 +0000[diff] [blame]2015Wikipedia: modulo operation</a>.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2016<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2017<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]2018</pre>
2019
2020</div>
2021<!-- _______________________________________________________________________ -->
2022<div class="doc_subsubsection"> <a name="i_frem">'<tt>frem</tt>'
2023Instruction</a> </div>
2024<div class="doc_text">
2025<h5>Syntax:</h5>
2026<pre> &lt;result&gt; = frem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2027</pre>
2028<h5>Overview:</h5>
2029<p>The '<tt>frem</tt>' instruction returns the remainder from the
2030division of its two operands.</p>
2031<h5>Arguments:</h5>
2032<p>The two arguments to the '<tt>frem</tt>' instruction must be
2033<a href="#t_floating">floating point</a> values. Both arguments must have
2034identical types.</p>
2035<h5>Semantics:</h5>
2036<p>This instruction returns the <i>remainder</i> of a division.</p>
2037<h5>Example:</h5>
2038<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]2039</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2040</div>
Robert Bocchino820bc75b2006-02-17 21:18:08 +0000[diff] [blame]2041
Reid Spencer2ab01932007-02-02 13:57:07 +0000[diff] [blame]2042<!-- ======================================================================= -->
2043<div class="doc_subsection"> <a name="bitwiseops">Bitwise Binary
2044Operations</a> </div>
2045<div class="doc_text">
2046<p>Bitwise binary operators are used to do various forms of
2047bit-twiddling in a program. They are generally very efficient
2048instructions and can commonly be strength reduced from other
2049instructions. They require two operands, execute an operation on them,
2050and produce a single value. The resulting value of the bitwise binary
2051operators is always the same type as its first operand.</p>
2052</div>
2053
Reid Spencer04e259b2007-01-31 21:39:12 +0000[diff] [blame]2054<!-- _______________________________________________________________________ -->
2055<div class="doc_subsubsection"> <a name="i_shl">'<tt>shl</tt>'
2056Instruction</a> </div>
2057<div class="doc_text">
2058<h5>Syntax:</h5>
2059<pre> &lt;result&gt; = shl &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2060</pre>
2061<h5>Overview:</h5>
2062<p>The '<tt>shl</tt>' instruction returns the first operand shifted to
2063the left a specified number of bits.</p>
2064<h5>Arguments:</h5>
2065<p>Both arguments to the '<tt>shl</tt>' instruction must be the same <a
2066 href="#t_integer">integer</a> type.</p>
2067<h5>Semantics:</h5>
2068<p>The value produced is <tt>var1</tt> * 2<sup><tt>var2</tt></sup>.</p>
2069<h5>Example:</h5><pre>
2070 &lt;result&gt; = shl i32 4, %var <i>; yields {i32}: 4 &lt;&lt; %var</i>
2071 &lt;result&gt; = shl i32 4, 2 <i>; yields {i32}: 16</i>
2072 &lt;result&gt; = shl i32 1, 10 <i>; yields {i32}: 1024</i>
2073</pre>
2074</div>
2075<!-- _______________________________________________________________________ -->
2076<div class="doc_subsubsection"> <a name="i_lshr">'<tt>lshr</tt>'
2077Instruction</a> </div>
2078<div class="doc_text">
2079<h5>Syntax:</h5>
2080<pre> &lt;result&gt; = lshr &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2081</pre>
2082
2083<h5>Overview:</h5>
2084<p>The '<tt>lshr</tt>' instruction (logical shift right) returns the first
2085operand shifted to the right a specified number of bits.</p>
2086
2087<h5>Arguments:</h5>
2088<p>Both arguments to the '<tt>lshr</tt>' instruction must be the same
2089<a href="#t_integer">integer</a> type.</p>
2090
2091<h5>Semantics:</h5>
2092<p>This instruction always performs a logical shift right operation. The most
2093significant bits of the result will be filled with zero bits after the
2094shift.</p>
2095
2096<h5>Example:</h5>
2097<pre>
2098 &lt;result&gt; = lshr i32 4, 1 <i>; yields {i32}:result = 2</i>
2099 &lt;result&gt; = lshr i32 4, 2 <i>; yields {i32}:result = 1</i>
2100 &lt;result&gt; = lshr i8 4, 3 <i>; yields {i8}:result = 0</i>
2101 &lt;result&gt; = lshr i8 -2, 1 <i>; yields {i8}:result = 0x7FFFFFFF </i>
2102</pre>
2103</div>
2104
Reid Spencer2ab01932007-02-02 13:57:07 +0000[diff] [blame]2105<!-- _______________________________________________________________________ -->
Reid Spencer04e259b2007-01-31 21:39:12 +0000[diff] [blame]2106<div class="doc_subsubsection"> <a name="i_ashr">'<tt>ashr</tt>'
2107Instruction</a> </div>
2108<div class="doc_text">
2109
2110<h5>Syntax:</h5>
2111<pre> &lt;result&gt; = ashr &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2112</pre>
2113
2114<h5>Overview:</h5>
2115<p>The '<tt>ashr</tt>' instruction (arithmetic shift right) returns the first
2116operand shifted to the right a specified number of bits.</p>
2117
2118<h5>Arguments:</h5>
2119<p>Both arguments to the '<tt>ashr</tt>' instruction must be the same
2120<a href="#t_integer">integer</a> type.</p>
2121
2122<h5>Semantics:</h5>
2123<p>This instruction always performs an arithmetic shift right operation,
2124The most significant bits of the result will be filled with the sign bit
2125of <tt>var1</tt>.</p>
2126
2127<h5>Example:</h5>
2128<pre>
2129 &lt;result&gt; = ashr i32 4, 1 <i>; yields {i32}:result = 2</i>
2130 &lt;result&gt; = ashr i32 4, 2 <i>; yields {i32}:result = 1</i>
2131 &lt;result&gt; = ashr i8 4, 3 <i>; yields {i8}:result = 0</i>
2132 &lt;result&gt; = ashr i8 -2, 1 <i>; yields {i8}:result = -1</i>
2133</pre>
2134</div>
2135
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2136<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2137<div class="doc_subsubsection"> <a name="i_and">'<tt>and</tt>'
2138Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2139<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2140<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2141<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]2142</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2143<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2144<p>The '<tt>and</tt>' instruction returns the bitwise logical and of
2145its two operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2146<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2147<p>The two arguments to the '<tt>and</tt>' instruction must be <a
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]2148 href="#t_integer">integer</a> values. Both arguments must have
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2149identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2150<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2151<p>The truth table used for the '<tt>and</tt>' instruction is:</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2152<p> </p>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]2153<div style="align: center">
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2154<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2155 <tbody>
2156 <tr>
2157 <td>In0</td>
2158 <td>In1</td>
2159 <td>Out</td>
2160 </tr>
2161 <tr>
2162 <td>0</td>
2163 <td>0</td>
2164 <td>0</td>
2165 </tr>
2166 <tr>
2167 <td>0</td>
2168 <td>1</td>
2169 <td>0</td>
2170 </tr>
2171 <tr>
2172 <td>1</td>
2173 <td>0</td>
2174 <td>0</td>
2175 </tr>
2176 <tr>
2177 <td>1</td>
2178 <td>1</td>
2179 <td>1</td>
2180 </tr>
2181 </tbody>
2182</table>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]2183</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2184<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2185<pre> &lt;result&gt; = and i32 4, %var <i>; yields {i32}:result = 4 &amp; %var</i>
2186 &lt;result&gt; = and i32 15, 40 <i>; yields {i32}:result = 8</i>
2187 &lt;result&gt; = and i32 4, 8 <i>; yields {i32}:result = 0</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2188</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2189</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2190<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2191<div class="doc_subsubsection"> <a name="i_or">'<tt>or</tt>' Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2192<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2193<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2194<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]2195</pre>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2196<h5>Overview:</h5>
2197<p>The '<tt>or</tt>' instruction returns the bitwise logical inclusive
2198or of its two operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2199<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2200<p>The two arguments to the '<tt>or</tt>' instruction must be <a
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]2201 href="#t_integer">integer</a> values. Both arguments must have
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2202identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2203<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2204<p>The truth table used for the '<tt>or</tt>' instruction is:</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2205<p> </p>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]2206<div style="align: center">
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2207<table border="1" cellspacing="0" cellpadding="4">
2208 <tbody>
2209 <tr>
2210 <td>In0</td>
2211 <td>In1</td>
2212 <td>Out</td>
2213 </tr>
2214 <tr>
2215 <td>0</td>
2216 <td>0</td>
2217 <td>0</td>
2218 </tr>
2219 <tr>
2220 <td>0</td>
2221 <td>1</td>
2222 <td>1</td>
2223 </tr>
2224 <tr>
2225 <td>1</td>
2226 <td>0</td>
2227 <td>1</td>
2228 </tr>
2229 <tr>
2230 <td>1</td>
2231 <td>1</td>
2232 <td>1</td>
2233 </tr>
2234 </tbody>
2235</table>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]2236</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2237<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2238<pre> &lt;result&gt; = or i32 4, %var <i>; yields {i32}:result = 4 | %var</i>
2239 &lt;result&gt; = or i32 15, 40 <i>; yields {i32}:result = 47</i>
2240 &lt;result&gt; = or i32 4, 8 <i>; yields {i32}:result = 12</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2241</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2242</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2243<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2244<div class="doc_subsubsection"> <a name="i_xor">'<tt>xor</tt>'
2245Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2246<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2247<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2248<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]2249</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2250<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2251<p>The '<tt>xor</tt>' instruction returns the bitwise logical exclusive
2252or of its two operands. The <tt>xor</tt> is used to implement the
2253"one's complement" operation, which is the "~" operator in C.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2254<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2255<p>The two arguments to the '<tt>xor</tt>' instruction must be <a
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]2256 href="#t_integer">integer</a> values. Both arguments must have
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2257identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2258<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2259<p>The truth table used for the '<tt>xor</tt>' instruction is:</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2260<p> </p>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]2261<div style="align: center">
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2262<table border="1" cellspacing="0" cellpadding="4">
2263 <tbody>
2264 <tr>
2265 <td>In0</td>
2266 <td>In1</td>
2267 <td>Out</td>
2268 </tr>
2269 <tr>
2270 <td>0</td>
2271 <td>0</td>
2272 <td>0</td>
2273 </tr>
2274 <tr>
2275 <td>0</td>
2276 <td>1</td>
2277 <td>1</td>
2278 </tr>
2279 <tr>
2280 <td>1</td>
2281 <td>0</td>
2282 <td>1</td>
2283 </tr>
2284 <tr>
2285 <td>1</td>
2286 <td>1</td>
2287 <td>0</td>
2288 </tr>
2289 </tbody>
2290</table>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]2291</div>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2292<p> </p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2293<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2294<pre> &lt;result&gt; = xor i32 4, %var <i>; yields {i32}:result = 4 ^ %var</i>
2295 &lt;result&gt; = xor i32 15, 40 <i>; yields {i32}:result = 39</i>
2296 &lt;result&gt; = xor i32 4, 8 <i>; yields {i32}:result = 12</i>
2297 &lt;result&gt; = xor i32 %V, -1 <i>; yields {i32}:result = ~%V</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2298</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2299</div>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2300
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2301<!-- ======================================================================= -->
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2302<div class="doc_subsection">
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2303 <a name="vectorops">Vector Operations</a>
2304</div>
2305
2306<div class="doc_text">
2307
2308<p>LLVM supports several instructions to represent vector operations in a
2309target-independent manner. This instructions cover the element-access and
2310vector-specific operations needed to process vectors effectively. While LLVM
2311does directly support these vector operations, many sophisticated algorithms
2312will want to use target-specific intrinsics to take full advantage of a specific
2313target.</p>
2314
2315</div>
2316
2317<!-- _______________________________________________________________________ -->
2318<div class="doc_subsubsection">
2319 <a name="i_extractelement">'<tt>extractelement</tt>' Instruction</a>
2320</div>
2321
2322<div class="doc_text">
2323
2324<h5>Syntax:</h5>
2325
2326<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2327 &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]2328</pre>
2329
2330<h5>Overview:</h5>
2331
2332<p>
2333The '<tt>extractelement</tt>' instruction extracts a single scalar
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]2334element from a vector at a specified index.
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2335</p>
2336
2337
2338<h5>Arguments:</h5>
2339
2340<p>
2341The first operand of an '<tt>extractelement</tt>' instruction is a
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]2342value of <a href="#t_vector">vector</a> type. The second operand is
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2343an index indicating the position from which to extract the element.
2344The index may be a variable.</p>
2345
2346<h5>Semantics:</h5>
2347
2348<p>
2349The result is a scalar of the same type as the element type of
2350<tt>val</tt>. Its value is the value at position <tt>idx</tt> of
2351<tt>val</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
2352results are undefined.
2353</p>
2354
2355<h5>Example:</h5>
2356
2357<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2358 %result = extractelement &lt;4 x i32&gt; %vec, i32 0 <i>; yields i32</i>
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2359</pre>
2360</div>
2361
2362
2363<!-- _______________________________________________________________________ -->
2364<div class="doc_subsubsection">
2365 <a name="i_insertelement">'<tt>insertelement</tt>' Instruction</a>
2366</div>
2367
2368<div class="doc_text">
2369
2370<h5>Syntax:</h5>
2371
2372<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2373 &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]2374</pre>
2375
2376<h5>Overview:</h5>
2377
2378<p>
2379The '<tt>insertelement</tt>' instruction inserts a scalar
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]2380element into a vector at a specified index.
