Gitiles
Code Review Sign In
gerrit-public.fairphone.software / toolchain / llvm-project / 7e80b0b31ef448877c60de9b3d518a2a79c0a8a7 / . / llvm / docs / LangRef.html
blob: 0ed371a749bede110407e6fece037fbeed8dbe2e [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>
27 <li><a href="#moduleasm">Module-Level Inline Assembly</a></li>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]28 </ol>
29 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]30 <li><a href="#typesystem">Type System</a>
31 <ol>
Robert Bocchino820bc75b2006-02-17 21:18:08 +0000[diff] [blame]32 <li><a href="#t_primitive">Primitive Types</a>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]33 <ol>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]34 <li><a href="#t_classifications">Type Classifications</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]35 </ol>
36 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]37 <li><a href="#t_derived">Derived Types</a>
38 <ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]39 <li><a href="#t_array">Array Type</a></li>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]40 <li><a href="#t_function">Function Type</a></li>
41 <li><a href="#t_pointer">Pointer Type</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]42 <li><a href="#t_struct">Structure Type</a></li>
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]43 <li><a href="#t_packed">Packed Type</a></li>
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]44 <li><a href="#t_opaque">Opaque Type</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]45 </ol>
46 </li>
47 </ol>
48 </li>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]49 <li><a href="#constants">Constants</a>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]50 <ol>
51 <li><a href="#simpleconstants">Simple Constants</a>
52 <li><a href="#aggregateconstants">Aggregate Constants</a>
53 <li><a href="#globalconstants">Global Variable and Function Addresses</a>
54 <li><a href="#undefvalues">Undefined Values</a>
55 <li><a href="#constantexprs">Constant Expressions</a>
56 </ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]57 </li>
Chris Lattner98f013c2006-01-25 23:47:57 +0000[diff] [blame]58 <li><a href="#othervalues">Other Values</a>
59 <ol>
60 <li><a href="#inlineasm">Inline Assembler Expressions</a>
61 </ol>
62 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]63 <li><a href="#instref">Instruction Reference</a>
64 <ol>
65 <li><a href="#terminators">Terminator Instructions</a>
66 <ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]67 <li><a href="#i_ret">'<tt>ret</tt>' Instruction</a></li>
68 <li><a href="#i_br">'<tt>br</tt>' Instruction</a></li>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]69 <li><a href="#i_switch">'<tt>switch</tt>' Instruction</a></li>
70 <li><a href="#i_invoke">'<tt>invoke</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]71 <li><a href="#i_unwind">'<tt>unwind</tt>' Instruction</a></li>
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]72 <li><a href="#i_unreachable">'<tt>unreachable</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]73 </ol>
74 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]75 <li><a href="#binaryops">Binary Operations</a>
76 <ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]77 <li><a href="#i_add">'<tt>add</tt>' Instruction</a></li>
78 <li><a href="#i_sub">'<tt>sub</tt>' Instruction</a></li>
79 <li><a href="#i_mul">'<tt>mul</tt>' Instruction</a></li>
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame^]80 <li><a href="#i_udiv">'<tt>udiv</tt>' Instruction</a></li>
81 <li><a href="#i_sdiv">'<tt>sdiv</tt>' Instruction</a></li>
82 <li><a href="#i_fdiv">'<tt>fdiv</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]83 <li><a href="#i_rem">'<tt>rem</tt>' Instruction</a></li>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]84 <li><a href="#i_setcc">'<tt>set<i>cc</i></tt>' Instructions</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]85 </ol>
86 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]87 <li><a href="#bitwiseops">Bitwise Binary Operations</a>
88 <ol>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]89 <li><a href="#i_and">'<tt>and</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]90 <li><a href="#i_or">'<tt>or</tt>' Instruction</a></li>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]91 <li><a href="#i_xor">'<tt>xor</tt>' Instruction</a></li>
92 <li><a href="#i_shl">'<tt>shl</tt>' Instruction</a></li>
93 <li><a href="#i_shr">'<tt>shr</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]94 </ol>
95 </li>
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]96 <li><a href="#vectorops">Vector Operations</a>
97 <ol>
98 <li><a href="#i_extractelement">'<tt>extractelement</tt>' Instruction</a></li>
99 <li><a href="#i_insertelement">'<tt>insertelement</tt>' Instruction</a></li>
100 <li><a href="#i_shufflevector">'<tt>shufflevector</tt>' Instruction</a></li>
Tanya Lattnerb138bbe2006-04-14 19:24:33 +0000[diff] [blame]101 <li><a href="#i_vsetint">'<tt>vsetint</tt>' Instruction</a></li>
102 <li><a href="#i_vsetfp">'<tt>vsetfp</tt>' Instruction</a></li>
103 <li><a href="#i_vselect">'<tt>vselect</tt>' Instruction</a></li>
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]104 </ol>
105 </li>
Chris Lattner6ab66722006-08-15 00:45:58 +0000[diff] [blame]106 <li><a href="#memoryops">Memory Access and Addressing Operations</a>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]107 <ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]108 <li><a href="#i_malloc">'<tt>malloc</tt>' Instruction</a></li>
109 <li><a href="#i_free">'<tt>free</tt>' Instruction</a></li>
110 <li><a href="#i_alloca">'<tt>alloca</tt>' Instruction</a></li>
Robert Bocchino820bc75b2006-02-17 21:18:08 +0000[diff] [blame]111 <li><a href="#i_load">'<tt>load</tt>' Instruction</a></li>
112 <li><a href="#i_store">'<tt>store</tt>' Instruction</a></li>
113 <li><a href="#i_getelementptr">'<tt>getelementptr</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]114 </ol>
115 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]116 <li><a href="#otherops">Other Operations</a>
117 <ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]118 <li><a href="#i_phi">'<tt>phi</tt>' Instruction</a></li>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]119 <li><a href="#i_cast">'<tt>cast .. to</tt>' Instruction</a></li>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]120 <li><a href="#i_select">'<tt>select</tt>' Instruction</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]121 <li><a href="#i_call">'<tt>call</tt>' Instruction</a></li>
Chris Lattner33337472006-01-13 23:26:01 +0000[diff] [blame]122 <li><a href="#i_va_arg">'<tt>va_arg</tt>' Instruction</a></li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]123 </ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]124 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]125 </ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]126 </li>
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]127 <li><a href="#intrinsics">Intrinsic Functions</a>
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]128 <ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]129 <li><a href="#int_varargs">Variable Argument Handling Intrinsics</a>
130 <ol>
131 <li><a href="#i_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a></li>
132 <li><a href="#i_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a></li>
133 <li><a href="#i_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a></li>
134 </ol>
135 </li>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]136 <li><a href="#int_gc">Accurate Garbage Collection Intrinsics</a>
137 <ol>
138 <li><a href="#i_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a></li>
139 <li><a href="#i_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a></li>
140 <li><a href="#i_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a></li>
141 </ol>
142 </li>
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]143 <li><a href="#int_codegen">Code Generator Intrinsics</a>
144 <ol>
145 <li><a href="#i_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a></li>
146 <li><a href="#i_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a></li>
Chris Lattner2f0f0012006-01-13 02:03:13 +0000[diff] [blame]147 <li><a href="#i_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a></li>
148 <li><a href="#i_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a></li>
Chris Lattnerc8a2c222005-02-28 19:24:19 +0000[diff] [blame]149 <li><a href="#i_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a></li>
Andrew Lenharthb4427912005-03-28 20:05:49 +0000[diff] [blame]150 <li><a href="#i_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a></li>
Andrew Lenharth01aa5632005-11-11 16:47:30 +0000[diff] [blame]151 <li><a href="#i_readcyclecounter"><tt>llvm.readcyclecounter</tt>' Intrinsic</a></li>
John Criswellaa1c3c12004-04-09 16:43:20 +0000[diff] [blame]152 </ol>
153 </li>
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]154 <li><a href="#int_libc">Standard C Library Intrinsics</a>
155 <ol>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]156 <li><a href="#i_memcpy">'<tt>llvm.memcpy.*</tt>' Intrinsic</a></li>
157 <li><a href="#i_memmove">'<tt>llvm.memmove.*</tt>' Intrinsic</a></li>
158 <li><a href="#i_memset">'<tt>llvm.memset.*</tt>' Intrinsic</a></li>
Chris Lattner069b5bd2006-01-16 22:38:59 +0000[diff] [blame]159 <li><a href="#i_isunordered">'<tt>llvm.isunordered.*</tt>' Intrinsic</a></li>
160 <li><a href="#i_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a></li>
Chris Lattner33b73f92006-09-08 06:34:02 +0000[diff] [blame]161 <li><a href="#i_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a></li>
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]162 </ol>
163 </li>
Nate Begeman0f223bb2006-01-13 23:26:38 +0000[diff] [blame]164 <li><a href="#int_manip">Bit Manipulation Intrinsics</a>
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]165 <ol>
Nate Begeman0f223bb2006-01-13 23:26:38 +0000[diff] [blame]166 <li><a href="#i_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a></li>
Chris Lattnerb748c672006-01-16 22:34:14 +0000[diff] [blame]167 <li><a href="#int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic </a></li>
168 <li><a href="#int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic </a></li>
169 <li><a href="#int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic </a></li>
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]170 </ol>
171 </li>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]172 <li><a href="#int_debugger">Debugger intrinsics</a></li>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]173 </ol>
174 </li>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]175</ol>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]176
177<div class="doc_author">
178 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>
179 and <a href="mailto:vadve@cs.uiuc.edu">Vikram Adve</a></p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]180</div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]181
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]182<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]183<div class="doc_section"> <a name="abstract">Abstract </a></div>
184<!-- *********************************************************************** -->
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]185
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]186<div class="doc_text">
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]187<p>This document is a reference manual for the LLVM assembly language.
188LLVM is an SSA based representation that provides type safety,
189low-level operations, flexibility, and the capability of representing
190'all' high-level languages cleanly. It is the common code
191representation used throughout all phases of the LLVM compilation
192strategy.</p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]193</div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]194
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]195<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]196<div class="doc_section"> <a name="introduction">Introduction</a> </div>
197<!-- *********************************************************************** -->
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]198
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]199<div class="doc_text">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]200
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]201<p>The LLVM code representation is designed to be used in three
202different forms: as an in-memory compiler IR, as an on-disk bytecode
203representation (suitable for fast loading by a Just-In-Time compiler),
204and as a human readable assembly language representation. This allows
205LLVM to provide a powerful intermediate representation for efficient
206compiler transformations and analysis, while providing a natural means
207to debug and visualize the transformations. The three different forms
208of LLVM are all equivalent. This document describes the human readable
209representation and notation.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]210
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]211<p>The LLVM representation aims to be light-weight and low-level
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]212while being expressive, typed, and extensible at the same time. It
213aims to be a "universal IR" of sorts, by being at a low enough level
214that high-level ideas may be cleanly mapped to it (similar to how
215microprocessors are "universal IR's", allowing many source languages to
216be mapped to them). By providing type information, LLVM can be used as
217the target of optimizations: for example, through pointer analysis, it
218can be proven that a C automatic variable is never accessed outside of
219the current function... allowing it to be promoted to a simple SSA
220value instead of a memory location.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]221
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]222</div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]223
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]224<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]225<div class="doc_subsubsection"> <a name="wellformed">Well-Formedness</a> </div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]226
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]227<div class="doc_text">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]228
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]229<p>It is important to note that this document describes 'well formed'
230LLVM assembly language. There is a difference between what the parser
231accepts and what is considered 'well formed'. For example, the
232following instruction is syntactically okay, but not well formed:</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]233
234<pre>
235 %x = <a href="#i_add">add</a> int 1, %x
236</pre>
237
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]238<p>...because the definition of <tt>%x</tt> does not dominate all of
239its uses. The LLVM infrastructure provides a verification pass that may
240be used to verify that an LLVM module is well formed. This pass is
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]241automatically run by the parser after parsing input assembly and by
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]242the optimizer before it outputs bytecode. The violations pointed out
243by the verifier pass indicate bugs in transformation passes or input to
244the parser.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]245
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]246<!-- Describe the typesetting conventions here. --> </div>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]247
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]248<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]249<div class="doc_section"> <a name="identifiers">Identifiers</a> </div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]250<!-- *********************************************************************** -->
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]251
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]252<div class="doc_text">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]253
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]254<p>LLVM uses three different forms of identifiers, for different
255purposes:</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]256
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]257<ol>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]258 <li>Named values are represented as a string of characters with a '%' prefix.
259 For example, %foo, %DivisionByZero, %a.really.long.identifier. The actual
260 regular expression used is '<tt>%[a-zA-Z$._][a-zA-Z$._0-9]*</tt>'.
261 Identifiers which require other characters in their names can be surrounded
262 with quotes. In this way, anything except a <tt>"</tt> character can be used
263 in a name.</li>
264
265 <li>Unnamed values are represented as an unsigned numeric value with a '%'
266 prefix. For example, %12, %2, %44.</li>
267
Reid Spencer8f08d802004-12-09 18:02:53 +0000[diff] [blame]268 <li>Constants, which are described in a <a href="#constants">section about
269 constants</a>, below.</li>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]270</ol>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]271
272<p>LLVM requires that values start with a '%' sign for two reasons: Compilers
273don't need to worry about name clashes with reserved words, and the set of
274reserved words may be expanded in the future without penalty. Additionally,
275unnamed identifiers allow a compiler to quickly come up with a temporary
276variable without having to avoid symbol table conflicts.</p>
277
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]278<p>Reserved words in LLVM are very similar to reserved words in other
279languages. There are keywords for different opcodes ('<tt><a
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]280href="#i_add">add</a></tt>', '<tt><a href="#i_cast">cast</a></tt>', '<tt><a
281href="#i_ret">ret</a></tt>', etc...), for primitive type names ('<tt><a
282href="#t_void">void</a></tt>', '<tt><a href="#t_uint">uint</a></tt>', etc...),
283and others. These reserved words cannot conflict with variable names, because
284none of them start with a '%' character.</p>
285
286<p>Here is an example of LLVM code to multiply the integer variable
287'<tt>%X</tt>' by 8:</p>
288
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]289<p>The easy way:</p>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]290
291<pre>
292 %result = <a href="#i_mul">mul</a> uint %X, 8
293</pre>
294
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]295<p>After strength reduction:</p>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]296
297<pre>
298 %result = <a href="#i_shl">shl</a> uint %X, ubyte 3
299</pre>
300
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]301<p>And the hard way:</p>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]302
303<pre>
304 <a href="#i_add">add</a> uint %X, %X <i>; yields {uint}:%0</i>
305 <a href="#i_add">add</a> uint %0, %0 <i>; yields {uint}:%1</i>
306 %result = <a href="#i_add">add</a> uint %1, %1
307</pre>
308
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]309<p>This last way of multiplying <tt>%X</tt> by 8 illustrates several
310important lexical features of LLVM:</p>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]311
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]312<ol>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]313
314 <li>Comments are delimited with a '<tt>;</tt>' and go until the end of
315 line.</li>
316
317 <li>Unnamed temporaries are created when the result of a computation is not
318 assigned to a named value.</li>
319
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]320 <li>Unnamed temporaries are numbered sequentially</li>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]321
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]322</ol>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]323
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]324<p>...and it also shows a convention that we follow in this document. When
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]325demonstrating instructions, we will follow an instruction with a comment that
326defines the type and name of value produced. Comments are shown in italic
327text.</p>
328
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]329</div>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]330
331<!-- *********************************************************************** -->
332<div class="doc_section"> <a name="highlevel">High Level Structure</a> </div>
333<!-- *********************************************************************** -->
334
335<!-- ======================================================================= -->
336<div class="doc_subsection"> <a name="modulestructure">Module Structure</a>
337</div>
338
339<div class="doc_text">
340
341<p>LLVM programs are composed of "Module"s, each of which is a
342translation unit of the input programs. Each module consists of
343functions, global variables, and symbol table entries. Modules may be
344combined together with the LLVM linker, which merges function (and
345global variable) definitions, resolves forward declarations, and merges
346symbol table entries. Here is an example of the "hello world" module:</p>
347
348<pre><i>; Declare the string constant as a global constant...</i>
349<a href="#identifiers">%.LC0</a> = <a href="#linkage_internal">internal</a> <a
350 href="#globalvars">constant</a> <a href="#t_array">[13 x sbyte]</a> c"hello world\0A\00" <i>; [13 x sbyte]*</i>
351
352<i>; External declaration of the puts function</i>
353<a href="#functionstructure">declare</a> int %puts(sbyte*) <i>; int(sbyte*)* </i>
354
Chris Lattnerd2d29a02006-06-13 03:05:47 +0000[diff] [blame]355<i>; Global variable / Function body section separator</i>
356implementation
357
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]358<i>; Definition of main function</i>
359int %main() { <i>; int()* </i>
360 <i>; Convert [13x sbyte]* to sbyte *...</i>
361 %cast210 = <a
362 href="#i_getelementptr">getelementptr</a> [13 x sbyte]* %.LC0, long 0, long 0 <i>; sbyte*</i>
363
364 <i>; Call puts function to write out the string to stdout...</i>
365 <a
366 href="#i_call">call</a> int %puts(sbyte* %cast210) <i>; int</i>
367 <a
368 href="#i_ret">ret</a> int 0<br>}<br></pre>
369
370<p>This example is made up of a <a href="#globalvars">global variable</a>
371named "<tt>.LC0</tt>", an external declaration of the "<tt>puts</tt>"
372function, and a <a href="#functionstructure">function definition</a>
373for "<tt>main</tt>".</p>
374
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]375<p>In general, a module is made up of a list of global values,
376where both functions and global variables are global values. Global values are
377represented by a pointer to a memory location (in this case, a pointer to an
378array of char, and a pointer to a function), and have one of the following <a
379href="#linkage">linkage types</a>.</p>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]380
Chris Lattnerd2d29a02006-06-13 03:05:47 +0000[diff] [blame]381<p>Due to a limitation in the current LLVM assembly parser (it is limited by
382one-token lookahead), modules are split into two pieces by the "implementation"
383keyword. Global variable prototypes and definitions must occur before the
384keyword, and function definitions must occur after it. Function prototypes may
385occur either before or after it. In the future, the implementation keyword may
386become a noop, if the parser gets smarter.</p>
387
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]388</div>
389
390<!-- ======================================================================= -->
391<div class="doc_subsection">
392 <a name="linkage">Linkage Types</a>
393</div>
394
395<div class="doc_text">
396
397<p>
398All Global Variables and Functions have one of the following types of linkage:
399</p>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]400
401<dl>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]402
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]403 <dt><tt><b><a name="linkage_internal">internal</a></b></tt> </dt>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]404
405 <dd>Global values with internal linkage are only directly accessible by
406 objects in the current module. In particular, linking code into a module with
407 an internal global value may cause the internal to be renamed as necessary to
408 avoid collisions. Because the symbol is internal to the module, all
409 references can be updated. This corresponds to the notion of the
410 '<tt>static</tt>' keyword in C, or the idea of "anonymous namespaces" in C++.
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]411 </dd>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]412
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]413 <dt><tt><b><a name="linkage_linkonce">linkonce</a></b></tt>: </dt>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]414
415 <dd>"<tt>linkonce</tt>" linkage is similar to <tt>internal</tt> linkage, with
416 the twist that linking together two modules defining the same
417 <tt>linkonce</tt> globals will cause one of the globals to be discarded. This
418 is typically used to implement inline functions. Unreferenced
419 <tt>linkonce</tt> globals are allowed to be discarded.
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]420 </dd>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]421
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]422 <dt><tt><b><a name="linkage_weak">weak</a></b></tt>: </dt>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]423
424 <dd>"<tt>weak</tt>" linkage is exactly the same as <tt>linkonce</tt> linkage,
425 except that unreferenced <tt>weak</tt> globals may not be discarded. This is
426 used to implement constructs in C such as "<tt>int X;</tt>" at global scope.
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]427 </dd>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]428
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]429 <dt><tt><b><a name="linkage_appending">appending</a></b></tt>: </dt>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]430
431 <dd>"<tt>appending</tt>" linkage may only be applied to global variables of
432 pointer to array type. When two global variables with appending linkage are
433 linked together, the two global arrays are appended together. This is the
434 LLVM, typesafe, equivalent of having the system linker append together
435 "sections" with identical names when .o files are linked.
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]436 </dd>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]437
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]438 <dt><tt><b><a name="linkage_external">externally visible</a></b></tt>:</dt>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]439
440 <dd>If none of the above identifiers are used, the global is externally
441 visible, meaning that it participates in linkage and can be used to resolve
442 external symbol references.
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]443 </dd>
Anton Korobeynikovd61d39e2006-09-14 18:23:27 +0000[diff] [blame]444
445 <dt><tt><b><a name="linkage_externweak">extern_weak</a></b></tt>: </dt>
446
447 <dd>"<tt>extern_weak</tt>" TBD
448 </dd>
449
450 <p>
451 The next two types of linkage are targeted for Microsoft Windows platform
452 only. They are designed to support importing (exporting) symbols from (to)
453 DLLs.
454 </p>
455
456 <dt><tt><b><a name="linkage_dllimport">dllimport</a></b></tt>: </dt>
457
458 <dd>"<tt>dllimport</tt>" linkage causes the compiler to reference a function
459 or variable via a global pointer to a pointer that is set up by the DLL
460 exporting the symbol. On Microsoft Windows targets, the pointer name is
461 formed by combining <code>_imp__</code> and the function or variable name.
462 </dd>
463
464 <dt><tt><b><a name="linkage_dllexport">dllexport</a></b></tt>: </dt>
465
466 <dd>"<tt>dllexport</tt>" linkage causes the compiler to provide a global
467 pointer to a pointer in a DLL, so that it can be referenced with the
468 <tt>dllimport</tt> attribute. On Microsoft Windows targets, the pointer
469 name is formed by combining <code>_imp__</code> and the function or variable
470 name.
471 </dd>
472
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]473</dl>
474
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]475<p><a name="linkage_external">For example, since the "<tt>.LC0</tt>"
476variable is defined to be internal, if another module defined a "<tt>.LC0</tt>"
477variable and was linked with this one, one of the two would be renamed,
478preventing a collision. Since "<tt>main</tt>" and "<tt>puts</tt>" are
479external (i.e., lacking any linkage declarations), they are accessible
480outside of the current module. It is illegal for a function <i>declaration</i>
481to have any linkage type other than "externally visible".</a></p>
Chris Lattnerd79749a2004-12-09 16:36:40 +0000[diff] [blame]482
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]483</div>
484
485<!-- ======================================================================= -->
486<div class="doc_subsection">
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]487 <a name="callingconv">Calling Conventions</a>
488</div>
489
490<div class="doc_text">
491
492<p>LLVM <a href="#functionstructure">functions</a>, <a href="#i_call">calls</a>
493and <a href="#i_invoke">invokes</a> can all have an optional calling convention
494specified for the call. The calling convention of any pair of dynamic
495caller/callee must match, or the behavior of the program is undefined. The
496following calling conventions are supported by LLVM, and more may be added in
497the future:</p>
498
499<dl>
500 <dt><b>"<tt>ccc</tt>" - The C calling convention</b>:</dt>
501
502 <dd>This calling convention (the default if no other calling convention is
503 specified) matches the target C calling conventions. This calling convention
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]504 supports varargs function calls and tolerates some mismatch in the declared
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]505 prototype and implemented declaration of the function (as does normal C).
