Gitiles
Code Review Sign In
gerrit-public.fairphone.software / toolchain / llvm-project / 2444a9adeb18faef5c08486ea5174cdf8d039b03 / . / llvm / docs / LangRef.html
blob: 8d2dc2f98a87a25ea9dea42922c8df0598bc31da [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>
Anton Korobeynikov546ea7e2007-04-29 18:02:48 +0000[diff] [blame]27 <li><a href="#aliasstructure">Aliases</a>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]28 <li><a href="#paramattrs">Parameter Attributes</a></li>
Chris Lattner91c15c42006-01-23 23:23:47 +0000[diff] [blame]29 <li><a href="#moduleasm">Module-Level Inline Assembly</a></li>
Reid Spencer50c723a2007-02-19 23:54:10 +0000[diff] [blame]30 <li><a href="#datalayout">Data Layout</a></li>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]31 </ol>
32 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]33 <li><a href="#typesystem">Type System</a>
34 <ol>
Robert Bocchino820bc75b2006-02-17 21:18:08 +0000[diff] [blame]35 <li><a href="#t_primitive">Primitive Types</a>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]36 <ol>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]37 <li><a href="#t_classifications">Type Classifications</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]38 </ol>
39 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]40 <li><a href="#t_derived">Derived Types</a>
41 <ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]42 <li><a href="#t_array">Array Type</a></li>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]43 <li><a href="#t_function">Function Type</a></li>
44 <li><a href="#t_pointer">Pointer Type</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]45 <li><a href="#t_struct">Structure Type</a></li>
Andrew Lenharth8df88e22006-12-08 17:13:00 +0000[diff] [blame]46 <li><a href="#t_pstruct">Packed Structure Type</a></li>
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]47 <li><a href="#t_vector">Vector Type</a></li>
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]48 <li><a href="#t_opaque">Opaque Type</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]49 </ol>
50 </li>
51 </ol>
52 </li>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]53 <li><a href="#constants">Constants</a>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]54 <ol>
55 <li><a href="#simpleconstants">Simple Constants</a>
56 <li><a href="#aggregateconstants">Aggregate Constants</a>
57 <li><a href="#globalconstants">Global Variable and Function Addresses</a>
58 <li><a href="#undefvalues">Undefined Values</a>
59 <li><a href="#constantexprs">Constant Expressions</a>
60 </ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]61 </li>
Chris Lattner98f013c2006-01-25 23:47:57 +0000[diff] [blame]62 <li><a href="#othervalues">Other Values</a>
63 <ol>
64 <li><a href="#inlineasm">Inline Assembler Expressions</a>
65 </ol>
66 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]67 <li><a href="#instref">Instruction Reference</a>
68 <ol>
69 <li><a href="#terminators">Terminator Instructions</a>
70 <ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]71 <li><a href="#i_ret">'<tt>ret</tt>' Instruction</a></li>
72 <li><a href="#i_br">'<tt>br</tt>' Instruction</a></li>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]73 <li><a href="#i_switch">'<tt>switch</tt>' Instruction</a></li>
74 <li><a href="#i_invoke">'<tt>invoke</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]75 <li><a href="#i_unwind">'<tt>unwind</tt>' Instruction</a></li>
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]76 <li><a href="#i_unreachable">'<tt>unreachable</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]77 </ol>
78 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]79 <li><a href="#binaryops">Binary Operations</a>
80 <ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]81 <li><a href="#i_add">'<tt>add</tt>' Instruction</a></li>
82 <li><a href="#i_sub">'<tt>sub</tt>' Instruction</a></li>
83 <li><a href="#i_mul">'<tt>mul</tt>' Instruction</a></li>
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame]84 <li><a href="#i_udiv">'<tt>udiv</tt>' Instruction</a></li>
85 <li><a href="#i_sdiv">'<tt>sdiv</tt>' Instruction</a></li>
86 <li><a href="#i_fdiv">'<tt>fdiv</tt>' Instruction</a></li>
Reid Spencer7eb55b32006-11-02 01:53:59 +0000[diff] [blame]87 <li><a href="#i_urem">'<tt>urem</tt>' Instruction</a></li>
88 <li><a href="#i_srem">'<tt>srem</tt>' Instruction</a></li>
89 <li><a href="#i_frem">'<tt>frem</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]90 </ol>
91 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]92 <li><a href="#bitwiseops">Bitwise Binary Operations</a>
93 <ol>
Reid Spencer2ab01932007-02-02 13:57:07 +0000[diff] [blame]94 <li><a href="#i_shl">'<tt>shl</tt>' Instruction</a></li>
95 <li><a href="#i_lshr">'<tt>lshr</tt>' Instruction</a></li>
96 <li><a href="#i_ashr">'<tt>ashr</tt>' Instruction</a></li>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]97 <li><a href="#i_and">'<tt>and</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]98 <li><a href="#i_or">'<tt>or</tt>' Instruction</a></li>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]99 <li><a href="#i_xor">'<tt>xor</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]100 </ol>
101 </li>
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]102 <li><a href="#vectorops">Vector Operations</a>
103 <ol>
104 <li><a href="#i_extractelement">'<tt>extractelement</tt>' Instruction</a></li>
105 <li><a href="#i_insertelement">'<tt>insertelement</tt>' Instruction</a></li>
106 <li><a href="#i_shufflevector">'<tt>shufflevector</tt>' Instruction</a></li>
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]107 </ol>
108 </li>
Chris Lattner6ab66722006-08-15 00:45:58 +0000[diff] [blame]109 <li><a href="#memoryops">Memory Access and Addressing Operations</a>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]110 <ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]111 <li><a href="#i_malloc">'<tt>malloc</tt>' Instruction</a></li>
112 <li><a href="#i_free">'<tt>free</tt>' Instruction</a></li>
113 <li><a href="#i_alloca">'<tt>alloca</tt>' Instruction</a></li>
Robert Bocchino820bc75b2006-02-17 21:18:08 +0000[diff] [blame]114 <li><a href="#i_load">'<tt>load</tt>' Instruction</a></li>
115 <li><a href="#i_store">'<tt>store</tt>' Instruction</a></li>
116 <li><a href="#i_getelementptr">'<tt>getelementptr</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]117 </ol>
118 </li>
Reid Spencer97c5fa42006-11-08 01:18:52 +0000[diff] [blame]119 <li><a href="#convertops">Conversion Operations</a>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]120 <ol>
121 <li><a href="#i_trunc">'<tt>trunc .. to</tt>' Instruction</a></li>
122 <li><a href="#i_zext">'<tt>zext .. to</tt>' Instruction</a></li>
123 <li><a href="#i_sext">'<tt>sext .. to</tt>' Instruction</a></li>
124 <li><a href="#i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a></li>
125 <li><a href="#i_fpext">'<tt>fpext .. to</tt>' Instruction</a></li>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]126 <li><a href="#i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a></li>
127 <li><a href="#i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a></li>
128 <li><a href="#i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a></li>
129 <li><a href="#i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a></li>
Reid Spencerb7344ff2006-11-11 21:00:47 +0000[diff] [blame]130 <li><a href="#i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a></li>
131 <li><a href="#i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a></li>
Reid Spencer5b950642006-11-11 23:08:07 +0000[diff] [blame]132 <li><a href="#i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a></li>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]133 </ol>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]134 <li><a href="#otherops">Other Operations</a>
135 <ol>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]136 <li><a href="#i_icmp">'<tt>icmp</tt>' Instruction</a></li>
137 <li><a href="#i_fcmp">'<tt>fcmp</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]138 <li><a href="#i_phi">'<tt>phi</tt>' Instruction</a></li>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]139 <li><a href="#i_select">'<tt>select</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]140 <li><a href="#i_call">'<tt>call</tt>' Instruction</a></li>
Chris Lattner33337472006-01-13 23:26:01 +0000[diff] [blame]141 <li><a href="#i_va_arg">'<tt>va_arg</tt>' Instruction</a></li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]142 </ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]143 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]144 </ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]145 </li>
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]146 <li><a href="#intrinsics">Intrinsic Functions</a>
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]147 <ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]148 <li><a href="#int_varargs">Variable Argument Handling Intrinsics</a>
149 <ol>
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]150 <li><a href="#int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a></li>
151 <li><a href="#int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a></li>
152 <li><a href="#int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]153 </ol>
154 </li>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]155 <li><a href="#int_gc">Accurate Garbage Collection Intrinsics</a>
156 <ol>
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]157 <li><a href="#int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a></li>
158 <li><a href="#int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a></li>
159 <li><a href="#int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a></li>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]160 </ol>
161 </li>
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]162 <li><a href="#int_codegen">Code Generator Intrinsics</a>
163 <ol>
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]164 <li><a href="#int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a></li>
165 <li><a href="#int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a></li>
166 <li><a href="#int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a></li>
167 <li><a href="#int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a></li>
168 <li><a href="#int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a></li>
169 <li><a href="#int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a></li>
170 <li><a href="#int_readcyclecounter"><tt>llvm.readcyclecounter</tt>' Intrinsic</a></li>
John Criswellaa1c3c12004-04-09 16:43:20 +0000[diff] [blame]171 </ol>
172 </li>
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]173 <li><a href="#int_libc">Standard C Library Intrinsics</a>
174 <ol>
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]175 <li><a href="#int_memcpy">'<tt>llvm.memcpy.*</tt>' Intrinsic</a></li>
176 <li><a href="#int_memmove">'<tt>llvm.memmove.*</tt>' Intrinsic</a></li>
177 <li><a href="#int_memset">'<tt>llvm.memset.*</tt>' Intrinsic</a></li>
178 <li><a href="#int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a></li>
179 <li><a href="#int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a></li>
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]180 </ol>
181 </li>
Nate Begeman0f223bb2006-01-13 23:26:38 +0000[diff] [blame]182 <li><a href="#int_manip">Bit Manipulation Intrinsics</a>
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]183 <ol>
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]184 <li><a href="#int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a></li>
Chris Lattnerb748c672006-01-16 22:34:14 +0000[diff] [blame]185 <li><a href="#int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic </a></li>
186 <li><a href="#int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic </a></li>
187 <li><a href="#int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic </a></li>
Reid Spencer5bf54c82007-04-11 23:23:49 +0000[diff] [blame]188 <li><a href="#int_part_select">'<tt>llvm.part.select.*</tt>' Intrinsic </a></li>
189 <li><a href="#int_part_set">'<tt>llvm.part.set.*</tt>' Intrinsic </a></li>
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]190 </ol>
191 </li>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]192 <li><a href="#int_debugger">Debugger intrinsics</a></li>
Jim Laskey2211f492007-03-14 19:31:19 +0000[diff] [blame]193 <li><a href="#int_eh">Exception Handling intrinsics</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]194 </ol>
195 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]196</ol>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]197
198<div class="doc_author">
199 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>
200 and <a href="mailto:vadve@cs.uiuc.edu">Vikram Adve</a></p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]201</div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]202
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]203<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]204<div class="doc_section"> <a name="abstract">Abstract </a></div>
205<!-- *********************************************************************** -->
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]206
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]207<div class="doc_text">
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]208<p>This document is a reference manual for the LLVM assembly language.
209LLVM is an SSA based representation that provides type safety,
210low-level operations, flexibility, and the capability of representing
211'all' high-level languages cleanly. It is the common code
212representation used throughout all phases of the LLVM compilation
213strategy.</p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]214</div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]215
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]216<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]217<div class="doc_section"> <a name="introduction">Introduction</a> </div>
218<!-- *********************************************************************** -->
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]219
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]220<div class="doc_text">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]221
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]222<p>The LLVM code representation is designed to be used in three
223different forms: as an in-memory compiler IR, as an on-disk bytecode
224representation (suitable for fast loading by a Just-In-Time compiler),
225and as a human readable assembly language representation. This allows
226LLVM to provide a powerful intermediate representation for efficient
227compiler transformations and analysis, while providing a natural means
228to debug and visualize the transformations. The three different forms
229of LLVM are all equivalent. This document describes the human readable
230representation and notation.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]231
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]232<p>The LLVM representation aims to be light-weight and low-level
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]233while being expressive, typed, and extensible at the same time. It
234aims to be a "universal IR" of sorts, by being at a low enough level
235that high-level ideas may be cleanly mapped to it (similar to how
236microprocessors are "universal IR's", allowing many source languages to
237be mapped to them). By providing type information, LLVM can be used as
238the target of optimizations: for example, through pointer analysis, it
239can be proven that a C automatic variable is never accessed outside of
240the current function... allowing it to be promoted to a simple SSA
241value instead of a memory location.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]242
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]243</div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]244
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]245<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]246<div class="doc_subsubsection"> <a name="wellformed">Well-Formedness</a> </div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]247
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]248<div class="doc_text">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]249
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]250<p>It is important to note that this document describes 'well formed'
251LLVM assembly language. There is a difference between what the parser
252accepts and what is considered 'well formed'. For example, the
253following instruction is syntactically okay, but not well formed:</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]254
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]255<div class="doc_code">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]256<pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]257%x = <a href="#i_add">add</a> i32 1, %x
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]258</pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]259</div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]260
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]261<p>...because the definition of <tt>%x</tt> does not dominate all of
262its uses. The LLVM infrastructure provides a verification pass that may
263be used to verify that an LLVM module is well formed. This pass is
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]264automatically run by the parser after parsing input assembly and by
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]265the optimizer before it outputs bytecode. The violations pointed out
266by the verifier pass indicate bugs in transformation passes or input to
267the parser.</p>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]268</div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]269
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]270<!-- Describe the typesetting conventions here. --> </div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]271
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]272<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]273<div class="doc_section"> <a name="identifiers">Identifiers</a> </div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]274<!-- *********************************************************************** -->
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]275
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]276<div class="doc_text">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]277
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]278<p>LLVM uses three different forms of identifiers, for different
279purposes:</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]280
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]281<ol>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]282 <li>Named values are represented as a string of characters with a '%' prefix.
283 For example, %foo, %DivisionByZero, %a.really.long.identifier. The actual
284 regular expression used is '<tt>%[a-zA-Z$._][a-zA-Z$._0-9]*</tt>'.
285 Identifiers which require other characters in their names can be surrounded
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]286 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]287 in a name.</li>
288
289 <li>Unnamed values are represented as an unsigned numeric value with a '%'
290 prefix. For example, %12, %2, %44.</li>
291
Reid Spencer8f08d802004-12-09 18:02:53 +0000[diff] [blame]292 <li>Constants, which are described in a <a href="#constants">section about
293 constants</a>, below.</li>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]294</ol>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]295
296<p>LLVM requires that values start with a '%' sign for two reasons: Compilers
297don't need to worry about name clashes with reserved words, and the set of
298reserved words may be expanded in the future without penalty. Additionally,
299unnamed identifiers allow a compiler to quickly come up with a temporary
300variable without having to avoid symbol table conflicts.</p>
301
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]302<p>Reserved words in LLVM are very similar to reserved words in other
Reid Spencer5b950642006-11-11 23:08:07 +0000[diff] [blame]303languages. There are keywords for different opcodes
304('<tt><a href="#i_add">add</a></tt>',
305 '<tt><a href="#i_bitcast">bitcast</a></tt>',
306 '<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]307href="#t_void">void</a></tt>', '<tt><a href="#t_primitive">i32</a></tt>', etc...),
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]308and others. These reserved words cannot conflict with variable names, because
309none of them start with a '%' character.</p>
310
311<p>Here is an example of LLVM code to multiply the integer variable
312'<tt>%X</tt>' by 8:</p>
313
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]314<p>The easy way:</p>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]315
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]316<div class="doc_code">
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]317<pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]318%result = <a href="#i_mul">mul</a> i32 %X, 8
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]319</pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]320</div>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]321
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]322<p>After strength reduction:</p>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]323
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]324<div class="doc_code">
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]325<pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]326%result = <a href="#i_shl">shl</a> i32 %X, i8 3
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]327</pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]328</div>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]329
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]330<p>And the hard way:</p>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]331
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]332<div class="doc_code">
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]333<pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]334<a href="#i_add">add</a> i32 %X, %X <i>; yields {i32}:%0</i>
335<a href="#i_add">add</a> i32 %0, %0 <i>; yields {i32}:%1</i>
336%result = <a href="#i_add">add</a> i32 %1, %1
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]337</pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]338</div>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]339
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]340<p>This last way of multiplying <tt>%X</tt> by 8 illustrates several
341important lexical features of LLVM:</p>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]342
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]343<ol>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]344
345 <li>Comments are delimited with a '<tt>;</tt>' and go until the end of
346 line.</li>
347
348 <li>Unnamed temporaries are created when the result of a computation is not
349 assigned to a named value.</li>
350
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]351 <li>Unnamed temporaries are numbered sequentially</li>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]352
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]353</ol>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]354
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]355<p>...and it also shows a convention that we follow in this document. When
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]356demonstrating instructions, we will follow an instruction with a comment that
357defines the type and name of value produced. Comments are shown in italic
358text.</p>
359
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]360</div>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]361
362<!-- *********************************************************************** -->
363<div class="doc_section"> <a name="highlevel">High Level Structure</a> </div>
364<!-- *********************************************************************** -->
365
366<!-- ======================================================================= -->
367<div class="doc_subsection"> <a name="modulestructure">Module Structure</a>
368</div>
369
370<div class="doc_text">
371
372<p>LLVM programs are composed of "Module"s, each of which is a
373translation unit of the input programs. Each module consists of
374functions, global variables, and symbol table entries. Modules may be
375combined together with the LLVM linker, which merges function (and
376global variable) definitions, resolves forward declarations, and merges
377symbol table entries. Here is an example of the "hello world" module:</p>
378
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]379<div class="doc_code">
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]380<pre><i>; Declare the string constant as a global constant...</i>
381<a href="#identifiers">%.LC0</a> = <a href="#linkage_internal">internal</a> <a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]382 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]383
384<i>; External declaration of the puts function</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]385<a href="#functionstructure">declare</a> i32 %puts(i8 *) <i>; i32(i8 *)* </i>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]386
387<i>; Definition of main function</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]388define i32 %main() { <i>; i32()* </i>
389 <i>; Convert [13x i8 ]* to i8 *...</i>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]390 %cast210 = <a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]391 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]392
393 <i>; Call puts function to write out the string to stdout...</i>
394 <a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]395 href="#i_call">call</a> i32 %puts(i8 * %cast210) <i>; i32</i>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]396 <a
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]397 href="#i_ret">ret</a> i32 0<br>}<br>
398</pre>
399</div>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]400
401<p>This example is made up of a <a href="#globalvars">global variable</a>
402named "<tt>.LC0</tt>", an external declaration of the "<tt>puts</tt>"
403function, and a <a href="#functionstructure">function definition</a>
404for "<tt>main</tt>".</p>
405
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]406<p>In general, a module is made up of a list of global values,
407where both functions and global variables are global values. Global values are
408represented by a pointer to a memory location (in this case, a pointer to an
409array of char, and a pointer to a function), and have one of the following <a
410href="#linkage">linkage types</a>.</p>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]411
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]412</div>
413
414<!-- ======================================================================= -->
415<div class="doc_subsection">
416 <a name="linkage">Linkage Types</a>
417</div>
418
419<div class="doc_text">
420
421<p>
422All Global Variables and Functions have one of the following types of linkage:
423</p>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]424
425<dl>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]426
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]427 <dt><tt><b><a name="linkage_internal">internal</a></b></tt> </dt>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]428
429 <dd>Global values with internal linkage are only directly accessible by
430 objects in the current module. In particular, linking code into a module with
431 an internal global value may cause the internal to be renamed as necessary to
432 avoid collisions. Because the symbol is internal to the module, all
433 references can be updated. This corresponds to the notion of the
Chris Lattnere20b4702007-01-14 06:51:48 +0000[diff] [blame]434 '<tt>static</tt>' keyword in C.
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]435 </dd>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]436
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]437 <dt><tt><b><a name="linkage_linkonce">linkonce</a></b></tt>: </dt>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]438
Chris Lattnere20b4702007-01-14 06:51:48 +0000[diff] [blame]439 <dd>Globals with "<tt>linkonce</tt>" linkage are merged with other globals of
440 the same name when linkage occurs. This is typically used to implement
441 inline functions, templates, or other code which must be generated in each
442 translation unit that uses it. Unreferenced <tt>linkonce</tt> globals are
443 allowed to be discarded.
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]444 </dd>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]445
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]446 <dt><tt><b><a name="linkage_weak">weak</a></b></tt>: </dt>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]447
448 <dd>"<tt>weak</tt>" linkage is exactly the same as <tt>linkonce</tt> linkage,
449 except that unreferenced <tt>weak</tt> globals may not be discarded. This is
Chris Lattnere20b4702007-01-14 06:51:48 +0000[diff] [blame]450 used for globals that may be emitted in multiple translation units, but that
451 are not guaranteed to be emitted into every translation unit that uses them.
452 One example of this are common globals in C, such as "<tt>int X;</tt>" at
453 global scope.
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]454 </dd>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]455
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]456 <dt><tt><b><a name="linkage_appending">appending</a></b></tt>: </dt>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]457
458 <dd>"<tt>appending</tt>" linkage may only be applied to global variables of
459 pointer to array type. When two global variables with appending linkage are
460 linked together, the two global arrays are appended together. This is the
461 LLVM, typesafe, equivalent of having the system linker append together
462 "sections" with identical names when .o files are linked.
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]463 </dd>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]464
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]465 <dt><tt><b><a name="linkage_externweak">extern_weak</a></b></tt>: </dt>
466 <dd>The semantics of this linkage follow the ELF model: the symbol is weak
467 until linked, if not linked, the symbol becomes null instead of being an
468 undefined reference.
469 </dd>
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]470
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]471 <dt><tt><b><a name="linkage_external">externally visible</a></b></tt>:</dt>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]472
473 <dd>If none of the above identifiers are used, the global is externally
474 visible, meaning that it participates in linkage and can be used to resolve
475 external symbol references.
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]476 </dd>
Reid Spencer7972c472007-04-11 23:49:50 +0000[diff] [blame]477</dl>
Anton Korobeynikovd61d39e2006-09-14 18:23:27 +0000[diff] [blame]478
Anton Korobeynikovd61d39e2006-09-14 18:23:27 +0000[diff] [blame]479 <p>
480 The next two types of linkage are targeted for Microsoft Windows platform
481 only. They are designed to support importing (exporting) symbols from (to)
482 DLLs.
483 </p>
484
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]485 <dl>
Anton Korobeynikovd61d39e2006-09-14 18:23:27 +0000[diff] [blame]486 <dt><tt><b><a name="linkage_dllimport">dllimport</a></b></tt>: </dt>
487
488 <dd>"<tt>dllimport</tt>" linkage causes the compiler to reference a function
489 or variable via a global pointer to a pointer that is set up by the DLL
490 exporting the symbol. On Microsoft Windows targets, the pointer name is
491 formed by combining <code>_imp__</code> and the function or variable name.
492 </dd>
493
494 <dt><tt><b><a name="linkage_dllexport">dllexport</a></b></tt>: </dt>
495
496 <dd>"<tt>dllexport</tt>" linkage causes the compiler to provide a global
497 pointer to a pointer in a DLL, so that it can be referenced with the
498 <tt>dllimport</tt> attribute. On Microsoft Windows targets, the pointer
499 name is formed by combining <code>_imp__</code> and the function or variable
500 name.
501 </dd>
502
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]503</dl>
504
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]505<p><a name="linkage_external"></a>For example, since the "<tt>.LC0</tt>"
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]506variable is defined to be internal, if another module defined a "<tt>.LC0</tt>"
507variable and was linked with this one, one of the two would be renamed,
508preventing a collision. Since "<tt>main</tt>" and "<tt>puts</tt>" are
509external (i.e., lacking any linkage declarations), they are accessible
Reid Spencer92c671e2007-01-05 00:59:10 +0000[diff] [blame]510outside of the current module.</p>
511<p>It is illegal for a function <i>declaration</i>
512to have any linkage type other than "externally visible", <tt>dllimport</tt>,
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]513or <tt>extern_weak</tt>.</p>
Anton Korobeynikova97b6942007-04-25 14:27:10 +0000[diff] [blame]514<p>Aliases can have only <tt>external</tt>, <tt>internal</tt> and <tt>weak</tt>
515linkages.
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]516</div>
517
518<!-- ======================================================================= -->
519<div class="doc_subsection">
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]520 <a name="callingconv">Calling Conventions</a>
521</div>
522
523<div class="doc_text">
524
525<p>LLVM <a href="#functionstructure">functions</a>, <a href="#i_call">calls</a>
526and <a href="#i_invoke">invokes</a> can all have an optional calling convention
527specified for the call. The calling convention of any pair of dynamic
528caller/callee must match, or the behavior of the program is undefined. The
529following calling conventions are supported by LLVM, and more may be added in
530the future:</p>
531
532<dl>
533 <dt><b>"<tt>ccc</tt>" - The C calling convention</b>:</dt>
534
535 <dd>This calling convention (the default if no other calling convention is
536 specified) matches the target C calling conventions. This calling convention
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]537 supports varargs function calls and tolerates some mismatch in the declared
Reid Spencer72ba4992006-12-31 21:30:18 +0000[diff] [blame]538 prototype and implemented declaration of the function (as does normal C).
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]539 </dd>
540
541 <dt><b>"<tt>fastcc</tt>" - The fast calling convention</b>:</dt>
542
543 <dd>This calling convention attempts to make calls as fast as possible
544 (e.g. by passing things in registers). This calling convention allows the
545 target to use whatever tricks it wants to produce fast code for the target,
Chris Lattnerc792eb32005-05-06 23:08:23 +0000[diff] [blame]546 without having to conform to an externally specified ABI. Implementations of
547 this convention should allow arbitrary tail call optimization to be supported.
548 This calling convention does not support varargs and requires the prototype of
549 all callees to exactly match the prototype of the function definition.
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]550 </dd>
551
552 <dt><b>"<tt>coldcc</tt>" - The cold calling convention</b>:</dt>
553
554 <dd>This calling convention attempts to make code in the caller as efficient
555 as possible under the assumption that the call is not commonly executed. As
556 such, these calls often preserve all registers so that the call does not break
557 any live ranges in the caller side. This calling convention does not support
558 varargs and requires the prototype of all callees to exactly match the
559 prototype of the function definition.
560 </dd>
561
Chris Lattner573f64e2005-05-07 01:46:40 +0000[diff] [blame]562 <dt><b>"<tt>cc &lt;<em>n</em>&gt;</tt>" - Numbered convention</b>:</dt>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]563
564 <dd>Any calling convention may be specified by number, allowing
565 target-specific calling conventions to be used. Target specific calling
566 conventions start at 64.
567 </dd>
Chris Lattner573f64e2005-05-07 01:46:40 +0000[diff] [blame]568</dl>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]569
570<p>More calling conventions can be added/defined on an as-needed basis, to
571support pascal conventions or any other well-known target-independent
572convention.</p>
573
574</div>
575
576<!-- ======================================================================= -->
577<div class="doc_subsection">
Anton Korobeynikovc7f9f3d2007-01-23 12:35:46 +0000[diff] [blame]578 <a name="visibility">Visibility Styles</a>
579</div>
580
581<div class="doc_text">
582
583<p>
584All Global Variables and Functions have one of the following visibility styles:
585</p>
586
587<dl>
588 <dt><b>"<tt>default</tt>" - Default style</b>:</dt>
589
590 <dd>On ELF, default visibility means that the declaration is visible to other
591 modules and, in shared libraries, means that the declared entity may be
592 overridden. On Darwin, default visibility means that the declaration is
593 visible to other modules. Default visibility corresponds to "external
594 linkage" in the language.