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2381</p>
2382
2383
2384<h5>Arguments:</h5>
2385
2386<p>
2387The first operand of an '<tt>insertelement</tt>' instruction is a
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]2388value of <a href="#t_vector">vector</a> type. The second operand is a
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2389scalar value whose type must equal the element type of the first
2390operand. The third operand is an index indicating the position at
2391which to insert the value. The index may be a variable.</p>
2392
2393<h5>Semantics:</h5>
2394
2395<p>
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]2396The result is a vector of the same type as <tt>val</tt>. Its
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2397element values are those of <tt>val</tt> except at position
2398<tt>idx</tt>, where it gets the value <tt>elt</tt>. If <tt>idx</tt>
2399exceeds the length of <tt>val</tt>, the results are undefined.
2400</p>
2401
2402<h5>Example:</h5>
2403
2404<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2405 %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]2406</pre>
2407</div>
2408
2409<!-- _______________________________________________________________________ -->
2410<div class="doc_subsubsection">
2411 <a name="i_shufflevector">'<tt>shufflevector</tt>' Instruction</a>
2412</div>
2413
2414<div class="doc_text">
2415
2416<h5>Syntax:</h5>
2417
2418<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2419 &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]2420</pre>
2421
2422<h5>Overview:</h5>
2423
2424<p>
2425The '<tt>shufflevector</tt>' instruction constructs a permutation of elements
2426from two input vectors, returning a vector of the same type.
2427</p>
2428
2429<h5>Arguments:</h5>
2430
2431<p>
2432The first two operands of a '<tt>shufflevector</tt>' instruction are vectors
2433with types that match each other and types that match the result of the
2434instruction. The third argument is a shuffle mask, which has the same number
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2435of elements as the other vector type, but whose element type is always 'i32'.
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2436</p>
2437
2438<p>
2439The shuffle mask operand is required to be a constant vector with either
2440constant integer or undef values.
2441</p>
2442
2443<h5>Semantics:</h5>
2444
2445<p>
2446The elements of the two input vectors are numbered from left to right across
2447both of the vectors. The shuffle mask operand specifies, for each element of
2448the result vector, which element of the two input registers the result element
2449gets. The element selector may be undef (meaning "don't care") and the second
2450operand may be undef if performing a shuffle from only one vector.
2451</p>
2452
2453<h5>Example:</h5>
2454
2455<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2456 %result = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; %v2,
2457 &lt;4 x i32&gt; &lt;i32 0, i32 4, i32 1, i32 5&gt; <i>; yields &lt;4 x i32&gt;</i>
2458 %result = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; undef,
2459 &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]2460</pre>
2461</div>
2462
Tanya Lattnerb138bbe2006-04-14 19:24:33 +0000[diff] [blame]2463
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2464<!-- ======================================================================= -->
2465<div class="doc_subsection">
Chris Lattner6ab66722006-08-15 00:45:58 +0000[diff] [blame]2466 <a name="memoryops">Memory Access and Addressing Operations</a>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2467</div>
2468
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2469<div class="doc_text">
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2470
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2471<p>A key design point of an SSA-based representation is how it
2472represents memory. In LLVM, no memory locations are in SSA form, which
2473makes things very simple. This section describes how to read, write,
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]2474allocate, and free memory in LLVM.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2475
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2476</div>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2477
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2478<!-- _______________________________________________________________________ -->
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2479<div class="doc_subsubsection">
2480 <a name="i_malloc">'<tt>malloc</tt>' Instruction</a>
2481</div>
2482
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2483<div class="doc_text">
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2484
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2485<h5>Syntax:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2486
2487<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2488 &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]2489</pre>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2490
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2491<h5>Overview:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2492
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2493<p>The '<tt>malloc</tt>' instruction allocates memory from the system
2494heap and returns a pointer to it.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2495
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2496<h5>Arguments:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2497
2498<p>The '<tt>malloc</tt>' instruction allocates
2499<tt>sizeof(&lt;type&gt;)*NumElements</tt>
John Criswella92e5862004-02-24 16:13:56 +0000[diff] [blame]2500bytes of memory from the operating system and returns a pointer of the
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2501appropriate type to the program. If "NumElements" is specified, it is the
2502number of elements allocated. If an alignment is specified, the value result
2503of the allocation is guaranteed to be aligned to at least that boundary. If
2504not specified, or if zero, the target can choose to align the allocation on any
2505convenient boundary.</p>
2506
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2507<p>'<tt>type</tt>' must be a sized type.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2508
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2509<h5>Semantics:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2510
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2511<p>Memory is allocated using the system "<tt>malloc</tt>" function, and
2512a pointer is returned.</p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2513
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2514<h5>Example:</h5>
2515
2516<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2517 %array = malloc [4 x i8 ] <i>; yields {[%4 x i8]*}:array</i>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2518
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2519 %size = <a href="#i_add">add</a> i32 2, 2 <i>; yields {i32}:size = i32 4</i>
2520 %array1 = malloc i8, i32 4 <i>; yields {i8*}:array1</i>
2521 %array2 = malloc [12 x i8], i32 %size <i>; yields {[12 x i8]*}:array2</i>
2522 %array3 = malloc i32, i32 4, align 1024 <i>; yields {i32*}:array3</i>
2523 %array4 = malloc i32, align 1024 <i>; yields {i32*}:array4</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2524</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2525</div>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2526
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2527<!-- _______________________________________________________________________ -->
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2528<div class="doc_subsubsection">
2529 <a name="i_free">'<tt>free</tt>' Instruction</a>
2530</div>
2531
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2532<div class="doc_text">
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2533
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2534<h5>Syntax:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2535
2536<pre>
2537 free &lt;type&gt; &lt;value&gt; <i>; yields {void}</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2538</pre>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2539
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2540<h5>Overview:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2541
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2542<p>The '<tt>free</tt>' instruction returns memory back to the unused
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]2543memory heap to be reallocated in the future.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2544
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2545<h5>Arguments:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2546
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2547<p>'<tt>value</tt>' shall be a pointer value that points to a value
2548that was allocated with the '<tt><a href="#i_malloc">malloc</a></tt>'
2549instruction.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2550
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2551<h5>Semantics:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2552
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]2553<p>Access to the memory pointed to by the pointer is no longer defined
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2554after this instruction executes.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2555
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2556<h5>Example:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2557
2558<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2559 %array = <a href="#i_malloc">malloc</a> [4 x i8] <i>; yields {[4 x i8]*}:array</i>
2560 free [4 x i8]* %array
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2561</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2562</div>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2563
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2564<!-- _______________________________________________________________________ -->
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2565<div class="doc_subsubsection">
2566 <a name="i_alloca">'<tt>alloca</tt>' Instruction</a>
2567</div>
2568
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2569<div class="doc_text">
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2570
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2571<h5>Syntax:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2572
2573<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2574 &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]2575</pre>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2576
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2577<h5>Overview:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2578
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2579<p>The '<tt>alloca</tt>' instruction allocates memory on the current
2580stack frame of the procedure that is live until the current function
2581returns to its caller.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2582
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2583<h5>Arguments:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2584
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]2585<p>The '<tt>alloca</tt>' instruction allocates <tt>sizeof(&lt;type&gt;)*NumElements</tt>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2586bytes of memory on the runtime stack, returning a pointer of the
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2587appropriate type to the program. If "NumElements" is specified, it is the
2588number of elements allocated. If an alignment is specified, the value result
2589of the allocation is guaranteed to be aligned to at least that boundary. If
2590not specified, or if zero, the target can choose to align the allocation on any
2591convenient boundary.</p>
2592
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2593<p>'<tt>type</tt>' may be any sized type.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2594
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2595<h5>Semantics:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2596
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]2597<p>Memory is allocated; a pointer is returned. '<tt>alloca</tt>'d
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2598memory is automatically released when the function returns. The '<tt>alloca</tt>'
2599instruction is commonly used to represent automatic variables that must
2600have an address available. When the function returns (either with the <tt><a
John Criswellc932bef2005-05-12 16:55:34 +0000[diff] [blame]2601 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]2602instructions), the memory is reclaimed.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2603
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2604<h5>Example:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2605
2606<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2607 %ptr = alloca i32 <i>; yields {i32*}:ptr</i>
2608 %ptr = alloca i32, i32 4 <i>; yields {i32*}:ptr</i>
2609 %ptr = alloca i32, i32 4, align 1024 <i>; yields {i32*}:ptr</i>
2610 %ptr = alloca i32, align 1024 <i>; yields {i32*}:ptr</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2611</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2612</div>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2613
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2614<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2615<div class="doc_subsubsection"> <a name="i_load">'<tt>load</tt>'
2616Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2617<div class="doc_text">
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2618<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2619<pre> &lt;result&gt; = load &lt;ty&gt;* &lt;pointer&gt;<br> &lt;result&gt; = volatile load &lt;ty&gt;* &lt;pointer&gt;<br></pre>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2620<h5>Overview:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2621<p>The '<tt>load</tt>' instruction is used to read from memory.</p>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2622<h5>Arguments:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2623<p>The argument to the '<tt>load</tt>' instruction specifies the memory
John Criswell4c0cf7f2005-10-24 16:17:18 +0000[diff] [blame]2624address from which to load. The pointer must point to a <a
Chris Lattner10ee9652004-06-03 22:57:15 +0000[diff] [blame]2625 href="#t_firstclass">first class</a> type. If the <tt>load</tt> is
John Criswell4c0cf7f2005-10-24 16:17:18 +0000[diff] [blame]2626marked as <tt>volatile</tt>, then the optimizer is not allowed to modify
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2627the number or order of execution of this <tt>load</tt> with other
2628volatile <tt>load</tt> and <tt><a href="#i_store">store</a></tt>
2629instructions. </p>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2630<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2631<p>The location of memory pointed to is loaded.</p>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2632<h5>Examples:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2633<pre> %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2634 <a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2635 href="#i_store">store</a> i32 3, i32* %ptr <i>; yields {void}</i>
2636 %val = load i32* %ptr <i>; yields {i32}:val = i32 3</i>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2637</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2638</div>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2639<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2640<div class="doc_subsubsection"> <a name="i_store">'<tt>store</tt>'
2641Instruction</a> </div>
Reid Spencera89fb182006-11-09 21:18:01 +0000[diff] [blame]2642<div class="doc_text">
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2643<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2644<pre> store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt; <i>; yields {void}</i>
Chris Lattner12d456c2003-09-08 18:27:49 +0000[diff] [blame]2645 volatile store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt; <i>; yields {void}</i>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2646</pre>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2647<h5>Overview:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2648<p>The '<tt>store</tt>' instruction is used to write to memory.</p>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2649<h5>Arguments:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2650<p>There are two arguments to the '<tt>store</tt>' instruction: a value
John Criswell4c0cf7f2005-10-24 16:17:18 +0000[diff] [blame]2651to store and an address in which to store it. The type of the '<tt>&lt;pointer&gt;</tt>'
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2652operand must be a pointer to the type of the '<tt>&lt;value&gt;</tt>'
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]2653operand. If the <tt>store</tt> is marked as <tt>volatile</tt>, then the
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2654optimizer is not allowed to modify the number or order of execution of
2655this <tt>store</tt> with other volatile <tt>load</tt> and <tt><a
2656 href="#i_store">store</a></tt> instructions.</p>
2657<h5>Semantics:</h5>
2658<p>The contents of memory are updated to contain '<tt>&lt;value&gt;</tt>'
2659at the location specified by the '<tt>&lt;pointer&gt;</tt>' operand.</p>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2660<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2661<pre> %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2662 <a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2663 href="#i_store">store</a> i32 3, i32* %ptr <i>; yields {void}</i>
2664 %val = load i32* %ptr <i>; yields {i32}:val = i32 3</i>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2665</pre>
Reid Spencer443460a2006-11-09 21:15:49 +0000[diff] [blame]2666</div>
2667
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2668<!-- _______________________________________________________________________ -->
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2669<div class="doc_subsubsection">
2670 <a name="i_getelementptr">'<tt>getelementptr</tt>' Instruction</a>
2671</div>
2672
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2673<div class="doc_text">
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]2674<h5>Syntax:</h5>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2675<pre>
2676 &lt;result&gt; = getelementptr &lt;ty&gt;* &lt;ptrval&gt;{, &lt;ty&gt; &lt;idx&gt;}*
2677</pre>
2678
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]2679<h5>Overview:</h5>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2680
2681<p>
2682The '<tt>getelementptr</tt>' instruction is used to get the address of a
2683subelement of an aggregate data structure.</p>
2684
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]2685<h5>Arguments:</h5>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2686
Reid Spencercee005c2006-12-04 21:29:24 +0000[diff] [blame]2687<p>This instruction takes a list of integer operands that indicate what
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2688elements of the aggregate object to index to. The actual types of the arguments
2689provided depend on the type of the first pointer argument. The
2690'<tt>getelementptr</tt>' instruction is used to index down through the type
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]2691levels 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]2692structure, only <tt>i32</tt> integer constants are allowed. When indexing
Reid Spencercee005c2006-12-04 21:29:24 +0000[diff] [blame]2693into an array or pointer, only integers of 32 or 64 bits are allowed, and will
2694be sign extended to 64-bit values.</p>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2695
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2696<p>For example, let's consider a C code fragment and how it gets
2697compiled to LLVM:</p>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2698
2699<pre>
2700 struct RT {
2701 char A;
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2702 i32 B[10][20];
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2703 char C;
2704 };
2705 struct ST {
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2706 i32 X;
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2707 double Y;
2708 struct RT Z;
2709 };
2710
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2711 define i32 *foo(struct ST *s) {
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2712 return &amp;s[1].Z.B[5][13];
2713 }
2714</pre>
2715
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2716<p>The LLVM code generated by the GCC frontend is:</p>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2717
2718<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2719 %RT = type { i8 , [10 x [20 x i32]], i8 }
2720 %ST = type { i32, double, %RT }
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2721
Brian Gaeke317ef962004-07-02 21:08:14 +0000[diff] [blame]2722 implementation
2723
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2724 define i32* %foo(%ST* %s) {
Brian Gaeke317ef962004-07-02 21:08:14 +0000[diff] [blame]2725 entry:
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2726 %reg = getelementptr %ST* %s, i32 1, i32 2, i32 1, i32 5, i32 13
2727 ret i32* %reg
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2728 }
2729</pre>
2730
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]2731<h5>Semantics:</h5>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2732
2733<p>The index types specified for the '<tt>getelementptr</tt>' instruction depend
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]2734on the pointer type that is being indexed into. <a href="#t_pointer">Pointer</a>
Reid Spencercee005c2006-12-04 21:29:24 +0000[diff] [blame]2735and <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]2736<a href="#t_integer">integer</a> type but the value will always be sign extended
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2737to 64-bits. <a href="#t_struct">Structure</a> types, require <tt>i32</tt>
Reid Spencerc0312692006-12-03 16:53:48 +0000[diff] [blame]2738<b>constants</b>.</p>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2739
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2740<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]2741type, which is a pointer, yielding a '<tt>%ST</tt>' = '<tt>{ i32, double, %RT
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2742}</tt>' type, a structure. The second index indexes into the third element of
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2743the structure, yielding a '<tt>%RT</tt>' = '<tt>{ i8 , [10 x [20 x i32]],
2744i8 }</tt>' type, another structure. The third index indexes into the second
2745element of the structure, yielding a '<tt>[10 x [20 x i32]]</tt>' type, an
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2746array. The two dimensions of the array are subscripted into, yielding an
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2747'<tt>i32</tt>' type. The '<tt>getelementptr</tt>' instruction returns a pointer
2748to this element, thus computing a value of '<tt>i32*</tt>' type.</p>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2749
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2750<p>Note that it is perfectly legal to index partially through a
2751structure, returning a pointer to an inner element. Because of this,
2752the LLVM code for the given testcase is equivalent to:</p>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2753
2754<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2755 define i32* %foo(%ST* %s) {
2756 %t1 = getelementptr %ST* %s, i32 1 <i>; yields %ST*:%t1</i>
2757 %t2 = getelementptr %ST* %t1, i32 0, i32 2 <i>; yields %RT*:%t2</i>
2758 %t3 = getelementptr %RT* %t2, i32 0, i32 1 <i>; yields [10 x [20 x i32]]*:%t3</i>
2759 %t4 = getelementptr [10 x [20 x i32]]* %t3, i32 0, i32 5 <i>; yields [20 x i32]*:%t4</i>
2760 %t5 = getelementptr [20 x i32]* %t4, i32 0, i32 13 <i>; yields i32*:%t5</i>
2761 ret i32* %t5
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2762 }
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]2763</pre>
Chris Lattnerc0ad71e2005-06-24 17:22:57 +0000[diff] [blame]2764
2765<p>Note that it is undefined to access an array out of bounds: array and
2766pointer indexes must always be within the defined bounds of the array type.