506 </dd>
507
Chris Lattner95ff1952006-05-19 21:15:36 +0000[diff] [blame]508 <dt><b>"<tt>csretcc</tt>" - The C struct return calling convention</b>:</dt>
509
510 <dd>This calling convention matches the target C calling conventions, except
511 that functions with this convention are required to take a pointer as their
512 first argument, and the return type of the function must be void. This is
513 used for C functions that return aggregates by-value. In this case, the
514 function has been transformed to take a pointer to the struct as the first
515 argument to the function. For targets where the ABI specifies specific
516 behavior for structure-return calls, the calling convention can be used to
517 distinguish between struct return functions and other functions that take a
518 pointer to a struct as the first argument.
519 </dd>
520
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]521 <dt><b>"<tt>fastcc</tt>" - The fast calling convention</b>:</dt>
522
523 <dd>This calling convention attempts to make calls as fast as possible
524 (e.g. by passing things in registers). This calling convention allows the
525 target to use whatever tricks it wants to produce fast code for the target,
Chris Lattnerc792eb32005-05-06 23:08:23 +0000[diff] [blame]526 without having to conform to an externally specified ABI. Implementations of
527 this convention should allow arbitrary tail call optimization to be supported.
528 This calling convention does not support varargs and requires the prototype of
529 all callees to exactly match the prototype of the function definition.
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]530 </dd>
531
532 <dt><b>"<tt>coldcc</tt>" - The cold calling convention</b>:</dt>
533
534 <dd>This calling convention attempts to make code in the caller as efficient
535 as possible under the assumption that the call is not commonly executed. As
536 such, these calls often preserve all registers so that the call does not break
537 any live ranges in the caller side. This calling convention does not support
538 varargs and requires the prototype of all callees to exactly match the
539 prototype of the function definition.
540 </dd>
541
Chris Lattner573f64e2005-05-07 01:46:40 +0000[diff] [blame]542 <dt><b>"<tt>cc &lt;<em>n</em>&gt;</tt>" - Numbered convention</b>:</dt>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]543
544 <dd>Any calling convention may be specified by number, allowing
545 target-specific calling conventions to be used. Target specific calling
546 conventions start at 64.
547 </dd>
Chris Lattner573f64e2005-05-07 01:46:40 +0000[diff] [blame]548</dl>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]549
550<p>More calling conventions can be added/defined on an as-needed basis, to
551support pascal conventions or any other well-known target-independent
552convention.</p>
553
554</div>
555
556<!-- ======================================================================= -->
557<div class="doc_subsection">
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]558 <a name="globalvars">Global Variables</a>
559</div>
560
561<div class="doc_text">
562
Chris Lattner5d5aede2005-02-12 19:30:21 +0000[diff] [blame]563<p>Global variables define regions of memory allocated at compilation time
Chris Lattner662c8722005-11-12 00:45:07 +0000[diff] [blame]564instead of run-time. Global variables may optionally be initialized, may have
565an explicit section to be placed in, and may
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]566have an optional explicit alignment specified. A
John Criswell4c0cf7f2005-10-24 16:17:18 +0000[diff] [blame]567variable may be defined as a global "constant," which indicates that the
Chris Lattner5d5aede2005-02-12 19:30:21 +0000[diff] [blame]568contents of the variable will <b>never</b> be modified (enabling better
569optimization, allowing the global data to be placed in the read-only section of
570an executable, etc). Note that variables that need runtime initialization
John Criswell4c0cf7f2005-10-24 16:17:18 +0000[diff] [blame]571cannot be marked "constant" as there is a store to the variable.</p>
Chris Lattner5d5aede2005-02-12 19:30:21 +0000[diff] [blame]572
573<p>
574LLVM explicitly allows <em>declarations</em> of global variables to be marked
575constant, even if the final definition of the global is not. This capability
576can be used to enable slightly better optimization of the program, but requires
577the language definition to guarantee that optimizations based on the
578'constantness' are valid for the translation units that do not include the
579definition.
580</p>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]581
582<p>As SSA values, global variables define pointer values that are in
583scope (i.e. they dominate) all basic blocks in the program. Global
584variables always define a pointer to their "content" type because they
585describe a region of memory, and all memory objects in LLVM are
586accessed through pointers.</p>
587
Chris Lattner662c8722005-11-12 00:45:07 +0000[diff] [blame]588<p>LLVM allows an explicit section to be specified for globals. If the target
589supports it, it will emit globals to the section specified.</p>
590
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]591<p>An explicit alignment may be specified for a global. If not present, or if
592the alignment is set to zero, the alignment of the global is set by the target
593to whatever it feels convenient. If an explicit alignment is specified, the
594global is forced to have at least that much alignment. All alignments must be
595a power of 2.</p>
596
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]597</div>
598
599
600<!-- ======================================================================= -->
601<div class="doc_subsection">
602 <a name="functionstructure">Functions</a>
603</div>
604
605<div class="doc_text">
606
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]607<p>LLVM function definitions consist of an optional <a href="#linkage">linkage
608type</a>, an optional <a href="#callingconv">calling convention</a>, a return
Chris Lattner662c8722005-11-12 00:45:07 +0000[diff] [blame]609type, a function name, a (possibly empty) argument list, an optional section,
610an optional alignment, an opening curly brace,
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]611a list of basic blocks, and a closing curly brace. LLVM function declarations
612are defined with the "<tt>declare</tt>" keyword, an optional <a
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]613href="#callingconv">calling convention</a>, a return type, a function name,
614a possibly empty list of arguments, and an optional alignment.</p>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]615
616<p>A function definition contains a list of basic blocks, forming the CFG for
617the function. Each basic block may optionally start with a label (giving the
618basic block a symbol table entry), contains a list of instructions, and ends
619with a <a href="#terminators">terminator</a> instruction (such as a branch or
620function return).</p>
621
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]622<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]623executed on entrance to the function, and it is not allowed to have predecessor
624basic blocks (i.e. there can not be any branches to the entry block of a
625function). Because the block can have no predecessors, it also cannot have any
626<a href="#i_phi">PHI nodes</a>.</p>
627
628<p>LLVM functions are identified by their name and type signature. Hence, two
629functions with the same name but different parameter lists or return values are
Chris Lattner455fc8c2005-03-07 22:13:59 +0000[diff] [blame]630considered different functions, and LLVM will resolve references to each
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]631appropriately.</p>
632
Chris Lattner662c8722005-11-12 00:45:07 +0000[diff] [blame]633<p>LLVM allows an explicit section to be specified for functions. If the target
634supports it, it will emit functions to the section specified.</p>
635
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]636<p>An explicit alignment may be specified for a function. If not present, or if
637the alignment is set to zero, the alignment of the function is set by the target
638to whatever it feels convenient. If an explicit alignment is specified, the
639function is forced to have at least that much alignment. All alignments must be
640a power of 2.</p>
641
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]642</div>
643
Chris Lattner91c15c42006-01-23 23:23:47 +0000[diff] [blame]644<!-- ======================================================================= -->
645<div class="doc_subsection">
Chris Lattner93564892006-04-08 04:40:53 +0000[diff] [blame]646 <a name="moduleasm">Module-Level Inline Assembly</a>
Chris Lattner91c15c42006-01-23 23:23:47 +0000[diff] [blame]647</div>
648
649<div class="doc_text">
650<p>
651Modules may contain "module-level inline asm" blocks, which corresponds to the
652GCC "file scope inline asm" blocks. These blocks are internally concatenated by
653LLVM and treated as a single unit, but may be separated in the .ll file if
654desired. The syntax is very simple:
655</p>
656
657<div class="doc_code"><pre>
Chris Lattnera1280ad2006-01-24 00:37:20 +0000[diff] [blame]658 module asm "inline asm code goes here"
659 module asm "more can go here"
Chris Lattner91c15c42006-01-23 23:23:47 +0000[diff] [blame]660</pre></div>
661
662<p>The strings can contain any character by escaping non-printable characters.
663 The escape sequence used is simply "\xx" where "xx" is the two digit hex code
664 for the number.
665</p>
666
667<p>
668 The inline asm code is simply printed to the machine code .s file when
669 assembly code is generated.
670</p>
671</div>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]672
673
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]674<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]675<div class="doc_section"> <a name="typesystem">Type System</a> </div>
676<!-- *********************************************************************** -->
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]677
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]678<div class="doc_text">
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]679
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]680<p>The LLVM type system is one of the most important features of the
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]681intermediate representation. Being typed enables a number of
682optimizations to be performed on the IR directly, without having to do
683extra analyses on the side before the transformation. A strong type
684system makes it easier to read the generated code and enables novel
685analyses and transformations that are not feasible to perform on normal
686three address code representations.</p>
Chris Lattner6af02f32004-12-09 16:11:40 +0000[diff] [blame]687
688</div>
689
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]690<!-- ======================================================================= -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]691<div class="doc_subsection"> <a name="t_primitive">Primitive Types</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]692<div class="doc_text">
John Criswell417228d2004-04-09 16:48:45 +0000[diff] [blame]693<p>The primitive types are the fundamental building blocks of the LLVM
Chris Lattner455fc8c2005-03-07 22:13:59 +0000[diff] [blame]694system. The current set of primitive types is as follows:</p>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]695
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]696<table class="layout">
697 <tr class="layout">
698 <td class="left">
699 <table>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]700 <tbody>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]701 <tr><th>Type</th><th>Description</th></tr>
702 <tr><td><tt>void</tt></td><td>No value</td></tr>
Misha Brukman36c6bc12005-04-22 18:02:52 +0000[diff] [blame]703 <tr><td><tt>ubyte</tt></td><td>Unsigned 8-bit value</td></tr>
704 <tr><td><tt>ushort</tt></td><td>Unsigned 16-bit value</td></tr>
705 <tr><td><tt>uint</tt></td><td>Unsigned 32-bit value</td></tr>
706 <tr><td><tt>ulong</tt></td><td>Unsigned 64-bit value</td></tr>
707 <tr><td><tt>float</tt></td><td>32-bit floating point value</td></tr>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]708 <tr><td><tt>label</tt></td><td>Branch destination</td></tr>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]709 </tbody>
710 </table>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]711 </td>
712 <td class="right">
713 <table>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]714 <tbody>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]715 <tr><th>Type</th><th>Description</th></tr>
716 <tr><td><tt>bool</tt></td><td>True or False value</td></tr>
Misha Brukman36c6bc12005-04-22 18:02:52 +0000[diff] [blame]717 <tr><td><tt>sbyte</tt></td><td>Signed 8-bit value</td></tr>
718 <tr><td><tt>short</tt></td><td>Signed 16-bit value</td></tr>
719 <tr><td><tt>int</tt></td><td>Signed 32-bit value</td></tr>
720 <tr><td><tt>long</tt></td><td>Signed 64-bit value</td></tr>
721 <tr><td><tt>double</tt></td><td>64-bit floating point value</td></tr>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]722 </tbody>
723 </table>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]724 </td>
725 </tr>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]726</table>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]727</div>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]728
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]729<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]730<div class="doc_subsubsection"> <a name="t_classifications">Type
731Classifications</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]732<div class="doc_text">
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]733<p>These different primitive types fall into a few useful
734classifications:</p>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]735
736<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]737 <tbody>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]738 <tr><th>Classification</th><th>Types</th></tr>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]739 <tr>
740 <td><a name="t_signed">signed</a></td>
741 <td><tt>sbyte, short, int, long, float, double</tt></td>
742 </tr>
743 <tr>
744 <td><a name="t_unsigned">unsigned</a></td>
745 <td><tt>ubyte, ushort, uint, ulong</tt></td>
746 </tr>
747 <tr>
748 <td><a name="t_integer">integer</a></td>
749 <td><tt>ubyte, sbyte, ushort, short, uint, int, ulong, long</tt></td>
750 </tr>
751 <tr>
752 <td><a name="t_integral">integral</a></td>
Misha Brukman20f9a622004-08-12 20:16:08 +0000[diff] [blame]753 <td><tt>bool, ubyte, sbyte, ushort, short, uint, int, ulong, long</tt>
754 </td>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]755 </tr>
756 <tr>
757 <td><a name="t_floating">floating point</a></td>
758 <td><tt>float, double</tt></td>
759 </tr>
760 <tr>
761 <td><a name="t_firstclass">first class</a></td>
Misha Brukman20f9a622004-08-12 20:16:08 +0000[diff] [blame]762 <td><tt>bool, ubyte, sbyte, ushort, short, uint, int, ulong, long,<br>
763 float, double, <a href="#t_pointer">pointer</a>,
764 <a href="#t_packed">packed</a></tt></td>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]765 </tr>
766 </tbody>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]767</table>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]768
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]769<p>The <a href="#t_firstclass">first class</a> types are perhaps the
770most important. Values of these types are the only ones which can be
771produced by instructions, passed as arguments, or used as operands to
772instructions. This means that all structures and arrays must be
773manipulated either by pointer or by component.</p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]774</div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]775
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]776<!-- ======================================================================= -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]777<div class="doc_subsection"> <a name="t_derived">Derived Types</a> </div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]778
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]779<div class="doc_text">
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]780
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]781<p>The real power in LLVM comes from the derived types in the system.
782This is what allows a programmer to represent arrays, functions,
783pointers, and other useful types. Note that these derived types may be
784recursive: For example, it is possible to have a two dimensional array.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]785
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]786</div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]787
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]788<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]789<div class="doc_subsubsection"> <a name="t_array">Array Type</a> </div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]790
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]791<div class="doc_text">
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]792
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]793<h5>Overview:</h5>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]794
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]795<p>The array type is a very simple derived type that arranges elements
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]796sequentially in memory. The array type requires a size (number of
797elements) and an underlying data type.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]798
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]799<h5>Syntax:</h5>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]800
801<pre>
802 [&lt;# elements&gt; x &lt;elementtype&gt;]
803</pre>
804
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]805<p>The number of elements is a constant integer value; elementtype may
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]806be any type with a size.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]807
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]808<h5>Examples:</h5>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]809<table class="layout">
810 <tr class="layout">
811 <td class="left">
812 <tt>[40 x int ]</tt><br/>
813 <tt>[41 x int ]</tt><br/>
814 <tt>[40 x uint]</tt><br/>
815 </td>
816 <td class="left">
817 Array of 40 integer values.<br/>
818 Array of 41 integer values.<br/>
819 Array of 40 unsigned integer values.<br/>
820 </td>
821 </tr>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]822</table>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]823<p>Here are some examples of multidimensional arrays:</p>
824<table class="layout">
825 <tr class="layout">
826 <td class="left">
827 <tt>[3 x [4 x int]]</tt><br/>
828 <tt>[12 x [10 x float]]</tt><br/>
829 <tt>[2 x [3 x [4 x uint]]]</tt><br/>
830 </td>
831 <td class="left">
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]832 3x4 array of integer values.<br/>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]833 12x10 array of single precision floating point values.<br/>
834 2x3x4 array of unsigned integer values.<br/>
835 </td>
836 </tr>
837</table>
Chris Lattnerc0ad71e2005-06-24 17:22:57 +0000[diff] [blame]838
John Criswell4c0cf7f2005-10-24 16:17:18 +0000[diff] [blame]839<p>Note that 'variable sized arrays' can be implemented in LLVM with a zero
840length array. Normally, accesses past the end of an array are undefined in
Chris Lattnerc0ad71e2005-06-24 17:22:57 +0000[diff] [blame]841LLVM (e.g. it is illegal to access the 5th element of a 3 element array).
842As a special case, however, zero length arrays are recognized to be variable
843length. This allows implementation of 'pascal style arrays' with the LLVM
844type "{ int, [0 x float]}", for example.</p>
845
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]846</div>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]847
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]848<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]849<div class="doc_subsubsection"> <a name="t_function">Function Type</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]850<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]851<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]852<p>The function type can be thought of as a function signature. It
853consists of a return type and a list of formal parameter types.
John Criswella0d50d22003-11-25 21:45:46 +0000[diff] [blame]854Function types are usually used to build virtual function tables
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]855(which are structures of pointers to functions), for indirect function
856calls, and when defining a function.</p>
John Criswella0d50d22003-11-25 21:45:46 +0000[diff] [blame]857<p>
858The return type of a function type cannot be an aggregate type.
859</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]860<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]861<pre> &lt;returntype&gt; (&lt;parameter list&gt;)<br></pre>
John Criswell4c0cf7f2005-10-24 16:17:18 +0000[diff] [blame]862<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]863specifiers. Optionally, the parameter list may include a type <tt>...</tt>,
Chris Lattner5ed60612003-09-03 00:41:47 +0000[diff] [blame]864which indicates that the function takes a variable number of arguments.
865Variable argument functions can access their arguments with the <a
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]866 href="#int_varargs">variable argument handling intrinsic</a> functions.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]867<h5>Examples:</h5>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]868<table class="layout">
869 <tr class="layout">
870 <td class="left">
871 <tt>int (int)</tt> <br/>
872 <tt>float (int, int *) *</tt><br/>
873 <tt>int (sbyte *, ...)</tt><br/>
874 </td>
875 <td class="left">
876 function taking an <tt>int</tt>, returning an <tt>int</tt><br/>
877 <a href="#t_pointer">Pointer</a> to a function that takes an
Misha Brukman20f9a622004-08-12 20:16:08 +0000[diff] [blame]878 <tt>int</tt> and a <a href="#t_pointer">pointer</a> to <tt>int</tt>,
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]879 returning <tt>float</tt>.<br/>
880 A vararg function that takes at least one <a href="#t_pointer">pointer</a>
881 to <tt>sbyte</tt> (signed char in C), which returns an integer. This is
882 the signature for <tt>printf</tt> in LLVM.<br/>
883 </td>
884 </tr>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]885</table>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]886
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]887</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]888<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]889<div class="doc_subsubsection"> <a name="t_struct">Structure Type</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]890<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]891<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]892<p>The structure type is used to represent a collection of data members
893together in memory. The packing of the field types is defined to match
894the ABI of the underlying processor. The elements of a structure may
895be any type that has a size.</p>
896<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
897and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
898field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
899instruction.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]900<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]901<pre> { &lt;type list&gt; }<br></pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]902<h5>Examples:</h5>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]903<table class="layout">
904 <tr class="layout">
905 <td class="left">
906 <tt>{ int, int, int }</tt><br/>
907 <tt>{ float, int (int) * }</tt><br/>
908 </td>
909 <td class="left">
910 a triple of three <tt>int</tt> values<br/>
911 A pair, where the first element is a <tt>float</tt> and the second element
912 is a <a href="#t_pointer">pointer</a> to a <a href="#t_function">function</a>
913 that takes an <tt>int</tt>, returning an <tt>int</tt>.<br/>
914 </td>
915 </tr>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]916</table>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]917</div>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]918
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]919<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]920<div class="doc_subsubsection"> <a name="t_pointer">Pointer Type</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]921<div class="doc_text">
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]922<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]923<p>As in many languages, the pointer type represents a pointer or
924reference to another object, which must live in memory.</p>
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]925<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]926<pre> &lt;type&gt; *<br></pre>
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]927<h5>Examples:</h5>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]928<table class="layout">
929 <tr class="layout">
930 <td class="left">
931 <tt>[4x int]*</tt><br/>
932 <tt>int (int *) *</tt><br/>
933 </td>
934 <td class="left">
935 A <a href="#t_pointer">pointer</a> to <a href="#t_array">array</a> of
936 four <tt>int</tt> values<br/>
937 A <a href="#t_pointer">pointer</a> to a <a
Chris Lattner16fb0032005-02-19 02:22:14 +0000[diff] [blame]938 href="#t_function">function</a> that takes an <tt>int*</tt>, returning an
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]939 <tt>int</tt>.<br/>
940 </td>
941 </tr>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]942</table>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]943</div>
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]944
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]945<!-- _______________________________________________________________________ -->
946<div class="doc_subsubsection"> <a name="t_packed">Packed Type</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]947<div class="doc_text">
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]948
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]949<h5>Overview:</h5>
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]950
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]951<p>A packed type is a simple derived type that represents a vector
952of elements. Packed types are used when multiple primitive data
953are operated in parallel using a single instruction (SIMD).
954A packed type requires a size (number of
Chris Lattner330ce692005-11-10 01:44:22 +0000[diff] [blame]955elements) and an underlying primitive data type. Vectors must have a power
956of two length (1, 2, 4, 8, 16 ...). Packed types are
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]957considered <a href="#t_firstclass">first class</a>.</p>
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]958
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]959<h5>Syntax:</h5>
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]960
961<pre>
962 &lt; &lt;# elements&gt; x &lt;elementtype&gt; &gt;
963</pre>
964
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]965<p>The number of elements is a constant integer value; elementtype may
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]966be any integral or floating point type.</p>
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]967
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]968<h5>Examples:</h5>
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]969
Reid Spencerc3c4c4f2004-11-01 08:19:36 +0000[diff] [blame]970<table class="layout">
971 <tr class="layout">
972 <td class="left">
973 <tt>&lt;4 x int&gt;</tt><br/>
974 <tt>&lt;8 x float&gt;</tt><br/>
975 <tt>&lt;2 x uint&gt;</tt><br/>
976 </td>
977 <td class="left">
978 Packed vector of 4 integer values.<br/>
979 Packed vector of 8 floating-point values.<br/>
980 Packed vector of 2 unsigned integer values.<br/>
981 </td>
982 </tr>
983</table>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]984</div>
985
Chris Lattner37b6b092005-04-25 17:34:15 +0000[diff] [blame]986<!-- _______________________________________________________________________ -->
987<div class="doc_subsubsection"> <a name="t_opaque">Opaque Type</a> </div>
988<div class="doc_text">
989
990<h5>Overview:</h5>
991
992<p>Opaque types are used to represent unknown types in the system. This
993corresponds (for example) to the C notion of a foward declared structure type.