595 </dd>
596
597 <dt><b>"<tt>hidden</tt>" - Hidden style</b>:</dt>
598
599 <dd>Two declarations of an object with hidden visibility refer to the same
600 object if they are in the same shared object. Usually, hidden visibility
601 indicates that the symbol will not be placed into the dynamic symbol table,
602 so no other module (executable or shared library) can reference it
603 directly.
604 </dd>
605
Anton Korobeynikov39f3cff2007-04-29 18:35:00 +0000[diff] [blame]606 <dt><b>"<tt>protected</tt>" - Protected style</b>:</dt>
607
608 <dd>On ELF, protected visibility indicates that the symbol will be placed in
609 the dynamic symbol table, but that references within the defining module will
610 bind to the local symbol. That is, the symbol cannot be overridden by another
611 module.
612 </dd>
Anton Korobeynikovc7f9f3d2007-01-23 12:35:46 +0000[diff] [blame]613</dl>
614
615</div>
616
617<!-- ======================================================================= -->
618<div class="doc_subsection">
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]619 <a name="globalvars">Global Variables</a>
620</div>
621
622<div class="doc_text">
623
Chris Lattner5d5aede2005-02-12 19:30:21 +0000[diff] [blame]624<p>Global variables define regions of memory allocated at compilation time
Chris Lattner662c8722005-11-12 00:45:07 +0000[diff] [blame]625instead of run-time. Global variables may optionally be initialized, may have
Anton Korobeynikova97b6942007-04-25 14:27:10 +0000[diff] [blame]626an explicit section to be placed in, and may have an optional explicit alignment
627specified. A variable may be defined as "thread_local", which means that it
628will not be shared by threads (each thread will have a separated copy of the
629variable). A variable may be defined as a global "constant," which indicates
630that the contents of the variable will <b>never</b> be modified (enabling better
Chris Lattner5d5aede2005-02-12 19:30:21 +0000[diff] [blame]631optimization, allowing the global data to be placed in the read-only section of
632an executable, etc). Note that variables that need runtime initialization
John Criswell4c0cf7f2005-10-24 16:17:18 +0000[diff] [blame]633cannot be marked "constant" as there is a store to the variable.</p>
Chris Lattner5d5aede2005-02-12 19:30:21 +0000[diff] [blame]634
635<p>
636LLVM explicitly allows <em>declarations</em> of global variables to be marked
637constant, even if the final definition of the global is not. This capability
638can be used to enable slightly better optimization of the program, but requires
639the language definition to guarantee that optimizations based on the
640'constantness' are valid for the translation units that do not include the
641definition.
642</p>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]643
644<p>As SSA values, global variables define pointer values that are in
645scope (i.e. they dominate) all basic blocks in the program. Global
646variables always define a pointer to their "content" type because they
647describe a region of memory, and all memory objects in LLVM are
648accessed through pointers.</p>
649
Chris Lattner662c8722005-11-12 00:45:07 +0000[diff] [blame]650<p>LLVM allows an explicit section to be specified for globals. If the target
651supports it, it will emit globals to the section specified.</p>
652
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]653<p>An explicit alignment may be specified for a global. If not present, or if
654the alignment is set to zero, the alignment of the global is set by the target
655to whatever it feels convenient. If an explicit alignment is specified, the
656global is forced to have at least that much alignment. All alignments must be
657a power of 2.</p>
658
Chris Lattner5760c502007-01-14 00:27:09 +0000[diff] [blame]659<p>For example, the following defines a global with an initializer, section,
660 and alignment:</p>
661
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]662<div class="doc_code">
Chris Lattner5760c502007-01-14 00:27:09 +0000[diff] [blame]663<pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]664%G = constant float 1.0, section "foo", align 4
Chris Lattner5760c502007-01-14 00:27:09 +0000[diff] [blame]665</pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]666</div>
Chris Lattner5760c502007-01-14 00:27:09 +0000[diff] [blame]667
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]668</div>
669
670
671<!-- ======================================================================= -->
672<div class="doc_subsection">
673 <a name="functionstructure">Functions</a>
674</div>
675
676<div class="doc_text">
677
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]678<p>LLVM function definitions consist of the "<tt>define</tt>" keyord,
679an optional <a href="#linkage">linkage type</a>, an optional
Anton Korobeynikovc7f9f3d2007-01-23 12:35:46 +0000[diff] [blame]680<a href="#visibility">visibility style</a>, an optional
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]681<a href="#callingconv">calling convention</a>, a return type, an optional
682<a href="#paramattrs">parameter attribute</a> for the return type, a function
683name, a (possibly empty) argument list (each with optional
Anton Korobeynikovc7f9f3d2007-01-23 12:35:46 +0000[diff] [blame]684<a href="#paramattrs">parameter attributes</a>), an optional section, an
685optional alignment, an opening curly brace, a list of basic blocks, and a
686closing curly brace.
687
688LLVM function declarations consist of the "<tt>declare</tt>" keyword, an
689optional <a href="#linkage">linkage type</a>, an optional
690<a href="#visibility">visibility style</a>, an optional
691<a href="#callingconv">calling convention</a>, a return type, an optional
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]692<a href="#paramattrs">parameter attribute</a> for the return type, a function
693name, a possibly empty list of arguments, and an optional alignment.</p>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]694
695<p>A function definition contains a list of basic blocks, forming the CFG for
696the function. Each basic block may optionally start with a label (giving the
697basic block a symbol table entry), contains a list of instructions, and ends
698with a <a href="#terminators">terminator</a> instruction (such as a branch or
699function return).</p>
700
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]701<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]702executed on entrance to the function, and it is not allowed to have predecessor
703basic blocks (i.e. there can not be any branches to the entry block of a
704function). Because the block can have no predecessors, it also cannot have any
705<a href="#i_phi">PHI nodes</a>.</p>
706
707<p>LLVM functions are identified by their name and type signature. Hence, two
708functions with the same name but different parameter lists or return values are
Chris Lattner455fc8c2005-03-07 22:13:59 +0000[diff] [blame]709considered different functions, and LLVM will resolve references to each
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]710appropriately.</p>
711
Chris Lattner662c8722005-11-12 00:45:07 +0000[diff] [blame]712<p>LLVM allows an explicit section to be specified for functions. If the target
713supports it, it will emit functions to the section specified.</p>
714
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]715<p>An explicit alignment may be specified for a function. If not present, or if
716the alignment is set to zero, the alignment of the function is set by the target
717to whatever it feels convenient. If an explicit alignment is specified, the
718function is forced to have at least that much alignment. All alignments must be
719a power of 2.</p>
720
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]721</div>
722
Anton Korobeynikova97b6942007-04-25 14:27:10 +0000[diff] [blame]723
724<!-- ======================================================================= -->
725<div class="doc_subsection">
726 <a name="aliasstructure">Aliases</a>
727</div>
728<div class="doc_text">
729 <p>Aliases act as "second name" for the aliasee value (which can be either
Anton Korobeynikovb18f8f82007-04-28 13:45:00 +0000[diff] [blame]730 function or global variable or bitcast of global value). Aliases may have an
Anton Korobeynikova97b6942007-04-25 14:27:10 +0000[diff] [blame]731 optional <a href="#linkage">linkage type</a>, and an
732 optional <a href="#visibility">visibility style</a>.</p>
733
734 <h5>Syntax:</h5>
735
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]736<div class="doc_code">
Bill Wendling2d8b9a82007-05-29 09:42:13 +0000[diff] [blame]737<pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]738@&lt;Name&gt; = [Linkage] [Visibility] alias &lt;AliaseeTy&gt; @&lt;Aliasee&gt;
Bill Wendling2d8b9a82007-05-29 09:42:13 +0000[diff] [blame]739</pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]740</div>
Anton Korobeynikova97b6942007-04-25 14:27:10 +0000[diff] [blame]741
742</div>
743
744
745
Chris Lattner91c15c42006-01-23 23:23:47 +0000[diff] [blame]746<!-- ======================================================================= -->
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]747<div class="doc_subsection"><a name="paramattrs">Parameter Attributes</a></div>
748<div class="doc_text">
749 <p>The return type and each parameter of a function type may have a set of
750 <i>parameter attributes</i> associated with them. Parameter attributes are
751 used to communicate additional information about the result or parameters of
752 a function. Parameter attributes are considered to be part of the function
753 type so two functions types that differ only by the parameter attributes
754 are different function types.</p>
755
Reid Spencercf7ebf52007-01-15 18:27:39 +0000[diff] [blame]756 <p>Parameter attributes are simple keywords that follow the type specified. If
757 multiple parameter attributes are needed, they are space separated. For
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]758 example:</p>
759
760<div class="doc_code">
761<pre>
762%someFunc = i16 (i8 sext %someParam) zext
763%someFunc = i16 (i8 zext %someParam) zext
764</pre>
765</div>
766
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]767 <p>Note that the two function types above are unique because the parameter has
Reid Spencercf7ebf52007-01-15 18:27:39 +0000[diff] [blame]768 a different attribute (sext in the first one, zext in the second). Also note
769 that the attribute for the function result (zext) comes immediately after the
770 argument list.</p>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]771
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]772 <p>Currently, only the following parameter attributes are defined:</p>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]773 <dl>
Reid Spencercf7ebf52007-01-15 18:27:39 +0000[diff] [blame]774 <dt><tt>zext</tt></dt>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]775 <dd>This indicates that the parameter should be zero extended just before
776 a call to this function.</dd>
Reid Spencercf7ebf52007-01-15 18:27:39 +0000[diff] [blame]777 <dt><tt>sext</tt></dt>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]778 <dd>This indicates that the parameter should be sign extended just before
779 a call to this function.</dd>
Anton Korobeynikove8166852007-01-28 14:30:45 +0000[diff] [blame]780 <dt><tt>inreg</tt></dt>
781 <dd>This indicates that the parameter should be placed in register (if
Anton Korobeynikove93c6e82007-01-28 15:27:21 +0000[diff] [blame]782 possible) during assembling function call. Support for this attribute is
783 target-specific</dd>
Anton Korobeynikove8166852007-01-28 14:30:45 +0000[diff] [blame]784 <dt><tt>sret</tt></dt>
Anton Korobeynikove93c6e82007-01-28 15:27:21 +0000[diff] [blame]785 <dd>This indicates that the parameter specifies the address of a structure
Reid Spencer05dbb9d2007-03-22 02:02:11 +0000[diff] [blame]786 that is the return value of the function in the source program.</dd>
Zhou Sheng2444a9a2007-06-05 05:28:26 +0000[diff] [blame^]787 <dt><tt>noalias</tt></dt>
788 <dd>This indicates that the parameter not alias any other object or any
789 other "noalias" objects during the function call.
Reid Spencer9d1700e2007-03-22 02:18:56 +0000[diff] [blame]790 <dt><tt>noreturn</tt></dt>
791 <dd>This function attribute indicates that the function never returns. This
792 indicates to LLVM that every call to this function should be treated as if
793 an <tt>unreachable</tt> instruction immediately followed the call.</dd>
Reid Spencer05dbb9d2007-03-22 02:02:11 +0000[diff] [blame]794 <dt><tt>nounwind</tt></dt>
795 <dd>This function attribute indicates that the function type does not use
796 the unwind instruction and does not allow stack unwinding to propagate
797 through it.</dd>
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]798 </dl>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]799
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]800</div>
801
802<!-- ======================================================================= -->
Chris Lattner91c15c42006-01-23 23:23:47 +0000[diff] [blame]803<div class="doc_subsection">
Chris Lattner93564892006-04-08 04:40:53 +0000[diff] [blame]804 <a name="moduleasm">Module-Level Inline Assembly</a>
Chris Lattner91c15c42006-01-23 23:23:47 +0000[diff] [blame]805</div>
806
807<div class="doc_text">
808<p>
809Modules may contain "module-level inline asm" blocks, which corresponds to the
810GCC "file scope inline asm" blocks. These blocks are internally concatenated by
811LLVM and treated as a single unit, but may be separated in the .ll file if
812desired. The syntax is very simple:
813</p>
814
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]815<div class="doc_code">
816<pre>
817module asm "inline asm code goes here"
818module asm "more can go here"
819</pre>
820</div>
Chris Lattner91c15c42006-01-23 23:23:47 +0000[diff] [blame]821
822<p>The strings can contain any character by escaping non-printable characters.
823 The escape sequence used is simply "\xx" where "xx" is the two digit hex code
824 for the number.
825</p>
826
827<p>
828 The inline asm code is simply printed to the machine code .s file when
829 assembly code is generated.
830</p>
831</div>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]832
Reid Spencer50c723a2007-02-19 23:54:10 +0000[diff] [blame]833<!-- ======================================================================= -->
834<div class="doc_subsection">
835 <a name="datalayout">Data Layout</a>
836</div>
837
838<div class="doc_text">
839<p>A module may specify a target specific data layout string that specifies how
Reid Spencer7972c472007-04-11 23:49:50 +0000[diff] [blame]840data is to be laid out in memory. The syntax for the data layout is simply:</p>
841<pre> target datalayout = "<i>layout specification</i>"</pre>
842<p>The <i>layout specification</i> consists of a list of specifications
843separated by the minus sign character ('-'). Each specification starts with a
844letter and may include other information after the letter to define some
845aspect of the data layout. The specifications accepted are as follows: </p>
Reid Spencer50c723a2007-02-19 23:54:10 +0000[diff] [blame]846<dl>
847 <dt><tt>E</tt></dt>
848 <dd>Specifies that the target lays out data in big-endian form. That is, the
849 bits with the most significance have the lowest address location.</dd>
850 <dt><tt>e</tt></dt>
851 <dd>Specifies that hte target lays out data in little-endian form. That is,
852 the bits with the least significance have the lowest address location.</dd>
853 <dt><tt>p:<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
854 <dd>This specifies the <i>size</i> of a pointer and its <i>abi</i> and
855 <i>preferred</i> alignments. All sizes are in bits. Specifying the <i>pref</i>
856 alignment is optional. If omitted, the preceding <tt>:</tt> should be omitted
857 too.</dd>
858 <dt><tt>i<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
859 <dd>This specifies the alignment for an integer type of a given bit
860 <i>size</i>. The value of <i>size</i> must be in the range [1,2^23).</dd>
861 <dt><tt>v<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
862 <dd>This specifies the alignment for a vector type of a given bit
863 <i>size</i>.</dd>
864 <dt><tt>f<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
865 <dd>This specifies the alignment for a floating point type of a given bit
866 <i>size</i>. The value of <i>size</i> must be either 32 (float) or 64
867 (double).</dd>
868 <dt><tt>a<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
869 <dd>This specifies the alignment for an aggregate type of a given bit
870 <i>size</i>.</dd>
871</dl>
872<p>When constructing the data layout for a given target, LLVM starts with a
873default set of specifications which are then (possibly) overriden by the
874specifications in the <tt>datalayout</tt> keyword. The default specifications
875are given in this list:</p>
876<ul>
877 <li><tt>E</tt> - big endian</li>
878 <li><tt>p:32:64:64</tt> - 32-bit pointers with 64-bit alignment</li>
879 <li><tt>i1:8:8</tt> - i1 is 8-bit (byte) aligned</li>
880 <li><tt>i8:8:8</tt> - i8 is 8-bit (byte) aligned</li>
881 <li><tt>i16:16:16</tt> - i16 is 16-bit aligned</li>
882 <li><tt>i32:32:32</tt> - i32 is 32-bit aligned</li>
883 <li><tt>i64:32:64</tt> - i64 has abi alignment of 32-bits but preferred
884 alignment of 64-bits</li>
885 <li><tt>f32:32:32</tt> - float is 32-bit aligned</li>
886 <li><tt>f64:64:64</tt> - double is 64-bit aligned</li>
887 <li><tt>v64:64:64</tt> - 64-bit vector is 64-bit aligned</li>
888 <li><tt>v128:128:128</tt> - 128-bit vector is 128-bit aligned</li>
889 <li><tt>a0:0:1</tt> - aggregates are 8-bit aligned</li>
890</ul>
891<p>When llvm is determining the alignment for a given type, it uses the
892following rules:
893<ol>
894 <li>If the type sought is an exact match for one of the specifications, that
895 specification is used.</li>
896 <li>If no match is found, and the type sought is an integer type, then the
897 smallest integer type that is larger than the bitwidth of the sought type is
898 used. If none of the specifications are larger than the bitwidth then the the
899 largest integer type is used. For example, given the default specifications
900 above, the i7 type will use the alignment of i8 (next largest) while both
901 i65 and i256 will use the alignment of i64 (largest specified).</li>
902 <li>If no match is found, and the type sought is a vector type, then the
903 largest vector type that is smaller than the sought vector type will be used
904 as a fall back. This happens because <128 x double> can be implemented in
905 terms of 64 <2 x double>, for example.</li>
906</ol>
907</div>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]908
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]909<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]910<div class="doc_section"> <a name="typesystem">Type System</a> </div>
911<!-- *********************************************************************** -->
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]912
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]913<div class="doc_text">
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]914
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]915<p>The LLVM type system is one of the most important features of the
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]916intermediate representation. Being typed enables a number of
917optimizations to be performed on the IR directly, without having to do
918extra analyses on the side before the transformation. A strong type
919system makes it easier to read the generated code and enables novel
920analyses and transformations that are not feasible to perform on normal
921three address code representations.</p>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]922
923</div>
924
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]925<!-- ======================================================================= -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]926<div class="doc_subsection"> <a name="t_primitive">Primitive Types</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]927<div class="doc_text">
John Criswell417228d2004-04-09 16:48:45 +0000[diff] [blame]928<p>The primitive types are the fundamental building blocks of the LLVM
Chris Lattner455fc8c2005-03-07 22:13:59 +0000[diff] [blame]929system. The current set of primitive types is as follows:</p>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]930
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]931<table class="layout">
932 <tr class="layout">
933 <td class="left">
934 <table>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]935 <tbody>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]936 <tr><th>Type</th><th>Description</th></tr>
Duncan Sands16f122e2007-03-30 12:22:09 +0000[diff] [blame]937 <tr><td><tt><a name="t_void">void</a></tt></td><td>No value</td></tr>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]938 <tr><td><tt>label</tt></td><td>Branch destination</td></tr>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]939 </tbody>
940 </table>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]941 </td>
942 <td class="right">
943 <table>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]944 <tbody>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]945 <tr><th>Type</th><th>Description</th></tr>
Reid Spencer138249b2007-05-16 18:44:01 +0000[diff] [blame]946 <tr><td><tt>float</tt></td><td>32-bit floating point value</td></tr>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]947 <tr><td><tt>double</tt></td><td>64-bit floating point value</td></tr>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]948 </tbody>
949 </table>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]950 </td>
951 </tr>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]952</table>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]953</div>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]954
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]955<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]956<div class="doc_subsubsection"> <a name="t_classifications">Type
957Classifications</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]958<div class="doc_text">
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]959<p>These different primitive types fall into a few useful
960classifications:</p>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]961
962<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]963 <tbody>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]964 <tr><th>Classification</th><th>Types</th></tr>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]965 <tr>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]966 <td><a name="t_integer">integer</a></td>
Reid Spencer138249b2007-05-16 18:44:01 +0000[diff] [blame]967 <td><tt>i1, i2, i3, ... i8, ... i16, ... i32, ... i64, ... </tt></td>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]968 </tr>
969 <tr>
970 <td><a name="t_floating">floating point</a></td>
971 <td><tt>float, double</tt></td>
972 </tr>
973 <tr>
974 <td><a name="t_firstclass">first class</a></td>
Reid Spencer138249b2007-05-16 18:44:01 +0000[diff] [blame]975 <td><tt>i1, ..., float, double, <br/>
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]976 <a href="#t_pointer">pointer</a>,<a href="#t_vector">vector</a></tt>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]977 </td>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]978 </tr>
979 </tbody>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]980</table>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]981
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]982<p>The <a href="#t_firstclass">first class</a> types are perhaps the
983most important. Values of these types are the only ones which can be
984produced by instructions, passed as arguments, or used as operands to
985instructions. This means that all structures and arrays must be
986manipulated either by pointer or by component.</p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]987</div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]988
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]989<!-- ======================================================================= -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]990<div class="doc_subsection"> <a name="t_derived">Derived Types</a> </div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]991
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]992<div class="doc_text">
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]993
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]994<p>The real power in LLVM comes from the derived types in the system.
995This is what allows a programmer to represent arrays, functions,
996pointers, and other useful types. Note that these derived types may be
997recursive: For example, it is possible to have a two dimensional array.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]998
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]999</div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1000
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1001<!-- _______________________________________________________________________ -->
Reid Spencer138249b2007-05-16 18:44:01 +0000[diff] [blame]1002<div class="doc_subsubsection"> <a name="t_integer">Integer Type</a> </div>
1003
1004<div class="doc_text">
1005
1006<h5>Overview:</h5>
1007<p>The integer type is a very simple derived type that simply specifies an
1008arbitrary bit width for the integer type desired. Any bit width from 1 bit to
10092^23-1 (about 8 million) can be specified.</p>
1010
1011<h5>Syntax:</h5>
1012
1013<pre>
1014 iN
1015</pre>
1016
1017<p>The number of bits the integer will occupy is specified by the <tt>N</tt>
1018value.</p>
1019
1020<h5>Examples:</h5>
1021<table class="layout">
1022 <tr class="layout">
1023 <td class="left">
1024 <tt>i1</tt><br/>
1025 <tt>i4</tt><br/>
1026 <tt>i8</tt><br/>
1027 <tt>i16</tt><br/>
1028 <tt>i32</tt><br/>
1029 <tt>i42</tt><br/>
1030 <tt>i64</tt><br/>
1031 <tt>i1942652</tt><br/>
1032 </td>
1033 <td class="left">
1034 A boolean integer of 1 bit<br/>
1035 A nibble sized integer of 4 bits.<br/>
1036 A byte sized integer of 8 bits.<br/>
1037 A half word sized integer of 16 bits.<br/>
1038 A word sized integer of 32 bits.<br/>
1039 An integer whose bit width is the answer. <br/>
1040 A double word sized integer of 64 bits.<br/>
1041 A really big integer of over 1 million bits.<br/>
1042 </td>
1043 </tr>
1044</table>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]1045</div>
Reid Spencer138249b2007-05-16 18:44:01 +0000[diff] [blame]1046
1047<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1048<div class="doc_subsubsection"> <a name="t_array">Array Type</a> </div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1049
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1050<div class="doc_text">
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1051
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1052<h5>Overview:</h5>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1053
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1054<p>The array type is a very simple derived type that arranges elements
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1055sequentially in memory. The array type requires a size (number of
1056elements) and an underlying data type.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1057
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]1058<h5>Syntax:</h5>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1059
1060<pre>
1061 [&lt;# elements&gt; x &lt;elementtype&gt;]
1062</pre>
1063
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]1064<p>The number of elements is a constant integer value; elementtype may
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1065be any type with a size.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1066
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]1067<h5>Examples:</h5>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1068<table class="layout">
1069 <tr class="layout">
1070 <td class="left">
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1071 <tt>[40 x i32 ]</tt><br/>
1072 <tt>[41 x i32 ]</tt><br/>
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]1073 <tt>[40 x i8]</tt><br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1074 </td>
1075 <td class="left">
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]1076 Array of 40 32-bit integer values.<br/>
1077 Array of 41 32-bit integer values.<br/>
1078 Array of 40 8-bit integer values.<br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1079 </td>
1080 </tr>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1081</table>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1082<p>Here are some examples of multidimensional arrays:</p>
1083<table class="layout">
1084 <tr class="layout">
1085 <td class="left">
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1086 <tt>[3 x [4 x i32]]</tt><br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1087 <tt>[12 x [10 x float]]</tt><br/>
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]1088 <tt>[2 x [3 x [4 x i16]]]</tt><br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1089 </td>
1090 <td class="left">
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]1091 3x4 array of 32-bit integer values.<br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1092 12x10 array of single precision floating point values.<br/>
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]1093 2x3x4 array of 16-bit integer values.<br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1094 </td>
1095 </tr>
1096</table>
Chris Lattnerc0ad71e2005-06-24 17:22:57 +0000[diff] [blame]1097
John Criswell4c0cf7f2005-10-24 16:17:18 +0000[diff] [blame]1098<p>Note that 'variable sized arrays' can be implemented in LLVM with a zero
1099length array. Normally, accesses past the end of an array are undefined in
Chris Lattnerc0ad71e2005-06-24 17:22:57 +0000[diff] [blame]1100LLVM (e.g. it is illegal to access the 5th element of a 3 element array).
1101As a special case, however, zero length arrays are recognized to be variable
1102length. This allows implementation of 'pascal style arrays' with the LLVM
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1103type "{ i32, [0 x float]}", for example.</p>
Chris Lattnerc0ad71e2005-06-24 17:22:57 +0000[diff] [blame]1104
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1105</div>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1106
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1107<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1108<div class="doc_subsubsection"> <a name="t_function">Function Type</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1109<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1110<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1111<p>The function type can be thought of as a function signature. It
1112consists of a return type and a list of formal parameter types.
John Criswella0d50d22003-11-25 21:45:46 +0000[diff] [blame]1113Function types are usually used to build virtual function tables
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1114(which are structures of pointers to functions), for indirect function
1115calls, and when defining a function.</p>
John Criswella0d50d22003-11-25 21:45:46 +0000[diff] [blame]1116<p>
1117The return type of a function type cannot be an aggregate type.
1118</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1119<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1120<pre> &lt;returntype&gt; (&lt;parameter list&gt;)<br></pre>
John Criswell4c0cf7f2005-10-24 16:17:18 +0000[diff] [blame]1121<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]1122specifiers. Optionally, the parameter list may include a type <tt>...</tt>,
Chris Lattner5ed60612003-09-03 00:41:47 +0000[diff] [blame]1123which indicates that the function takes a variable number of arguments.
1124Variable argument functions can access their arguments with the <a
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1125 href="#int_varargs">variable argument handling intrinsic</a> functions.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1126<h5>Examples:</h5>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1127<table class="layout">
1128 <tr class="layout">
Reid Spencer58c08712006-12-31 07:18:34 +0000[diff] [blame]1129 <td class="left"><tt>i32 (i32)</tt></td>
1130 <td class="left">function taking an <tt>i32</tt>, returning an <tt>i32</tt>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1131 </td>
Reid Spencer58c08712006-12-31 07:18:34 +0000[diff] [blame]1132 </tr><tr class="layout">
Reid Spencere6a338d2007-01-15 18:28:34 +0000[diff] [blame]1133 <td class="left"><tt>float&nbsp;(i16&nbsp;sext,&nbsp;i32&nbsp;*)&nbsp;*
Reid Spencer655dcc62006-12-31 07:20:23 +0000[diff] [blame]1134 </tt></td>
Reid Spencer58c08712006-12-31 07:18:34 +0000[diff] [blame]1135 <td class="left"><a href="#t_pointer">Pointer</a> to a function that takes
1136 an <tt>i16</tt> that should be sign extended and a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1137 <a href="#t_pointer">pointer</a> to <tt>i32</tt>, returning
Reid Spencer58c08712006-12-31 07:18:34 +0000[diff] [blame]1138 <tt>float</tt>.