2767The one exception for this rules is zero length arrays. These arrays are
2768defined to be accessible as variable length arrays, which requires access
2769beyond the zero'th element.</p>
2770
Chris Lattner6ab66722006-08-15 00:45:58 +0000[diff] [blame]2771<p>The getelementptr instruction is often confusing. For some more insight
2772into how it works, see <a href="GetElementPtr.html">the getelementptr
2773FAQ</a>.</p>
2774
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]2775<h5>Example:</h5>
Chris Lattnerc0ad71e2005-06-24 17:22:57 +0000[diff] [blame]2776
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2777<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2778 <i>; yields [12 x i8]*:aptr</i>
2779 %aptr = getelementptr {i32, [12 x i8]}* %sptr, i64 0, i32 1
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2780</pre>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2781</div>
Reid Spencer443460a2006-11-09 21:15:49 +0000[diff] [blame]2782
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2783<!-- ======================================================================= -->
Reid Spencer97c5fa42006-11-08 01:18:52 +0000[diff] [blame]2784<div class="doc_subsection"> <a name="convertops">Conversion Operations</a>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2785</div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2786<div class="doc_text">
Reid Spencer97c5fa42006-11-08 01:18:52 +0000[diff] [blame]2787<p>The instructions in this category are the conversion instructions (casting)
2788which all take a single operand and a type. They perform various bit conversions
2789on the operand.</p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2790</div>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]2791
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]2792<!-- _______________________________________________________________________ -->
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]2793<div class="doc_subsubsection">
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2794 <a name="i_trunc">'<tt>trunc .. to</tt>' Instruction</a>
2795</div>
2796<div class="doc_text">
2797
2798<h5>Syntax:</h5>
2799<pre>
2800 &lt;result&gt; = trunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2801</pre>
2802
2803<h5>Overview:</h5>
2804<p>
2805The '<tt>trunc</tt>' instruction truncates its operand to the type <tt>ty2</tt>.
2806</p>
2807
2808<h5>Arguments:</h5>
2809<p>
2810The '<tt>trunc</tt>' instruction takes a <tt>value</tt> to trunc, which must
2811be 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]2812and type of the result, which must be an <a href="#t_integer">integer</a>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]2813type. The bit size of <tt>value</tt> must be larger than the bit size of
2814<tt>ty2</tt>. Equal sized types are not allowed.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2815
2816<h5>Semantics:</h5>
2817<p>
2818The '<tt>trunc</tt>' instruction truncates the high order bits in <tt>value</tt>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]2819and converts the remaining bits to <tt>ty2</tt>. Since the source size must be
2820larger than the destination size, <tt>trunc</tt> cannot be a <i>no-op cast</i>.
2821It will always truncate bits.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2822
2823<h5>Example:</h5>
2824<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2825 %X = trunc i32 257 to i8 <i>; yields i8:1</i>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]2826 %Y = trunc i32 123 to i1 <i>; yields i1:true</i>
2827 %Y = trunc i32 122 to i1 <i>; yields i1:false</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2828</pre>
2829</div>
2830
2831<!-- _______________________________________________________________________ -->
2832<div class="doc_subsubsection">
2833 <a name="i_zext">'<tt>zext .. to</tt>' Instruction</a>
2834</div>
2835<div class="doc_text">
2836
2837<h5>Syntax:</h5>
2838<pre>
2839 &lt;result&gt; = zext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2840</pre>
2841
2842<h5>Overview:</h5>
2843<p>The '<tt>zext</tt>' instruction zero extends its operand to type
2844<tt>ty2</tt>.</p>
2845
2846
2847<h5>Arguments:</h5>
2848<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]2849<a href="#t_integer">integer</a> type, and a type to cast it to, which must
2850also be of <a href="#t_integer">integer</a> type. The bit size of the
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]2851<tt>value</tt> must be smaller than the bit size of the destination type,
2852<tt>ty2</tt>.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2853
2854<h5>Semantics:</h5>
2855<p>The <tt>zext</tt> fills the high order bits of the <tt>value</tt> with zero
2856bits until it reaches the size of the destination type, <tt>ty2</tt>. When the
2857the operand and the type are the same size, no bit filling is done and the
2858cast is considered a <i>no-op cast</i> because no bits change (only the type
2859changes).</p>
2860
Reid Spencer07c9c682007-01-12 15:46:11 +0000[diff] [blame]2861<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]2862
2863<h5>Example:</h5>
2864<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2865 %X = zext i32 257 to i64 <i>; yields i64:257</i>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]2866 %Y = zext i1 true to i32 <i>; yields i32:1</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2867</pre>
2868</div>
2869
2870<!-- _______________________________________________________________________ -->
2871<div class="doc_subsubsection">
2872 <a name="i_sext">'<tt>sext .. to</tt>' Instruction</a>
2873</div>
2874<div class="doc_text">
2875
2876<h5>Syntax:</h5>
2877<pre>
2878 &lt;result&gt; = sext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2879</pre>
2880
2881<h5>Overview:</h5>
2882<p>The '<tt>sext</tt>' sign extends <tt>value</tt> to the type <tt>ty2</tt>.</p>
2883
2884<h5>Arguments:</h5>
2885<p>
2886The '<tt>sext</tt>' instruction takes a value to cast, which must be of
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]2887<a href="#t_integer">integer</a> type, and a type to cast it to, which must
2888also be of <a href="#t_integer">integer</a> type. The bit size of the
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]2889<tt>value</tt> must be smaller than the bit size of the destination type,
2890<tt>ty2</tt>.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2891
2892<h5>Semantics:</h5>
2893<p>
2894The '<tt>sext</tt>' instruction performs a sign extension by copying the sign
2895bit (highest order bit) of the <tt>value</tt> until it reaches the bit size of
2896the type <tt>ty2</tt>. When the the operand and the type are the same size,
2897no bit filling is done and the cast is considered a <i>no-op cast</i> because
2898no bits change (only the type changes).</p>
2899
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]2900<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]2901
2902<h5>Example:</h5>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2903<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2904 %X = sext i8 -1 to i16 <i>; yields i16 :65535</i>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]2905 %Y = sext i1 true to i32 <i>; yields i32:-1</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2906</pre>
2907</div>
2908
2909<!-- _______________________________________________________________________ -->
2910<div class="doc_subsubsection">
Reid Spencer2e2740d2006-11-09 21:48:10 +0000[diff] [blame]2911 <a name="i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a>
2912</div>
2913
2914<div class="doc_text">
2915
2916<h5>Syntax:</h5>
2917
2918<pre>
2919 &lt;result&gt; = fptrunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2920</pre>
2921
2922<h5>Overview:</h5>
2923<p>The '<tt>fptrunc</tt>' instruction truncates <tt>value</tt> to type
2924<tt>ty2</tt>.</p>
2925
2926
2927<h5>Arguments:</h5>
2928<p>The '<tt>fptrunc</tt>' instruction takes a <a href="#t_floating">floating
2929 point</a> value to cast and a <a href="#t_floating">floating point</a> type to
2930cast it to. The size of <tt>value</tt> must be larger than the size of
2931<tt>ty2</tt>. This implies that <tt>fptrunc</tt> cannot be used to make a
2932<i>no-op cast</i>.</p>
2933
2934<h5>Semantics:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]2935<p> The '<tt>fptrunc</tt>' instruction truncates a <tt>value</tt> from a larger
2936<a href="#t_floating">floating point</a> type to a smaller
2937<a href="#t_floating">floating point</a> type. If the value cannot fit within
2938the destination type, <tt>ty2</tt>, then the results are undefined.</p>
Reid Spencer2e2740d2006-11-09 21:48:10 +0000[diff] [blame]2939
2940<h5>Example:</h5>
2941<pre>
2942 %X = fptrunc double 123.0 to float <i>; yields float:123.0</i>
2943 %Y = fptrunc double 1.0E+300 to float <i>; yields undefined</i>
2944</pre>
2945</div>
2946
2947<!-- _______________________________________________________________________ -->
2948<div class="doc_subsubsection">
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2949 <a name="i_fpext">'<tt>fpext .. to</tt>' Instruction</a>
2950</div>
2951<div class="doc_text">
2952
2953<h5>Syntax:</h5>
2954<pre>
2955 &lt;result&gt; = fpext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2956</pre>
2957
2958<h5>Overview:</h5>
2959<p>The '<tt>fpext</tt>' extends a floating point <tt>value</tt> to a larger
2960floating point value.</p>
2961
2962<h5>Arguments:</h5>
2963<p>The '<tt>fpext</tt>' instruction takes a
2964<a href="#t_floating">floating point</a> <tt>value</tt> to cast,
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]2965and a <a href="#t_floating">floating point</a> type to cast it to. The source
2966type must be smaller than the destination type.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2967
2968<h5>Semantics:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]2969<p>The '<tt>fpext</tt>' instruction extends the <tt>value</tt> from a smaller
2970<a href="t_floating">floating point</a> type to a larger
2971<a href="t_floating">floating point</a> type. The <tt>fpext</tt> cannot be
2972used to make a <i>no-op cast</i> because it always changes bits. Use
Reid Spencer5b950642006-11-11 23:08:07 +0000[diff] [blame]2973<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]2974
2975<h5>Example:</h5>
2976<pre>
2977 %X = fpext float 3.1415 to double <i>; yields double:3.1415</i>
2978 %Y = fpext float 1.0 to float <i>; yields float:1.0 (no-op)</i>
2979</pre>
2980</div>
2981
2982<!-- _______________________________________________________________________ -->
2983<div class="doc_subsubsection">
Reid Spencer2eadb532007-01-21 00:29:26 +0000[diff] [blame]2984 <a name="i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2985</div>
2986<div class="doc_text">
2987
2988<h5>Syntax:</h5>
2989<pre>
2990 &lt;result&gt; = fp2uint &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2991</pre>
2992
2993<h5>Overview:</h5>
2994<p>The '<tt>fp2uint</tt>' converts a floating point <tt>value</tt> to its
2995unsigned integer equivalent of type <tt>ty2</tt>.
2996</p>
2997
2998<h5>Arguments:</h5>
2999<p>The '<tt>fp2uint</tt>' instruction takes a value to cast, which must be a
3000<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]3001must be an <a href="#t_integer">integer</a> type.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3002
3003<h5>Semantics:</h5>
3004<p> The '<tt>fp2uint</tt>' instruction converts its
3005<a href="#t_floating">floating point</a> operand into the nearest (rounding
3006towards zero) unsigned integer value. If the value cannot fit in <tt>ty2</tt>,
3007the results are undefined.</p>
3008
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]3009<p>When converting to i1, the conversion is done as a comparison against
3010zero. If the <tt>value</tt> was zero, the i1 result will be <tt>false</tt>.