994In LLVM, opaque types can eventually be resolved to any type (not just a
995structure type).</p>
996
997<h5>Syntax:</h5>
998
999<pre>
1000 opaque
1001</pre>
1002
1003<h5>Examples:</h5>
1004
1005<table class="layout">
1006 <tr class="layout">
1007 <td class="left">
1008 <tt>opaque</tt>
1009 </td>
1010 <td class="left">
1011 An opaque type.<br/>
1012 </td>
1013 </tr>
1014</table>
1015</div>
1016
1017
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1018<!-- *********************************************************************** -->
1019<div class="doc_section"> <a name="constants">Constants</a> </div>
1020<!-- *********************************************************************** -->
1021
1022<div class="doc_text">
1023
1024<p>LLVM has several different basic types of constants. This section describes
1025them all and their syntax.</p>
1026
1027</div>
1028
1029<!-- ======================================================================= -->
Reid Spencer8f08d802004-12-09 18:02:53 +0000[diff] [blame]1030<div class="doc_subsection"><a name="simpleconstants">Simple Constants</a></div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1031
1032<div class="doc_text">
1033
1034<dl>
1035 <dt><b>Boolean constants</b></dt>
1036
1037 <dd>The two strings '<tt>true</tt>' and '<tt>false</tt>' are both valid
1038 constants of the <tt><a href="#t_primitive">bool</a></tt> type.
1039 </dd>
1040
1041 <dt><b>Integer constants</b></dt>
1042
Reid Spencer8f08d802004-12-09 18:02:53 +0000[diff] [blame]1043 <dd>Standard integers (such as '4') are constants of the <a
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1044 href="#t_integer">integer</a> type. Negative numbers may be used with signed
1045 integer types.
1046 </dd>
1047
1048 <dt><b>Floating point constants</b></dt>
1049
1050 <dd>Floating point constants use standard decimal notation (e.g. 123.421),
1051 exponential notation (e.g. 1.23421e+2), or a more precise hexadecimal
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1052 notation (see below). Floating point constants must have a <a
1053 href="#t_floating">floating point</a> type. </dd>
1054
1055 <dt><b>Null pointer constants</b></dt>
1056
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]1057 <dd>The identifier '<tt>null</tt>' is recognized as a null pointer constant
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1058 and must be of <a href="#t_pointer">pointer type</a>.</dd>
1059
1060</dl>
1061
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]1062<p>The one non-intuitive notation for constants is the optional hexadecimal form
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1063of floating point constants. For example, the form '<tt>double
10640x432ff973cafa8000</tt>' is equivalent to (but harder to read than) '<tt>double
10654.5e+15</tt>'. The only time hexadecimal floating point constants are required
Reid Spencer8f08d802004-12-09 18:02:53 +0000[diff] [blame]1066(and the only time that they are generated by the disassembler) is when a
1067floating point constant must be emitted but it cannot be represented as a
1068decimal floating point number. For example, NaN's, infinities, and other
1069special values are represented in their IEEE hexadecimal format so that
1070assembly and disassembly do not cause any bits to change in the constants.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1071
1072</div>
1073
1074<!-- ======================================================================= -->
1075<div class="doc_subsection"><a name="aggregateconstants">Aggregate Constants</a>
1076</div>
1077
1078<div class="doc_text">
Chris Lattner455fc8c2005-03-07 22:13:59 +0000[diff] [blame]1079<p>Aggregate constants arise from aggregation of simple constants
1080and smaller aggregate constants.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1081
1082<dl>
1083 <dt><b>Structure constants</b></dt>
1084
1085 <dd>Structure constants are represented with notation similar to structure
1086 type definitions (a comma separated list of elements, surrounded by braces
Chris Lattner455fc8c2005-03-07 22:13:59 +0000[diff] [blame]1087 (<tt>{}</tt>)). For example: "<tt>{ int 4, float 17.0, int* %G }</tt>",
1088 where "<tt>%G</tt>" is declared as "<tt>%G = external global int</tt>". Structure constants
1089 must have <a href="#t_struct">structure type</a>, and the number and
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1090 types of elements must match those specified by the type.
1091 </dd>
1092
1093 <dt><b>Array constants</b></dt>
1094
1095 <dd>Array constants are represented with notation similar to array type
1096 definitions (a comma separated list of elements, surrounded by square brackets
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]1097 (<tt>[]</tt>)). For example: "<tt>[ int 42, int 11, int 74 ]</tt>". Array
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1098 constants must have <a href="#t_array">array type</a>, and the number and
1099 types of elements must match those specified by the type.
1100 </dd>
1101
1102 <dt><b>Packed constants</b></dt>
1103
1104 <dd>Packed constants are represented with notation similar to packed type
1105 definitions (a comma separated list of elements, surrounded by
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]1106 less-than/greater-than's (<tt>&lt;&gt;</tt>)). For example: "<tt>&lt; int 42,
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1107 int 11, int 74, int 100 &gt;</tt>". Packed constants must have <a
1108 href="#t_packed">packed type</a>, and the number and types of elements must
1109 match those specified by the type.
1110 </dd>
1111
1112 <dt><b>Zero initialization</b></dt>
1113
1114 <dd>The string '<tt>zeroinitializer</tt>' can be used to zero initialize a
1115 value to zero of <em>any</em> type, including scalar and aggregate types.
1116 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]1117 large arrays) and is always exactly equivalent to using explicit zero
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1118 initializers.
1119 </dd>
1120</dl>
1121
1122</div>
1123
1124<!-- ======================================================================= -->
1125<div class="doc_subsection">
1126 <a name="globalconstants">Global Variable and Function Addresses</a>
1127</div>
1128
1129<div class="doc_text">
1130
1131<p>The addresses of <a href="#globalvars">global variables</a> and <a
1132href="#functionstructure">functions</a> are always implicitly valid (link-time)
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]1133constants. These constants are explicitly referenced when the <a
1134href="#identifiers">identifier for the global</a> is used and always have <a
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1135href="#t_pointer">pointer</a> type. For example, the following is a legal LLVM
1136file:</p>
1137
1138<pre>
1139 %X = global int 17
1140 %Y = global int 42
1141 %Z = global [2 x int*] [ int* %X, int* %Y ]
1142</pre>
1143
1144</div>
1145
1146<!-- ======================================================================= -->
Reid Spencer641f5c92004-12-09 18:13:12 +0000[diff] [blame]1147<div class="doc_subsection"><a name="undefvalues">Undefined Values</a></div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1148<div class="doc_text">
Reid Spencer641f5c92004-12-09 18:13:12 +0000[diff] [blame]1149 <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]1150 no specific value. Undefined values may be of any type and be used anywhere
Reid Spencer641f5c92004-12-09 18:13:12 +0000[diff] [blame]1151 a constant is permitted.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1152
Reid Spencer641f5c92004-12-09 18:13:12 +0000[diff] [blame]1153 <p>Undefined values indicate to the compiler that the program is well defined
1154 no matter what value is used, giving the compiler more freedom to optimize.
1155 </p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1156</div>
1157
1158<!-- ======================================================================= -->
1159<div class="doc_subsection"><a name="constantexprs">Constant Expressions</a>
1160</div>
1161
1162<div class="doc_text">
1163
1164<p>Constant expressions are used to allow expressions involving other constants
1165to be used as constants. Constant expressions may be of any <a
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]1166href="#t_firstclass">first class</a> type and may involve any LLVM operation
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1167that does not have side effects (e.g. load and call are not supported). The
1168following is the syntax for constant expressions:</p>
1169
1170<dl>
1171 <dt><b><tt>cast ( CST to TYPE )</tt></b></dt>
1172
1173 <dd>Cast a constant to another type.</dd>
1174
1175 <dt><b><tt>getelementptr ( CSTPTR, IDX0, IDX1, ... )</tt></b></dt>
1176
1177 <dd>Perform the <a href="#i_getelementptr">getelementptr operation</a> on
1178 constants. As with the <a href="#i_getelementptr">getelementptr</a>
1179 instruction, the index list may have zero or more indexes, which are required
1180 to make sense for the type of "CSTPTR".</dd>
1181
Robert Bocchino7e97a6d2006-01-10 19:31:34 +0000[diff] [blame]1182 <dt><b><tt>select ( COND, VAL1, VAL2 )</tt></b></dt>
1183
1184 <dd>Perform the <a href="#i_select">select operation</a> on
1185 constants.
1186
1187 <dt><b><tt>extractelement ( VAL, IDX )</tt></b></dt>
1188
1189 <dd>Perform the <a href="#i_extractelement">extractelement
1190 operation</a> on constants.
1191
Robert Bocchinof72fdfe2006-01-15 20:48:27 +0000[diff] [blame]1192 <dt><b><tt>insertelement ( VAL, ELT, IDX )</tt></b></dt>
1193
1194 <dd>Perform the <a href="#i_insertelement">insertelement
1195 operation</a> on constants.
1196
Chris Lattner016a0e52006-04-08 00:13:41 +0000[diff] [blame]1197
1198 <dt><b><tt>shufflevector ( VEC1, VEC2, IDXMASK )</tt></b></dt>
1199
1200 <dd>Perform the <a href="#i_shufflevector">shufflevector
1201 operation</a> on constants.
1202
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1203 <dt><b><tt>OPCODE ( LHS, RHS )</tt></b></dt>
1204
Reid Spencer641f5c92004-12-09 18:13:12 +0000[diff] [blame]1205 <dd>Perform the specified operation of the LHS and RHS constants. OPCODE may
1206 be any of the <a href="#binaryops">binary</a> or <a href="#bitwiseops">bitwise
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1207 binary</a> operations. The constraints on operands are the same as those for
1208 the corresponding instruction (e.g. no bitwise operations on floating point
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]1209 values are allowed).</dd>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1210</dl>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1211</div>
Chris Lattnerb1652612004-03-08 16:49:10 +0000[diff] [blame]1212
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1213<!-- *********************************************************************** -->
Chris Lattner98f013c2006-01-25 23:47:57 +0000[diff] [blame]1214<div class="doc_section"> <a name="othervalues">Other Values</a> </div>
1215<!-- *********************************************************************** -->
1216
1217<!-- ======================================================================= -->
1218<div class="doc_subsection">
1219<a name="inlineasm">Inline Assembler Expressions</a>
1220</div>
1221
1222<div class="doc_text">
1223
1224<p>
1225LLVM supports inline assembler expressions (as opposed to <a href="#moduleasm">
1226Module-Level Inline Assembly</a>) through the use of a special value. This
1227value represents the inline assembler as a string (containing the instructions
1228to emit), a list of operand constraints (stored as a string), and a flag that
1229indicates whether or not the inline asm expression has side effects. An example
1230inline assembler expression is:
1231</p>
1232
1233<pre>
1234 int(int) asm "bswap $0", "=r,r"
1235</pre>
1236
1237<p>
1238Inline assembler expressions may <b>only</b> be used as the callee operand of
1239a <a href="#i_call"><tt>call</tt> instruction</a>. Thus, typically we have:
1240</p>
1241
1242<pre>
1243 %X = call int asm "<a href="#i_bswap">bswap</a> $0", "=r,r"(int %Y)
1244</pre>
1245
1246<p>
1247Inline asms with side effects not visible in the constraint list must be marked
1248as having side effects. This is done through the use of the
1249'<tt>sideeffect</tt>' keyword, like so:
1250</p>
1251
1252<pre>
1253 call void asm sideeffect "eieio", ""()
1254</pre>
1255
1256<p>TODO: The format of the asm and constraints string still need to be
1257documented here. Constraints on what can be done (e.g. duplication, moving, etc
1258need to be documented).
1259</p>
1260
1261</div>
1262
1263<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1264<div class="doc_section"> <a name="instref">Instruction Reference</a> </div>
1265<!-- *********************************************************************** -->
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1266
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1267<div class="doc_text">
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1268
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1269<p>The LLVM instruction set consists of several different
1270classifications of instructions: <a href="#terminators">terminator
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]1271instructions</a>, <a href="#binaryops">binary instructions</a>,
1272<a href="#bitwiseops">bitwise binary instructions</a>, <a
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1273 href="#memoryops">memory instructions</a>, and <a href="#otherops">other
1274instructions</a>.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1275
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1276</div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1277
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1278<!-- ======================================================================= -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1279<div class="doc_subsection"> <a name="terminators">Terminator
1280Instructions</a> </div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1281
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1282<div class="doc_text">
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1283
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1284<p>As mentioned <a href="#functionstructure">previously</a>, every
1285basic block in a program ends with a "Terminator" instruction, which
1286indicates which block should be executed after the current block is
1287finished. These terminator instructions typically yield a '<tt>void</tt>'
1288value: they produce control flow, not values (the one exception being
1289the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction).</p>
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]1290<p>There are six different terminator instructions: the '<a
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1291 href="#i_ret"><tt>ret</tt></a>' instruction, the '<a href="#i_br"><tt>br</tt></a>'
1292instruction, the '<a href="#i_switch"><tt>switch</tt></a>' instruction,
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1293the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction, the '<a
1294 href="#i_unwind"><tt>unwind</tt></a>' instruction, and the '<a
1295 href="#i_unreachable"><tt>unreachable</tt></a>' instruction.</p>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1296
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1297</div>
Chris Lattner74d3f822004-12-09 17:30:23 +0000[diff] [blame]1298
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1299<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1300<div class="doc_subsubsection"> <a name="i_ret">'<tt>ret</tt>'
1301Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1302<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1303<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1304<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]1305 ret void <i>; Return from void function</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1306</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1307<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1308<p>The '<tt>ret</tt>' instruction is used to return control flow (and a
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]1309value) from a function back to the caller.</p>
John Criswell417228d2004-04-09 16:48:45 +0000[diff] [blame]1310<p>There are two forms of the '<tt>ret</tt>' instruction: one that
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1311returns a value and then causes control flow, and one that just causes
1312control flow to occur.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1313<h5>Arguments:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1314<p>The '<tt>ret</tt>' instruction may return any '<a
1315 href="#t_firstclass">first class</a>' type. Notice that a function is
1316not <a href="#wellformed">well formed</a> if there exists a '<tt>ret</tt>'
1317instruction inside of the function that returns a value that does not
1318match the return type of the function.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1319<h5>Semantics:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1320<p>When the '<tt>ret</tt>' instruction is executed, control flow
1321returns back to the calling function's context. If the caller is a "<a
John Criswell40db33f2004-06-25 15:16:57 +0000[diff] [blame]1322 href="#i_call"><tt>call</tt></a>" instruction, execution continues at
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1323the instruction after the call. If the caller was an "<a
1324 href="#i_invoke"><tt>invoke</tt></a>" instruction, execution continues
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]1325at the beginning of the "normal" destination block. If the instruction
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1326returns a value, that value shall set the call or invoke instruction's
1327return value.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1328<h5>Example:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1329<pre> ret int 5 <i>; Return an integer value of 5</i>
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]1330 ret void <i>; Return from a void function</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1331</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1332</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1333<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1334<div class="doc_subsubsection"> <a name="i_br">'<tt>br</tt>' Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1335<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1336<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1337<pre> br bool &lt;cond&gt;, label &lt;iftrue&gt;, label &lt;iffalse&gt;<br> br label &lt;dest&gt; <i>; Unconditional branch</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1338</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1339<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1340<p>The '<tt>br</tt>' instruction is used to cause control flow to
1341transfer to a different basic block in the current function. There are
1342two forms of this instruction, corresponding to a conditional branch
1343and an unconditional branch.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1344<h5>Arguments:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1345<p>The conditional branch form of the '<tt>br</tt>' instruction takes a
1346single '<tt>bool</tt>' value and two '<tt>label</tt>' values. The
1347unconditional form of the '<tt>br</tt>' instruction takes a single '<tt>label</tt>'
1348value as a target.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1349<h5>Semantics:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1350<p>Upon execution of a conditional '<tt>br</tt>' instruction, the '<tt>bool</tt>'
1351argument is evaluated. If the value is <tt>true</tt>, control flows
1352to the '<tt>iftrue</tt>' <tt>label</tt> argument. If "cond" is <tt>false</tt>,
1353control flows to the '<tt>iffalse</tt>' <tt>label</tt> argument.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1354<h5>Example:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1355<pre>Test:<br> %cond = <a href="#i_setcc">seteq</a> int %a, %b<br> br bool %cond, label %IfEqual, label %IfUnequal<br>IfEqual:<br> <a
1356 href="#i_ret">ret</a> int 1<br>IfUnequal:<br> <a href="#i_ret">ret</a> int 0<br></pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1357</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1358<!-- _______________________________________________________________________ -->
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1359<div class="doc_subsubsection">
1360 <a name="i_switch">'<tt>switch</tt>' Instruction</a>
1361</div>
1362
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1363<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1364<h5>Syntax:</h5>
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1365
1366<pre>
1367 switch &lt;intty&gt; &lt;value&gt;, label &lt;defaultdest&gt; [ &lt;intty&gt; &lt;val&gt;, label &lt;dest&gt; ... ]
1368</pre>
1369
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1370<h5>Overview:</h5>
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1371
1372<p>The '<tt>switch</tt>' instruction is used to transfer control flow to one of
1373several different places. It is a generalization of the '<tt>br</tt>'
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1374instruction, allowing a branch to occur to one of many possible
1375destinations.</p>
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1376
1377
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1378<h5>Arguments:</h5>
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1379
1380<p>The '<tt>switch</tt>' instruction uses three parameters: an integer
1381comparison value '<tt>value</tt>', a default '<tt>label</tt>' destination, and
1382an array of pairs of comparison value constants and '<tt>label</tt>'s. The
1383table is not allowed to contain duplicate constant entries.</p>
1384
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1385<h5>Semantics:</h5>
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1386
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1387<p>The <tt>switch</tt> instruction specifies a table of values and
1388destinations. When the '<tt>switch</tt>' instruction is executed, this
John Criswellbcbb18c2004-06-25 16:05:06 +0000[diff] [blame]1389table is searched for the given value. If the value is found, control flow is
1390transfered to the corresponding destination; otherwise, control flow is
1391transfered to the default destination.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1392
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1393<h5>Implementation:</h5>
1394
1395<p>Depending on properties of the target machine and the particular
1396<tt>switch</tt> instruction, this instruction may be code generated in different
John Criswellbcbb18c2004-06-25 16:05:06 +0000[diff] [blame]1397ways. For example, it could be generated as a series of chained conditional
1398branches or with a lookup table.</p>
Chris Lattnercf96c6c2004-02-24 04:54:45 +0000[diff] [blame]1399
1400<h5>Example:</h5>
1401
1402<pre>
1403 <i>; Emulate a conditional br instruction</i>
1404 %Val = <a href="#i_cast">cast</a> bool %value to int
1405 switch int %Val, label %truedest [int 0, label %falsedest ]
1406
1407 <i>; Emulate an unconditional br instruction</i>
1408 switch uint 0, label %dest [ ]
1409
1410 <i>; Implement a jump table:</i>
1411 switch uint %val, label %otherwise [ uint 0, label %onzero
1412 uint 1, label %onone
1413 uint 2, label %ontwo ]
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1414</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1415</div>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1416
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1417<!-- _______________________________________________________________________ -->
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1418<div class="doc_subsubsection">
1419 <a name="i_invoke">'<tt>invoke</tt>' Instruction</a>
1420</div>
1421
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1422<div class="doc_text">
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1423
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1424<h5>Syntax:</h5>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1425
1426<pre>
1427 &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]1428 to label &lt;normal label&gt; unwind label &lt;exception label&gt;
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1429</pre>
1430
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]1431<h5>Overview:</h5>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1432
1433<p>The '<tt>invoke</tt>' instruction causes control to transfer to a specified
1434function, with the possibility of control flow transfer to either the
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]1435'<tt>normal</tt>' label or the
1436'<tt>exception</tt>' label. If the callee function returns with the
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1437"<tt><a href="#i_ret">ret</a></tt>" instruction, control flow will return to the
1438"normal" label. If the callee (or any indirect callees) returns with the "<a
John Criswell02fdc6f2005-05-12 16:52:32 +0000[diff] [blame]1439href="#i_unwind"><tt>unwind</tt></a>" instruction, control is interrupted and
1440continued at the dynamically nearest "exception" label.</p>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1441
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1442<h5>Arguments:</h5>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1443
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1444<p>This instruction requires several arguments:</p>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1445
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1446<ol>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1447 <li>
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]1448 The optional "cconv" marker indicates which <a href="callingconv">calling
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1449 convention</a> the call should use. If none is specified, the call defaults
1450 to using C calling conventions.
1451 </li>
1452 <li>'<tt>ptr to function ty</tt>': shall be the signature of the pointer to
1453 function value being invoked. In most cases, this is a direct function
1454 invocation, but indirect <tt>invoke</tt>s are just as possible, branching off
1455 an arbitrary pointer to function value.