1139 </td>
1140 </tr><tr class="layout">
1141 <td class="left"><tt>i32 (i8*, ...)</tt></td>
1142 <td class="left">A vararg function that takes at least one
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]1143 <a href="#t_pointer">pointer</a> to <tt>i8 </tt> (char in C),
Reid Spencer58c08712006-12-31 07:18:34 +0000[diff] [blame]1144 which returns an integer. This is the signature for <tt>printf</tt> in
1145 LLVM.
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1146 </td>
1147 </tr>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1148</table>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]1149
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1150</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1151<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1152<div class="doc_subsubsection"> <a name="t_struct">Structure Type</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1153<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1154<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1155<p>The structure type is used to represent a collection of data members
1156together in memory. The packing of the field types is defined to match
1157the ABI of the underlying processor. The elements of a structure may
1158be any type that has a size.</p>
1159<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
1160and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
1161field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
1162instruction.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1163<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1164<pre> { &lt;type list&gt; }<br></pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1165<h5>Examples:</h5>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1166<table class="layout">
1167 <tr class="layout">
Jeff Cohen5819f182007-04-22 01:17:39 +0000[diff] [blame]1168 <td class="left"><tt>{ i32, i32, i32 }</tt></td>
1169 <td class="left">A triple of three <tt>i32</tt> values</td>
1170 </tr><tr class="layout">
1171 <td class="left"><tt>{&nbsp;float,&nbsp;i32&nbsp;(i32)&nbsp;*&nbsp;}</tt></td>
1172 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1173 second element is a <a href="#t_pointer">pointer</a> to a
1174 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1175 an <tt>i32</tt>.</td>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1176 </tr>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1177</table>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1178</div>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1179
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1180<!-- _______________________________________________________________________ -->
Andrew Lenharth8df88e22006-12-08 17:13:00 +0000[diff] [blame]1181<div class="doc_subsubsection"> <a name="t_pstruct">Packed Structure Type</a>
1182</div>
1183<div class="doc_text">
1184<h5>Overview:</h5>
1185<p>The packed structure type is used to represent a collection of data members
1186together in memory. There is no padding between fields. Further, the alignment
1187of a packed structure is 1 byte. The elements of a packed structure may
1188be any type that has a size.</p>
1189<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
1190and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
1191field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
1192instruction.</p>
1193<h5>Syntax:</h5>
1194<pre> &lt; { &lt;type list&gt; } &gt; <br></pre>
1195<h5>Examples:</h5>
1196<table class="layout">
1197 <tr class="layout">
Jeff Cohen5819f182007-04-22 01:17:39 +0000[diff] [blame]1198 <td class="left"><tt>&lt; { i32, i32, i32 } &gt;</tt></td>
1199 <td class="left">A triple of three <tt>i32</tt> values</td>
1200 </tr><tr class="layout">
1201 <td class="left"><tt>&lt;&nbsp;{&nbsp;float,&nbsp;i32&nbsp;(i32)&nbsp;*&nbsp;}&nbsp;&gt;</tt></td>
1202 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1203 second element is a <a href="#t_pointer">pointer</a> to a
1204 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1205 an <tt>i32</tt>.</td>
Andrew Lenharth8df88e22006-12-08 17:13:00 +0000[diff] [blame]1206 </tr>
1207</table>
1208</div>
1209
1210<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1211<div class="doc_subsubsection"> <a name="t_pointer">Pointer Type</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1212<div class="doc_text">
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]1213<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1214<p>As in many languages, the pointer type represents a pointer or
1215reference to another object, which must live in memory.</p>
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]1216<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1217<pre> &lt;type&gt; *<br></pre>
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]1218<h5>Examples:</h5>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1219<table class="layout">
1220 <tr class="layout">
1221 <td class="left">
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1222 <tt>[4x i32]*</tt><br/>
1223 <tt>i32 (i32 *) *</tt><br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1224 </td>
1225 <td class="left">
1226 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]1227 four <tt>i32</tt> values<br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1228 A <a href="#t_pointer">pointer</a> to a <a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1229 href="#t_function">function</a> that takes an <tt>i32*</tt>, returning an
1230 <tt>i32</tt>.<br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1231 </td>
1232 </tr>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1233</table>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1234</div>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1235
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1236<!-- _______________________________________________________________________ -->
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1237<div class="doc_subsubsection"> <a name="t_vector">Vector Type</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1238<div class="doc_text">
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]1239
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1240<h5>Overview:</h5>
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]1241
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1242<p>A vector type is a simple derived type that represents a vector
1243of elements. Vector types are used when multiple primitive data
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1244are operated in parallel using a single instruction (SIMD).
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1245A vector type requires a size (number of
Chris Lattner330ce692005-11-10 01:44:22 +0000[diff] [blame]1246elements) and an underlying primitive data type. Vectors must have a power
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1247of two length (1, 2, 4, 8, 16 ...). Vector types are
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1248considered <a href="#t_firstclass">first class</a>.</p>
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]1249
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1250<h5>Syntax:</h5>
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]1251
1252<pre>
1253 &lt; &lt;# elements&gt; x &lt;elementtype&gt; &gt;
1254</pre>
1255
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]1256<p>The number of elements is a constant integer value; elementtype may
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]1257be any integer or floating point type.</p>
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]1258
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1259<h5>Examples:</h5>
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]1260
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1261<table class="layout">
1262 <tr class="layout">
1263 <td class="left">
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1264 <tt>&lt;4 x i32&gt;</tt><br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1265 <tt>&lt;8 x float&gt;</tt><br/>
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]1266 <tt>&lt;2 x i64&gt;</tt><br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1267 </td>
1268 <td class="left">
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1269 Vector of 4 32-bit integer values.<br/>
1270 Vector of 8 floating-point values.<br/>
1271 Vector of 2 64-bit integer values.<br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]1272 </td>
1273 </tr>
1274</table>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1275</div>
1276
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]1277<!-- _______________________________________________________________________ -->
1278<div class="doc_subsubsection"> <a name="t_opaque">Opaque Type</a> </div>
1279<div class="doc_text">
1280
1281<h5>Overview:</h5>
1282
1283<p>Opaque types are used to represent unknown types in the system. This
1284corresponds (for example) to the C notion of a foward declared structure type.
1285In LLVM, opaque types can eventually be resolved to any type (not just a
1286structure type).</p>
1287
1288<h5>Syntax:</h5>
1289
1290<pre>
1291 opaque
1292</pre>
1293
1294<h5>Examples:</h5>
1295
1296<table class="layout">
1297 <tr class="layout">
1298 <td class="left">
1299 <tt>opaque</tt>
1300 </td>
1301 <td class="left">
1302 An opaque type.<br/>
1303 </td>
1304 </tr>
1305</table>
1306</div>
1307
1308
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1309<!-- *********************************************************************** -->
1310<div class="doc_section"> <a name="constants">Constants</a> </div>
1311<!-- *********************************************************************** -->
1312
1313<div class="doc_text">
1314
1315<p>LLVM has several different basic types of constants. This section describes
1316them all and their syntax.</p>
1317
1318</div>
1319
1320<!-- ======================================================================= -->
Reid Spencer8f08d802004-12-09 18:02:53 +0000[diff] [blame]1321<div class="doc_subsection"><a name="simpleconstants">Simple Constants</a></div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1322
1323<div class="doc_text">
1324
1325<dl>
1326 <dt><b>Boolean constants</b></dt>
1327
1328 <dd>The two strings '<tt>true</tt>' and '<tt>false</tt>' are both valid
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]1329 constants of the <tt><a href="#t_primitive">i1</a></tt> type.
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1330 </dd>
1331
1332 <dt><b>Integer constants</b></dt>
1333
Reid Spencer8f08d802004-12-09 18:02:53 +0000[diff] [blame]1334 <dd>Standard integers (such as '4') are constants of the <a
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]1335 href="#t_integer">integer</a> type. Negative numbers may be used with
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1336 integer types.
1337 </dd>
1338
1339 <dt><b>Floating point constants</b></dt>
1340
1341 <dd>Floating point constants use standard decimal notation (e.g. 123.421),
1342 exponential notation (e.g. 1.23421e+2), or a more precise hexadecimal
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1343 notation (see below). Floating point constants must have a <a
1344 href="#t_floating">floating point</a> type. </dd>
1345
1346 <dt><b>Null pointer constants</b></dt>
1347
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]1348 <dd>The identifier '<tt>null</tt>' is recognized as a null pointer constant
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1349 and must be of <a href="#t_pointer">pointer type</a>.</dd>
1350
1351</dl>
1352
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]1353<p>The one non-intuitive notation for constants is the optional hexadecimal form
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1354of floating point constants. For example, the form '<tt>double
13550x432ff973cafa8000</tt>' is equivalent to (but harder to read than) '<tt>double
13564.5e+15</tt>'. The only time hexadecimal floating point constants are required
Reid Spencer8f08d802004-12-09 18:02:53 +0000[diff] [blame]1357(and the only time that they are generated by the disassembler) is when a
1358floating point constant must be emitted but it cannot be represented as a
1359decimal floating point number. For example, NaN's, infinities, and other
1360special values are represented in their IEEE hexadecimal format so that
1361assembly and disassembly do not cause any bits to change in the constants.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1362
1363</div>
1364
1365<!-- ======================================================================= -->
1366<div class="doc_subsection"><a name="aggregateconstants">Aggregate Constants</a>
1367</div>
1368
1369<div class="doc_text">
Chris Lattner455fc8c2005-03-07 22:13:59 +0000[diff] [blame]1370<p>Aggregate constants arise from aggregation of simple constants
1371and smaller aggregate constants.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1372
1373<dl>
1374 <dt><b>Structure constants</b></dt>
1375
1376 <dd>Structure constants are represented with notation similar to structure
1377 type definitions (a comma separated list of elements, surrounded by braces
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1378 (<tt>{}</tt>)). For example: "<tt>{ i32 4, float 17.0, i32* %G }</tt>",
1379 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]1380 must have <a href="#t_struct">structure type</a>, and the number and
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1381 types of elements must match those specified by the type.
1382 </dd>
1383
1384 <dt><b>Array constants</b></dt>
1385
1386 <dd>Array constants are represented with notation similar to array type
1387 definitions (a comma separated list of elements, surrounded by square brackets
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1388 (<tt>[]</tt>)). For example: "<tt>[ i32 42, i32 11, i32 74 ]</tt>". Array
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1389 constants must have <a href="#t_array">array type</a>, and the number and
1390 types of elements must match those specified by the type.
1391 </dd>
1392
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1393 <dt><b>Vector constants</b></dt>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1394
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1395 <dd>Vector constants are represented with notation similar to vector type
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1396 definitions (a comma separated list of elements, surrounded by
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1397 less-than/greater-than's (<tt>&lt;&gt;</tt>)). For example: "<tt>&lt; i32 42,
Jeff Cohen5819f182007-04-22 01:17:39 +0000[diff] [blame]1398 i32 11, i32 74, i32 100 &gt;</tt>". Vector constants must have <a
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1399 href="#t_vector">vector type</a>, and the number and types of elements must
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1400 match those specified by the type.
1401 </dd>
1402
1403 <dt><b>Zero initialization</b></dt>
1404
1405 <dd>The string '<tt>zeroinitializer</tt>' can be used to zero initialize a
1406 value to zero of <em>any</em> type, including scalar and aggregate types.
1407 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]1408 large arrays) and is always exactly equivalent to using explicit zero
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1409 initializers.
1410 </dd>
1411</dl>
1412
1413</div>
1414
1415<!-- ======================================================================= -->
1416<div class="doc_subsection">
1417 <a name="globalconstants">Global Variable and Function Addresses</a>
1418</div>
1419
1420<div class="doc_text">
1421
1422<p>The addresses of <a href="#globalvars">global variables</a> and <a
1423href="#functionstructure">functions</a> are always implicitly valid (link-time)
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]1424constants. These constants are explicitly referenced when the <a
1425href="#identifiers">identifier for the global</a> is used and always have <a
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1426href="#t_pointer">pointer</a> type. For example, the following is a legal LLVM
1427file:</p>
1428
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]1429<div class="doc_code">
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1430<pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]1431%X = global i32 17
1432%Y = global i32 42
1433%Z = global [2 x i32*] [ i32* %X, i32* %Y ]
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1434</pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]1435</div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1436
1437</div>
1438
1439<!-- ======================================================================= -->
Reid Spencer641f5c92004-12-09 18:13:12 +0000[diff] [blame]1440<div class="doc_subsection"><a name="undefvalues">Undefined Values</a></div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1441<div class="doc_text">
Reid Spencer641f5c92004-12-09 18:13:12 +0000[diff] [blame]1442 <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]1443 no specific value. Undefined values may be of any type and be used anywhere
Reid Spencer641f5c92004-12-09 18:13:12 +0000[diff] [blame]1444 a constant is permitted.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1445
Reid Spencer641f5c92004-12-09 18:13:12 +0000[diff] [blame]1446 <p>Undefined values indicate to the compiler that the program is well defined
1447 no matter what value is used, giving the compiler more freedom to optimize.
1448 </p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1449</div>
1450
1451<!-- ======================================================================= -->
1452<div class="doc_subsection"><a name="constantexprs">Constant Expressions</a>
1453</div>
1454
1455<div class="doc_text">
1456
1457<p>Constant expressions are used to allow expressions involving other constants
1458to be used as constants. Constant expressions may be of any <a
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]1459href="#t_firstclass">first class</a> type and may involve any LLVM operation
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1460that does not have side effects (e.g. load and call are not supported). The
1461following is the syntax for constant expressions:</p>
1462
1463<dl>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]1464 <dt><b><tt>trunc ( CST to TYPE )</tt></b></dt>
1465 <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]1466 than the bit size of TYPE. Both types must be integers.</dd>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1467
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]1468 <dt><b><tt>zext ( CST to TYPE )</tt></b></dt>
1469 <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]1470 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]1471
1472 <dt><b><tt>sext ( CST to TYPE )</tt></b></dt>
1473 <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]1474 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]1475
1476 <dt><b><tt>fptrunc ( CST to TYPE )</tt></b></dt>
1477 <dd>Truncate a floating point constant to another floating point type. The
1478 size of CST must be larger than the size of TYPE. Both types must be
1479 floating point.</dd>
1480
1481 <dt><b><tt>fpext ( CST to TYPE )</tt></b></dt>
1482 <dd>Floating point extend a constant to another type. The size of CST must be
1483 smaller or equal to the size of TYPE. Both types must be floating point.</dd>
1484
1485 <dt><b><tt>fp2uint ( CST to TYPE )</tt></b></dt>
1486 <dd>Convert a floating point constant to the corresponding unsigned integer
1487 constant. TYPE must be an integer type. CST must be floating point. If the
1488 value won't fit in the integer type, the results are undefined.</dd>
1489
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]1490 <dt><b><tt>fptosi ( CST to TYPE )</tt></b></dt>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]1491 <dd>Convert a floating point constant to the corresponding signed integer
1492 constant. TYPE must be an integer type. CST must be floating point. If the
1493 value won't fit in the integer type, the results are undefined.</dd>
1494
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]1495 <dt><b><tt>uitofp ( CST to TYPE )</tt></b></dt>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]1496 <dd>Convert an unsigned integer constant to the corresponding floating point
1497 constant. TYPE must be floating point. CST must be of integer type. If the
Jeff Cohenbeccb742007-04-22 14:56:37 +0000[diff] [blame]1498 value won't fit in the floating point type, the results are undefined.</dd>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]1499
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]1500 <dt><b><tt>sitofp ( CST to TYPE )</tt></b></dt>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]1501 <dd>Convert a signed integer constant to the corresponding floating point
1502 constant. TYPE must be floating point. CST must be of integer type. If the
Jeff Cohenbeccb742007-04-22 14:56:37 +0000[diff] [blame]1503 value won't fit in the floating point type, the results are undefined.</dd>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]1504
Reid Spencer5b950642006-11-11 23:08:07 +0000[diff] [blame]1505 <dt><b><tt>ptrtoint ( CST to TYPE )</tt></b></dt>
1506 <dd>Convert a pointer typed constant to the corresponding integer constant
1507 TYPE must be an integer type. CST must be of pointer type. The CST value is
1508 zero extended, truncated, or unchanged to make it fit in TYPE.</dd>
1509
1510 <dt><b><tt>inttoptr ( CST to TYPE )</tt></b></dt>
1511 <dd>Convert a integer constant to a pointer constant. TYPE must be a
1512 pointer type. CST must be of integer type. The CST value is zero extended,
1513 truncated, or unchanged to make it fit in a pointer size. This one is
1514 <i>really</i> dangerous!</dd>
1515
1516 <dt><b><tt>bitcast ( CST to TYPE )</tt></b></dt>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]1517 <dd>Convert a constant, CST, to another TYPE. The size of CST and TYPE must be
1518 identical (same number of bits). The conversion is done as if the CST value
1519 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]1520 with this operator, just the type. This can be used for conversion of
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1521 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]1522 pointers it is only valid to cast to another pointer type.
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]1523 </dd>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1524
1525 <dt><b><tt>getelementptr ( CSTPTR, IDX0, IDX1, ... )</tt></b></dt>
1526
1527 <dd>Perform the <a href="#i_getelementptr">getelementptr operation</a> on
1528 constants. As with the <a href="#i_getelementptr">getelementptr</a>
1529 instruction, the index list may have zero or more indexes, which are required
1530 to make sense for the type of "CSTPTR".</dd>
1531
Robert Bocchino7e97a6d2006-01-10 19:31:34 +0000[diff] [blame]1532 <dt><b><tt>select ( COND, VAL1, VAL2 )</tt></b></dt>
1533
1534 <dd>Perform the <a href="#i_select">select operation</a> on
Reid Spencer9965ee72006-12-04 19:23:19 +0000[diff] [blame]1535 constants.</dd>
1536
1537 <dt><b><tt>icmp COND ( VAL1, VAL2 )</tt></b></dt>
1538 <dd>Performs the <a href="#i_icmp">icmp operation</a> on constants.</dd>
1539
1540 <dt><b><tt>fcmp COND ( VAL1, VAL2 )</tt></b></dt>
1541 <dd>Performs the <a href="#i_fcmp">fcmp operation</a> on constants.</dd>
Robert Bocchino7e97a6d2006-01-10 19:31:34 +0000[diff] [blame]1542
1543 <dt><b><tt>extractelement ( VAL, IDX )</tt></b></dt>
1544
1545 <dd>Perform the <a href="#i_extractelement">extractelement
1546 operation</a> on constants.
1547
Robert Bocchinof72fdfe2006-01-15 20:48:27 +0000[diff] [blame]1548 <dt><b><tt>insertelement ( VAL, ELT, IDX )</tt></b></dt>
1549
1550 <dd>Perform the <a href="#i_insertelement">insertelement
Reid Spencer9965ee72006-12-04 19:23:19 +0000[diff] [blame]1551 operation</a> on constants.</dd>
Robert Bocchinof72fdfe2006-01-15 20:48:27 +0000[diff] [blame]1552
Chris Lattner016a0e52006-04-08 00:13:41 +0000[diff] [blame]1553
1554 <dt><b><tt>shufflevector ( VEC1, VEC2, IDXMASK )</tt></b></dt>
1555
1556 <dd>Perform the <a href="#i_shufflevector">shufflevector
Reid Spencer9965ee72006-12-04 19:23:19 +0000[diff] [blame]1557 operation</a> on constants.</dd>
Chris Lattner016a0e52006-04-08 00:13:41 +0000[diff] [blame]1558
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1559 <dt><b><tt>OPCODE ( LHS, RHS )</tt></b></dt>
1560
Reid Spencer641f5c92004-12-09 18:13:12 +0000[diff] [blame]1561 <dd>Perform the specified operation of the LHS and RHS constants. OPCODE may
1562 be any of the <a href="#binaryops">binary</a> or <a href="#bitwiseops">bitwise
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1563 binary</a> operations. The constraints on operands are the same as those for
1564 the corresponding instruction (e.g. no bitwise operations on floating point
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]1565 values are allowed).</dd>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1566</dl>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1567</div>
Chris Lattnerb1652612004-03-08 16:49:10 +0000[diff] [blame]1568
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1569<!-- *********************************************************************** -->
Chris Lattner98f013c2006-01-25 23:47:57 +0000[diff] [blame]1570<div class="doc_section"> <a name="othervalues">Other Values</a> </div>
1571<!-- *********************************************************************** -->
1572
1573<!-- ======================================================================= -->
1574<div class="doc_subsection">
1575<a name="inlineasm">Inline Assembler Expressions</a>
1576</div>
1577
1578<div class="doc_text">
1579
1580<p>
1581LLVM supports inline assembler expressions (as opposed to <a href="#moduleasm">
1582Module-Level Inline Assembly</a>) through the use of a special value. This
1583value represents the inline assembler as a string (containing the instructions
1584to emit), a list of operand constraints (stored as a string), and a flag that
1585indicates whether or not the inline asm expression has side effects. An example
1586inline assembler expression is:
1587</p>
1588
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]1589<div class="doc_code">
Chris Lattner98f013c2006-01-25 23:47:57 +0000[diff] [blame]1590<pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]1591i32 (i32) asm "bswap $0", "=r,r"
Chris Lattner98f013c2006-01-25 23:47:57 +0000[diff] [blame]1592</pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]1593</div>
Chris Lattner98f013c2006-01-25 23:47:57 +0000[diff] [blame]1594
1595<p>
1596Inline assembler expressions may <b>only</b> be used as the callee operand of
1597a <a href="#i_call"><tt>call</tt> instruction</a>. Thus, typically we have:
1598</p>
1599
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]1600<div class="doc_code">
Chris Lattner98f013c2006-01-25 23:47:57 +0000[diff] [blame]1601<pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]1602%X = call i32 asm "<a href="#int_bswap">bswap</a> $0", "=r,r"(i32 %Y)
Chris Lattner98f013c2006-01-25 23:47:57 +0000[diff] [blame]1603</pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]1604</div>
Chris Lattner98f013c2006-01-25 23:47:57 +0000[diff] [blame]1605
1606<p>
1607Inline asms with side effects not visible in the constraint list must be marked
1608as having side effects. This is done through the use of the
1609'<tt>sideeffect</tt>' keyword, like so:
1610</p>
1611
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]1612<div class="doc_code">
Chris Lattner98f013c2006-01-25 23:47:57 +0000[diff] [blame]1613<pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]1614call void asm sideeffect "eieio", ""()
Chris Lattner98f013c2006-01-25 23:47:57 +0000[diff] [blame]1615</pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]1616</div>
Chris Lattner98f013c2006-01-25 23:47:57 +0000[diff] [blame]1617
1618<p>TODO: The format of the asm and constraints string still need to be
1619documented here. Constraints on what can be done (e.g. duplication, moving, etc
1620need to be documented).
1621</p>
1622
1623</div>
1624
1625<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1626<div class="doc_section"> <a name="instref">Instruction Reference</a> </div>
1627<!-- *********************************************************************** -->
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1628
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1629<div class="doc_text">
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1630
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1631<p>The LLVM instruction set consists of several different
1632classifications of instructions: <a href="#terminators">terminator
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]1633instructions</a>, <a href="#binaryops">binary instructions</a>,
1634<a href="#bitwiseops">bitwise binary instructions</a>, <a
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1635 href="#memoryops">memory instructions</a>, and <a href="#otherops">other
1636instructions</a>.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1637
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1638</div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1639
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1640<!-- ======================================================================= -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1641<div class="doc_subsection"> <a name="terminators">Terminator
1642Instructions</a> </div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1643
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1644<div class="doc_text">
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1645
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1646<p>As mentioned <a href="#functionstructure">previously</a>, every
1647basic block in a program ends with a "Terminator" instruction, which
1648indicates which block should be executed after the current block is
1649finished. These terminator instructions typically yield a '<tt>void</tt>'
1650value: they produce control flow, not values (the one exception being
1651the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction).</p>
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]1652<p>There are six different terminator instructions: the '<a
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1653 href="#i_ret"><tt>ret</tt></a>' instruction, the '<a href="#i_br"><tt>br</tt></a>'
1654instruction, the '<a href="#i_switch"><tt>switch</tt></a>' instruction,
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1655the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction, the '<a
1656 href="#i_unwind"><tt>unwind</tt></a>' instruction, and the '<a
1657 href="#i_unreachable"><tt>unreachable</tt></a>' instruction.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1658
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1659</div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1660
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1661<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1662<div class="doc_subsubsection"> <a name="i_ret">'<tt>ret</tt>'
1663Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1664<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1665<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1666<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]1667 ret void <i>; Return from void function</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1668</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1669<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1670<p>The '<tt>ret</tt>' instruction is used to return control flow (and a
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]1671value) from a function back to the caller.</p>
John Criswell417228d2004-04-09 16:48:45 +0000[diff] [blame]1672<p>There are two forms of the '<tt>ret</tt>' instruction: one that
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1673returns a value and then causes control flow, and one that just causes
1674control flow to occur.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1675<h5>Arguments:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1676<p>The '<tt>ret</tt>' instruction may return any '<a
1677 href="#t_firstclass">first class</a>' type. Notice that a function is
1678not <a href="#wellformed">well formed</a> if there exists a '<tt>ret</tt>'
1679instruction inside of the function that returns a value that does not
1680match the return type of the function.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1681<h5>Semantics:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1682<p>When the '<tt>ret</tt>' instruction is executed, control flow
1683returns back to the calling function's context. If the caller is a "<a
John Criswell40db33f2004-06-25 15:16:57 +0000[diff] [blame]1684 href="#i_call"><tt>call</tt></a>" instruction, execution continues at
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1685the instruction after the call. If the caller was an "<a
1686 href="#i_invoke"><tt>invoke</tt></a>" instruction, execution continues
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]1687at the beginning of the "normal" destination block. If the instruction
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1688returns a value, that value shall set the call or invoke instruction's
1689return value.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1690<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1691<pre> ret i32 5 <i>; Return an integer value of 5</i>
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]1692 ret void <i>; Return from a void function</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1693</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1694</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1695<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1696<div class="doc_subsubsection"> <a name="i_br">'<tt>br</tt>' Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1697<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1698<h5>Syntax:</h5>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]1699<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]1700</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1701<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1702<p>The '<tt>br</tt>' instruction is used to cause control flow to
1703transfer to a different basic block in the current function. There are
1704two forms of this instruction, corresponding to a conditional branch
1705and an unconditional branch.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1706<h5>Arguments:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1707<p>The conditional branch form of the '<tt>br</tt>' instruction takes a
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]1708single '<tt>i1</tt>' value and two '<tt>label</tt>' values. The
Reid Spencer50c723a2007-02-19 23:54:10 +0000[diff] [blame]1709unconditional form of the '<tt>br</tt>' instruction takes a single
1710'<tt>label</tt>' value as a target.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1711<h5>Semantics:</h5>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]1712<p>Upon execution of a conditional '<tt>br</tt>' instruction, the '<tt>i1</tt>'
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1713argument is evaluated. If the value is <tt>true</tt>, control flows
1714to the '<tt>iftrue</tt>' <tt>label</tt> argument. If "cond" is <tt>false</tt>,
1715control flows to the '<tt>iffalse</tt>' <tt>label</tt> argument.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1716<h5>Example:</h5>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]1717<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]1718 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]1719</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1720<!-- _______________________________________________________________________ -->
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1721<div class="doc_subsubsection">
1722 <a name="i_switch">'<tt>switch</tt>' Instruction</a>
1723</div>
1724
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1725<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1726<h5>Syntax:</h5>
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1727
1728<pre>
1729 switch &lt;intty&gt; &lt;value&gt;, label &lt;defaultdest&gt; [ &lt;intty&gt; &lt;val&gt;, label &lt;dest&gt; ... ]
1730</pre>
1731
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1732<h5>Overview:</h5>
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1733
1734<p>The '<tt>switch</tt>' instruction is used to transfer control flow to one of
1735several different places. It is a generalization of the '<tt>br</tt>'
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1736instruction, allowing a branch to occur to one of many possible
1737destinations.</p>
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1738
1739
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1740<h5>Arguments:</h5>
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1741
1742<p>The '<tt>switch</tt>' instruction uses three parameters: an integer
1743comparison value '<tt>value</tt>', a default '<tt>label</tt>' destination, and
1744an array of pairs of comparison value constants and '<tt>label</tt>'s. The
1745table is not allowed to contain duplicate constant entries.</p>
1746
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1747<h5>Semantics:</h5>
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1748
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1749<p>The <tt>switch</tt> instruction specifies a table of values and
1750destinations. When the '<tt>switch</tt>' instruction is executed, this
John Criswellbcbb18c2004-06-25 16:05:06 +0000[diff] [blame]1751table is searched for the given value. If the value is found, control flow is
1752transfered to the corresponding destination; otherwise, control flow is
1753transfered to the default destination.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1754
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1755<h5>Implementation:</h5>
1756
1757<p>Depending on properties of the target machine and the particular
1758<tt>switch</tt> instruction, this instruction may be code generated in different
John Criswellbcbb18c2004-06-25 16:05:06 +0000[diff] [blame]1759ways. For example, it could be generated as a series of chained conditional
1760branches or with a lookup table.</p>
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1761
1762<h5>Example:</h5>
1763
1764<pre>
1765 <i>; Emulate a conditional br instruction</i>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]1766 %Val = <a href="#i_zext">zext</a> i1 %value to i32
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1767 switch i32 %Val, label %truedest [i32 0, label %falsedest ]
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1768
1769 <i>; Emulate an unconditional br instruction</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1770 switch i32 0, label %dest [ ]
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1771
1772 <i>; Implement a jump table:</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1773 switch i32 %val, label %otherwise [ i32 0, label %onzero
1774 i32 1, label %onone
1775 i32 2, label %ontwo ]
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1776</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1777</div>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1778
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1779<!-- _______________________________________________________________________ -->
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1780<div class="doc_subsubsection">
1781 <a name="i_invoke">'<tt>invoke</tt>' Instruction</a>
1782</div>
1783
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1784<div class="doc_text">
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1785
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1786<h5>Syntax:</h5>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1787
1788<pre>
1789 &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]1790 to label &lt;normal label&gt; unwind label &lt;exception label&gt;
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1791</pre>
1792
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]1793<h5>Overview:</h5>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1794
1795<p>The '<tt>invoke</tt>' instruction causes control to transfer to a specified
1796function, with the possibility of control flow transfer to either the
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]1797'<tt>normal</tt>' label or the
1798'<tt>exception</tt>' label. If the callee function returns with the
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1799"<tt><a href="#i_ret">ret</a></tt>" instruction, control flow will return to the
1800"normal" label. If the callee (or any indirect callees) returns with the "<a
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]1801href="#i_unwind"><tt>unwind</tt></a>" instruction, control is interrupted and
1802continued at the dynamically nearest "exception" label.</p>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1803
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1804<h5>Arguments:</h5>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1805
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1806<p>This instruction requires several arguments:</p>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1807
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1808<ol>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1809 <li>
Duncan Sands16f122e2007-03-30 12:22:09 +0000[diff] [blame]1810 The optional "cconv" marker indicates which <a href="#callingconv">calling
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1811 convention</a> the call should use. If none is specified, the call defaults
1812 to using C calling conventions.