3011If 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]3012
3013<h5>Example:</h5>
3014<pre>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]3015 %X = fp2uint double 123.0 to i32 <i>; yields i32:123</i>
3016 %Y = fp2uint float 1.0E+300 to i1 <i>; yields i1:true</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3017 %X = fp2uint float 1.04E+17 to i8 <i>; yields undefined:1</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3018</pre>
3019</div>
3020
3021<!-- _______________________________________________________________________ -->
3022<div class="doc_subsubsection">
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3023 <a name="i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3024</div>
3025<div class="doc_text">
3026
3027<h5>Syntax:</h5>
3028<pre>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3029 &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]3030</pre>
3031
3032<h5>Overview:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3033<p>The '<tt>fptosi</tt>' instruction converts
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3034<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]3035</p>
3036
3037
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]3038<h5>Arguments:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3039<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]3040<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]3041must also be an <a href="#t_integer">integer</a> type.</p>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]3042
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]3043<h5>Semantics:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3044<p>The '<tt>fptosi</tt>' instruction converts its
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3045<a href="#t_floating">floating point</a> operand into the nearest (rounding
3046towards zero) signed integer value. If the value cannot fit in <tt>ty2</tt>,
3047the results are undefined.</p>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]3048
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]3049<p>When converting to i1, the conversion is done as a comparison against
3050zero. If the <tt>value</tt> was zero, the i1 result will be <tt>false</tt>.
3051If 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]3052
Chris Lattner70de6632001-07-09 00:26:23 +0000[diff] [blame]3053<h5>Example:</h5>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]3054<pre>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]3055 %X = fptosi double -123.0 to i32 <i>; yields i32:-123</i>
3056 %Y = fptosi float 1.0E-247 to i1 <i>; yields i1:true</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3057 %X = fptosi float 1.04E+17 to i8 <i>; yields undefined:1</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3058</pre>
3059</div>
3060
3061<!-- _______________________________________________________________________ -->
3062<div class="doc_subsubsection">
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3063 <a name="i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3064</div>
3065<div class="doc_text">
3066
3067<h5>Syntax:</h5>
3068<pre>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3069 &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]3070</pre>
3071
3072<h5>Overview:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3073<p>The '<tt>uitofp</tt>' instruction regards <tt>value</tt> as an unsigned
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3074integer and converts that value to the <tt>ty2</tt> type.</p>
3075
3076
3077<h5>Arguments:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3078<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]3079<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]3080be a <a href="#t_floating">floating point</a> type.</p>
3081
3082<h5>Semantics:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3083<p>The '<tt>uitofp</tt>' instruction interprets its operand as an unsigned
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3084integer quantity and converts it to the corresponding floating point value. If
3085the value cannot fit in the floating point value, the results are undefined.</p>
3086
3087
3088<h5>Example:</h5>
3089<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3090 %X = uitofp i32 257 to float <i>; yields float:257.0</i>
3091 %Y = uitofp i8 -1 to double <i>; yields double:255.0</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3092</pre>
3093</div>
3094
3095<!-- _______________________________________________________________________ -->
3096<div class="doc_subsubsection">
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3097 <a name="i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3098</div>
3099<div class="doc_text">
3100
3101<h5>Syntax:</h5>
3102<pre>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3103 &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]3104</pre>
3105
3106<h5>Overview:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3107<p>The '<tt>sitofp</tt>' instruction regards <tt>value</tt> as a signed
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3108integer and converts that value to the <tt>ty2</tt> type.</p>
3109
3110<h5>Arguments:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3111<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]3112<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]3113a <a href="#t_floating">floating point</a> type.</p>
3114
3115<h5>Semantics:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3116<p>The '<tt>sitofp</tt>' instruction interprets its operand as a signed
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3117integer quantity and converts it to the corresponding floating point value. If
3118the value cannot fit in the floating point value, the results are undefined.</p>
3119
3120<h5>Example:</h5>
3121<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3122 %X = sitofp i32 257 to float <i>; yields float:257.0</i>
3123 %Y = sitofp i8 -1 to double <i>; yields double:-1.0</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3124</pre>
3125</div>
3126
3127<!-- _______________________________________________________________________ -->
3128<div class="doc_subsubsection">
Reid Spencerb7344ff2006-11-11 21:00:47 +0000[diff] [blame]3129 <a name="i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a>
3130</div>
3131<div class="doc_text">
3132
3133<h5>Syntax:</h5>
3134<pre>
3135 &lt;result&gt; = ptrtoint &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3136</pre>
3137
3138<h5>Overview:</h5>
3139<p>The '<tt>ptrtoint</tt>' instruction converts the pointer <tt>value</tt> to
3140the integer type <tt>ty2</tt>.</p>
3141
3142<h5>Arguments:</h5>
3143<p>The '<tt>ptrtoint</tt>' instruction takes a <tt>value</tt> to cast, which
3144must be a <a href="t_pointer">pointer</a> value, and a type to cast it to
3145<tt>ty2</tt>, which must be an <a href="#t_integer">integer</a> type.
3146
3147<h5>Semantics:</h5>
3148<p>The '<tt>ptrtoint</tt>' instruction converts <tt>value</tt> to integer type
3149<tt>ty2</tt> by interpreting the pointer value as an integer and either
3150truncating or zero extending that value to the size of the integer type. If
3151<tt>value</tt> is smaller than <tt>ty2</tt> then a zero extension is done. If
3152<tt>value</tt> is larger than <tt>ty2</tt> then a truncation is done. If they
3153are the same size, then nothing is done (<i>no-op cast</i>).</p>
3154
3155<h5>Example:</h5>
3156<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3157 %X = ptrtoint i32* %X to i8 <i>; yields truncation on 32-bit</i>
3158 %Y = ptrtoint i32* %x to i64 <i>; yields zero extend on 32-bit</i>
Reid Spencerb7344ff2006-11-11 21:00:47 +0000[diff] [blame]3159</pre>
3160</div>
3161
3162<!-- _______________________________________________________________________ -->
3163<div class="doc_subsubsection">
3164 <a name="i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a>
3165</div>
3166<div class="doc_text">
3167
3168<h5>Syntax:</h5>
3169<pre>
3170 &lt;result&gt; = inttoptr &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3171</pre>
3172
3173<h5>Overview:</h5>
3174<p>The '<tt>inttoptr</tt>' instruction converts an integer <tt>value</tt> to
3175a pointer type, <tt>ty2</tt>.</p>
3176
3177<h5>Arguments:</h5>
3178<p>The '<tt>inttoptr</tt>' instruction takes an <a href="i_integer">integer</a>
3179value to cast, and a type to cast it to, which must be a
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]3180<a href="#t_pointer">pointer</a> type.
Reid Spencerb7344ff2006-11-11 21:00:47 +0000[diff] [blame]3181
3182<h5>Semantics:</h5>
3183<p>The '<tt>inttoptr</tt>' instruction converts <tt>value</tt> to type
3184<tt>ty2</tt> by applying either a zero extension or a truncation depending on
3185the size of the integer <tt>value</tt>. If <tt>value</tt> is larger than the
3186size of a pointer then a truncation is done. If <tt>value</tt> is smaller than
3187the size of a pointer then a zero extension is done. If they are the same size,
3188nothing is done (<i>no-op cast</i>).</p>
3189
3190<h5>Example:</h5>
3191<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3192 %X = inttoptr i32 255 to i32* <i>; yields zero extend on 64-bit</i>
3193 %X = inttoptr i32 255 to i32* <i>; yields no-op on 32-bit </i>
3194 %Y = inttoptr i16 0 to i32* <i>; yields zero extend on 32-bit</i>
Reid Spencerb7344ff2006-11-11 21:00:47 +0000[diff] [blame]3195</pre>
3196</div>
3197
3198<!-- _______________________________________________________________________ -->
3199<div class="doc_subsubsection">
Reid Spencer5b950642006-11-11 23:08:07 +0000[diff] [blame]3200 <a name="i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3201</div>
3202<div class="doc_text">
3203
3204<h5>Syntax:</h5>
3205<pre>
Reid Spencer5b950642006-11-11 23:08:07 +0000[diff] [blame]3206 &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]3207</pre>
3208
3209<h5>Overview:</h5>
Reid Spencer5b950642006-11-11 23:08:07 +0000[diff] [blame]3210<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3211<tt>ty2</tt> without changing any bits.</p>
3212
3213<h5>Arguments:</h5>
Reid Spencer5b950642006-11-11 23:08:07 +0000[diff] [blame]3214<p>The '<tt>bitcast</tt>' instruction takes a value to cast, which must be
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3215a first class value, and a type to cast it to, which must also be a <a
3216 href="#t_firstclass">first class</a> type. The bit sizes of <tt>value</tt>
Reid Spencere3db84c2007-01-09 20:08:58 +0000[diff] [blame]3217and the destination type, <tt>ty2</tt>, must be identical. If the source
3218type is a pointer, the destination type must also be a pointer.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3219
3220<h5>Semantics:</h5>
Reid Spencer5b950642006-11-11 23:08:07 +0000[diff] [blame]3221<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
Reid Spencerb7344ff2006-11-11 21:00:47 +0000[diff] [blame]3222<tt>ty2</tt>. It is always a <i>no-op cast</i> because no bits change with
3223this conversion. The conversion is done as if the <tt>value</tt> had been
3224stored to memory and read back as type <tt>ty2</tt>. Pointer types may only be
3225converted to other pointer types with this instruction. To convert pointers to
3226other types, use the <a href="#i_inttoptr">inttoptr</a> or
3227<a href="#i_ptrtoint">ptrtoint</a> instructions first.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3228
3229<h5>Example:</h5>
3230<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3231 %X = bitcast i8 255 to i8 <i>; yields i8 :-1</i>
3232 %Y = bitcast i32* %x to sint* <i>; yields sint*:%x</i>
3233 %Z = bitcast <2xint> %V to i64; <i>; yields i64: %V</i>
Chris Lattner70de6632001-07-09 00:26:23 +0000[diff] [blame]3234</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3235</div>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]3236
Reid Spencer97c5fa42006-11-08 01:18:52 +0000[diff] [blame]3237<!-- ======================================================================= -->
3238<div class="doc_subsection"> <a name="otherops">Other Operations</a> </div>
3239<div class="doc_text">
3240<p>The instructions in this category are the "miscellaneous"
3241instructions, which defy better classification.</p>
3242</div>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3243
3244<!-- _______________________________________________________________________ -->
3245<div class="doc_subsubsection"><a name="i_icmp">'<tt>icmp</tt>' Instruction</a>
3246</div>
3247<div class="doc_text">
3248<h5>Syntax:</h5>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]3249<pre> &lt;result&gt; = icmp &lt;cond&gt; &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt;
3250<i>; yields {i1}:result</i>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3251</pre>
3252<h5>Overview:</h5>
3253<p>The '<tt>icmp</tt>' instruction returns a boolean value based on comparison
3254of its two integer operands.</p>
3255<h5>Arguments:</h5>
3256<p>The '<tt>icmp</tt>' instruction takes three operands. The first operand is
3257the condition code which indicates the kind of comparison to perform. It is not
3258a value, just a keyword. The possibilities for the condition code are:
3259<ol>
3260 <li><tt>eq</tt>: equal</li>
3261 <li><tt>ne</tt>: not equal </li>
3262 <li><tt>ugt</tt>: unsigned greater than</li>
3263 <li><tt>uge</tt>: unsigned greater or equal</li>
3264 <li><tt>ult</tt>: unsigned less than</li>
3265 <li><tt>ule</tt>: unsigned less or equal</li>
3266 <li><tt>sgt</tt>: signed greater than</li>
3267 <li><tt>sge</tt>: signed greater or equal</li>
3268 <li><tt>slt</tt>: signed less than</li>
3269 <li><tt>sle</tt>: signed less or equal</li>
3270</ol>
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]3271<p>The remaining two arguments must be <a href="#t_integer">integer</a> or
Reid Spencer784ef792007-01-04 05:19:58 +0000[diff] [blame]3272<a href="#t_pointer">pointer</a> typed. They must also be identical types.</p>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3273<h5>Semantics:</h5>
3274<p>The '<tt>icmp</tt>' compares <tt>var1</tt> and <tt>var2</tt> according to
3275the condition code given as <tt>cond</tt>. The comparison performed always
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]3276yields a <a href="#t_primitive">i1</a> result, as follows:
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3277<ol>
3278 <li><tt>eq</tt>: yields <tt>true</tt> if the operands are equal,
3279 <tt>false</tt> otherwise. No sign interpretation is necessary or performed.
3280 </li>
3281 <li><tt>ne</tt>: yields <tt>true</tt> if the operands are unequal,
3282 <tt>false</tt> otherwise. No sign interpretation is necessary or performed.