1456 </li>
1457
1458 <li>'<tt>function ptr val</tt>': An LLVM value containing a pointer to a
1459 function to be invoked. </li>
1460
1461 <li>'<tt>function args</tt>': argument list whose types match the function
1462 signature argument types. If the function signature indicates the function
1463 accepts a variable number of arguments, the extra arguments can be
1464 specified. </li>
1465
1466 <li>'<tt>normal label</tt>': the label reached when the called function
1467 executes a '<tt><a href="#i_ret">ret</a></tt>' instruction. </li>
1468
1469 <li>'<tt>exception label</tt>': the label reached when a callee returns with
1470 the <a href="#i_unwind"><tt>unwind</tt></a> instruction. </li>
1471
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1472</ol>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1473
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1474<h5>Semantics:</h5>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1475
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1476<p>This instruction is designed to operate as a standard '<tt><a
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1477href="#i_call">call</a></tt>' instruction in most regards. The primary
1478difference is that it establishes an association with a label, which is used by
1479the runtime library to unwind the stack.</p>
1480
1481<p>This instruction is used in languages with destructors to ensure that proper
1482cleanup is performed in the case of either a <tt>longjmp</tt> or a thrown
1483exception. Additionally, this is important for implementation of
1484'<tt>catch</tt>' clauses in high-level languages that support them.</p>
1485
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1486<h5>Example:</h5>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1487<pre>
1488 %retval = invoke int %Test(int 15) to label %Continue
Chris Lattner6b7a0082006-05-14 18:23:06 +0000[diff] [blame]1489 unwind label %TestCleanup <i>; {int}:retval set</i>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]1490 %retval = invoke <a href="#callingconv">coldcc</a> int %Test(int 15) to label %Continue
Chris Lattner6b7a0082006-05-14 18:23:06 +0000[diff] [blame]1491 unwind label %TestCleanup <i>; {int}:retval set</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1492</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1493</div>
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1494
1495
Chris Lattner5ed60612003-09-03 00:41:47 +0000[diff] [blame]1496<!-- _______________________________________________________________________ -->
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1497
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1498<div class="doc_subsubsection"> <a name="i_unwind">'<tt>unwind</tt>'
1499Instruction</a> </div>
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1500
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1501<div class="doc_text">
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1502
Chris Lattner5ed60612003-09-03 00:41:47 +0000[diff] [blame]1503<h5>Syntax:</h5>
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1504<pre>
1505 unwind
1506</pre>
1507
Chris Lattner5ed60612003-09-03 00:41:47 +0000[diff] [blame]1508<h5>Overview:</h5>
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1509
1510<p>The '<tt>unwind</tt>' instruction unwinds the stack, continuing control flow
1511at the first callee in the dynamic call stack which used an <a
1512href="#i_invoke"><tt>invoke</tt></a> instruction to perform the call. This is
1513primarily used to implement exception handling.</p>
1514
Chris Lattner5ed60612003-09-03 00:41:47 +0000[diff] [blame]1515<h5>Semantics:</h5>
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1516
1517<p>The '<tt>unwind</tt>' intrinsic causes execution of the current function to
1518immediately halt. The dynamic call stack is then searched for the first <a
1519href="#i_invoke"><tt>invoke</tt></a> instruction on the call stack. Once found,
1520execution continues at the "exceptional" destination block specified by the
1521<tt>invoke</tt> instruction. If there is no <tt>invoke</tt> instruction in the
1522dynamic call chain, undefined behavior results.</p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1523</div>
Chris Lattner08b7d5b2004-10-16 18:04:13 +0000[diff] [blame]1524
1525<!-- _______________________________________________________________________ -->
1526
1527<div class="doc_subsubsection"> <a name="i_unreachable">'<tt>unreachable</tt>'
1528Instruction</a> </div>
1529
1530<div class="doc_text">
1531
1532<h5>Syntax:</h5>
1533<pre>
1534 unreachable
1535</pre>
1536
1537<h5>Overview:</h5>
1538
1539<p>The '<tt>unreachable</tt>' instruction has no defined semantics. This
1540instruction is used to inform the optimizer that a particular portion of the
1541code is not reachable. This can be used to indicate that the code after a
1542no-return function cannot be reached, and other facts.</p>
1543
1544<h5>Semantics:</h5>
1545
1546<p>The '<tt>unreachable</tt>' instruction has no defined semantics.</p>
1547</div>
1548
1549
1550
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1551<!-- ======================================================================= -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1552<div class="doc_subsection"> <a name="binaryops">Binary Operations</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1553<div class="doc_text">
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1554<p>Binary operators are used to do most of the computation in a
1555program. They require two operands, execute an operation on them, and
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]1556produce a single value. The operands might represent
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1557multiple data, as is the case with the <a href="#t_packed">packed</a> data type.
1558The result value of a binary operator is not
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1559necessarily the same type as its operands.</p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1560<p>There are several different binary operators:</p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1561</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1562<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1563<div class="doc_subsubsection"> <a name="i_add">'<tt>add</tt>'
1564Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1565<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1566<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1567<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]1568</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1569<h5>Overview:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1570<p>The '<tt>add</tt>' instruction returns the sum of its two operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1571<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1572<p>The two arguments to the '<tt>add</tt>' instruction must be either <a
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1573 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a> values.
1574 This instruction can also take <a href="#t_packed">packed</a> versions of the values.
1575Both arguments must have identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1576<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1577<p>The value produced is the integer or floating point sum of the two
1578operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1579<h5>Example:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1580<pre> &lt;result&gt; = add int 4, %var <i>; yields {int}:result = 4 + %var</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1581</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1582</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1583<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1584<div class="doc_subsubsection"> <a name="i_sub">'<tt>sub</tt>'
1585Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1586<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1587<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1588<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]1589</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1590<h5>Overview:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1591<p>The '<tt>sub</tt>' instruction returns the difference of its two
1592operands.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1593<p>Note that the '<tt>sub</tt>' instruction is used to represent the '<tt>neg</tt>'
1594instruction present in most other intermediate representations.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1595<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1596<p>The two arguments to the '<tt>sub</tt>' instruction must be either <a
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1597 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1598values.
1599This instruction can also take <a href="#t_packed">packed</a> versions of the values.
1600Both arguments must have identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1601<h5>Semantics:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1602<p>The value produced is the integer or floating point difference of
1603the two operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1604<h5>Example:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1605<pre> &lt;result&gt; = sub int 4, %var <i>; yields {int}:result = 4 - %var</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1606 &lt;result&gt; = sub int 0, %val <i>; yields {int}:result = -%var</i>
1607</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1608</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1609<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1610<div class="doc_subsubsection"> <a name="i_mul">'<tt>mul</tt>'
1611Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1612<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1613<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1614<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]1615</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1616<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1617<p>The '<tt>mul</tt>' instruction returns the product of its two
1618operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1619<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1620<p>The two arguments to the '<tt>mul</tt>' instruction must be either <a
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1621 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1622values.
1623This instruction can also take <a href="#t_packed">packed</a> versions of the values.
1624Both arguments must have identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1625<h5>Semantics:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1626<p>The value produced is the integer or floating point product of the
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1627two operands.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1628<p>There is no signed vs unsigned multiplication. The appropriate
1629action is taken based on the type of the operand.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1630<h5>Example:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1631<pre> &lt;result&gt; = mul int 4, %var <i>; yields {int}:result = 4 * %var</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1632</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1633</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1634<!-- _______________________________________________________________________ -->
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame^]1635<div class="doc_subsubsection"> <a name="i_udiv">'<tt>udiv</tt>' Instruction
1636</a></div>
1637<div class="doc_text">
1638<h5>Syntax:</h5>
1639<pre> &lt;result&gt; = udiv &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
1640</pre>
1641<h5>Overview:</h5>
1642<p>The '<tt>udiv</tt>' instruction returns the quotient of its two
1643operands.</p>
1644<h5>Arguments:</h5>
1645<p>The two arguments to the '<tt>udiv</tt>' instruction must be
1646<a href="#t_integer">integer</a> values. Both arguments must have identical
1647types. This instruction can also take <a href="#t_packed">packed</a> versions
1648of the values in which case the elements must be integers.</p>
1649<h5>Semantics:</h5>
1650<p>The value produced is the unsigned integer quotient of the two operands. This
1651instruction always performs an unsigned division operation, regardless of
1652whether the arguments are unsigned or not.</p>
1653<h5>Example:</h5>
1654<pre> &lt;result&gt; = udiv uint 4, %var <i>; yields {uint}:result = 4 / %var</i>
1655</pre>
1656</div>
1657<!-- _______________________________________________________________________ -->
1658<div class="doc_subsubsection"> <a name="i_sdiv">'<tt>sdiv</tt>' Instruction
1659</a> </div>
1660<div class="doc_text">
1661<h5>Syntax:</h5>
1662<pre> &lt;result&gt; = sdiv &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
1663</pre>
1664<h5>Overview:</h5>
1665<p>The '<tt>sdiv</tt>' instruction returns the quotient of its two
1666operands.</p>
1667<h5>Arguments:</h5>
1668<p>The two arguments to the '<tt>sdiv</tt>' instruction must be
1669<a href="#t_integer">integer</a> values. Both arguments must have identical
1670types. This instruction can also take <a href="#t_packed">packed</a> versions
1671of the values in which case the elements must be integers.</p>
1672<h5>Semantics:</h5>
1673<p>The value produced is the signed integer quotient of the two operands. This
1674instruction always performs a signed division operation, regardless of whether
1675the arguments are signed or not.</p>
1676<h5>Example:</h5>
1677<pre> &lt;result&gt; = sdiv int 4, %var <i>; yields {int}:result = 4 / %var</i>
1678</pre>
1679</div>
1680<!-- _______________________________________________________________________ -->
1681<div class="doc_subsubsection"> <a name="i_fdiv">'<tt>fdiv</tt>'
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1682Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1683<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1684<h5>Syntax:</h5>
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame^]1685<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]1686</pre>
1687<h5>Overview:</h5>
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame^]1688<p>The '<tt>fdiv</tt>' instruction returns the quotient of its two
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1689operands.</p>
1690<h5>Arguments:</h5>
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame^]1691<p>The two arguments to the '<tt>div</tt>' instruction must be
1692<a href="#t_floating">floating point</a> values. Both arguments must have
1693identical types. This instruction can also take <a href="#t_packed">packed</a>
1694versions of the values in which case the elements must be floating point.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1695<h5>Semantics:</h5>
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame^]1696<p>The value produced is the floating point quotient of the two operands.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1697<h5>Example:</h5>
Reid Spencer7e80b0b2006-10-26 06:15:43 +0000[diff] [blame^]1698<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]1699</pre>
1700</div>
1701<!-- _______________________________________________________________________ -->
1702<div class="doc_subsubsection"> <a name="i_rem">'<tt>rem</tt>'
1703Instruction</a> </div>
1704<div class="doc_text">
1705<h5>Syntax:</h5>
1706<pre> &lt;result&gt; = rem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
1707</pre>
1708<h5>Overview:</h5>
1709<p>The '<tt>rem</tt>' instruction returns the remainder from the
1710division of its two operands.</p>
1711<h5>Arguments:</h5>
1712<p>The two arguments to the '<tt>rem</tt>' instruction must be either <a
1713 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
Chris Lattnerc8cb6952004-08-12 19:12:28 +0000[diff] [blame]1714values.
1715This instruction can also take <a href="#t_packed">packed</a> versions of the values.
1716Both arguments must have identical types.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1717<h5>Semantics:</h5>
1718<p>This returns the <i>remainder</i> of a division (where the result
1719has the same sign as the divisor), not the <i>modulus</i> (where the
1720result has the same sign as the dividend) of a value. For more
John Criswell4c0cf7f2005-10-24 16:17:18 +0000[diff] [blame]1721information about the difference, see <a
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1722 href="http://mathforum.org/dr.math/problems/anne.4.28.99.html">The
1723Math Forum</a>.</p>
1724<h5>Example:</h5>
1725<pre> &lt;result&gt; = rem int 4, %var <i>; yields {int}:result = 4 % %var</i>
1726</pre>
Robert Bocchino820bc75b2006-02-17 21:18:08 +0000[diff] [blame]1727
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1728</div>
1729<!-- _______________________________________________________________________ -->
1730<div class="doc_subsubsection"> <a name="i_setcc">'<tt>set<i>cc</i></tt>'
1731Instructions</a> </div>
1732<div class="doc_text">
1733<h5>Syntax:</h5>
1734<pre> &lt;result&gt; = seteq &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {bool}:result</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1735 &lt;result&gt; = setne &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {bool}:result</i>
1736 &lt;result&gt; = setlt &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {bool}:result</i>
1737 &lt;result&gt; = setgt &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {bool}:result</i>
1738 &lt;result&gt; = setle &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {bool}:result</i>
1739 &lt;result&gt; = setge &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {bool}:result</i>
1740</pre>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1741<h5>Overview:</h5>
1742<p>The '<tt>set<i>cc</i></tt>' family of instructions returns a boolean
1743value based on a comparison of their two operands.</p>
1744<h5>Arguments:</h5>
1745<p>The two arguments to the '<tt>set<i>cc</i></tt>' instructions must
1746be of <a href="#t_firstclass">first class</a> type (it is not possible
1747to compare '<tt>label</tt>'s, '<tt>array</tt>'s, '<tt>structure</tt>'
1748or '<tt>void</tt>' values, etc...). Both arguments must have identical
1749types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1750<h5>Semantics:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1751<p>The '<tt>seteq</tt>' instruction yields a <tt>true</tt> '<tt>bool</tt>'
1752value if both operands are equal.<br>
1753The '<tt>setne</tt>' instruction yields a <tt>true</tt> '<tt>bool</tt>'
1754value if both operands are unequal.<br>
1755The '<tt>setlt</tt>' instruction yields a <tt>true</tt> '<tt>bool</tt>'
1756value if the first operand is less than the second operand.<br>
1757The '<tt>setgt</tt>' instruction yields a <tt>true</tt> '<tt>bool</tt>'
1758value if the first operand is greater than the second operand.<br>
1759The '<tt>setle</tt>' instruction yields a <tt>true</tt> '<tt>bool</tt>'
1760value if the first operand is less than or equal to the second operand.<br>
1761The '<tt>setge</tt>' instruction yields a <tt>true</tt> '<tt>bool</tt>'
1762value if the first operand is greater than or equal to the second
1763operand.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1764<h5>Example:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1765<pre> &lt;result&gt; = seteq int 4, 5 <i>; yields {bool}:result = false</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1766 &lt;result&gt; = setne float 4, 5 <i>; yields {bool}:result = true</i>
1767 &lt;result&gt; = setlt uint 4, 5 <i>; yields {bool}:result = true</i>
1768 &lt;result&gt; = setgt sbyte 4, 5 <i>; yields {bool}:result = false</i>
1769 &lt;result&gt; = setle sbyte 4, 5 <i>; yields {bool}:result = true</i>
1770 &lt;result&gt; = setge sbyte 4, 5 <i>; yields {bool}:result = false</i>
1771</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1772</div>
Robert Bocchino820bc75b2006-02-17 21:18:08 +0000[diff] [blame]1773
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1774<!-- ======================================================================= -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1775<div class="doc_subsection"> <a name="bitwiseops">Bitwise Binary
1776Operations</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1777<div class="doc_text">
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1778<p>Bitwise binary operators are used to do various forms of
1779bit-twiddling in a program. They are generally very efficient
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]1780instructions and can commonly be strength reduced from other
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1781instructions. They require two operands, execute an operation on them,
1782and produce a single value. The resulting value of the bitwise binary
1783operators is always the same type as its first operand.</p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1784</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1785<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1786<div class="doc_subsubsection"> <a name="i_and">'<tt>and</tt>'
1787Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1788<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1789<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1790<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]1791</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1792<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1793<p>The '<tt>and</tt>' instruction returns the bitwise logical and of
1794its two operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1795<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1796<p>The two arguments to the '<tt>and</tt>' instruction must be <a
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1797 href="#t_integral">integral</a> values. Both arguments must have
1798identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1799<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1800<p>The truth table used for the '<tt>and</tt>' instruction is:</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1801<p> </p>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]1802<div style="align: center">
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1803<table border="1" cellspacing="0" cellpadding="4">
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1804 <tbody>
1805 <tr>
1806 <td>In0</td>
1807 <td>In1</td>
1808 <td>Out</td>
1809 </tr>
1810 <tr>
1811 <td>0</td>
1812 <td>0</td>
1813 <td>0</td>
1814 </tr>
1815 <tr>
1816 <td>0</td>
1817 <td>1</td>
1818 <td>0</td>
1819 </tr>
1820 <tr>
1821 <td>1</td>
1822 <td>0</td>
1823 <td>0</td>
1824 </tr>
1825 <tr>
1826 <td>1</td>
1827 <td>1</td>
1828 <td>1</td>
1829 </tr>
1830 </tbody>
1831</table>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]1832</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1833<h5>Example:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1834<pre> &lt;result&gt; = and int 4, %var <i>; yields {int}:result = 4 &amp; %var</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1835 &lt;result&gt; = and int 15, 40 <i>; yields {int}:result = 8</i>
1836 &lt;result&gt; = and int 4, 8 <i>; yields {int}:result = 0</i>
1837</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1838</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1839<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1840<div class="doc_subsubsection"> <a name="i_or">'<tt>or</tt>' Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1841<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1842<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1843<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]1844</pre>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1845<h5>Overview:</h5>
1846<p>The '<tt>or</tt>' instruction returns the bitwise logical inclusive
1847or of its two operands.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1848<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1849<p>The two arguments to the '<tt>or</tt>' instruction must be <a
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1850 href="#t_integral">integral</a> values. Both arguments must have
1851identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1852<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1853<p>The truth table used for the '<tt>or</tt>' instruction is:</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1854<p> </p>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]1855<div style="align: center">
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1856<table border="1" cellspacing="0" cellpadding="4">
1857 <tbody>
1858 <tr>
1859 <td>In0</td>
1860 <td>In1</td>
1861 <td>Out</td>
1862 </tr>
1863 <tr>
1864 <td>0</td>
1865 <td>0</td>
1866 <td>0</td>
1867 </tr>
1868 <tr>
1869 <td>0</td>
1870 <td>1</td>
1871 <td>1</td>
1872 </tr>
1873 <tr>
1874 <td>1</td>
1875 <td>0</td>
1876 <td>1</td>
1877 </tr>
1878 <tr>
1879 <td>1</td>
1880 <td>1</td>
1881 <td>1</td>
1882 </tr>
1883 </tbody>
1884</table>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]1885</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1886<h5>Example:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1887<pre> &lt;result&gt; = or int 4, %var <i>; yields {int}:result = 4 | %var</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1888 &lt;result&gt; = or int 15, 40 <i>; yields {int}:result = 47</i>
1889 &lt;result&gt; = or int 4, 8 <i>; yields {int}:result = 12</i>
1890</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1891</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1892<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1893<div class="doc_subsubsection"> <a name="i_xor">'<tt>xor</tt>'
1894Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1895<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1896<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1897<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]1898</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1899<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1900<p>The '<tt>xor</tt>' instruction returns the bitwise logical exclusive
1901or of its two operands. The <tt>xor</tt> is used to implement the
1902"one's complement" operation, which is the "~" operator in C.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1903<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1904<p>The two arguments to the '<tt>xor</tt>' instruction must be <a
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1905 href="#t_integral">integral</a> values. Both arguments must have
1906identical types.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1907<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1908<p>The truth table used for the '<tt>xor</tt>' instruction is:</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1909<p> </p>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]1910<div style="align: center">
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1911<table border="1" cellspacing="0" cellpadding="4">
1912 <tbody>
1913 <tr>
1914 <td>In0</td>
1915 <td>In1</td>
1916 <td>Out</td>
1917 </tr>
1918 <tr>
1919 <td>0</td>
1920 <td>0</td>
1921 <td>0</td>
1922 </tr>
1923 <tr>
1924 <td>0</td>
1925 <td>1</td>
1926 <td>1</td>
1927 </tr>
1928 <tr>
1929 <td>1</td>
1930 <td>0</td>
1931 <td>1</td>
1932 </tr>
1933 <tr>
1934 <td>1</td>
1935 <td>1</td>
1936 <td>0</td>
1937 </tr>
1938 </tbody>
1939</table>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]1940</div>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1941<p> </p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1942<h5>Example:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1943<pre> &lt;result&gt; = xor int 4, %var <i>; yields {int}:result = 4 ^ %var</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1944 &lt;result&gt; = xor int 15, 40 <i>; yields {int}:result = 39</i>
1945 &lt;result&gt; = xor int 4, 8 <i>; yields {int}:result = 12</i>
Chris Lattner5ed60612003-09-03 00:41:47 +0000[diff] [blame]1946 &lt;result&gt; = xor int %V, -1 <i>; yields {int}:result = ~%V</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1947</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1948</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1949<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1950<div class="doc_subsubsection"> <a name="i_shl">'<tt>shl</tt>'
1951Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1952<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1953<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1954<pre> &lt;result&gt; = shl &lt;ty&gt; &lt;var1&gt;, ubyte &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1955</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1956<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1957<p>The '<tt>shl</tt>' instruction returns the first operand shifted to
1958the left a specified number of bits.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1959<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1960<p>The first argument to the '<tt>shl</tt>' instruction must be an <a
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1961 href="#t_integer">integer</a> type. The second argument must be an '<tt>ubyte</tt>'
1962type.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1963<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1964<p>The value produced is <tt>var1</tt> * 2<sup><tt>var2</tt></sup>.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1965<h5>Example:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1966<pre> &lt;result&gt; = shl int 4, ubyte %var <i>; yields {int}:result = 4 &lt;&lt; %var</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1967 &lt;result&gt; = shl int 4, ubyte 2 <i>; yields {int}:result = 16</i>
1968 &lt;result&gt; = shl int 1, ubyte 10 <i>; yields {int}:result = 1024</i>
1969</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1970</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1971<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1972<div class="doc_subsubsection"> <a name="i_shr">'<tt>shr</tt>'
1973Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1974<div class="doc_text">
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1975<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1976<pre> &lt;result&gt; = shr &lt;ty&gt; &lt;var1&gt;, ubyte &lt;var2&gt; <i>; yields {ty}:result</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1977</pre>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1978<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1979<p>The '<tt>shr</tt>' instruction returns the first operand shifted to
1980the right a specified number of bits.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1981<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]1982<p>The first argument to the '<tt>shr</tt>' instruction must be an <a
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1983 href="#t_integer">integer</a> type. The second argument must be an '<tt>ubyte</tt>'
1984type.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1985<h5>Semantics:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1986<p>If the first argument is a <a href="#t_signed">signed</a> type, the
1987most significant bit is duplicated in the newly free'd bit positions.
1988If the first argument is unsigned, zero bits shall fill the empty
1989positions.</p>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1990<h5>Example:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]1991<pre> &lt;result&gt; = shr int 4, ubyte %var <i>; yields {int}:result = 4 &gt;&gt; %var</i>
Chris Lattner33426d92003-06-18 21:30:51 +0000[diff] [blame]1992 &lt;result&gt; = shr uint 4, ubyte 1 <i>; yields {uint}:result = 2</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1993 &lt;result&gt; = shr int 4, ubyte 2 <i>; yields {int}:result = 1</i>
Chris Lattner33426d92003-06-18 21:30:51 +0000[diff] [blame]1994 &lt;result&gt; = shr sbyte 4, ubyte 3 <i>; yields {sbyte}:result = 0</i>
1995 &lt;result&gt; = shr sbyte -2, ubyte 1 <i>; yields {sbyte}:result = -1</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 Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]1998
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]1999<!-- ======================================================================= -->
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2000<div class="doc_subsection">
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2001 <a name="vectorops">Vector Operations</a>
2002</div>
2003
2004<div class="doc_text">
2005
2006<p>LLVM supports several instructions to represent vector operations in a
2007target-independent manner. This instructions cover the element-access and
2008vector-specific operations needed to process vectors effectively. While LLVM
2009does directly support these vector operations, many sophisticated algorithms
2010will want to use target-specific intrinsics to take full advantage of a specific
2011target.</p>
2012
2013</div>
2014
2015<!-- _______________________________________________________________________ -->
2016<div class="doc_subsubsection">
2017 <a name="i_extractelement">'<tt>extractelement</tt>' Instruction</a>
2018</div>
2019
2020<div class="doc_text">
2021
2022<h5>Syntax:</h5>
2023
2024<pre>
2025 &lt;result&gt; = extractelement &lt;n x &lt;ty&gt;&gt; &lt;val&gt;, uint &lt;idx&gt; <i>; yields &lt;ty&gt;</i>
2026</pre>
2027
2028<h5>Overview:</h5>
2029
2030<p>
2031The '<tt>extractelement</tt>' instruction extracts a single scalar
2032element from a packed vector at a specified index.