1813 </li>
1814 <li>'<tt>ptr to function ty</tt>': shall be the signature of the pointer to
1815 function value being invoked. In most cases, this is a direct function
1816 invocation, but indirect <tt>invoke</tt>s are just as possible, branching off
1817 an arbitrary pointer to function value.
1818 </li>
1819
1820 <li>'<tt>function ptr val</tt>': An LLVM value containing a pointer to a
1821 function to be invoked. </li>
1822
1823 <li>'<tt>function args</tt>': argument list whose types match the function
1824 signature argument types. If the function signature indicates the function
1825 accepts a variable number of arguments, the extra arguments can be
1826 specified. </li>
1827
1828 <li>'<tt>normal label</tt>': the label reached when the called function
1829 executes a '<tt><a href="#i_ret">ret</a></tt>' instruction. </li>
1830
1831 <li>'<tt>exception label</tt>': the label reached when a callee returns with
1832 the <a href="#i_unwind"><tt>unwind</tt></a> instruction. </li>
1833
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1834</ol>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1835
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1836<h5>Semantics:</h5>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1837
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1838<p>This instruction is designed to operate as a standard '<tt><a
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1839href="#i_call">call</a></tt>' instruction in most regards. The primary
1840difference is that it establishes an association with a label, which is used by
1841the runtime library to unwind the stack.</p>
1842
1843<p>This instruction is used in languages with destructors to ensure that proper
1844cleanup is performed in the case of either a <tt>longjmp</tt> or a thrown
1845exception. Additionally, this is important for implementation of
1846'<tt>catch</tt>' clauses in high-level languages that support them.</p>
1847
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1848<h5>Example:</h5>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1849<pre>
Jeff Cohen5819f182007-04-22 01:17:39 +0000[diff] [blame]1850 %retval = invoke i32 %Test(i32 15) to label %Continue
1851 unwind label %TestCleanup <i>; {i32}:retval set</i>
1852 %retval = invoke <a href="#callingconv">coldcc</a> i32 %Test(i32 15) to label %Continue
1853 unwind label %TestCleanup <i>; {i32}:retval set</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1854</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1855</div>
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1856
1857
Chris Lattner5ed60612003-09-03 00:41:47 +0000[diff] [blame]1858<!-- _______________________________________________________________________ -->
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1859
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1860<div class="doc_subsubsection"> <a name="i_unwind">'<tt>unwind</tt>'
1861Instruction</a> </div>
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1862
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1863<div class="doc_text">
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1864
Chris Lattner5ed60612003-09-03 00:41:47 +0000[diff] [blame]1865<h5>Syntax:</h5>
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1866<pre>
1867 unwind
1868</pre>
1869
Chris Lattner5ed60612003-09-03 00:41:47 +0000[diff] [blame]1870<h5>Overview:</h5>
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1871
1872<p>The '<tt>unwind</tt>' instruction unwinds the stack, continuing control flow
1873at the first callee in the dynamic call stack which used an <a
1874href="#i_invoke"><tt>invoke</tt></a> instruction to perform the call. This is
1875primarily used to implement exception handling.</p>
1876
Chris Lattner5ed60612003-09-03 00:41:47 +0000[diff] [blame]1877<h5>Semantics:</h5>
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1878
1879<p>The '<tt>unwind</tt>' intrinsic causes execution of the current function to
1880immediately halt. The dynamic call stack is then searched for the first <a
1881href="#i_invoke"><tt>invoke</tt></a> instruction on the call stack. Once found,
1882execution continues at the "exceptional" destination block specified by the
1883<tt>invoke</tt> instruction. If there is no <tt>invoke</tt> instruction in the
1884dynamic call chain, undefined behavior results.</p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1885</div>
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1886
1887<!-- _______________________________________________________________________ -->
1888
1889<div class="doc_subsubsection"> <a name="i_unreachable">'<tt>unreachable</tt>'
1890Instruction</a> </div>
1891
1892<div class="doc_text">
1893
1894<h5>Syntax:</h5>
1895<pre>
1896 unreachable
1897</pre>
1898
1899<h5>Overview:</h5>
1900
1901<p>The '<tt>unreachable</tt>' instruction has no defined semantics. This
1902instruction is used to inform the optimizer that a particular portion of the
1903code is not reachable. This can be used to indicate that the code after a
1904no-return function cannot be reached, and other facts.</p>
1905
1906<h5>Semantics:</h5>
1907
1908<p>The '<tt>unreachable</tt>' instruction has no defined semantics.</p>
1909</div>
1910
1911
1912
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1913<!-- ======================================================================= -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1914<div class="doc_subsection"> <a name="binaryops">Binary Operations</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1915<div class="doc_text">
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1916<p>Binary operators are used to do most of the computation in a
1917program. They require two operands, execute an operation on them, and
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]1918produce a single value. The operands might represent
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1919multiple 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]1920The result value of a binary operator is not
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1921necessarily the same type as its operands.</p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1922<p>There are several different binary operators:</p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1923</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1924<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1925<div class="doc_subsubsection"> <a name="i_add">'<tt>add</tt>'
1926Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1927<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1928<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1929<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]1930</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1931<h5>Overview:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1932<p>The '<tt>add</tt>' instruction returns the sum of its two operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1933<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1934<p>The two arguments to the '<tt>add</tt>' instruction must be either <a
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1935 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a> values.
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1936 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]1937Both arguments must have identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1938<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1939<p>The value produced is the integer or floating point sum of the two
1940operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1941<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1942<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]1943</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1944</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1945<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1946<div class="doc_subsubsection"> <a name="i_sub">'<tt>sub</tt>'
1947Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1948<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1949<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1950<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]1951</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1952<h5>Overview:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1953<p>The '<tt>sub</tt>' instruction returns the difference of its two
1954operands.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1955<p>Note that the '<tt>sub</tt>' instruction is used to represent the '<tt>neg</tt>'
1956instruction present in most other intermediate representations.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1957<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1958<p>The two arguments to the '<tt>sub</tt>' instruction must be either <a
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1959 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1960values.
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1961This instruction can also take <a href="#t_vector">vector</a> versions of the values.
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1962Both arguments must have identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1963<h5>Semantics:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1964<p>The value produced is the integer or floating point difference of
1965the two operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1966<h5>Example:</h5>
Bill Wendling2d8b9a82007-05-29 09:42:13 +0000[diff] [blame]1967<pre>
1968 &lt;result&gt; = sub i32 4, %var <i>; yields {i32}:result = 4 - %var</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1969 &lt;result&gt; = sub i32 0, %val <i>; yields {i32}:result = -%var</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1970</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1971</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1972<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1973<div class="doc_subsubsection"> <a name="i_mul">'<tt>mul</tt>'
1974Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1975<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1976<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1977<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]1978</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1979<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1980<p>The '<tt>mul</tt>' instruction returns the product of its two
1981operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1982<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1983<p>The two arguments to the '<tt>mul</tt>' instruction must be either <a
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1984 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1985values.
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]1986This instruction can also take <a href="#t_vector">vector</a> versions of the values.
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1987Both arguments must have identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1988<h5>Semantics:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1989<p>The value produced is the integer or floating point product of the
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1990two operands.</p>
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]1991<p>Because the operands are the same width, the result of an integer
1992multiplication is the same whether the operands should be deemed unsigned or
1993signed.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1994<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]1995<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]1996</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1997</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1998<!-- _______________________________________________________________________ -->
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame]1999<div class="doc_subsubsection"> <a name="i_udiv">'<tt>udiv</tt>' Instruction
2000</a></div>
2001<div class="doc_text">
2002<h5>Syntax:</h5>
2003<pre> &lt;result&gt; = udiv &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2004</pre>
2005<h5>Overview:</h5>
2006<p>The '<tt>udiv</tt>' instruction returns the quotient of its two
2007operands.</p>
2008<h5>Arguments:</h5>
2009<p>The two arguments to the '<tt>udiv</tt>' instruction must be
2010<a href="#t_integer">integer</a> values. Both arguments must have identical
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]2011types. This instruction can also take <a href="#t_vector">vector</a> versions
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame]2012of the values in which case the elements must be integers.</p>
2013<h5>Semantics:</h5>
2014<p>The value produced is the unsigned integer quotient of the two operands. This
2015instruction always performs an unsigned division operation, regardless of
2016whether the arguments are unsigned or not.</p>
2017<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2018<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]2019</pre>
2020</div>
2021<!-- _______________________________________________________________________ -->
2022<div class="doc_subsubsection"> <a name="i_sdiv">'<tt>sdiv</tt>' Instruction
2023</a> </div>
2024<div class="doc_text">
2025<h5>Syntax:</h5>
2026<pre> &lt;result&gt; = sdiv &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2027</pre>
2028<h5>Overview:</h5>
2029<p>The '<tt>sdiv</tt>' instruction returns the quotient of its two
2030operands.</p>
2031<h5>Arguments:</h5>
2032<p>The two arguments to the '<tt>sdiv</tt>' instruction must be
2033<a href="#t_integer">integer</a> values. Both arguments must have identical
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]2034types. This instruction can also take <a href="#t_vector">vector</a> versions
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame]2035of the values in which case the elements must be integers.</p>
2036<h5>Semantics:</h5>
2037<p>The value produced is the signed integer quotient of the two operands. This
2038instruction always performs a signed division operation, regardless of whether
2039the arguments are signed or not.</p>
2040<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2041<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]2042</pre>
2043</div>
2044<!-- _______________________________________________________________________ -->
2045<div class="doc_subsubsection"> <a name="i_fdiv">'<tt>fdiv</tt>'
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2046Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2047<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2048<h5>Syntax:</h5>
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame]2049<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]2050</pre>
2051<h5>Overview:</h5>
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame]2052<p>The '<tt>fdiv</tt>' instruction returns the quotient of its two
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2053operands.</p>
2054<h5>Arguments:</h5>
Jeff Cohen5819f182007-04-22 01:17:39 +0000[diff] [blame]2055<p>The two arguments to the '<tt>fdiv</tt>' instruction must be
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame]2056<a href="#t_floating">floating point</a> values. Both arguments must have
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]2057identical types. This instruction can also take <a href="#t_vector">vector</a>
Jeff Cohen5819f182007-04-22 01:17:39 +0000[diff] [blame]2058versions of floating point values.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2059<h5>Semantics:</h5>
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame]2060<p>The value produced is the floating point quotient of the two operands.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2061<h5>Example:</h5>
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame]2062<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]2063</pre>
2064</div>
2065<!-- _______________________________________________________________________ -->
Reid Spencer7eb55b32006-11-02 01:53:59 +0000[diff] [blame]2066<div class="doc_subsubsection"> <a name="i_urem">'<tt>urem</tt>' Instruction</a>
2067</div>
2068<div class="doc_text">
2069<h5>Syntax:</h5>
2070<pre> &lt;result&gt; = urem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2071</pre>
2072<h5>Overview:</h5>
2073<p>The '<tt>urem</tt>' instruction returns the remainder from the
2074unsigned division of its two arguments.</p>
2075<h5>Arguments:</h5>
2076<p>The two arguments to the '<tt>urem</tt>' instruction must be
2077<a href="#t_integer">integer</a> values. Both arguments must have identical
2078types.</p>
2079<h5>Semantics:</h5>
2080<p>This instruction returns the unsigned integer <i>remainder</i> of a division.
2081This instruction always performs an unsigned division to get the remainder,
2082regardless of whether the arguments are unsigned or not.</p>
2083<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2084<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]2085</pre>
2086
2087</div>
2088<!-- _______________________________________________________________________ -->
2089<div class="doc_subsubsection"> <a name="i_srem">'<tt>srem</tt>'
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2090Instruction</a> </div>
2091<div class="doc_text">
2092<h5>Syntax:</h5>
Reid Spencer7eb55b32006-11-02 01:53:59 +0000[diff] [blame]2093<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]2094</pre>
2095<h5>Overview:</h5>
Reid Spencer7eb55b32006-11-02 01:53:59 +0000[diff] [blame]2096<p>The '<tt>srem</tt>' instruction returns the remainder from the
2097signed division of its two operands.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2098<h5>Arguments:</h5>
Reid Spencer7eb55b32006-11-02 01:53:59 +0000[diff] [blame]2099<p>The two arguments to the '<tt>srem</tt>' instruction must be
2100<a href="#t_integer">integer</a> values. Both arguments must have identical
2101types.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2102<h5>Semantics:</h5>
Reid Spencer7eb55b32006-11-02 01:53:59 +0000[diff] [blame]2103<p>This instruction returns the <i>remainder</i> of a division (where the result
Reid Spencer806ad6a2007-03-24 22:23:39 +0000[diff] [blame]2104has the same sign as the dividend, <tt>var1</tt>), not the <i>modulo</i>
2105operator (where the result has the same sign as the divisor, <tt>var2</tt>) of
2106a value. For more information about the difference, see <a
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2107 href="http://mathforum.org/dr.math/problems/anne.4.28.99.html">The
Reid Spencer806ad6a2007-03-24 22:23:39 +0000[diff] [blame]2108Math Forum</a>. For a table of how this is implemented in various languages,
Reid Spencerdb3b93b2007-03-24 22:40:44 +0000[diff] [blame]2109please see <a href="http://en.wikipedia.org/wiki/Modulo_operation">
Reid Spencer806ad6a2007-03-24 22:23:39 +0000[diff] [blame]2110Wikipedia: modulo operation</a>.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2111<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2112<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]2113</pre>
2114
2115</div>
2116<!-- _______________________________________________________________________ -->
2117<div class="doc_subsubsection"> <a name="i_frem">'<tt>frem</tt>'
2118Instruction</a> </div>
2119<div class="doc_text">
2120<h5>Syntax:</h5>
2121<pre> &lt;result&gt; = frem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2122</pre>
2123<h5>Overview:</h5>
2124<p>The '<tt>frem</tt>' instruction returns the remainder from the
2125division of its two operands.</p>
2126<h5>Arguments:</h5>
2127<p>The two arguments to the '<tt>frem</tt>' instruction must be
2128<a href="#t_floating">floating point</a> values. Both arguments must have
2129identical types.</p>
2130<h5>Semantics:</h5>
2131<p>This instruction returns the <i>remainder</i> of a division.</p>
2132<h5>Example:</h5>
2133<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]2134</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2135</div>
Robert Bocchino820bc75b2006-02-17 21:18:08 +0000[diff] [blame]2136
Reid Spencer2ab01932007-02-02 13:57:07 +0000[diff] [blame]2137<!-- ======================================================================= -->
2138<div class="doc_subsection"> <a name="bitwiseops">Bitwise Binary
2139Operations</a> </div>
2140<div class="doc_text">
2141<p>Bitwise binary operators are used to do various forms of
2142bit-twiddling in a program. They are generally very efficient
2143instructions and can commonly be strength reduced from other
2144instructions. They require two operands, execute an operation on them,
2145and produce a single value. The resulting value of the bitwise binary
2146operators is always the same type as its first operand.</p>
2147</div>
2148
Reid Spencer04e259b2007-01-31 21:39:12 +0000[diff] [blame]2149<!-- _______________________________________________________________________ -->
2150<div class="doc_subsubsection"> <a name="i_shl">'<tt>shl</tt>'
2151Instruction</a> </div>
2152<div class="doc_text">
2153<h5>Syntax:</h5>
2154<pre> &lt;result&gt; = shl &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2155</pre>
2156<h5>Overview:</h5>
2157<p>The '<tt>shl</tt>' instruction returns the first operand shifted to
2158the left a specified number of bits.</p>
2159<h5>Arguments:</h5>
2160<p>Both arguments to the '<tt>shl</tt>' instruction must be the same <a
2161 href="#t_integer">integer</a> type.</p>
2162<h5>Semantics:</h5>
2163<p>The value produced is <tt>var1</tt> * 2<sup><tt>var2</tt></sup>.</p>
2164<h5>Example:</h5><pre>
2165 &lt;result&gt; = shl i32 4, %var <i>; yields {i32}: 4 &lt;&lt; %var</i>
2166 &lt;result&gt; = shl i32 4, 2 <i>; yields {i32}: 16</i>
2167 &lt;result&gt; = shl i32 1, 10 <i>; yields {i32}: 1024</i>
2168</pre>
2169</div>
2170<!-- _______________________________________________________________________ -->
2171<div class="doc_subsubsection"> <a name="i_lshr">'<tt>lshr</tt>'
2172Instruction</a> </div>
2173<div class="doc_text">
2174<h5>Syntax:</h5>
2175<pre> &lt;result&gt; = lshr &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2176</pre>
2177
2178<h5>Overview:</h5>
2179<p>The '<tt>lshr</tt>' instruction (logical shift right) returns the first
Jeff Cohen5819f182007-04-22 01:17:39 +0000[diff] [blame]2180operand shifted to the right a specified number of bits with zero fill.</p>
Reid Spencer04e259b2007-01-31 21:39:12 +0000[diff] [blame]2181
2182<h5>Arguments:</h5>
2183<p>Both arguments to the '<tt>lshr</tt>' instruction must be the same
2184<a href="#t_integer">integer</a> type.</p>
2185
2186<h5>Semantics:</h5>
2187<p>This instruction always performs a logical shift right operation. The most
2188significant bits of the result will be filled with zero bits after the
2189shift.</p>
2190
2191<h5>Example:</h5>
2192<pre>
2193 &lt;result&gt; = lshr i32 4, 1 <i>; yields {i32}:result = 2</i>
2194 &lt;result&gt; = lshr i32 4, 2 <i>; yields {i32}:result = 1</i>
2195 &lt;result&gt; = lshr i8 4, 3 <i>; yields {i8}:result = 0</i>
2196 &lt;result&gt; = lshr i8 -2, 1 <i>; yields {i8}:result = 0x7FFFFFFF </i>
2197</pre>
2198</div>
2199
Reid Spencer2ab01932007-02-02 13:57:07 +0000[diff] [blame]2200<!-- _______________________________________________________________________ -->
Reid Spencer04e259b2007-01-31 21:39:12 +0000[diff] [blame]2201<div class="doc_subsubsection"> <a name="i_ashr">'<tt>ashr</tt>'
2202Instruction</a> </div>
2203<div class="doc_text">
2204
2205<h5>Syntax:</h5>
2206<pre> &lt;result&gt; = ashr &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2207</pre>
2208
2209<h5>Overview:</h5>
2210<p>The '<tt>ashr</tt>' instruction (arithmetic shift right) returns the first
Jeff Cohen5819f182007-04-22 01:17:39 +0000[diff] [blame]2211operand shifted to the right a specified number of bits with sign extension.</p>
Reid Spencer04e259b2007-01-31 21:39:12 +0000[diff] [blame]2212
2213<h5>Arguments:</h5>
2214<p>Both arguments to the '<tt>ashr</tt>' instruction must be the same
2215<a href="#t_integer">integer</a> type.</p>
2216
2217<h5>Semantics:</h5>
2218<p>This instruction always performs an arithmetic shift right operation,
2219The most significant bits of the result will be filled with the sign bit
2220of <tt>var1</tt>.</p>
2221
2222<h5>Example:</h5>
2223<pre>
2224 &lt;result&gt; = ashr i32 4, 1 <i>; yields {i32}:result = 2</i>
2225 &lt;result&gt; = ashr i32 4, 2 <i>; yields {i32}:result = 1</i>
2226 &lt;result&gt; = ashr i8 4, 3 <i>; yields {i8}:result = 0</i>
2227 &lt;result&gt; = ashr i8 -2, 1 <i>; yields {i8}:result = -1</i>
2228</pre>
2229</div>
2230
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2231<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2232<div class="doc_subsubsection"> <a name="i_and">'<tt>and</tt>'
2233Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2234<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2235<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2236<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]2237</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2238<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2239<p>The '<tt>and</tt>' instruction returns the bitwise logical and of
2240its two operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2241<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2242<p>The two arguments to the '<tt>and</tt>' instruction must be <a
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]2243 href="#t_integer">integer</a> values. Both arguments must have
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2244identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2245<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2246<p>The truth table used for the '<tt>and</tt>' instruction is:</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2247<p> </p>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]2248<div style="align: center">
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2249<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2250 <tbody>
2251 <tr>
2252 <td>In0</td>
2253 <td>In1</td>
2254 <td>Out</td>
2255 </tr>
2256 <tr>
2257 <td>0</td>
2258 <td>0</td>
2259 <td>0</td>
2260 </tr>
2261 <tr>
2262 <td>0</td>
2263 <td>1</td>
2264 <td>0</td>
2265 </tr>
2266 <tr>
2267 <td>1</td>
2268 <td>0</td>
2269 <td>0</td>
2270 </tr>
2271 <tr>
2272 <td>1</td>
2273 <td>1</td>
2274 <td>1</td>
2275 </tr>
2276 </tbody>
2277</table>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]2278</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2279<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2280<pre> &lt;result&gt; = and i32 4, %var <i>; yields {i32}:result = 4 &amp; %var</i>
2281 &lt;result&gt; = and i32 15, 40 <i>; yields {i32}:result = 8</i>
2282 &lt;result&gt; = and i32 4, 8 <i>; yields {i32}:result = 0</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2283</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2284</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2285<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2286<div class="doc_subsubsection"> <a name="i_or">'<tt>or</tt>' Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2287<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2288<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2289<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]2290</pre>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2291<h5>Overview:</h5>
2292<p>The '<tt>or</tt>' instruction returns the bitwise logical inclusive
2293or of its two operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2294<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2295<p>The two arguments to the '<tt>or</tt>' instruction must be <a
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]2296 href="#t_integer">integer</a> values. Both arguments must have
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2297identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2298<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2299<p>The truth table used for the '<tt>or</tt>' instruction is:</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2300<p> </p>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]2301<div style="align: center">
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2302<table border="1" cellspacing="0" cellpadding="4">
2303 <tbody>
2304 <tr>
2305 <td>In0</td>
2306 <td>In1</td>
2307 <td>Out</td>
2308 </tr>
2309 <tr>
2310 <td>0</td>
2311 <td>0</td>
2312 <td>0</td>
2313 </tr>
2314 <tr>
2315 <td>0</td>
2316 <td>1</td>
2317 <td>1</td>
2318 </tr>
2319 <tr>
2320 <td>1</td>
2321 <td>0</td>
2322 <td>1</td>
2323 </tr>
2324 <tr>
2325 <td>1</td>
2326 <td>1</td>
2327 <td>1</td>
2328 </tr>
2329 </tbody>
2330</table>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]2331</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2332<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2333<pre> &lt;result&gt; = or i32 4, %var <i>; yields {i32}:result = 4 | %var</i>
2334 &lt;result&gt; = or i32 15, 40 <i>; yields {i32}:result = 47</i>
2335 &lt;result&gt; = or i32 4, 8 <i>; yields {i32}:result = 12</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2336</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2337</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2338<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2339<div class="doc_subsubsection"> <a name="i_xor">'<tt>xor</tt>'
2340Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2341<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2342<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2343<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]2344</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2345<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2346<p>The '<tt>xor</tt>' instruction returns the bitwise logical exclusive
2347or of its two operands. The <tt>xor</tt> is used to implement the
2348"one's complement" operation, which is the "~" operator in C.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2349<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2350<p>The two arguments to the '<tt>xor</tt>' instruction must be <a
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]2351 href="#t_integer">integer</a> values. Both arguments must have
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2352identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2353<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2354<p>The truth table used for the '<tt>xor</tt>' instruction is:</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2355<p> </p>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]2356<div style="align: center">
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2357<table border="1" cellspacing="0" cellpadding="4">
2358 <tbody>
2359 <tr>
2360 <td>In0</td>
2361 <td>In1</td>
2362 <td>Out</td>
2363 </tr>
2364 <tr>
2365 <td>0</td>
2366 <td>0</td>
2367 <td>0</td>
2368 </tr>
2369 <tr>
2370 <td>0</td>
2371 <td>1</td>
2372 <td>1</td>
2373 </tr>
2374 <tr>
2375 <td>1</td>
2376 <td>0</td>
2377 <td>1</td>
2378 </tr>
2379 <tr>
2380 <td>1</td>
2381 <td>1</td>
2382 <td>0</td>
2383 </tr>
2384 </tbody>
2385</table>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]2386</div>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2387<p> </p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2388<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2389<pre> &lt;result&gt; = xor i32 4, %var <i>; yields {i32}:result = 4 ^ %var</i>
2390 &lt;result&gt; = xor i32 15, 40 <i>; yields {i32}:result = 39</i>
2391 &lt;result&gt; = xor i32 4, 8 <i>; yields {i32}:result = 12</i>
2392 &lt;result&gt; = xor i32 %V, -1 <i>; yields {i32}:result = ~%V</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2393</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2394</div>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2395
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2396<!-- ======================================================================= -->
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2397<div class="doc_subsection">
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2398 <a name="vectorops">Vector Operations</a>
2399</div>
2400
2401<div class="doc_text">
2402
2403<p>LLVM supports several instructions to represent vector operations in a
Jeff Cohen5819f182007-04-22 01:17:39 +0000[diff] [blame]2404target-independent manner. These instructions cover the element-access and
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2405vector-specific operations needed to process vectors effectively. While LLVM
2406does directly support these vector operations, many sophisticated algorithms
2407will want to use target-specific intrinsics to take full advantage of a specific
2408target.</p>
2409
2410</div>
2411
2412<!-- _______________________________________________________________________ -->
2413<div class="doc_subsubsection">
2414 <a name="i_extractelement">'<tt>extractelement</tt>' Instruction</a>
2415</div>
2416
2417<div class="doc_text">
2418
2419<h5>Syntax:</h5>
2420
2421<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2422 &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]2423</pre>
2424
2425<h5>Overview:</h5>
2426
2427<p>
2428The '<tt>extractelement</tt>' instruction extracts a single scalar
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]2429element from a vector at a specified index.