3283 <li><tt>ugt</tt>: interprets the operands as unsigned values and yields
3284 <tt>true</tt> if <tt>var1</tt> is greater than <tt>var2</tt>.</li>
3285 <li><tt>uge</tt>: interprets the operands as unsigned values and yields
3286 <tt>true</tt> if <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
3287 <li><tt>ult</tt>: interprets the operands as unsigned values and yields
3288 <tt>true</tt> if <tt>var1</tt> is less than <tt>var2</tt>.</li>
3289 <li><tt>ule</tt>: interprets the operands as unsigned values and yields
3290 <tt>true</tt> if <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
3291 <li><tt>sgt</tt>: interprets the operands as signed values and yields
3292 <tt>true</tt> if <tt>var1</tt> is greater than <tt>var2</tt>.</li>
3293 <li><tt>sge</tt>: interprets the operands as signed values and yields
3294 <tt>true</tt> if <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
3295 <li><tt>slt</tt>: interprets the operands as signed values and yields
3296 <tt>true</tt> if <tt>var1</tt> is less than <tt>var2</tt>.</li>
3297 <li><tt>sle</tt>: interprets the operands as signed values and yields
3298 <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]3299</ol>
3300<p>If the operands are <a href="#t_pointer">pointer</a> typed, the pointer
3301values are treated as integers and then compared.</p>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3302
3303<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3304<pre> &lt;result&gt; = icmp eq i32 4, 5 <i>; yields: result=false</i>
3305 &lt;result&gt; = icmp ne float* %X, %X <i>; yields: result=false</i>
3306 &lt;result&gt; = icmp ult i16 4, 5 <i>; yields: result=true</i>
3307 &lt;result&gt; = icmp sgt i16 4, 5 <i>; yields: result=false</i>
3308 &lt;result&gt; = icmp ule i16 -4, 5 <i>; yields: result=false</i>
3309 &lt;result&gt; = icmp sge i16 4, 5 <i>; yields: result=false</i>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3310</pre>
3311</div>
3312
3313<!-- _______________________________________________________________________ -->
3314<div class="doc_subsubsection"><a name="i_fcmp">'<tt>fcmp</tt>' Instruction</a>
3315</div>
3316<div class="doc_text">
3317<h5>Syntax:</h5>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]3318<pre> &lt;result&gt; = fcmp &lt;cond&gt; &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt;
3319<i>; yields {i1}:result</i>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3320</pre>
3321<h5>Overview:</h5>
3322<p>The '<tt>fcmp</tt>' instruction returns a boolean value based on comparison
3323of its floating point operands.</p>
3324<h5>Arguments:</h5>
3325<p>The '<tt>fcmp</tt>' instruction takes three operands. The first operand is
3326the condition code which indicates the kind of comparison to perform. It is not
3327a value, just a keyword. The possibilities for the condition code are:
3328<ol>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3329 <li><tt>false</tt>: no comparison, always returns false</li>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3330 <li><tt>oeq</tt>: ordered and equal</li>
3331 <li><tt>ogt</tt>: ordered and greater than </li>
3332 <li><tt>oge</tt>: ordered and greater than or equal</li>
3333 <li><tt>olt</tt>: ordered and less than </li>
3334 <li><tt>ole</tt>: ordered and less than or equal</li>
3335 <li><tt>one</tt>: ordered and not equal</li>
3336 <li><tt>ord</tt>: ordered (no nans)</li>
3337 <li><tt>ueq</tt>: unordered or equal</li>
3338 <li><tt>ugt</tt>: unordered or greater than </li>
3339 <li><tt>uge</tt>: unordered or greater than or equal</li>
3340 <li><tt>ult</tt>: unordered or less than </li>
3341 <li><tt>ule</tt>: unordered or less than or equal</li>
3342 <li><tt>une</tt>: unordered or not equal</li>
3343 <li><tt>uno</tt>: unordered (either nans)</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3344 <li><tt>true</tt>: no comparison, always returns true</li>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3345</ol>
Reid Spencer02e0d1d2006-12-06 07:08:07 +0000[diff] [blame]3346<p>In the preceding, <i>ordered</i> means that neither operand is a QNAN while
3347<i>unordered</i> means that either operand may be a QNAN.</p>
Reid Spencer784ef792007-01-04 05:19:58 +0000[diff] [blame]3348<p>The <tt>val1</tt> and <tt>val2</tt> arguments must be
3349<a href="#t_floating">floating point</a> typed. They must have identical
3350types.</p>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3351<p>In the foregoing, <i>ordered</i> means that neither operand is a QNAN and
3352<i>unordered</i> means that either operand is a QNAN.</p>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3353<h5>Semantics:</h5>
3354<p>The '<tt>fcmp</tt>' compares <tt>var1</tt> and <tt>var2</tt> according to
3355the condition code given as <tt>cond</tt>. The comparison performed always
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]3356yields a <a href="#t_primitive">i1</a> result, as follows:
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3357<ol>
3358 <li><tt>false</tt>: always yields <tt>false</tt>, regardless of operands.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3359 <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]3360 <tt>var1</tt> is equal to <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3361 <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]3362 <tt>var1</tt> is greather than <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3363 <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]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>olt</tt>: yields <tt>true</tt> if both operands are not a QNAN and
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>ole</tt>: yields <tt>true</tt> if both operands are not a QNAN and
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>one</tt>: yields <tt>true</tt> if both operands are not a QNAN and
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>ord</tt>: yields <tt>true</tt> if both operands are not a QNAN.</li>
3372 <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]3373 <tt>var1</tt> is equal to <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3374 <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]3375 <tt>var1</tt> is greater than <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3376 <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]3377 <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3378 <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]3379 <tt>var1</tt> is less than <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3380 <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]3381 <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3382 <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]3383 <tt>var1</tt> is not equal to <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3384 <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]3385 <li><tt>true</tt>: always yields <tt>true</tt>, regardless of operands.</li>
3386</ol>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3387
3388<h5>Example:</h5>
3389<pre> &lt;result&gt; = fcmp oeq float 4.0, 5.0 <i>; yields: result=false</i>
3390 &lt;result&gt; = icmp one float 4.0, 5.0 <i>; yields: result=true</i>
3391 &lt;result&gt; = icmp olt float 4.0, 5.0 <i>; yields: result=true</i>
3392 &lt;result&gt; = icmp ueq double 1.0, 2.0 <i>; yields: result=false</i>
3393</pre>
3394</div>
3395
Reid Spencer97c5fa42006-11-08 01:18:52 +0000[diff] [blame]3396<!-- _______________________________________________________________________ -->
3397<div class="doc_subsubsection"> <a name="i_phi">'<tt>phi</tt>'
3398Instruction</a> </div>
3399<div class="doc_text">
3400<h5>Syntax:</h5>
3401<pre> &lt;result&gt; = phi &lt;ty&gt; [ &lt;val0&gt;, &lt;label0&gt;], ...<br></pre>
3402<h5>Overview:</h5>
3403<p>The '<tt>phi</tt>' instruction is used to implement the &#966; node in
3404the SSA graph representing the function.</p>
3405<h5>Arguments:</h5>
3406<p>The type of the incoming values are specified with the first type
3407field. After this, the '<tt>phi</tt>' instruction takes a list of pairs
3408as arguments, with one pair for each predecessor basic block of the
3409current block. Only values of <a href="#t_firstclass">first class</a>
3410type may be used as the value arguments to the PHI node. Only labels
3411may be used as the label arguments.</p>
3412<p>There must be no non-phi instructions between the start of a basic
3413block and the PHI instructions: i.e. PHI instructions must be first in
3414a basic block.</p>
3415<h5>Semantics:</h5>
3416<p>At runtime, the '<tt>phi</tt>' instruction logically takes on the
3417value specified by the parameter, depending on which basic block we
3418came from in the last <a href="#terminators">terminator</a> instruction.</p>
3419<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3420<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]3421</div>
3422
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]3423<!-- _______________________________________________________________________ -->
3424<div class="doc_subsubsection">
3425 <a name="i_select">'<tt>select</tt>' Instruction</a>
3426</div>
3427
3428<div class="doc_text">
3429
3430<h5>Syntax:</h5>
3431
3432<pre>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]3433 &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]3434</pre>
3435
3436<h5>Overview:</h5>
3437
3438<p>
3439The '<tt>select</tt>' instruction is used to choose one value based on a
3440condition, without branching.
3441</p>
3442
3443
3444<h5>Arguments:</h5>
3445
3446<p>
3447The '<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.
3448</p>
3449
3450<h5>Semantics:</h5>
3451
3452<p>
3453If the boolean condition evaluates to true, the instruction returns the first
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]3454value argument; otherwise, it returns the second value argument.
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]3455</p>
3456
3457<h5>Example:</h5>
3458
3459<pre>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]3460 %X = select i1 true, i8 17, i8 42 <i>; yields i8:17</i>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]3461</pre>
3462</div>
3463
Robert Bocchinof72fdfe2006-01-15 20:48:27 +0000[diff] [blame]3464
3465<!-- _______________________________________________________________________ -->
3466<div class="doc_subsubsection">
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3467 <a name="i_call">'<tt>call</tt>' Instruction</a>
3468</div>
3469
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3470<div class="doc_text">
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3471
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]3472<h5>Syntax:</h5>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3473<pre>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]3474 &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]3475</pre>
3476
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]3477<h5>Overview:</h5>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3478
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3479<p>The '<tt>call</tt>' instruction represents a simple function call.</p>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3480
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]3481<h5>Arguments:</h5>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3482
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3483<p>This instruction requires several arguments:</p>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3484
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]3485<ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3486 <li>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]3487 <p>The optional "tail" marker indicates whether the callee function accesses
3488 any allocas or varargs in the caller. If the "tail" marker is present, the
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3489 function call is eligible for tail call optimization. Note that calls may
3490 be marked "tail" even if they do not occur before a <a
3491 href="#i_ret"><tt>ret</tt></a> instruction.
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3492 </li>
3493 <li>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]3494 <p>The optional "cconv" marker indicates which <a href="callingconv">calling
3495 convention</a> the call should use. If none is specified, the call defaults
3496 to using C calling conventions.
3497 </li>
3498 <li>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3499 <p>'<tt>ty</tt>': shall be the signature of the pointer to function value
3500 being invoked. The argument types must match the types implied by this
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]3501 signature. This type can be omitted if the function is not varargs and
3502 if the function type does not return a pointer to a function.</p>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3503 </li>
3504 <li>
3505 <p>'<tt>fnptrval</tt>': An LLVM value containing a pointer to a function to
3506 be invoked. In most cases, this is a direct function invocation, but
3507 indirect <tt>call</tt>s are just as possible, calling an arbitrary pointer
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]3508 to function value.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3509 </li>
3510 <li>
3511 <p>'<tt>function args</tt>': argument list whose types match the
Reid Spencerd845d162005-05-01 22:22:57 +0000[diff] [blame]3512 function signature argument types. All arguments must be of
3513 <a href="#t_firstclass">first class</a> type. If the function signature
3514 indicates the function accepts a variable number of arguments, the extra
3515 arguments can be specified.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3516 </li>
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]3517</ol>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3518
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]3519<h5>Semantics:</h5>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3520
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3521<p>The '<tt>call</tt>' instruction is used to cause control flow to
3522transfer to a specified function, with its incoming arguments bound to
3523the specified values. Upon a '<tt><a href="#i_ret">ret</a></tt>'
3524instruction in the called function, control flow continues with the
3525instruction after the function call, and the return value of the
3526function is bound to the result argument. This is a simpler case of
3527the <a href="#i_invoke">invoke</a> instruction.</p>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3528
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]3529<h5>Example:</h5>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3530
3531<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3532 %retval = call i32 %test(i32 %argc)
3533 call i32(i8 *, ...) *%printf(i8 * %msg, i32 12, i8 42);
3534 %X = tail call i32 %foo()
3535 %Y = tail call <a href="#callingconv">fastcc</a> i32 %foo()
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3536</pre>
3537
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3538</div>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3539
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3540<!-- _______________________________________________________________________ -->
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3541<div class="doc_subsubsection">
Chris Lattner33337472006-01-13 23:26:01 +0000[diff] [blame]3542 <a name="i_va_arg">'<tt>va_arg</tt>' Instruction</a>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3543</div>
3544
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3545<div class="doc_text">
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3546
Chris Lattner26ca62e2003-10-18 05:51:36 +0000[diff] [blame]3547<h5>Syntax:</h5>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3548
3549<pre>
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3550 &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]3551</pre>
3552
Chris Lattner26ca62e2003-10-18 05:51:36 +0000[diff] [blame]3553<h5>Overview:</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 is used to access arguments passed through
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3556the "variable argument" area of a function call. It is used to implement the
3557<tt>va_arg</tt> macro in C.</p>
3558
Chris Lattner26ca62e2003-10-18 05:51:36 +0000[diff] [blame]3559<h5>Arguments:</h5>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3560
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3561<p>This instruction takes a <tt>va_list*</tt> value and the type of
3562the argument. It returns a value of the specified argument type and
Jeff Cohendc6bfea2005-11-11 02:15:27 +0000[diff] [blame]3563increments the <tt>va_list</tt> to point to the next argument. Again, the
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3564actual type of <tt>va_list</tt> is target specific.</p>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3565
Chris Lattner26ca62e2003-10-18 05:51:36 +0000[diff] [blame]3566<h5>Semantics:</h5>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3567
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3568<p>The '<tt>va_arg</tt>' instruction loads an argument of the specified
3569type from the specified <tt>va_list</tt> and causes the
3570<tt>va_list</tt> to point to the next argument. For more information,
3571see the variable argument handling <a href="#int_varargs">Intrinsic
3572Functions</a>.</p>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3573
3574<p>It is legal for this instruction to be called in a function which does not
3575take a variable number of arguments, for example, the <tt>vfprintf</tt>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3576function.</p>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3577
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3578<p><tt>va_arg</tt> is an LLVM instruction instead of an <a
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]3579href="#intrinsics">intrinsic function</a> because it takes a type as an
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3580argument.</p>
3581
Chris Lattner26ca62e2003-10-18 05:51:36 +0000[diff] [blame]3582<h5>Example:</h5>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3583
3584<p>See the <a href="#int_varargs">variable argument processing</a> section.</p>
3585
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3586</div>
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3587
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3588<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3589<div class="doc_section"> <a name="intrinsics">Intrinsic Functions</a> </div>
3590<!-- *********************************************************************** -->
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3591
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3592<div class="doc_text">
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3593
3594<p>LLVM supports the notion of an "intrinsic function". These functions have
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]3595well known names and semantics and are required to follow certain
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3596restrictions. Overall, these instructions represent an extension mechanism for
3597the LLVM language that does not require changing all of the transformations in
3598LLVM to add to the language (or the bytecode reader/writer, the parser,
3599etc...).</p>
3600
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]3601<p>Intrinsic function names must all start with an "<tt>llvm.</tt>" prefix. This
3602prefix is reserved in LLVM for intrinsic names; thus, functions may not be named
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3603this. Intrinsic functions must always be external functions: you cannot define
3604the body of intrinsic functions. Intrinsic functions may only be used in call
3605or invoke instructions: it is illegal to take the address of an intrinsic
3606function. Additionally, because intrinsic functions are part of the LLVM
3607language, it is required that they all be documented here if any are added.</p>
3608
3609
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]3610<p>To learn how to add an intrinsic function, please see the <a
Chris Lattner90391c12005-05-11 03:35:57 +0000[diff] [blame]3611href="ExtendingLLVM.html">Extending LLVM Guide</a>.