2033</p>
2034
2035
2036<h5>Arguments:</h5>
2037
2038<p>
2039The first operand of an '<tt>extractelement</tt>' instruction is a
2040value of <a href="#t_packed">packed</a> type. The second operand is
2041an index indicating the position from which to extract the element.
2042The index may be a variable.</p>
2043
2044<h5>Semantics:</h5>
2045
2046<p>
2047The result is a scalar of the same type as the element type of
2048<tt>val</tt>. Its value is the value at position <tt>idx</tt> of
2049<tt>val</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
2050results are undefined.
2051</p>
2052
2053<h5>Example:</h5>
2054
2055<pre>
2056 %result = extractelement &lt;4 x int&gt; %vec, uint 0 <i>; yields int</i>
2057</pre>
2058</div>
2059
2060
2061<!-- _______________________________________________________________________ -->
2062<div class="doc_subsubsection">
2063 <a name="i_insertelement">'<tt>insertelement</tt>' Instruction</a>
2064</div>
2065
2066<div class="doc_text">
2067
2068<h5>Syntax:</h5>
2069
2070<pre>
2071 &lt;result&gt; = insertelement &lt;n x &lt;ty&gt;&gt; &lt;val&gt;, &lt;ty&gt; &lt;elt&gt, uint &lt;idx&gt; <i>; yields &lt;n x &lt;ty&gt;&gt;</i>
2072</pre>
2073
2074<h5>Overview:</h5>
2075
2076<p>
2077The '<tt>insertelement</tt>' instruction inserts a scalar
2078element into a packed vector at a specified index.
2079</p>
2080
2081
2082<h5>Arguments:</h5>
2083
2084<p>
2085The first operand of an '<tt>insertelement</tt>' instruction is a
2086value of <a href="#t_packed">packed</a> type. The second operand is a
2087scalar value whose type must equal the element type of the first
2088operand. The third operand is an index indicating the position at
2089which to insert the value. The index may be a variable.</p>
2090
2091<h5>Semantics:</h5>
2092
2093<p>
2094The result is a packed vector of the same type as <tt>val</tt>. Its
2095element values are those of <tt>val</tt> except at position
2096<tt>idx</tt>, where it gets the value <tt>elt</tt>. If <tt>idx</tt>
2097exceeds the length of <tt>val</tt>, the results are undefined.
2098</p>
2099
2100<h5>Example:</h5>
2101
2102<pre>
2103 %result = insertelement &lt;4 x int&gt; %vec, int 1, uint 0 <i>; yields &lt;4 x int&gt;</i>
2104</pre>
2105</div>
2106
2107<!-- _______________________________________________________________________ -->
2108<div class="doc_subsubsection">
2109 <a name="i_shufflevector">'<tt>shufflevector</tt>' Instruction</a>
2110</div>
2111
2112<div class="doc_text">
2113
2114<h5>Syntax:</h5>
2115
2116<pre>
2117 &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 uint&gt; &lt;mask&gt; <i>; yields &lt;n x &lt;ty&gt;&gt;</i>
2118</pre>
2119
2120<h5>Overview:</h5>
2121
2122<p>
2123The '<tt>shufflevector</tt>' instruction constructs a permutation of elements
2124from two input vectors, returning a vector of the same type.
2125</p>
2126
2127<h5>Arguments:</h5>
2128
2129<p>
2130The first two operands of a '<tt>shufflevector</tt>' instruction are vectors
2131with types that match each other and types that match the result of the
2132instruction. The third argument is a shuffle mask, which has the same number
2133of elements as the other vector type, but whose element type is always 'uint'.
2134</p>
2135
2136<p>
2137The shuffle mask operand is required to be a constant vector with either
2138constant integer or undef values.
2139</p>
2140
2141<h5>Semantics:</h5>
2142
2143<p>
2144The elements of the two input vectors are numbered from left to right across
2145both of the vectors. The shuffle mask operand specifies, for each element of
2146the result vector, which element of the two input registers the result element
2147gets. The element selector may be undef (meaning "don't care") and the second
2148operand may be undef if performing a shuffle from only one vector.
2149</p>
2150
2151<h5>Example:</h5>
2152
2153<pre>
2154 %result = shufflevector &lt;4 x int&gt; %v1, &lt;4 x int&gt; %v2,
2155 &lt;4 x uint&gt; &lt;uint 0, uint 4, uint 1, uint 5&gt; <i>; yields &lt;4 x int&gt;</i>
2156 %result = shufflevector &lt;4 x int&gt; %v1, &lt;4 x int&gt; undef,
2157 &lt;4 x uint&gt; &lt;uint 0, uint 1, uint 2, uint 3&gt; <i>; yields &lt;4 x int&gt;</i> - Identity shuffle.
2158</pre>
2159</div>
2160
Tanya Lattnerb138bbe2006-04-14 19:24:33 +0000[diff] [blame]2161
2162<!-- _______________________________________________________________________ -->
2163<div class="doc_subsubsection"> <a name="i_vsetint">'<tt>vsetint</tt>'
2164Instruction</a> </div>
2165<div class="doc_text">
2166<h5>Syntax:</h5>
2167<pre>&lt;result&gt; = vsetint &lt;op&gt;, &lt;n x &lt;ty&gt;&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields &lt;n x bool&gt;</i>
2168</pre>
2169
2170<h5>Overview:</h5>
2171
2172<p>The '<tt>vsetint</tt>' instruction takes two integer vectors and
2173returns a vector of boolean values representing, at each position, the
2174result of the comparison between the values at that position in the
2175two operands.</p>
2176
2177<h5>Arguments:</h5>
2178
2179<p>The arguments to a '<tt>vsetint</tt>' instruction are a comparison
2180operation and two value arguments. The value arguments must be of <a
2181href="#t_integral">integral</a> <a href="#t_packed">packed</a> type,
2182and they must have identical types. The operation argument must be
2183one of <tt>eq</tt>, <tt>ne</tt>, <tt>slt</tt>, <tt>sgt</tt>,
2184<tt>sle</tt>, <tt>sge</tt>, <tt>ult</tt>, <tt>ugt</tt>, <tt>ule</tt>,
2185<tt>uge</tt>, <tt>true</tt>, and <tt>false</tt>. The result is a
2186packed <tt>bool</tt> value with the same length as each operand.</p>
2187
2188<h5>Semantics:</h5>
2189
2190<p>The following table shows the semantics of '<tt>vsetint</tt>'. For
2191each position of the result, the comparison is done on the
2192corresponding positions of the two value arguments. Note that the
2193signedness of the comparison depends on the comparison opcode and
2194<i>not</i> on the signedness of the value operands. E.g., <tt>vsetint
2195slt <4 x unsigned> %x, %y</tt> does an elementwise <i>signed</i>
2196comparison of <tt>%x</tt> and <tt>%y</tt>.</p>
2197
2198<table border="1" cellspacing="0" cellpadding="4">
2199 <tbody>
2200 <tr><th>Operation</th><th>Result is true iff</th><th>Comparison is</th></tr>
2201 <tr><td><tt>eq</tt></td><td>var1 == var2</td><td>--</td></tr>
2202 <tr><td><tt>ne</tt></td><td>var1 != var2</td><td>--</td></tr>
2203 <tr><td><tt>slt</tt></td><td>var1 &lt; var2</td><td>signed</td></tr>
2204 <tr><td><tt>sgt</tt></td><td>var1 &gt; var2</td><td>signed</td></tr>
2205 <tr><td><tt>sle</tt></td><td>var1 &lt;= var2</td><td>signed</td></tr>
2206 <tr><td><tt>sge</tt></td><td>var1 &gt;= var2</td><td>signed</td></tr>
2207 <tr><td><tt>ult</tt></td><td>var1 &lt; var2</td><td>unsigned</td></tr>
2208 <tr><td><tt>ugt</tt></td><td>var1 &gt; var2</td><td>unsigned</td></tr>
2209 <tr><td><tt>ule</tt></td><td>var1 &lt;= var2</td><td>unsigned</td></tr>
2210 <tr><td><tt>uge</tt></td><td>var1 &gt;= var2</td><td>unsigned</td></tr>
2211 <tr><td><tt>true</tt></td><td>always</td><td>--</td></tr>
2212 <tr><td><tt>false</tt></td><td>never</td><td>--</td></tr>
2213 </tbody>
2214</table>
2215
2216<h5>Example:</h5>
2217<pre> &lt;result&gt; = vsetint eq &lt;2 x int&gt; &lt;int 0, int 1&gt;, &lt;int 1, int 0&gt; <i>; yields {&lt;2 x bool&gt;}:result = false, false</i>
2218 &lt;result&gt; = vsetint ne &lt;2 x int&gt; &lt;int 0, int 1&gt;, &lt;int 1, int 0&gt; <i>; yields {&lt;2 x bool&gt;}:result = true, true</i>
2219 &lt;result&gt; = vsetint slt &lt;2 x int&gt; &lt;int 0, int 1&gt;, &lt;int 1, int 0&gt; <i>; yields {&lt;2 x bool&gt;}:result = true, false</i>
2220 &lt;result&gt; = vsetint sgt &lt;2 x int&gt; &lt;int 0, int 1&gt;, &lt;int 1, int 0&gt; <i>; yields {&lt;2 x bool&gt;}:result = false, true</i>
2221 &lt;result&gt; = vsetint sle &lt;2 x int&gt; &lt;int 0, int 1&gt;, &lt;int 1, int 0&gt; <i>; yields {&lt;2 x bool&gt;}:result = true, false</i>
2222 &lt;result&gt; = vsetint sge &lt;2 x int&gt; &lt;int 0, int 1&gt;, &lt;int 1, int 0&gt; <i>; yields {&lt;2 x bool&gt;}:result = false, true</i>
2223</pre>
2224</div>
2225
2226<!-- _______________________________________________________________________ -->
2227<div class="doc_subsubsection"> <a name="i_vsetfp">'<tt>vsetfp</tt>'
2228Instruction</a> </div>
2229<div class="doc_text">
2230<h5>Syntax:</h5>
2231<pre>&lt;result&gt; = vsetfp &lt;op&gt;, &lt;n x &lt;ty&gt;&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields &lt;n x bool&gt;</i>
2232</pre>
2233
2234<h5>Overview:</h5>
2235
2236<p>The '<tt>vsetfp</tt>' instruction takes two floating point vector
2237arguments and returns a vector of boolean values representing, at each
2238position, the result of the comparison between the values at that
2239position in the two operands.</p>
2240
2241<h5>Arguments:</h5>
2242
2243<p>The arguments to a '<tt>vsetfp</tt>' instruction are a comparison
2244operation and two value arguments. The value arguments must be of <a
2245href="t_floating">floating point</a> <a href="#t_packed">packed</a>
2246type, and they must have identical types. The operation argument must
2247be one of <tt>eq</tt>, <tt>ne</tt>, <tt>lt</tt>, <tt>gt</tt>,
2248<tt>le</tt>, <tt>ge</tt>, <tt>oeq</tt>, <tt>one</tt>, <tt>olt</tt>,
2249<tt>ogt</tt>, <tt>ole</tt>, <tt>oge</tt>, <tt>ueq</tt>, <tt>une</tt>,
2250<tt>ult</tt>, <tt>ugt</tt>, <tt>ule</tt>, <tt>uge</tt>, <tt>o</tt>,
2251<tt>u</tt>, <tt>true</tt>, and <tt>false</tt>. The result is a packed
2252<tt>bool</tt> value with the same length as each operand.</p>
2253
2254<h5>Semantics:</h5>
2255
2256<p>The following table shows the semantics of '<tt>vsetfp</tt>' for
2257floating point types. If either operand is a floating point Not a
2258Number (NaN) value, the operation is unordered, and the value in the
2259first column below is produced at that position. Otherwise, the
2260operation is ordered, and the value in the second column is
2261produced.</p>
2262
2263<table border="1" cellspacing="0" cellpadding="4">
2264 <tbody>
2265 <tr><th>Operation</th><th>If unordered<th>Otherwise true iff</th></tr>
2266 <tr><td><tt>eq</tt></td><td>undefined</td><td>var1 == var2</td></tr>
2267 <tr><td><tt>ne</tt></td><td>undefined</td><td>var1 != var2</td></tr>
2268 <tr><td><tt>lt</tt></td><td>undefined</td><td>var1 &lt; var2</td></tr>
2269 <tr><td><tt>gt</tt></td><td>undefined</td><td>var1 &gt; var2</td></tr>
2270 <tr><td><tt>le</tt></td><td>undefined</td><td>var1 &lt;= var2</td></tr>
2271 <tr><td><tt>ge</tt></td><td>undefined</td><td>var1 &gt;= var2</td></tr>
2272 <tr><td><tt>oeq</tt></td><td>false</td><td>var1 == var2</td></tr>
2273 <tr><td><tt>one</tt></td><td>false</td><td>var1 != var2</td></tr>
2274 <tr><td><tt>olt</tt></td><td>false</td><td>var1 &lt; var2</td></tr>
2275 <tr><td><tt>ogt</tt></td><td>false</td><td>var1 &gt; var2</td></tr>
2276 <tr><td><tt>ole</tt></td><td>false</td><td>var1 &lt;= var2</td></tr>
2277 <tr><td><tt>oge</tt></td><td>false</td><td>var1 &gt;= var2</td></tr>
2278 <tr><td><tt>ueq</tt></td><td>true</td><td>var1 == var2</td></tr>
2279 <tr><td><tt>une</tt></td><td>true</td><td>var1 != var2</td></tr>
2280 <tr><td><tt>ult</tt></td><td>true</td><td>var1 &lt; var2</td></tr>
2281 <tr><td><tt>ugt</tt></td><td>true</td><td>var1 &gt; var2</td></tr>
2282 <tr><td><tt>ule</tt></td><td>true</td><td>var1 &lt;= var2</td></tr>
2283 <tr><td><tt>uge</tt></td><td>true</td><td>var1 &gt;= var2</td></tr>
2284 <tr><td><tt>o</tt></td><td>false</td><td>always</td></tr>
2285 <tr><td><tt>u</tt></td><td>true</td><td>never</td></tr>
2286 <tr><td><tt>true</tt></td><td>true</td><td>always</td></tr>
2287 <tr><td><tt>false</tt></td><td>false</td><td>never</td></tr>
2288 </tbody>
2289</table>
2290
2291<h5>Example:</h5>
2292<pre> &lt;result&gt; = vsetfp eq &lt;2 x float&gt; &lt;float 0.0, float 1.0&gt;, &lt;float 1.0, float 0.0&gt; <i>; yields {&lt;2 x bool&gt;}:result = false, false</i>
2293 &lt;result&gt; = vsetfp ne &lt;2 x float&gt; &lt;float 0.0, float 1.0&gt;, &lt;float 1.0, float 0.0&gt; <i>; yields {&lt;2 x bool&gt;}:result = true, true</i>
2294 &lt;result&gt; = vsetfp lt &lt;2 x float&gt; &lt;float 0.0, float 1.0&gt;, &lt;float 1.0, float 0.0&gt; <i>; yields {&lt;2 x bool&gt;}:result = true, false</i>
2295 &lt;result&gt; = vsetfp gt &lt;2 x float&gt; &lt;float 0.0, float 1.0&gt;, &lt;float 1.0, float 0.0&gt; <i>; yields {&lt;2 x bool&gt;}:result = false, true</i>
2296 &lt;result&gt; = vsetfp le &lt;2 x float&gt; &lt;float 0.0, float 1.0&gt;, &lt;float 1.0, float 0.0&gt; <i>; yields {&lt;2 x bool&gt;}:result = true, false</i>
2297 &lt;result&gt; = vsetfp ge &lt;2 x float&gt; &lt;float 0.0, float 1.0&gt;, &lt;float 1.0, float 0.0&gt; <i>; yields {&lt;2 x bool&gt;}:result = false, true</i>
2298</pre>
2299</div>
2300
2301<!-- _______________________________________________________________________ -->
2302<div class="doc_subsubsection">
2303 <a name="i_vselect">'<tt>vselect</tt>' Instruction</a>
2304</div>
2305
2306<div class="doc_text">
2307
2308<h5>Syntax:</h5>
2309
2310<pre>
2311 &lt;result&gt; = vselect &lt;n x bool&gt; &lt;cond&gt;, &lt;n x &lt;ty&gt;&gt; &lt;val1&gt;, &lt;n x &lt;ty&gt;&gt; &lt;val2&gt; <i>; yields &lt;n x &lt;ty&gt;&gt;</i>
2312</pre>
2313
2314<h5>Overview:</h5>
2315
2316<p>
2317The '<tt>vselect</tt>' instruction chooses one value at each position
2318of a vector based on a condition.
2319</p>
2320
2321
2322<h5>Arguments:</h5>
2323
2324<p>
2325The '<tt>vselect</tt>' instruction requires a <a
2326href="#t_packed">packed</a> <tt>bool</tt> value indicating the
2327condition at each vector position, and two values of the same packed
2328type. All three operands must have the same length. The type of the
2329result is the same as the type of the two value operands.</p>
2330
2331<h5>Semantics:</h5>
2332
2333<p>
2334At each position where the <tt>bool</tt> vector is true, that position
2335of the result gets its value from the first value argument; otherwise,
2336it gets its value from the second value argument.