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2430</p>
2431
2432
2433<h5>Arguments:</h5>
2434
2435<p>
2436The first operand of an '<tt>extractelement</tt>' instruction is a
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]2437value of <a href="#t_vector">vector</a> type. The second operand is
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2438an index indicating the position from which to extract the element.
2439The index may be a variable.</p>
2440
2441<h5>Semantics:</h5>
2442
2443<p>
2444The result is a scalar of the same type as the element type of
2445<tt>val</tt>. Its value is the value at position <tt>idx</tt> of
2446<tt>val</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
2447results are undefined.
2448</p>
2449
2450<h5>Example:</h5>
2451
2452<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2453 %result = extractelement &lt;4 x i32&gt; %vec, i32 0 <i>; yields i32</i>
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2454</pre>
2455</div>
2456
2457
2458<!-- _______________________________________________________________________ -->
2459<div class="doc_subsubsection">
2460 <a name="i_insertelement">'<tt>insertelement</tt>' Instruction</a>
2461</div>
2462
2463<div class="doc_text">
2464
2465<h5>Syntax:</h5>
2466
2467<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2468 &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]2469</pre>
2470
2471<h5>Overview:</h5>
2472
2473<p>
2474The '<tt>insertelement</tt>' instruction inserts a scalar
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]2475element into a vector at a specified index.
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2476</p>
2477
2478
2479<h5>Arguments:</h5>
2480
2481<p>
2482The first operand of an '<tt>insertelement</tt>' instruction is a
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]2483value of <a href="#t_vector">vector</a> type. The second operand is a
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2484scalar value whose type must equal the element type of the first
2485operand. The third operand is an index indicating the position at
2486which to insert the value. The index may be a variable.</p>
2487
2488<h5>Semantics:</h5>
2489
2490<p>
Reid Spencer404a3252007-02-15 03:07:05 +0000[diff] [blame]2491The result is a vector of the same type as <tt>val</tt>. Its
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2492element values are those of <tt>val</tt> except at position
2493<tt>idx</tt>, where it gets the value <tt>elt</tt>. If <tt>idx</tt>
2494exceeds the length of <tt>val</tt>, the results are undefined.
2495</p>
2496
2497<h5>Example:</h5>
2498
2499<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2500 %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]2501</pre>
2502</div>
2503
2504<!-- _______________________________________________________________________ -->
2505<div class="doc_subsubsection">
2506 <a name="i_shufflevector">'<tt>shufflevector</tt>' Instruction</a>
2507</div>
2508
2509<div class="doc_text">
2510
2511<h5>Syntax:</h5>
2512
2513<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2514 &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]2515</pre>
2516
2517<h5>Overview:</h5>
2518
2519<p>
2520The '<tt>shufflevector</tt>' instruction constructs a permutation of elements
2521from two input vectors, returning a vector of the same type.
2522</p>
2523
2524<h5>Arguments:</h5>
2525
2526<p>
2527The first two operands of a '<tt>shufflevector</tt>' instruction are vectors
2528with types that match each other and types that match the result of the
2529instruction. The third argument is a shuffle mask, which has the same number
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2530of elements as the other vector type, but whose element type is always 'i32'.
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2531</p>
2532
2533<p>
2534The shuffle mask operand is required to be a constant vector with either
2535constant integer or undef values.
2536</p>
2537
2538<h5>Semantics:</h5>
2539
2540<p>
2541The elements of the two input vectors are numbered from left to right across
2542both of the vectors. The shuffle mask operand specifies, for each element of
2543the result vector, which element of the two input registers the result element
2544gets. The element selector may be undef (meaning "don't care") and the second
2545operand may be undef if performing a shuffle from only one vector.
2546</p>
2547
2548<h5>Example:</h5>
2549
2550<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2551 %result = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; %v2,
Jeff Cohen5819f182007-04-22 01:17:39 +0000[diff] [blame]2552 &lt;4 x i32&gt; &lt;i32 0, i32 4, i32 1, i32 5&gt; <i>; yields &lt;4 x i32&gt;</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2553 %result = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; undef,
2554 &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]2555</pre>
2556</div>
2557
Tanya Lattnerb138bbe2006-04-14 19:24:33 +0000[diff] [blame]2558
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2559<!-- ======================================================================= -->
2560<div class="doc_subsection">
Chris Lattner6ab66722006-08-15 00:45:58 +0000[diff] [blame]2561 <a name="memoryops">Memory Access and Addressing Operations</a>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2562</div>
2563
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2564<div class="doc_text">
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2565
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2566<p>A key design point of an SSA-based representation is how it
2567represents memory. In LLVM, no memory locations are in SSA form, which
2568makes things very simple. This section describes how to read, write,
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]2569allocate, and free memory in LLVM.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2570
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2571</div>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2572
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2573<!-- _______________________________________________________________________ -->
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2574<div class="doc_subsubsection">
2575 <a name="i_malloc">'<tt>malloc</tt>' Instruction</a>
2576</div>
2577
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2578<div class="doc_text">
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2579
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2580<h5>Syntax:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2581
2582<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2583 &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]2584</pre>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2585
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2586<h5>Overview:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2587
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2588<p>The '<tt>malloc</tt>' instruction allocates memory from the system
2589heap and returns a pointer to it.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2590
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2591<h5>Arguments:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2592
2593<p>The '<tt>malloc</tt>' instruction allocates
2594<tt>sizeof(&lt;type&gt;)*NumElements</tt>
John Criswella92e5862004-02-24 16:13:56 +0000[diff] [blame]2595bytes of memory from the operating system and returns a pointer of the
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2596appropriate type to the program. If "NumElements" is specified, it is the
2597number of elements allocated. If an alignment is specified, the value result
2598of the allocation is guaranteed to be aligned to at least that boundary. If
2599not specified, or if zero, the target can choose to align the allocation on any
2600convenient boundary.</p>
2601
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2602<p>'<tt>type</tt>' must be a sized type.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2603
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2604<h5>Semantics:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2605
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2606<p>Memory is allocated using the system "<tt>malloc</tt>" function, and
2607a pointer is returned.</p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2608
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2609<h5>Example:</h5>
2610
2611<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2612 %array = malloc [4 x i8 ] <i>; yields {[%4 x i8]*}:array</i>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2613
Bill Wendling2d8b9a82007-05-29 09:42:13 +0000[diff] [blame]2614 %size = <a href="#i_add">add</a> i32 2, 2 <i>; yields {i32}:size = i32 4</i>
2615 %array1 = malloc i8, i32 4 <i>; yields {i8*}:array1</i>
2616 %array2 = malloc [12 x i8], i32 %size <i>; yields {[12 x i8]*}:array2</i>
2617 %array3 = malloc i32, i32 4, align 1024 <i>; yields {i32*}:array3</i>
2618 %array4 = malloc i32, align 1024 <i>; yields {i32*}:array4</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2619</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2620</div>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2621
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2622<!-- _______________________________________________________________________ -->
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2623<div class="doc_subsubsection">
2624 <a name="i_free">'<tt>free</tt>' Instruction</a>
2625</div>
2626
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2627<div class="doc_text">
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2628
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2629<h5>Syntax:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2630
2631<pre>
2632 free &lt;type&gt; &lt;value&gt; <i>; yields {void}</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2633</pre>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2634
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2635<h5>Overview:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2636
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2637<p>The '<tt>free</tt>' instruction returns memory back to the unused
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]2638memory heap to be reallocated in the future.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2639
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2640<h5>Arguments:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2641
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2642<p>'<tt>value</tt>' shall be a pointer value that points to a value
2643that was allocated with the '<tt><a href="#i_malloc">malloc</a></tt>'
2644instruction.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2645
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2646<h5>Semantics:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2647
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]2648<p>Access to the memory pointed to by the pointer is no longer defined
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2649after this instruction executes.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2650
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2651<h5>Example:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2652
2653<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2654 %array = <a href="#i_malloc">malloc</a> [4 x i8] <i>; yields {[4 x i8]*}:array</i>
2655 free [4 x i8]* %array
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2656</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2657</div>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2658
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2659<!-- _______________________________________________________________________ -->
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2660<div class="doc_subsubsection">
2661 <a name="i_alloca">'<tt>alloca</tt>' Instruction</a>
2662</div>
2663
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2664<div class="doc_text">
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2665
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2666<h5>Syntax:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2667
2668<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2669 &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]2670</pre>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2671
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2672<h5>Overview:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2673
Jeff Cohen5819f182007-04-22 01:17:39 +0000[diff] [blame]2674<p>The '<tt>alloca</tt>' instruction allocates memory on the stack frame of the
2675currently executing function, to be automatically released when this function
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2676returns to its caller.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2677
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2678<h5>Arguments:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2679
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]2680<p>The '<tt>alloca</tt>' instruction allocates <tt>sizeof(&lt;type&gt;)*NumElements</tt>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2681bytes of memory on the runtime stack, returning a pointer of the
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2682appropriate type to the program. If "NumElements" is specified, it is the
2683number of elements allocated. If an alignment is specified, the value result
2684of the allocation is guaranteed to be aligned to at least that boundary. If
2685not specified, or if zero, the target can choose to align the allocation on any
2686convenient boundary.</p>
2687
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2688<p>'<tt>type</tt>' may be any sized type.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2689
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2690<h5>Semantics:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2691
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]2692<p>Memory is allocated; a pointer is returned. '<tt>alloca</tt>'d
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2693memory is automatically released when the function returns. The '<tt>alloca</tt>'
2694instruction is commonly used to represent automatic variables that must
2695have an address available. When the function returns (either with the <tt><a
John Criswellc932bef2005-05-12 16:55:34 +0000[diff] [blame]2696 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]2697instructions), the memory is reclaimed.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2698
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2699<h5>Example:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2700
2701<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2702 %ptr = alloca i32 <i>; yields {i32*}:ptr</i>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]2703 %ptr = alloca i32, i32 4 <i>; yields {i32*}:ptr</i>
2704 %ptr = alloca i32, i32 4, align 1024 <i>; yields {i32*}:ptr</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2705 %ptr = alloca i32, align 1024 <i>; yields {i32*}:ptr</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2706</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2707</div>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2708
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2709<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2710<div class="doc_subsubsection"> <a name="i_load">'<tt>load</tt>'
2711Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2712<div class="doc_text">
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2713<h5>Syntax:</h5>
Christopher Lambbff50202007-04-21 08:16:25 +0000[diff] [blame]2714<pre> &lt;result&gt; = load &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;]<br> &lt;result&gt; = volatile load &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;]<br></pre>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2715<h5>Overview:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2716<p>The '<tt>load</tt>' instruction is used to read from memory.</p>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2717<h5>Arguments:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2718<p>The argument to the '<tt>load</tt>' instruction specifies the memory
John Criswell4c0cf7f2005-10-24 16:17:18 +0000[diff] [blame]2719address from which to load. The pointer must point to a <a
Chris Lattner10ee9652004-06-03 22:57:15 +0000[diff] [blame]2720 href="#t_firstclass">first class</a> type. If the <tt>load</tt> is
John Criswell4c0cf7f2005-10-24 16:17:18 +0000[diff] [blame]2721marked as <tt>volatile</tt>, then the optimizer is not allowed to modify
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2722the number or order of execution of this <tt>load</tt> with other
2723volatile <tt>load</tt> and <tt><a href="#i_store">store</a></tt>
2724instructions. </p>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2725<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2726<p>The location of memory pointed to is loaded.</p>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2727<h5>Examples:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2728<pre> %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2729 <a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2730 href="#i_store">store</a> i32 3, i32* %ptr <i>; yields {void}</i>
2731 %val = load i32* %ptr <i>; yields {i32}:val = i32 3</i>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2732</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2733</div>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2734<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2735<div class="doc_subsubsection"> <a name="i_store">'<tt>store</tt>'
2736Instruction</a> </div>
Reid Spencera89fb182006-11-09 21:18:01 +0000[diff] [blame]2737<div class="doc_text">
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2738<h5>Syntax:</h5>
Christopher Lambbff50202007-04-21 08:16:25 +0000[diff] [blame]2739<pre> store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;] <i>; yields {void}</i>
2740 volatile store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;] <i>; yields {void}</i>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2741</pre>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2742<h5>Overview:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2743<p>The '<tt>store</tt>' instruction is used to write to memory.</p>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2744<h5>Arguments:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2745<p>There are two arguments to the '<tt>store</tt>' instruction: a value
Jeff Cohen5819f182007-04-22 01:17:39 +0000[diff] [blame]2746to store and an address at which to store it. The type of the '<tt>&lt;pointer&gt;</tt>'
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2747operand must be a pointer to the type of the '<tt>&lt;value&gt;</tt>'
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]2748operand. If the <tt>store</tt> is marked as <tt>volatile</tt>, then the
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2749optimizer is not allowed to modify the number or order of execution of
2750this <tt>store</tt> with other volatile <tt>load</tt> and <tt><a
2751 href="#i_store">store</a></tt> instructions.</p>
2752<h5>Semantics:</h5>
2753<p>The contents of memory are updated to contain '<tt>&lt;value&gt;</tt>'
2754at the location specified by the '<tt>&lt;pointer&gt;</tt>' operand.</p>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2755<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2756<pre> %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2757 <a
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2758 href="#i_store">store</a> i32 3, i32* %ptr <i>; yields {void}</i>
2759 %val = load i32* %ptr <i>; yields {i32}:val = i32 3</i>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2760</pre>
Reid Spencer443460a2006-11-09 21:15:49 +0000[diff] [blame]2761</div>
2762
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2763<!-- _______________________________________________________________________ -->
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2764<div class="doc_subsubsection">
2765 <a name="i_getelementptr">'<tt>getelementptr</tt>' Instruction</a>
2766</div>
2767
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2768<div class="doc_text">
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]2769<h5>Syntax:</h5>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2770<pre>
2771 &lt;result&gt; = getelementptr &lt;ty&gt;* &lt;ptrval&gt;{, &lt;ty&gt; &lt;idx&gt;}*
2772</pre>
2773
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]2774<h5>Overview:</h5>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2775
2776<p>
2777The '<tt>getelementptr</tt>' instruction is used to get the address of a
2778subelement of an aggregate data structure.</p>
2779
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]2780<h5>Arguments:</h5>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2781
Reid Spencercee005c2006-12-04 21:29:24 +0000[diff] [blame]2782<p>This instruction takes a list of integer operands that indicate what
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2783elements of the aggregate object to index to. The actual types of the arguments
2784provided depend on the type of the first pointer argument. The
2785'<tt>getelementptr</tt>' instruction is used to index down through the type
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]2786levels 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]2787structure, only <tt>i32</tt> integer constants are allowed. When indexing
Reid Spencercee005c2006-12-04 21:29:24 +0000[diff] [blame]2788into an array or pointer, only integers of 32 or 64 bits are allowed, and will
2789be sign extended to 64-bit values.</p>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2790
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2791<p>For example, let's consider a C code fragment and how it gets
2792compiled to LLVM:</p>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2793
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]2794<div class="doc_code">
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2795<pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]2796struct RT {
2797 char A;
Chris Lattnera446f1b2007-05-29 15:43:56 +0000[diff] [blame]2798 int B[10][20];
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]2799 char C;
2800};
2801struct ST {
Chris Lattnera446f1b2007-05-29 15:43:56 +0000[diff] [blame]2802 int X;
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]2803 double Y;
2804 struct RT Z;
2805};
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2806
Chris Lattnera446f1b2007-05-29 15:43:56 +0000[diff] [blame]2807int *foo(struct ST *s) {
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]2808 return &amp;s[1].Z.B[5][13];
2809}
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2810</pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]2811</div>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2812
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2813<p>The LLVM code generated by the GCC frontend is:</p>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2814
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]2815<div class="doc_code">
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2816<pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]2817%RT = type { i8 , [10 x [20 x i32]], i8 }
2818%ST = type { i32, double, %RT }
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2819
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]2820define i32* %foo(%ST* %s) {
2821entry:
2822 %reg = getelementptr %ST* %s, i32 1, i32 2, i32 1, i32 5, i32 13
2823 ret i32* %reg
2824}
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2825</pre>
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]2826</div>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2827
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]2828<h5>Semantics:</h5>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2829
2830<p>The index types specified for the '<tt>getelementptr</tt>' instruction depend
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]2831on the pointer type that is being indexed into. <a href="#t_pointer">Pointer</a>
Reid Spencercee005c2006-12-04 21:29:24 +0000[diff] [blame]2832and <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]2833<a href="#t_integer">integer</a> type but the value will always be sign extended
Jeff Cohen5819f182007-04-22 01:17:39 +0000[diff] [blame]2834to 64-bits. <a href="#t_struct">Structure</a> types require <tt>i32</tt>
Reid Spencerc0312692006-12-03 16:53:48 +0000[diff] [blame]2835<b>constants</b>.</p>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2836
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2837<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]2838type, which is a pointer, yielding a '<tt>%ST</tt>' = '<tt>{ i32, double, %RT
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2839}</tt>' type, a structure. The second index indexes into the third element of
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2840the structure, yielding a '<tt>%RT</tt>' = '<tt>{ i8 , [10 x [20 x i32]],
2841i8 }</tt>' type, another structure. The third index indexes into the second
2842element of the structure, yielding a '<tt>[10 x [20 x i32]]</tt>' type, an
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2843array. The two dimensions of the array are subscripted into, yielding an
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2844'<tt>i32</tt>' type. The '<tt>getelementptr</tt>' instruction returns a pointer
2845to this element, thus computing a value of '<tt>i32*</tt>' type.</p>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2846
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2847<p>Note that it is perfectly legal to index partially through a
2848structure, returning a pointer to an inner element. Because of this,
2849the LLVM code for the given testcase is equivalent to:</p>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2850
2851<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2852 define i32* %foo(%ST* %s) {
2853 %t1 = getelementptr %ST* %s, i32 1 <i>; yields %ST*:%t1</i>
Jeff Cohen5819f182007-04-22 01:17:39 +0000[diff] [blame]2854 %t2 = getelementptr %ST* %t1, i32 0, i32 2 <i>; yields %RT*:%t2</i>
2855 %t3 = getelementptr %RT* %t2, i32 0, i32 1 <i>; yields [10 x [20 x i32]]*:%t3</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2856 %t4 = getelementptr [10 x [20 x i32]]* %t3, i32 0, i32 5 <i>; yields [20 x i32]*:%t4</i>
2857 %t5 = getelementptr [20 x i32]* %t4, i32 0, i32 13 <i>; yields i32*:%t5</i>
2858 ret i32* %t5
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2859 }
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]2860</pre>
Chris Lattnerc0ad71e2005-06-24 17:22:57 +0000[diff] [blame]2861
2862<p>Note that it is undefined to access an array out of bounds: array and
2863pointer indexes must always be within the defined bounds of the array type.
2864The one exception for this rules is zero length arrays. These arrays are
2865defined to be accessible as variable length arrays, which requires access
2866beyond the zero'th element.</p>
2867
Chris Lattner6ab66722006-08-15 00:45:58 +0000[diff] [blame]2868<p>The getelementptr instruction is often confusing. For some more insight
2869into how it works, see <a href="GetElementPtr.html">the getelementptr
2870FAQ</a>.</p>
2871
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]2872<h5>Example:</h5>
Chris Lattnerc0ad71e2005-06-24 17:22:57 +0000[diff] [blame]2873
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2874<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2875 <i>; yields [12 x i8]*:aptr</i>
2876 %aptr = getelementptr {i32, [12 x i8]}* %sptr, i64 0, i32 1
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2877</pre>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2878</div>
Reid Spencer443460a2006-11-09 21:15:49 +0000[diff] [blame]2879
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2880<!-- ======================================================================= -->
Reid Spencer97c5fa42006-11-08 01:18:52 +0000[diff] [blame]2881<div class="doc_subsection"> <a name="convertops">Conversion Operations</a>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2882</div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2883<div class="doc_text">
Reid Spencer97c5fa42006-11-08 01:18:52 +0000[diff] [blame]2884<p>The instructions in this category are the conversion instructions (casting)
2885which all take a single operand and a type. They perform various bit conversions
2886on the operand.</p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2887</div>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]2888
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]2889<!-- _______________________________________________________________________ -->
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]2890<div class="doc_subsubsection">
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2891 <a name="i_trunc">'<tt>trunc .. to</tt>' Instruction</a>
2892</div>
2893<div class="doc_text">
2894
2895<h5>Syntax:</h5>
2896<pre>
2897 &lt;result&gt; = trunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2898</pre>
2899
2900<h5>Overview:</h5>
2901<p>
2902The '<tt>trunc</tt>' instruction truncates its operand to the type <tt>ty2</tt>.
2903</p>
2904
2905<h5>Arguments:</h5>
2906<p>
2907The '<tt>trunc</tt>' instruction takes a <tt>value</tt> to trunc, which must
2908be 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]2909and type of the result, which must be an <a href="#t_integer">integer</a>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]2910type. The bit size of <tt>value</tt> must be larger than the bit size of
2911<tt>ty2</tt>. Equal sized types are not allowed.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2912
2913<h5>Semantics:</h5>
2914<p>
2915The '<tt>trunc</tt>' instruction truncates the high order bits in <tt>value</tt>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]2916and converts the remaining bits to <tt>ty2</tt>. Since the source size must be
2917larger than the destination size, <tt>trunc</tt> cannot be a <i>no-op cast</i>.
2918It will always truncate bits.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2919
2920<h5>Example:</h5>
2921<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2922 %X = trunc i32 257 to i8 <i>; yields i8:1</i>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]2923 %Y = trunc i32 123 to i1 <i>; yields i1:true</i>
2924 %Y = trunc i32 122 to i1 <i>; yields i1:false</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2925</pre>
2926</div>
2927
2928<!-- _______________________________________________________________________ -->
2929<div class="doc_subsubsection">
2930 <a name="i_zext">'<tt>zext .. to</tt>' Instruction</a>
2931</div>
2932<div class="doc_text">
2933
2934<h5>Syntax:</h5>
2935<pre>
2936 &lt;result&gt; = zext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2937</pre>
2938
2939<h5>Overview:</h5>
2940<p>The '<tt>zext</tt>' instruction zero extends its operand to type
2941<tt>ty2</tt>.</p>
2942
2943
2944<h5>Arguments:</h5>
2945<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]2946<a href="#t_integer">integer</a> type, and a type to cast it to, which must
2947also be of <a href="#t_integer">integer</a> type. The bit size of the
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]2948<tt>value</tt> must be smaller than the bit size of the destination type,
2949<tt>ty2</tt>.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2950
2951<h5>Semantics:</h5>
2952<p>The <tt>zext</tt> fills the high order bits of the <tt>value</tt> with zero
Chris Lattnerc87f3df2007-05-24 19:13:27 +0000[diff] [blame]2953bits until it reaches the size of the destination type, <tt>ty2</tt>.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2954
Reid Spencer07c9c682007-01-12 15:46:11 +0000[diff] [blame]2955<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]2956
2957<h5>Example:</h5>
2958<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2959 %X = zext i32 257 to i64 <i>; yields i64:257</i>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]2960 %Y = zext i1 true to i32 <i>; yields i32:1</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2961</pre>
2962</div>
2963
2964<!-- _______________________________________________________________________ -->
2965<div class="doc_subsubsection">
2966 <a name="i_sext">'<tt>sext .. to</tt>' Instruction</a>
2967</div>
2968<div class="doc_text">
2969
2970<h5>Syntax:</h5>
2971<pre>
2972 &lt;result&gt; = sext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
2973</pre>
2974
2975<h5>Overview:</h5>
2976<p>The '<tt>sext</tt>' sign extends <tt>value</tt> to the type <tt>ty2</tt>.</p>
2977
2978<h5>Arguments:</h5>
2979<p>
2980The '<tt>sext</tt>' instruction takes a value to cast, which must be of
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]2981<a href="#t_integer">integer</a> type, and a type to cast it to, which must
2982also be of <a href="#t_integer">integer</a> type. The bit size of the
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]2983<tt>value</tt> must be smaller than the bit size of the destination type,
2984<tt>ty2</tt>.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2985
2986<h5>Semantics:</h5>
2987<p>
2988The '<tt>sext</tt>' instruction performs a sign extension by copying the sign
2989bit (highest order bit) of the <tt>value</tt> until it reaches the bit size of
Chris Lattnerc87f3df2007-05-24 19:13:27 +0000[diff] [blame]2990the type <tt>ty2</tt>.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2991
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]2992<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]2993
2994<h5>Example:</h5>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2995<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]2996 %X = sext i8 -1 to i16 <i>; yields i16 :65535</i>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]2997 %Y = sext i1 true to i32 <i>; yields i32:-1</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]2998</pre>
2999</div>
3000
3001<!-- _______________________________________________________________________ -->
3002<div class="doc_subsubsection">
Reid Spencer2e2740d2006-11-09 21:48:10 +0000[diff] [blame]3003 <a name="i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a>
3004</div>
3005
3006<div class="doc_text">
3007
3008<h5>Syntax:</h5>
3009
3010<pre>
3011 &lt;result&gt; = fptrunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3012</pre>
3013
3014<h5>Overview:</h5>
3015<p>The '<tt>fptrunc</tt>' instruction truncates <tt>value</tt> to type
3016<tt>ty2</tt>.</p>
3017
3018
3019<h5>Arguments:</h5>
3020<p>The '<tt>fptrunc</tt>' instruction takes a <a href="#t_floating">floating
3021 point</a> value to cast and a <a href="#t_floating">floating point</a> type to
3022cast it to. The size of <tt>value</tt> must be larger than the size of
3023<tt>ty2</tt>. This implies that <tt>fptrunc</tt> cannot be used to make a
3024<i>no-op cast</i>.</p>
3025
3026<h5>Semantics:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3027<p> The '<tt>fptrunc</tt>' instruction truncates a <tt>value</tt> from a larger
3028<a href="#t_floating">floating point</a> type to a smaller
3029<a href="#t_floating">floating point</a> type. If the value cannot fit within
3030the destination type, <tt>ty2</tt>, then the results are undefined.</p>
Reid Spencer2e2740d2006-11-09 21:48:10 +0000[diff] [blame]3031
3032<h5>Example:</h5>
3033<pre>
3034 %X = fptrunc double 123.0 to float <i>; yields float:123.0</i>
3035 %Y = fptrunc double 1.0E+300 to float <i>; yields undefined</i>
3036</pre>
3037</div>
3038
3039<!-- _______________________________________________________________________ -->
3040<div class="doc_subsubsection">
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3041 <a name="i_fpext">'<tt>fpext .. to</tt>' Instruction</a>
3042</div>
3043<div class="doc_text">
3044
3045<h5>Syntax:</h5>
3046<pre>
3047 &lt;result&gt; = fpext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3048</pre>
3049
3050<h5>Overview:</h5>
3051<p>The '<tt>fpext</tt>' extends a floating point <tt>value</tt> to a larger
3052floating point value.</p>
3053
3054<h5>Arguments:</h5>
3055<p>The '<tt>fpext</tt>' instruction takes a
3056<a href="#t_floating">floating point</a> <tt>value</tt> to cast,
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3057and a <a href="#t_floating">floating point</a> type to cast it to. The source
3058type must be smaller than the destination type.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3059
3060<h5>Semantics:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3061<p>The '<tt>fpext</tt>' instruction extends the <tt>value</tt> from a smaller
Duncan Sands16f122e2007-03-30 12:22:09 +0000[diff] [blame]3062<a href="#t_floating">floating point</a> type to a larger
3063<a href="#t_floating">floating point</a> type. The <tt>fpext</tt> cannot be
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3064used to make a <i>no-op cast</i> because it always changes bits. Use
Reid Spencer5b950642006-11-11 23:08:07 +0000[diff] [blame]3065<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]3066
3067<h5>Example:</h5>
3068<pre>
3069 %X = fpext float 3.1415 to double <i>; yields double:3.1415</i>
3070 %Y = fpext float 1.0 to float <i>; yields float:1.0 (no-op)</i>
3071</pre>
3072</div>
3073
3074<!-- _______________________________________________________________________ -->
3075<div class="doc_subsubsection">
Reid Spencer2eadb532007-01-21 00:29:26 +0000[diff] [blame]3076 <a name="i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3077</div>
3078<div class="doc_text">
3079
3080<h5>Syntax:</h5>
3081<pre>
3082 &lt;result&gt; = fp2uint &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3083</pre>
3084
3085<h5>Overview:</h5>
3086<p>The '<tt>fp2uint</tt>' converts a floating point <tt>value</tt> to its
3087unsigned integer equivalent of type <tt>ty2</tt>.