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3612</p>
3613
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3614</div>
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3615
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3616<!-- ======================================================================= -->
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3617<div class="doc_subsection">
3618 <a name="int_varargs">Variable Argument Handling Intrinsics</a>
3619</div>
3620
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3621<div class="doc_text">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3622
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3623<p>Variable argument support is defined in LLVM with the <a
Chris Lattner33337472006-01-13 23:26:01 +0000[diff] [blame]3624 href="#i_va_arg"><tt>va_arg</tt></a> instruction and these three
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3625intrinsic functions. These functions are related to the similarly
3626named macros defined in the <tt>&lt;stdarg.h&gt;</tt> header file.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3627
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3628<p>All of these functions operate on arguments that use a
3629target-specific value type "<tt>va_list</tt>". The LLVM assembly
3630language reference manual does not define what this type is, so all
3631transformations should be prepared to handle intrinsics with any type
3632used.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3633
Chris Lattner30b868d2006-05-15 17:26:46 +0000[diff] [blame]3634<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]3635instruction and the variable argument handling intrinsic functions are
3636used.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3637
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3638<pre>
Anton Korobeynikov640bbe02007-03-21 23:58:04 +0000[diff] [blame]3639define i32 @test(i32 %X, ...) {
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3640 ; Initialize variable argument processing
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3641 %ap = alloca i8 *
Chris Lattnerdb0790c2007-01-08 07:55:15 +0000[diff] [blame]3642 %ap2 = bitcast i8** %ap to i8*
Anton Korobeynikov640bbe02007-03-21 23:58:04 +0000[diff] [blame]3643 call void @llvm.va_start(i8* %ap2)
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3644
3645 ; Read a single integer argument
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3646 %tmp = va_arg i8 ** %ap, i32
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3647
3648 ; Demonstrate usage of llvm.va_copy and llvm.va_end
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3649 %aq = alloca i8 *
Chris Lattnerdb0790c2007-01-08 07:55:15 +0000[diff] [blame]3650 %aq2 = bitcast i8** %aq to i8*
Anton Korobeynikov640bbe02007-03-21 23:58:04 +0000[diff] [blame]3651 call void @llvm.va_copy(i8 *%aq2, i8* %ap2)
3652 call void @llvm.va_end(i8* %aq2)
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3653
3654 ; Stop processing of arguments.
Anton Korobeynikov640bbe02007-03-21 23:58:04 +0000[diff] [blame]3655 call void @llvm.va_end(i8* %ap2)
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3656 ret i32 %tmp
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3657}
Anton Korobeynikov640bbe02007-03-21 23:58:04 +0000[diff] [blame]3658
3659declare void @llvm.va_start(i8*)
3660declare void @llvm.va_copy(i8*, i8*)
3661declare void @llvm.va_end(i8*)
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3662</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3663</div>
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3664
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3665<!-- _______________________________________________________________________ -->
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3666<div class="doc_subsubsection">
3667 <a name="i_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a>
3668</div>
3669
3670
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3671<div class="doc_text">
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3672<h5>Syntax:</h5>
Chris Lattnerdb0790c2007-01-08 07:55:15 +0000[diff] [blame]3673<pre> declare void %llvm.va_start(i8* &lt;arglist&gt;)<br></pre>
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3674<h5>Overview:</h5>
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3675<P>The '<tt>llvm.va_start</tt>' intrinsic initializes
3676<tt>*&lt;arglist&gt;</tt> for subsequent use by <tt><a
3677href="#i_va_arg">va_arg</a></tt>.</p>
3678
3679<h5>Arguments:</h5>
3680
3681<P>The argument is a pointer to a <tt>va_list</tt> element to initialize.</p>
3682
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3683<h5>Semantics:</h5>
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3684
3685<P>The '<tt>llvm.va_start</tt>' intrinsic works just like the <tt>va_start</tt>
3686macro available in C. In a target-dependent way, it initializes the
3687<tt>va_list</tt> element the argument points to, so that the next call to
3688<tt>va_arg</tt> will produce the first variable argument passed to the function.
3689Unlike the C <tt>va_start</tt> macro, this intrinsic does not need to know the
3690last argument of the function, the compiler can figure that out.</p>
3691
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3692</div>
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3693
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3694<!-- _______________________________________________________________________ -->
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3695<div class="doc_subsubsection">
3696 <a name="i_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a>
3697</div>
3698
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3699<div class="doc_text">
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3700<h5>Syntax:</h5>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]3701<pre> declare void @llvm.va_end(i8* &lt;arglist&gt;)<br></pre>
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3702<h5>Overview:</h5>
Chris Lattnerdb0790c2007-01-08 07:55:15 +0000[diff] [blame]3703
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3704<p>The '<tt>llvm.va_end</tt>' intrinsic destroys <tt>&lt;arglist&gt;</tt>
3705which has been initialized previously with <tt><a href="#i_va_start">llvm.va_start</a></tt>
3706or <tt><a href="#i_va_copy">llvm.va_copy</a></tt>.</p>
Chris Lattnerdb0790c2007-01-08 07:55:15 +0000[diff] [blame]3707
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3708<h5>Arguments:</h5>
Chris Lattnerdb0790c2007-01-08 07:55:15 +0000[diff] [blame]3709
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3710<p>The argument is a <tt>va_list</tt> to destroy.</p>
Chris Lattnerdb0790c2007-01-08 07:55:15 +0000[diff] [blame]3711
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3712<h5>Semantics:</h5>
Chris Lattnerdb0790c2007-01-08 07:55:15 +0000[diff] [blame]3713
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3714<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]3715macro available in C. In a target-dependent way, it destroys the <tt>va_list</tt>.
3716Calls to <a href="#i_va_start"><tt>llvm.va_start</tt></a> and <a
3717 href="#i_va_copy"><tt>llvm.va_copy</tt></a> must be matched exactly
3718with calls to <tt>llvm.va_end</tt>.</p>
Chris Lattnerdb0790c2007-01-08 07:55:15 +0000[diff] [blame]3719
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3720</div>
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3721
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3722<!-- _______________________________________________________________________ -->
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3723<div class="doc_subsubsection">
3724 <a name="i_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a>
3725</div>
3726
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3727<div class="doc_text">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3728
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3729<h5>Syntax:</h5>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3730
3731<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]3732 declare void @llvm.va_copy(i8* &lt;destarglist&gt;, i8* &lt;srcarglist&gt;)
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3733</pre>
3734
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3735<h5>Overview:</h5>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3736
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3737<p>The '<tt>llvm.va_copy</tt>' intrinsic copies the current argument position from
3738the source argument list to the destination argument list.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3739
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3740<h5>Arguments:</h5>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3741
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3742<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]3743The 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]3744
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3745
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3746<h5>Semantics:</h5>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3747
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3748<p>The '<tt>llvm.va_copy</tt>' intrinsic works just like the <tt>va_copy</tt> macro
3749available in C. In a target-dependent way, it copies the source
3750<tt>va_list</tt> element into the destination list. This intrinsic is necessary
3751because the <tt><a href="i_va_begin">llvm.va_begin</a></tt> intrinsic may be
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3752arbitrarily complex and require memory allocation, for example.</p>
3753
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3754</div>
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3755
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3756<!-- ======================================================================= -->
3757<div class="doc_subsection">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3758 <a name="int_gc">Accurate Garbage Collection Intrinsics</a>
3759</div>
3760
3761<div class="doc_text">
3762
3763<p>
3764LLVM support for <a href="GarbageCollection.html">Accurate Garbage
3765Collection</a> requires the implementation and generation of these intrinsics.
3766These intrinsics allow identification of <a href="#i_gcroot">GC roots on the
3767stack</a>, as well as garbage collector implementations that require <a
3768href="#i_gcread">read</a> and <a href="#i_gcwrite">write</a> barriers.
3769Front-ends for type-safe garbage collected languages should generate these
3770intrinsics to make use of the LLVM garbage collectors. For more details, see <a
3771href="GarbageCollection.html">Accurate Garbage Collection with LLVM</a>.
3772</p>
3773</div>
3774
3775<!-- _______________________________________________________________________ -->
3776<div class="doc_subsubsection">
3777 <a name="i_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a>
3778</div>
3779
3780<div class="doc_text">
3781
3782<h5>Syntax:</h5>
3783
3784<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]3785 declare void @llvm.gcroot(&lt;ty&gt;** %ptrloc, &lt;ty2&gt;* %metadata)
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3786</pre>
3787
3788<h5>Overview:</h5>
3789
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]3790<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]3791the code generator, and allows some metadata to be associated with it.</p>
3792
3793<h5>Arguments:</h5>
3794
3795<p>The first argument specifies the address of a stack object that contains the
3796root pointer. The second pointer (which must be either a constant or a global
3797value address) contains the meta-data to be associated with the root.</p>
3798
3799<h5>Semantics:</h5>
3800
3801<p>At runtime, a call to this intrinsics stores a null pointer into the "ptrloc"
3802location. At compile-time, the code generator generates information to allow
3803the runtime to find the pointer at GC safe points.
3804</p>
3805
3806</div>
3807
3808
3809<!-- _______________________________________________________________________ -->
3810<div class="doc_subsubsection">
3811 <a name="i_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a>
3812</div>
3813
3814<div class="doc_text">
3815
3816<h5>Syntax:</h5>
3817
3818<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]3819 declare i8 * @llvm.gcread(i8 * %ObjPtr, i8 ** %Ptr)
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3820</pre>
3821
3822<h5>Overview:</h5>
3823
3824<p>The '<tt>llvm.gcread</tt>' intrinsic identifies reads of references from heap
3825locations, allowing garbage collector implementations that require read
3826barriers.</p>
3827
3828<h5>Arguments:</h5>
3829
Chris Lattnerf9228072006-03-14 20:02:51 +0000[diff] [blame]3830<p>The second argument is the address to read from, which should be an address
3831allocated from the garbage collector. The first object is a pointer to the
3832start of the referenced object, if needed by the language runtime (otherwise
3833null).</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3834
3835<h5>Semantics:</h5>
3836
3837<p>The '<tt>llvm.gcread</tt>' intrinsic has the same semantics as a load
3838instruction, but may be replaced with substantially more complex code by the
3839garbage collector runtime, as needed.</p>
3840
3841</div>
3842
3843
3844<!-- _______________________________________________________________________ -->
3845<div class="doc_subsubsection">
3846 <a name="i_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a>
3847</div>
3848
3849<div class="doc_text">
3850
3851<h5>Syntax:</h5>
3852
3853<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]3854 declare void @llvm.gcwrite(i8 * %P1, i8 * %Obj, i8 ** %P2)
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3855</pre>
3856
3857<h5>Overview:</h5>
3858
3859<p>The '<tt>llvm.gcwrite</tt>' intrinsic identifies writes of references to heap
3860locations, allowing garbage collector implementations that require write
3861barriers (such as generational or reference counting collectors).</p>
3862
3863<h5>Arguments:</h5>
3864
Chris Lattnerf9228072006-03-14 20:02:51 +0000[diff] [blame]3865<p>The first argument is the reference to store, the second is the start of the
3866object to store it to, and the third is the address of the field of Obj to
3867store to. If the runtime does not require a pointer to the object, Obj may be
3868null.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3869
3870<h5>Semantics:</h5>
3871
3872<p>The '<tt>llvm.gcwrite</tt>' intrinsic has the same semantics as a store
3873instruction, but may be replaced with substantially more complex code by the
3874garbage collector runtime, as needed.</p>
3875
3876</div>
3877
3878
3879
3880<!-- ======================================================================= -->
3881<div class="doc_subsection">
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3882 <a name="int_codegen">Code Generator Intrinsics</a>
3883</div>
3884
3885<div class="doc_text">
3886<p>
3887These intrinsics are provided by LLVM to expose special features that may only
3888be implemented with code generator support.
3889</p>
3890
3891</div>
3892
3893<!-- _______________________________________________________________________ -->
3894<div class="doc_subsubsection">
3895 <a name="i_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a>
3896</div>
3897
3898<div class="doc_text">
3899
3900<h5>Syntax:</h5>
3901<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]3902 declare i8 *@llvm.returnaddress(i32 &lt;level&gt;)
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3903</pre>
3904
3905<h5>Overview:</h5>
3906
3907<p>
Chris Lattnerc1fb4262006-10-15 20:05:59 +0000[diff] [blame]3908The '<tt>llvm.returnaddress</tt>' intrinsic attempts to compute a
3909target-specific value indicating the return address of the current function
3910or one of its callers.
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3911</p>
3912
3913<h5>Arguments:</h5>
3914
3915<p>
3916The argument to this intrinsic indicates which function to return the address
3917for. Zero indicates the calling function, one indicates its caller, etc. The
3918argument is <b>required</b> to be a constant integer value.
3919</p>
3920
3921<h5>Semantics:</h5>
3922
3923<p>
3924The '<tt>llvm.returnaddress</tt>' intrinsic either returns a pointer indicating
3925the return address of the specified call frame, or zero if it cannot be
3926identified. The value returned by this intrinsic is likely to be incorrect or 0
3927for arguments other than zero, so it should only be used for debugging purposes.
3928</p>
3929
3930<p>
3931Note that calling this intrinsic does not prevent function inlining or other
Chris Lattner2e6eb5f2005-03-07 20:30:51 +0000[diff] [blame]3932aggressive transformations, so the value returned may not be that of the obvious
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3933source-language caller.