2337</p>
2338
2339<h5>Example:</h5>
2340
2341<pre>
2342 %X = vselect bool &lt;2 x bool&gt; &lt;bool true, bool false&gt;, &lt;2 x ubyte&gt; &lt;ubyte 17, ubyte 17&gt;,
2343 &lt;2 x ubyte&gt; &lt;ubyte 42, ubyte 42&gt; <i>; yields &lt;2 x ubyte&gt;:17, 42</i>
2344</pre>
2345</div>
2346
2347
2348
Chris Lattnerce83bff2006-04-08 23:07:04 +0000[diff] [blame]2349<!-- ======================================================================= -->
2350<div class="doc_subsection">
Chris Lattner6ab66722006-08-15 00:45:58 +0000[diff] [blame]2351 <a name="memoryops">Memory Access and Addressing Operations</a>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2352</div>
2353
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2354<div class="doc_text">
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2355
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2356<p>A key design point of an SSA-based representation is how it
2357represents memory. In LLVM, no memory locations are in SSA form, which
2358makes things very simple. This section describes how to read, write,
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]2359allocate, and free memory in LLVM.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2360
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2361</div>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2362
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2363<!-- _______________________________________________________________________ -->
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2364<div class="doc_subsubsection">
2365 <a name="i_malloc">'<tt>malloc</tt>' Instruction</a>
2366</div>
2367
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2368<div class="doc_text">
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2369
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2370<h5>Syntax:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2371
2372<pre>
2373 &lt;result&gt; = malloc &lt;type&gt;[, uint &lt;NumElements&gt;][, align &lt;alignment&gt;] <i>; yields {type*}:result</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2374</pre>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2375
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2376<h5>Overview:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2377
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2378<p>The '<tt>malloc</tt>' instruction allocates memory from the system
2379heap and returns a pointer to it.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2380
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2381<h5>Arguments:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2382
2383<p>The '<tt>malloc</tt>' instruction allocates
2384<tt>sizeof(&lt;type&gt;)*NumElements</tt>
John Criswella92e5862004-02-24 16:13:56 +0000[diff] [blame]2385bytes of memory from the operating system and returns a pointer of the
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2386appropriate type to the program. If "NumElements" is specified, it is the
2387number of elements allocated. If an alignment is specified, the value result
2388of the allocation is guaranteed to be aligned to at least that boundary. If
2389not specified, or if zero, the target can choose to align the allocation on any
2390convenient boundary.</p>
2391
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2392<p>'<tt>type</tt>' must be a sized type.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2393
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2394<h5>Semantics:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2395
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2396<p>Memory is allocated using the system "<tt>malloc</tt>" function, and
2397a pointer is returned.</p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2398
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2399<h5>Example:</h5>
2400
2401<pre>
2402 %array = malloc [4 x ubyte ] <i>; yields {[%4 x ubyte]*}:array</i>
2403
2404 %size = <a href="#i_add">add</a> uint 2, 2 <i>; yields {uint}:size = uint 4</i>
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]2405 %array1 = malloc ubyte, uint 4 <i>; yields {ubyte*}:array1</i>
2406 %array2 = malloc [12 x ubyte], uint %size <i>; yields {[12 x ubyte]*}:array2</i>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2407 %array3 = malloc int, uint 4, align 1024 <i>; yields {int*}:array3</i>
2408 %array4 = malloc int, align 1024 <i>; yields {int*}:array4</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2409</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2410</div>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2411
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2412<!-- _______________________________________________________________________ -->
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2413<div class="doc_subsubsection">
2414 <a name="i_free">'<tt>free</tt>' Instruction</a>
2415</div>
2416
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2417<div class="doc_text">
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2418
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2419<h5>Syntax:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2420
2421<pre>
2422 free &lt;type&gt; &lt;value&gt; <i>; yields {void}</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2423</pre>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2424
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2425<h5>Overview:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2426
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2427<p>The '<tt>free</tt>' instruction returns memory back to the unused
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]2428memory heap to be reallocated in the future.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2429
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2430<h5>Arguments:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2431
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2432<p>'<tt>value</tt>' shall be a pointer value that points to a value
2433that was allocated with the '<tt><a href="#i_malloc">malloc</a></tt>'
2434instruction.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2435
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2436<h5>Semantics:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2437
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]2438<p>Access to the memory pointed to by the pointer is no longer defined
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2439after this instruction executes.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2440
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2441<h5>Example:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2442
2443<pre>
2444 %array = <a href="#i_malloc">malloc</a> [4 x ubyte] <i>; yields {[4 x ubyte]*}:array</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2445 free [4 x ubyte]* %array
2446</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2447</div>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2448
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2449<!-- _______________________________________________________________________ -->
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2450<div class="doc_subsubsection">
2451 <a name="i_alloca">'<tt>alloca</tt>' Instruction</a>
2452</div>
2453
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2454<div class="doc_text">
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2455
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2456<h5>Syntax:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2457
2458<pre>
2459 &lt;result&gt; = alloca &lt;type&gt;[, uint &lt;NumElements&gt;][, align &lt;alignment&gt;] <i>; yields {type*}:result</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2460</pre>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2461
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2462<h5>Overview:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2463
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2464<p>The '<tt>alloca</tt>' instruction allocates memory on the current
2465stack frame of the procedure that is live until the current function
2466returns to its caller.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2467
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2468<h5>Arguments:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2469
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]2470<p>The '<tt>alloca</tt>' instruction allocates <tt>sizeof(&lt;type&gt;)*NumElements</tt>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2471bytes of memory on the runtime stack, returning a pointer of the
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2472appropriate type to the program. If "NumElements" is specified, it is the
2473number of elements allocated. If an alignment is specified, the value result
2474of the allocation is guaranteed to be aligned to at least that boundary. If
2475not specified, or if zero, the target can choose to align the allocation on any
2476convenient boundary.</p>
2477
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2478<p>'<tt>type</tt>' may be any sized type.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2479
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2480<h5>Semantics:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2481
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]2482<p>Memory is allocated; a pointer is returned. '<tt>alloca</tt>'d
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2483memory is automatically released when the function returns. The '<tt>alloca</tt>'
2484instruction is commonly used to represent automatic variables that must
2485have an address available. When the function returns (either with the <tt><a
John Criswellc932bef2005-05-12 16:55:34 +0000[diff] [blame]2486 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]2487instructions), the memory is reclaimed.</p>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2488
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2489<h5>Example:</h5>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2490
2491<pre>
2492 %ptr = alloca int <i>; yields {int*}:ptr</i>
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]2493 %ptr = alloca int, uint 4 <i>; yields {int*}:ptr</i>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2494 %ptr = alloca int, uint 4, align 1024 <i>; yields {int*}:ptr</i>
2495 %ptr = alloca int, align 1024 <i>; yields {int*}:ptr</i>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2496</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2497</div>
Chris Lattner54611b42005-11-06 08:02:57 +0000[diff] [blame]2498
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2499<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2500<div class="doc_subsubsection"> <a name="i_load">'<tt>load</tt>'
2501Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2502<div class="doc_text">
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2503<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2504<pre> &lt;result&gt; = load &lt;ty&gt;* &lt;pointer&gt;<br> &lt;result&gt; = volatile load &lt;ty&gt;* &lt;pointer&gt;<br></pre>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2505<h5>Overview:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2506<p>The '<tt>load</tt>' instruction is used to read from memory.</p>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2507<h5>Arguments:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2508<p>The argument to the '<tt>load</tt>' instruction specifies the memory
John Criswell4c0cf7f2005-10-24 16:17:18 +0000[diff] [blame]2509address from which to load. The pointer must point to a <a
Chris Lattner10ee9652004-06-03 22:57:15 +0000[diff] [blame]2510 href="#t_firstclass">first class</a> type. If the <tt>load</tt> is
John Criswell4c0cf7f2005-10-24 16:17:18 +0000[diff] [blame]2511marked as <tt>volatile</tt>, then the optimizer is not allowed to modify
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2512the number or order of execution of this <tt>load</tt> with other
2513volatile <tt>load</tt> and <tt><a href="#i_store">store</a></tt>
2514instructions. </p>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2515<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2516<p>The location of memory pointed to is loaded.</p>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2517<h5>Examples:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2518<pre> %ptr = <a href="#i_alloca">alloca</a> int <i>; yields {int*}:ptr</i>
2519 <a
2520 href="#i_store">store</a> int 3, int* %ptr <i>; yields {void}</i>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2521 %val = load int* %ptr <i>; yields {int}:val = int 3</i>
2522</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2523</div>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2524<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2525<div class="doc_subsubsection"> <a name="i_store">'<tt>store</tt>'
2526Instruction</a> </div>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2527<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2528<pre> store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt; <i>; yields {void}</i>
Chris Lattner12d456c2003-09-08 18:27:49 +0000[diff] [blame]2529 volatile store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt; <i>; yields {void}</i>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2530</pre>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2531<h5>Overview:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2532<p>The '<tt>store</tt>' instruction is used to write to memory.</p>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2533<h5>Arguments:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2534<p>There are two arguments to the '<tt>store</tt>' instruction: a value
John Criswell4c0cf7f2005-10-24 16:17:18 +0000[diff] [blame]2535to store and an address in which to store it. The type of the '<tt>&lt;pointer&gt;</tt>'
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2536operand must be a pointer to the type of the '<tt>&lt;value&gt;</tt>'
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]2537operand. If the <tt>store</tt> is marked as <tt>volatile</tt>, then the
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2538optimizer is not allowed to modify the number or order of execution of
2539this <tt>store</tt> with other volatile <tt>load</tt> and <tt><a
2540 href="#i_store">store</a></tt> instructions.</p>
2541<h5>Semantics:</h5>
2542<p>The contents of memory are updated to contain '<tt>&lt;value&gt;</tt>'
2543at the location specified by the '<tt>&lt;pointer&gt;</tt>' operand.</p>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2544<h5>Example:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2545<pre> %ptr = <a href="#i_alloca">alloca</a> int <i>; yields {int*}:ptr</i>
2546 <a
2547 href="#i_store">store</a> int 3, int* %ptr <i>; yields {void}</i>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2548 %val = load int* %ptr <i>; yields {int}:val = int 3</i>
2549</pre>
Chris Lattner095735d2002-05-06 03:03:22 +0000[diff] [blame]2550<!-- _______________________________________________________________________ -->
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2551<div class="doc_subsubsection">
2552 <a name="i_getelementptr">'<tt>getelementptr</tt>' Instruction</a>
2553</div>
2554
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2555<div class="doc_text">
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]2556<h5>Syntax:</h5>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2557<pre>
2558 &lt;result&gt; = getelementptr &lt;ty&gt;* &lt;ptrval&gt;{, &lt;ty&gt; &lt;idx&gt;}*
2559</pre>
2560
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]2561<h5>Overview:</h5>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2562
2563<p>
2564The '<tt>getelementptr</tt>' instruction is used to get the address of a
2565subelement of an aggregate data structure.</p>
2566
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]2567<h5>Arguments:</h5>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2568
2569<p>This instruction takes a list of integer constants that indicate what
2570elements of the aggregate object to index to. The actual types of the arguments
2571provided depend on the type of the first pointer argument. The
2572'<tt>getelementptr</tt>' instruction is used to index down through the type
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]2573levels of a structure or to a specific index in an array. When indexing into a
2574structure, only <tt>uint</tt>
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]2575integer constants are allowed. When indexing into an array or pointer,
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2576<tt>int</tt> and <tt>long</tt> indexes are allowed of any sign.</p>
2577
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2578<p>For example, let's consider a C code fragment and how it gets
2579compiled to LLVM:</p>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2580
2581<pre>
2582 struct RT {
2583 char A;
2584 int B[10][20];
2585 char C;
2586 };
2587 struct ST {
2588 int X;
2589 double Y;
2590 struct RT Z;
2591 };
2592
2593 int *foo(struct ST *s) {
2594 return &amp;s[1].Z.B[5][13];
2595 }
2596</pre>
2597
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2598<p>The LLVM code generated by the GCC frontend is:</p>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2599
2600<pre>
2601 %RT = type { sbyte, [10 x [20 x int]], sbyte }
2602 %ST = type { int, double, %RT }
2603
Brian Gaeke317ef962004-07-02 21:08:14 +0000[diff] [blame]2604 implementation
2605
2606 int* %foo(%ST* %s) {
2607 entry:
2608 %reg = getelementptr %ST* %s, int 1, uint 2, uint 1, int 5, int 13
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2609 ret int* %reg
2610 }
2611</pre>
2612
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]2613<h5>Semantics:</h5>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2614
2615<p>The index types specified for the '<tt>getelementptr</tt>' instruction depend
John Criswell4a3327e2005-05-13 22:25:59 +0000[diff] [blame]2616on the pointer type that is being indexed into. <a href="#t_pointer">Pointer</a>
Chris Lattner10ee9652004-06-03 22:57:15 +0000[diff] [blame]2617and <a href="#t_array">array</a> types require <tt>uint</tt>, <tt>int</tt>,
2618<tt>ulong</tt>, or <tt>long</tt> values, and <a href="#t_struct">structure</a>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2619types require <tt>uint</tt> <b>constants</b>.</p>
2620
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2621<p>In the example above, the first index is indexing into the '<tt>%ST*</tt>'
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2622type, which is a pointer, yielding a '<tt>%ST</tt>' = '<tt>{ int, double, %RT
2623}</tt>' type, a structure. The second index indexes into the third element of
2624the structure, yielding a '<tt>%RT</tt>' = '<tt>{ sbyte, [10 x [20 x int]],
2625sbyte }</tt>' type, another structure. The third index indexes into the second
2626element of the structure, yielding a '<tt>[10 x [20 x int]]</tt>' type, an
2627array. The two dimensions of the array are subscripted into, yielding an
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]2628'<tt>int</tt>' type. The '<tt>getelementptr</tt>' instruction returns a pointer
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2629to this element, thus computing a value of '<tt>int*</tt>' type.</p>
2630
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2631<p>Note that it is perfectly legal to index partially through a
2632structure, returning a pointer to an inner element. Because of this,
2633the LLVM code for the given testcase is equivalent to:</p>
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2634
2635<pre>
Chris Lattner455fc8c2005-03-07 22:13:59 +0000[diff] [blame]2636 int* %foo(%ST* %s) {
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2637 %t1 = getelementptr %ST* %s, int 1 <i>; yields %ST*:%t1</i>
2638 %t2 = getelementptr %ST* %t1, int 0, uint 2 <i>; yields %RT*:%t2</i>
2639 %t3 = getelementptr %RT* %t2, int 0, uint 1 <i>; yields [10 x [20 x int]]*:%t3</i>
2640 %t4 = getelementptr [10 x [20 x int]]* %t3, int 0, int 5 <i>; yields [20 x int]*:%t4</i>
2641 %t5 = getelementptr [20 x int]* %t4, int 0, int 13 <i>; yields int*:%t5</i>
2642 ret int* %t5
2643 }
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]2644</pre>
Chris Lattnerc0ad71e2005-06-24 17:22:57 +0000[diff] [blame]2645
2646<p>Note that it is undefined to access an array out of bounds: array and
2647pointer indexes must always be within the defined bounds of the array type.
2648The one exception for this rules is zero length arrays. These arrays are
2649defined to be accessible as variable length arrays, which requires access
2650beyond the zero'th element.</p>
2651
Chris Lattner6ab66722006-08-15 00:45:58 +0000[diff] [blame]2652<p>The getelementptr instruction is often confusing. For some more insight
2653into how it works, see <a href="GetElementPtr.html">the getelementptr
2654FAQ</a>.</p>
2655
Chris Lattner590645f2002-04-14 06:13:44 +0000[diff] [blame]2656<h5>Example:</h5>
Chris Lattnerc0ad71e2005-06-24 17:22:57 +0000[diff] [blame]2657
Chris Lattner33fd7022004-04-05 01:30:49 +0000[diff] [blame]2658<pre>
2659 <i>; yields [12 x ubyte]*:aptr</i>
2660 %aptr = getelementptr {int, [12 x ubyte]}* %sptr, long 0, uint 1
2661</pre>
2662
2663</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2664<!-- ======================================================================= -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2665<div class="doc_subsection"> <a name="otherops">Other Operations</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2666<div class="doc_text">
John Criswell417228d2004-04-09 16:48:45 +0000[diff] [blame]2667<p>The instructions in this category are the "miscellaneous"
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2668instructions, which defy better classification.</p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2669</div>
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2670<!-- _______________________________________________________________________ -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2671<div class="doc_subsubsection"> <a name="i_phi">'<tt>phi</tt>'
2672Instruction</a> </div>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2673<div class="doc_text">
Chris Lattner70de6632001-07-09 00:26:23 +0000[diff] [blame]2674<h5>Syntax:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2675<pre> &lt;result&gt; = phi &lt;ty&gt; [ &lt;val0&gt;, &lt;label0&gt;], ...<br></pre>
Chris Lattner70de6632001-07-09 00:26:23 +0000[diff] [blame]2676<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2677<p>The '<tt>phi</tt>' instruction is used to implement the &#966; node in
2678the SSA graph representing the function.</p>
Chris Lattner70de6632001-07-09 00:26:23 +0000[diff] [blame]2679<h5>Arguments:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2680<p>The type of the incoming values are specified with the first type
2681field. After this, the '<tt>phi</tt>' instruction takes a list of pairs
2682as arguments, with one pair for each predecessor basic block of the
2683current block. Only values of <a href="#t_firstclass">first class</a>
2684type may be used as the value arguments to the PHI node. Only labels
2685may be used as the label arguments.</p>
2686<p>There must be no non-phi instructions between the start of a basic
2687block and the PHI instructions: i.e. PHI instructions must be first in
2688a basic block.</p>
Chris Lattner70de6632001-07-09 00:26:23 +0000[diff] [blame]2689<h5>Semantics:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2690<p>At runtime, the '<tt>phi</tt>' instruction logically takes on the
2691value specified by the parameter, depending on which basic block we
2692came from in the last <a href="#terminators">terminator</a> instruction.</p>
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]2693<h5>Example:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2694<pre>Loop: ; Infinite loop that counts from 0 on up...<br> %indvar = phi uint [ 0, %LoopHeader ], [ %nextindvar, %Loop ]<br> %nextindvar = add uint %indvar, 1<br> br label %Loop<br></pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2695</div>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]2696
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]2697<!-- _______________________________________________________________________ -->
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]2698<div class="doc_subsubsection">
2699 <a name="i_cast">'<tt>cast .. to</tt>' Instruction</a>
2700</div>
2701
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2702<div class="doc_text">
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]2703
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]2704<h5>Syntax:</h5>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]2705
2706<pre>
2707 &lt;result&gt; = cast &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]2708</pre>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]2709
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]2710<h5>Overview:</h5>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]2711
2712<p>
2713The '<tt>cast</tt>' instruction is used as the primitive means to convert
2714integers to floating point, change data type sizes, and break type safety (by
2715casting pointers).
2716</p>
2717
2718
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]2719<h5>Arguments:</h5>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]2720
2721<p>
2722The '<tt>cast</tt>' instruction takes a value to cast, which must be a first
2723class value, and a type to cast it to, which must also be a <a
2724href="#t_firstclass">first class</a> type.
2725</p>
2726
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]2727<h5>Semantics:</h5>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]2728
2729<p>
2730This instruction follows the C rules for explicit casts when determining how the
2731data being cast must change to fit in its new container.
2732</p>
2733
2734<p>
2735When casting to bool, any value that would be considered true in the context of
2736a C '<tt>if</tt>' condition is converted to the boolean '<tt>true</tt>' values,
2737all else are '<tt>false</tt>'.
2738</p>
2739
2740<p>
2741When extending an integral value from a type of one signness to another (for
2742example '<tt>sbyte</tt>' to '<tt>ulong</tt>'), the value is sign-extended if the
2743<b>source</b> value is signed, and zero-extended if the source value is
2744unsigned. <tt>bool</tt> values are always zero extended into either zero or
2745one.
2746</p>
2747
Chris Lattner70de6632001-07-09 00:26:23 +0000[diff] [blame]2748<h5>Example:</h5>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]2749
2750<pre>
2751 %X = cast int 257 to ubyte <i>; yields ubyte:1</i>
Chris Lattnerd8f8ede2002-06-25 18:03:17 +0000[diff] [blame]2752 %Y = cast int 123 to bool <i>; yields bool:true</i>
Chris Lattner70de6632001-07-09 00:26:23 +0000[diff] [blame]2753</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2754</div>
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]2755
2756<!-- _______________________________________________________________________ -->
2757<div class="doc_subsubsection">
2758 <a name="i_select">'<tt>select</tt>' Instruction</a>
2759</div>
2760
2761<div class="doc_text">
2762
2763<h5>Syntax:</h5>
2764
2765<pre>
2766 &lt;result&gt; = select bool &lt;cond&gt;, &lt;ty&gt; &lt;val1&gt;, &lt;ty&gt; &lt;val2&gt; <i>; yields ty</i>
2767</pre>
2768
2769<h5>Overview:</h5>
2770
2771<p>
2772The '<tt>select</tt>' instruction is used to choose one value based on a
2773condition, without branching.
2774</p>
2775
2776
2777<h5>Arguments:</h5>
2778
2779<p>
2780The '<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.
2781</p>
2782
2783<h5>Semantics:</h5>
2784
2785<p>
2786If the boolean condition evaluates to true, the instruction returns the first
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]2787value argument; otherwise, it returns the second value argument.
Chris Lattnerb53c28d2004-03-12 05:50:16 +0000[diff] [blame]2788</p>
2789
2790<h5>Example:</h5>
2791
2792<pre>
2793 %X = select bool true, ubyte 17, ubyte 42 <i>; yields ubyte:17</i>
2794</pre>
2795</div>
2796
Robert Bocchinof72fdfe2006-01-15 20:48:27 +0000[diff] [blame]2797
2798<!-- _______________________________________________________________________ -->
2799<div class="doc_subsubsection">
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]2800 <a name="i_call">'<tt>call</tt>' Instruction</a>
2801</div>
2802
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2803<div class="doc_text">
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]2804
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2805<h5>Syntax:</h5>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]2806<pre>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]2807 &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]2808</pre>
2809
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2810<h5>Overview:</h5>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]2811
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2812<p>The '<tt>call</tt>' instruction represents a simple function call.</p>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]2813
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2814<h5>Arguments:</h5>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]2815
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2816<p>This instruction requires several arguments:</p>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]2817
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]2818<ol>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2819 <li>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]2820 <p>The optional "tail" marker indicates whether the callee function accesses
2821 any allocas or varargs in the caller. If the "tail" marker is present, the
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]2822 function call is eligible for tail call optimization. Note that calls may
2823 be marked "tail" even if they do not occur before a <a
2824 href="#i_ret"><tt>ret</tt></a> instruction.
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2825 </li>
2826 <li>
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]2827 <p>The optional "cconv" marker indicates which <a href="callingconv">calling
2828 convention</a> the call should use. If none is specified, the call defaults
2829 to using C calling conventions.
2830 </li>
2831 <li>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]2832 <p>'<tt>ty</tt>': shall be the signature of the pointer to function value
2833 being invoked. The argument types must match the types implied by this
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]2834 signature. This type can be omitted if the function is not varargs and
2835 if the function type does not return a pointer to a function.</p>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]2836 </li>
2837 <li>
2838 <p>'<tt>fnptrval</tt>': An LLVM value containing a pointer to a function to
2839 be invoked. In most cases, this is a direct function invocation, but
2840 indirect <tt>call</tt>s are just as possible, calling an arbitrary pointer
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]2841 to function value.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2842 </li>
2843 <li>
2844 <p>'<tt>function args</tt>': argument list whose types match the
Reid Spencerd845d162005-05-01 22:22:57 +0000[diff] [blame]2845 function signature argument types. All arguments must be of
2846 <a href="#t_firstclass">first class</a> type. If the function signature
2847 indicates the function accepts a variable number of arguments, the extra
2848 arguments can be specified.</p>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2849 </li>
Chris Lattnera8292f32002-05-06 22:08:29 +0000[diff] [blame]2850</ol>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]2851
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2852<h5>Semantics:</h5>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]2853
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2854<p>The '<tt>call</tt>' instruction is used to cause control flow to
2855transfer to a specified function, with its incoming arguments bound to
2856the specified values. Upon a '<tt><a href="#i_ret">ret</a></tt>'
2857instruction in the called function, control flow continues with the
2858instruction after the function call, and the return value of the
2859function is bound to the result argument. This is a simpler case of
2860the <a href="#i_invoke">invoke</a> instruction.</p>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]2861
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]2862<h5>Example:</h5>
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]2863
2864<pre>
2865 %retval = call int %test(int %argc)
2866 call int(sbyte*, ...) *%printf(sbyte* %msg, int 12, sbyte 42);
2867 %X = tail call int %foo()
Chris Lattner0132aff2005-05-06 22:57:40 +0000[diff] [blame]2868 %Y = tail call <a href="#callingconv">fastcc</a> int %foo()
Chris Lattnere23c1392005-05-06 05:47:36 +0000[diff] [blame]2869</pre>
2870
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2871</div>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]2872
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]2873<!-- _______________________________________________________________________ -->
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]2874<div class="doc_subsubsection">
Chris Lattner33337472006-01-13 23:26:01 +0000[diff] [blame]2875 <a name="i_va_arg">'<tt>va_arg</tt>' Instruction</a>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]2876</div>
2877
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2878<div class="doc_text">
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]2879
Chris Lattner26ca62e2003-10-18 05:51:36 +0000[diff] [blame]2880<h5>Syntax:</h5>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]2881
2882<pre>
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]2883 &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]2884</pre>
2885
Chris Lattner26ca62e2003-10-18 05:51:36 +0000[diff] [blame]2886<h5>Overview:</h5>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]2887
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]2888<p>The '<tt>va_arg</tt>' instruction is used to access arguments passed through
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]2889the "variable argument" area of a function call. It is used to implement the
2890<tt>va_arg</tt> macro in C.</p>
2891
Chris Lattner26ca62e2003-10-18 05:51:36 +0000[diff] [blame]2892<h5>Arguments:</h5>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]2893
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]2894<p>This instruction takes a <tt>va_list*</tt> value and the type of
2895the argument. It returns a value of the specified argument type and
Jeff Cohendc6bfea2005-11-11 02:15:27 +0000[diff] [blame]2896increments the <tt>va_list</tt> to point to the next argument. Again, the
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]2897actual type of <tt>va_list</tt> is target specific.</p>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]2898
Chris Lattner26ca62e2003-10-18 05:51:36 +0000[diff] [blame]2899<h5>Semantics:</h5>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]2900
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]2901<p>The '<tt>va_arg</tt>' instruction loads an argument of the specified
2902type from the specified <tt>va_list</tt> and causes the
2903<tt>va_list</tt> to point to the next argument. For more information,
2904see the variable argument handling <a href="#int_varargs">Intrinsic
2905Functions</a>.</p>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]2906
2907<p>It is legal for this instruction to be called in a function which does not
2908take a variable number of arguments, for example, the <tt>vfprintf</tt>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2909function.</p>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]2910
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]2911<p><tt>va_arg</tt> is an LLVM instruction instead of an <a
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]2912href="#intrinsics">intrinsic function</a> because it takes a type as an
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]2913argument.</p>
2914
Chris Lattner26ca62e2003-10-18 05:51:36 +0000[diff] [blame]2915<h5>Example:</h5>
Chris Lattner6a4a0492004-09-27 21:51:25 +0000[diff] [blame]2916
2917<p>See the <a href="#int_varargs">variable argument processing</a> section.</p>
2918
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2919</div>
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]2920
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]2921<!-- *********************************************************************** -->
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2922<div class="doc_section"> <a name="intrinsics">Intrinsic Functions</a> </div>
2923<!-- *********************************************************************** -->
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]2924
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2925<div class="doc_text">
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]2926
2927<p>LLVM supports the notion of an "intrinsic function". These functions have
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]2928well known names and semantics and are required to follow certain
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]2929restrictions. Overall, these instructions represent an extension mechanism for
2930the LLVM language that does not require changing all of the transformations in
2931LLVM to add to the language (or the bytecode reader/writer, the parser,
2932etc...).</p>
2933
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]2934<p>Intrinsic function names must all start with an "<tt>llvm.</tt>" prefix. This
2935prefix is reserved in LLVM for intrinsic names; thus, functions may not be named
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]2936this. Intrinsic functions must always be external functions: you cannot define
2937the body of intrinsic functions. Intrinsic functions may only be used in call
2938or invoke instructions: it is illegal to take the address of an intrinsic
2939function. Additionally, because intrinsic functions are part of the LLVM
2940language, it is required that they all be documented here if any are added.</p>
2941
2942
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]2943<p>To learn how to add an intrinsic function, please see the <a
Chris Lattner90391c12005-05-11 03:35:57 +0000[diff] [blame]2944href="ExtendingLLVM.html">Extending LLVM Guide</a>.