3088</p>
3089
3090<h5>Arguments:</h5>
3091<p>The '<tt>fp2uint</tt>' instruction takes a value to cast, which must be a
3092<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]3093must be an <a href="#t_integer">integer</a> type.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3094
3095<h5>Semantics:</h5>
3096<p> The '<tt>fp2uint</tt>' instruction converts its
3097<a href="#t_floating">floating point</a> operand into the nearest (rounding
3098towards zero) unsigned integer value. If the value cannot fit in <tt>ty2</tt>,
3099the results are undefined.</p>
3100
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]3101<p>When converting to i1, the conversion is done as a comparison against
3102zero. If the <tt>value</tt> was zero, the i1 result will be <tt>false</tt>.
3103If 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]3104
3105<h5>Example:</h5>
3106<pre>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]3107 %X = fp2uint double 123.0 to i32 <i>; yields i32:123</i>
3108 %Y = fp2uint float 1.0E+300 to i1 <i>; yields i1:true</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3109 %X = fp2uint float 1.04E+17 to i8 <i>; yields undefined:1</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3110</pre>
3111</div>
3112
3113<!-- _______________________________________________________________________ -->
3114<div class="doc_subsubsection">
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3115 <a name="i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3116</div>
3117<div class="doc_text">
3118
3119<h5>Syntax:</h5>
3120<pre>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3121 &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]3122</pre>
3123
3124<h5>Overview:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3125<p>The '<tt>fptosi</tt>' instruction converts
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3126<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]3127</p>
3128
3129
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]3130<h5>Arguments:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3131<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]3132<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]3133must also be an <a href="#t_integer">integer</a> type.</p>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]3134
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]3135<h5>Semantics:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3136<p>The '<tt>fptosi</tt>' instruction converts its
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3137<a href="#t_floating">floating point</a> operand into the nearest (rounding
3138towards zero) signed integer value. If the value cannot fit in <tt>ty2</tt>,
3139the results are undefined.</p>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]3140
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]3141<p>When converting to i1, the conversion is done as a comparison against
3142zero. If the <tt>value</tt> was zero, the i1 result will be <tt>false</tt>.
3143If 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]3144
Chris Lattner70de6632001-07-09 00:26:23 +0000[diff] [blame]3145<h5>Example:</h5>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]3146<pre>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]3147 %X = fptosi double -123.0 to i32 <i>; yields i32:-123</i>
3148 %Y = fptosi float 1.0E-247 to i1 <i>; yields i1:true</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3149 %X = fptosi float 1.04E+17 to i8 <i>; yields undefined:1</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3150</pre>
3151</div>
3152
3153<!-- _______________________________________________________________________ -->
3154<div class="doc_subsubsection">
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3155 <a name="i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3156</div>
3157<div class="doc_text">
3158
3159<h5>Syntax:</h5>
3160<pre>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3161 &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]3162</pre>
3163
3164<h5>Overview:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3165<p>The '<tt>uitofp</tt>' instruction regards <tt>value</tt> as an unsigned
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3166integer and converts that value to the <tt>ty2</tt> type.</p>
3167
3168
3169<h5>Arguments:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3170<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]3171<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]3172be a <a href="#t_floating">floating point</a> type.</p>
3173
3174<h5>Semantics:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3175<p>The '<tt>uitofp</tt>' instruction interprets its operand as an unsigned
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3176integer quantity and converts it to the corresponding floating point value. If
Jeff Cohenbeccb742007-04-22 14:56:37 +0000[diff] [blame]3177the value cannot fit in the floating point value, the results are undefined.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3178
3179
3180<h5>Example:</h5>
3181<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3182 %X = uitofp i32 257 to float <i>; yields float:257.0</i>
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3183 %Y = uitofp i8 -1 to double <i>; yields double:255.0</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3184</pre>
3185</div>
3186
3187<!-- _______________________________________________________________________ -->
3188<div class="doc_subsubsection">
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3189 <a name="i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3190</div>
3191<div class="doc_text">
3192
3193<h5>Syntax:</h5>
3194<pre>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3195 &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]3196</pre>
3197
3198<h5>Overview:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3199<p>The '<tt>sitofp</tt>' instruction regards <tt>value</tt> as a signed
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3200integer and converts that value to the <tt>ty2</tt> type.</p>
3201
3202<h5>Arguments:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3203<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]3204<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]3205a <a href="#t_floating">floating point</a> type.</p>
3206
3207<h5>Semantics:</h5>
Reid Spencer51b07252006-11-09 23:03:26 +0000[diff] [blame]3208<p>The '<tt>sitofp</tt>' instruction interprets its operand as a signed
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3209integer quantity and converts it to the corresponding floating point value. If
Jeff Cohenbeccb742007-04-22 14:56:37 +0000[diff] [blame]3210the value cannot fit in the floating point value, the results are undefined.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3211
3212<h5>Example:</h5>
3213<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3214 %X = sitofp i32 257 to float <i>; yields float:257.0</i>
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3215 %Y = sitofp i8 -1 to double <i>; yields double:-1.0</i>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3216</pre>
3217</div>
3218
3219<!-- _______________________________________________________________________ -->
3220<div class="doc_subsubsection">
Reid Spencerb7344ff2006-11-11 21:00:47 +0000[diff] [blame]3221 <a name="i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a>
3222</div>
3223<div class="doc_text">
3224
3225<h5>Syntax:</h5>
3226<pre>
3227 &lt;result&gt; = ptrtoint &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3228</pre>
3229
3230<h5>Overview:</h5>
3231<p>The '<tt>ptrtoint</tt>' instruction converts the pointer <tt>value</tt> to
3232the integer type <tt>ty2</tt>.</p>
3233
3234<h5>Arguments:</h5>
3235<p>The '<tt>ptrtoint</tt>' instruction takes a <tt>value</tt> to cast, which
Duncan Sands16f122e2007-03-30 12:22:09 +0000[diff] [blame]3236must be a <a href="#t_pointer">pointer</a> value, and a type to cast it to
Reid Spencerb7344ff2006-11-11 21:00:47 +0000[diff] [blame]3237<tt>ty2</tt>, which must be an <a href="#t_integer">integer</a> type.
3238
3239<h5>Semantics:</h5>
3240<p>The '<tt>ptrtoint</tt>' instruction converts <tt>value</tt> to integer type
3241<tt>ty2</tt> by interpreting the pointer value as an integer and either
3242truncating or zero extending that value to the size of the integer type. If
3243<tt>value</tt> is smaller than <tt>ty2</tt> then a zero extension is done. If
3244<tt>value</tt> is larger than <tt>ty2</tt> then a truncation is done. If they
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3245are the same size, then nothing is done (<i>no-op cast</i>) other than a type
3246change.</p>
Reid Spencerb7344ff2006-11-11 21:00:47 +0000[diff] [blame]3247
3248<h5>Example:</h5>
3249<pre>
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3250 %X = ptrtoint i32* %X to i8 <i>; yields truncation on 32-bit architecture</i>
3251 %Y = ptrtoint i32* %x to i64 <i>; yields zero extension on 32-bit architecture</i>
Reid Spencerb7344ff2006-11-11 21:00:47 +0000[diff] [blame]3252</pre>
3253</div>
3254
3255<!-- _______________________________________________________________________ -->
3256<div class="doc_subsubsection">
3257 <a name="i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a>
3258</div>
3259<div class="doc_text">
3260
3261<h5>Syntax:</h5>
3262<pre>
3263 &lt;result&gt; = inttoptr &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3264</pre>
3265
3266<h5>Overview:</h5>
3267<p>The '<tt>inttoptr</tt>' instruction converts an integer <tt>value</tt> to
3268a pointer type, <tt>ty2</tt>.</p>
3269
3270<h5>Arguments:</h5>
Duncan Sands16f122e2007-03-30 12:22:09 +0000[diff] [blame]3271<p>The '<tt>inttoptr</tt>' instruction takes an <a href="#t_integer">integer</a>
Reid Spencerb7344ff2006-11-11 21:00:47 +0000[diff] [blame]3272value to cast, and a type to cast it to, which must be a
Anton Korobeynikova0554d92007-01-12 19:20:47 +0000[diff] [blame]3273<a href="#t_pointer">pointer</a> type.
Reid Spencerb7344ff2006-11-11 21:00:47 +0000[diff] [blame]3274
3275<h5>Semantics:</h5>
3276<p>The '<tt>inttoptr</tt>' instruction converts <tt>value</tt> to type
3277<tt>ty2</tt> by applying either a zero extension or a truncation depending on
3278the size of the integer <tt>value</tt>. If <tt>value</tt> is larger than the
3279size of a pointer then a truncation is done. If <tt>value</tt> is smaller than
3280the size of a pointer then a zero extension is done. If they are the same size,
3281nothing is done (<i>no-op cast</i>).</p>
3282
3283<h5>Example:</h5>
3284<pre>
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3285 %X = inttoptr i32 255 to i32* <i>; yields zero extension on 64-bit architecture</i>
3286 %X = inttoptr i32 255 to i32* <i>; yields no-op on 32-bit architecture</i>
3287 %Y = inttoptr i64 0 to i32* <i>; yields truncation on 32-bit architecture</i>
Reid Spencerb7344ff2006-11-11 21:00:47 +0000[diff] [blame]3288</pre>
3289</div>
3290
3291<!-- _______________________________________________________________________ -->
3292<div class="doc_subsubsection">
Reid Spencer5b950642006-11-11 23:08:07 +0000[diff] [blame]3293 <a name="i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3294</div>
3295<div class="doc_text">
3296
3297<h5>Syntax:</h5>
3298<pre>
Reid Spencer5b950642006-11-11 23:08:07 +0000[diff] [blame]3299 &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]3300</pre>
3301
3302<h5>Overview:</h5>
Reid Spencer5b950642006-11-11 23:08:07 +0000[diff] [blame]3303<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3304<tt>ty2</tt> without changing any bits.</p>
3305
3306<h5>Arguments:</h5>
Reid Spencer5b950642006-11-11 23:08:07 +0000[diff] [blame]3307<p>The '<tt>bitcast</tt>' instruction takes a value to cast, which must be
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3308a first class value, and a type to cast it to, which must also be a <a
3309 href="#t_firstclass">first class</a> type. The bit sizes of <tt>value</tt>
Reid Spencere3db84c2007-01-09 20:08:58 +0000[diff] [blame]3310and the destination type, <tt>ty2</tt>, must be identical. If the source
3311type is a pointer, the destination type must also be a pointer.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3312
3313<h5>Semantics:</h5>
Reid Spencer5b950642006-11-11 23:08:07 +0000[diff] [blame]3314<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
Reid Spencerb7344ff2006-11-11 21:00:47 +0000[diff] [blame]3315<tt>ty2</tt>. It is always a <i>no-op cast</i> because no bits change with
3316this conversion. The conversion is done as if the <tt>value</tt> had been
3317stored to memory and read back as type <tt>ty2</tt>. Pointer types may only be
3318converted to other pointer types with this instruction. To convert pointers to
3319other types, use the <a href="#i_inttoptr">inttoptr</a> or
3320<a href="#i_ptrtoint">ptrtoint</a> instructions first.</p>
Reid Spencer59b6b7d2006-11-08 01:11:31 +0000[diff] [blame]3321
3322<h5>Example:</h5>
3323<pre>
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3324 %X = bitcast i8 255 to i8 <i>; yields i8 :-1</i>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3325 %Y = bitcast i32* %x to sint* <i>; yields sint*:%x</i>
3326 %Z = bitcast <2xint> %V to i64; <i>; yields i64: %V</i>
Chris Lattner70de6632001-07-09 00:26:23 +0000[diff] [blame]3327</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3328</div>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]3329
Reid Spencer97c5fa42006-11-08 01:18:52 +0000[diff] [blame]3330<!-- ======================================================================= -->
3331<div class="doc_subsection"> <a name="otherops">Other Operations</a> </div>
3332<div class="doc_text">
3333<p>The instructions in this category are the "miscellaneous"
3334instructions, which defy better classification.</p>
3335</div>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3336
3337<!-- _______________________________________________________________________ -->
3338<div class="doc_subsubsection"><a name="i_icmp">'<tt>icmp</tt>' Instruction</a>
3339</div>
3340<div class="doc_text">
3341<h5>Syntax:</h5>
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3342<pre> &lt;result&gt; = icmp &lt;cond&gt; &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {i1}:result</i>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3343</pre>
3344<h5>Overview:</h5>
3345<p>The '<tt>icmp</tt>' instruction returns a boolean value based on comparison
3346of its two integer operands.</p>
3347<h5>Arguments:</h5>
3348<p>The '<tt>icmp</tt>' instruction takes three operands. The first operand is
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3349the condition code indicating the kind of comparison to perform. It is not
3350a value, just a keyword. The possible condition code are:
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3351<ol>
3352 <li><tt>eq</tt>: equal</li>
3353 <li><tt>ne</tt>: not equal </li>
3354 <li><tt>ugt</tt>: unsigned greater than</li>
3355 <li><tt>uge</tt>: unsigned greater or equal</li>
3356 <li><tt>ult</tt>: unsigned less than</li>
3357 <li><tt>ule</tt>: unsigned less or equal</li>
3358 <li><tt>sgt</tt>: signed greater than</li>
3359 <li><tt>sge</tt>: signed greater or equal</li>
3360 <li><tt>slt</tt>: signed less than</li>
3361 <li><tt>sle</tt>: signed less or equal</li>
3362</ol>
Chris Lattnerc0f423a2007-01-15 01:54:13 +0000[diff] [blame]3363<p>The remaining two arguments must be <a href="#t_integer">integer</a> or
Reid Spencer784ef792007-01-04 05:19:58 +0000[diff] [blame]3364<a href="#t_pointer">pointer</a> typed. They must also be identical types.</p>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3365<h5>Semantics:</h5>
3366<p>The '<tt>icmp</tt>' compares <tt>var1</tt> and <tt>var2</tt> according to
3367the condition code given as <tt>cond</tt>. The comparison performed always
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]3368yields a <a href="#t_primitive">i1</a> result, as follows:
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3369<ol>
3370 <li><tt>eq</tt>: yields <tt>true</tt> if the operands are equal,
3371 <tt>false</tt> otherwise. No sign interpretation is necessary or performed.
3372 </li>
3373 <li><tt>ne</tt>: yields <tt>true</tt> if the operands are unequal,
3374 <tt>false</tt> otherwise. No sign interpretation is necessary or performed.
3375 <li><tt>ugt</tt>: interprets the operands as unsigned values and yields
3376 <tt>true</tt> if <tt>var1</tt> is greater than <tt>var2</tt>.</li>
3377 <li><tt>uge</tt>: interprets the operands as unsigned values and yields
3378 <tt>true</tt> if <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
3379 <li><tt>ult</tt>: interprets the operands as unsigned values and yields
3380 <tt>true</tt> if <tt>var1</tt> is less than <tt>var2</tt>.</li>
3381 <li><tt>ule</tt>: interprets the operands as unsigned values and yields
3382 <tt>true</tt> if <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
3383 <li><tt>sgt</tt>: interprets the operands as signed values and yields
3384 <tt>true</tt> if <tt>var1</tt> is greater than <tt>var2</tt>.</li>
3385 <li><tt>sge</tt>: interprets the operands as signed values and yields
3386 <tt>true</tt> if <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
3387 <li><tt>slt</tt>: interprets the operands as signed values and yields
3388 <tt>true</tt> if <tt>var1</tt> is less than <tt>var2</tt>.</li>
3389 <li><tt>sle</tt>: interprets the operands as signed values and yields
3390 <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]3391</ol>
3392<p>If the operands are <a href="#t_pointer">pointer</a> typed, the pointer
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3393values are compared as if they were integers.</p>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3394
3395<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3396<pre> &lt;result&gt; = icmp eq i32 4, 5 <i>; yields: result=false</i>
3397 &lt;result&gt; = icmp ne float* %X, %X <i>; yields: result=false</i>
3398 &lt;result&gt; = icmp ult i16 4, 5 <i>; yields: result=true</i>
3399 &lt;result&gt; = icmp sgt i16 4, 5 <i>; yields: result=false</i>
3400 &lt;result&gt; = icmp ule i16 -4, 5 <i>; yields: result=false</i>
3401 &lt;result&gt; = icmp sge i16 4, 5 <i>; yields: result=false</i>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3402</pre>
3403</div>
3404
3405<!-- _______________________________________________________________________ -->
3406<div class="doc_subsubsection"><a name="i_fcmp">'<tt>fcmp</tt>' Instruction</a>
3407</div>
3408<div class="doc_text">
3409<h5>Syntax:</h5>
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3410<pre> &lt;result&gt; = fcmp &lt;cond&gt; &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {i1}:result</i>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3411</pre>
3412<h5>Overview:</h5>
3413<p>The '<tt>fcmp</tt>' instruction returns a boolean value based on comparison
3414of its floating point operands.</p>
3415<h5>Arguments:</h5>
3416<p>The '<tt>fcmp</tt>' instruction takes three operands. The first operand is
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3417the condition code indicating the kind of comparison to perform. It is not
3418a value, just a keyword. The possible condition code are:
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3419<ol>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3420 <li><tt>false</tt>: no comparison, always returns false</li>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3421 <li><tt>oeq</tt>: ordered and equal</li>
3422 <li><tt>ogt</tt>: ordered and greater than </li>
3423 <li><tt>oge</tt>: ordered and greater than or equal</li>
3424 <li><tt>olt</tt>: ordered and less than </li>
3425 <li><tt>ole</tt>: ordered and less than or equal</li>
3426 <li><tt>one</tt>: ordered and not equal</li>
3427 <li><tt>ord</tt>: ordered (no nans)</li>
3428 <li><tt>ueq</tt>: unordered or equal</li>
3429 <li><tt>ugt</tt>: unordered or greater than </li>
3430 <li><tt>uge</tt>: unordered or greater than or equal</li>
3431 <li><tt>ult</tt>: unordered or less than </li>
3432 <li><tt>ule</tt>: unordered or less than or equal</li>
3433 <li><tt>une</tt>: unordered or not equal</li>
3434 <li><tt>uno</tt>: unordered (either nans)</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3435 <li><tt>true</tt>: no comparison, always returns true</li>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3436</ol>
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3437<p><i>Ordered</i> means that neither operand is a QNAN while
Reid Spencer02e0d1d2006-12-06 07:08:07 +0000[diff] [blame]3438<i>unordered</i> means that either operand may be a QNAN.</p>
Reid Spencer784ef792007-01-04 05:19:58 +0000[diff] [blame]3439<p>The <tt>val1</tt> and <tt>val2</tt> arguments must be
3440<a href="#t_floating">floating point</a> typed. They must have identical
3441types.</p>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3442<h5>Semantics:</h5>
3443<p>The '<tt>fcmp</tt>' compares <tt>var1</tt> and <tt>var2</tt> according to
3444the condition code given as <tt>cond</tt>. The comparison performed always
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]3445yields a <a href="#t_primitive">i1</a> result, as follows:
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3446<ol>
3447 <li><tt>false</tt>: always yields <tt>false</tt>, regardless of operands.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3448 <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]3449 <tt>var1</tt> is equal to <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3450 <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]3451 <tt>var1</tt> is greather than <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3452 <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]3453 <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3454 <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]3455 <tt>var1</tt> is less than <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3456 <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]3457 <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3458 <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]3459 <tt>var1</tt> is not equal to <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3460 <li><tt>ord</tt>: yields <tt>true</tt> if both operands are not a QNAN.</li>
3461 <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]3462 <tt>var1</tt> is equal to <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3463 <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]3464 <tt>var1</tt> is greater than <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3465 <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]3466 <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3467 <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]3468 <tt>var1</tt> is less than <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3469 <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]3470 <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3471 <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]3472 <tt>var1</tt> is not equal to <tt>var2</tt>.</li>
Reid Spencerf69acf32006-11-19 03:00:14 +0000[diff] [blame]3473 <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]3474 <li><tt>true</tt>: always yields <tt>true</tt>, regardless of operands.</li>
3475</ol>
Reid Spencerc828a0e2006-11-18 21:50:54 +0000[diff] [blame]3476
3477<h5>Example:</h5>
3478<pre> &lt;result&gt; = fcmp oeq float 4.0, 5.0 <i>; yields: result=false</i>
3479 &lt;result&gt; = icmp one float 4.0, 5.0 <i>; yields: result=true</i>
3480 &lt;result&gt; = icmp olt float 4.0, 5.0 <i>; yields: result=true</i>
3481 &lt;result&gt; = icmp ueq double 1.0, 2.0 <i>; yields: result=false</i>
3482</pre>
3483</div>
3484
Reid Spencer97c5fa42006-11-08 01:18:52 +0000[diff] [blame]3485<!-- _______________________________________________________________________ -->
3486<div class="doc_subsubsection"> <a name="i_phi">'<tt>phi</tt>'
3487Instruction</a> </div>
3488<div class="doc_text">
3489<h5>Syntax:</h5>
3490<pre> &lt;result&gt; = phi &lt;ty&gt; [ &lt;val0&gt;, &lt;label0&gt;], ...<br></pre>
3491<h5>Overview:</h5>
3492<p>The '<tt>phi</tt>' instruction is used to implement the &#966; node in
3493the SSA graph representing the function.</p>
3494<h5>Arguments:</h5>
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3495<p>The type of the incoming values is specified with the first type
Reid Spencer97c5fa42006-11-08 01:18:52 +0000[diff] [blame]3496field. After this, the '<tt>phi</tt>' instruction takes a list of pairs
3497as arguments, with one pair for each predecessor basic block of the
3498current block. Only values of <a href="#t_firstclass">first class</a>
3499type may be used as the value arguments to the PHI node. Only labels
3500may be used as the label arguments.</p>
3501<p>There must be no non-phi instructions between the start of a basic
3502block and the PHI instructions: i.e. PHI instructions must be first in
3503a basic block.</p>
3504<h5>Semantics:</h5>
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3505<p>At runtime, the '<tt>phi</tt>' instruction logically takes on the value
3506specified by the pair corresponding to the predecessor basic block that executed
3507just prior to the current block.</p>
Reid Spencer97c5fa42006-11-08 01:18:52 +0000[diff] [blame]3508<h5>Example:</h5>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3509<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]3510</div>
3511
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]3512<!-- _______________________________________________________________________ -->
3513<div class="doc_subsubsection">
3514 <a name="i_select">'<tt>select</tt>' Instruction</a>
3515</div>
3516
3517<div class="doc_text">
3518
3519<h5>Syntax:</h5>
3520
3521<pre>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]3522 &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]3523</pre>
3524
3525<h5>Overview:</h5>
3526
3527<p>
3528The '<tt>select</tt>' instruction is used to choose one value based on a
3529condition, without branching.
3530</p>
3531
3532
3533<h5>Arguments:</h5>
3534
3535<p>
3536The '<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.
3537</p>
3538
3539<h5>Semantics:</h5>
3540
3541<p>
3542If the boolean condition evaluates to true, the instruction returns the first
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]3543value argument; otherwise, it returns the second value argument.
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]3544</p>
3545
3546<h5>Example:</h5>
3547
3548<pre>
Reid Spencer36a15422007-01-12 03:35:51 +0000[diff] [blame]3549 %X = select i1 true, i8 17, i8 42 <i>; yields i8:17</i>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]3550</pre>
3551</div>
3552
Robert Bocchinof72fdfe2006-01-15 20:48:27 +0000[diff] [blame]3553
3554<!-- _______________________________________________________________________ -->
3555<div class="doc_subsubsection">
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3556 <a name="i_call">'<tt>call</tt>' Instruction</a>
3557</div>
3558
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3559<div class="doc_text">
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3560
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]3561<h5>Syntax:</h5>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3562<pre>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]3563 &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]3564</pre>
3565
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]3566<h5>Overview:</h5>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3567
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3568<p>The '<tt>call</tt>' instruction represents a simple function call.</p>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3569
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]3570<h5>Arguments:</h5>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3571
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3572<p>This instruction requires several arguments:</p>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3573
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]3574<ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3575 <li>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]3576 <p>The optional "tail" marker indicates whether the callee function accesses
3577 any allocas or varargs in the caller. If the "tail" marker is present, the
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3578 function call is eligible for tail call optimization. Note that calls may
3579 be marked "tail" even if they do not occur before a <a
3580 href="#i_ret"><tt>ret</tt></a> instruction.
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3581 </li>
3582 <li>
Duncan Sands16f122e2007-03-30 12:22:09 +0000[diff] [blame]3583 <p>The optional "cconv" marker indicates which <a href="#callingconv">calling
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]3584 convention</a> the call should use. If none is specified, the call defaults
3585 to using C calling conventions.