3934</p>
3935</div>
3936
3937
3938<!-- _______________________________________________________________________ -->
3939<div class="doc_subsubsection">
3940 <a name="i_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a>
3941</div>
3942
3943<div class="doc_text">
3944
3945<h5>Syntax:</h5>
3946<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]3947 declare i8 *@llvm.frameaddress(i32 &lt;level&gt;)
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3948</pre>
3949
3950<h5>Overview:</h5>
3951
3952<p>
Chris Lattnerc1fb4262006-10-15 20:05:59 +0000[diff] [blame]3953The '<tt>llvm.frameaddress</tt>' intrinsic attempts to return the
3954target-specific frame pointer value for the specified stack frame.
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3955</p>
3956
3957<h5>Arguments:</h5>
3958
3959<p>
3960The argument to this intrinsic indicates which function to return the frame
3961pointer for. Zero indicates the calling function, one indicates its caller,
3962etc. The argument is <b>required</b> to be a constant integer value.
3963</p>
3964
3965<h5>Semantics:</h5>
3966
3967<p>
3968The '<tt>llvm.frameaddress</tt>' intrinsic either returns a pointer indicating
3969the frame address of the specified call frame, or zero if it cannot be
3970identified. The value returned by this intrinsic is likely to be incorrect or 0
3971for arguments other than zero, so it should only be used for debugging purposes.
3972</p>
3973
3974<p>
3975Note that calling this intrinsic does not prevent function inlining or other
Chris Lattner2e6eb5f2005-03-07 20:30:51 +0000[diff] [blame]3976aggressive transformations, so the value returned may not be that of the obvious
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3977source-language caller.
3978</p>
3979</div>
3980
Chris Lattnerc8a2c222005-02-28 19:24:19 +0000[diff] [blame]3981<!-- _______________________________________________________________________ -->
3982<div class="doc_subsubsection">
Chris Lattner2f0f0012006-01-13 02:03:13 +0000[diff] [blame]3983 <a name="i_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a>
3984</div>
3985
3986<div class="doc_text">
3987
3988<h5>Syntax:</h5>
3989<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]3990 declare i8 *@llvm.stacksave()
Chris Lattner2f0f0012006-01-13 02:03:13 +0000[diff] [blame]3991</pre>
3992
3993<h5>Overview:</h5>
3994
3995<p>
3996The '<tt>llvm.stacksave</tt>' intrinsic is used to remember the current state of
3997the function stack, for use with <a href="#i_stackrestore">
3998<tt>llvm.stackrestore</tt></a>. This is useful for implementing language
3999features like scoped automatic variable sized arrays in C99.
4000</p>
4001
4002<h5>Semantics:</h5>
4003
4004<p>
4005This intrinsic returns a opaque pointer value that can be passed to <a
4006href="#i_stackrestore"><tt>llvm.stackrestore</tt></a>. When an
4007<tt>llvm.stackrestore</tt> intrinsic is executed with a value saved from
4008<tt>llvm.stacksave</tt>, it effectively restores the state of the stack to the
4009state it was in when the <tt>llvm.stacksave</tt> intrinsic executed. In
4010practice, this pops any <a href="#i_alloca">alloca</a> blocks from the stack
4011that were allocated after the <tt>llvm.stacksave</tt> was executed.
4012</p>
4013
4014</div>
4015
4016<!-- _______________________________________________________________________ -->
4017<div class="doc_subsubsection">
4018 <a name="i_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a>
4019</div>
4020
4021<div class="doc_text">
4022
4023<h5>Syntax:</h5>
4024<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4025 declare void @llvm.stackrestore(i8 * %ptr)
Chris Lattner2f0f0012006-01-13 02:03:13 +0000[diff] [blame]4026</pre>
4027
4028<h5>Overview:</h5>
4029
4030<p>
4031The '<tt>llvm.stackrestore</tt>' intrinsic is used to restore the state of
4032the function stack to the state it was in when the corresponding <a
4033href="#llvm.stacksave"><tt>llvm.stacksave</tt></a> intrinsic executed. This is
4034useful for implementing language features like scoped automatic variable sized
4035arrays in C99.
4036</p>
4037
4038<h5>Semantics:</h5>
4039
4040<p>
4041See the description for <a href="#i_stacksave"><tt>llvm.stacksave</tt></a>.
4042</p>
4043
4044</div>
4045
4046
4047<!-- _______________________________________________________________________ -->
4048<div class="doc_subsubsection">
Chris Lattnerc8a2c222005-02-28 19:24:19 +0000[diff] [blame]4049 <a name="i_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a>
4050</div>
4051
4052<div class="doc_text">
4053
4054<h5>Syntax:</h5>
4055<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4056 declare void @llvm.prefetch(i8 * &lt;address&gt;,
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]4057 i32 &lt;rw&gt;, i32 &lt;locality&gt;)
Chris Lattnerc8a2c222005-02-28 19:24:19 +0000[diff] [blame]4058</pre>
4059
4060<h5>Overview:</h5>
4061
4062
4063<p>
4064The '<tt>llvm.prefetch</tt>' intrinsic is a hint to the code generator to insert
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]4065a prefetch instruction if supported; otherwise, it is a noop. Prefetches have
4066no
4067effect on the behavior of the program but can change its performance
Chris Lattnerff851072005-02-28 19:47:14 +0000[diff] [blame]4068characteristics.
Chris Lattnerc8a2c222005-02-28 19:24:19 +0000[diff] [blame]4069</p>
4070
4071<h5>Arguments:</h5>
4072
4073<p>
4074<tt>address</tt> is the address to be prefetched, <tt>rw</tt> is the specifier
4075determining if the fetch should be for a read (0) or write (1), and
4076<tt>locality</tt> is a temporal locality specifier ranging from (0) - no
Chris Lattnerd3e641c2005-03-07 20:31:38 +0000[diff] [blame]4077locality, to (3) - extremely local keep in cache. The <tt>rw</tt> and
Chris Lattnerc8a2c222005-02-28 19:24:19 +0000[diff] [blame]4078<tt>locality</tt> arguments must be constant integers.
4079</p>
4080
4081<h5>Semantics:</h5>
4082
4083<p>
4084This intrinsic does not modify the behavior of the program. In particular,
4085prefetches cannot trap and do not produce a value. On targets that support this
4086intrinsic, the prefetch can provide hints to the processor cache for better
4087performance.
4088</p>
4089
4090</div>
4091
Andrew Lenharthb4427912005-03-28 20:05:49 +0000[diff] [blame]4092<!-- _______________________________________________________________________ -->
4093<div class="doc_subsubsection">
4094 <a name="i_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a>
4095</div>
4096
4097<div class="doc_text">
4098
4099<h5>Syntax:</h5>
4100<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4101 declare void @llvm.pcmarker( i32 &lt;id&gt; )
Andrew Lenharthb4427912005-03-28 20:05:49 +0000[diff] [blame]4102</pre>
4103
4104<h5>Overview:</h5>
4105
4106
4107<p>
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]4108The '<tt>llvm.pcmarker</tt>' intrinsic is a method to export a Program Counter
4109(PC) in a region of
Andrew Lenharthb4427912005-03-28 20:05:49 +0000[diff] [blame]4110code to simulators and other tools. The method is target specific, but it is
4111expected that the marker will use exported symbols to transmit the PC of the marker.
Jeff Cohendc6bfea2005-11-11 02:15:27 +0000[diff] [blame]4112The marker makes no guarantees that it will remain with any specific instruction
Chris Lattnere64d41d2005-11-15 06:07:55 +0000[diff] [blame]4113after optimizations. It is possible that the presence of a marker will inhibit
Chris Lattnerb40261e2006-03-24 07:16:10 +0000[diff] [blame]4114optimizations. The intended use is to be inserted after optimizations to allow
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]4115correlations of simulation runs.
Andrew Lenharthb4427912005-03-28 20:05:49 +0000[diff] [blame]4116</p>
4117
4118<h5>Arguments:</h5>
4119
4120<p>
4121<tt>id</tt> is a numerical id identifying the marker.
4122</p>
4123
4124<h5>Semantics:</h5>
4125
4126<p>
4127This intrinsic does not modify the behavior of the program. Backends that do not
4128support this intrinisic may ignore it.
4129</p>
4130
4131</div>
4132
Andrew Lenharth01aa5632005-11-11 16:47:30 +0000[diff] [blame]4133<!-- _______________________________________________________________________ -->
4134<div class="doc_subsubsection">
4135 <a name="i_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a>
4136</div>
4137
4138<div class="doc_text">
4139
4140<h5>Syntax:</h5>
4141<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4142 declare i64 @llvm.readcyclecounter( )
Andrew Lenharth01aa5632005-11-11 16:47:30 +0000[diff] [blame]4143</pre>
4144
4145<h5>Overview:</h5>
4146
4147
4148<p>
4149The '<tt>llvm.readcyclecounter</tt>' intrinsic provides access to the cycle
4150counter register (or similar low latency, high accuracy clocks) on those targets
4151that support it. On X86, it should map to RDTSC. On Alpha, it should map to RPCC.
4152As the backing counters overflow quickly (on the order of 9 seconds on alpha), this
4153should only be used for small timings.
4154</p>
4155
4156<h5>Semantics:</h5>
4157
4158<p>
4159When directly supported, reading the cycle counter should not modify any memory.
4160Implementations are allowed to either return a application specific value or a
4161system wide value. On backends without support, this is lowered to a constant 0.
4162</p>
4163
4164</div>
4165
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4166<!-- ======================================================================= -->
4167<div class="doc_subsection">
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]4168 <a name="int_libc">Standard C Library Intrinsics</a>
4169</div>
4170
4171<div class="doc_text">
4172<p>
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4173LLVM provides intrinsics for a few important standard C library functions.
4174These intrinsics allow source-language front-ends to pass information about the
4175alignment of the pointer arguments to the code generator, providing opportunity
4176for more efficient code generation.
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]4177</p>
4178
4179</div>
4180
4181<!-- _______________________________________________________________________ -->
4182<div class="doc_subsubsection">
4183 <a name="i_memcpy">'<tt>llvm.memcpy</tt>' Intrinsic</a>
4184</div>
4185
4186<div class="doc_text">
4187
4188<h5>Syntax:</h5>
4189<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4190 declare void @llvm.memcpy.i32(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]4191 i32 &lt;len&gt;, i32 &lt;align&gt;)
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4192 declare void @llvm.memcpy.i64(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]4193 i64 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]4194</pre>
4195
4196<h5>Overview:</h5>
4197
4198<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4199The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]4200location to the destination location.
4201</p>
4202
4203<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4204Note that, unlike the standard libc function, the <tt>llvm.memcpy.*</tt>
4205intrinsics do not return a value, and takes an extra alignment argument.
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]4206</p>
4207
4208<h5>Arguments:</h5>
4209
4210<p>
4211The first argument is a pointer to the destination, the second is a pointer to
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4212the source. The third argument is an integer argument
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]4213specifying the number of bytes to copy, and the fourth argument is the alignment
4214of the source and destination locations.
4215</p>
4216
Chris Lattner4c67c482004-02-12 21:18:15 +0000[diff] [blame]4217<p>
4218If 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]4219the caller guarantees that both the source and destination pointers are aligned
4220to that boundary.
Chris Lattner4c67c482004-02-12 21:18:15 +0000[diff] [blame]4221</p>
4222
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]4223<h5>Semantics:</h5>
4224
4225<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4226The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]4227location to the destination location, which are not allowed to overlap. It
4228copies "len" bytes of memory over. If the argument is known to be aligned to
4229some boundary, this can be specified as the fourth argument, otherwise it should
4230be set to 0 or 1.
4231</p>
4232</div>
4233
4234
Chris Lattnerf30152e2004-02-12 18:10:10 +0000[diff] [blame]4235<!-- _______________________________________________________________________ -->
4236<div class="doc_subsubsection">
4237 <a name="i_memmove">'<tt>llvm.memmove</tt>' Intrinsic</a>
4238</div>
4239
4240<div class="doc_text">
4241
4242<h5>Syntax:</h5>
4243<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4244 declare void @llvm.memmove.i32(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]4245 i32 &lt;len&gt;, i32 &lt;align&gt;)
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4246 declare void @llvm.memmove.i64(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]4247 i64 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattnerf30152e2004-02-12 18:10:10 +0000[diff] [blame]4248</pre>
4249
4250<h5>Overview:</h5>
4251
4252<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4253The '<tt>llvm.memmove.*</tt>' intrinsics move a block of memory from the source
4254location to the destination location. It is similar to the
4255'<tt>llvm.memcmp</tt>' intrinsic but allows the two memory locations to overlap.
Chris Lattnerf30152e2004-02-12 18:10:10 +0000[diff] [blame]4256</p>
4257
4258<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4259Note that, unlike the standard libc function, the <tt>llvm.memmove.*</tt>
4260intrinsics do not return a value, and takes an extra alignment argument.
Chris Lattnerf30152e2004-02-12 18:10:10 +0000[diff] [blame]4261</p>
4262
4263<h5>Arguments:</h5>
4264
4265<p>
4266The first argument is a pointer to the destination, the second is a pointer to
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4267the source. The third argument is an integer argument
Chris Lattnerf30152e2004-02-12 18:10:10 +0000[diff] [blame]4268specifying the number of bytes to copy, and the fourth argument is the alignment
4269of the source and destination locations.
4270</p>
4271
Chris Lattner4c67c482004-02-12 21:18:15 +0000[diff] [blame]4272<p>
4273If 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]4274the caller guarantees that the source and destination pointers are aligned to
4275that boundary.