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]2945</p>
2946
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2947</div>
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]2948
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]2949<!-- ======================================================================= -->
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]2950<div class="doc_subsection">
2951 <a name="int_varargs">Variable Argument Handling Intrinsics</a>
2952</div>
2953
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2954<div class="doc_text">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]2955
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2956<p>Variable argument support is defined in LLVM with the <a
Chris Lattner33337472006-01-13 23:26:01 +0000[diff] [blame]2957 href="#i_va_arg"><tt>va_arg</tt></a> instruction and these three
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2958intrinsic functions. These functions are related to the similarly
2959named macros defined in the <tt>&lt;stdarg.h&gt;</tt> header file.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]2960
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]2961<p>All of these functions operate on arguments that use a
2962target-specific value type "<tt>va_list</tt>". The LLVM assembly
2963language reference manual does not define what this type is, so all
2964transformations should be prepared to handle intrinsics with any type
2965used.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]2966
Chris Lattner30b868d2006-05-15 17:26:46 +0000[diff] [blame]2967<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]2968instruction and the variable argument handling intrinsic functions are
2969used.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]2970
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]2971<pre>
2972int %test(int %X, ...) {
2973 ; Initialize variable argument processing
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]2974 %ap = alloca sbyte*
2975 call void %<a href="#i_va_start">llvm.va_start</a>(sbyte** %ap)
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]2976
2977 ; Read a single integer argument
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]2978 %tmp = va_arg sbyte** %ap, int
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]2979
2980 ; Demonstrate usage of llvm.va_copy and llvm.va_end
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]2981 %aq = alloca sbyte*
Andrew Lenharth5305ea52005-06-22 20:38:11 +0000[diff] [blame]2982 call void %<a href="#i_va_copy">llvm.va_copy</a>(sbyte** %aq, sbyte** %ap)
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]2983 call void %<a href="#i_va_end">llvm.va_end</a>(sbyte** %aq)
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]2984
2985 ; Stop processing of arguments.
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]2986 call void %<a href="#i_va_end">llvm.va_end</a>(sbyte** %ap)
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]2987 ret int %tmp
2988}
2989</pre>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2990</div>
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]2991
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]2992<!-- _______________________________________________________________________ -->
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]2993<div class="doc_subsubsection">
2994 <a name="i_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a>
2995</div>
2996
2997
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]2998<div class="doc_text">
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]2999<h5>Syntax:</h5>
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3000<pre> declare void %llvm.va_start(&lt;va_list&gt;* &lt;arglist&gt;)<br></pre>
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3001<h5>Overview:</h5>
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3002<P>The '<tt>llvm.va_start</tt>' intrinsic initializes
3003<tt>*&lt;arglist&gt;</tt> for subsequent use by <tt><a
3004href="#i_va_arg">va_arg</a></tt>.</p>
3005
3006<h5>Arguments:</h5>
3007
3008<P>The argument is a pointer to a <tt>va_list</tt> element to initialize.</p>
3009
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3010<h5>Semantics:</h5>
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3011
3012<P>The '<tt>llvm.va_start</tt>' intrinsic works just like the <tt>va_start</tt>
3013macro available in C. In a target-dependent way, it initializes the
3014<tt>va_list</tt> element the argument points to, so that the next call to
3015<tt>va_arg</tt> will produce the first variable argument passed to the function.
3016Unlike the C <tt>va_start</tt> macro, this intrinsic does not need to know the
3017last argument of the function, the compiler can figure that out.</p>
3018
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3019</div>
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3020
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3021<!-- _______________________________________________________________________ -->
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3022<div class="doc_subsubsection">
3023 <a name="i_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a>
3024</div>
3025
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3026<div class="doc_text">
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3027<h5>Syntax:</h5>
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3028<pre> declare void %llvm.va_end(&lt;va_list*&gt; &lt;arglist&gt;)<br></pre>
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3029<h5>Overview:</h5>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3030<p>The '<tt>llvm.va_end</tt>' intrinsic destroys <tt>&lt;arglist&gt;</tt>
3031which has been initialized previously with <tt><a href="#i_va_start">llvm.va_start</a></tt>
3032or <tt><a href="#i_va_copy">llvm.va_copy</a></tt>.</p>
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3033<h5>Arguments:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3034<p>The argument is a <tt>va_list</tt> to destroy.</p>
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3035<h5>Semantics:</h5>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3036<p>The '<tt>llvm.va_end</tt>' intrinsic works just like the <tt>va_end</tt>
Chris Lattner48b383b02003-11-25 01:02:51 +0000[diff] [blame]3037macro available in C. In a target-dependent way, it destroys the <tt>va_list</tt>.
3038Calls to <a href="#i_va_start"><tt>llvm.va_start</tt></a> and <a
3039 href="#i_va_copy"><tt>llvm.va_copy</tt></a> must be matched exactly
3040with calls to <tt>llvm.va_end</tt>.</p>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3041</div>
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3042
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3043<!-- _______________________________________________________________________ -->
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3044<div class="doc_subsubsection">
3045 <a name="i_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a>
3046</div>
3047
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3048<div class="doc_text">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3049
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3050<h5>Syntax:</h5>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3051
3052<pre>
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3053 declare void %llvm.va_copy(&lt;va_list&gt;* &lt;destarglist&gt;,
Andrew Lenharth5305ea52005-06-22 20:38:11 +0000[diff] [blame]3054 &lt;va_list&gt;* &lt;srcarglist&gt;)
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3055</pre>
3056
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3057<h5>Overview:</h5>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3058
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3059<p>The '<tt>llvm.va_copy</tt>' intrinsic copies the current argument position from
3060the source argument list to the destination argument list.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3061
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3062<h5>Arguments:</h5>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3063
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3064<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]3065The 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]3066
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3067
Chris Lattnerbd64b4e2003-05-08 04:57:36 +0000[diff] [blame]3068<h5>Semantics:</h5>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3069
Andrew Lenharth5fb787c2005-06-18 18:28:17 +0000[diff] [blame]3070<p>The '<tt>llvm.va_copy</tt>' intrinsic works just like the <tt>va_copy</tt> macro
3071available in C. In a target-dependent way, it copies the source
3072<tt>va_list</tt> element into the destination list. This intrinsic is necessary
3073because the <tt><a href="i_va_begin">llvm.va_begin</a></tt> intrinsic may be
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3074arbitrarily complex and require memory allocation, for example.</p>
3075
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3076</div>
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3077
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3078<!-- ======================================================================= -->
3079<div class="doc_subsection">
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3080 <a name="int_gc">Accurate Garbage Collection Intrinsics</a>
3081</div>
3082
3083<div class="doc_text">
3084
3085<p>
3086LLVM support for <a href="GarbageCollection.html">Accurate Garbage
3087Collection</a> requires the implementation and generation of these intrinsics.
3088These intrinsics allow identification of <a href="#i_gcroot">GC roots on the
3089stack</a>, as well as garbage collector implementations that require <a
3090href="#i_gcread">read</a> and <a href="#i_gcwrite">write</a> barriers.
3091Front-ends for type-safe garbage collected languages should generate these
3092intrinsics to make use of the LLVM garbage collectors. For more details, see <a
3093href="GarbageCollection.html">Accurate Garbage Collection with LLVM</a>.
3094</p>
3095</div>
3096
3097<!-- _______________________________________________________________________ -->
3098<div class="doc_subsubsection">
3099 <a name="i_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a>
3100</div>
3101
3102<div class="doc_text">
3103
3104<h5>Syntax:</h5>
3105
3106<pre>
Reid Spencer7821d062005-04-26 20:50:44 +0000[diff] [blame]3107 declare void %llvm.gcroot(&lt;ty&gt;** %ptrloc, &lt;ty2&gt;* %metadata)
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3108</pre>
3109
3110<h5>Overview:</h5>
3111
John Criswelldfe6a862004-12-10 15:51:16 +0000[diff] [blame]3112<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]3113the code generator, and allows some metadata to be associated with it.</p>
3114
3115<h5>Arguments:</h5>
3116
3117<p>The first argument specifies the address of a stack object that contains the
3118root pointer. The second pointer (which must be either a constant or a global
3119value address) contains the meta-data to be associated with the root.</p>
3120
3121<h5>Semantics:</h5>
3122
3123<p>At runtime, a call to this intrinsics stores a null pointer into the "ptrloc"
3124location. At compile-time, the code generator generates information to allow
3125the runtime to find the pointer at GC safe points.
3126</p>
3127
3128</div>
3129
3130
3131<!-- _______________________________________________________________________ -->
3132<div class="doc_subsubsection">
3133 <a name="i_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a>
3134</div>
3135
3136<div class="doc_text">
3137
3138<h5>Syntax:</h5>
3139
3140<pre>
Chris Lattnerf9228072006-03-14 20:02:51 +0000[diff] [blame]3141 declare sbyte* %llvm.gcread(sbyte* %ObjPtr, sbyte** %Ptr)
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3142</pre>
3143
3144<h5>Overview:</h5>
3145
3146<p>The '<tt>llvm.gcread</tt>' intrinsic identifies reads of references from heap
3147locations, allowing garbage collector implementations that require read
3148barriers.</p>
3149
3150<h5>Arguments:</h5>
3151
Chris Lattnerf9228072006-03-14 20:02:51 +0000[diff] [blame]3152<p>The second argument is the address to read from, which should be an address
3153allocated from the garbage collector. The first object is a pointer to the
3154start of the referenced object, if needed by the language runtime (otherwise
3155null).</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3156
3157<h5>Semantics:</h5>
3158
3159<p>The '<tt>llvm.gcread</tt>' intrinsic has the same semantics as a load
3160instruction, but may be replaced with substantially more complex code by the
3161garbage collector runtime, as needed.</p>
3162
3163</div>
3164
3165
3166<!-- _______________________________________________________________________ -->
3167<div class="doc_subsubsection">
3168 <a name="i_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a>
3169</div>
3170
3171<div class="doc_text">
3172
3173<h5>Syntax:</h5>
3174
3175<pre>
Chris Lattnerf9228072006-03-14 20:02:51 +0000[diff] [blame]3176 declare void %llvm.gcwrite(sbyte* %P1, sbyte* %Obj, sbyte** %P2)
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3177</pre>
3178
3179<h5>Overview:</h5>
3180
3181<p>The '<tt>llvm.gcwrite</tt>' intrinsic identifies writes of references to heap
3182locations, allowing garbage collector implementations that require write
3183barriers (such as generational or reference counting collectors).</p>
3184
3185<h5>Arguments:</h5>
3186
Chris Lattnerf9228072006-03-14 20:02:51 +0000[diff] [blame]3187<p>The first argument is the reference to store, the second is the start of the
3188object to store it to, and the third is the address of the field of Obj to
3189store to. If the runtime does not require a pointer to the object, Obj may be
3190null.</p>
Chris Lattner757528b0b2004-05-23 21:06:01 +0000[diff] [blame]3191
3192<h5>Semantics:</h5>
3193
3194<p>The '<tt>llvm.gcwrite</tt>' intrinsic has the same semantics as a store
3195instruction, but may be replaced with substantially more complex code by the
3196garbage collector runtime, as needed.</p>
3197
3198</div>
3199
3200
3201
3202<!-- ======================================================================= -->
3203<div class="doc_subsection">
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3204 <a name="int_codegen">Code Generator Intrinsics</a>
3205</div>
3206
3207<div class="doc_text">
3208<p>
3209These intrinsics are provided by LLVM to expose special features that may only
3210be implemented with code generator support.
3211</p>
3212
3213</div>
3214
3215<!-- _______________________________________________________________________ -->
3216<div class="doc_subsubsection">
3217 <a name="i_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a>
3218</div>
3219
3220<div class="doc_text">
3221
3222<h5>Syntax:</h5>
3223<pre>
Chris Lattnerb3d430e2006-01-13 01:20:27 +0000[diff] [blame]3224 declare sbyte *%llvm.returnaddress(uint &lt;level&gt;)
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3225</pre>
3226
3227<h5>Overview:</h5>
3228
3229<p>
Chris Lattnerc1fb4262006-10-15 20:05:59 +0000[diff] [blame]3230The '<tt>llvm.returnaddress</tt>' intrinsic attempts to compute a
3231target-specific value indicating the return address of the current function
3232or one of its callers.
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3233</p>
3234
3235<h5>Arguments:</h5>
3236
3237<p>
3238The argument to this intrinsic indicates which function to return the address
3239for. Zero indicates the calling function, one indicates its caller, etc. The
3240argument is <b>required</b> to be a constant integer value.
3241</p>
3242
3243<h5>Semantics:</h5>
3244
3245<p>
3246The '<tt>llvm.returnaddress</tt>' intrinsic either returns a pointer indicating
3247the return address of the specified call frame, or zero if it cannot be
3248identified. The value returned by this intrinsic is likely to be incorrect or 0
3249for arguments other than zero, so it should only be used for debugging purposes.
3250</p>
3251
3252<p>
3253Note that calling this intrinsic does not prevent function inlining or other
Chris Lattner2e6eb5f2005-03-07 20:30:51 +0000[diff] [blame]3254aggressive transformations, so the value returned may not be that of the obvious
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3255source-language caller.
3256</p>
3257</div>
3258
3259
3260<!-- _______________________________________________________________________ -->
3261<div class="doc_subsubsection">
3262 <a name="i_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a>
3263</div>
3264
3265<div class="doc_text">
3266
3267<h5>Syntax:</h5>
3268<pre>
Chris Lattnerb3d430e2006-01-13 01:20:27 +0000[diff] [blame]3269 declare sbyte *%llvm.frameaddress(uint &lt;level&gt;)
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3270</pre>
3271
3272<h5>Overview:</h5>
3273
3274<p>
Chris Lattnerc1fb4262006-10-15 20:05:59 +0000[diff] [blame]3275The '<tt>llvm.frameaddress</tt>' intrinsic attempts to return the
3276target-specific frame pointer value for the specified stack frame.
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3277</p>
3278
3279<h5>Arguments:</h5>
3280
3281<p>
3282The argument to this intrinsic indicates which function to return the frame
3283pointer for. Zero indicates the calling function, one indicates its caller,
3284etc. The argument is <b>required</b> to be a constant integer value.
3285</p>
3286
3287<h5>Semantics:</h5>
3288
3289<p>
3290The '<tt>llvm.frameaddress</tt>' intrinsic either returns a pointer indicating
3291the frame address of the specified call frame, or zero if it cannot be
3292identified. The value returned by this intrinsic is likely to be incorrect or 0
3293for arguments other than zero, so it should only be used for debugging purposes.
3294</p>
3295
3296<p>
3297Note that calling this intrinsic does not prevent function inlining or other
Chris Lattner2e6eb5f2005-03-07 20:30:51 +0000[diff] [blame]3298aggressive transformations, so the value returned may not be that of the obvious
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3299source-language caller.
3300</p>
3301</div>
3302
Chris Lattnerc8a2c222005-02-28 19:24:19 +0000[diff] [blame]3303<!-- _______________________________________________________________________ -->
3304<div class="doc_subsubsection">
Chris Lattner2f0f0012006-01-13 02:03:13 +0000[diff] [blame]3305 <a name="i_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a>
3306</div>
3307
3308<div class="doc_text">
3309
3310<h5>Syntax:</h5>
3311<pre>
3312 declare sbyte *%llvm.stacksave()
3313</pre>
3314
3315<h5>Overview:</h5>
3316
3317<p>
3318The '<tt>llvm.stacksave</tt>' intrinsic is used to remember the current state of
3319the function stack, for use with <a href="#i_stackrestore">
3320<tt>llvm.stackrestore</tt></a>. This is useful for implementing language
3321features like scoped automatic variable sized arrays in C99.
3322</p>
3323
3324<h5>Semantics:</h5>
3325
3326<p>
3327This intrinsic returns a opaque pointer value that can be passed to <a
3328href="#i_stackrestore"><tt>llvm.stackrestore</tt></a>. When an
3329<tt>llvm.stackrestore</tt> intrinsic is executed with a value saved from
3330<tt>llvm.stacksave</tt>, it effectively restores the state of the stack to the
3331state it was in when the <tt>llvm.stacksave</tt> intrinsic executed. In
3332practice, this pops any <a href="#i_alloca">alloca</a> blocks from the stack
3333that were allocated after the <tt>llvm.stacksave</tt> was executed.
3334</p>
3335
3336</div>
3337
3338<!-- _______________________________________________________________________ -->
3339<div class="doc_subsubsection">
3340 <a name="i_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a>
3341</div>
3342
3343<div class="doc_text">
3344
3345<h5>Syntax:</h5>
3346<pre>
3347 declare void %llvm.stackrestore(sbyte* %ptr)
3348</pre>
3349
3350<h5>Overview:</h5>
3351
3352<p>
3353The '<tt>llvm.stackrestore</tt>' intrinsic is used to restore the state of
3354the function stack to the state it was in when the corresponding <a
3355href="#llvm.stacksave"><tt>llvm.stacksave</tt></a> intrinsic executed. This is
3356useful for implementing language features like scoped automatic variable sized
3357arrays in C99.
3358</p>
3359
3360<h5>Semantics:</h5>
3361
3362<p>
3363See the description for <a href="#i_stacksave"><tt>llvm.stacksave</tt></a>.
3364</p>
3365
3366</div>
3367
3368
3369<!-- _______________________________________________________________________ -->
3370<div class="doc_subsubsection">
Chris Lattnerc8a2c222005-02-28 19:24:19 +0000[diff] [blame]3371 <a name="i_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a>
3372</div>
3373
3374<div class="doc_text">
3375
3376<h5>Syntax:</h5>
3377<pre>
Reid Spencer7821d062005-04-26 20:50:44 +0000[diff] [blame]3378 declare void %llvm.prefetch(sbyte * &lt;address&gt;,
3379 uint &lt;rw&gt;, uint &lt;locality&gt;)
Chris Lattnerc8a2c222005-02-28 19:24:19 +0000[diff] [blame]3380</pre>
3381
3382<h5>Overview:</h5>
3383
3384
3385<p>
3386The '<tt>llvm.prefetch</tt>' intrinsic is a hint to the code generator to insert
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]3387a prefetch instruction if supported; otherwise, it is a noop. Prefetches have
3388no
3389effect on the behavior of the program but can change its performance
Chris Lattnerff851072005-02-28 19:47:14 +0000[diff] [blame]3390characteristics.
Chris Lattnerc8a2c222005-02-28 19:24:19 +0000[diff] [blame]3391</p>
3392
3393<h5>Arguments:</h5>
3394
3395<p>
3396<tt>address</tt> is the address to be prefetched, <tt>rw</tt> is the specifier
3397determining if the fetch should be for a read (0) or write (1), and
3398<tt>locality</tt> is a temporal locality specifier ranging from (0) - no
Chris Lattnerd3e641c2005-03-07 20:31:38 +0000[diff] [blame]3399locality, to (3) - extremely local keep in cache. The <tt>rw</tt> and
Chris Lattnerc8a2c222005-02-28 19:24:19 +0000[diff] [blame]3400<tt>locality</tt> arguments must be constant integers.
3401</p>
3402
3403<h5>Semantics:</h5>
3404
3405<p>
3406This intrinsic does not modify the behavior of the program. In particular,
3407prefetches cannot trap and do not produce a value. On targets that support this
3408intrinsic, the prefetch can provide hints to the processor cache for better
3409performance.
3410</p>
3411
3412</div>
3413
Andrew Lenharthb4427912005-03-28 20:05:49 +0000[diff] [blame]3414<!-- _______________________________________________________________________ -->
3415<div class="doc_subsubsection">
3416 <a name="i_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a>
3417</div>
3418
3419<div class="doc_text">
3420
3421<h5>Syntax:</h5>
3422<pre>
Reid Spencer7821d062005-04-26 20:50:44 +0000[diff] [blame]3423 declare void %llvm.pcmarker( uint &lt;id&gt; )
Andrew Lenharthb4427912005-03-28 20:05:49 +0000[diff] [blame]3424</pre>
3425
3426<h5>Overview:</h5>
3427
3428
3429<p>
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]3430The '<tt>llvm.pcmarker</tt>' intrinsic is a method to export a Program Counter
3431(PC) in a region of
Andrew Lenharthb4427912005-03-28 20:05:49 +0000[diff] [blame]3432code to simulators and other tools. The method is target specific, but it is
3433expected that the marker will use exported symbols to transmit the PC of the marker.
Jeff Cohendc6bfea2005-11-11 02:15:27 +0000[diff] [blame]3434The marker makes no guarantees that it will remain with any specific instruction
Chris Lattnere64d41d2005-11-15 06:07:55 +0000[diff] [blame]3435after optimizations. It is possible that the presence of a marker will inhibit
Chris Lattnerb40261e2006-03-24 07:16:10 +0000[diff] [blame]3436optimizations. The intended use is to be inserted after optimizations to allow
John Criswell88190562005-05-16 16:17:45 +0000[diff] [blame]3437correlations of simulation runs.
Andrew Lenharthb4427912005-03-28 20:05:49 +0000[diff] [blame]3438</p>
3439
3440<h5>Arguments:</h5>
3441
3442<p>
3443<tt>id</tt> is a numerical id identifying the marker.
3444</p>
3445
3446<h5>Semantics:</h5>
3447
3448<p>
3449This intrinsic does not modify the behavior of the program. Backends that do not
3450support this intrinisic may ignore it.