3586 </li>
3587 <li>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3588 <p>'<tt>ty</tt>': shall be the signature of the pointer to function value
3589 being invoked. The argument types must match the types implied by this
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]3590 signature. This type can be omitted if the function is not varargs and
3591 if the function type does not return a pointer to a function.</p>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3592 </li>
3593 <li>
3594 <p>'<tt>fnptrval</tt>': An LLVM value containing a pointer to a function to
3595 be invoked. In most cases, this is a direct function invocation, but
3596 indirect <tt>call</tt>s are just as possible, calling an arbitrary pointer
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]3597 to function value.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3598 </li>
3599 <li>
3600 <p>'<tt>function args</tt>': argument list whose types match the
Reid Spencerd845d162005-05-01 22:22:57 +0000[diff] [blame]3601 function signature argument types. All arguments must be of
3602 <a href="#t_firstclass">first class</a> type. If the function signature
3603 indicates the function accepts a variable number of arguments, the extra
3604 arguments can be specified.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3605 </li>
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]3606</ol>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3607
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]3608<h5>Semantics:</h5>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3609
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3610<p>The '<tt>call</tt>' instruction is used to cause control flow to
3611transfer to a specified function, with its incoming arguments bound to
3612the specified values. Upon a '<tt><a href="#i_ret">ret</a></tt>'
3613instruction in the called function, control flow continues with the
3614instruction after the function call, and the return value of the
3615function is bound to the result argument. This is a simpler case of
3616the <a href="#i_invoke">invoke</a> instruction.</p>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3617
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]3618<h5>Example:</h5>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3619
3620<pre>
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3621 %retval = call i32 %test(i32 %argc)
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3622 call i32(i8 *, ...) *%printf(i8 * %msg, i32 12, i8 42);
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3623 %X = tail call i32 %foo()
3624 %Y = tail call <a href="#callingconv">fastcc</a> i32 %foo()
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]3625</pre>
3626
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3627</div>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3628
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3629<!-- _______________________________________________________________________ -->
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3630<div class="doc_subsubsection">
Chris Lattner33337472006-01-13 23:26:01 +0000[diff] [blame]3631 <a name="i_va_arg">'<tt>va_arg</tt>' Instruction</a>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3632</div>
3633
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3634<div class="doc_text">
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3635
Chris Lattner26ca62e2003-10-18 05:51:36 +0000[diff] [blame]3636<h5>Syntax:</h5>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3637
3638<pre>
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3639 &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]3640</pre>
3641
Chris Lattner26ca62e2003-10-18 05:51:36 +0000[diff] [blame]3642<h5>Overview:</h5>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3643
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3644<p>The '<tt>va_arg</tt>' instruction is used to access arguments passed through
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3645the "variable argument" area of a function call. It is used to implement the
3646<tt>va_arg</tt> macro in C.</p>
3647
Chris Lattner26ca62e2003-10-18 05:51:36 +0000[diff] [blame]3648<h5>Arguments:</h5>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3649
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3650<p>This instruction takes a <tt>va_list*</tt> value and the type of
3651the argument. It returns a value of the specified argument type and
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3652increments the <tt>va_list</tt> to point to the next argument. The
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3653actual type of <tt>va_list</tt> is target specific.</p>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3654
Chris Lattner26ca62e2003-10-18 05:51:36 +0000[diff] [blame]3655<h5>Semantics:</h5>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3656
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3657<p>The '<tt>va_arg</tt>' instruction loads an argument of the specified
3658type from the specified <tt>va_list</tt> and causes the
3659<tt>va_list</tt> to point to the next argument. For more information,
3660see the variable argument handling <a href="#int_varargs">Intrinsic
3661Functions</a>.</p>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3662
3663<p>It is legal for this instruction to be called in a function which does not
3664take a variable number of arguments, for example, the <tt>vfprintf</tt>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3665function.</p>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3666
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3667<p><tt>va_arg</tt> is an LLVM instruction instead of an <a
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]3668href="#intrinsics">intrinsic function</a> because it takes a type as an
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3669argument.</p>
3670
Chris Lattner26ca62e2003-10-18 05:51:36 +0000[diff] [blame]3671<h5>Example:</h5>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]3672
3673<p>See the <a href="#int_varargs">variable argument processing</a> section.</p>
3674
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3675</div>
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3676
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3677<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3678<div class="doc_section"> <a name="intrinsics">Intrinsic Functions</a> </div>
3679<!-- *********************************************************************** -->
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3680
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3681<div class="doc_text">
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3682
3683<p>LLVM supports the notion of an "intrinsic function". These functions have
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]3684well known names and semantics and are required to follow certain restrictions.
3685Overall, these intrinsics represent an extension mechanism for the LLVM
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3686language that does not require changing all of the transformations in LLVM when
3687adding to the language (or the bytecode reader/writer, the parser, etc...).</p>
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3688
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]3689<p>Intrinsic function names must all start with an "<tt>llvm.</tt>" prefix. This
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3690prefix is reserved in LLVM for intrinsic names; thus, function names may not
3691begin with this prefix. Intrinsic functions must always be external functions:
3692you cannot define the body of intrinsic functions. Intrinsic functions may
3693only be used in call or invoke instructions: it is illegal to take the address
3694of an intrinsic function. Additionally, because intrinsic functions are part
3695of the LLVM language, it is required if any are added that they be documented
3696here.</p>
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3697
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3698<p>Some intrinsic functions can be overloaded, i.e., the intrinsic represents
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]3699a family of functions that perform the same operation but on different data
3700types. This is most frequent with the integer types. Since LLVM can represent
3701over 8 million different integer types, there is a way to declare an intrinsic
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3702that can be overloaded based on its arguments. Such an intrinsic will have the
3703names of its argument types encoded into its function name, each
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]3704preceded by a period. For example, the <tt>llvm.ctpop</tt> function can take an
3705integer of any width. This leads to a family of functions such as
3706<tt>i32 @llvm.ctpop.i8(i8 %val)</tt> and <tt>i32 @llvm.ctpop.i29(i29 %val)</tt>.
3707</p>
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3708
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]3709
3710<p>To learn how to add an intrinsic function, please see the
3711<a href="ExtendingLLVM.html">Extending LLVM Guide</a>.
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3712</p>
3713
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3714</div>
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3715
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3716<!-- ======================================================================= -->
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3717<div class="doc_subsection">
3718 <a name="int_varargs">Variable Argument Handling Intrinsics</a>
3719</div>
3720
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3721<div class="doc_text">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3722
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3723<p>Variable argument support is defined in LLVM with the <a
Chris Lattner33337472006-01-13 23:26:01 +0000[diff] [blame]3724 href="#i_va_arg"><tt>va_arg</tt></a> instruction and these three
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3725intrinsic functions. These functions are related to the similarly
3726named macros defined in the <tt>&lt;stdarg.h&gt;</tt> header file.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3727
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3728<p>All of these functions operate on arguments that use a
3729target-specific value type "<tt>va_list</tt>". The LLVM assembly
3730language reference manual does not define what this type is, so all
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3731transformations should be prepared to handle these functions regardless of
3732the type used.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3733
Chris Lattner30b868d2006-05-15 17:26:46 +0000[diff] [blame]3734<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]3735instruction and the variable argument handling intrinsic functions are
3736used.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3737
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]3738<div class="doc_code">
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3739<pre>
Anton Korobeynikov640bbe02007-03-21 23:58:04 +0000[diff] [blame]3740define i32 @test(i32 %X, ...) {
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3741 ; Initialize variable argument processing
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3742 %ap = alloca i8*
Chris Lattnerdb0790c2007-01-08 07:55:15 +0000[diff] [blame]3743 %ap2 = bitcast i8** %ap to i8*
Anton Korobeynikov640bbe02007-03-21 23:58:04 +0000[diff] [blame]3744 call void @llvm.va_start(i8* %ap2)
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3745
3746 ; Read a single integer argument
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3747 %tmp = va_arg i8** %ap, i32
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3748
3749 ; Demonstrate usage of llvm.va_copy and llvm.va_end
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3750 %aq = alloca i8*
Chris Lattnerdb0790c2007-01-08 07:55:15 +0000[diff] [blame]3751 %aq2 = bitcast i8** %aq to i8*
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3752 call void @llvm.va_copy(i8* %aq2, i8* %ap2)
Anton Korobeynikov640bbe02007-03-21 23:58:04 +0000[diff] [blame]3753 call void @llvm.va_end(i8* %aq2)
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3754
3755 ; Stop processing of arguments.
Anton Korobeynikov640bbe02007-03-21 23:58:04 +0000[diff] [blame]3756 call void @llvm.va_end(i8* %ap2)
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]3757 ret i32 %tmp
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3758}
Anton Korobeynikov640bbe02007-03-21 23:58:04 +0000[diff] [blame]3759
3760declare void @llvm.va_start(i8*)
3761declare void @llvm.va_copy(i8*, i8*)
3762declare void @llvm.va_end(i8*)
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3763</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3764</div>
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3765
Bill Wendling3716c5d2007-05-29 09:04:49 +0000[diff] [blame]3766</div>
3767
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3768<!-- _______________________________________________________________________ -->
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3769<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]3770 <a name="int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a>
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3771</div>
3772
3773
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3774<div class="doc_text">
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3775<h5>Syntax:</h5>
Chris Lattnerdb0790c2007-01-08 07:55:15 +0000[diff] [blame]3776<pre> declare void %llvm.va_start(i8* &lt;arglist&gt;)<br></pre>
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3777<h5>Overview:</h5>
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3778<P>The '<tt>llvm.va_start</tt>' intrinsic initializes
3779<tt>*&lt;arglist&gt;</tt> for subsequent use by <tt><a
3780href="#i_va_arg">va_arg</a></tt>.</p>
3781
3782<h5>Arguments:</h5>
3783
3784<P>The argument is a pointer to a <tt>va_list</tt> element to initialize.</p>
3785
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3786<h5>Semantics:</h5>
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3787
3788<P>The '<tt>llvm.va_start</tt>' intrinsic works just like the <tt>va_start</tt>
3789macro available in C. In a target-dependent way, it initializes the
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3790<tt>va_list</tt> element to which the argument points, so that the next call to
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3791<tt>va_arg</tt> will produce the first variable argument passed to the function.
3792Unlike the C <tt>va_start</tt> macro, this intrinsic does not need to know the
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3793last argument of the function as the compiler can figure that out.</p>
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3794
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3795</div>
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3796
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3797<!-- _______________________________________________________________________ -->
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3798<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]3799 <a name="int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a>
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3800</div>
3801
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3802<div class="doc_text">
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3803<h5>Syntax:</h5>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]3804<pre> declare void @llvm.va_end(i8* &lt;arglist&gt;)<br></pre>
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3805<h5>Overview:</h5>
Chris Lattnerdb0790c2007-01-08 07:55:15 +0000[diff] [blame]3806
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3807<p>The '<tt>llvm.va_end</tt>' intrinsic destroys <tt>*&lt;arglist&gt;</tt>,
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]3808which has been initialized previously with <tt><a href="#int_va_start">llvm.va_start</a></tt>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3809or <tt><a href="#i_va_copy">llvm.va_copy</a></tt>.</p>
Chris Lattnerdb0790c2007-01-08 07:55:15 +0000[diff] [blame]3810
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3811<h5>Arguments:</h5>
Chris Lattnerdb0790c2007-01-08 07:55:15 +0000[diff] [blame]3812
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3813<p>The argument is a pointer to a <tt>va_list</tt> to destroy.</p>
Chris Lattnerdb0790c2007-01-08 07:55:15 +0000[diff] [blame]3814
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3815<h5>Semantics:</h5>
Chris Lattnerdb0790c2007-01-08 07:55:15 +0000[diff] [blame]3816
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3817<p>The '<tt>llvm.va_end</tt>' intrinsic works just like the <tt>va_end</tt>
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3818macro available in C. In a target-dependent way, it destroys the
3819<tt>va_list</tt> element to which the argument points. Calls to <a
3820href="#int_va_start"><tt>llvm.va_start</tt></a> and <a href="#int_va_copy">
3821<tt>llvm.va_copy</tt></a> must be matched exactly with calls to
3822<tt>llvm.va_end</tt>.</p>
Chris Lattnerdb0790c2007-01-08 07:55:15 +0000[diff] [blame]3823
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3824</div>
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3825
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3826<!-- _______________________________________________________________________ -->
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3827<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]3828 <a name="int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a>
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3829</div>
3830
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3831<div class="doc_text">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3832
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3833<h5>Syntax:</h5>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3834
3835<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]3836 declare void @llvm.va_copy(i8* &lt;destarglist&gt;, i8* &lt;srcarglist&gt;)
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3837</pre>
3838
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3839<h5>Overview:</h5>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3840
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3841<p>The '<tt>llvm.va_copy</tt>' intrinsic copies the current argument position
3842from the source argument list to the destination argument list.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3843
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3844<h5>Arguments:</h5>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3845
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3846<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]3847The 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]3848
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3849
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3850<h5>Semantics:</h5>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3851
Jeff Cohen222a8a42007-04-29 01:07:00 +0000[diff] [blame]3852<p>The '<tt>llvm.va_copy</tt>' intrinsic works just like the <tt>va_copy</tt>
3853macro available in C. In a target-dependent way, it copies the source
3854<tt>va_list</tt> element into the destination <tt>va_list</tt> element. This
3855intrinsic is necessary because the <tt><a href="#int_va_start">
3856llvm.va_start</a></tt> intrinsic may be arbitrarily complex and require, for
3857example, memory allocation.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3858
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3859</div>
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3860
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3861<!-- ======================================================================= -->
3862<div class="doc_subsection">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3863 <a name="int_gc">Accurate Garbage Collection Intrinsics</a>
3864</div>
3865
3866<div class="doc_text">
3867
3868<p>
3869LLVM support for <a href="GarbageCollection.html">Accurate Garbage
3870Collection</a> requires the implementation and generation of these intrinsics.
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]3871These intrinsics allow identification of <a href="#int_gcroot">GC roots on the
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3872stack</a>, as well as garbage collector implementations that require <a
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]3873href="#int_gcread">read</a> and <a href="#int_gcwrite">write</a> barriers.
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3874Front-ends for type-safe garbage collected languages should generate these
3875intrinsics to make use of the LLVM garbage collectors. For more details, see <a
3876href="GarbageCollection.html">Accurate Garbage Collection with LLVM</a>.
3877</p>
3878</div>
3879
3880<!-- _______________________________________________________________________ -->
3881<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]3882 <a name="int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3883</div>
3884
3885<div class="doc_text">
3886
3887<h5>Syntax:</h5>
3888
3889<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]3890 declare void @llvm.gcroot(&lt;ty&gt;** %ptrloc, &lt;ty2&gt;* %metadata)
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3891</pre>
3892
3893<h5>Overview:</h5>
3894
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]3895<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]3896the code generator, and allows some metadata to be associated with it.</p>
3897
3898<h5>Arguments:</h5>
3899
3900<p>The first argument specifies the address of a stack object that contains the
3901root pointer. The second pointer (which must be either a constant or a global
3902value address) contains the meta-data to be associated with the root.</p>
3903
3904<h5>Semantics:</h5>
3905
3906<p>At runtime, a call to this intrinsics stores a null pointer into the "ptrloc"
3907location. At compile-time, the code generator generates information to allow
3908the runtime to find the pointer at GC safe points.
3909</p>
3910
3911</div>
3912
3913
3914<!-- _______________________________________________________________________ -->
3915<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]3916 <a name="int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3917</div>
3918
3919<div class="doc_text">
3920
3921<h5>Syntax:</h5>
3922
3923<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]3924 declare i8 * @llvm.gcread(i8 * %ObjPtr, i8 ** %Ptr)
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3925</pre>
3926
3927<h5>Overview:</h5>
3928
3929<p>The '<tt>llvm.gcread</tt>' intrinsic identifies reads of references from heap
3930locations, allowing garbage collector implementations that require read
3931barriers.</p>
3932
3933<h5>Arguments:</h5>
3934
Chris Lattnerf9228072006-03-14 20:02:51 +0000[diff] [blame]3935<p>The second argument is the address to read from, which should be an address
3936allocated from the garbage collector. The first object is a pointer to the
3937start of the referenced object, if needed by the language runtime (otherwise
3938null).</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3939
3940<h5>Semantics:</h5>
3941
3942<p>The '<tt>llvm.gcread</tt>' intrinsic has the same semantics as a load
3943instruction, but may be replaced with substantially more complex code by the
3944garbage collector runtime, as needed.</p>
3945
3946</div>
3947
3948
3949<!-- _______________________________________________________________________ -->
3950<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]3951 <a name="int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3952</div>
3953
3954<div class="doc_text">
3955
3956<h5>Syntax:</h5>
3957
3958<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]3959 declare void @llvm.gcwrite(i8 * %P1, i8 * %Obj, i8 ** %P2)
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3960</pre>
3961
3962<h5>Overview:</h5>
3963
3964<p>The '<tt>llvm.gcwrite</tt>' intrinsic identifies writes of references to heap
3965locations, allowing garbage collector implementations that require write
3966barriers (such as generational or reference counting collectors).</p>
3967
3968<h5>Arguments:</h5>
3969
Chris Lattnerf9228072006-03-14 20:02:51 +0000[diff] [blame]3970<p>The first argument is the reference to store, the second is the start of the
3971object to store it to, and the third is the address of the field of Obj to
3972store to. If the runtime does not require a pointer to the object, Obj may be
3973null.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3974
3975<h5>Semantics:</h5>
3976
3977<p>The '<tt>llvm.gcwrite</tt>' intrinsic has the same semantics as a store
3978instruction, but may be replaced with substantially more complex code by the
3979garbage collector runtime, as needed.</p>
3980
3981</div>
3982
3983
3984
3985<!-- ======================================================================= -->
3986<div class="doc_subsection">
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3987 <a name="int_codegen">Code Generator Intrinsics</a>
3988</div>
3989
3990<div class="doc_text">
3991<p>
3992These intrinsics are provided by LLVM to expose special features that may only
3993be implemented with code generator support.
3994</p>
3995
3996</div>
3997
3998<!-- _______________________________________________________________________ -->
3999<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4000 <a name="int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a>
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4001</div>
4002
4003<div class="doc_text">
4004
4005<h5>Syntax:</h5>
4006<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4007 declare i8 *@llvm.returnaddress(i32 &lt;level&gt;)
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4008</pre>
4009
4010<h5>Overview:</h5>
4011
4012<p>
Chris Lattnerc1fb4262006-10-15 20:05:59 +0000[diff] [blame]4013The '<tt>llvm.returnaddress</tt>' intrinsic attempts to compute a
4014target-specific value indicating the return address of the current function
4015or one of its callers.
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4016</p>
4017
4018<h5>Arguments:</h5>
4019
4020<p>
4021The argument to this intrinsic indicates which function to return the address
4022for. Zero indicates the calling function, one indicates its caller, etc. The
4023argument is <b>required</b> to be a constant integer value.
4024</p>
4025
4026<h5>Semantics:</h5>
4027
4028<p>
4029The '<tt>llvm.returnaddress</tt>' intrinsic either returns a pointer indicating
4030the return address of the specified call frame, or zero if it cannot be
4031identified. The value returned by this intrinsic is likely to be incorrect or 0
4032for arguments other than zero, so it should only be used for debugging purposes.
4033</p>
4034
4035<p>
4036Note that calling this intrinsic does not prevent function inlining or other
Chris Lattner2e6eb5f2005-03-07 20:30:51 +0000[diff] [blame]4037aggressive transformations, so the value returned may not be that of the obvious
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4038source-language caller.
4039</p>
4040</div>
4041
4042
4043<!-- _______________________________________________________________________ -->
4044<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4045 <a name="int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a>
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4046</div>
4047
4048<div class="doc_text">
4049
4050<h5>Syntax:</h5>
4051<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4052 declare i8 *@llvm.frameaddress(i32 &lt;level&gt;)
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4053</pre>
4054
4055<h5>Overview:</h5>
4056
4057<p>
Chris Lattnerc1fb4262006-10-15 20:05:59 +0000[diff] [blame]4058The '<tt>llvm.frameaddress</tt>' intrinsic attempts to return the
4059target-specific frame pointer value for the specified stack frame.
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4060</p>
4061
4062<h5>Arguments:</h5>
4063
4064<p>
4065The argument to this intrinsic indicates which function to return the frame
4066pointer for. Zero indicates the calling function, one indicates its caller,
4067etc. The argument is <b>required</b> to be a constant integer value.
4068</p>
4069
4070<h5>Semantics:</h5>
4071
4072<p>
4073The '<tt>llvm.frameaddress</tt>' intrinsic either returns a pointer indicating
4074the frame address of the specified call frame, or zero if it cannot be
4075identified. The value returned by this intrinsic is likely to be incorrect or 0
4076for arguments other than zero, so it should only be used for debugging purposes.
4077</p>
4078
4079<p>
4080Note that calling this intrinsic does not prevent function inlining or other
Chris Lattner2e6eb5f2005-03-07 20:30:51 +0000[diff] [blame]4081aggressive transformations, so the value returned may not be that of the obvious
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4082source-language caller.
4083</p>
4084</div>
4085
Chris Lattnerc8a2c222005-02-28 19:24:19 +0000[diff] [blame]4086<!-- _______________________________________________________________________ -->
4087<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4088 <a name="int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a>
Chris Lattner2f0f0012006-01-13 02:03:13 +0000[diff] [blame]4089</div>
4090
4091<div class="doc_text">
4092
4093<h5>Syntax:</h5>
4094<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4095 declare i8 *@llvm.stacksave()
Chris Lattner2f0f0012006-01-13 02:03:13 +0000[diff] [blame]4096</pre>
4097
4098<h5>Overview:</h5>
4099
4100<p>
4101The '<tt>llvm.stacksave</tt>' intrinsic is used to remember the current state of
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4102the function stack, for use with <a href="#int_stackrestore">
Chris Lattner2f0f0012006-01-13 02:03:13 +0000[diff] [blame]4103<tt>llvm.stackrestore</tt></a>. This is useful for implementing language
4104features like scoped automatic variable sized arrays in C99.
4105</p>
4106
4107<h5>Semantics:</h5>
4108
4109<p>
4110This intrinsic returns a opaque pointer value that can be passed to <a
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4111href="#int_stackrestore"><tt>llvm.stackrestore</tt></a>. When an
Chris Lattner2f0f0012006-01-13 02:03:13 +0000[diff] [blame]4112<tt>llvm.stackrestore</tt> intrinsic is executed with a value saved from
4113<tt>llvm.stacksave</tt>, it effectively restores the state of the stack to the
4114state it was in when the <tt>llvm.stacksave</tt> intrinsic executed. In
4115practice, this pops any <a href="#i_alloca">alloca</a> blocks from the stack
4116that were allocated after the <tt>llvm.stacksave</tt> was executed.
4117</p>
4118
4119</div>
4120
4121<!-- _______________________________________________________________________ -->
4122<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4123 <a name="int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a>
Chris Lattner2f0f0012006-01-13 02:03:13 +0000[diff] [blame]4124</div>
4125
4126<div class="doc_text">
4127
4128<h5>Syntax:</h5>
4129<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4130 declare void @llvm.stackrestore(i8 * %ptr)
Chris Lattner2f0f0012006-01-13 02:03:13 +0000[diff] [blame]4131</pre>
4132
4133<h5>Overview:</h5>
4134
4135<p>
4136The '<tt>llvm.stackrestore</tt>' intrinsic is used to restore the state of
4137the function stack to the state it was in when the corresponding <a
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4138href="#int_stacksave"><tt>llvm.stacksave</tt></a> intrinsic executed. This is
Chris Lattner2f0f0012006-01-13 02:03:13 +0000[diff] [blame]4139useful for implementing language features like scoped automatic variable sized
4140arrays in C99.
4141</p>
4142
4143<h5>Semantics:</h5>
4144
4145<p>
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4146See the description for <a href="#int_stacksave"><tt>llvm.stacksave</tt></a>.
Chris Lattner2f0f0012006-01-13 02:03:13 +0000[diff] [blame]4147</p>
4148
4149</div>
4150
4151
4152<!-- _______________________________________________________________________ -->
4153<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4154 <a name="int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a>
Chris Lattnerc8a2c222005-02-28 19:24:19 +0000[diff] [blame]4155</div>
4156
4157<div class="doc_text">
4158
4159<h5>Syntax:</h5>
4160<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4161 declare void @llvm.prefetch(i8 * &lt;address&gt;,
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]4162 i32 &lt;rw&gt;, i32 &lt;locality&gt;)
Chris Lattnerc8a2c222005-02-28 19:24:19 +0000[diff] [blame]4163</pre>
4164
4165<h5>Overview:</h5>
4166
4167
4168<p>
4169The '<tt>llvm.prefetch</tt>' intrinsic is a hint to the code generator to insert
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]4170a prefetch instruction if supported; otherwise, it is a noop. Prefetches have
4171no
4172effect on the behavior of the program but can change its performance
Chris Lattnerff851072005-02-28 19:47:14 +0000[diff] [blame]4173characteristics.
Chris Lattnerc8a2c222005-02-28 19:24:19 +0000[diff] [blame]4174</p>
4175
4176<h5>Arguments:</h5>
4177
4178<p>
4179<tt>address</tt> is the address to be prefetched, <tt>rw</tt> is the specifier
4180determining if the fetch should be for a read (0) or write (1), and
4181<tt>locality</tt> is a temporal locality specifier ranging from (0) - no
Chris Lattnerd3e641c2005-03-07 20:31:38 +0000[diff] [blame]4182locality, to (3) - extremely local keep in cache. The <tt>rw</tt> and
Chris Lattnerc8a2c222005-02-28 19:24:19 +0000[diff] [blame]4183<tt>locality</tt> arguments must be constant integers.
4184</p>
4185
4186<h5>Semantics:</h5>
4187
4188<p>
4189This intrinsic does not modify the behavior of the program. In particular,
4190prefetches cannot trap and do not produce a value. On targets that support this
4191intrinsic, the prefetch can provide hints to the processor cache for better
4192performance.
4193</p>
4194
4195</div>
4196
Andrew Lenharthb4427912005-03-28 20:05:49 +0000[diff] [blame]4197<!-- _______________________________________________________________________ -->
4198<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4199 <a name="int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a>
Andrew Lenharthb4427912005-03-28 20:05:49 +0000[diff] [blame]4200</div>
4201
4202<div class="doc_text">
4203
4204<h5>Syntax:</h5>
4205<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4206 declare void @llvm.pcmarker( i32 &lt;id&gt; )
Andrew Lenharthb4427912005-03-28 20:05:49 +0000[diff] [blame]4207</pre>
4208
4209<h5>Overview:</h5>
4210
4211
4212<p>
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]4213The '<tt>llvm.pcmarker</tt>' intrinsic is a method to export a Program Counter
4214(PC) in a region of
Andrew Lenharthb4427912005-03-28 20:05:49 +0000[diff] [blame]4215code to simulators and other tools. The method is target specific, but it is
4216expected that the marker will use exported symbols to transmit the PC of the marker.
Jeff Cohendc6bfea2005-11-11 02:15:27 +0000[diff] [blame]4217The marker makes no guarantees that it will remain with any specific instruction
Chris Lattnere64d41d2005-11-15 06:07:55 +0000[diff] [blame]4218after optimizations. It is possible that the presence of a marker will inhibit
Chris Lattnerb40261e2006-03-24 07:16:10 +0000[diff] [blame]4219optimizations. The intended use is to be inserted after optimizations to allow
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]4220correlations of simulation runs.
Andrew Lenharthb4427912005-03-28 20:05:49 +0000[diff] [blame]4221</p>
4222
4223<h5>Arguments:</h5>
4224
4225<p>
4226<tt>id</tt> is a numerical id identifying the marker.
4227</p>
4228
4229<h5>Semantics:</h5>
4230
4231<p>
4232This intrinsic does not modify the behavior of the program. Backends that do not
4233support this intrinisic may ignore it.
4234</p>
4235
4236</div>
4237
Andrew Lenharth01aa5632005-11-11 16:47:30 +0000[diff] [blame]4238<!-- _______________________________________________________________________ -->
4239<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4240 <a name="int_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a>
Andrew Lenharth01aa5632005-11-11 16:47:30 +0000[diff] [blame]4241</div>
4242
4243<div class="doc_text">
4244
4245<h5>Syntax:</h5>
4246<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4247 declare i64 @llvm.readcyclecounter( )
Andrew Lenharth01aa5632005-11-11 16:47:30 +0000[diff] [blame]4248</pre>
4249
4250<h5>Overview:</h5>
4251
4252
4253<p>
4254The '<tt>llvm.readcyclecounter</tt>' intrinsic provides access to the cycle
4255counter register (or similar low latency, high accuracy clocks) on those targets
4256that support it. On X86, it should map to RDTSC. On Alpha, it should map to RPCC.
4257As the backing counters overflow quickly (on the order of 9 seconds on alpha), this
4258should only be used for small timings.
4259</p>
4260
4261<h5>Semantics:</h5>
4262
4263<p>
4264When directly supported, reading the cycle counter should not modify any memory.