Chris Lattner4c67c482004-02-12 21:18:15 +0000[diff] [blame]4276</p>
4277
Chris Lattnerf30152e2004-02-12 18:10:10 +0000[diff] [blame]4278<h5>Semantics:</h5>
4279
4280<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4281The '<tt>llvm.memmove.*</tt>' intrinsics copy a block of memory from the source
Chris Lattnerf30152e2004-02-12 18:10:10 +0000[diff] [blame]4282location to the destination location, which may overlap. It
4283copies "len" bytes of memory over. If the argument is known to be aligned to
4284some boundary, this can be specified as the fourth argument, otherwise it should
4285be set to 0 or 1.
4286</p>
4287</div>
4288
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]4289
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4290<!-- _______________________________________________________________________ -->
4291<div class="doc_subsubsection">
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4292 <a name="i_memset">'<tt>llvm.memset.*</tt>' Intrinsics</a>
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4293</div>
4294
4295<div class="doc_text">
4296
4297<h5>Syntax:</h5>
4298<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4299 declare void @llvm.memset.i32(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]4300 i32 &lt;len&gt;, i32 &lt;align&gt;)
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4301 declare void @llvm.memset.i64(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]4302 i64 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4303</pre>
4304
4305<h5>Overview:</h5>
4306
4307<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4308The '<tt>llvm.memset.*</tt>' intrinsics fill a block of memory with a particular
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4309byte value.
4310</p>
4311
4312<p>
4313Note that, unlike the standard libc function, the <tt>llvm.memset</tt> intrinsic
4314does not return a value, and takes an extra alignment argument.
4315</p>
4316
4317<h5>Arguments:</h5>
4318
4319<p>
4320The first argument is a pointer to the destination to fill, the second is the
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4321byte value to fill it with, the third argument is an integer
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4322argument specifying the number of bytes to fill, and the fourth argument is the
4323known alignment of destination location.
4324</p>
4325
4326<p>
4327If 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]4328the caller guarantees that the destination pointer is aligned to that boundary.
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4329</p>
4330
4331<h5>Semantics:</h5>
4332
4333<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4334The '<tt>llvm.memset.*</tt>' intrinsics fill "len" bytes of memory starting at
4335the
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4336destination location. If the argument is known to be aligned to some boundary,
4337this can be specified as the fourth argument, otherwise it should be set to 0 or
43381.
4339</p>
4340</div>
4341
4342
Chris Lattner3b4f4372004-06-11 02:28:03 +0000[diff] [blame]4343<!-- _______________________________________________________________________ -->
4344<div class="doc_subsubsection">
Chris Lattner069b5bd2006-01-16 22:38:59 +0000[diff] [blame]4345 <a name="i_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a>
Chris Lattner8a8f2e52005-07-21 01:29:16 +0000[diff] [blame]4346</div>
4347
4348<div class="doc_text">
4349
4350<h5>Syntax:</h5>
4351<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4352 declare float @llvm.sqrt.f32(float %Val)
4353 declare double @llvm.sqrt.f64(double %Val)
Chris Lattner8a8f2e52005-07-21 01:29:16 +0000[diff] [blame]4354</pre>
4355
4356<h5>Overview:</h5>
4357
4358<p>
Reid Spencerb4f9a6f2006-01-16 21:12:35 +0000[diff] [blame]4359The '<tt>llvm.sqrt</tt>' intrinsics return the sqrt of the specified operand,
Chris Lattner8a8f2e52005-07-21 01:29:16 +0000[diff] [blame]4360returning the same value as the libm '<tt>sqrt</tt>' function would. Unlike
4361<tt>sqrt</tt> in libm, however, <tt>llvm.sqrt</tt> has undefined behavior for
4362negative numbers (which allows for better optimization).
4363</p>
4364
4365<h5>Arguments:</h5>
4366
4367<p>
4368The argument and return value are floating point numbers of the same type.
4369</p>
4370
4371<h5>Semantics:</h5>
4372
4373<p>
4374This function returns the sqrt of the specified operand if it is a positive
4375floating point number.
4376</p>
4377</div>
4378
Chris Lattner33b73f92006-09-08 06:34:02 +0000[diff] [blame]4379<!-- _______________________________________________________________________ -->
4380<div class="doc_subsubsection">
4381 <a name="i_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a>
4382</div>
4383
4384<div class="doc_text">
4385
4386<h5>Syntax:</h5>
4387<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4388 declare float @llvm.powi.f32(float %Val, i32 %power)
4389 declare double @llvm.powi.f64(double %Val, i32 %power)
Chris Lattner33b73f92006-09-08 06:34:02 +0000[diff] [blame]4390</pre>
4391
4392<h5>Overview:</h5>
4393
4394<p>
4395The '<tt>llvm.powi.*</tt>' intrinsics return the first operand raised to the
4396specified (positive or negative) power. The order of evaluation of
4397multiplications is not defined.
4398</p>
4399
4400<h5>Arguments:</h5>
4401
4402<p>
4403The second argument is an integer power, and the first is a value to raise to
4404that power.
4405</p>
4406
4407<h5>Semantics:</h5>
4408
4409<p>
4410This function returns the first value raised to the second power with an
4411unspecified sequence of rounding operations.</p>
4412</div>
4413
4414
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4415<!-- ======================================================================= -->
4416<div class="doc_subsection">
Nate Begeman0f223bb2006-01-13 23:26:38 +0000[diff] [blame]4417 <a name="int_manip">Bit Manipulation Intrinsics</a>
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4418</div>
4419
4420<div class="doc_text">
4421<p>
Nate Begeman0f223bb2006-01-13 23:26:38 +0000[diff] [blame]4422LLVM provides intrinsics for a few important bit manipulation operations.
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4423These allow efficient code generation for some algorithms.
4424</p>
4425
4426</div>
4427
4428<!-- _______________________________________________________________________ -->
4429<div class="doc_subsubsection">
Nate Begeman0f223bb2006-01-13 23:26:38 +0000[diff] [blame]4430 <a name="i_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a>
4431</div>
4432
4433<div class="doc_text">
4434
4435<h5>Syntax:</h5>
4436<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4437 declare i16 @llvm.bswap.i16(i16 &lt;id&gt;)
4438 declare i32 @llvm.bswap.i32(i32 &lt;id&gt;)
4439 declare i64 @llvm.bswap.i64(i64 &lt;id&gt;)
Nate Begeman0f223bb2006-01-13 23:26:38 +0000[diff] [blame]4440</pre>
4441
4442<h5>Overview:</h5>
4443
4444<p>
4445The '<tt>llvm.bwsap</tt>' family of intrinsics is used to byteswap a 16, 32 or
444664 bit quantity. These are useful for performing operations on data that is not
4447in the target's native byte order.
4448</p>
4449
4450<h5>Semantics:</h5>
4451
4452<p>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]4453The <tt>llvm.bswap.16</tt> intrinsic returns an i16 value that has the high
4454and low byte of the input i16 swapped. Similarly, the <tt>llvm.bswap.i32</tt>
4455intrinsic returns an i32 value that has the four bytes of the input i32
4456swapped, so that if the input bytes are numbered 0, 1, 2, 3 then the returned
4457i32 will have its bytes in 3, 2, 1, 0 order. The <tt>llvm.bswap.i64</tt>
4458intrinsic extends this concept to 64 bits.
Nate Begeman0f223bb2006-01-13 23:26:38 +0000[diff] [blame]4459</p>
4460
4461</div>
4462
4463<!-- _______________________________________________________________________ -->
4464<div class="doc_subsubsection">
Reid Spencerb4f9a6f2006-01-16 21:12:35 +0000[diff] [blame]4465 <a name="int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic</a>
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4466</div>
4467
4468<div class="doc_text">
4469
4470<h5>Syntax:</h5>
4471<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4472 declare i8 @llvm.ctpop.i8 (i8 &lt;src&gt;)
4473 declare i16 @llvm.ctpop.i16(i16 &lt;src&gt;)
4474 declare i32 @llvm.ctpop.i32(i32 &lt;src&gt;)
4475 declare i64 @llvm.ctpop.i64(i64 &lt;src&gt;)
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4476</pre>
4477
4478<h5>Overview:</h5>
4479
4480<p>
Chris Lattner069b5bd2006-01-16 22:38:59 +0000[diff] [blame]4481The '<tt>llvm.ctpop</tt>' family of intrinsics counts the number of bits set in a
4482value.
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4483</p>
4484
4485<h5>Arguments:</h5>
4486
4487<p>
Chris Lattner573f64e2005-05-07 01:46:40 +0000[diff] [blame]4488The only argument is the value to be counted. The argument may be of any
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]4489integer type. The return type must match the argument type.
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4490</p>
4491
4492<h5>Semantics:</h5>
4493
4494<p>
4495The '<tt>llvm.ctpop</tt>' intrinsic counts the 1's in a variable.
4496</p>
4497</div>
4498
4499<!-- _______________________________________________________________________ -->
4500<div class="doc_subsubsection">
Chris Lattnerb748c672006-01-16 22:34:14 +0000[diff] [blame]4501 <a name="int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic</a>
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4502</div>
4503
4504<div class="doc_text">
4505
4506<h5>Syntax:</h5>
4507<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4508 declare i8 @llvm.ctlz.i8 (i8 &lt;src&gt;)
4509 declare i16 @llvm.ctlz.i16(i16 &lt;src&gt;)
4510 declare i32 @llvm.ctlz.i32(i32 &lt;src&gt;)
4511 declare i64 @llvm.ctlz.i64(i64 &lt;src&gt;)
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4512</pre>
4513
4514<h5>Overview:</h5>
4515
4516<p>
Reid Spencerb4f9a6f2006-01-16 21:12:35 +0000[diff] [blame]4517The '<tt>llvm.ctlz</tt>' family of intrinsic functions counts the number of
4518leading zeros in a variable.
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4519</p>
4520
4521<h5>Arguments:</h5>
4522
4523<p>
Chris Lattner573f64e2005-05-07 01:46:40 +0000[diff] [blame]4524The only argument is the value to be counted. The argument may be of any
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]4525integer type. The return type must match the argument type.
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4526</p>
4527
4528<h5>Semantics:</h5>
4529
4530<p>
Chris Lattnerefa20fa2005-05-15 19:39:26 +0000[diff] [blame]4531The '<tt>llvm.ctlz</tt>' intrinsic counts the leading (most significant) zeros
4532in 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]4533of src. For example, <tt>llvm.ctlz(i32 2) = 30</tt>.
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4534</p>
4535</div>
Chris Lattner3b4f4372004-06-11 02:28:03 +0000[diff] [blame]4536
4537
Chris Lattnerefa20fa2005-05-15 19:39:26 +0000[diff] [blame]4538
4539<!-- _______________________________________________________________________ -->
4540<div class="doc_subsubsection">
Chris Lattnerb748c672006-01-16 22:34:14 +0000[diff] [blame]4541 <a name="int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic</a>
Chris Lattnerefa20fa2005-05-15 19:39:26 +0000[diff] [blame]4542</div>
4543
4544<div class="doc_text">
4545
4546<h5>Syntax:</h5>
4547<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4548 declare i8 @llvm.cttz.i8 (i8 &lt;src&gt;)
4549 declare i16 @llvm.cttz.i16(i16 &lt;src&gt;)
4550 declare i32 @llvm.cttz.i32(i32 &lt;src&gt;)
4551 declare i64 @llvm.cttz.i64(i64 &lt;src&gt;)
Chris Lattnerefa20fa2005-05-15 19:39:26 +0000[diff] [blame]4552</pre>
4553
4554<h5>Overview:</h5>
4555
4556<p>
Reid Spencerb4f9a6f2006-01-16 21:12:35 +0000[diff] [blame]4557The '<tt>llvm.cttz</tt>' family of intrinsic functions counts the number of
4558trailing zeros.
Chris Lattnerefa20fa2005-05-15 19:39:26 +0000[diff] [blame]4559</p>
4560
4561<h5>Arguments:</h5>
4562
4563<p>
4564The only argument is the value to be counted. The argument may be of any
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]4565integer type. The return type must match the argument type.
Chris Lattnerefa20fa2005-05-15 19:39:26 +0000[diff] [blame]4566</p>
4567
4568<h5>Semantics:</h5>
4569
4570<p>
4571The '<tt>llvm.cttz</tt>' intrinsic counts the trailing (least significant) zeros
4572in a variable. If the src == 0 then the result is the size in bits of the type
4573of src. For example, <tt>llvm.cttz(2) = 1</tt>.
4574</p>
4575</div>
4576
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]4577<!-- ======================================================================= -->
4578<div class="doc_subsection">
4579 <a name="int_debugger">Debugger Intrinsics</a>
4580</div>
4581
4582<div class="doc_text">
4583<p>
4584The LLVM debugger intrinsics (which all start with <tt>llvm.dbg.</tt> prefix),
4585are described in the <a
4586href="SourceLevelDebugging.html#format_common_intrinsics">LLVM Source Level
4587Debugging</a> document.
4588</p>
4589</div>
4590
4591
Jim Laskey2211f492007-03-14 19:31:19 +0000[diff] [blame]4592<!-- ======================================================================= -->
4593<div class="doc_subsection">
4594 <a name="int_eh">Exception Handling Intrinsics</a>
4595</div>
4596
4597<div class="doc_text">
4598<p> The LLVM exception handling intrinsics (which all start with
4599<tt>llvm.eh.</tt> prefix), are described in the <a
4600href="ExceptionHandling.html#format_common_intrinsics">LLVM Exception
4601Handling</a> document. </p>
4602</div>
4603
4604
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]4605<!-- *********************************************************************** -->
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]4606<hr>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]4607<address>
4608 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
4609 src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
4610 <a href="http://validator.w3.org/check/referer"><img
4611 src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!" /></a>
4612
4613 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
Reid Spencerca058542006-03-14 05:39:39 +0000[diff] [blame]4614 <a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]4615 Last modified: $Date$
4616</address>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]4617</body>
4618</html>