3451</p>
3452
3453</div>
3454
Andrew Lenharth01aa5632005-11-11 16:47:30 +0000[diff] [blame]3455<!-- _______________________________________________________________________ -->
3456<div class="doc_subsubsection">
3457 <a name="i_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a>
3458</div>
3459
3460<div class="doc_text">
3461
3462<h5>Syntax:</h5>
3463<pre>
3464 declare ulong %llvm.readcyclecounter( )
3465</pre>
3466
3467<h5>Overview:</h5>
3468
3469
3470<p>
3471The '<tt>llvm.readcyclecounter</tt>' intrinsic provides access to the cycle
3472counter register (or similar low latency, high accuracy clocks) on those targets
3473that support it. On X86, it should map to RDTSC. On Alpha, it should map to RPCC.
3474As the backing counters overflow quickly (on the order of 9 seconds on alpha), this
3475should only be used for small timings.
3476</p>
3477
3478<h5>Semantics:</h5>
3479
3480<p>
3481When directly supported, reading the cycle counter should not modify any memory.
3482Implementations are allowed to either return a application specific value or a
3483system wide value. On backends without support, this is lowered to a constant 0.
3484</p>
3485
3486</div>
3487
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3488<!-- ======================================================================= -->
3489<div class="doc_subsection">
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3490 <a name="int_libc">Standard C Library Intrinsics</a>
3491</div>
3492
3493<div class="doc_text">
3494<p>
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3495LLVM provides intrinsics for a few important standard C library functions.
3496These intrinsics allow source-language front-ends to pass information about the
3497alignment of the pointer arguments to the code generator, providing opportunity
3498for more efficient code generation.
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3499</p>
3500
3501</div>
3502
3503<!-- _______________________________________________________________________ -->
3504<div class="doc_subsubsection">
3505 <a name="i_memcpy">'<tt>llvm.memcpy</tt>' Intrinsic</a>
3506</div>
3507
3508<div class="doc_text">
3509
3510<h5>Syntax:</h5>
3511<pre>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]3512 declare void %llvm.memcpy.i32(sbyte* &lt;dest&gt;, sbyte* &lt;src&gt;,
3513 uint &lt;len&gt;, uint &lt;align&gt;)
3514 declare void %llvm.memcpy.i64(sbyte* &lt;dest&gt;, sbyte* &lt;src&gt;,
3515 ulong &lt;len&gt;, uint &lt;align&gt;)
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3516</pre>
3517
3518<h5>Overview:</h5>
3519
3520<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]3521The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3522location to the destination location.
3523</p>
3524
3525<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]3526Note that, unlike the standard libc function, the <tt>llvm.memcpy.*</tt>
3527intrinsics do not return a value, and takes an extra alignment argument.
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3528</p>
3529
3530<h5>Arguments:</h5>
3531
3532<p>
3533The first argument is a pointer to the destination, the second is a pointer to
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]3534the source. The third argument is an integer argument
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3535specifying the number of bytes to copy, and the fourth argument is the alignment
3536of the source and destination locations.
3537</p>
3538
Chris Lattner4c67c482004-02-12 21:18:15 +0000[diff] [blame]3539<p>
3540If 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]3541the caller guarantees that both the source and destination pointers are aligned
3542to that boundary.
Chris Lattner4c67c482004-02-12 21:18:15 +0000[diff] [blame]3543</p>
3544
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3545<h5>Semantics:</h5>
3546
3547<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]3548The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
Chris Lattnerfee11462004-02-12 17:01:32 +0000[diff] [blame]3549location to the destination location, which are not allowed to overlap. It
3550copies "len" bytes of memory over. If the argument is known to be aligned to
3551some boundary, this can be specified as the fourth argument, otherwise it should
3552be set to 0 or 1.
3553</p>
3554</div>
3555
3556
Chris Lattnerf30152e2004-02-12 18:10:10 +0000[diff] [blame]3557<!-- _______________________________________________________________________ -->
3558<div class="doc_subsubsection">
3559 <a name="i_memmove">'<tt>llvm.memmove</tt>' Intrinsic</a>
3560</div>
3561
3562<div class="doc_text">
3563
3564<h5>Syntax:</h5>
3565<pre>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]3566 declare void %llvm.memmove.i32(sbyte* &lt;dest&gt;, sbyte* &lt;src&gt;,
3567 uint &lt;len&gt;, uint &lt;align&gt;)
3568 declare void %llvm.memmove.i64(sbyte* &lt;dest&gt;, sbyte* &lt;src&gt;,
3569 ulong &lt;len&gt;, uint &lt;align&gt;)
Chris Lattnerf30152e2004-02-12 18:10:10 +0000[diff] [blame]3570</pre>
3571
3572<h5>Overview:</h5>
3573
3574<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]3575The '<tt>llvm.memmove.*</tt>' intrinsics move a block of memory from the source
3576location to the destination location. It is similar to the
3577'<tt>llvm.memcmp</tt>' intrinsic but allows the two memory locations to overlap.
Chris Lattnerf30152e2004-02-12 18:10:10 +0000[diff] [blame]3578</p>
3579
3580<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]3581Note that, unlike the standard libc function, the <tt>llvm.memmove.*</tt>
3582intrinsics do not return a value, and takes an extra alignment argument.
Chris Lattnerf30152e2004-02-12 18:10:10 +0000[diff] [blame]3583</p>
3584
3585<h5>Arguments:</h5>
3586
3587<p>
3588The first argument is a pointer to the destination, the second is a pointer to
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]3589the source. The third argument is an integer argument
Chris Lattnerf30152e2004-02-12 18:10:10 +0000[diff] [blame]3590specifying the number of bytes to copy, and the fourth argument is the alignment
3591of the source and destination locations.
3592</p>
3593
Chris Lattner4c67c482004-02-12 21:18:15 +0000[diff] [blame]3594<p>
3595If 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]3596the caller guarantees that the source and destination pointers are aligned to
3597that boundary.
Chris Lattner4c67c482004-02-12 21:18:15 +0000[diff] [blame]3598</p>
3599
Chris Lattnerf30152e2004-02-12 18:10:10 +0000[diff] [blame]3600<h5>Semantics:</h5>
3601
3602<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]3603The '<tt>llvm.memmove.*</tt>' intrinsics copy a block of memory from the source
Chris Lattnerf30152e2004-02-12 18:10:10 +0000[diff] [blame]3604location to the destination location, which may overlap. It
3605copies "len" bytes of memory over. If the argument is known to be aligned to
3606some boundary, this can be specified as the fourth argument, otherwise it should
3607be set to 0 or 1.
3608</p>
3609</div>
3610
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3611
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3612<!-- _______________________________________________________________________ -->
3613<div class="doc_subsubsection">
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]3614 <a name="i_memset">'<tt>llvm.memset.*</tt>' Intrinsics</a>
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3615</div>
3616
3617<div class="doc_text">
3618
3619<h5>Syntax:</h5>
3620<pre>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]3621 declare void %llvm.memset.i32(sbyte* &lt;dest&gt;, ubyte &lt;val&gt;,
3622 uint &lt;len&gt;, uint &lt;align&gt;)
3623 declare void %llvm.memset.i64(sbyte* &lt;dest&gt;, ubyte &lt;val&gt;,
3624 ulong &lt;len&gt;, uint &lt;align&gt;)
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3625</pre>
3626
3627<h5>Overview:</h5>
3628
3629<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]3630The '<tt>llvm.memset.*</tt>' intrinsics fill a block of memory with a particular
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3631byte value.
3632</p>
3633
3634<p>
3635Note that, unlike the standard libc function, the <tt>llvm.memset</tt> intrinsic
3636does not return a value, and takes an extra alignment argument.
3637</p>
3638
3639<h5>Arguments:</h5>
3640
3641<p>
3642The first argument is a pointer to the destination to fill, the second is the
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]3643byte value to fill it with, the third argument is an integer
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3644argument specifying the number of bytes to fill, and the fourth argument is the
3645known alignment of destination location.
3646</p>
3647
3648<p>
3649If 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]3650the caller guarantees that the destination pointer is aligned to that boundary.
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3651</p>
3652
3653<h5>Semantics:</h5>
3654
3655<p>
Chris Lattner0c8b2592006-03-03 00:07:20 +0000[diff] [blame]3656The '<tt>llvm.memset.*</tt>' intrinsics fill "len" bytes of memory starting at
3657the
Chris Lattner3649c3a2004-02-14 04:08:35 +0000[diff] [blame]3658destination location. If the argument is known to be aligned to some boundary,
3659this can be specified as the fourth argument, otherwise it should be set to 0 or
36601.
3661</p>
3662</div>
3663
3664
Chris Lattner3b4f4372004-06-11 02:28:03 +0000[diff] [blame]3665<!-- _______________________________________________________________________ -->
3666<div class="doc_subsubsection">
Reid Spencerb4f9a6f2006-01-16 21:12:35 +0000[diff] [blame]3667 <a name="i_isunordered">'<tt>llvm.isunordered.*</tt>' Intrinsic</a>
Alkis Evlogimenos0fa39232004-06-13 01:16:15 +0000[diff] [blame]3668</div>
3669
3670<div class="doc_text">
3671
3672<h5>Syntax:</h5>
3673<pre>
Reid Spencerb4f9a6f2006-01-16 21:12:35 +0000[diff] [blame]3674 declare bool %llvm.isunordered.f32(float Val1, float Val2)
3675 declare bool %llvm.isunordered.f64(double Val1, double Val2)
Alkis Evlogimenos0fa39232004-06-13 01:16:15 +0000[diff] [blame]3676</pre>
3677
3678<h5>Overview:</h5>
3679
3680<p>
Reid Spencerb4f9a6f2006-01-16 21:12:35 +0000[diff] [blame]3681The '<tt>llvm.isunordered</tt>' intrinsics return true if either or both of the
Alkis Evlogimenos0fa39232004-06-13 01:16:15 +0000[diff] [blame]3682specified floating point values is a NAN.
3683</p>
3684
3685<h5>Arguments:</h5>
3686
3687<p>
3688The arguments are floating point numbers of the same type.
3689</p>
3690
3691<h5>Semantics:</h5>
3692
3693<p>
3694If either or both of the arguments is a SNAN or QNAN, it returns true, otherwise
3695false.
3696</p>
3697</div>
3698
3699
Chris Lattner8a8f2e52005-07-21 01:29:16 +0000[diff] [blame]3700<!-- _______________________________________________________________________ -->
3701<div class="doc_subsubsection">
Chris Lattner069b5bd2006-01-16 22:38:59 +0000[diff] [blame]3702 <a name="i_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a>
Chris Lattner8a8f2e52005-07-21 01:29:16 +0000[diff] [blame]3703</div>
3704
3705<div class="doc_text">
3706
3707<h5>Syntax:</h5>
3708<pre>
Chris Lattner33b73f92006-09-08 06:34:02 +0000[diff] [blame]3709 declare float %llvm.sqrt.f32(float %Val)
3710 declare double %llvm.sqrt.f64(double %Val)
Chris Lattner8a8f2e52005-07-21 01:29:16 +0000[diff] [blame]3711</pre>
3712
3713<h5>Overview:</h5>
3714
3715<p>
Reid Spencerb4f9a6f2006-01-16 21:12:35 +0000[diff] [blame]3716The '<tt>llvm.sqrt</tt>' intrinsics return the sqrt of the specified operand,
Chris Lattner8a8f2e52005-07-21 01:29:16 +0000[diff] [blame]3717returning the same value as the libm '<tt>sqrt</tt>' function would. Unlike
3718<tt>sqrt</tt> in libm, however, <tt>llvm.sqrt</tt> has undefined behavior for
3719negative numbers (which allows for better optimization).
3720</p>
3721
3722<h5>Arguments:</h5>
3723
3724<p>
3725The argument and return value are floating point numbers of the same type.
3726</p>
3727
3728<h5>Semantics:</h5>
3729
3730<p>
3731This function returns the sqrt of the specified operand if it is a positive
3732floating point number.
3733</p>
3734</div>
3735
Chris Lattner33b73f92006-09-08 06:34:02 +0000[diff] [blame]3736<!-- _______________________________________________________________________ -->
3737<div class="doc_subsubsection">
3738 <a name="i_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a>
3739</div>
3740
3741<div class="doc_text">
3742
3743<h5>Syntax:</h5>
3744<pre>
3745 declare float %llvm.powi.f32(float %Val, int %power)
3746 declare double %llvm.powi.f64(double %Val, int %power)
3747</pre>
3748
3749<h5>Overview:</h5>
3750
3751<p>
3752The '<tt>llvm.powi.*</tt>' intrinsics return the first operand raised to the
3753specified (positive or negative) power. The order of evaluation of
3754multiplications is not defined.
3755</p>
3756
3757<h5>Arguments:</h5>
3758
3759<p>
3760The second argument is an integer power, and the first is a value to raise to
3761that power.
3762</p>
3763
3764<h5>Semantics:</h5>
3765
3766<p>
3767This function returns the first value raised to the second power with an
3768unspecified sequence of rounding operations.</p>
3769</div>
3770
3771
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]3772<!-- ======================================================================= -->
3773<div class="doc_subsection">
Nate Begeman0f223bb2006-01-13 23:26:38 +0000[diff] [blame]3774 <a name="int_manip">Bit Manipulation Intrinsics</a>
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]3775</div>
3776
3777<div class="doc_text">
3778<p>
Nate Begeman0f223bb2006-01-13 23:26:38 +0000[diff] [blame]3779LLVM provides intrinsics for a few important bit manipulation operations.
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]3780These allow efficient code generation for some algorithms.
3781</p>
3782
3783</div>
3784
3785<!-- _______________________________________________________________________ -->
3786<div class="doc_subsubsection">
Nate Begeman0f223bb2006-01-13 23:26:38 +0000[diff] [blame]3787 <a name="i_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a>
3788</div>
3789
3790<div class="doc_text">
3791
3792<h5>Syntax:</h5>
3793<pre>
Chris Lattner069b5bd2006-01-16 22:38:59 +0000[diff] [blame]3794 declare ushort %llvm.bswap.i16(ushort &lt;id&gt;)
3795 declare uint %llvm.bswap.i32(uint &lt;id&gt;)
3796 declare ulong %llvm.bswap.i64(ulong &lt;id&gt;)
Nate Begeman0f223bb2006-01-13 23:26:38 +0000[diff] [blame]3797</pre>
3798
3799<h5>Overview:</h5>
3800
3801<p>
3802The '<tt>llvm.bwsap</tt>' family of intrinsics is used to byteswap a 16, 32 or
380364 bit quantity. These are useful for performing operations on data that is not
3804in the target's native byte order.
3805</p>
3806
3807<h5>Semantics:</h5>
3808
3809<p>
Chris Lattner069b5bd2006-01-16 22:38:59 +0000[diff] [blame]3810The <tt>llvm.bswap.16</tt> intrinsic returns a ushort value that has the high and low
3811byte of the input ushort swapped. Similarly, the <tt>llvm.bswap.i32</tt> intrinsic
Nate Begeman0f223bb2006-01-13 23:26:38 +0000[diff] [blame]3812returns a uint value that has the four bytes of the input uint swapped, so that
3813if the input bytes are numbered 0, 1, 2, 3 then the returned uint will have its
Chris Lattner069b5bd2006-01-16 22:38:59 +0000[diff] [blame]3814bytes in 3, 2, 1, 0 order. The <tt>llvm.bswap.i64</tt> intrinsic extends this concept
Nate Begeman0f223bb2006-01-13 23:26:38 +0000[diff] [blame]3815to 64 bits.
3816</p>
3817
3818</div>
3819
3820<!-- _______________________________________________________________________ -->
3821<div class="doc_subsubsection">
Reid Spencerb4f9a6f2006-01-16 21:12:35 +0000[diff] [blame]3822 <a name="int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic</a>
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]3823</div>
3824
3825<div class="doc_text">
3826
3827<h5>Syntax:</h5>
3828<pre>
Chris Lattner069b5bd2006-01-16 22:38:59 +0000[diff] [blame]3829 declare ubyte %llvm.ctpop.i8 (ubyte &lt;src&gt;)
3830 declare ushort %llvm.ctpop.i16(ushort &lt;src&gt;)
3831 declare uint %llvm.ctpop.i32(uint &lt;src&gt;)
3832 declare ulong %llvm.ctpop.i64(ulong &lt;src&gt;)
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]3833</pre>
3834
3835<h5>Overview:</h5>
3836
3837<p>
Chris Lattner069b5bd2006-01-16 22:38:59 +0000[diff] [blame]3838The '<tt>llvm.ctpop</tt>' family of intrinsics counts the number of bits set in a
3839value.
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]3840</p>
3841
3842<h5>Arguments:</h5>
3843
3844<p>
Chris Lattner573f64e2005-05-07 01:46:40 +0000[diff] [blame]3845The only argument is the value to be counted. The argument may be of any
Chris Lattner069b5bd2006-01-16 22:38:59 +0000[diff] [blame]3846unsigned integer type. The return type must match the argument type.
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]3847</p>
3848
3849<h5>Semantics:</h5>
3850
3851<p>
3852The '<tt>llvm.ctpop</tt>' intrinsic counts the 1's in a variable.
3853</p>
3854</div>
3855
3856<!-- _______________________________________________________________________ -->
3857<div class="doc_subsubsection">
Chris Lattnerb748c672006-01-16 22:34:14 +0000[diff] [blame]3858 <a name="int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic</a>
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]3859</div>
3860
3861<div class="doc_text">
3862
3863<h5>Syntax:</h5>
3864<pre>
Chris Lattner069b5bd2006-01-16 22:38:59 +0000[diff] [blame]3865 declare ubyte %llvm.ctlz.i8 (ubyte &lt;src&gt;)
3866 declare ushort %llvm.ctlz.i16(ushort &lt;src&gt;)
3867 declare uint %llvm.ctlz.i32(uint &lt;src&gt;)
3868 declare ulong %llvm.ctlz.i64(ulong &lt;src&gt;)
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]3869</pre>
3870
3871<h5>Overview:</h5>
3872
3873<p>
Reid Spencerb4f9a6f2006-01-16 21:12:35 +0000[diff] [blame]3874The '<tt>llvm.ctlz</tt>' family of intrinsic functions counts the number of
3875leading zeros in a variable.
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]3876</p>
3877
3878<h5>Arguments:</h5>
3879
3880<p>
Chris Lattner573f64e2005-05-07 01:46:40 +0000[diff] [blame]3881The only argument is the value to be counted. The argument may be of any
Chris Lattner069b5bd2006-01-16 22:38:59 +0000[diff] [blame]3882unsigned integer type. The return type must match the argument type.
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]3883</p>
3884
3885<h5>Semantics:</h5>
3886
3887<p>
Chris Lattnerefa20fa2005-05-15 19:39:26 +0000[diff] [blame]3888The '<tt>llvm.ctlz</tt>' intrinsic counts the leading (most significant) zeros
3889in a variable. If the src == 0 then the result is the size in bits of the type
Chris Lattner905bd172006-04-21 21:37:40 +0000[diff] [blame]3890of src. For example, <tt>llvm.ctlz(int 2) = 30</tt>.
Andrew Lenharth1d463522005-05-03 18:01:48 +0000[diff] [blame]3891</p>
3892</div>
Chris Lattner3b4f4372004-06-11 02:28:03 +0000[diff] [blame]3893
3894
Chris Lattnerefa20fa2005-05-15 19:39:26 +0000[diff] [blame]3895
3896<!-- _______________________________________________________________________ -->
3897<div class="doc_subsubsection">
Chris Lattnerb748c672006-01-16 22:34:14 +0000[diff] [blame]3898 <a name="int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic</a>
Chris Lattnerefa20fa2005-05-15 19:39:26 +0000[diff] [blame]3899</div>
3900
3901<div class="doc_text">
3902
3903<h5>Syntax:</h5>
3904<pre>
Chris Lattner069b5bd2006-01-16 22:38:59 +0000[diff] [blame]3905 declare ubyte %llvm.cttz.i8 (ubyte &lt;src&gt;)
3906 declare ushort %llvm.cttz.i16(ushort &lt;src&gt;)
3907 declare uint %llvm.cttz.i32(uint &lt;src&gt;)
3908 declare ulong %llvm.cttz.i64(ulong &lt;src&gt;)
Chris Lattnerefa20fa2005-05-15 19:39:26 +0000[diff] [blame]3909</pre>
3910
3911<h5>Overview:</h5>
3912
3913<p>
Reid Spencerb4f9a6f2006-01-16 21:12:35 +0000[diff] [blame]3914The '<tt>llvm.cttz</tt>' family of intrinsic functions counts the number of
3915trailing zeros.
Chris Lattnerefa20fa2005-05-15 19:39:26 +0000[diff] [blame]3916</p>
3917
3918<h5>Arguments:</h5>
3919
3920<p>
3921The only argument is the value to be counted. The argument may be of any
Chris Lattner069b5bd2006-01-16 22:38:59 +0000[diff] [blame]3922unsigned integer type. The return type must match the argument type.
Chris Lattnerefa20fa2005-05-15 19:39:26 +0000[diff] [blame]3923</p>
3924
3925<h5>Semantics:</h5>
3926
3927<p>
3928The '<tt>llvm.cttz</tt>' intrinsic counts the trailing (least significant) zeros
3929in a variable. If the src == 0 then the result is the size in bits of the type
3930of src. For example, <tt>llvm.cttz(2) = 1</tt>.
3931</p>
3932</div>
3933
Chris Lattner941515c2004-01-06 05:31:32 +0000[diff] [blame]3934<!-- ======================================================================= -->
3935<div class="doc_subsection">
3936 <a name="int_debugger">Debugger Intrinsics</a>
3937</div>
3938
3939<div class="doc_text">
3940<p>
3941The LLVM debugger intrinsics (which all start with <tt>llvm.dbg.</tt> prefix),
3942are described in the <a
3943href="SourceLevelDebugging.html#format_common_intrinsics">LLVM Source Level
3944Debugging</a> document.
3945</p>
3946</div>
3947
3948
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]3949<!-- *********************************************************************** -->
Chris Lattner2f7c9632001-06-06 20:29:01 +0000[diff] [blame]3950<hr>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]3951<address>
3952 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
3953 src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
3954 <a href="http://validator.w3.org/check/referer"><img
3955 src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!" /></a>
3956
3957 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
Reid Spencerca058542006-03-14 05:39:39 +0000[diff] [blame]3958 <a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br>
Misha Brukmanc501f552004-03-01 17:47:27 +0000[diff] [blame]3959 Last modified: $Date$
3960</address>
Misha Brukman76307852003-11-08 01:05:38 +0000[diff] [blame]3961</body>
3962</html>