4265Implementations are allowed to either return a application specific value or a
4266system wide value. On backends without support, this is lowered to a constant 0.
4267</p>
4268
4269</div>
4270
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4271<!-- ======================================================================= -->
4272<div class="doc_subsection">
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]4273 <a name="int_libc">Standard C Library Intrinsics</a>
4274</div>
4275
4276<div class="doc_text">
4277<p>
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4278LLVM provides intrinsics for a few important standard C library functions.
4279These intrinsics allow source-language front-ends to pass information about the
4280alignment of the pointer arguments to the code generator, providing opportunity
4281for more efficient code generation.
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]4282</p>
4283
4284</div>
4285
4286<!-- _______________________________________________________________________ -->
4287<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4288 <a name="int_memcpy">'<tt>llvm.memcpy</tt>' Intrinsic</a>
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]4289</div>
4290
4291<div class="doc_text">
4292
4293<h5>Syntax:</h5>
4294<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4295 declare void @llvm.memcpy.i32(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]4296 i32 &lt;len&gt;, i32 &lt;align&gt;)
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4297 declare void @llvm.memcpy.i64(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]4298 i64 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]4299</pre>
4300
4301<h5>Overview:</h5>
4302
4303<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4304The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]4305location to the destination location.
4306</p>
4307
4308<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4309Note that, unlike the standard libc function, the <tt>llvm.memcpy.*</tt>
4310intrinsics do not return a value, and takes an extra alignment argument.
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]4311</p>
4312
4313<h5>Arguments:</h5>
4314
4315<p>
4316The first argument is a pointer to the destination, the second is a pointer to
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4317the source. The third argument is an integer argument
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]4318specifying the number of bytes to copy, and the fourth argument is the alignment
4319of the source and destination locations.
4320</p>
4321
Chris Lattner4c67c482004-02-12 21:18:15 +0000[diff] [blame]4322<p>
4323If 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]4324the caller guarantees that both the source and destination pointers are aligned
4325to that boundary.
Chris Lattner4c67c482004-02-12 21:18:15 +0000[diff] [blame]4326</p>
4327
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]4328<h5>Semantics:</h5>
4329
4330<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4331The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]4332location to the destination location, which are not allowed to overlap. It
4333copies "len" bytes of memory over. If the argument is known to be aligned to
4334some boundary, this can be specified as the fourth argument, otherwise it should
4335be set to 0 or 1.
4336</p>
4337</div>
4338
4339
Chris Lattnerf30152e2004-02-12 18:10:10 +0000[diff] [blame]4340<!-- _______________________________________________________________________ -->
4341<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4342 <a name="int_memmove">'<tt>llvm.memmove</tt>' Intrinsic</a>
Chris Lattnerf30152e2004-02-12 18:10:10 +0000[diff] [blame]4343</div>
4344
4345<div class="doc_text">
4346
4347<h5>Syntax:</h5>
4348<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4349 declare void @llvm.memmove.i32(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]4350 i32 &lt;len&gt;, i32 &lt;align&gt;)
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4351 declare void @llvm.memmove.i64(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]4352 i64 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattnerf30152e2004-02-12 18:10:10 +0000[diff] [blame]4353</pre>
4354
4355<h5>Overview:</h5>
4356
4357<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4358The '<tt>llvm.memmove.*</tt>' intrinsics move a block of memory from the source
4359location to the destination location. It is similar to the
4360'<tt>llvm.memcmp</tt>' intrinsic but allows the two memory locations to overlap.
Chris Lattnerf30152e2004-02-12 18:10:10 +0000[diff] [blame]4361</p>
4362
4363<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4364Note that, unlike the standard libc function, the <tt>llvm.memmove.*</tt>
4365intrinsics do not return a value, and takes an extra alignment argument.
Chris Lattnerf30152e2004-02-12 18:10:10 +0000[diff] [blame]4366</p>
4367
4368<h5>Arguments:</h5>
4369
4370<p>
4371The first argument is a pointer to the destination, the second is a pointer to
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4372the source. The third argument is an integer argument
Chris Lattnerf30152e2004-02-12 18:10:10 +0000[diff] [blame]4373specifying the number of bytes to copy, and the fourth argument is the alignment
4374of the source and destination locations.
4375</p>
4376
Chris Lattner4c67c482004-02-12 21:18:15 +0000[diff] [blame]4377<p>
4378If 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]4379the caller guarantees that the source and destination pointers are aligned to
4380that boundary.
Chris Lattner4c67c482004-02-12 21:18:15 +0000[diff] [blame]4381</p>
4382
Chris Lattnerf30152e2004-02-12 18:10:10 +0000[diff] [blame]4383<h5>Semantics:</h5>
4384
4385<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4386The '<tt>llvm.memmove.*</tt>' intrinsics copy a block of memory from the source
Chris Lattnerf30152e2004-02-12 18:10:10 +0000[diff] [blame]4387location to the destination location, which may overlap. It
4388copies "len" bytes of memory over. If the argument is known to be aligned to
4389some boundary, this can be specified as the fourth argument, otherwise it should
4390be set to 0 or 1.
4391</p>
4392</div>
4393
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]4394
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4395<!-- _______________________________________________________________________ -->
4396<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4397 <a name="int_memset">'<tt>llvm.memset.*</tt>' Intrinsics</a>
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4398</div>
4399
4400<div class="doc_text">
4401
4402<h5>Syntax:</h5>
4403<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4404 declare void @llvm.memset.i32(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]4405 i32 &lt;len&gt;, i32 &lt;align&gt;)
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4406 declare void @llvm.memset.i64(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]4407 i64 &lt;len&gt;, i32 &lt;align&gt;)
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4408</pre>
4409
4410<h5>Overview:</h5>
4411
4412<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4413The '<tt>llvm.memset.*</tt>' intrinsics fill a block of memory with a particular
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4414byte value.
4415</p>
4416
4417<p>
4418Note that, unlike the standard libc function, the <tt>llvm.memset</tt> intrinsic
4419does not return a value, and takes an extra alignment argument.
4420</p>
4421
4422<h5>Arguments:</h5>
4423
4424<p>
4425The first argument is a pointer to the destination to fill, the second is the
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4426byte value to fill it with, the third argument is an integer
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4427argument specifying the number of bytes to fill, and the fourth argument is the
4428known alignment of destination location.
4429</p>
4430
4431<p>
4432If 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]4433the caller guarantees that the destination pointer is aligned to that boundary.
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4434</p>
4435
4436<h5>Semantics:</h5>
4437
4438<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]4439The '<tt>llvm.memset.*</tt>' intrinsics fill "len" bytes of memory starting at
4440the
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]4441destination location. If the argument is known to be aligned to some boundary,
4442this can be specified as the fourth argument, otherwise it should be set to 0 or
44431.
4444</p>
4445</div>
4446
4447
Chris Lattner3b4f4372004-06-11 02:28:03 +0000[diff] [blame]4448<!-- _______________________________________________________________________ -->
4449<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4450 <a name="int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a>
Chris Lattner8a8f2e52005-07-21 01:29:16 +0000[diff] [blame]4451</div>
4452
4453<div class="doc_text">
4454
4455<h5>Syntax:</h5>
4456<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4457 declare float @llvm.sqrt.f32(float %Val)
4458 declare double @llvm.sqrt.f64(double %Val)
Chris Lattner8a8f2e52005-07-21 01:29:16 +0000[diff] [blame]4459</pre>
4460
4461<h5>Overview:</h5>
4462
4463<p>
Reid Spencerb4f9a6f2006-01-16 21:12:35 +0000[diff] [blame]4464The '<tt>llvm.sqrt</tt>' intrinsics return the sqrt of the specified operand,
Chris Lattner8a8f2e52005-07-21 01:29:16 +0000[diff] [blame]4465returning the same value as the libm '<tt>sqrt</tt>' function would. Unlike
4466<tt>sqrt</tt> in libm, however, <tt>llvm.sqrt</tt> has undefined behavior for
4467negative numbers (which allows for better optimization).
4468</p>
4469
4470<h5>Arguments:</h5>
4471
4472<p>
4473The argument and return value are floating point numbers of the same type.
4474</p>
4475
4476<h5>Semantics:</h5>
4477
4478<p>
4479This function returns the sqrt of the specified operand if it is a positive
4480floating point number.
4481</p>
4482</div>
4483
Chris Lattner33b73f92006-09-08 06:34:02 +0000[diff] [blame]4484<!-- _______________________________________________________________________ -->
4485<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4486 <a name="int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a>
Chris Lattner33b73f92006-09-08 06:34:02 +0000[diff] [blame]4487</div>
4488
4489<div class="doc_text">
4490
4491<h5>Syntax:</h5>
4492<pre>
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4493 declare float @llvm.powi.f32(float %Val, i32 %power)
4494 declare double @llvm.powi.f64(double %Val, i32 %power)
Chris Lattner33b73f92006-09-08 06:34:02 +0000[diff] [blame]4495</pre>
4496
4497<h5>Overview:</h5>
4498
4499<p>
4500The '<tt>llvm.powi.*</tt>' intrinsics return the first operand raised to the
4501specified (positive or negative) power. The order of evaluation of
4502multiplications is not defined.
4503</p>
4504
4505<h5>Arguments:</h5>
4506
4507<p>
4508The second argument is an integer power, and the first is a value to raise to
4509that power.
4510</p>
4511
4512<h5>Semantics:</h5>
4513
4514<p>
4515This function returns the first value raised to the second power with an
4516unspecified sequence of rounding operations.</p>
4517</div>
4518
4519
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4520<!-- ======================================================================= -->
4521<div class="doc_subsection">
Nate Begeman0f223bb2006-01-13 23:26:38 +0000[diff] [blame]4522 <a name="int_manip">Bit Manipulation Intrinsics</a>
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4523</div>
4524
4525<div class="doc_text">
4526<p>
Nate Begeman0f223bb2006-01-13 23:26:38 +0000[diff] [blame]4527LLVM provides intrinsics for a few important bit manipulation operations.
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4528These allow efficient code generation for some algorithms.
4529</p>
4530
4531</div>
4532
4533<!-- _______________________________________________________________________ -->
4534<div class="doc_subsubsection">
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4535 <a name="int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a>
Nate Begeman0f223bb2006-01-13 23:26:38 +0000[diff] [blame]4536</div>
4537
4538<div class="doc_text">
4539
4540<h5>Syntax:</h5>
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]4541<p>This is an overloaded intrinsic function. You can use bswap on any integer
4542type that is an even number of bytes (i.e. BitWidth % 16 == 0). Note the suffix
4543that includes the type for the result and the operand.
Nate Begeman0f223bb2006-01-13 23:26:38 +0000[diff] [blame]4544<pre>
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]4545 declare i16 @llvm.bswap.i16.i16(i16 &lt;id&gt;)
4546 declare i32 @llvm.bswap.i32.i32(i32 &lt;id&gt;)
Reid Spencer403a1c42007-04-02 00:19:52 +0000[diff] [blame]4547 declare i64 @llvm.bswap.i64.i64(i64 &lt;id&gt;)
Nate Begeman0f223bb2006-01-13 23:26:38 +0000[diff] [blame]4548</pre>
4549
4550<h5>Overview:</h5>
4551
4552<p>
Reid Spencerf361c4f2007-04-02 02:25:19 +0000[diff] [blame]4553The '<tt>llvm.bswap</tt>' family of intrinsics is used to byte swap integer
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]4554values with an even number of bytes (positive multiple of 16 bits). These are
4555useful for performing operations on data that is not in the target's native
4556byte order.
Nate Begeman0f223bb2006-01-13 23:26:38 +0000[diff] [blame]4557</p>
4558
4559<h5>Semantics:</h5>
4560
4561<p>
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]4562The <tt>llvm.bswap.16.i16</tt> intrinsic returns an i16 value that has the high
Reid Spencerb5ebf3d2006-12-31 07:07:53 +0000[diff] [blame]4563and low byte of the input i16 swapped. Similarly, the <tt>llvm.bswap.i32</tt>
4564intrinsic returns an i32 value that has the four bytes of the input i32
4565swapped, so that if the input bytes are numbered 0, 1, 2, 3 then the returned
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]4566i32 will have its bytes in 3, 2, 1, 0 order. The <tt>llvm.bswap.i48.i48</tt>,
4567<tt>llvm.bswap.i64.i64</tt> and other intrinsics extend this concept to
4568additional even-byte lengths (6 bytes, 8 bytes and more, respectively).
Nate Begeman0f223bb2006-01-13 23:26:38 +0000[diff] [blame]4569</p>
4570
4571</div>
4572
4573<!-- _______________________________________________________________________ -->
4574<div class="doc_subsubsection">
Reid Spencerb4f9a6f2006-01-16 21:12:35 +0000[diff] [blame]4575 <a name="int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic</a>
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4576</div>
4577
4578<div class="doc_text">
4579
4580<h5>Syntax:</h5>
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]4581<p>This is an overloaded intrinsic. You can use llvm.ctpop on any integer bit
4582width. Not all targets support all bit widths however.
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4583<pre>
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]4584 declare i32 @llvm.ctpop.i8 (i8 &lt;src&gt;)
4585 declare i32 @llvm.ctpop.i16(i16 &lt;src&gt;)
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4586 declare i32 @llvm.ctpop.i32(i32 &lt;src&gt;)
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]4587 declare i32 @llvm.ctpop.i64(i64 &lt;src&gt;)
4588 declare i32 @llvm.ctpop.i256(i256 &lt;src&gt;)
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4589</pre>
4590
4591<h5>Overview:</h5>
4592
4593<p>
Chris Lattner069b5bd2006-01-16 22:38:59 +0000[diff] [blame]4594The '<tt>llvm.ctpop</tt>' family of intrinsics counts the number of bits set in a
4595value.
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4596</p>
4597
4598<h5>Arguments:</h5>
4599
4600<p>
Chris Lattner573f64e2005-05-07 01:46:40 +0000[diff] [blame]4601The only argument is the value to be counted. The argument may be of any
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]4602integer type. The return type must match the argument type.
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4603</p>
4604
4605<h5>Semantics:</h5>
4606
4607<p>
4608The '<tt>llvm.ctpop</tt>' intrinsic counts the 1's in a variable.
4609</p>
4610</div>
4611
4612<!-- _______________________________________________________________________ -->
4613<div class="doc_subsubsection">
Chris Lattnerb748c672006-01-16 22:34:14 +0000[diff] [blame]4614 <a name="int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic</a>
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4615</div>
4616
4617<div class="doc_text">
4618
4619<h5>Syntax:</h5>
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]4620<p>This is an overloaded intrinsic. You can use <tt>llvm.ctlz</tt> on any
4621integer bit width. Not all targets support all bit widths however.
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4622<pre>
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]4623 declare i32 @llvm.ctlz.i8 (i8 &lt;src&gt;)
4624 declare i32 @llvm.ctlz.i16(i16 &lt;src&gt;)
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4625 declare i32 @llvm.ctlz.i32(i32 &lt;src&gt;)
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]4626 declare i32 @llvm.ctlz.i64(i64 &lt;src&gt;)
4627 declare i32 @llvm.ctlz.i256(i256 &lt;src&gt;)
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4628</pre>
4629
4630<h5>Overview:</h5>
4631
4632<p>
Reid Spencerb4f9a6f2006-01-16 21:12:35 +0000[diff] [blame]4633The '<tt>llvm.ctlz</tt>' family of intrinsic functions counts the number of
4634leading zeros in a variable.
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4635</p>
4636
4637<h5>Arguments:</h5>
4638
4639<p>
Chris Lattner573f64e2005-05-07 01:46:40 +0000[diff] [blame]4640The only argument is the value to be counted. The argument may be of any
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]4641integer type. The return type must match the argument type.
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4642</p>
4643
4644<h5>Semantics:</h5>
4645
4646<p>
Chris Lattnerefa20fa2005-05-15 19:39:26 +0000[diff] [blame]4647The '<tt>llvm.ctlz</tt>' intrinsic counts the leading (most significant) zeros
4648in 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]4649of src. For example, <tt>llvm.ctlz(i32 2) = 30</tt>.
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]4650</p>
4651</div>
Chris Lattner3b4f4372004-06-11 02:28:03 +0000[diff] [blame]4652
4653
Chris Lattnerefa20fa2005-05-15 19:39:26 +0000[diff] [blame]4654
4655<!-- _______________________________________________________________________ -->
4656<div class="doc_subsubsection">
Chris Lattnerb748c672006-01-16 22:34:14 +0000[diff] [blame]4657 <a name="int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic</a>
Chris Lattnerefa20fa2005-05-15 19:39:26 +0000[diff] [blame]4658</div>
4659
4660<div class="doc_text">
4661
4662<h5>Syntax:</h5>
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]4663<p>This is an overloaded intrinsic. You can use <tt>llvm.cttz</tt> on any
4664integer bit width. Not all targets support all bit widths however.
Chris Lattnerefa20fa2005-05-15 19:39:26 +0000[diff] [blame]4665<pre>
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]4666 declare i32 @llvm.cttz.i8 (i8 &lt;src&gt;)
4667 declare i32 @llvm.cttz.i16(i16 &lt;src&gt;)
Anton Korobeynikovbe9c93c2007-03-22 00:02:17 +0000[diff] [blame]4668 declare i32 @llvm.cttz.i32(i32 &lt;src&gt;)
Reid Spencer4eefaab2007-04-01 08:04:23 +0000[diff] [blame]4669 declare i32 @llvm.cttz.i64(i64 &lt;src&gt;)
4670 declare i32 @llvm.cttz.i256(i256 &lt;src&gt;)
Chris Lattnerefa20fa2005-05-15 19:39:26 +0000[diff] [blame]4671</pre>
4672
4673<h5>Overview:</h5>
4674
4675<p>
Reid Spencerb4f9a6f2006-01-16 21:12:35 +0000[diff] [blame]4676The '<tt>llvm.cttz</tt>' family of intrinsic functions counts the number of
4677trailing zeros.
Chris Lattnerefa20fa2005-05-15 19:39:26 +0000[diff] [blame]4678</p>
4679
4680<h5>Arguments:</h5>
4681
4682<p>
4683The only argument is the value to be counted. The argument may be of any
Reid Spencer3e628eb92007-01-04 16:43:23 +0000[diff] [blame]4684integer type. The return type must match the argument type.
Chris Lattnerefa20fa2005-05-15 19:39:26 +0000[diff] [blame]4685</p>
4686
4687<h5>Semantics:</h5>
4688
4689<p>
4690The '<tt>llvm.cttz</tt>' intrinsic counts the trailing (least significant) zeros
4691in a variable. If the src == 0 then the result is the size in bits of the type
4692of src. For example, <tt>llvm.cttz(2) = 1</tt>.
4693</p>
4694</div>
4695
Reid Spencer8a5799f2007-04-01 08:27:01 +0000[diff] [blame]4696<!-- _______________________________________________________________________ -->
4697<div class="doc_subsubsection">
Reid Spencerea2945e2007-04-10 02:51:31 +0000[diff] [blame]4698 <a name="int_part_select">'<tt>llvm.part.select.*</tt>' Intrinsic</a>
Reid Spencer8bc7d952007-04-01 19:00:37 +0000[diff] [blame]4699</div>
4700
4701<div class="doc_text">
4702
4703<h5>Syntax:</h5>
Reid Spencerea2945e2007-04-10 02:51:31 +0000[diff] [blame]4704<p>This is an overloaded intrinsic. You can use <tt>llvm.part.select</tt>
Reid Spencer8bc7d952007-04-01 19:00:37 +0000[diff] [blame]4705on any integer bit width.
4706<pre>
Reid Spencerea2945e2007-04-10 02:51:31 +0000[diff] [blame]4707 declare i17 @llvm.part.select.i17.i17 (i17 %val, i32 %loBit, i32 %hiBit)
4708 declare i29 @llvm.part.select.i29.i29 (i29 %val, i32 %loBit, i32 %hiBit)
Reid Spencer8bc7d952007-04-01 19:00:37 +0000[diff] [blame]4709</pre>
4710
4711<h5>Overview:</h5>
Reid Spencerea2945e2007-04-10 02:51:31 +0000[diff] [blame]4712<p>The '<tt>llvm.part.select</tt>' family of intrinsic functions selects a
Reid Spencer8bc7d952007-04-01 19:00:37 +0000[diff] [blame]4713range of bits from an integer value and returns them in the same bit width as
4714the original value.</p>
4715
4716<h5>Arguments:</h5>
4717<p>The first argument, <tt>%val</tt> and the result may be integer types of
4718any bit width but they must have the same bit width. The second and third
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4719arguments must be <tt>i32</tt> type since they specify only a bit index.</p>
Reid Spencer8bc7d952007-04-01 19:00:37 +0000[diff] [blame]4720
4721<h5>Semantics:</h5>
Reid Spencerea2945e2007-04-10 02:51:31 +0000[diff] [blame]4722<p>The operation of the '<tt>llvm.part.select</tt>' intrinsic has two modes
Reid Spencer96a5f022007-04-04 02:42:35 +0000[diff] [blame]4723of operation: forwards and reverse. If <tt>%loBit</tt> is greater than
4724<tt>%hiBits</tt> then the intrinsic operates in reverse mode. Otherwise it
4725operates in forward mode.</p>
4726<p>In forward mode, this intrinsic is the equivalent of shifting <tt>%val</tt>
4727right by <tt>%loBit</tt> bits and then ANDing it with a mask with
Reid Spencer8bc7d952007-04-01 19:00:37 +0000[diff] [blame]4728only the <tt>%hiBit - %loBit</tt> bits set, as follows:</p>
4729<ol>
4730 <li>The <tt>%val</tt> is shifted right (LSHR) by the number of bits specified
4731 by <tt>%loBits</tt>. This normalizes the value to the low order bits.</li>
4732 <li>The <tt>%loBits</tt> value is subtracted from the <tt>%hiBits</tt> value
4733 to determine the number of bits to retain.</li>
4734 <li>A mask of the retained bits is created by shifting a -1 value.</li>
4735 <li>The mask is ANDed with <tt>%val</tt> to produce the result.
4736</ol>
Reid Spencer70845c02007-05-14 16:14:57 +0000[diff] [blame]4737<p>In reverse mode, a similar computation is made except that the bits are
4738returned in the reverse order. So, for example, if <tt>X</tt> has the value
4739<tt>i16 0x0ACF (101011001111)</tt> and we apply
4740<tt>part.select(i16 X, 8, 3)</tt> to it, we get back the value
4741<tt>i16 0x0026 (000000100110)</tt>.</p>
Reid Spencer8bc7d952007-04-01 19:00:37 +0000[diff] [blame]4742</div>
4743
Reid Spencer5bf54c82007-04-11 23:23:49 +0000[diff] [blame]4744<div class="doc_subsubsection">
4745 <a name="int_part_set">'<tt>llvm.part.set.*</tt>' Intrinsic</a>
4746</div>
4747
4748<div class="doc_text">
4749
4750<h5>Syntax:</h5>
4751<p>This is an overloaded intrinsic. You can use <tt>llvm.part.set</tt>
4752on any integer bit width.
4753<pre>
4754 declare i17 @llvm.part.set.i17.i17.i9 (i17 %val, i9 %repl, i32 %lo, i32 %hi)
4755 declare i29 @llvm.part.set.i29.i29.i9 (i29 %val, i9 %repl, i32 %lo, i32 %hi)
4756</pre>
4757
4758<h5>Overview:</h5>
4759<p>The '<tt>llvm.part.set</tt>' family of intrinsic functions replaces a range
4760of bits in an integer value with another integer value. It returns the integer
4761with the replaced bits.</p>
4762
4763<h5>Arguments:</h5>
4764<p>The first argument, <tt>%val</tt> and the result may be integer types of
4765any bit width but they must have the same bit width. <tt>%val</tt> is the value
4766whose bits will be replaced. The second argument, <tt>%repl</tt> may be an
4767integer of any bit width. The third and fourth arguments must be <tt>i32</tt>
4768type since they specify only a bit index.</p>
4769
4770<h5>Semantics:</h5>
4771<p>The operation of the '<tt>llvm.part.set</tt>' intrinsic has two modes
4772of operation: forwards and reverse. If <tt>%lo</tt> is greater than
4773<tt>%hi</tt> then the intrinsic operates in reverse mode. Otherwise it
4774operates in forward mode.</p>
4775<p>For both modes, the <tt>%repl</tt> value is prepared for use by either
4776truncating it down to the size of the replacement area or zero extending it
4777up to that size.</p>
4778<p>In forward mode, the bits between <tt>%lo</tt> and <tt>%hi</tt> (inclusive)
4779are replaced with corresponding bits from <tt>%repl</tt>. That is the 0th bit
4780in <tt>%repl</tt> replaces the <tt>%lo</tt>th bit in <tt>%val</tt> and etc. up
4781to the <tt>%hi</tt>th bit.
Reid Spencer146281c2007-05-14 16:50:20 +0000[diff] [blame]4782<p>In reverse mode, a similar computation is made except that the bits are
4783reversed. That is, the <tt>0</tt>th bit in <tt>%repl</tt> replaces the
4784<tt>%hi</tt> bit in <tt>%val</tt> and etc. down to the <tt>%lo</tt>th bit.
Reid Spencer5bf54c82007-04-11 23:23:49 +0000[diff] [blame]4785<h5>Examples:</h5>
4786<pre>
Reid Spencerc70afc32007-04-12 01:03:03 +0000[diff] [blame]4787 llvm.part.set(0xFFFF, 0, 4, 7) -&gt; 0xFF0F
Reid Spencer146281c2007-05-14 16:50:20 +0000[diff] [blame]4788 llvm.part.set(0xFFFF, 0, 7, 4) -&gt; 0xFF0F
4789 llvm.part.set(0xFFFF, 1, 7, 4) -&gt; 0xFF8F
4790 llvm.part.set(0xFFFF, F, 8, 3) -&gt; 0xFFE7
Reid Spencerc70afc32007-04-12 01:03:03 +0000[diff] [blame]4791 llvm.part.set(0xFFFF, 0, 3, 8) -&gt; 0xFE07
Reid Spencer7972c472007-04-11 23:49:50 +0000[diff] [blame]4792</pre>
Reid Spencer5bf54c82007-04-11 23:23:49 +0000[diff] [blame]4793</div>
4794
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]4795<!-- ======================================================================= -->
4796<div class="doc_subsection">
4797 <a name="int_debugger">Debugger Intrinsics</a>
4798</div>
4799
4800<div class="doc_text">
4801<p>
4802The LLVM debugger intrinsics (which all start with <tt>llvm.dbg.</tt> prefix),
4803are described in the <a
4804href="SourceLevelDebugging.html#format_common_intrinsics">LLVM Source Level
4805Debugging</a> document.
4806</p>
4807</div>
4808
4809
Jim Laskey2211f492007-03-14 19:31:19 +0000[diff] [blame]4810<!-- ======================================================================= -->
4811<div class="doc_subsection">
4812 <a name="int_eh">Exception Handling Intrinsics</a>
4813</div>
4814
4815<div class="doc_text">
4816<p> The LLVM exception handling intrinsics (which all start with
4817<tt>llvm.eh.</tt> prefix), are described in the <a
4818href="ExceptionHandling.html#format_common_intrinsics">LLVM Exception
4819Handling</a> document. </p>
4820</div>
4821
4822
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]4823<!-- *********************************************************************** -->
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]4824<hr>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]4825<address>
4826 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
4827 src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
4828 <a href="http://validator.w3.org/check/referer"><img
4829 src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!" /></a>
4830
4831 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
Reid Spencerca058542006-03-14 05:39:39 +0000[diff] [blame]4832 <a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]4833 Last modified: $Date$
4834</address>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]4835</body>
4836</html>